2022-12-01 04:23:56 <cache_return> is there a way to roll back updates?
2022-12-01 05:37:04 <Apachez> sure, restore from your backup ;)
2022-12-01 05:47:36 <jhjacobs[m]> 🙈
2022-12-01 06:08:44 <ForeverNoob[m]> @cache_return: You might want to look at Nix or Guix package managers for that.
2022-12-01 06:15:23 <ForeverNoob[m]> @jhjacobs: I was oogling this board recently: https://news.ycombinator.com/item?id=33109979
2022-12-01 07:33:04 <jhjacobs[m]> ForeverNoob: thats inderd the article that got me interested :-)
2022-12-01 11:03:11 <wyk72> Hello
2022-12-01 11:03:33 <wyk72> I need help with IPv6 configuration with Alpine 3.16
2022-12-01 11:03:56 <wyk72> I have a VPS with Alpine 3.16
2022-12-01 11:04:12 <wyk72> dhcp gets correct values from ipV4, but fails with ipv6
2022-12-01 11:04:34 <wyk72> I installed dhcpd
2022-12-01 11:04:48 <wyk72> but I get no IPv6
2022-12-01 11:05:43 <wyk72> tried everything, even static, but is very confusing
2022-12-01 11:06:59 <wyk72> I get "no ADDRESS FOR THIS INTERFACE" from dhcpd
2022-12-01 11:07:15 <wyk72> from dhcpcd
2022-12-01 11:11:15 <wyk72> I always get "Dec  1 11:19:01 vps-ionos3 daemon.err dhcpcd[3198]: eth0: DHCPv6 REPLY: No addresses available for this interface."
2022-12-01 11:16:57 <lopid> is the dhcp server configured to assign one?
2022-12-01 11:19:52 <wyk72> ipv4 is opk
2022-12-01 11:20:02 <wyk72> the ISP tells me that DHCP6 is enabled
2022-12-01 11:23:22 <wyk72> tried static
2022-12-01 11:23:26 <wyk72> does not work
2022-12-01 11:23:31 <wyk72> I am puzzled
2022-12-01 11:26:14 <lopid> if you have static ip's, you don't need dhcp, right?
2022-12-01 11:30:13 <wyk72> the VPS is a VMWARE server: it assigns the virtual machine a static IPs via  DHCP usually
2022-12-01 11:30:37 <wyk72> works with IPv4, the ISP tells me that should work with IPv6 but it doesn't
2022-12-01 11:31:06 <wyk72> on the manual of Alpine I read: "IPv6 DHCP Configuration  Alpine's use of ifupdown-ng supports three DHCP clients: udhcpc, dhclient, and dhcpcd. Of these, only dhcpcd can interact with both DHCP and DHCPv6 from the same process, which ifupdown-ng requires. Thus the IPv4 DHCP configuration given above will also result in the use of DHCPv6, but only if you install the dhcpcd package. (The ifupdown-ng scripts prioritize dhclient over u
2022-12-01 11:31:23 <wyk72> so I installed dhcpcd
2022-12-01 11:31:35 <wyk72> I get v4 ip but not v6 ip
2022-12-01 11:31:41 <Shiz> why do you want dhcp on ipv6?
2022-12-01 11:31:47 <Shiz> it's usually not the kinda setup you want
2022-12-01 11:32:00 <lopid> vmware may assign its vm ip's using dhcp, but they should be static to the os inside
2022-12-01 11:32:59 <wyk72> my ISP tell me that is providing the IPv6 via DHCP - tried STATIC and it does not work either
2022-12-01 11:33:19 <wyk72> they sent me the UBUNTU config
2022-12-01 11:33:27 <lopid> ok
2022-12-01 11:34:06 <wyk72> auto lo eth0 iface lo inet loopback  allow-hotplug eth0 iface eth0 inet static  address INDIRIZZO-IPv4-PRINCIPALE/32  iface eth0 inet6 manual       pre-up sleep 5       up dhclient -6 -nw -v eth0       post-up sleep 5; /sbin/ip -family inet6 route add default via fe80::1 dev eth0       pre-down /sbin/ip -family inet6 route del default via fe80::1 dev eth0       down dhclient -6 -r -v eth0
2022-12-01 11:34:38 <ikke> wyk72: then they are probably giving you a single /128 and not routing a /64 to you like it was mean to
2022-12-01 11:34:54 <wyk72> this is the UBUNTU config 
2022-12-01 11:35:30 <wyk72> they give me      2001:8d8:1800:8213::/64 as subnet
2022-12-01 11:36:13 <wyk72> and 2001:8d8:1800:8213::1 as static IP
2022-12-01 11:36:31 <ikke> But it does not work if you give yourself a static IP in that range?
2022-12-01 11:36:34 <wyk72> next-hop should be fe80::1
2022-12-01 11:36:38 <wyk72> no
2022-12-01 11:37:03 <ikke> lesigh
2022-12-01 11:48:39 <wyk72> is it normal that I get "inet6 addr: fe80::250:56ff:fe0d:c027/64 Scope:Link" as ipv6 ip ?
2022-12-01 11:48:53 <wyk72> makes no sense to me
2022-12-01 11:50:14 <psykose> it's a normal link-local so yes
2022-12-01 11:50:26 <ikke> The kernel automatically assigns you one
2022-12-01 11:50:43 <ikke> with ipv6, you communicate with the router over the link-local subnet
2022-12-01 11:51:06 <wyk72> oh I see
2022-12-01 11:51:39 <wyk72> it's a mistery to me: Ubuntu works Alpine no
2022-12-01 11:51:58 <wyk72> I can try with dhclient
2022-12-01 11:52:01 <ikke> What does not work?
2022-12-01 11:52:32 <wyk72> dhcpv6 works with Ubuntu 20.20 (dhclient)
2022-12-01 11:53:08 <wyk72> a line like: dhclient -6 -nw -v eth0
2022-12-01 11:53:40 <wyk72> no idea how I should I use it on alpine, package is present but I get an error
2022-12-01 11:54:10 <wyk72> Can't bind to dhcp address: Address in use Please make sure there is no other dhcp server running and that there's no entry for dhcp or bootp in /etc/inetd.conf.   Also make sure you are not running HP JetAdmin software, which includes a bootp server.
2022-12-01 11:54:34 <M32573i12[m]> How much Microsoft is behind Linux development these days? 
2022-12-01 11:55:33 <M32573i12[m]> They had that one Linux distro about ten years ago... Opensuse or something like that.
2022-12-01 11:56:16 <wyk72> WTF
2022-12-01 11:56:21 <wyk72> dhclient works
2022-12-01 11:56:32 <wyk72> I get this adress:  inet6 addr: 2001:8d8:1800:8213::1/128 Scope:Global
2022-12-01 11:56:44 <wyk72> dhcpcd no
2022-12-01 11:56:49 <wyk72> dhclient gets an ip
2022-12-01 11:56:54 <wyk72> wtf
2022-12-01 11:58:57 <wyk72> and gets resolv.conf correctly
2022-12-01 11:59:50 <Shiz> yeah, dhcpv6 is a mess
2022-12-01 12:01:25 <M32573i12[m]> About 5G... what it means to Linux? Nothing new?
2022-12-01 12:01:40 <wyk72> so I should remove dhcpcd and use dhclient...there is a way not to modify ifupdown-ng ?
2022-12-01 12:02:13 <Shiz> M32573i12[m]: nothing at all, as none of that stuff happens on the host computer
2022-12-01 12:05:59 <M32573i12[m]> Oh... my bad. I thought it was all about MS.
2022-12-01 12:06:12 <Shiz> cache_return: it depends on how you configured stuff, apk may have re-run your initramfs generation
2022-12-01 12:06:43 <Shiz> and if you made some changes there that didn't persist in that way, the initramfs was likely re-generated without
2022-12-01 12:06:51 <Shiz> (bit of a usability failure...)
2022-12-01 12:06:59 <cache_return> Shiz: it did, but why should that bork everything?
2022-12-01 12:07:08 <M32573i12[m]> Wikimafia lol.. 
2022-12-01 12:07:41 <Shiz> random snarking is not particularly desired here, keep it to yourself
2022-12-01 12:07:54 <cache_return> I seem to remember that the wifi worked "out of the box"
2022-12-01 12:07:56 <Shiz> cache_return: if the initramfs was regenerated without the mdadm and wifi modules you need, it will bork stuff
2022-12-01 12:07:57 <M32573i12[m]> Yes sir
2022-12-01 12:08:20 <cache_return> dmesg doesnt even show it trying to load the wifi
2022-12-01 12:08:30 <Shiz> i am a bit vague on the specifics, but check your features= in /etc/mkinitfs/mkinitfs.conf
2022-12-01 12:08:38 <Shiz> (been a while since I used alpine's stock initramfs)
2022-12-01 12:10:30 <Shiz> and your modules= in /etc/update-extlinux.conf
2022-12-01 12:10:50 <Shiz> (presuming you use extlinux as bootloader)
2022-12-01 12:17:43 <cache_return> I will look at this, but it seems so weird. Why would it fail to build the wifi module that was in the initial installation? ugh.
2022-12-01 12:17:56 <cache_return> makes total sense for raid, though.
2022-12-01 12:20:21 <M32573i12[m]> Do you still own huge ITX case computers? 
2022-12-01 12:49:32 <Shiz> cache_return: yeah, it seems like a weird bug for sure
2022-12-01 12:49:49 <Shiz> hopefully someone more familiar with alpine-conf mechanics can chime in :)
2022-12-01 14:00:48 <ForeverNoob[m]> Q: How do I tell crond to only run stuff in `/etc/periodic/15min/` ? I've already deleted all dirs there except that one and restarted crond.
2022-12-01 14:02:08 <PureTryOut> ForeverNoob: remove them from regular crontab
2022-12-01 14:02:19 <PureTryOut> # crontab -e
2022-12-01 14:04:47 <ForeverNoob[m]> Cool, I'll try that, thanks!
2022-12-01 14:09:08 <M32573i12[m]> Do you have old alpine distros installed? I have old ubuntu from 2006. Are those old distros any good today?
2022-12-01 14:14:40 <M32573i12[m]> <M32573i12[m]> "Do you have old alpine distros..." <- Probably not... maybe I have to try... if it works. 
2022-12-01 14:15:33 <M32573i12[m]> I feel so old because I was playing with Linux already in 2003. We talked about same stuff you guys today :)
2022-12-01 14:16:22 <minimal> M32573i12[m]: perhaps move to the alpine-offtopic channel?
2022-12-01 14:17:32 <M32573i12[m]> minimal: Oh. You are tough. Alpine only and nothing related. Ok.
2022-12-01 14:23:29 <PureTryOut> Yes, like everywhere there are offtopic channels for such things
2022-12-01 14:25:23 <ikke> M32573i12[m]: it's a courtesy for people here trying to respond to people who need help
2022-12-01 16:38:27 <scorpion2185[m]1> I am making my first package, newapkbuild packagename made a folder with the APKBUILD and a folder src
2022-12-01 16:38:38 <scorpion2185[m]1> do I put the unpacked source in src?
2022-12-01 16:40:56 <ikke> scorpion2185[m]1: no, you define the source in the APKBUILD (the url to the source archive)
2022-12-01 16:41:03 <ikke> apk will fetch and unpack it
2022-12-01 16:42:14 <ikke> If this is something just for yourself you want to build locally, then you could put it manually there
2022-12-01 16:42:46 <scorpion2185[m]1> in /var/cache/distfiles/ ?
2022-12-01 16:42:48 <ikke> abuild clean (which is automatically run after abuild -r, will remove the src dir for example
2022-12-01 16:43:27 <scorpion2185[m]1> make: *** No targets specified and no makefile found.  Stop.
2022-12-01 16:43:27 <scorpion2185[m]1> >>> ERROR: smlock: build failed
2022-12-01 16:43:57 <ikke> scorpion2185[m]1: you need to set builddir to the directory where the unpacked source is located
2022-12-01 16:44:14 <ikke> the default is "$srcdir/$pkgname-$pkgver/
2022-12-01 16:45:21 <scorpion2185[m]1> I have that set, how do I know where is it unpacked ?
2022-12-01 16:45:56 <ikke> scorpion2185[m]1: run abuild unpack, then check the src/ directory
2022-12-01 16:47:12 <scorpion2185[m]1> only $pkgname
2022-12-01 16:47:37 <ikke> then set pkgdir="$srcdir/$pkgname"
2022-12-01 16:47:43 <ikke> sorry
2022-12-01 16:47:46 <ikke> builddir=".."
2022-12-01 16:48:16 <scorpion2185[m]1> `WARNING: Ignoring ~/packages//apps: No such file or directory`
2022-12-01 16:48:16 <scorpion2185[m]1> What's the double /?
2022-12-01 16:49:06 <ikke> linux treats double // as /
2022-12-01 16:51:25 <scorpion2185[m]1> is it simply ~/packages/apps then?
2022-12-01 16:52:51 <ikke> yes
2022-12-01 16:53:20 <scorpion2185[m]1> I am missing:   `17 | #include <linux/vt.h>`
2022-12-01 16:53:44 <scorpion2185[m]1> do I need the headers? 
2022-12-01 16:54:10 <ikke> https://pkgs.alpinelinux.org/contents?file=vt.h&path=%2Fusr%2Finclude%2Flinux&name=&branch=edge
2022-12-01 16:57:05 <scorpion2185[m]1> should that go to the makedepends?
2022-12-01 16:57:20 <scorpion2185[m]1> is there a wildcard for the headers package?
2022-12-01 16:57:45 <ikke> yes, it should go in makedepends, no, there is no wildcard
2022-12-01 16:58:17 <ikke> or do you mean for pkgs.alpinelinux.org?
2022-12-01 16:59:23 <scorpion2185[m]1> yes I'd like to make it in a way that it is suitable for official alpine
2022-12-01 16:59:27 <ikke> But for makedepends, you need to list concrete specific packages
2022-12-01 17:00:11 <scorpion2185[m]1> because for example on pmOS you have linux-postmarketos-allwinner-dev as headers
2022-12-01 17:00:26 <psykose> it is linux-headers
2022-12-01 17:01:19 <ikke> linux-headers should not be kernel flavor specific (afaik)
2022-12-01 17:08:48 <scorpion2185[m]1> it seems stuck at `>>> smlock*: Running postcheck for smlock`
2022-12-01 17:18:46 <scorpion2185[m]1> something wrong with the postcheck
2022-12-01 17:22:16 <ikke> can you show the complete output?
2022-12-01 17:24:52 <scorpion2185[m]1> ikke: https://paste.centos.org/view/c00a38a5
2022-12-01 17:26:08 <scorpion2185[m]1> I do have packages/apps/ in my home now but it states "No such file...."
2022-12-01 17:26:30 <scorpion2185[m]1> ah wait do I miss `options="!check" # has no tests` ?
2022-12-01 17:27:54 <ikke> for now it's still a warning, so it should not cause the build to fail
2022-12-01 17:28:45 <ikke> in any case, normally postcheck should tell you what the issue is if it found any
2022-12-01 17:28:56 <ikke> that it just exits means something unpexected happens
2022-12-01 17:30:09 <scorpion2185[m]1> it is stuck there
2022-12-01 17:30:29 <ikke> scorpion2185[m]1: can you try to run abuild with -v
2022-12-01 17:30:45 <scorpion2185[m]1> and -r?
2022-12-01 17:30:49 <ikke> yes
2022-12-01 17:35:04 <scorpion2185[m]1> https://paste.centos.org/view/8ef66b21
2022-12-01 17:35:50 <ikke> that does not help..
2022-12-01 17:36:15 <ikke> wondering what's going on
2022-12-01 17:37:28 <psykose> what environment are you building this in
2022-12-01 17:39:58 <scorpion2185[m]1> env?
2022-12-01 17:40:21 <scorpion2185[m]1> do you need to set one?
2022-12-01 17:51:09 <scorpion2185[m]1> abuild-keygen -a -i ?
2022-12-01 17:57:52 <scorpion2185[m]1> ikke: other part https://paste.centos.org/view/5381e72b
2022-12-01 18:06:40 <marshallishere> hey guys
2022-12-01 18:08:43 <marshallishere> I was trying the full disk encryption secure boot guide from the wiki, but I'm not doing with lvm, instead I'm doing with btrfs. I got everything setted up but I can't figure out one step, the creation of /root/grub-pre.cfg. I don't know what to put on the root part, can anyone help me?
2022-12-01 18:09:24 <marshallishere> I'm basically with a btrfs partition instead of lvm, haven't created subvolumes yet
2022-12-01 18:10:00 <marshallishere> Actually I was planning to do an installation without subvolumes first
2022-12-01 18:10:10 <marshallishere> I'm on a vm btw, testing before installing 
2022-12-01 18:10:15 <lopid> can you experiment with a vm?
2022-12-01 18:10:19 <lopid> o
2022-12-01 18:10:48 <marshallishere> I never made this type of installation lol
2022-12-01 18:11:15 <elly> I tried to follow that guide and failed, for what it's worth
2022-12-01 18:11:19 <marshallishere> I opened the luks partition as cryptroot, so it's on /dev/mapper/cryptroot
2022-12-01 18:11:20 <elly> I think it might be a bit out of date
2022-12-01 18:11:39 <marshallishere> last edit was november 30
2022-12-01 18:11:56 <elly> I ended up doing a regular install instead with `setup-alpine` and asking it to set up an encrypted root for me, and that worked
2022-12-01 18:11:57 <marshallishere> I don't even want swap, just the root and boot
2022-12-01 18:12:33 <marshallishere> elly: I thought on doing this, but as the alpine has a tutorial for fde I decided to try it out
2022-12-01 18:12:42 <elly> setup-alpine does not set up lvm for you (I think) so if you do want lvm you probably will need that guide
2022-12-01 18:12:56 <elly> but if you don't want lvm, just encrypted root, `setup-alpine -m sys` and tell it you want an encrypted root when it asks
2022-12-01 18:13:03 <marshallishere> it has a lvm option actually
2022-12-01 18:13:40 <marshallishere> elly: I want a boot and btrfs partition, going to do the rest with subvolumes
2022-12-01 18:13:49 <marshallishere> no swap
2022-12-01 18:13:59 <marshallishere> maybe zram later
2022-12-01 18:14:02 <ikke> scorpion2185[m]1: can you try: FAKEROOT= abuild -r
2022-12-01 18:14:05 <marshallishere> or swapfile
2022-12-01 18:14:17 <elly> gotcha
2022-12-01 18:14:23 <elly> good luck :)
2022-12-01 18:14:31 <marshallishere> thanks
2022-12-01 18:18:20 <minimal> marshallishere: that wiki article looks suspect to me, I don't understand why it is creating a grub-pre.cfg file
2022-12-01 18:20:32 <elly> are we talking about https://wiki.alpinelinux.org/wiki/LVM_on_LUKS ?
2022-12-01 18:20:58 <minimal> elly: I assume this: https://wiki.alpinelinux.org/wiki/Full_disk_encryption_secure_boot
2022-12-01 18:21:09 <elly> or about https://wiki.alpinelinux.org/wiki/Setting_up_a_laptop
2022-12-01 18:21:13 <elly> ha, there are several guides :)
2022-12-01 18:21:33 <minimal> wel lhe said "secure boot", i.e. UEFI Secure boot
2022-12-01 18:21:33 <elly> imo, all of these are obsolete since setup-alpine got encrypted root support
2022-12-01 18:21:57 <minimal> elly@ setup-disk doesn't do UEFI Secure Boot
2022-12-01 18:22:36 <marshallishere> elly: but setup alpine doesn't allow me to do btrfs and creates a swap partition
2022-12-01 18:22:43 <marshallishere> I don't want this to happen
2022-12-01 18:23:15 <minimal> marshallishere: you can pass a swap size of "0" to setup-disk AFAIK to have no swap partition
2022-12-01 18:23:24 <marshallishere> hmm
2022-12-01 18:24:15 <elly> it doesn't do btrfs though I think
2022-12-01 18:24:16 <minimal> anyway that guide in the "Configuring LUKS" section is calling "cryptsetup luksFormat" with no version specified, which means its LUKS v2 (from memory)
2022-12-01 18:24:20 <elly> I am using ext4
2022-12-01 18:25:11 <minimal> but later in the gruide with the "custom" grub-pre.cfg file it loading Grub's "luks" module which is for LUKS v1, not v2
2022-12-01 18:26:39 <marshallishere> hmm suspect
2022-12-01 18:27:16 <minimal> and its having /boot as part of the rootfs partition (and so encrypted) but booting from a encrypted LUKS *v2* partition does not work with current Grub...
2022-12-01 18:28:20 <marshallishere> actually I think it works but you need to set custom encryption algorithms, I think that the luksFormat command does this
2022-12-01 18:28:25 <minimal> support for encrypted v2 boot was supposed to have been introduced in Grub 2.06 but it doesn't work and only some (but not all required) patches are in Grub mainline and so are unlikely to appear in next Grub release (2.16?) due to shortly
2022-12-01 18:32:00 <minimal> marskallishere: with luksFormat you can do "--type luks1 or "--type luks2"
2022-12-01 18:32:25 <minimal> if you don't specify either I believe the version of cryptsetup in Alpine 3.17/Edge defaults to luks2
2022-12-01 18:32:59 <marshallishere> I saw on 'setup-disk --help' that it can support btrfs with some sort of ROOTFS, I tried using this as a variable but it didn't work, what is this option?
2022-12-01 18:33:33 <marshallishere> minimal: If the setup works with btrfs I will use the setup
2022-12-01 18:33:38 <marshallishere> easier lol
2022-12-01 18:34:32 <ikke> marshallishere: ROOTFS=btrfs setup-disk ..
2022-12-01 18:35:48 <ikke> https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-disk.in#L9
2022-12-01 18:36:44 <marshallishere> oh
2022-12-01 18:36:46 <marshallishere> makes sense
2022-12-01 18:36:49 <marshallishere> going to try
2022-12-01 18:39:10 <minimal> just checked cryptsetup source, luks v2 is the default
2022-12-01 18:39:35 <minimal> so that wiki guide makes no sense
2022-12-01 18:41:58 <marshallishere> it worked
2022-12-01 18:43:45 <marshallishere> so, there's any way of passing these parameters for the setup-alpine? This way the system installs properly configured
2022-12-01 18:44:45 <ikke> marshallishere: as they are environment variables, you can do the same for setup-alpine
2022-12-01 18:44:46 <minimal> marshallishere: ROOTFS=btrfs SWAP_SIZE=0 USE_CRYPT=1 setup-disk ?
2022-12-01 18:45:00 <minimal> those are mentioned in "setup-disk -h" from memory
2022-12-01 18:47:03 <marshallishere> thanks guys
2022-12-01 18:47:32 <marshallishere> one more question, when I use kde and install the partitionmanager it doesn't seems to recognize the disks, but gparted does recognizes
2022-12-01 18:47:36 <marshallishere> what could be the problem?
2022-12-01 18:48:27 <minimal> what/which partition manager?
2022-12-01 18:49:02 <marshallishere> the kde one
2022-12-01 18:49:07 <elly> I wonder if it's something to do with access to the disk devices?
2022-12-01 18:49:30 <marshallishere> hmm
2022-12-01 18:52:25 <marshallishere> the btrfs and swap worked as expected, thanks guys
2022-12-01 19:23:08 <M32573i12[m]> Is it possible to install Alpine Linux on Android devices? 
2022-12-01 19:23:29 <ikke> M32573i12[m]: there is a related project called postmarketOS, which does
2022-12-01 19:23:52 <M32573i12[m]> Cool
2022-12-01 19:26:24 <Solarbaby> M32573i12[m], not exactly what you asked but I would like to tell you I installed Alpine Linux on Android (Meta Quest 2 VR Headset) by enabling developer mode, sideloading android apk of termux, qemu, emulate aarch64 AlpineLinux.iso
2022-12-01 19:27:14 <ikke> So you can install apks, while you install apks
2022-12-01 19:27:36 <Solarbaby> weirdness huh?
2022-12-01 19:27:54 <Solarbaby> sometimes I dont even hear what I'm saying sometimes
2022-12-01 19:28:02 <blt[m]> I have a brandnew laptop where I installed Alpine Linux and sway. I realize that sway is not starting and a lot of kernel modules as well but doing a service udev-trigger restart is fixing everything as all the discovered modules are started. so I added udev in default but if I don't add udev-trigger in default I got the same situation after everyboot, which make sense to me, so my question is a good practice to add udev-trigger in default
2022-12-01 19:28:02 <blt[m]> and so let udev detect the required module for the system to work perflectly or better to add all the module manually?
2022-12-01 19:29:50 <jh81> ugh! im too tired for this shit :)
2022-12-01 19:30:10 <scorpion2185[m]1> ikke: that doesn't help?
2022-12-01 19:30:33 <Habbie> Solarbaby, wouldn't chroot be a lot faster than qemu?
2022-12-01 19:31:19 <psykose> blt[m]: setup-devd udev indeed does all of that for you
2022-12-01 19:31:28 <psykose> trigger+settle+postmount
2022-12-01 19:31:36 <Solarbaby> Habbie, Yeah totally faster but I don't have root on the quest 2.. just what they call developer mode is the stripped down access they give me
2022-12-01 19:31:45 <Habbie> oh, right, chroot doesn't work in termux
2022-12-01 19:32:18 <Solarbaby> I'm rooting for pkvm to come to the quest kernel and then maybe qemu will get faster
2022-12-01 19:32:39 <Ermine> Habbie: one needs proot for termux.
2022-12-01 19:32:54 <blt[m]> psykose: yes thanks I saw this script but it is a one time script or doest it populate something?
2022-12-01 19:33:03 <psykose> it does what you are doing
2022-12-01 19:33:30 <blt[m]> how should I call that after every boot?
2022-12-01 19:34:17 <marshallishere> hey, I installed librewolf through flatpak and my font is really bad, can anyone help?
2022-12-01 19:34:18 <marshallishere> https://envs.sh/Qsr.png
2022-12-01 19:34:24 <Habbie> Ermine, ah, learned a thing, thanks!
2022-12-01 19:35:14 <scorpion2185[m]1> ikke: I tried `FAKEROOT= abuild -r` same as before. 
2022-12-01 19:35:33 <scorpion2185[m]1> can I remove the postcheck since it is stuck there?
2022-12-01 19:35:41 <ikke> no
2022-12-01 19:36:52 <psykose> blt[m]: you call it once? it adds what you are to the correct runlevels and you never use it again
2022-12-01 19:36:55 <psykose> it's a 5 line script
2022-12-01 19:36:56 <blt[m]> psykose: ok I am reading the code, it seems indeed that it does what I need and push that into sysinit 
2022-12-01 19:37:40 <blt[m]> it seems that I have to force udev-trigger after the reboot but I will try it again, like my modules are not loaded
2022-12-01 19:39:09 <blt[m]> psykose: it is supposed to add the udev-* but it does not
2022-12-01 19:40:15 <psykose> what exactly do you type
2022-12-01 19:40:40 <blt[m]> setup-devd
2022-12-01 19:40:40 <blt[m]> udev
2022-12-01 19:40:40 <blt[m]> y
2022-12-01 19:41:06 <psykose> and what does rc-update show after
2022-12-01 19:41:23 <blt[m]> it doesnt show anyting related to udev....
2022-12-01 19:41:38 <minimal> blt[m]: are you using Alpine in run-from-ram or run-from-disk (Sys-mode)
2022-12-01 19:41:40 <blt[m]> but I see the code it is suposed to do line 82 to 85
2022-12-01 19:41:51 <blt[m]> minimal: sys-mode
2022-12-01 19:42:22 <psykose> could you pastebin your whole console output
2022-12-01 19:42:28 <blt[m]> should I try mdev instead?
2022-12-01 19:45:41 <minimal> blt[m]: so you have the eudev package installed?
2022-12-01 19:45:58 <blt[m]> yes
2022-12-01 19:46:14 <minimal> and "setup-devd udev" gave no errors?
2022-12-01 19:46:27 <blt[m]> https://paste.debian.net/1262552
2022-12-01 19:47:24 <minimal> try is as root directly?
2022-12-01 19:47:27 <minimal> s/is/it/
2022-12-01 19:47:28 <alpine-meetbot> minimal meant to say: try it as root directly?
2022-12-01 19:47:31 <blt[m]> https://paste.debian.net/1262553
2022-12-01 19:47:50 <blt[m]> yes I did the exact same command as root just before same result
2022-12-01 19:48:02 <minimal> which alpine version?
2022-12-01 19:48:08 <blt[m]> 3.17
2022-12-01 19:48:41 <minimal> strange, I use "setup-devd udev" often
2022-12-01 19:50:23 <psykose> if you run `rc-update add --quiet udev` yourself does it change anything
2022-12-01 19:51:09 <blt[m]> oh yes it solve the issue, I know how to do it manually but my point is that setup-devd is supposed to do that...
2022-12-01 19:53:46 <psykose> strange
2022-12-01 19:53:50 <psykose> i tested it myself and it works
2022-12-01 19:54:11 <marshallishere> I solved the font issue installing xdg-desktop-portal-gtk
2022-12-01 19:54:22 <psykose> i guess you can set -x in the script and see why it fails :)
2022-12-01 19:57:33 <blt[m]> it doesn't seem -x to work?
2022-12-01 19:57:53 <blt[m]> just for info, I just rebooted and yes adding theses manually fix all the issues :)
2022-12-01 19:58:10 <blt[m]> but still don't understant why setup-devd is not working on my case
2022-12-01 19:58:47 <minimal> blt[m]: what happened when you tried "-x" exactly?
2022-12-01 20:01:58 <blt[m]> minimal: so you meant setup-devd -x correct ?
2022-12-01 20:02:49 <minimal> blt[m]: nope, psykose was suggesting setting the "shebang" line at the top of setup-devd to include "-x" to trace the shellscript
2022-12-01 20:03:25 <psykose> you can also do sh -x /sbin/setup/devd
2022-12-01 20:03:37 <psykose> setup-devd*
2022-12-01 20:17:22 <scorpion2185[m]1> ikke: can you try to build it? it is a little package or someone else
2022-12-01 20:25:07 <ikke> scorpion2185[m]1: can you share it?
2022-12-01 20:34:04 <scorpion2185[m]1> ikke: https://paste.centos.org/view/d6ee09a9
2022-12-01 20:36:17 <ikke> Not sure if that comes from paste.centos.org, but if I download the raw output of that paste, it appears it has dos line endings
2022-12-01 20:36:21 <ikke> does it have for you?
2022-12-01 20:36:52 <ikke> Also, it appears the checksums are missing
2022-12-01 20:37:41 <ikke> also, checking the source, '
2022-12-01 20:37:44 <ikke> 'cannot retrieve password entry (make sure to suid smlock)'
2022-12-01 20:37:53 <ikke> suid programs require a security review
2022-12-01 20:38:03 <ikke> (we try to avoid them)
2022-12-01 20:39:35 <scorpion2185[m]1> suid? didn't see that , I didn't copy the checksum
2022-12-01 20:40:30 <ikke> scorpion2185[m]1: for me it outputs >>> ERROR: smlock*: Found executable files with SUID bit set:
2022-12-01 20:40:34 <ikke> That's from postcheck
2022-12-01 20:40:39 <ikke> not sure why you didn't see the message
2022-12-01 20:40:45 <scorpion2185[m]1> ssh?
2022-12-01 20:41:00 <ikke> sorry?
2022-12-01 20:41:18 <scorpion2185[m]1> I did it from ssh 
2022-12-01 20:41:26 <ikke> Me too
2022-12-01 20:42:32 <scorpion2185[m]1> how do you fix the executable file with SUID?
2022-12-01 20:43:12 <M32573i12[m]> Centos doesn't seems to be Alpine Linux discussion 
2022-12-01 20:43:26 <ikke> M32573i12[m]: this has nothing to do with centos
2022-12-01 20:43:40 <scorpion2185[m]1> i find their pastebin nice
2022-12-01 20:44:13 <M32573i12[m]> ikke: Sure, you make the rules
2022-12-01 20:44:33 <scorpion2185[m]1> it is just a pastebin service 
2022-12-01 20:50:02 <ikke> scorpion2185[m]1: You can remove suid, but that means you need to run it as root
2022-12-01 20:59:47 <marshallishere> hey guys, I was trying to follow the secure boot guide but when I get to the 'apk fix kernel-hooks' it gives me an error because I don't have /boot/efi, but I've used the script setup-disk to install alpine
2022-12-01 21:02:16 <marshallishere> do I need to make manual partitions?
2022-12-01 21:03:04 <marshallishere> https://wiki.alpinelinux.org/wiki/UEFI_Secure_Boot
2022-12-01 21:03:06 <marshallishere> the guide
2022-12-01 21:04:09 <Ermine> Do you have EFI system partition? If yes, mount it at /boot/efi .
2022-12-01 21:04:27 <marshallishere> I have a vfat partition mounted at /boot
2022-12-01 21:05:33 <marshallishere> I think the installer didn't create the EFI partition
2022-12-01 21:05:52 <Ermine> I guess this is the EFI partition.
2022-12-01 21:05:56 <psykose> M32573i12[m]: where is the centos discussion
2022-12-01 21:11:56 <marshallishere> Ermine: I tried to set the /etc/fstab to be /boot/efi and it didn't work, also I think this is the normal boot, alpine isn't creating an efi partition by default
2022-12-01 21:12:44 <Ermine> do you have EFI directory on that partition by the way?
2022-12-01 21:12:49 <marshallishere> yep
2022-12-01 21:13:34 <Ermine> what is its contents?
2022-12-01 21:13:55 <Ermine> s/is/are/
2022-12-01 21:13:55 <alpine-meetbot> Ermine meant to say: what are its contents?
2022-12-01 21:15:25 <marshallishere> there's a directory called boot with a bootx64.efi and a directory called alpine with a grubx64.efi
2022-12-01 21:16:04 <Ermine> That means this is EFI system partition and it was detected by the installer.
2022-12-01 21:16:35 <marshallishere> but it didn't create the partition itself, it just placed in the folder
2022-12-01 21:16:40 <Ermine> kernel-hooks except it to be mounted at /boot/efi, though.
2022-12-01 21:16:50 <marshallishere> so, what can I do?
2022-12-01 21:20:06 <Ermine> did you add /boot/efi to your fstab?
2022-12-01 21:20:13 <marshallishere> nope
2022-12-01 21:20:14 <minimal> marshallishere: you installed (via setup-disk) with encryption, right? so the ESP partition is used for both ESP and as /boot
2022-12-01 21:20:44 <marshallishere> hmm
2022-12-01 21:20:54 <marshallishere> and how can I fix the problem with kernel-hooks?
2022-12-01 21:21:14 <minimal> marshallishere: as was mentioned earlier you are using a "guide" that is outdated and also may not be accurate
2022-12-01 21:22:14 <marshallishere> makes sense
2022-12-01 21:22:17 <minimal> marshallishere: you didn't answer my question, you are using EFI with encryption installed via setup-disk, correct?
2022-12-01 21:22:30 <marshallishere> yes
2022-12-01 21:22:53 <minimal> so you have 2 partitions, an ESP partition and a LUKS partition
2022-12-01 21:22:58 <marshallishere> yes
2022-12-01 21:23:21 <minimal> the 1st partition is mounted as what, /boot ?
2022-12-01 21:23:29 <marshallishere> yes
2022-12-01 21:24:04 <minimal> and inside that you have a EFI directory (so visible as /boot/EFI" ?
2022-12-01 21:24:22 <Ermine> marshallishere: open /etc/kernel-hooks.d/secureboot.conf and set output_dir to /boot/EFI/alpine ?
2022-12-01 21:24:32 <Ermine> s/?/./
2022-12-01 21:24:32 <alpine-meetbot> Ermine meant to say: marshallishere: open /etc/kernel-hooks.d/secureboot.conf and set output_dir to /boot/EFI/alpine .
2022-12-01 21:25:32 <marshallishere> okay
2022-12-01 21:28:09 <minimal> it a case issue, the directory inside /boot is "EFI" correct, but secureboot.hook is referring to it in lowercase
2022-12-01 21:28:36 <minimal> ?
2022-12-01 21:29:08 <marshallishere> that's it
2022-12-01 21:29:16 <Ermine> Worked?
2022-12-01 21:29:31 <marshallishere> I haven't tested, need to leave
2022-12-01 21:29:35 <marshallishere> later I'll come back
2022-12-01 21:29:39 <marshallishere> bye
2022-12-01 21:29:44 <Ermine> Well, bye
2022-12-01 21:31:20 <Ermine> Auto-detection would be cool, but it requires parsing /etc/fstab or /proc/mounts iiuc.
2022-12-01 21:32:46 <minimal> if/when secure boot support gets added to setup-disk this will all be ironed out
2022-12-01 21:33:12 <cache_return> can someone help me figure out how to recover from apk update? I have no modules loaded after update and reboot.
2022-12-01 21:33:57 <cache_return> at least according to lsmod
2022-12-01 21:34:33 <Ermine> did your system boot, or are you in emergency shell?
2022-12-01 21:34:46 <cache_return> it booted
2022-12-01 21:35:13 <cache_return> but no wifi, so I'm down in the dungeon with usb keyboard
2022-12-01 21:35:13 <Ermine> does dmesg or syslog say why aren't modules loaded?
2022-12-01 21:35:42 <cache_return> dmesg doesnt look like its trying to load e.g. the wireless card
2022-12-01 21:35:42 <minimal> cache_return: standard kernel? I'd assume it would need some modules loaded in order to boot
2022-12-01 21:36:28 <psykose> could you pastebin all the logs and all your commands
2022-12-01 21:36:48 <cache_return> minimal: yes, 5.15.55-0-rpi4
2022-12-01 21:36:49 <Piraty> how can i get a list of top-level packages that nothing else depends on ? i.e. what would be the minimum of what i need for an equivalent re-install
2022-12-01 21:37:06 <cache_return> psykose: let me figure out how to do that without internet
2022-12-01 21:37:14 <minimal> ah, RPI has a lot of stuff compiled in rather than loadable
2022-12-01 21:37:20 <Ermine> Piraty: cat /etc/apk/world
2022-12-01 21:37:45 <Ermine> cache_return: did you try loading something manually?
2022-12-01 21:37:55 <Piraty> isn't that the list of manual packages, which then would contain the ones that other depend on still ?
2022-12-01 21:37:58 <bl4ckb0ne> psykose: any blocker for mesa 22.3.0 i can help with?
2022-12-01 21:38:01 <psykose> there are probably 20 ways to do that without internet on the actual device, given how you are currently on it
2022-12-01 21:38:04 <psykose> bl4ckb0ne: it being .1
2022-12-01 21:38:42 <Ermine> Piraty: afaik this is a list of packages you ordered to install.
2022-12-01 21:38:47 <psykose> (as always we wait for the .1, that's all)
2022-12-01 21:38:52 <bl4ckb0ne> alrighty
2022-12-01 21:39:12 <Ermine> cache_return: can you paste approximate relevent error messages?
2022-12-01 21:39:17 <minimal> cache_return: which Alpine version are you on? 3.16? also, aarch64 or armv7?
2022-12-01 21:39:52 <cache_return> minimal: 3.16 aarch64
2022-12-01 21:39:54 <Ermine> minimal: btw does secure boot infrastructure exist on Alpine for a year now?
2022-12-01 21:41:13 <minimal> cache_return: v3.16 currently has the 5.15.78 kernel, so how did you update to 5.15.55?
2022-12-01 21:43:48 <minimal> Ermine: there have been some tools packaged for a while (secureboot-hook first packages July last year) but secure boot is not official supported (there's an Issue open in alpine-conf to have it added AFAIK)
2022-12-01 21:44:22 <psykose> minimal: it's the kernel in the 3.16.1 iso iirc
2022-12-01 21:44:53 <minimal> psykose: ah, ok, but an update would have used the repos though, right?
2022-12-01 21:44:59 <psykose> shrug
2022-12-01 21:46:06 <minimal> s/update/upgrade/
2022-12-01 21:46:06 <alpine-meetbot> minimal meant to say: psykose: ah, ok, but an upgrade would have used the repos though, right?
2022-12-01 21:46:27 <minimal> as "apk update" only updates the package indexes
2022-12-01 22:07:56 <cache_return> all I know is that I installed a month or two ago and just recently did update and upgrade because I was installing nextcloud
2022-12-01 22:08:18 <cache_return> it rebuilt the initramfs and niw here I am
2022-12-01 22:08:32 <cache_return> what logs do people want besides dmesg?
2022-12-01 22:34:18 <cache_return> dmesg: http://sprunge.us/ecJ6e8
2022-12-01 22:34:35 <cache_return> psykose: ^
2022-12-01 22:36:36 <garo> I am on a 3.17 and I can't seem to save the nftables status, this is what i do and what the result is: https://pastebin.com/sCv82vKc
2022-12-01 22:38:01 <psykose> cache_return: hmm
2022-12-01 22:38:20 <psykose> not sure why it wouldn't have brcm stuff there
2022-12-01 22:39:52 <psykose> all i could really suggest is upgrading to the latest versions (the classic) but that is a little hard without direct networking
2022-12-01 22:40:25 <psykose> for reference i also have a pi4 and mine looks like https://img.ayaya.dev/uxfZI7FMyiUA
2022-12-01 22:40:51 <dwfreed> garo: what do 'nft list counters' 'nft list limits' 'nft list quotas' output?
2022-12-01 22:48:12 <dwfreed> if you look at the initscript and the conf.d file, I think you'll find the explanation; since 3.17, the initscript does not save the ruleset by default; it can be told to do this, but equally you can just write ruleset files by hand and load them, since the syntax to the nft command is basically the same as what goes in the file (though you can define tables and chains in the file without having to 
2022-12-01 22:48:12 <dwfreed> use the add command)
2022-12-01 22:49:25 <dwfreed> the initscript defaults to saving counters, limits, and quotas, and from the commands in your pastebin, you have none of those, and thus the file is empty, because there's nothing to save of those types
2022-12-01 22:53:53 <dwfreed> the default /etc/nftables.nft includes /etc/nftables.d/*.nft, and you can put whatever snippets you want in there; you can redefine the same tables and chains multiple times, and as long as they don't have conflicting definitions, nft will merge them in definition order; so you can have little snippets for each logical grouping of rules
2022-12-01 22:58:15 <garo> You mention that it doesn't save it by default, but i did run "/etc/init.d/nftables save" manually
2022-12-01 22:59:37 <garo> Then which command should i run to create the .nft files ?
2022-12-01 23:04:06 <cache_return> psykose: this was _caused by_ updating, though.
2022-12-01 23:10:27 <psykose> i am not sure how you can "update" into a kernel from 5 months ago that is not even in the repositories
2022-12-01 23:10:43 <psykose> the only way you can update to that kernel version is by flashing the 3.16.1 iso
2022-12-01 23:11:22 <psykose> or having a really out of date mirror somewhere, i guess
2022-12-01 23:12:27 <psykose> debugging why that didn't actually update would probably yield the answer
2022-12-01 23:15:02 <dwfreed> garo: the save function is used whether you execute it as a command or during stop, the behavior is the same
2022-12-01 23:15:40 <dwfreed> (stop does some extra things, of course, and save is only run on stop if enabled)
2022-12-01 23:17:05 <garo> But that still doesn't generate the .nft files :-(
2022-12-01 23:17:16 <dwfreed> and that's intentional
2022-12-01 23:17:25 <garo> So which command does ?
2022-12-01 23:17:38 <dwfreed> you can write them by hand, the format is not bad
2022-12-01 23:20:53 <dwfreed> the output of 'nft list ruleset' can also be used, but note that if your input file has regular file comments (not rule comments), those would be missing from the list ruleset output, so it's not going to be a loseless roundtrip
2022-12-01 23:21:09 <garo> Would literally copying line 5 to 8 to a new file /etc/nftables.d/forward-443-t-o10443.nft work ?
2022-12-01 23:21:24 <dwfreed> yes
2022-12-01 23:22:50 <dwfreed> but you can also do it in a better style: table ip nat { chain prerouting { type nat hook prerouting priority -100; tcp dport 443 redirect to :1443; } }
2022-12-01 23:22:52 <garo> The syntax of my 3 commands is not the same as the one that 'nft list ruleset' outputs, do both work ?
2022-12-01 23:23:00 <dwfreed> yes
2022-12-01 23:23:02 <dwfreed> with newlines as appropriate
2022-12-01 23:23:58 <garo> Ok, thanks !
2022-12-01 23:34:52 <minimal> cache_return: which repo (site) did you use for updating? as well as the kernel being old for v3.16 the init (from mkinitfs) is old too...
2022-12-01 23:46:52 <minimal> cache_return: also note it is telling you to run fsck on the boot partition
2022-12-02 00:00:13 <cache_return> minimal: I will check the repo and get back to you. the fsck warning has been there since intial install. I would fsck, clean the dirty bit, and it would be back next boot every time.
2022-12-02 00:04:32 <cache_return> mirror.clarkson.edu
2022-12-02 00:05:23 <psykose> it has linux-rpi4-5.15.78 currently for 3.16 and seems to be up to date
2022-12-02 00:22:43 <minimal> cache_return: sounds like it's not being cleanly unomunted during shutdown/reboot
2022-12-02 00:29:37 <cache_return> maybe it updated something other than the kernel which broke module loading. I can try to get it some internet later and try ro update again.
2022-12-02 00:30:34 <cache_return> minimal: yes, I agree not unmounting correctly, but it wasn't causing a problem so I passed on examining it.
2022-12-02 00:33:47 <minimal> cache_return: wasn't causing a problem.....yet. Continuing to mount/use it without fsck could gradually make things worse
2022-12-02 00:35:42 <minimal> cache_return: "apk -Ua upgrade" ?
2022-12-02 00:42:15 <marshallishere> I'm back
2022-12-02 00:56:25 <cache_return> will try
2022-12-02 01:27:08 <cache_return> is this correct/up to date? https://wiki.alpinelinux.org/wiki/Alpine_Package_Keeper#Upgrading_"diskless"_and_"data"_disk_mode_installs
2022-12-02 01:27:46 <cache_return> was I supposed to do update-kernel after apk upgrade?
2022-12-02 01:27:58 <cache_return> because I did not do that.
2022-12-02 01:33:23 <minimal> cache_return: ah, you're using run-from-ram, kernels are handled differently for that.........don't use run-from-ram myself
2022-12-02 01:33:51 <cache_return> minimal: no, sysmode
2022-12-02 01:34:05 <cache_return> so this doesn't apply?
2022-12-02 01:35:51 <minimal> update-kernel is only for run-from-ram AFAIK
2022-12-02 01:36:11 <minimal> that wiki does mention "diskless"
2022-12-02 01:38:24 <cache_return> I didn't know what data disk referred to
2022-12-02 01:38:56 <cache_return> although I will say that sysmode conversion was pretty iffy for this device, so that cpuld be a cause of this.
2022-12-02 01:42:20 <minimal> cache_retrurn: it is still unclear how your update ended up with out-of-date packages
2022-12-02 01:44:08 <cache_return> my working theory is that the kernel did not update
2022-12-02 01:44:29 <cache_return> that the running kernel is the pre-update kernel
2022-12-02 01:45:03 <minimal> does the /lib/modules/<current kernel version> directory exist?
2022-12-02 02:15:18 <cache_return> it does
2022-12-02 02:16:54 <cache_return> Wait, let me check the version
2022-12-02 02:21:50 <cache_return> nope! it's /lib/modules/5.15.78
2022-12-02 02:22:05 <cache_return> and kernel *.55
2022-12-02 02:22:34 <minimal> so it updated the modules but not the kernel itself
2022-12-02 02:22:52 <minimal> that's why modules don't load - there are no modules for the current kernel version
2022-12-02 02:23:10 <cache_return> yes
2022-12-02 02:23:24 <minimal> the RPI's boot partition is mounted as /boot ?
2022-12-02 02:23:40 <cache_return> it should be, let me check
2022-12-02 02:25:52 <cache_return> /dev/mmcblk0p1/ is mounted at /media/mmcblk0p1 and /boot is a symlink to that
2022-12-02 02:26:35 <minimal> so a "ls -l /boot" shows initramfs-rpi4 and vmlinuz-rpi4 files?
2022-12-02 02:27:10 <minimal> oops "ls -l /boot/" (slash on the end
2022-12-02 02:27:17 <cache_return> yes.
2022-12-02 02:27:42 <minimal> and the dates of those files are old or look like when you did the upgrade?
2022-12-02 02:27:55 <cache_return> one was last updated yesterday when I did the update (initramfs) and the other last updated on 11/11
2022-12-02 02:28:09 <cache_return> vmlinuz
2022-12-02 02:28:15 <minimal> that's the kernel
2022-12-02 02:29:22 <minimal> so then in the /boot/config.txt file what's the values of "kernel=" and "initramfs=" ?
2022-12-02 02:29:47 <minimal> in the "[cm4]" section if there is one
2022-12-02 03:07:52 <cache_return> it says "kernel=boot/vmlinuz-rpi4" and "initramfs boot/initramfs-rpi4"
2022-12-02 03:08:14 <cache_return> not /boot and not initramfs=
2022-12-02 03:09:39 <minimal> cache_return: ok, so if you do a "ls -l /boot/boot/" are you seeing DIFFERENT (size / dated) vmlinuz and initramfs files?
2022-12-02 03:34:52 <cache_return> minimal: both files are there, but with different sizes and dated 10/16
2022-12-02 03:35:14 <minimal> cache_return: right, that's the problem then
2022-12-02 03:35:57 <minimal> so (a) when the new kernel was installed it was written to /boot/ and a generated/regenerated initfs also to /boot
2022-12-02 03:36:26 <minimal> however the RPI's bootloader is configured to load kernel & initramfs from "boot/" (relative to the boot partition)
2022-12-02 03:36:38 <minimal> so it loads an older set of files ;-)
2022-12-02 03:36:58 <minimal> meanwhile the installation of the newer kernel resulted in the old version kernel modules being deleted
2022-12-02 03:40:20 <cache_return> I can try replacing with /boot in config.txt
2022-12-02 03:40:52 <cache_return> would like to just remove /boot/boot. Not sure why it exists.
2022-12-02 03:41:23 <minimal> cache_return: no, you don't use "/boot" in config.txt as that would refer to the existing boot directory on the boot partition
2022-12-02 03:42:07 <minimal> also you need to be careful changing where you tell config.txt to load kernel and initram as there may be other files in there it wants to load also
2022-12-02 03:42:19 <cache_return> because it is read before rootfs is mounted?
2022-12-02 03:42:29 <cache_return> okay, then do what?
2022-12-02 03:42:39 <minimal> it is not read by Linux, it is read by the RPI's own bootloader
2022-12-02 03:43:09 <cache_return> change to kernel=vmlinuz-rpi4?
2022-12-02 03:44:21 <cache_return> is there a way to test these vmlinuzes to see what kernel version is in them?
2022-12-02 03:44:46 <minimal> that would read the vmlinuz from the top-level of the boot partition - note my point about other files though.....you'd have to check config.txt, usercfg.txt, etc to see what files are loaded from where
2022-12-02 03:45:05 <minimal> you don't want to end up with a non-booting system
2022-12-02 03:45:22 <cache_return> honestly the problem seems to be that config.txt will be overqritten by updates
2022-12-02 03:45:44 <minimal> "file /boot/vmlinuz-rpi4"
2022-12-02 03:46:31 <cache_return> file: not found
2022-12-02 03:46:58 <minimal> I'd need to check my notes but (from memory) config.txt is not changed once an installation is done (as it is not packaged by Alpine, it is created by the setup scripts instead)
2022-12-02 03:47:11 <minimal> I'd have to doublecheck though, I'm just relying on memory
2022-12-02 03:47:19 <minimal> ah, you don't have "file" installed
2022-12-02 03:47:29 <cache_return> the text of it says it will be overwritten
2022-12-02 03:47:44 <cache_return> file didn't come with the image
2022-12-02 03:47:58 <minimal> cache_return: don't believe everything you read ;-) Though like I said I'd have to doublecheck
2022-12-02 03:49:21 <psykose> the default does have it like that
2022-12-02 03:49:23 <cache_return> sounds like either kernel=vmlinuz-rpi4 or removing /boot/boot and doing ln -s /media/mmcblk0p1 /boot/boot
2022-12-02 03:49:24 <psykose> don't remember why though
2022-12-02 03:49:32 <psykose> ah right
2022-12-02 03:49:43 <psykose> there is a boot -> . symlink isn't there
2022-12-02 03:49:44 <minimal> psykose: yeah I know it has that message, but I don't think it is actually accurate
2022-12-02 03:49:54 <psykose> so boot/vmlinuz just points to vmlinuz in cwd
2022-12-02 03:50:04 <psykose> and for some reason it's an actual folder here
2022-12-02 03:50:10 <psykose> no idea why it was set up with the symlink though
2022-12-02 03:50:41 <cache_return> so I should replace it with the symlink?
2022-12-02 03:50:45 <psykose> minimal: yeah it's only mkimg that makes it so nothing overrides it
2022-12-02 03:50:48 <psykose> cache_return: sure
2022-12-02 03:50:49 <minimal> cache_return: older Alpine versions had stuff in boot subdirectory, then a RPI Bootloader bug was discovered that caused problems if the boot partition was labelled "boot" and so a change was made to but stuff in top-level
2022-12-02 03:50:57 <psykose> aha
2022-12-02 03:50:59 <psykose> that was the one
2022-12-02 03:51:03 <psykose> good memory
2022-12-02 03:51:15 <cache_return> I suffeted from that bug on initial install
2022-12-02 03:51:31 <psykose> personally i just edit the file to have no boot/ prefix
2022-12-02 03:51:44 <psykose> and have no boot anything in there
2022-12-02 03:51:46 <minimal> cache_return: your problem is as it's a CM4 you couldn't install from scratch using SDcard with setup-disk
2022-12-02 03:51:52 <psykose> but i'm more confident with doing random stuff :)
2022-12-02 03:52:18 <cache_return> oh, I know
2022-12-02 03:52:30 <minimal> I'm trying to remember what settings control where to find overlays (as I'm guessing you have an them inside the boot subdir)
2022-12-02 03:53:07 <minimal> psykose: for a RPI4 is easy to try risk stuff as you just pull the SDcard, in his case he's boot off builtin flash
2022-12-02 03:53:11 <cache_return> I'm going to nuke /boot/boot for a symlink
2022-12-02 03:53:23 <cache_return> minimal: no overlays in /boot/boot
2022-12-02 03:53:31 <cache_return> only in /boot
2022-12-02 03:53:44 <minimal> so what else is in /boot/boot apart from the kernel and initramfs files?
2022-12-02 03:53:52 <psykose> sure, rm -r boot, ln -s . boot
2022-12-02 03:54:37 <cache_return> minimal: system map, bootcode, config-rpi4, fixup.dat, start.elf
2022-12-02 03:54:54 <minimal> psykose: *selectively* good memory, I switched to zswap to fit more in ;-)
2022-12-02 03:55:11 <psykose> ;)
2022-12-02 03:55:51 <minimal> cache_return: is there bootcode, fixup, and start.elf also in the top-level of boot partition?
2022-12-02 03:56:30 <cache_return> yes but different ones
2022-12-02 03:56:46 <minimal> I assume newer in top-level
2022-12-02 03:57:06 <minimal> those are files the bootloader uses, rather important ;-)
2022-12-02 03:57:09 <cache_return> actually, some could be same but with different dates.
2022-12-02 03:58:00 <minimal> the system map and config-rpi4 come with the kernel package so there'll be versions already in top-level
2022-12-02 03:59:00 <cache_return> no system map in /boot
2022-12-02 03:59:20 <minimal> so *in theory* changing the config.txt references to remove "boot/" from both kernel and initramfs should be ok
2022-12-02 03:59:37 <cache_return> ln -s . boot operarion not permitted
2022-12-02 03:59:58 <minimal> no system map? strange, its not critical AFAIK, just used for "debug info" if the kernel panics, right psykose?
2022-12-02 04:00:01 <psykose> ole root permissions needed just like every other step here
2022-12-02 04:00:08 <psykose> minimal: yeah, used for nothing
2022-12-02 04:00:10 <minimal> cache_return: boot is FAT, you can't symlink
2022-12-02 04:00:23 <psykose> ah that too
2022-12-02 04:00:24 <psykose> hmm
2022-12-02 04:00:29 <cache_return> I am root
2022-12-02 04:00:52 <cache_return> oh, wow I did not know that.
2022-12-02 04:00:57 <psykose> i guess you just edit the files then and remove the boot
2022-12-02 04:01:03 <minimal> cache_return: no, I mean symlinking ("ln -s" is a Unix/Linux filesystem thing, not a FAT filesystem thing
2022-12-02 04:01:04 <psykose> i forgot the whole rest of the other setup :)
2022-12-02 04:01:09 <cache_return> I guess so.
2022-12-02 04:01:39 <minimal> psykose: yeah I'm working from memory too, I don't often tinker with RPI bootloader setup once I got things the way I like them
2022-12-02 04:01:59 <psykose> more stable than i expect it to be really
2022-12-02 04:02:06 <psykose> i just set some overclock ages ago and it's still there
2022-12-02 04:02:26 <cache_return> here we go...
2022-12-02 04:08:24 <minimal> "now playing Simon & Garfunkel's The Sound Of Silence" ;-)
2022-12-02 04:08:37 <minimal> cache_return: so?
2022-12-02 04:19:07 <cache_return> that was it. It worked.
2022-12-02 04:20:24 <minimal> good news
2022-12-02 04:25:25 <cache_return> thanks for the help
2022-12-02 04:25:38 <cache_return> as usual
2022-12-02 05:13:15 <Sheila> clandmeter: ping. inquiry: for whatever reason packages in Adélie which had contributions from you had elided your email address. is this something you requested, and if so is it something you'd like to remain the case?
2022-12-02 08:26:46 <ForeverNoob[m]> Yo, when is the next LTS kernel update gonna hit? I'm trying to test my unattended upgrade script on real hardware lol
2022-12-02 08:28:19 <psykose> probably after it is released and not before it exists
2022-12-02 08:31:01 <ForeverNoob[m]> Hmm, I think you're on to something. I must investigate this further.
2022-12-02 09:09:44 <eeijat> Hello, I have a question. Is there an option to not to install any bootloader during initial setup?
2022-12-02 10:04:41 <markand> plop
2022-12-02 10:05:01 <markand> installing shadow-login does not seem to read /etc/environment (at least from ssh)
2022-12-02 10:06:06 <grawity> normally sshd doesn't invoke `login` at all (it used to be an option, but no longer is)
2022-12-02 10:06:12 <markand> even though pam_env is included in /etc/pam.d/system-login 
2022-12-02 10:06:17 <markand> I've tried in tty as well
2022-12-02 10:17:55 <marshallishere> Ermine: hey, it worked.
2022-12-02 10:18:18 <marshallishere> I think I'm ready to try on real hardware
2022-12-02 15:01:46 <msrd0> Hi! Can I find a list of alpine repository signing keys for a specific alpine version/architecture somewhere?
2022-12-02 15:08:20 <minimal> eeijat: setup-disk currently always installs a bootloader (well apart from for Raspberry Pi which has one builtin, but it does create config for that).
2022-12-02 15:08:55 <minimal> eeijat: why would you not want a bootloader installed/configured? what exactly are you trying to do?
2022-12-02 16:13:04 <ptrc> msrd0: usually it's in an alpine-keys package in the repos
2022-12-02 16:14:05 <ptrc> see here: https://gitlab.alpinelinux.org/alpine/aports/-/blob/aab68f8c/main/alpine-keys/APKBUILD#L13
2022-12-02 16:14:13 <ptrc> you can also select a specific version tag from the dropdown
2022-12-02 16:59:57 <eeijat> minimal, I use EFI stub and so any bootloader on my machines is dead weight (I'm not sure if its the right words anyway I just don't need them)
2022-12-02 17:02:37 <minimal> eeijat: right, I assumed that's what you were doing. At some point in the future setup-alpine/setup-disk might support using EFISTUB but it currently does not
2022-12-02 17:03:33 <minimal> so for now you'd have to install with Grub used and then afterwards remove it and manually setup EFISTUB instead
2022-12-02 17:11:48 <eeijat> minimal, yes that's what I'm doing for now. And such an option could be very helpful, thanks for respond
2022-12-02 17:19:03 <scorpion2185[m]1> ikke:  how can I remove suid, so I can package it? that app needs root indeed IIRC
2022-12-02 17:19:38 <ikke> chmod u-s $pkgdir/path/to/smlock
2022-12-02 17:36:39 <scorpion2185[m]1> to the bin?
2022-12-02 17:38:09 <ikke> yes
2022-12-02 17:39:53 <scorpion2185[m]1> the bin was in pkg
2022-12-02 17:40:13 <ikke> yes, so $pkgdir
2022-12-02 17:40:13 <scorpion2185[m]1> pkg is in the same folder of the APKBUILD. it is stuck in the postprocess again
2022-12-02 17:40:38 <ikke> scorpion2185[m]1: is it hanging or does it stop?
2022-12-02 17:40:46 <scorpion2185[m]1> hanging
2022-12-02 17:46:20 <ikke> can you check with (h)top what is happening (in another session, while it's hanging)?
2022-12-02 18:03:07 <eeijat> Another question, can I somehow uninstall unused firmware from a working system? Not by hand and one by one I mean
2022-12-02 18:06:38 <minimal> eeijat: are you talking about multiple linux-firmware-* packages?
2022-12-02 18:07:07 <minimal> if so then "apk add linux-firmware-none" and then you can remove whichever of the other linux-firmware-* packages you do not need
2022-12-02 18:16:28 <eeijat> minimal, thanks again I'll try it later
2022-12-02 19:44:44 <scorpion2185[m]1> ikke: what should I search?
2022-12-02 19:50:48 <scorpion2185[m]1> 
ACTION uploaded an image: (30KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/rFHBHGKQicLNKVcBRnssTQOh/2022-12-02_20-50-26.png >

2022-12-02 19:50:54 <scorpion2185[m]1> ikke: I found that ^
2022-12-02 19:53:37 <ikke> If you press `t`, or `f5`, you get a hierarchical view of the processes
2022-12-02 20:03:03 <scorpion2185[m]1> xargs  scanel --nobanner --etype ET_EXEC
2022-12-02 20:03:03 <scorpion2185[m]1> scanel --nobanner --etype ET_EXEC /path/smlock/usr/bin/smlock
2022-12-02 20:03:03 <scorpion2185[m]1> sed s|ET_EXEC ~/path/smlock/pkg/smlock|\t|
2022-12-02 20:27:17 <ikke> running scanelf, wondering what it's hanging on
2022-12-02 20:28:05 <ikke> can you install strace, and then try: strace -p <PID> -f -o scanelf.strace
2022-12-02 20:28:18 <ikke> maybe after 10 seconds or so cancel it
2022-12-02 20:30:05 <scorpion2185[m]1> cancel strace or abuild -r?
2022-12-02 20:30:41 <ikke> strace
2022-12-02 20:57:31 <scorpion2185[m]1> ikke:  `8419  semop(0, [{sem_num=0, sem_op=-1, sem_flg=SEM_UNDO}], 1 <detached ...>`
2022-12-02 21:08:02 <garo> Raspberry Pi 1 is that armhf or armv7 ?
2022-12-02 21:08:16 <garo> I am already assuming aarch64 won't work...
2022-12-02 21:13:18 <ikke> armhf
2022-12-02 21:13:48 <ikke> and no, aarch64 would not work
2022-12-02 21:20:15 <garo> armhf is with armv7 with extra floating point right ?
2022-12-02 21:21:27 <ikke> no, it's armv6
2022-12-02 21:26:27 <garo> so in short: aarch64 > armv7 > armhf and a pi1 only supports armhf
2022-12-02 21:34:19 <ikke> yes
2022-12-03 01:45:17 <pwr> hello, how do I disable the bold-is-bright in the console?
2022-12-03 01:45:31 <pwr> I think it's known as ANSI bold or something...
2022-12-03 01:45:57 <pwr> my aim is to have the regular text display as bright white so I can see it in daylight
2022-12-03 02:11:03 <pwr> (and by console I mean vt not terminal emulator in X)
2022-12-03 08:36:49 <uniqNuuid> plz plz plz enable amlogic on linux-lts aarch64, thank you 
2022-12-03 08:42:28 <psykose> you can open an issue for what you want enabled
2022-12-03 09:55:29 <vkrishn> re: ^ aarch64 > armv7 > armhf <-- and where does armv7l fit ?
2022-12-03 10:01:47 <psykose> the names don't mean anything
2022-12-03 10:02:52 <psykose> armv7l is just armv7, the l is 32-bit
2022-12-03 10:18:03 <vkrishn> ok, thanks
2022-12-03 13:04:45 <boombim> Hello. After some of updates screensharing stop works in my firefox on linux edge. Is it known issue?
2022-12-03 13:04:55 <boombim> or maybe something wrong with my setup...
2022-12-03 14:31:51 <cowboy_spike> hi all - have a host machine on 3.17.0.  trying to get octoprint working but i can't seem to see the usb connected printer via /dev/tty?  can anyone help?
2022-12-03 14:32:23 <ikke> cowboy_spike: do you see anything appearing in dmesg when you connect the printer?
2022-12-03 14:34:10 <cowboy_spike> ikke:  yes - can see the 3d printer
2022-12-03 14:35:09 <cowboy_spike> sorry not familiar with tty / serial / usb stuff
2022-12-03 14:37:21 <ikke> cowboy_spike: I'm not familiar with octoprint
2022-12-03 14:37:27 <ikke> What do you expect to happen exactly?
2022-12-03 14:37:41 <ikke> Usually a serial device is something like /dev/ttyS*
2022-12-03 14:38:21 <cowboy_spike> in setting up octoprint, it's looking to connect via /dev/tty
2022-12-03 14:38:37 <cowboy_spike> default is /dev/ttyASM0
2022-12-03 14:39:02 <ikke> cowboy_spike: in a terminal, can you show the output from `ls /dev/tty*`
2022-12-03 14:39:31 <cowboy_spike> yes
2022-12-03 14:41:15 <cowboy_spike> all the tty are not linked to anything
2022-12-03 14:42:56 <cowboy_spike> i got it working with a different machine (debian) using the default octoprint config using /dev/ttyASM0
2022-12-03 14:43:17 <cowboy_spike> my guess is i need to do something to get the usb connection to appear via /dev/tty ?
2022-12-03 14:44:33 <cowboy_spike> under alpine there is no /dev/ttyASM0
2022-12-03 14:45:20 <ikke> cowboy_spike: can you show the output that appears in dmesg on the debian system?
2022-12-03 14:46:01 <ikke> cowboy_spike: perhaps some kernel module, but we'd need to know which
2022-12-03 14:47:41 <cowboy_spike> [899078.369667] usb 2-1.6: new full-speed USB device number 3 using ehci-pci
2022-12-03 14:47:41 <cowboy_spike> [899078.469683] usb 2-1.6: Product: Original Prusa i3 MK3
2022-12-03 14:47:41 <cowboy_spike> [899078.469671] usb 2-1.6: New USB device found, idVendor=2c99, idProduct=0002, bcdDevice= 1.30
2022-12-03 14:47:41 <cowboy_spike> [899078.469679] usb 2-1.6: New USB device strings: Mfr=1, Product=2, SerialNumber=3
2022-12-03 14:47:41 <cowboy_spike> [899078.469685] usb 2-1.6: Manufacturer: Prusa Research (prusa3d.com)
2022-12-03 14:47:41 <cowboy_spike> [899078.469687] usb 2-1.6: SerialNumber: CZPX3322X004XK78960
2022-12-03 14:47:45 <cowboy_spike> sorry for the spam
2022-12-03 14:50:59 <ikke> does this work: ls /lib/modules/$(uname -r)
2022-12-03 14:54:13 <cowboy_spike> 5.15.80-0-lts
2022-12-03 14:55:26 <ikke> but does that directory exist?
2022-12-03 15:04:12 <cowboy_spike> yes
2022-12-03 15:04:45 <cowboy_spike> oh wait no, it doesn't
2022-12-03 15:05:11 <ikke> ok, it means you have to reboot
2022-12-03 15:05:20 <cowboy_spike> oh ok i see it now
2022-12-03 15:05:27 <cowboy_spike> it has a bunch of modules
2022-12-03 15:05:59 <ikke> So the output of uname -r is 5.15.80-0-lts, and /lib/modules/5.15.80-0-lts exists?
2022-12-03 15:06:09 <cowboy_spike> yes
2022-12-03 15:06:18 <cowboy_spike> it also a kernel dir
2022-12-03 15:06:35 <ikke> ok
2022-12-03 15:11:22 <ikke> cowboy_spike: can you run: modinfo ehci-pci?
2022-12-03 15:12:01 <cowboy_spike> modinfo: ERROR: Module ehci-pci not found.
2022-12-03 15:12:10 <cowboy_spike> guessing i need to add that module
2022-12-03 15:12:45 <ikke> cowboy_spike: ls -l /lib/modules/5.15.80-0-lts/kernel/drivers/usb/host/ehci-pci.ko.gz
2022-12-03 15:13:24 <cowboy_spike> it's there
2022-12-03 15:13:40 <ikke> ok, try modprobe ehci-pci as root
2022-12-03 15:14:23 <cowboy_spike> modprobe: FATAL: Module ehci-pci not found in directory /lib/modules/5.15.78-0-lts
2022-12-03 15:15:35 <ikke> cowboy_spike: can you please verify again the output of uname -0r
2022-12-03 15:15:39 <ikke> uname -r
2022-12-03 15:16:24 <ikke> You need to reboot
2022-12-03 15:16:30 <ikke> the installed kernel is newer then the running kernel
2022-12-03 15:17:10 <ikke> cowboy_spike: I suspect the output of uname -r is 5.15.78-0-lts, not 5.15.80-0-lts
2022-12-03 15:17:10 <cowboy_spike> rebooting now
2022-12-03 15:18:29 <cowboy_spike> modprobe: FATAL: Module echi-pci not found in directory /lib/modules/5.15.80-0-lts
2022-12-03 15:19:09 <ikke> ehci-pci
2022-12-03 15:19:38 <ikke> But I suspect it will automatically load it now if you plugin the printer
2022-12-03 15:19:51 <cowboy_spike> yeah nothing comes back
2022-12-03 15:20:00 <psykose> ecchi pci
2022-12-03 15:20:10 <psykose> cowboy_spike: that means it works
2022-12-03 15:20:16 <psykose> the modprobe, that is
2022-12-03 15:20:20 <psykose> but just replugging would be fine
2022-12-03 15:20:32 <ikke> cowboy_spike: can you verify that /dev/ttyASM0 exists now?
2022-12-03 15:20:55 <cowboy_spike> yes!
2022-12-03 15:20:58 <cowboy_spike> thanks all!
2022-12-03 15:21:17 <ikke> cowboy_spike: Details matter
2022-12-03 15:21:25 <cowboy_spike> yes thanks ikke
2022-12-03 16:15:40 <Ermine> I see ibm-plex-nerd fonts being prepared here: https://git.alpinelinux.org/aports/tree/community/nerd-fonts-stable/APKBUILD, but I don't see them at pkgs.a.o
2022-12-03 16:18:02 <ikke> Ermine: https://pkgs.alpinelinux.org/package/edge/community/x86_64/font-ibm-plex-mono-nerd
2022-12-03 16:18:45 <ikke> font-ibm-plex-nerd is provided by font-ibm-plex-mono-nerd
2022-12-03 16:18:59 <Ermine> ikke: ah, my bad, thank you
2022-12-03 19:39:15 <Saijin_Naib> If anyone's struggling with no auto-mount of NAS/SMB/NFS volumes at boot despite being in FSTAB (like I did for months), you just need rc-update add netmount boot. Got that one at random in my Local feed on Mastodon, so that was a gift from the Universe
2022-12-03 19:39:39 <Saijin_Naib> I thought I needed pam_mount and the associated rules and have not been able to figure out packaging that
2022-12-03 19:41:07 <blt[m]> I don't see an initd for boltctl, is there a bestpractice to start it automaticaly ? I realized that my usb port and other are not working if I am not doing a botcl list or any other command as it is starting /usr/libexec/boltctl, it is not a problem of authorize nor enroll but just that the service is not started, I was thinkg on adding an udev rule but there is maybe a better way?
2022-12-03 19:44:27 <minimal> Saijin_Naib: wouldn't pam_mount be for doing automounts upon user login, rather than on system boot?
2022-12-03 19:45:24 <Saijin_Naib> I think so, like USB drives and other GVFS stuff, right?
2022-12-03 19:45:48 <Saijin_Naib> I was fine with that, anything to not have to drop into a root shell before logging in, running mount -a, and then going back to normal login
2022-12-03 19:46:49 <minimal> blt[m]: the bolt package does install a udev rule, 90-bolt.rules
2022-12-03 19:48:16 <minimal> Saijin_Naib: haven't looked at it, was just pointing out that I'd not expect to use login-time functionality to mount at boot time
2022-12-03 19:53:02 <blt[m]> minimal: thanks, in fact I see it now (I was looking to /etc/udev/rules.d and not in /lib/udev/rules.d, so it is strange that it is not working... I am checking the rule/name to understand the reason
2022-12-03 19:55:35 <minimal> blt[m]: the rule appears to be written with systemd in mind
2022-12-03 19:56:06 <blt[m]> yes I saw the TAG as well...
2022-12-03 19:56:28 <minimal> blt[m]: not just the tag, it refers to a systemd bolt.service
2022-12-03 19:56:41 <minimal> which obviously doesn't exist on Alpine
2022-12-03 19:57:09 <blt[m]> I see so that is the issue
2022-12-03 19:59:05 <minimal> blkt[m]: also /usr/share/dbus-1/system-services/org.freedesktop.bolt.service refers to systemd bolt.service
2022-12-03 20:03:05 <Ermine> So, I have an Alpine system with grub and os-prober installed, but it doesn't want to detect other systems.
2022-12-03 20:08:08 <blt[m]> minimal: I made a new rule SUBSYSTEM=="thunderbolt", RUN+="/usr/libexec/boltd" which is working but am not an udev expert so it is perhaps a dirty way
2022-12-03 20:12:13 <minimal> Ermine: is grub-mkconfig running os-prober ok?
2022-12-03 20:20:25 <minimal> Ermine: in /etc/default/grub did you set GRUB_DISABLE_OS_PROBER=false ?
2022-12-03 20:26:42 <Ermine> minimal: yes, I've added this line. And there are log lines from os-prober in /var/log/messages.
2022-12-03 20:27:11 <minimal> Ermine: what happens if you run "os-prober" manually? does it discover the other OSes?
2022-12-03 20:28:17 <minimal> you should get output of 1 line per OS
2022-12-03 20:38:13 <Ermine> minimal: it discovers Arch Linux (but there is another Alpine and Adélie on the disc)
2022-12-03 20:39:21 <minimal> Ermine: so it is not a Grub issue, it is a os-prober issue
2022-12-03 20:40:08 <minimal> I suspect it might not find either of those due to the "non standard" naming of kernel and/or initramfs files
2022-12-03 20:41:19 <Ermine> Arch's os-prober works and finds everything.
2022-12-03 20:41:55 <minimal> Ermine: that proves nothing, it might have different logic for searching
2022-12-03 20:43:49 <minimal> Ermine: e.g. the 90fallback file has rules for various forms of initramfs naming
2022-12-03 20:44:07 <minimal> perhaps the Arch and Alpine versions of that file have differing rules
2022-12-03 20:48:04 <minimal> Ermine: os-prober is a shell script, so run with "-x" to debug it?
2022-12-03 20:50:02 <Ermine> minimal: tried diff on /usr/lib/os-prober in Alpine and Arch, no difference except newns binary.
2022-12-03 20:53:47 <Ermine> minimal: here's the log: https://pastesite.org/view/b3585c5b
2022-12-03 20:57:54 <unrznbl[m]> Any reports of wifi  working worse/crash a lot in edge versus 3.16? I use iwl driver on an old Lenovo laptop (x220 tablet)
2022-12-03 20:58:10 <ikke> unrznbl[m]: nothing reported so far
2022-12-03 21:00:12 <unrznbl[m]> Cool. I was getting lots of noise in dmesg/syslog and Wi-Fi was definitely flaky on and off. I’ll try and reinstall 3.16 maybe :) thanks!
2022-12-03 21:01:59 <minimal> Ermine: no idea, I don't have any Alpine machines with other OSes/distros on them to test
2022-12-03 21:02:20 <minimal> but when I run it in a Alpine-only system it (obviously) gives no results
2022-12-03 21:03:58 <minimal> Ermine: try modifying the 1st line of /usr/bin/os-prober to append " +x" and run it, you'll get more output (will need to do " 2> logfile")
2022-12-03 21:07:51 <Ermine> minimal: ok, will do it tomorrow.
2022-12-04 06:03:17 <valerius> uhhh... did the plugdev group not exist in Alpine before and now it needs to?
2022-12-04 08:20:10 <psykose> existed for a while
2022-12-04 08:20:25 <psykose> not that many things use it
2022-12-04 12:05:06 <Duck[m]> why are some dependencies of inkscape left out?
2022-12-04 12:05:06 <Duck[m]> specifically the ones for the extraction as an image
2022-12-04 12:05:07 <Duck[m]> It depends on some python packages
2022-12-04 13:46:00 <Saklad5> Does anyone have any experience with running dual-stack IP?
2022-12-04 13:46:58 <ikke> If mean with having both ipv4 and ipv6 present, sure
2022-12-04 13:47:01 <ikke> if you mean*
2022-12-04 13:47:30 <Saklad5> I'm trying to get Alpine Linux working in an IPv6-only Vultr server, and while I don't have a public IPv4 address, I should be able to access other IPv4 addresses. But I can't.
2022-12-04 13:47:43 <Saklad5> And I'm wondering if anyone has any idea how that would happen
2022-12-04 13:48:07 <Saklad5> I can connect to the IPv6 Internet just fine
2022-12-04 13:48:15 <ikke> Does vultr use NAT64
2022-12-04 13:48:17 <ikke> ?
2022-12-04 13:48:42 <psykose> don't think so
2022-12-04 13:48:51 <Saklad5> Yeah, I'm pretty sure it doesn't
2022-12-04 13:48:52 <psykose> you have to set up your own things to reach v4
2022-12-04 13:48:59 <psykose> can't reach any v4 addresses otherwise
2022-12-04 13:49:14 <Saklad5> Oh, so that's normal?
2022-12-04 13:49:17 <psykose> yeah
2022-12-04 13:49:35 <psykose> v4 and v6 are completely distinct and don't imply anything to eachother
2022-12-04 13:49:46 <psykose> plenty of ways to work with allowing one to reach the other but not by default
2022-12-04 13:49:52 <Saklad5> I thought I just had to share an IPv4 address with a bunch of other servers, like a personal computer behind your average home router
2022-12-04 13:50:04 <ikke> that implies nat64
2022-12-04 13:50:04 <psykose> that would imply you have a v4 address, but you don't
2022-12-04 13:50:14 <Saklad5> It says I have one in Vultr settings
2022-12-04 13:50:19 <Saklad5> 100.68.85.140
2022-12-04 13:50:31 <ikke> and did you add that to your VPS?
2022-12-04 13:50:39 <psykose> that would mean it's not v6-only :p
2022-12-04 13:50:40 <ikke> That sounds like cgnat
2022-12-04 13:50:49 <psykose> you have to somehow get that to your interface configuration
2022-12-04 13:51:05 <Saklad5> Perfect, this sounds like something I can actually work on solving
2022-12-04 13:51:19 <psykose> 100.64-127 is cgnat yeah
2022-12-04 13:51:44 <psykose> i guess what they mean by v6only is only v6 publically routable
2022-12-04 13:51:55 <Saklad5> That's what I expected
2022-12-04 13:52:19 <psykose> i would except the usual dhcp or whatever to work and give that as a v4 address for you, so if not you have to look around there
2022-12-04 13:52:23 <psykose> expect*
2022-12-04 13:53:18 <Saklad5> I have DHCP set up for both IPv4 and IPv6, though I've also tried configuring it statically based on the Debian configuration Vultr provides
2022-12-04 13:53:47 <Saklad5> Thanks for the help, I just wanted to make sure this wasn't an Alpine thing before bugging Vultr about an OS they don't officially support
2022-12-04 13:54:51 <psykose> if `ip a` shows that 100.x address on the iface you should be able to reach v4
2022-12-04 13:55:04 <ikke> You also need a default route
2022-12-04 13:55:11 <Saklad5> inet 100.68.85.140/18 brd 100.68.127.255 scope global eth0
2022-12-04 13:55:23 <ikke> ip route show default
2022-12-04 13:55:30 <psykose> ah yes
2022-12-04 13:55:46 <Saklad5> baza:~# ip route show default100.68.64.0/18 dev eth0 scope link  src 100.68.85.140
2022-12-04 13:56:01 <Saklad5> OK, so I forgot you can't do multiline IRC messages, but you get the idea
2022-12-04 13:56:05 <psykose> now try something like `ping -4 google.com` and ping -6 to compare
2022-12-04 13:56:15 <ikke> that's not a default route
2022-12-04 13:56:16 <Saklad5> IPv6 works, IPv4 doesn't
2022-12-04 13:56:30 <Saklad5> When I type "ip route show default", that's the result
2022-12-04 13:56:39 <ikke> it should show 'default via <gateway> dev <iface> ..'
2022-12-04 13:57:01 <Saklad5> What's the IPv6 equivalent?
2022-12-04 13:57:04 <ikke> `ip route add default via <gateway>`
2022-12-04 13:57:15 <ikke> ip -6 route show default
2022-12-04 13:57:40 <Saklad5> Got it
2022-12-04 13:57:51 <Saklad5> So, how do I designate a default route?
2022-12-04 13:57:56 <ikke> ^^
2022-12-04 13:58:10 <Saklad5> Wait, does that persist?
2022-12-04 13:58:21 <ikke> no
2022-12-04 13:58:38 <ikke> typically you add that to /etc/network/interfaces
2022-12-04 14:11:07 <apkcacher> hello there everyone
2022-12-04 14:11:47 <apkcacher> i wanted to ask if anyone around here knows what the hex characters represent in /var/cache/apk/APKINDEX.ac15ed62.tar.gz?
2022-12-04 14:12:13 <apkcacher> I have these APKINDEX.*.tar.gz files on different systems, where the hex characters are different, but the file sha256 is the same
2022-12-04 14:12:31 <apkcacher> any idea what those hex characters refer to? thanks a bunch folks!
2022-12-04 14:24:53 <ikke> apkcacher: I'm not entirely sure myself
2022-12-04 14:25:09 <ikke> trying to find it in the source, but cannot find exactly where it is done
2022-12-04 14:26:50 <ikke> oh, just found it
2022-12-04 14:26:52 <ikke> https://gitlab.alpinelinux.org/alpine/apk-tools/-/blob/master/src/database.c#L578
2022-12-04 14:27:25 <ikke> So it is a checksum
2022-12-04 14:28:48 <ikke> apkcacher: but no idea why it might be different even when the index files are equal
2022-12-04 14:35:52 <apkcacher> ikke: that is really strange. let me upload two screenshots, one sec
2022-12-04 14:45:08 <apkcacher> ikke: this is from a docker container after running apk update: https://ibb.co/XW6W3Bs
2022-12-04 14:45:36 <apkcacher> ikke: this is from a live system booted from an usb stick after running apk update: https://ibb.co/Y0bt2mZ
2022-12-04 14:46:35 <apkcacher> ikke: both are alpine 3.17. the hex characters in filename differ, md5sum is the same...
2022-12-04 15:22:02 <vkrishn> apkcacher: I noticed that too, how many diffrent mirror links are there in repositories file ?
2022-12-04 15:27:19 <apkcacher> vkrishn: Just the main and community repos for x86_64 (default mirror)
2022-12-04 15:29:35 <vkrishn> does changing mirror + apk update add another APKINDEX.. file ?
2022-12-04 15:31:32 <apkcacher> vkrishn: yeah maybe the repo endpoint is mixed into the checksum. let me check that quick
2022-12-04 15:35:33 <vkrishn> I can think of one reason for this behaviour, sometimes mirrors are not in sync, so this is needed to track what went wrong
2022-12-04 15:40:26 <vkrishn> but, fun part is, now you have three of them
2022-12-04 15:47:11 <apkcacher> vkrishn: I added the yandex mirror and got another APKINDEX.$hex.tar.gz. Same sha256. Here is a pic of my repos file and the APKINDEX files on my system: https://ibb.co/QJv1KLv
2022-12-04 15:47:44 <panekj> it's not sha256 of the file itself
2022-12-04 15:48:22 <apkcacher> panekj: yeah I figured that. otherwise of course that part would be identical too like the sha256 checksum is.
2022-12-04 15:48:38 <apkcacher> most likely the repo endpoint is somehow incorporated into that hash input
2022-12-04 16:01:03 <unrznbl[m]> I tried 3.17 with my thinkpad x220 and it didn’t find my wifi access points. I was having unstable iwlwifi with edge, so far 3.16 is stable and finds my APs just fine. So workaround is good just maybe back of mind that there might be a regression somewhere there. :) thanks for awesome alpine! Love using it day to day on this laptop and all my phones (well, except my iPhone:()
2022-12-04 16:17:13 <donoban> hi, is there some dropbear or similar alternative for unlock LUKS remotely officially supported?
2022-12-04 16:19:11 <minimal> donoban: nope, I raised a MR quite some time ago to add this to mkinitfs but it was never merged
2022-12-04 16:20:14 <minimal> quite some time == more than 1 year ago
2022-12-04 16:20:42 <donoban> oh, do you think it stills working?
2022-12-04 16:21:46 <minimal> donoban: the patches would probably need some minor changes, however I would not recomend using it - if you do then any update of the mkinitfs or dropbear packages would "remove" the changes
2022-12-04 16:22:47 <donoban> well, I need to find some solution
2022-12-04 16:23:18 <minimal> there is no solution in Alpine presently AFAIK
2022-12-04 16:24:07 <minimal> you could perhaps look at switching from mkinitfs to dracut and use hooks for dracut, but dracut is in testing - i.e. only available on Alpine Edge
2022-12-04 16:25:35 <donoban> I will try to look at your MR first, there was no expectation for merging it?
2022-12-04 16:26:38 <donoban> is https://gitlab.alpinelinux.org/alpine/mkinitfs/-/merge_requests/86 ?
2022-12-04 16:26:41 <minimal> the issue was around lack of a testsuite for mkinitfsm and the risk of changes breaking it
2022-12-04 16:27:27 <minimal> yes that's the MR, plus there is another change to Dropbear to go along with it (no MR raised for that as I was waiting for the mkinitfs one to be merged first)
2022-12-04 16:28:16 <minimal> donoban: have a look at https://github.com/dermotbradley/alpine-remote-LUKS-unlock
2022-12-04 16:28:58 <minimal> but as I say, once you start making local changes to mkinitfs then how will you cope with mkinitfs package updates?
2022-12-04 16:31:27 <donoban> hmm.., looking at the git history, probably once it works the majority of changes won't affect my setup, so I could stick with the working version
2022-12-04 16:33:33 <minimal> donoban: then you'd miss any fixes/exhancements that would appear in newer mkinitfs versions, plus miss any dropbear fixes or version updates
2022-12-04 16:36:27 <donoban> yes, I just feel that is not too frequently
2022-12-04 16:37:03 <donoban> maybe I can to setup a cryptkey initfs, reboot and remove the initfs containing the key
2022-12-04 16:37:32 <minimal> donoban: your choice but I would not rely on such a feature unless it was part of the official Alpine packages.
2022-12-04 16:37:48 <donoban> it has it's downsides but it seems more simple to achieve
2022-12-04 16:37:55 <minimal> how does that permit remote unlocking?
2022-12-04 16:38:16 <donoban> storing the key on the unecrypted boot partition
2022-12-04 16:38:22 <donoban> just before doing the reboot
2022-12-04 16:39:32 <minimal> I thought you specifically wanted remote unlocking
2022-12-04 16:41:28 <elly> that seems like a very bad idea? once the key hits disk in plain reliably removing it is very hard
2022-12-04 16:41:41 <elly> seems like it will undermine the utility of FDE
2022-12-04 16:41:42 <donoban> yeah, I log remotely via ssh, create the initramfs with the keys, reboot, crossfingers, log again and remove the initramfs
2022-12-04 16:42:06 <minimal> elly: well it is *not* FDE if boot is unencrypted...
2022-12-04 16:42:07 <donoban> yes, that it's the most important downside
2022-12-04 16:42:59 <elly> minimal: somehow, somewhere, there must be an unencrypted thing to load and run to do the decryption, so I think that's still in the spirit of FDE
2022-12-04 16:43:36 <minimal> donoban: how will you log in remotely unless you modify the initramfs to run dropbear? and if you modify it to do so then why bother with copying key when you can just prompt for passphrase (which is what my MR does)
2022-12-04 16:43:36 <donoban> uhM, luks slots are not reusable?
2022-12-04 16:43:57 <donoban> minimal: I mean logging into an already booted system
2022-12-04 16:44:20 <donoban> wich I booted with the keyboard
2022-12-04 16:44:47 <minimal> elly: it is PDE, partial disk encryption, unless you use Grub's FDE support for an encrypted fs holding /boot. Even then if the machine is UEFI-based it is still PE as the ESP partition has to be unencrypted
2022-12-04 16:45:00 <elly> I see
2022-12-04 16:45:21 <elly> but the grub bootloader itself is still unencrypted on disk?
2022-12-04 16:45:44 <minimal> so a BIOS-based system (assuming 1 disk) could be FDE using Grub, a UEFI system would be PDE
2022-12-04 16:45:55 <grawity> yeah, grub still remains unencrypted no matter what
2022-12-04 16:46:04 <grawity> whether it's in the bios mbr or in the uefi partition
2022-12-04 16:46:12 <elly> that still sounds partial under your definition
2022-12-04 16:46:12 <grawity> but it's still useful against "steal disks and never return them" attacks
2022-12-04 16:46:14 <raven523> at least the core image needs to be unencrypted so the UEFI/BIOS can load it
2022-12-04 16:46:20 <psykose> these distinctions are meaningless
2022-12-04 16:46:24 <elly> indeed
2022-12-04 16:46:40 <raven523> UEFI provides Secure Boot if you need to ensure that GRUB won't boot if it was tampered with
2022-12-04 16:46:46 <grawity> to avoid tampering, involve a TPM or something (either with secure boot, or without)
2022-12-04 16:46:54 <elly> what really matters is whether your private data is encrypted when the machine is off, the rest is semantics :P
2022-12-04 16:47:35 <minimal> grawity: in a BIOS system Grub's in the MBR, whereas with UEFI it's in the ESP partition/filesystem. So I was talking in terms of filesystems :-)
2022-12-04 16:47:46 <grawity> sure, sure
2022-12-04 16:47:53 <minimal> yes I'm backpedalling ;-)
2022-12-04 16:52:41 <minimal> the big difference is if you are using a key stored in the initramfs to automatically unlock other filesystems - if the filesystem your /boot is on is unencrypted then someone with physical access to the machine when it's powered down could get hold of the key, whereas if /boot is part of an encrypted filesystem it prevents that. Obviously secure boot should then be used if available to detect tampering of the bootloader
2022-12-04 16:54:08 <grawity> right, but the initramfs doesn't have to be on the EFI System Partition -- that's a separate thing from /boot
2022-12-04 16:55:36 <minimal> grawity: yes it doesn't have to be (though currently setup-disk for UEFI with an encrypted rootfs does place /boot on the ESP partition)
2022-12-04 16:56:36 <minimal> frawity: plus support for encrypted /boot is only in Grub and only if its "FDE" support is activated and if LUKSv1 is used
2022-12-04 17:31:33 <minimal> Saklad5: are you using cloud-init on Alpine with Vultr?
2022-12-04 17:40:46 <bananicorn> Hello! I have a question regarding diskless installation, which I am currently trying to set up with the generic ARM image on an SBC
2022-12-04 17:41:19 <bananicorn> At the end of the setup-alpine script (basically just picking default options) I get a warning
2022-12-04 17:44:34 <bananicorn> "WARNING: Ignoring http://dl-cdn.alpinelinux.org/alpine/v3.17/main: No such file or directory"
2022-12-04 17:45:06 <bananicorn> Is this intentional? As far as I can see, it comes from the "apk cache sync" at the end of the script
2022-12-04 17:46:35 <bananicorn> This warning does not happen if I run "apk update" and re-run "apk cache sync"
2022-12-04 17:59:52 <ikke> I don't think that's normal
2022-12-04 18:14:22 <bananicorn> I just noticed that while going through setup-apkrepos, it gives me this error:
2022-12-04 18:14:34 <bananicorn> "Updating repository indexes... ERROR: http://ftp.halifax.rwth-aachen.de/alpine/v3.17/main: temporary error (try again later)"
2022-12-04 18:14:56 <bananicorn> Any idea what may have caused it? It successfully got a dhcp lease
2022-12-04 18:15:29 <uuidNuniq> bananicorn: dns?
2022-12-04 18:15:30 <ikke> dns issues?
2022-12-04 18:15:40 <uuidNuniq> bananicorn: did you disable ipv6?
2022-12-04 18:15:49 <minimal> bananicorn: isn't it obvious? the "try again later" message indicates that particular server was not reachable then
2022-12-04 18:16:18 <psykose> no, it's not obvious that a "try again later" means anything to do with dns for people that have not read every errno in the book :-)
2022-12-04 18:16:25 <bananicorn> minimal: The error is obvious, the cause is not :)
2022-12-04 18:16:30 <grawity> yeah but there are like 10 things that might have caused that unreachability
2022-12-04 18:16:39 <bananicorn> As I mentioned, running apk update afterwards worked fine
2022-12-04 18:17:08 <ikke> bananicorn: it sounds like you had some temporary network issue (or network was not fully working yet)
2022-12-04 18:17:34 <bananicorn> I vaguely remember having an issue like this with the board before
2022-12-04 18:17:54 <bananicorn> It would be strange to have this repeatedly occur if it was random, I tried over multiple times
2022-12-04 18:18:05 <uuidNuniq> i had such issues before, it was due to my local dns server disabled ipv6 AAAA reply 
2022-12-04 18:18:17 <uuidNuniq> and musl it not happy with that 
2022-12-04 18:18:30 <grawity> disabled as in it started replying with "no such domain"?
2022-12-04 18:19:00 <uuidNuniq> i have dnsproxy set --no-ipv6 
2022-12-04 18:20:00 <uuidNuniq> my local dns server is dnsmasq + dnsproxy(provides DOH) and i set them --ipv6-disabled  
2022-12-04 18:20:14 <uuidNuniq> remove that, then everything works again 
2022-12-04 18:21:23 <bananicorn> I will try setting a different dns server during setup
2022-12-04 18:22:19 <uuidNuniq> musl seems to be very strict on dns reply.
2022-12-04 18:23:18 <uuidNuniq> just checked my note, the error was : bad address mirrors.alpinelinux.org
2022-12-04 18:24:49 <uuidNuniq> ^^ just my piece of share. i am not sure your issue was dns related or not. 
2022-12-04 18:26:29 <psykose> how do you disable ipv6 in dnsmasq
2022-12-04 18:26:46 <uuidNuniq> filter-aaaa something like that
2022-12-04 18:27:17 <uuidNuniq> --filter-AAAA
2022-12-04 18:27:25 <uuidNuniq> Remove AAAA records from answers. No IPv6 addresses will be returned.
2022-12-04 18:27:43 <uuidNuniq> i think it was added since 2.86?
2022-12-04 18:28:05 <psykose> just looked at dnsproxy
2022-12-04 18:28:07 <psykose> > --ipv6-disabled     If specified, all AAAA requests will be replied with NoError RCode and empty answer
2022-12-04 18:28:09 <psykose> well
2022-12-04 18:28:18 <psykose> that is beyond completely wrong so i am not surprised
2022-12-04 18:28:23 <psykose> jesus christ
2022-12-04 18:29:00 <grawity> well, at least that's better than faking a NxDomain rcode
2022-12-04 18:29:08 <grawity> and I've seen some dns resolvers do that
2022-12-04 18:29:16 <uuidNuniq> i don't know the underlying details. but yeah, i removed them. 
2022-12-04 18:29:21 <psykose> hah, sure
2022-12-04 18:29:59 <grawity> NoError + empty answer is what you normally get when the domain legitimately has no v6 addresses, so ... it's ugly in principle but I can't see it causing problems
2022-12-04 18:30:14 <grawity> (why *do* people use filter-ipv6 though?)
2022-12-04 18:30:43 <Saklad5> I hope, someday, that IPv4 dies in a fire
2022-12-04 18:31:41 <uuidNuniq> grawity: i thought disable ipv6 will prevent AAAA query to save time, but i was wrong. the query will be sent anyways 
2022-12-04 18:31:50 <bananicorn> The behaviour I am seeing is quite bizzare, I don't think it's DNS because I can ping random IPs just fine, over both v4 and v6
2022-12-04 18:32:11 <Saklad5> Please, for the love of God, never turn off IPv6
2022-12-04 18:32:12 <bananicorn> Is there anything hardware or driver related that could be causing this?
2022-12-04 18:32:13 <uuidNuniq> bananicorn: ping ip has nothing to do with DNS
2022-12-04 18:32:20 <uuidNuniq> you should ping domain 
2022-12-04 18:32:23 <bananicorn> sorry I mean domain
2022-12-04 18:32:26 <bananicorn> not IPs
2022-12-04 18:34:23 <uuidNuniq> bananicorn: can you reproduce the issue everytime?
2022-12-04 18:34:45 <bananicorn> Unfortunately it seems like yes, but I'll give it another round for good measure
2022-12-04 18:35:31 <uuidNuniq> bananicorn: you may want to turn on verbose or debug message on your dns servers if there is any 
2022-12-04 18:35:55 <uuidNuniq> and watch them while reproducing the issue
2022-12-04 18:36:50 <uuidNuniq> or you can force /etc/resolv.conf to use "nameserver 1.1.1.1" and try again 
2022-12-04 18:37:05 <Ermine> minimal: here's the log of os-prober: http://erminea.space/paste/os-prober.log
2022-12-04 18:37:06 <bananicorn> I will do the latter, I don't have any special dns setup
2022-12-04 18:40:17 <bananicorn> Happened again, even with the google nameserver
2022-12-04 18:41:41 <bananicorn> Is there a hardware or driver issue that could cause the first request to fail?
2022-12-04 18:41:50 <uuidNuniq> bananicorn: which mirror? the official one? 
2022-12-04 18:42:12 <bananicorn> any mirror, tried the official one and the rwth aachen mirror
2022-12-04 18:42:36 <uuidNuniq> try https?
2022-12-04 18:43:21 <bananicorn> will try :)
2022-12-04 18:58:17 <bananicorn> Okay that just worked for some reason 🤷
2022-12-04 18:58:23 <bananicorn> I'll try a few more times, wish me luck
2022-12-04 18:59:20 <uuidNuniq> well, if https works fine, then your ISP is doing something weird. 
2022-12-04 19:00:51 <uuidNuniq> the world is full of dns hijacking and plain http violation. be careful 
2022-12-04 19:03:12 <bananicorn> I remember having this issue with this particular board several years ago, in a different state in Germany with a different ISP
2022-12-04 19:03:29 <bananicorn> I don't know how https solved it, but I'll take it
2022-12-04 19:06:07 <bananicorn> Thanks for the idea :)
2022-12-04 19:11:34 <minimal> Ermine: are you running with eudev? do you have the "full" version (rather than Busybox one) of blkid installed?
2022-12-04 19:12:10 <Ermine> eudev - yes, blkid - not sure, most probably it is busybox version.
2022-12-04 19:12:48 <uuidNuniq> Ermine: realpath $(command -v blkid)
2022-12-04 19:16:57 <minimal> Ermine: try installing blkid package and see if that makes any difference
2022-12-04 19:17:33 <minimal> Ermine: I see common.sh using "blkid -o value -s TYPE" which busybox's blkid will not handle
2022-12-04 19:27:18 <Ermine> minimal: it didn't help.
2022-12-04 19:27:43 <callum1> Hi, I've got a problem with udev and sound devices. I use sway and recently re-read the wiki page (had some other problem which is now fine... I hope). I did `setup-devd udev` as instructed but now /dev/snd/* is owned by root:root not root:audio and the permissions are 600 not 660 so non-root users can't access sound devices even if they're part of the audio group. Changing the
2022-12-04 19:27:45 <callum1> permissions manually works but isn't saved. Changing (back?) to mdev makes the permissions right, but then sway and udev needing stuff doesn't work. I can't see anything obviously wrong in udev rules (don't think I've changed anything). Any ideas?
2022-12-04 19:28:32 <minimal> Ermine: II think it helps to have proper blkid installed but suspect there's other issues as well
2022-12-04 19:29:26 <minimal> Ermine: which devices/partitions do you have other Linux distros on?
2022-12-04 19:30:37 <minimal> Ermine: and what filesystem type(s) are you using? e.g. I don't think it handles ZFS
2022-12-04 19:34:31 <Ermine> sda1 is ESP, sda2 is Adélie Linux, sda3 is Alpine Linux, sda4 is another Alpine Linux from which I manage bootloader, sda5 is Arch Linux. Every distro uses ext4 file system.
2022-12-04 19:53:20 <minimal> Ermine: do you have the grub-mount package installed? if not then try installing that and run the test again
2022-12-04 19:59:54 <Ermine> minimal: it worked, thank you!
2022-12-04 20:03:59 <minimal> so grub-mount and blkid are missing dependencies for os-prober
2022-12-04 20:04:52 <minimal> basically it seems to use grub-mount (which uses fuse) to mount unmounted partitions to check them - without grub-mount installed then obviously it cannot discover other OSes on partitions it cannot mount
2022-12-04 20:07:12 <minimal> surprised no-one's noticed this before now as it's been packaged for 2+ years
2022-12-04 20:08:20 <ikke> I do regularly see messages about os-prober not working for whatever reason, but I never cared about it because I have no dual-boot systems
2022-12-04 20:09:49 <minimal> ikke: me neither, so Ermine: will you raise an MR to add the missing deps to os-prober?
2022-12-04 20:12:30 <minimal> ah, indeed there's an issue from 1 year ago that mentioned the missing grub-mount dep: #12956
2022-12-04 20:21:15 <Ermine> minimal: yes, I'll make it.
2022-12-05 04:04:22 <obsed_> Does Alpine Linux include GRSec/PaX kernel hardening btw ?
2022-12-05 05:04:55 <Ermine> obsed_: no.
2022-12-05 07:44:24 <jh81> so, can it?
2022-12-05 08:14:51 <ribalgz[m]> Since recent updates, sound isn't working anymore using pipewire
2022-12-05 08:15:55 <ribalgz[m]> pipewire-launcher outputs this code from pipewire-pulse, 536870912, and Pavucrontol only shows a dummy device
2022-12-05 08:16:35 <ribalgz[m]> `aplay -l` also outputs "no soundcards found..."
2022-12-05 08:20:18 <bananicorn> What's pipewire-launcher?
2022-12-05 08:20:58 <bananicorn> Nvm
2022-12-05 08:22:04 <bananicorn> Do you use wireplumber or pipewire-media-session?
2022-12-05 08:22:47 <bananicorn> I am running edge on latest updates and pipewire is working fine for me
2022-12-05 08:23:02 <bananicorn> With wireplumber
2022-12-05 08:29:12 <ribalgz[m]> I've tried both
2022-12-05 08:29:55 <ribalgz[m]> pipewire-launcher is the alpine script for managing pipewire
2022-12-05 08:32:15 <ribalgz[m]> Well, aplay -l is working for root, but my user is in the audio group
2022-12-05 08:33:47 <bananicorn> Which audio-related packages do you have installed? Could there be a conflict?
2022-12-05 08:38:26 <ribalgz[m]> alsa stuff and pipewire + pipewire-pulse
2022-12-05 08:39:10 <bananicorn> The alsa stuff might be the issue, I think I had problems with pipewire when I still had alsa installed
2022-12-05 08:39:47 <ribalgz[m]> Well, it wasn't recognizing the audio group because somehow elogind was installed
2022-12-05 08:39:58 <ribalgz[m]> Removed it and is working again
2022-12-05 08:40:05 <bananicorn> Alright, glad to be wrong^^
2022-12-05 08:40:22 <ribalgz[m]> Thank you bananicorn anyways :D
2022-12-05 08:40:26 <bananicorn> Np :P
2022-12-05 09:44:38 <michalkotyla> Hey, maybe someone has experience with integrating RAUC update system into Alpine Linux? I see it is here package for that but is this all that I need? what about A/B partitioning and SW/HW watchdog for error detection after an update? Do I need to configure this manually?
2022-12-05 11:58:06 <Ermine> Is it expected that apk info --contents works only for installed packages?
2022-12-05 12:00:31 <ikke> Yes
2022-12-05 17:31:42 <jh81> oh.. yeah.. audio.. back to my own audio problems, brb
2022-12-05 17:34:45 <jh81> so, alsa is out, pipewire is in?
2022-12-05 17:41:23 <jh81> nice! got sound!
2022-12-05 17:41:37 <jh81> now, Alpine can be my daily driver for my desktops <3
2022-12-05 17:46:03 <Ermine> Can I force apk to not install package dependencies?
2022-12-05 17:46:29 <ikke> no
2022-12-05 17:49:42 <jh81> do flatpak programs? access the local .config dir? or do i need to modify that?
2022-12-05 18:00:36 <jh81> hmz, audio seems a bit of a hit and miss thingy
2022-12-05 18:02:25 <Ermine> What's wrong with your audio?
2022-12-05 18:02:50 <jh81> it depends :) now it works.. but when i reboot, it may not find the audio device
2022-12-05 18:02:58 <jh81> and when i reboot then, suddenly it works again :)
2022-12-05 18:03:24 <Ermine> Do you have udev installed and how do you launch pipewire?
2022-12-05 18:03:46 <jh81> oh! thank you :) udev was not installed
2022-12-05 18:04:06 <jh81> and piwire launches through the pipewire-session package?
2022-12-05 18:04:21 <jh81> pipewire-media-session :)
2022-12-05 18:04:29 <jh81> i followed the wiki on that
2022-12-05 18:04:44 <Ermine> Use wireplumber. I've written this in bold.
2022-12-05 18:05:05 <jh81> yes, thats installed too
2022-12-05 18:07:41 <jh81> oh, i dont have the pipewire-media-session :|
2022-12-05 18:07:46 <jh81> been messing so much with it..
2022-12-05 18:08:01 <Ermine> You don't need pipewire-media-session.
2022-12-05 18:09:13 <jh81> bno, i read that :)
2022-12-05 18:09:36 <Ermine> 
ACTION waves hands like Obi-Man.

2022-12-05 18:10:08 <jh81> 
ACTION doesnt know Obi :) thats Star Wars, not my thing

2022-12-05 18:10:27 <lopid> obiwan picard, from stargate
2022-12-05 18:10:54 <jh81> :D
2022-12-05 18:11:04 <Habbie> somehow, obiwan returned
2022-12-05 18:11:04 <jh81> that would be obiwan teal'c
2022-12-05 18:11:15 <jh81> obiwan picard is from star trek
2022-12-05 18:11:27 <ikke> jh81: it's a joke
2022-12-05 18:11:37 <jh81> hm?
2022-12-05 18:11:58 <lopid> indeed
2022-12-05 18:12:02 <ikke> jh81: conflating 3 series, provoking all of the fans
2022-12-05 18:12:27 <jh81> humor isnt something i am good in, as you see ;-)
2022-12-05 18:13:39 <Ermine> Well, anyway, don't use pipewire-media-session and use wireplumber.
2022-12-05 18:13:50 <jh81> got it :)
2022-12-05 18:16:31 <Ermine> jh81: for reference, https://www.youtube.com/watch?v=k1rlThKe1qo
2022-12-05 18:17:44 <minimal> Jeff Vader? ;-) Eddie Izzard reference that I guess many people won't get
2022-12-05 18:17:57 <minimal> "Death Star Canteen"
2022-12-05 18:18:11 <jh81> if only i got what i want that easily :)
2022-12-05 18:24:55 <Ermine> Anyway, now my problems: os-prober doesn't seem to generate other Alpine's configuration correctly.
2022-12-05 18:27:15 <Ermine> 1) It does not append necessary kernel parameters (modules= and rootfstype=); 2) It does not remove '^' between initrd files (so config has e.g. 'initrd /boot/intel-ucode.img^/boot/initramfs-lts)
2022-12-05 18:27:42 <Ermine> As for the second part, I've seen fixes upstream.
2022-12-05 18:30:24 <minimal> Ermine: (1) would be a case of grub's template, in /etc/grub.d/, for other OS's not "knowing" to do this I assume
2022-12-05 18:49:11 <jh81> kill 7847
2022-12-05 18:52:15 <psykose> unit 7847 destroyed.
2022-12-05 19:07:23 <jh81> yeah sorry, i failed to see which window had focus
2022-12-05 19:08:01 <Ermine> psykose: *explosion sounds and effects*
2022-12-05 19:08:14 <psykose> :)
2022-12-05 19:24:02 <feuerteufel> Hello @ all
2022-12-05 19:25:36 <feuerteufel> I'm Using a edge installation on one of my VM, there I'm not able to start crond.
2022-12-05 19:27:06 <feuerteufel> The reason is, the machine say's, syslod is not runing but it is > 1503 root      0:00 /sbin/syslogd -t
2022-12-05 19:28:53 <feuerteufel> any idea's?
2022-12-05 19:31:34 <Ermine> How do you start crond?
2022-12-05 19:33:59 <feuerteufel> I tried "rc-service crond [start|restart]" both of them dosn't worked
2022-12-05 19:35:26 <feuerteufel> https://tpaste.us/KeQb < here with "start"
2022-12-05 19:36:55 <feuerteufel> https://tpaste.us/vy0L < that is up when I try to start syslogd
2022-12-05 19:42:15 <minimal> what happens if you do "/etc/init.d/syslog stop; /etc/init.d/syslog start" ?
2022-12-05 19:44:28 <feuerteufel> Moment pls.
2022-12-05 19:47:19 <feuerteufel> https://tpaste.us/DBdo
2022-12-05 19:47:52 <feuerteufel> https://tpaste.us/BkZy
2022-12-05 19:48:07 <feuerteufel> Same thing ...
2022-12-05 19:48:25 <nangel> Is it just me, or is something seriously messed up with libav in 3.17?   ffmpeg/mpv/vlc all have unresolved symbols.
2022-12-05 19:51:09 <minimal> feuerteufel: how can it already be stooped when you try and stop it yet already running when you try and start it?
2022-12-05 19:51:41 <minimal> are you sure that "stop" is actually stopping it?
2022-12-05 19:52:17 <nangel> it can happen at edge cases where "stop" doesn't remove the pid file or other.   "/etc/init.d/<foo> zap"
2022-12-05 19:52:38 <psykose> nangel: that would probably be caused by having some library from a different release like when mixing stable with edge
2022-12-05 19:52:40 <feuerteufel> I'm not
2022-12-05 19:52:59 <psykose> should check if the thing that normally provides the missing symbols comes from the same repos
2022-12-05 19:53:28 <ekristen> Hello. Is there a place where I can view the package policy? Looking for when alpine will or will not introduce major versions of pacakges, like will bind only rev to a major on a new alpine release or could 3.16 for example allow bind to get a major revision with breaking changes. I believe it's the former, only new versions like edge and 3.18 would get major revisions of packages, and then older versions like 3.16 get patches an
2022-12-05 19:53:29 <ekristen> maybe minor release updates? 
2022-12-05 19:53:59 <psykose> generally no majors in release branches
2022-12-05 19:54:03 <nangel> ok, I am religious about not running edge on my production machine - "edge" is only for the dev vm's.   But as long as it is "just me", something to look into
2022-12-05 19:54:22 <psykose> you can post some output of the missing symbols real quick in a pastebin
2022-12-05 19:55:07 <feuerteufel> Is killing syslogd an option?
2022-12-05 19:55:44 <ikke> nangel: just to be sure, did you use --available when upgrading?
2022-12-05 19:56:16 <ikke> It can happen that some packages are held back during upgrade to satisfy dependencies
2022-12-05 19:57:00 <feuerteufel> Btw. syslog and cron are both from Busybox ...
2022-12-05 19:57:28 <nangel> ikke: no
2022-12-05 19:57:36 <nangel> psykose: https://pastebin.com/FzdDhmBf
2022-12-05 19:57:46 <ikke> nangel: so it can be you still have some packages from 3.16
2022-12-05 19:58:00 <psykose> now `apk policy mpv ffmpeg-libs`
2022-12-05 19:58:25 <psykose> ah and ffmpeg
2022-12-05 19:59:29 <nangel> ikke:  --available replaced a bunch of stuff, but mpv and ffmpeg still are not happy
2022-12-05 20:00:23 <nangel> AFICT, "--available" just replaced the existing packages with the ones installed.
2022-12-05 20:00:37 <ikke> That's possible
2022-12-05 20:00:55 <psykose> no, if it replaced anything it means it now installed the 3.17 ones instead of the 3.16 ones (even if the versions look the same)
2022-12-05 20:01:10 <psykose> but if they worked anyway there isn't much of a difference there
2022-12-05 20:01:28 <ikke> nangel: what about those policy commands that psykose asked about?
2022-12-05 20:01:48 <nangel> (heh... ok... /me is working on 'documentation' in place of SpaceToast... LOTS of catching up here. :) thanks)
2022-12-05 20:02:46 <psykose> (the reason that happens is when you remove the 3.16 repositories and add the 3.17 ones, things don't actually reinstall to the 3.17 versions unless there is an actual reason (dependency constraints for sonames, etc))
2022-12-05 20:02:55 <psykose> with -a it swaps everything
2022-12-05 20:03:19 <ikke> nangel: our release notes always advise people to use --available ;-)
2022-12-05 20:03:47 <psykose> if we kept perfect track for every dependency constraint one wouldn't even need -a, but sadly aside from so: links there aren't many
2022-12-05 20:03:59 <ikke> and it's a ton of work to get right
2022-12-05 20:04:03 <psykose> yeah
2022-12-05 20:04:27 <ekristen> @psykose thanks that is what I figured
2022-12-05 20:04:32 <nangel> apk policy:  https://pastebin.com/3uqDe31U
2022-12-05 20:04:50 <ikke> so ffmpeg is held back
2022-12-05 20:04:57 <nangel> ikke:  Yeah, I've been ... distracted ... for a couple years. :)
2022-12-05 20:04:58 <psykose> just the libs
2022-12-05 20:05:06 <psykose> you have ffmpeg-libs from 5.0
2022-12-05 20:05:08 <psykose> rest on 5.1
2022-12-05 20:05:27 <psykose> now as for why, ehh
2022-12-05 20:05:27 <ikke> try apk add  ffmpeg-libs=5.1.2-r1
2022-12-05 20:07:04 <nangel> That will break my x264-libs... ok.   You gave me enough pointers to figure this out.  
2022-12-05 20:07:17 <nangel> @psykose, @ikke   Thanks!
2022-12-05 20:07:33 <psykose> now you have to see why x264-libs doesn't upgrade, same process :)
2022-12-05 20:07:57 <nangel> @psykose:  :D
2022-12-05 20:09:16 <nangel> +100 points it is handbrake.... which I am listed as "Maintainer"  
2022-12-05 20:09:26 <ikke> heh
2022-12-05 20:09:30 <nangel> Just put the "cone of shame on me... I will be ok"
2022-12-05 20:09:40 <feuerteufel> minimal: Ermine: --> https://pastebin.com/sYny4RrM
2022-12-05 20:10:02 <ikke> nangel: handbrake is testing, which implies edge :P
2022-12-05 20:10:41 <ikke> maybe you built it yourself locally and it has never been rebuilt
2022-12-05 20:10:42 <feuerteufel> I killed syslogd after that starting of crond works ...
2022-12-05 20:12:19 <psykose> would be, on edge it's built against ffmpeg4 instead so it would be separate
2022-12-05 20:14:10 <vkrishn> feuerteufel: which version of busybox ?
2022-12-05 20:14:31 <nangel> ikke, psykose:  THANKS: apk del handbrake-gtk; apk upgrade --available; all is good
2022-12-05 20:14:45 <psykose> glad to help :)
2022-12-05 20:14:52 <nangel> Now to think where in the "offical user documentation" this will go
2022-12-05 20:15:05 <ikke> Upgrading to new releases
2022-12-05 20:15:08 <vkrishn> try, cat /etc/conf.d/syslog, then do, syslogd -h
2022-12-05 20:15:09 <feuerteufel> vkrishn: busybox-1.35.0-r29
2022-12-05 20:15:27 <ikke> (in-place upgrades)
2022-12-05 20:15:39 <vkrishn> then check if default option given in conf.d/syslog is available
2022-12-05 20:17:25 <feuerteufel> SYSLOGD_OPTS="-t"
2022-12-05 20:17:54 <feuerteufel> 5415 root      0:00 /sbin/syslogd -t
2022-12-05 20:18:22 <feuerteufel> Is it that what you mean?
2022-12-05 20:18:41 <vkrishn> there had been changes in busybox 1.24 and 1.31
2022-12-05 20:19:11 <vkrishn> if syslogd -h shows its available then thats not the problem
2022-12-05 20:19:33 <scorpion2185[m]1> ikke: did you see the xtrace log?
2022-12-05 20:19:39 <vkrishn> busybox syslogd -h
2022-12-05 20:19:53 <ikke> scorpion2185[m]1: no, I have not
2022-12-05 20:20:17 <scorpion2185[m]1> `8419  semop(0, [{sem_num=0, sem_op=-1, sem_flg=SEM_UNDO}], 1 <detached ...>`
2022-12-05 20:20:20 <ikke> oh ok
2022-12-05 20:20:24 <ikke> that I have seen
2022-12-05 20:20:39 <scorpion2185[m]1> that's all the content of the file
2022-12-05 20:20:50 <feuerteufel> vkrishn: -t		Strip client-generated timestamps
2022-12-05 20:21:15 <ikke> scorpion2185[m]1: I believe that's a semaphore it's waiting on, which implies some kind of deadlock, but not sure
2022-12-05 20:22:01 <ikke> scorpion2185[m]1: do you recall what process had pid 8419>
2022-12-05 20:22:44 <scorpion2185[m]1> scanelf
2022-12-05 20:25:09 <ikke> scorpion2185[m]1: what kind of environment are you running this in?
2022-12-05 20:25:49 <scorpion2185[m]1> I didn't set any env I simply put the files in my home
2022-12-05 20:26:16 <ikke> I mean host / os
2022-12-05 20:26:46 <scorpion2185[m]1> OS: postmarketOS v22.06.3 aarch64 
2022-12-05 20:26:46 <scorpion2185[m]1> Host: Pine64 PinePhone (1.2) 
2022-12-05 20:27:15 <psykose> are you by chance running this inside a pmbootstrap chroot
2022-12-05 20:27:38 <scorpion2185[m]1> no
2022-12-05 20:32:01 <ikke> scorpion2185[m]1: can you try FAKEROOT=true abuild -r`
2022-12-05 20:32:17 <ikke> apparently defining it null (`FAKEROOT= abuild -r`, does not override it)
2022-12-05 20:33:34 <feuerteufel> And it also stays working after a reboot! Tks to everybody for helping!
2022-12-05 20:35:10 <ikke> scorpion2185[m]1: not sure, but I suspect some kind of fakeroot issue, hence trying to bypass it
2022-12-05 20:35:35 <scorpion2185[m]1> it finished I think
2022-12-05 20:36:08 <scorpion2185[m]1> where is the package if ~/packages//apps/aarch64: No such file or directory
2022-12-05 20:36:39 <ikke> right, problem is probably that setting fakeroot to true kind of disabled it
2022-12-05 20:36:52 <ikke> I mean, disabled the packaging part
2022-12-05 20:37:24 <ikke> As a test, can you change this line? https://gitlab.alpinelinux.org/alpine/abuild/-/blob/master/abuild.in#L25
2022-12-05 20:37:31 <ikke> remove the 2nd :
2022-12-05 20:37:38 <ikke> so : ${FAKEROOT="fakeroot"}
2022-12-05 20:37:41 <scorpion2185[m]1> /usr/bin/abuild: cd: line 1800: can't cd to ...
2022-12-05 20:38:39 <ikke> after you did that, you should be able to do `FAKEROOT= abuild -r`
2022-12-05 20:39:45 <scorpion2185[m]1> now i have the error that you had `>>> ERROR: smlock: Found executable files with SUID bit set:`
2022-12-05 20:40:08 <ikke> yeah, so seems like fakeroot is the issue then
2022-12-05 20:40:33 <scorpion2185[m]1> only abuild -r works too
2022-12-05 20:43:55 <scorpion2185[m]1> In this case of SUID what would you do?
2022-12-05 20:44:15 <ikke> removing it is one opton
2022-12-05 20:44:18 <ikke> option
2022-12-05 20:44:39 <scorpion2185[m]1> patching the code? 
2022-12-05 20:45:18 <ikke> in package() chmod u-s $pkgdir/path/to/smlock
2022-12-05 20:45:23 <ikke> after installing it
2022-12-05 20:45:56 <feuerteufel> GN8 @ all
2022-12-05 20:46:00 <ikke> o/
2022-12-05 20:50:24 <jh81> so, one could take one of those N4 system thingy's with 4 ethernet ports on them, put Alpine on it, and build a switch?
2022-12-05 20:50:59 <scorpion2185[m]1> thanks it made it now.
2022-12-05 20:51:39 <ikke> jh81: sure, should be possible
2022-12-05 20:51:53 <ikke> alpine would mostly be the control plane
2022-12-05 20:52:06 <jh81> I'm Alpinizing my house :)
2022-12-05 20:56:45 <scorpion2185[m]1> what app can I use to open a device on vnc? X11
2022-12-05 20:57:13 <ikke> remmina
2022-12-05 20:57:14 <scorpion2185[m]1> on ssh xfontsel shows the fonts of the guest
2022-12-05 20:57:29 <scorpion2185[m]1> doesn't remmina need a vnc server?
2022-12-05 20:58:45 <scorpion2185[m]1> On the guest do I set VNC plugin and on the pinephone VNC listener?
2022-12-05 20:59:29 <scorpion2185[m]1> I was told that remmina needs it 
2022-12-05 21:03:49 <scorpion2185[m]1> https://askubuntu.com/a/248194
2022-12-05 23:00:14 <estudiante> hola
2022-12-06 08:54:16 <ok> hi there!... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/XgeVHajVejMnpgKBrhUQFZcK>)
2022-12-06 09:02:47 <psykose> is your alpine-keys up to date
2022-12-06 09:03:02 <Ermine> ok: I think you're running old system with old alpine-keys.
2022-12-06 09:10:02 <ok> actually i am trying to run https://github.com/gliderlabs/docker-alpine/blob/master/builder/scripts/mkimage-alpine.bash inside docker controller
2022-12-06 09:10:23 <alpine-meetbot> ok meant to say: actually i am trying to run https://github.com/gliderlabs/docker-alpine/blob/master/builder/scripts/mkimage-alpine.bash inside docker https:
2022-12-06 09:10:23 <ok> s/controller/https://github.com/gliderlabs/docker-alpine/blob/master/builder/Dockerfile/
2022-12-06 09:14:30 <psykose> your 'alpine:latest' image is probably out of date
2022-12-06 09:14:47 <ok> will change to edge
2022-12-06 09:14:57 <psykose> no, you just have to pull it again
2022-12-06 09:15:14 <ok> i pulled it 1 hour ago
2022-12-06 09:15:29 <ok> so from this point it is fresh
2022-12-06 09:15:30 <psykose> then it's not out of date and something else is the issue :)
2022-12-06 09:18:07 <Ermine> but it pulls edge repositories, not 3.17
2022-12-06 09:19:23 <psykose> that part is fine
2022-12-06 09:39:21 <ok> I don't have such issue if run script mkimage-alpine.bash with ARCH=x86.
2022-12-06 09:42:34 <ok> will try to add to container the follwoing keys: https://pkgs.alpinelinux.org/package/edge/main/riscv64/alpine-keys
2022-12-06 09:50:08 <psykose> ah right
2022-12-06 09:50:15 <psykose> indeed they are only present on that arch
2022-12-06 09:51:57 <ok> it looks that after downloading and untar of these keys the scipt stops to work for both x86 and riscv arch )
2022-12-06 10:10:43 <jh81> anyone happen to know how the node_exporter service is called after you install prometheus-node-exporter ?
2022-12-06 10:10:53 <jh81> or is there a way i can find this out?
2022-12-06 10:11:44 <jh81> oh, its called node-exporter (duh)
2022-12-06 10:12:23 <psykose> the -_ alternation is confusing :p
2022-12-06 10:14:28 <lucidiot> probably needs some ‾
2022-12-06 10:15:02 <lucidiot> prometheus‾node_exporter
2022-12-06 10:15:35 <jh81> :D
2022-12-06 10:28:00 <jh81> oh well, all things to write documentation for :-)
2022-12-06 10:40:14 <liske> jh81: apk info -L prometheus-node-exporter
2022-12-06 10:40:34 <liske> shows the file list of a installed package
2022-12-06 10:41:13 <liske> (apk info -L prometheus-node-exporter-openrc for the init stuff)
2022-12-06 11:22:18 <jh81> thanks :)
2022-12-06 11:23:00 <jh81> now im busy with powerdns
2022-12-06 11:34:48 <jh81> hmz, no wiki for the powerdns server
2022-12-06 11:42:34 <jh81> someone needs to teach me how to write proper wiki pages :-)\
2022-12-06 11:45:30 <psykose> you look at an existing one that looks correct
2022-12-06 11:45:40 <psykose> and just copy the markup for the correct layout
2022-12-06 11:45:51 <jh81> good point :-)
2022-12-06 11:47:08 <ikke> It's mediawiki, so there should be plenty of sources
2022-12-06 11:59:22 <Habbie> jh81, anything i can help with regarding powerdns?
2022-12-06 12:13:44 <ok> why it can be that after installing some packages it is trying to do post-install or trigger action? Is it possible to skip these post-install and trigger?... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/BRbmCAdyMnjxRTeLiTESyvgU>)
2022-12-06 12:22:17 <ok> where can i get the information about post-install and other scripts? it looks like I can use option "--no-scripts" to skip them but I would like to understand what is going on in these scripts and can I skip them
2022-12-06 12:25:12 <jh81> Habbie no, it works flawlessly it seems :)
2022-12-06 12:25:19 <Habbie> jh81, always good to hear :D
2022-12-06 12:25:44 <jh81> why?
2022-12-06 12:28:00 <Habbie> i'm the maintainer of pdns/pdnsrec/dnsdist in alpine, upstream maintainer of pdns, former upstream maintainer of rec and dnsdist
2022-12-06 12:28:05 <Habbie> (i work at powerdns)
2022-12-06 12:28:12 <jh81> ahhh nice :)
2022-12-06 12:28:52 <Habbie> :)
2022-12-06 12:29:16 <jh81> its my first real work with dns servers :)
2022-12-06 12:31:45 <jh81> on Alpine, anyway
2022-12-06 12:37:46 <jh81> i only had problems with the sql file thats missing, but i found it online :) thats why i want to make a wiki page for powerdns :)
2022-12-06 12:56:07 <markand> there is no apk-tools in official alpine docker image?
2022-12-06 12:57:02 <markand> ah yes there is, it must be something wrong in the gitlab ci/cd runner 
2022-12-06 15:30:15 <ok> which one package provides /usr/include/sys/types.h ? If believe https://pkgs.alpinelinux.org/contents?file=%2Fusr%2Finclude%2Fsys%2Ftypes.h&path=&name=&branch=edge then there is no one package that provides types.h header.
2022-12-06 15:30:34 <psykose> musl-dev
2022-12-06 15:32:23 <ikke> markand: hmm? 
2022-12-06 16:15:47 <dsl400> hello world :) 
2022-12-06 16:18:17 <dsl400> can someone help me figure this one out ? lsof -i on linux mint shows only the pid number and on alpine shows the complete info 
2022-12-06 16:19:24 <dsl400> correction lsof -t -i <port>
2022-12-06 16:23:43 <ikke> dsl400: try apk add lsof
2022-12-06 16:24:12 <Habbie> ah, busybox?
2022-12-06 16:25:11 <dsl400> woooot ?! so lsof is not lsof  until you add lsof :)))
2022-12-06 16:26:50 <minimal> dsl400: one of the reasons that Alpine is such a small distro is that by default it uses Busybox versions of a lot of common commands that may not be as feature-full as the "usual" versions of those commands
2022-12-06 16:27:38 <minimal> dsl400: if you do "lsof --help" note the 1st line mentioning Busybox...
2022-12-06 16:28:09 <dsl400> thanks guys ... you gave me back a few days 
2022-12-06 16:42:53 <jh81> sooo.. do i need to add anything special for usb to work? today at work i tried connecting a dock with usb-c to which an external monitor is hooked (i'm on a laptop) but that didnt work..
2022-12-06 16:43:06 <jh81> thankfully i still have my arch system, but heh
2022-12-06 16:44:34 <minimal> jh81: dock with external monitor connected via USB-C? so then the dock contains a graphics chip
2022-12-06 16:45:22 <jh81> hmm no i think not?
2022-12-06 16:45:52 <minimal> jh81: the laptop only connects to dock via USB-C?
2022-12-06 16:46:00 <jh81> i'll show you a pic next week :D
2022-12-06 16:46:03 <jh81> yes
2022-12-06 16:46:13 <minimal> so it is graphics via USB
2022-12-06 16:46:29 <minimal> which is either old proprietary DisplayLink
2022-12-06 16:46:35 <jh81> ok well, i'll believe whatever you say on that front :)
2022-12-06 16:46:36 <Habbie> or it's something else tunneled over the connector
2022-12-06 16:46:40 <Habbie> in which case it's not USB at all
2022-12-06 16:46:47 <minimal> or newer graphics-over-USB-C - which needs specific support
2022-12-06 16:46:56 <jh81> i think that
2022-12-06 16:47:06 <jh81> so that will not work with alpine?
2022-12-06 16:47:26 <minimal> so if its working with Arch then find out what graphics driver it is using there
2022-12-06 16:47:44 <jh81> *makes notes*
2022-12-06 16:48:15 <grawity> hmm afaik, if it's displayport-over-usbc, it should behave like just another DP output of the iGPU ... I wonder if docks do pcie-over-thunderbolt though
2022-12-06 16:49:42 <minimal> jh81: werre you talking about a separate laptop running Arch? or Arch running on the same laptop as Alpine that worked?
2022-12-06 16:49:49 <jh81> arch on the same laptop :)
2022-12-06 16:50:01 <jh81> i'm in dual boot for the time beeing
2022-12-06 16:50:14 <jh81> anyways, i'll ask my collegue to make some pics tomorrow :)
2022-12-06 16:50:21 <minimal> ok, was just checking as not all laptops with USB-C ports support graphics via that
2022-12-06 16:50:45 <jh81> but in theory, *some* dock should work?
2022-12-06 16:50:59 <jh81> si then its a matter of figuring out which one, and getting it?
2022-12-06 16:51:01 <minimal> jh81: pics? you just need to look at stuff like "dmesg" and "lsmod" in Arch to see what it uses for graphics
2022-12-06 16:51:23 <jh81> yes, but i wont be back at the officve for another two weeks or so :)
2022-12-06 16:51:52 <minimal> jh81: pics won't be much use to solve the issue
2022-12-06 16:52:43 <jh81> not true, it will show me the model :)
2022-12-06 16:53:52 <jh81> but i was thinking, maybe i needed to install a package and enable a service :)
2022-12-06 17:08:45 <jh81> hmz
2022-12-06 17:53:39 <garo> What is the "clean" way to turn off a Alpine ? poweroff, init 0, halt, shutdown -h now or maybe just politely asking "Goodevening, computer. Would you be so kind to stop working so hard ? You deserve some rest."
2022-12-06 17:54:11 <ikke> poweroff is what I use
2022-12-06 17:56:14 <minimal> garo: there is no "shutdown" command provided by busybox, only poweroff/halt/reboot
2022-12-06 17:56:48 <minimal> garo: you could also use a ACPI shutdown signal (e.g. via power button) if you have acpid running
2022-12-06 17:57:28 <garo> Does it cleanly stop all running processes ?
2022-12-06 17:57:33 <ikke> yes
2022-12-06 17:58:30 <garo> And what is "clean" ? Like "systemctl stop", "kill" or maybe even "kill -9" ?
2022-12-06 17:58:46 <minimal> just make sure you have acpid running (or tiny-power-button kernel module if a VM)
2022-12-06 17:59:02 <minimal> gare: there is no "systemctl" on Alpine, it does not use Systemd
2022-12-06 17:59:31 <garo> "service stop" in that case
2022-12-06 17:59:59 <garo> I am using too many distro's... Hard to keep apart :-(
2022-12-06 18:00:18 <minimal> What exactly are you asking? an ACPI shutdown via acpid requests init to stop all services in a "clean" fashion
2022-12-06 18:00:36 <garo> acpid wouldn't work. It's on a pi where i have no shutdown button
2022-12-06 18:00:55 <minimal> so then use poweroff
2022-12-06 18:01:10 <garo> ok, thx
2022-12-06 18:02:03 <minimal> though technically on a RPI a poweroff will not turn off power to the RPI, it will sit in a low-power state
2022-12-06 18:06:52 <garo> Couldn't the same be said about a modern pc ?
2022-12-06 18:07:21 <garo> Their clock is still running and they will liston for a wol packet
2022-12-06 18:08:56 <heat> no, the cpu turns off and so does the dram
2022-12-06 18:09:02 <minimal> garo: not to the same extent, where you power off a PC the CPU is powered down (it's the NIC that handles watching for WoL packets, the RTC runs off battery). On a RPI after "power down" the CPU is still receiving power and in some sort of "loop" state
2022-12-06 18:09:11 <heat> and all/most of the other peripherals
2022-12-06 18:38:16 <jh81> too many distro's is why i want to Alpinize everything :D
2022-12-06 18:55:35 <Saklad5> What is the intended method of binding privileged ports in OpenRC services?
2022-12-06 18:55:42 <Saklad5> In this case, Caddy?
2022-12-06 18:59:26 <minimal> Saklad5: you configure Caddy to bind to the port?
2022-12-06 18:59:45 <minimal> the caddy binary has cap_net_bind_Service capability set
2022-12-06 19:00:37 <Saklad5> How do I set that on a manually-built binary?
2022-12-06 19:00:54 <Habbie> setcap?
2022-12-06 19:01:01 <Saklad5> Great, thanks
2022-12-06 19:01:25 <Habbie> (the question mark is because i don't really use alpine much and might be missing something)
2022-12-06 19:01:41 <psykose> setcap cap_net_bind_service=+ep 
2022-12-06 19:01:42 <Saklad5> And the "Great" is because I can probably work it out from there
2022-12-06 19:02:07 <Saklad5> I'm not about to run a command without verifying for myself that it does what I think it does, after all
2022-12-06 19:02:07 <Habbie> good :)
2022-12-06 19:02:13 <Habbie> very good
2022-12-06 19:03:56 <minimal> get your stunt double to try it for you first ;-)
2022-12-06 19:04:17 <jh81> good, no? https://wiki.alpinelinux.org/wiki/Power_DNS
2022-12-06 19:05:30 <Habbie> jh81, can i suggest PowerDNS for the title?
2022-12-06 19:06:18 <jh81> well sure, but i dont think i can change that now anymore?
2022-12-06 19:06:35 <Habbie> no idea
2022-12-06 19:06:36 <jh81> never mind, found it
2022-12-06 19:06:44 <Habbie> pdns-backend-sql3 should be pdns-backend-sqlite3
2022-12-06 19:06:56 <psykose> pdns-doc has /usr/share/doc/pdns/schema.sqlite3.sql
2022-12-06 19:07:03 <jh81> https://wiki.alpinelinux.org/wiki/PowerDNS
2022-12-06 19:07:05 <psykose> which is the sample schema i assume you say needs downloading
2022-12-06 19:07:06 <psykose> :)
2022-12-06 19:07:09 <jh81> oh!
2022-12-06 19:07:11 <jh81> nice
2022-12-06 19:07:12 <Habbie> indeed
2022-12-06 19:07:34 <psykose> also contains the pgsql/mysql/mssql ones
2022-12-06 19:07:48 <jh81> typo's galore
2022-12-06 19:08:41 <Habbie> jh81, you forgot 'mkdir /var/lib/powerdns'
2022-12-06 19:09:45 <Habbie> a link to https://doc.powerdns.com/authoritative/guides/basic-database.html would make sense - "here's the non-alpine version of what we're doing here, it's slightly longer"
2022-12-06 19:09:56 <jh81> yeah, its still a WIP :)
2022-12-06 19:10:07 <Habbie> yes, certainly not calling it lame, just trying to help :)
2022-12-06 19:10:51 <jh81> much apreciated :) its my first time contributing any documentation at all :)
2022-12-06 19:13:04 <Habbie> :)
2022-12-06 19:13:41 <jh81> Do you have a website of your own?
2022-12-06 19:15:34 <jh81> well.. that didnt go well..
2022-12-06 19:29:44 <Habbie> hmm?
2022-12-06 19:35:58 <jh81> never mind :) did some cosmetic changes, added some links to web gui's
2022-12-06 19:54:31 <jh81> am i correct in thinking, that all installed packages are listed in /etc/apk/world ?
2022-12-06 19:56:01 <jh81> lots of questions, this paddawan has
2022-12-06 19:56:35 <psykose> no, that's the list of world packages
2022-12-06 19:56:40 <jh81> ow
2022-12-06 19:56:50 <elly> dependencies of those packages are not listed there though
2022-12-06 19:56:59 <psykose> all installed is `apk info`
2022-12-06 19:57:08 <jh81> ahh okay :)
2022-12-06 19:58:01 <jh81> i can pipe that to a file,. then use it in puppet for compliancy :)
2022-12-06 20:36:46 <feuerteufel> Hello @ all
2022-12-06 20:40:11 <feuerteufel> I found on one more server the problem "crond is not running" but this time on a stable one "3.17.0" and this time , I also fond log entry's in "/var/log/messages"
2022-12-06 20:40:45 <feuerteufel> https://pastebin.com/2GNwq23G <-- Here are the log entry's
2022-12-06 20:42:55 <feuerteufel> Is there something you guy's want me to do before I run the workaround (kill syslog)?
2022-12-06 21:12:11 <Ermine> Nothing, I guess.
2022-12-06 21:17:22 <feuerteufel> Ok, so fyi ...
2022-12-06 21:17:44 <feuerteufel> Workarround done
2022-12-06 21:17:52 <feuerteufel> https://pastebin.com/VaGTSg9n
2022-12-06 21:23:15 <feuerteufel> Better paste --> https://pastebin.com/TTdid08S ... ;)
2022-12-06 21:49:43 <mr_czerwinsk> hello all!
2022-12-06 22:36:06 <Saklad5> Is there a better way to automate upgrading than sticking an Ash script with "apk upgrade" in /etc/periodic/daily?
2022-12-06 22:44:39 <ptrc> why would you ever want automatic upgrades?
2022-12-06 22:45:02 <ptrc> (out of curiosity; as for the question, i don't think there's a better way)
2022-12-06 22:48:49 <Saklad5> Because I don't want to babysit a server to get security updates
2022-12-06 22:49:50 <lopid> alpine doesn't have a security updates only feature, afaik
2022-12-06 22:49:57 <Saklad5> I'll pin versions and use RSS/Atom feeds to deal with breaking changes, but for the most part I want to let things update themselves'
2022-12-06 22:50:58 <Saklad5> I'm not going to use the edge repositories or anything
2022-12-06 22:51:22 <Saklad5> Anyway, that brings me to my next question: is there a way to randomize the time of day that cron jobs run?
2022-12-06 22:51:33 <Saklad5> I know some versions of crontab have that, but I'm not sure if Busybox does
2022-12-06 22:51:55 <Saklad5> This is to avoid hammering the repositories, you see
2022-12-06 22:52:03 <lopid> sleep and $RANDOM, if your shell has that
2022-12-06 22:52:31 <Saklad5> Never knew that existed
2022-12-06 22:52:38 <elly> ya, you can just tack a random sleep onto the start of the command in your crontab
2022-12-06 22:52:58 <lopid> yes, some cron's already float their execution time if you run them "daily"
2022-12-06 22:53:04 <lopid> -'
2022-12-06 22:53:28 <Saklad5> What are the bounds on $RANDOM, then?
2022-12-06 22:53:42 <Saklad5> I'm certain that isn't part of POSIX
2022-12-06 22:53:44 <lopid> that would depend on the shell's implementation
2022-12-06 22:53:55 <elly> it's generally a lot more than you would want a job to maybe wait
2022-12-06 22:54:12 <lopid> bash's is 0-32767, but you can modulus it
2022-12-06 22:54:36 <Habbie> (or multiply it, but usually not)
2022-12-06 22:54:56 <elly> I don't know of a posix-y way to do this; if it was me and I cared a great deal about posix strictness I might write the obvious tiny C program to sleep for a random interval
2022-12-06 22:59:11 <Saklad5> I only care about POSIX insofar as I don't want to care about things breaking with updates
2022-12-06 22:59:57 <lopid> proceeds to automate updates
2022-12-06 23:02:21 <lopid> anyway, something like awk 'BEGIN{ srand(); printf "%d\n", rand() * 256 }' should be posix compatible
2022-12-06 23:02:27 <ptrc> lopid: alpine has *mostly* security updates on stable releases
2022-12-06 23:03:39 <Saklad5> > proceeds to automate updates
2022-12-06 23:03:44 <Saklad5> I'm pinning stable versions
2022-12-06 23:04:48 <Saklad5> I don't need POSIX compatibility here, I just need documentation telling me exactly what it is right now
2022-12-06 23:06:26 <Saklad5> …I'm going to have to check Busybox's source code for this, aren't I
2022-12-06 23:07:46 <ptrc> Saklad5: `# define RAND_BASH_MASK 0x7fff`
2022-12-06 23:07:48 <psykose> it's 32-bit signed positive
2022-12-06 23:07:54 <psykose> 16*
2022-12-06 23:07:58 <psykose> 32767 is the max
2022-12-06 23:08:00 <ptrc> so 0-32767
2022-12-06 23:09:30 <Saklad5> You're sure the minimum is 0?
2022-12-06 23:10:10 <Saklad5> That seems to be what Bash does, so I'll assume that's true here
2022-12-06 23:10:12 <Saklad5> Thanks
2022-12-06 23:14:52 <psykose> i'm not sure, no
2022-12-06 23:15:07 <psykose> but it wouldn't really matter unless you're trying to divide by the number or something
2022-12-06 23:15:25 <psykose> a mere modulo 1000 and sleep with it would be ok
2022-12-06 23:16:14 <Saklad5> Does "apk upgrade" run updates if the cache is expired?
2022-12-06 23:16:33 <psykose> no, pass -U
2022-12-06 23:17:19 <Saklad5> That says it is an alias for "--cache-max-age 1"
2022-12-06 23:19:04 <Saklad5> What's the default cache max age?
2022-12-06 23:20:13 <Saklad5> This would be easier if "man 8 apk-upgrade" actually seemed to work. I'm guessing I have to install that documentation separately
2022-12-06 23:21:27 <psykose> 4 hours
2022-12-06 23:21:44 <Saklad5> Which means a daily "apk upgrade" will be fine without an argument
2022-12-06 23:22:15 <Saklad5> Where was that information, by the way?
2022-12-06 23:23:59 <psykose> believe it or not, the source
2022-12-06 23:24:07 <psykose> (there are no docs for it)
2022-12-06 23:31:04 <Saklad5> Yeah, I figured
2022-12-06 23:31:16 <Saklad5> Well, here's hoping it doesn't change in a patch release
2022-12-06 23:31:55 <Saklad5> …On second thought, I'll explicitly set it to a day
2022-12-06 23:32:01 <Saklad5> No need to risk it
2022-12-07 00:48:04 <blt[m]> got big screen flickering on external monitor with 'i915 0000:00:02.0: [drm] ERROR CPU pipe D FIFO underrun: soft,transcoder' on Alder Lake, am using enable_psr=0 and playing with multiple intel_idle.max_cstate but can't make it work, it is almost unusable, anyone on the same situation who found a working solution ? (am on 3.17 with latest 5.15.81)
2022-12-07 09:24:20 <hoel> Hi, I’m trying to setup an alpine instance, I need to add some users when some packages are installed and lock them when the packages are removed. I’m used to the GLEP81 (https://www.gentoo.org/glep/glep-0081.html) way of doing things, is there any standard way on alpine or should I hack my own solution with existing tools?
2022-12-07 09:30:49 <Ermine> hoel: if you are trying to create your own package, you may want to use pre-install and post-delete scripts to create and lock users respectively.
2022-12-07 09:31:21 <hoel> Ermine: And what if several packages depend on the same user?
2022-12-07 09:34:46 <Ermine> hoel: hmm... this question is harder to ask. But, as I see from the link you provided, Gentoo makes special packages for users/groups. I don't think this is impossible here, but somebody more clever than me may provide a better solution.
2022-12-07 10:33:07 <hoel> Ermine: okay, thanks a lot for your help, I’ll wait and see if someone has another solution, and in the meantime I’ll make a package per user/group. :)
2022-12-07 11:07:27 <liske> hoel: create a meta package only providing the user and let all packages depend on it?
2022-12-07 11:08:18 <liske> since it is not in world (as long you do not use apk add to install it), it would be removed if no other package depends on it
2022-12-07 11:54:55 <hoel> liske: Yeah, that’s what I’m heading towards, indeed.
2022-12-07 13:08:05 <Habbie> jh81, s/pdns-docs/pdns-doc/
2022-12-07 13:08:30 <Habbie> jh81, and I'm not sure the packagelist link at the bottom should point to 3.17 specifically
2022-12-07 13:08:42 <Habbie> jh81, for web frontends, you could link to https://github.com/PowerDNS/pdns/wiki/WebFrontends
2022-12-07 13:39:43 <chereskata> Has anyone a clue how to get gdb inside gnome builder going? I am using CMake and even with config-opts=-DCMAKE_BUILD_TYPE=Debug and loading the symbold from the .o file i get a permission denied error 
2022-12-07 13:42:59 <chereskata> i got it working :)
2022-12-07 13:43:49 <chereskata> but i still have to manually add "file ...../program" to gdm console in gnome-builder
2022-12-07 14:40:21 <tick> Hi, anyone knows how do do USB-C video out on pinebook pro?
2022-12-07 14:42:37 <tick> *to do
2022-12-07 14:48:36 <tick> I'm using postmarketOS, but I think a lot of things are the same as alpine linux
2022-12-07 14:53:30 <ok> is it ok that in musl-dev libc.so is link to ../../lib/ld-musl-x86_64.so.1 instead of ../../;lib/libc.musl-riscv64.so.1
2022-12-07 14:53:49 <ok> s/x86_64/x86\_64/, s/-riscv64./-*./
2022-12-07 14:58:30 <Habbie> ok, at this point, just repeat the corrected question i think :)
2022-12-07 15:02:06 <ok> the question is why libc.so from musl-dev points to ../../lib/ld-musl-x86_64.so.1 instead of ../../lib/libc.musl-x86_64.so.1? 
2022-12-07 15:05:24 <q66> they are all the same file 
2022-12-07 15:05:46 <q66> so it doesn't really matter 
2022-12-07 15:06:02 <markand> looks like there is a regression somewhat in 3.17 either in mpv, linux or I don't know but mpv is stuttering a bit on a raspberry pi 4 looping a video
2022-12-07 15:06:05 <markand> it wasn't on 3.16
2022-12-07 15:06:05 <ok> got it thanks.
2022-12-07 17:31:17 <scorpion2185[m]1> I have a packages/apps/aarch64/APKINDEX.tar.gz in my home. can I add it as a repo?
2022-12-07 17:31:32 <panekj> yes
2022-12-07 17:31:58 <ikke> add /home/<username>/packages/apps to /etc/apk/repositories
2022-12-07 17:32:20 <scorpion2185[m]1> I packaged smlock, simple mobile lock, fork of a fork of slock.  so maybe is someone interested?
2022-12-07 17:33:15 <scorpion2185[m]1> no aarch64 part? ikke 
2022-12-07 17:34:00 <ikke> scorpion2185[m]1: no, apk will automatically append that based on the arch it's built for
2022-12-07 17:34:31 <panekj> more like what is in /etc/apk/arch
2022-12-07 17:34:38 <ikke> no
2022-12-07 17:34:55 <ikke> You can change that, but it will not fetch for another arch
2022-12-07 17:35:13 <panekj> 🤔
2022-12-07 17:35:44 <scorpion2185[m]1> everyone can make packages for alpine? maybe this is more a pmos app
2022-12-07 17:41:33 <psykose> this application looks like a shitpost
2022-12-07 17:42:23 <panekj> it comes from suckless so....
2022-12-07 17:43:35 <panekj> ikke: is that expected behaviour?
2022-12-07 17:48:45 <ikke> I don't think it's anywhere documented that apk uses /etc/apk/arch
2022-12-07 17:48:58 <panekj> it's documented in the code
2022-12-07 17:49:06 <panekj> it uses it
2022-12-07 17:49:33 <ikke> hmm, it seems it does now
2022-12-07 17:49:37 <ikke> I recall it didn't
2022-12-07 17:49:41 <panekj> :>
2022-12-07 17:49:46 <ikke> fetch https://dl-cdn.alpinelinux.org/alpine/edge/main/aarch64/APKINDEX.tar.gz
2022-12-07 17:49:54 <ikke> (on x86_64)
2022-12-07 17:50:11 <ikke> ok, TIL
2022-12-07 17:50:17 <panekj> https://gitlab.alpinelinux.org/alpine/apk-tools/-/issues/10761
2022-12-07 17:50:24 <panekj> i believe you refer to that?
2022-12-07 17:51:17 <ikke> No, I recall testing changing /etc/apk/arch
2022-12-07 18:14:52 <jh81> hm
2022-12-07 20:44:12 <jh81> so, my display conumdrum.. turns out its a situation where the usb devices go into the back of the monitor, which then connects with the laptop through USB-C
2022-12-07 20:49:39 <minimal> jh81: the monitor connects to laptop via USB-C? not a docking station as you said yesterday?
2022-12-07 20:49:52 <jh81> exactly
2022-12-07 20:50:03 <jh81> i thought there was a dock, but there is none
2022-12-07 20:50:17 <jh81> i should pay more atention at the office..
2022-12-07 20:50:59 <minimal> and this is used to send display output to monitor as well as monitor providing USB hub for other devices?
2022-12-07 20:52:09 <jh81> exactly
2022-12-07 20:52:45 <minimal> jh81: well you said it worked with Arch so see what/how it worked with Arch in order to try and get it working with Alpine
2022-12-07 20:52:51 <jh81> and this works OOB on Arch, so i should figure out what driver arch uses?
2022-12-07 20:52:58 <jh81> got it :)
2022-12-07 20:53:01 <minimal> yes I said this yesterday
2022-12-07 20:53:06 <jh81> exactly
2022-12-07 20:53:07 <minimal> lsmod, dmesg, etc
2022-12-07 20:53:15 <jh81> you did, but then i was still talking dock ;-)
2022-12-07 20:53:30 <minimal> also X or Wayland config
2022-12-07 20:57:53 <jh81> noted :)
2022-12-08 10:34:17 <michalkotyla> Hey, do you have any tips on how to install Alpine (I use 3.17.0) on x86_64 device without monitor output? I have access only to serial. I tried to boot from isolinux by typing console=ttyS0,115200n8 but I didn't see any output on that - isolinux is the last boot stage where i can do or see anything
2022-12-08 10:38:22 <Piraty> michalkotyla: check https://www.kernel.org/doc/html/latest/admin-guide/serial-console.html  and make sure you got use the correct device (you can specify multiple)
2022-12-08 10:39:41 <Piraty> > So if you don’t have a VGA card in your system the first serial port will automatically become the console.
2022-12-08 10:42:03 <michalkotyla> Piraty: thanks, I will try all of available options from your link
2022-12-08 10:44:52 <michalkotyla> Piraty: can you confirm that, but I am not sure now: in isolinux shell (after "boot: " prompt) i should type <label> <kernel parameters> ?
2022-12-08 10:45:11 <michalkotyla> and name of my label i can get after pressing "Tab"? it is "lts" in my case
2022-12-08 11:22:14 <ikke> correct
2022-12-08 11:27:39 <michalkotyla> hmm, that's weir. adding kernel parameter doesnt help me, but when I tried to reproduce it on qemu then i was able to boot to grub menu
2022-12-08 11:28:21 <ikke> What did you add?
2022-12-08 11:31:54 <michalkotyla> ikke: I tried with `lts console=tty0`also i do this same with a lot of different combinations like ttys0, tty0, tty1, lp0. With adding baudrate too
2022-12-08 11:33:30 <ikke> What kind of device is it?
2022-12-08 11:33:38 <ikke> Usually the serial device is something like ttyS*
2022-12-08 11:33:57 <michalkotyla> ikke: it is APU2
2022-12-08 11:34:40 <michalkotyla> ikke: I see that someone install Alpine on that device before: https://martin-m.org/2019/11/28/alpinelinux-pcengines-apu2/, but this doesnt work for me
2022-12-08 11:35:03 <ikke> hmm telmich here uses those I believe
2022-12-08 11:35:42 <ikke> From https://pcengines.github.io/apu2-documentation/serial_console/, at seems that it indeed should by ttyS0
2022-12-08 11:35:52 <ikke> Did you type it like that, with the capital S?
2022-12-08 11:36:55 <michalkotyla> ikke: yes, i tried that. I can boot on apu2 with any other OS, for example with Yocto image
2022-12-08 11:38:22 <ikke> At least for the devices I used (HP servers over serial console), appending something like console=ttyS0,9600 has worked
2022-12-08 11:47:54 <michalkotyla> I got it: do not know why, but it works when I removed ssd disk connected over mPCIe. Maybe isolinux after executing label command tried to boot from another device
2022-12-08 11:48:03 <michalkotyla> anyway, thans for help
2022-12-08 11:48:11 <ikke> glad to hear
2022-12-08 11:51:30 <minimal> michalkotyla: you can select boot device in APU BIOS
2022-12-08 11:52:25 <minimal> michalkotyla: also note there are issues with booting from USB3 ports, I've had to use onboard USB2 headers to connect flash drives to boot from them
2022-12-08 12:08:01 <michalkotyla> minimal: right, I selected SD card in APU SeaBios boot menu and it properly boot to SD card with Alpine installer, but serial output stuck after isolinux shell
2022-12-08 12:13:28 <minimal> michalkoyla: MMC (i.e. SDcard) in the linux-lts is not built in, it is a *loadable* kernel module, perhaps you need to add "modules=" entries to the bootloader cmdline to load the appropriate modules?
2022-12-08 14:15:25 <cits> hi, i wanted to play a bit with lxc in a vm running alpine but for some reason all my connections are dropped as soon as i do brlctl addif br0 eth0
2022-12-08 14:15:38 <cits> any idea what that's about?
2022-12-08 14:42:25 <minimal> cits: well you changed eth0 from being a "native" interface to being layered on top of br0.
2022-12-08 14:50:54 <ikke> cits: is there a specific reason you need to add eth0 to br0?
2022-12-08 14:57:25 <cits> ikke: that's what the wiki said
2022-12-08 14:57:27 <cits> :)
2022-12-08 14:58:15 <ikke> It's not necessary
2022-12-08 14:58:22 <cits> i just wanted to have a few containers to test a thing or two and i followed the wiki not thinking much about anything really
2022-12-08 14:59:05 <cits> but i was suprised that it axed all my connections and made the entire machine inaccessible from outside
2022-12-08 14:59:05 <ikke> I guess they do that so that your containers will be on your lan
2022-12-08 14:59:11 <cits> hm
2022-12-08 14:59:19 <ikke> after you do that, you need to add the ip address on the bridge, not eth0
2022-12-08 14:59:54 <ikke> You can give your bridge a different subnet, but then you need to enable natting
2022-12-08 15:05:28 <cits> as soon as i set the network config and reboot i can't reach anything
2022-12-08 15:08:50 <cits> https://pastebin.com/rPFA5Xca
2022-12-08 15:08:56 <cits> that's pretty much all there is to it
2022-12-08 15:09:38 <ikke> The config of eth0 is no longer relevant once you add it to the bridge
2022-12-08 15:09:44 <ikke> it becomes a layer 2 port
2022-12-08 15:10:28 <ikke> So you need a 192.168.122.0/24 address to br0
2022-12-08 15:11:05 <ikke> You now added a 192.168.122.69.0/24 address with a gateway set to 192.168.122.1
2022-12-08 15:12:16 <cits> aaaaah ok, i was not aware of that being necessary
2022-12-08 15:12:21 <cits> thanks for your help :)
2022-12-08 15:16:14 <esteticmonster> hi, first time using alpine as well as irc. hope this is whre I type my question!? I setup Aline on a RaspberryPi but first a bit confused about the folder structure where there is no boot folder where config.txt is. I am trying to add a usercfg.txt but not permitted. what can I do?
2022-12-08 15:16:41 <ikke> esteticmonster: welcome
2022-12-08 15:16:51 <esteticmonster> thanks you:)
2022-12-08 15:18:35 <esteticmonster> I also try to edit the governor files but not permitted, even with sudo.....
2022-12-08 15:19:49 <ikke> esteticmonster: are you logged in as root?
2022-12-08 15:20:30 <esteticmonster> i tried su, sudo as user, but also login as root
2022-12-08 15:22:59 <ikke> Just to verify, what file (path) are you trying to write to?
2022-12-08 15:24:33 <esteticmonster> echo powersave > /sys/devices/system/cpu/cpufreq/policy0/scaling_governor
2022-12-08 15:24:48 <esteticmonster> this come back
2022-12-08 15:24:49 <esteticmonster>  bash: echo: write error: Invalid argument
2022-12-08 15:25:16 <ikke> That's not a permission issue
2022-12-08 15:25:47 <esteticmonster> ok, but I also tried to use vim and edit the file manually to "powersave"
2022-12-08 15:26:12 <ikke> Note that these are not regular files
2022-12-08 15:26:18 <esteticmonster> and vim did not let me save and exit(wq)
2022-12-08 15:26:23 <ikke> They are virtual files that the kernel responds to
2022-12-08 15:26:28 <esteticmonster> ok
2022-12-08 15:27:06 <ikke> https://unix.stackexchange.com/questions/434735/sys-devices-system-cpu-cpu0-cpufreq-scaling-setspeed-invalid-argument
2022-12-08 15:27:23 <esteticmonster> is this becuase of the os loaded into ram?
2022-12-08 15:27:46 <esteticmonster> how do I control all of this then?
2022-12-08 15:29:21 <ikke> can you try `cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor` (without the `)
2022-12-08 15:30:43 <esteticmonster> yes, I did that before and it spits out: schedutil
2022-12-08 15:31:18 <ikke> ok, now try `echo powersave >/sys/devices/system/cpu/cpu0/cpufreq/scaling_governor`
2022-12-08 15:31:24 <ikke> (note that the path is different from what you provided)
2022-12-08 15:31:52 <ikke> https://lists.alpinelinux.org/~alpine/devel/%3Ctrinity-717de0ab-f0c9-478e-b1d5-2cab993fbbed-1626545955450%403c-app-mailcom-lxa12%3E
2022-12-08 15:32:07 <esteticmonster> bash: echo: write error: Invalid argument
2022-12-08 15:34:45 <esteticmonster> am I missing echo as a command or something?
2022-12-08 15:35:05 <ikke> esteticmonster: wondering if you are running into this: https://github.com/raspberrypi/linux/issues/4875
2022-12-08 15:37:09 <esteticmonster> what I am actually trying to do is set the base clock to be fixed at 600Hz. But I cant even edit files in the /media/mmcblk0. I should be able to do that no?
2022-12-08 15:37:38 <ikke> esteticmonster: what does `mount | grep mmcblk0` return?
2022-12-08 15:38:39 <esteticmonster> give me a sec
2022-12-08 15:39:19 <esteticmonster> it refuse to post that string here. I will try to devide it
2022-12-08 15:39:37 <esteticmonster>  on /media/mmcblk0p1 type vfat (ro,relatime,fmask=0022,dmask=0022,codepage=437,iocharset=ascii,shortname=mixed,errors=remount-ro) /media/mmcblk0p1/boot/modloop-rpi4 on /.modloop type squashfs (ro,relatime,errors=continue)
2022-12-08 15:40:22 <esteticmonster> it returns: /dev/mmcblk0p1 on /media/mmcblk0p1 type vfat (ro,relatime,fmask=0022,dmask=0022,codepage=437,iocharset=ascii,shortname=mixed,errors=remount-ro) /media/mmcblk0p1/boot/modloop-rpi4 on /.modloop type squashfs (ro,relatime,errors=continue)
2022-12-08 15:41:01 <esteticmonster> (I could not start my post with /dev/mmcblk0p1.. for some reason)
2022-12-08 15:49:36 <jh81> so, what would really be awesome, if i could use a ybikey to unlock my alpine installation.. is that even possible? Alpine beeing a secure distro and all :D
2022-12-08 15:55:14 <bananicorn> Hello, just finished writing an article on how to get Alpine running on an Olimex A20-OLinuXino-Lime2 ARM SBC
2022-12-08 15:55:25 <bananicorn> Thought it might be interesting to some people here :)
2022-12-08 15:55:38 <bananicorn> https://regrow.earth/blog/2022-12-08_alpine-a20-olinuxino-lime2.html
2022-12-08 15:56:00 <bananicorn> Will also add this information to the wiki, since it is quite sparse on information about ARM boards
2022-12-08 15:57:48 <bananicorn> Let me know if you have suggestions for things to add/change
2022-12-08 16:06:12 <esteticmonster> should I reinstall alpine and set the config files to some other device?
2022-12-08 16:07:16 <minimal> jh81: you mean a Yubikey to unlock a LUKS-encrypted rootfs?
2022-12-08 16:11:22 <scorpion2185[m]1> If you run `gem install jekyll bundler` where are the files installed? 
2022-12-08 16:11:22 <scorpion2185[m]1> In  ~/.local/share/gem/ ?
2022-12-08 16:12:14 <ikke> jh81: depends on what you want to unlock
2022-12-08 16:12:37 <ikke> I create for a non-alpine system a mechanism to unlock it based on a challenge/response from the yubikey
2022-12-08 16:12:48 <ikke> But that's just the screen lock
2022-12-08 16:13:31 <ikke> esteticmonster: /media/mmcblk0p1 is mounted read-only, which means you cannot change any files there
2022-12-08 16:18:36 <esteticmonster> ikke: does this mean I should install in data disk mode instead?
2022-12-08 16:19:08 <ikke> esteticmonster: it would probably match better with what you expect
2022-12-08 16:20:10 <esteticmonster> ok. and you dont know in what way you edit all the files in diskless mode?
2022-12-08 16:22:46 <esteticmonster> also, when I do ' lbu commit -d ', does all stored files in my home folder for example load into ram?
2022-12-08 16:23:06 <ikke> everything is already in ram
2022-12-08 16:25:31 <ikke> note that the error you are getting with changing the scaling governor has nothing to do with running from ram
2022-12-08 16:25:49 <ikke> It's just the kernel saying that what you want to write is not valid for some reason
2022-12-08 16:27:31 <esteticmonster> ok, how can you advice me on going forward?
2022-12-08 16:28:04 <ikke> I think someone with more experience running alpine on an rpi might be of better help
2022-12-08 16:28:14 <esteticmonster> how do I edit the files in the write only oartition?
2022-12-08 16:28:24 <durrendal> ikke: that yubikey unlock is interesting, how'd you go about it?
2022-12-08 16:28:31 <esteticmonster> ok, is there a irc for that?
2022-12-08 16:28:41 <ikke> try: mount -oremount,rw /media/mmcblk0p1
2022-12-08 16:28:49 <ikke> esteticmonster: this is the right place, you just have to wait a bit
2022-12-08 16:29:26 <ikke> durrendal: udev rule that runs a script when a yubikey is inserted, that script sends a challenge, verifies the response, and then stops the lock screen service
2022-12-08 16:29:30 <esteticmonster> a great, then I wait. thank you so much for your effort, very kind of you
2022-12-08 16:31:47 <esteticmonster> ikke: mount -oremount,rw /media/mmcblk0p1 did not help
2022-12-08 16:32:49 <ikke> did you by chance enable the read-only switch on the sd card?
2022-12-08 16:35:43 <durrendal> ikke: nice and simple, wicked smart way to handle that :)
2022-12-08 16:35:47 <minimal> ikke: RPI (unless its a really old RPI 1) uses uSD card which don't have physical switches (or at least I've never seen uSD with one)
2022-12-08 16:36:22 <minimal> ikke@: unless you're referring to "mount" ro switch in which case I misunderstood ;-)
2022-12-08 16:36:32 <ikke> right, was wondering whether the microsd card hard on
2022-12-08 16:36:43 <ikke> had one *
2022-12-08 16:36:47 <ikke> wut :P
2022-12-08 16:37:17 <minimal> ikke: also I don't know if the RPI has logic to check the switch - SDcard switches are *not* hardware enforced, its done via software
2022-12-08 16:37:41 <ikke> I think it's linux that checks the state
2022-12-08 16:38:00 <ikke> You see it in dmesg
2022-12-08 16:38:42 <esteticmonster> ikke: there is no such switch on the card
2022-12-08 16:38:46 <ikke> ok
2022-12-08 16:40:23 <esteticmonster> ikke: the alpine raspberry pi site state that diskless is default and "Diskless installations still allow you to install packages, save local files, and tune the system to your needs. " so I guess I am just doing it the wrong way 
2022-12-08 16:42:50 <ikke> Diskless setups are meant for systems that you rarely change. You set it up once, commit the changes, and then let it run
2022-12-08 16:43:44 <ikke> there is a list of files / directories lbu tracks. Changes to files that are not tracked and committed will be gone after the next boot
2022-12-08 16:48:49 <esteticmonster> got it, but whre do I config things then? I placed userdfg.txt manually in the root folder on the sd-card and it worked. is this where I should continue edit things?
2022-12-08 16:50:29 <jh81> ikke is that in the wiki? 
2022-12-08 16:50:40 <ikke> esteticmonster: Things related to booting the pi, sure
2022-12-08 16:50:58 <ikke> jh81: no, I have not documented it, and the way I set it up does not work for alpine
2022-12-08 16:51:59 <ikke> The udev script itself is embarsingly simple
2022-12-08 16:52:57 <jh81> Well, thats what i like! KISS principles! Fits perfectly with Alpine i would say :)
2022-12-08 16:53:37 <bananicorn> I have a question regarding the u-boot binaries included in the generic ARM image, is it possible to provide them for more boards?
2022-12-08 16:57:31 <minimal> jh81: so you weren't asking about unlocking disk encryption then?
2022-12-08 16:58:08 <jh81> Well, i was :-) but i’ll settle gor anything i can get :)
2022-12-08 16:59:33 <minimal> jh81: is encryption unlocking being handled by Grub or by initramfs' init?
2022-12-08 16:59:44 <esteticmonster> ikke: this is the error message in vim when I try to edit the file manually: E514:write error (File system full?)
2022-12-08 17:00:35 <minimal> jh81: in other words is your /boot encrypted or not?
2022-12-08 17:00:53 <jh81> In my case? Nothing. I havent gotten to that part yet, which is why i have every chance to get it right :)
2022-12-08 17:02:00 <minimal> jh81: ok, wel if your /boot was encrypted (and so Grub would need to unlock it during boot to get to kernel and initramfs files then no, I don't believe Grub currently supports unlocking via Yubikey
2022-12-08 17:02:34 <jh81> Except, i’m not using grub :) i’m using Refind
2022-12-08 17:03:38 <minimal> ok, so then any unlocking of rootfs would be done by the initramfs' init and I'm not sure it supports Yubikey or can be easily modified to do so
2022-12-08 17:04:45 <jh81> Hmmz, thats a shame
2022-12-08 17:05:24 <minimal> jh81: I didn't say it couldn't be implemented......but someone would have to do so
2022-12-08 17:06:07 <jh81> True :-) unfortunatly, it would have to be someone with knowledge :-) 
2022-12-08 17:06:25 <minimal> and it would have to then be accepted for merging inot mkinitfs package - my own remote (SSH) unlock patches for initramfs' init have been sitting for over 1 year unmerged :-(
2022-12-08 17:06:55 <jh81> Unless someone where to setup his own package repository?
2022-12-08 17:08:12 <minimal> jh81: that would only help that 1 person though, and they'd have to track other mkinitfs package changes/fixes and ensure their own machine didn't upgrade to "official" Alpine mkinitfs packages...
2022-12-08 17:08:41 <esteticmonster> ikke: could it be that I need to do all of these settings before the first reebot after setup-alpine?
2022-12-08 17:08:48 <jh81> True :-) but change often starts with one person ;-)
2022-12-08 17:09:39 <ikke> esteticmonster: no, should not be necessary
2022-12-08 17:09:53 <jh81> One person started Alpine.. ;-) and look where it is now :)
2022-12-08 17:10:12 <esteticmonster> ukke: but could it be the way to actually do it?
2022-12-08 17:11:54 <esteticmonster> ikke: is this correct, that the mmcblk0 is encrypted and mounts every time att boot? when I look at the files directly on the card in my other mashine, this mmcblk mount is identical to that...
2022-12-08 17:12:52 <ikke> it's not encrypted
2022-12-08 17:13:29 <esteticmonster> ikke: but it "unfolds" into RAM?
2022-12-08 17:14:02 <ikke> / is mounted as tmpfs, yes
2022-12-08 17:17:02 <esteticmonster> ikke: I think I understand now. tmpfs is a RAM device?
2022-12-08 17:17:35 <ikke> yes, what you write on tmpfs only is stored in RAM
2022-12-08 17:17:39 <esteticmonster> ikke: and everything on tmpfs is read only?
2022-12-08 17:17:44 <ikke> no
2022-12-08 17:17:54 <esteticmonster> you lost me...
2022-12-08 17:18:21 <ikke> what is read-only for you is the sdcard
2022-12-08 17:19:08 <esteticmonster> but I am trying to edit a file in /sys. then there should not be a problem?
2022-12-08 17:19:31 <ikke> /sys is a special filesystem
2022-12-08 17:19:34 <esteticmonster> because its mounted on tmpfs...
2022-12-08 17:19:50 <ikke> don't think you are writing to regular files
2022-12-08 17:20:18 <ikke> They are not config files
2022-12-08 17:20:27 <ikke> when you change something there and reboot, the change is gone again
2022-12-08 17:21:05 <esteticmonster> i understand, but I cant even change them to begin with as we all know....
2022-12-08 17:21:38 <ikke> not because of permissions or that it's read-only
2022-12-08 17:21:45 <esteticmonster> my plan was to change the file and then type lbu commit -d, and it would be saved
2022-12-08 17:21:52 <ikke> It doesn't work like that
2022-12-08 17:22:25 <ikke> It's a virtual file system, they are not real files
2022-12-08 17:22:28 <ikke> https://www.thegeekdiary.com/understanding-the-sysfs-file-system-in-linux/
2022-12-08 17:22:54 <esteticmonster> is it because of the allocated space of tmpfs? the erroe code in vim say no space left..
2022-12-08 17:22:59 <ikke> No
2022-12-08 17:23:17 <ikke> /sys is it's own filesystem, it's not part of / (tmpfs)
2022-12-08 17:23:50 <ikke> It's something that the kernel exposes where you can change how it behaves
2022-12-08 17:23:56 <esteticmonster> in a refular desktop install it is a refular file right?
2022-12-08 17:24:01 <ikke> no
2022-12-08 17:24:07 <ikke> it would be exactly the same
2022-12-08 17:24:22 <esteticmonster> ok, but I changed the governor file on void linux on my desktop,,
2022-12-08 17:24:49 <ikke> Different hardware, different kernel
2022-12-08 17:25:31 <esteticmonster> ok
2022-12-08 17:26:55 <esteticmonster> so what folders are viritual? I did manage to edit files in my home directory, so I guess some files are actually still real files
2022-12-08 17:27:24 <ikke> /sys and /proc
2022-12-08 17:27:28 <ikke> and to some extent, /dev as well
2022-12-08 17:27:35 <esteticmonster> ok
2022-12-08 17:28:08 <esteticmonster> do you think this is a alpine thing, not beeing able to edit the files?
2022-12-08 17:28:32 <esteticmonster> or a general rpi thing?
2022-12-08 18:06:09 <esteticmonster> ikke: I found something. Felling a bit stupid, but powersave is not listed in avalable governors list when cat /sys/devices/system/cpu/cpufreq/policy*/scaling_available_governors
2022-12-08 18:06:47 <esteticmonster> ikke: but listed in the man page on alpine site, just asumed it was there...
2022-12-08 18:14:04 <ikke> esteticmonster: ah yes, that would make sense
2022-12-08 18:14:16 <esteticmonster> haha yes
2022-12-08 18:14:35 <ikke> It can be that it changed
2022-12-08 18:14:39 <ikke> what is available I mean
2022-12-08 18:14:52 <esteticmonster> i see
2022-12-08 18:15:06 <ikke> esteticmonster: And that vim does not work most likely has to do with how vim saves files
2022-12-08 18:15:31 <esteticmonster> is userspace the best if I want to control everything, to fix all aspects?
2022-12-08 18:15:38 <ikke> userspace?
2022-12-08 18:16:02 <esteticmonster> its governor
2022-12-08 18:16:10 <esteticmonster> its A governor
2022-12-08 18:16:11 <ikke> oh, I'm not familiar with it
2022-12-08 18:17:36 <esteticmonster> performance / powersave / ondemand / conservative /userspace
2022-12-08 18:17:54 <ikke> I suppose userspace means that is a service running telling the kernel what to do
2022-12-08 18:18:38 <esteticmonster> https://wiki.alpinelinux.org/wiki/CPU_frequency_scaling
2022-12-08 18:19:05 <esteticmonster> to me it looks like all custom settings
2022-12-08 18:19:26 <ikke> https://unix.stackexchange.com/questions/153693/cant-use-userspace-cpufreq-governor-and-set-cpu-frequency
2022-12-08 18:19:54 <esteticmonster> hmm bad news then... :(
2022-12-08 18:20:18 <ikke> So it depends on the power driver what options are available
2022-12-08 18:20:30 <esteticmonster> all I wanted was set it to minimum
2022-12-08 18:20:47 <esteticmonster> do you think its possible to install powersave governor?
2022-12-08 18:21:13 <esteticmonster> can understand why they took it away
2022-12-08 18:22:04 <ikke> You don't install governors, the power driver determines what governors are available. And not sure if you can choose anything for an rpi
2022-12-08 18:24:16 <minimal> esteticmonster: powersave appears to NOT be enabled for the Alpine RPI kernels
2022-12-08 18:25:32 <esteticmonster> minimal: I can see that, thanks!
2022-12-08 18:26:19 <esteticmonster> btw, I need to go into ' su ' to echo any changes
2022-12-08 18:26:31 <esteticmonster> is this normal?
2022-12-08 18:27:07 <ikke> you need to be root to make those kinds of changes
2022-12-08 18:27:22 <ikke> and sudo does not affect redirects (>)
2022-12-08 18:27:43 <ikke> it would execute the echo as root, but the shell redirect would still be done as the current user
2022-12-08 18:28:05 <ikke> a trick to get around that is to use tee
2022-12-08 18:28:20 <ikke> echo foo | sudo tee /path/to/file
2022-12-08 18:30:20 <esteticmonster> ok!
2022-12-08 18:31:46 <esteticmonster> whats the difference between setting governor on cpu0 and policy0 ?
2022-12-08 18:32:18 <esteticmonster> ikke: the wiki say policy0 but you sent me cpu0
2022-12-08 18:32:34 <ikke> I'm not sure, I seem to locally only have policy0, seems like that changed
2022-12-08 18:46:38 <esteticmonster> I set userspace as governor and both max and min freq to 600000, and now now its fixed on 600Mhz. so all good now. GREAT!!
2022-12-08 18:49:25 <esteticmonster> and now I know all about governors in Alpine, braging YES, and just call me if you want to know anything about governos....exept for powersave. Thank you ikke!!!!
2022-12-08 18:49:39 <ikke> np
2022-12-08 19:06:23 <elly> write up what you know on the wiki :)
2022-12-08 19:11:06 <jh81> indeed
2022-12-08 19:28:37 <esteticmonster> elly: never done it, do I add stuff here:
2022-12-08 19:28:39 <esteticmonster> https://wiki.alpinelinux.org/wiki
2022-12-08 19:28:41 <esteticmonster> ?
2022-12-08 19:29:34 <esteticmonster> this would be my first contributuion to open source. Yay:)
2022-12-08 19:30:32 <jh81> yes :) its fun :)
2022-12-08 19:30:50 <esteticmonster> do I add on the "discussion"?
2022-12-08 19:31:04 <jh81> simply go to https://wiki.alpinelinux.org/wiki/page_name 
2022-12-08 19:31:15 <jh81> where page_name is what you want to name it
2022-12-08 19:31:23 <jh81> then it will say the page doesnt exist and you can edit it
2022-12-08 19:31:39 <jh81> https://wiki.alpinelinux.org/wiki/PowerDNS <-- i did this one
2022-12-08 19:33:21 <jh81> i think this is how i want my future articles to be :) a General section to explain what is and why, then the install, configure and use sections for.. well  installing, configuring and using it, and then ofcourse some troubleshooting! and a section with resources, and links to other helpful pages :)
2022-12-08 19:37:55 <ikke> If your interested in writing documentation, #alpine-docs could use help (though there is not a lot going on there right now)
2022-12-08 20:19:12 <esteticmonster> jh81: Thank you for showing me. I will collect a bit more end then make an edit. Shall I call for review on this irc?
2022-12-08 21:01:34 <jh81> I think jts always good to have multiple eyes look at it :)
2022-12-08 21:01:46 <jh81> we’re all humans, we can not know everything
2022-12-09 11:42:38 <michalkotyla> Hey, I still fighting with APU2 and Alpine  I tried to run network router on alpine (eth0 connected to internet, eth1 is connected to my PC, and PC should get internet access over APU) Can you give me some general hints on how to do that? I tried with https://cylab.be/blog/221/a-light-nat-router-and-dhcp-server-with-alpine-linux. My PC got IP addr over DHCP, but routing do not work
2022-12-09 11:43:35 <michalkotyla> I should install another package/service to route traffic from my PC over APU, to the internet network? Or it can be done only with correct configuration of firewal (iptables) and DHCP server (dhcpd)?
2022-12-09 11:44:39 <ikke> michalkotyla: You need to enable forwarding on the relevant interfaces
2022-12-09 11:44:43 <ikke> not sure if you did that
2022-12-09 11:45:25 <michalkotyla> I add "net.ipv4.ip_forward=1" to /etc/sysctl.conf
2022-12-09 11:47:48 <grawity> did you verify that the sysctl gets applied to the actual system?
2022-12-09 11:48:28 <ikke> michalkotyla: what does `cat /proc/sys/net/ipv4/conf/eth0/forwarding` and cat `/proc/sys/net/ipv4/conf/eth0/forwarding` return?
2022-12-09 11:51:21 <michalkotyla> let me check, brb
2022-12-09 11:57:33 <michalkotyla> i got 1 from cat /proc/sys/net/ipv4/conf/eth0/forwarding
2022-12-09 11:58:16 <liske> michalkotyla: you could start `tcpdump -ni any icmp` on the firewall and run a ping at the PC
2022-12-09 11:58:39 <liske> ping something like 9.9.9.9
2022-12-09 11:59:52 <liske> you should see one echo request packet incoming and one outgoing for every ping
2022-12-09 12:01:49 <jh81> oh!
2022-12-09 12:02:06 <jh81> i got routing to work on alpine, let me see
2022-12-09 12:02:39 <michalkotyla> liske: I tried with ping 9.9.9.9 and i got "Destination Host Unreachable", and there was nothing on firewall side when i have started tcpdump
2022-12-09 12:02:53 <jh81> look at these scripts: https://github.com/bobfraser1/alpine-router
2022-12-09 12:03:54 <liske> did you provide a default route via dhcp (gateway)?
2022-12-09 12:04:50 <liske> destination host unreachable means there is no arp resolution somehere
2022-12-09 12:04:53 <jh81> actually, we use static ip adresses, but normally dhcp should provide gateway and dns information yes
2022-12-09 12:05:05 <jh81> oh sorry, not for me :)
2022-12-09 12:05:15 <liske> :)
2022-12-09 12:07:25 <michalkotyla> jh81: thanks, it may be useful - i will try with these scripts too
2022-12-09 12:07:41 <jh81> do you happen to have awall installed?
2022-12-09 12:07:57 <jh81> i am still strugling getting routing top work through awall
2022-12-09 12:15:03 <michalkotyla> liske: it should be set on alpine?
2022-12-09 12:15:28 <michalkotyla> jh81: no, I tried with iptables only
2022-12-09 12:16:13 <liske> your PC needs a route, does it have a default route?
2022-12-09 12:16:26 <liske> can you ping the gateway (firewall) from the PC?
2022-12-09 12:17:10 <ikke> jh81: look at the awall-policy package, it has a set of predefined policies for different situations 
2022-12-09 12:17:51 <liske> the "destination host unreachable" is a indication there is a arp resolution problem for the gatway (firewall) which is happens one layer below iptables
2022-12-09 12:20:37 <michalkotyla> liske: yes, i can ping firewall from PC, its wokrs
2022-12-09 12:20:42 <michalkotyla> works*
2022-12-09 12:21:52 <liske> what does `ip ro get 9.9.9.9` on PC show? Should something like 9.9.9.9 via $IP
2022-12-09 12:22:00 <liske> where $IP is your firewall's inside IP
2022-12-09 12:22:30 <jh81> ikke i know it has ;-) still i struggled with it :) 
2022-12-09 12:23:29 <jh81> actually, wait.. i dont think ive seen awall-policy
2022-12-09 12:23:33 <jh81> let me try
2022-12-09 12:25:42 <michalkotyla> liske: i got "... via 192.168.1.1" where it is not IP of my firewall - it should be 192.168.1.100
2022-12-09 12:27:28 <rakslice> maybe if you could explain how your addresses setup is different from that cylab.be guide?
2022-12-09 12:27:31 <liske> michalkotyla: this matches the observations (nothing in tcpdump + destination host unreachable)
2022-12-09 12:29:56 <michalkotyla> rakslice: hm, I think only one place where I have configured something different than this guide is /etc/network/interfaces, because this step was not described
2022-12-09 12:30:11 <michalkotyla> i got configured eth0 as dhcp and eth1 as inet static with address 192.168.1.100
2022-12-09 12:30:31 <rakslice> ok, i just mean, in the guide it has .100-.200 as addresses the dhcp server will hand out to clients so you can't put your router on that
2022-12-09 12:31:34 <michalkotyla> ok, so i try with setting eth1 on my apu as 192.168.1.1
2022-12-09 12:31:35 <rakslice> anyway whatever the inside address of your router is should be what you put in the option routers line in dhcpd.conf so the clients know about the router :D
2022-12-09 12:33:04 <michalkotyla> hah, thats works! :D
2022-12-09 12:33:27 <michalkotyla> thanks you guys a lot for help!!
2022-12-09 13:06:18 <hoel> Hi, I’m trying to write a meta-package whose purpose is to create a user or a group, but abuild complains about the url field not being set. What should I put into it?
2022-12-09 13:30:00 <terrorjack54> Hi, it seems the clang/llvm packages have been removed in edge? Does anyone knows what's going on
2022-12-09 13:35:19 <Shiz> terrorjack54: generally easiest to search on https://pkgs.alpinelinux.org/packages?name=llvm*&branch=edge&repo=&arch=&maintainer=
2022-12-09 13:35:34 <Shiz> it seems the unversioned ones got removed and now it's just e.g. clang15 and llvm15
2022-12-09 13:36:05 <Shiz> as per https://git.alpinelinux.org/aports/commit/main/clang15?id=2e6d95b6252824f9b28ef50814d2d846d0491911, but that happened two months ago :p
2022-12-09 13:37:11 <terrorjack54> ooh thanks a lot
2022-12-09 13:58:36 <jh81> hoel set it to the alpine wiki? :-) 
2022-12-09 14:03:53 <hoel> jh81: I’ll do that, sounds like a decent idea.
2022-12-09 14:13:16 <jh81> every few hundred years, i have one of those :)
2022-12-09 16:58:51 <iggy> Not sure if this interests anyone, but I created an Alpine SD image for the new Radxa Rock5 SBC: https://github.com/iggy/rock5-alpine
2022-12-09 23:01:17 <Saijin_Naib> iggy: that is awesome! Did you have to make custom dtb? 
2022-12-10 00:03:27 <iggy> Saijin_Naib: I cheated for now and pulled the kernel/uboot/dtbs from the debian packages. Building those is next up on my list
2022-12-10 00:16:13 <Saijin_Naib> I had no idea that was an option... Hmmm. I've been struggling with trying to get the RetroCenter RockBox R-Cade booting Alpine but this is all way above my head
2022-12-10 06:00:16 <bananicorn> iggy: You motivated me to write a script that creates a bootable Alpine uSD card for the A20-OLinuXino-Lime2, including building U-Boot inside a chroot
2022-12-10 06:03:30 <bananicorn> It will interactively ask for the board revision etc and partitioning options
2022-12-10 06:04:01 <bananicorn> A lot of fun to write, will share it once it is complete :)
2022-12-10 06:05:45 <bananicorn> Thankfully it does not need a custom kernel, just some u-boot config changes depending on the board revision
2022-12-10 12:52:17 <esteticmonster> ikke: I left the chat the other day and was kicked out when I got back. why did this happen? also, if I want to create a new wiki page for the governor on rpi you know, how do I go about it(all the info was gone after termination)?
2022-12-10 12:52:48 <esteticmonster> or if someone else know thats fine to haha
2022-12-10 12:53:24 <ikke> I don't see you being kicked
2022-12-10 12:54:15 <lopid> 2022-12-08 22:18:31 <-- esteticmonster (~oftc-webi@195-252-61-150-no254.tbcn.telia.com) has quit (Remote host closed the connection)
2022-12-10 12:56:38 <ikke> esteticmonster: https://tpaste.us/LeJ4 
2022-12-10 13:06:53 <psykose> irc doesn't give you message history, have to be connected to see stuff (or even keep history)
2022-12-10 13:07:37 <psykose> sorry what i mean to say is this is scandinavian telia's fault they should stop being shit
2022-12-10 13:07:38 <psykose> (/s)
2022-12-10 13:07:59 <panekj> ideal timing
2022-12-10 13:08:14 <psykose> :)
2022-12-10 13:14:10 <esteticmonster> psykose: you say its because of Telia?
2022-12-10 13:14:21 <psykose> no, i'm just trolling
2022-12-10 13:14:38 <psykose> (i also have some telia issues so i'm just poking fun)
2022-12-10 13:14:41 <psykose> the actual reason is right before it
2022-12-10 13:14:55 <ikke> "Remote host closed the connection"
2022-12-10 13:15:13 <esteticmonster> psykeose: kul i jul
2022-12-10 13:15:42 <esteticmonster> ikke: I did not...
2022-12-10 13:16:11 <psykose> :)
2022-12-10 13:16:15 <psykose> esteticmonster: did you close the browser tab
2022-12-10 13:16:19 <esteticmonster> psykose: oh sorry, I meant Fun in CHristmas 
2022-12-10 13:16:35 <psykose> i know what it means
2022-12-10 13:18:08 <esteticmonster> no when I got back I got a it scared because there was only red posts and it also looked like it was terminated after my computer posted this 5 times "/dev/mmcblk/", but I was on windows at that point
2022-12-10 13:19:24 <esteticmonster> a part of the discussion was about /dev/mmcblk but anyway...
2022-12-10 13:20:44 <esteticmonster> ikke: Ofc I dont know everything about Governors on Alpine RPI, shall I do it anyway, like writing what I know?
2022-12-10 13:21:09 <esteticmonster> psykose: it ws me beeing funny
2022-12-10 13:26:37 <esteticmonster> ikke: can you help me name it, is "Governors on Raspberry Pi" good you think?
2022-12-10 14:03:58 <psykose> you can add it as a section to the raspberrypi page probably
2022-12-10 14:09:45 <esteticmonster> can someone cat /sys/devices/system/cpu/cpufreq/policy*/scaling_available_governors on a sys install, and see if it list  ' powersave '
2022-12-10 14:11:16 <psykose> https://img.ayaya.dev/ImRF0fzOumNo
2022-12-10 14:11:18 <psykose> on my rpi4
2022-12-10 14:12:15 <esteticmonster> i have the same, but on the wiki powersave is listed so I just wanted to see if a x86_64 have powersave
2022-12-10 14:12:36 <esteticmonster> https://wiki.alpinelinux.org/wiki/CPU_frequency_scaling
2022-12-10 14:29:44 <nickersonm> esteticmonster: on my Ryzen 5800X I only have 'ondemand performance schedutil'
2022-12-10 14:30:16 <esteticmonster> nickersinm: so its based on a driver?
2022-12-10 14:32:39 <nickersonm> esteticmonster: unsure. I have removed noncritical firmware.
2022-12-10 14:33:15 <ikke> esteticmonster: from what I have read, yes
2022-12-10 14:33:31 <ikke> but, it also depends on the hw what drivers are available
2022-12-10 14:33:48 <esteticmonster> ikke: ok!
2022-12-10 14:36:25 <psykose> for me it's only performance/schedutil via amd_pstate :)
2022-12-10 14:36:57 <psykose> i think ondemand would also be there if i had it enabled actually, so it's a mix of the scaling driver and enabled governors
2022-12-10 14:38:19 <esteticmonster> how do I check current scaling driver?
2022-12-10 14:47:39 <nickersonm> cat /sys/devices/system/cpu/cpufreq/policy*/scaling_governor
2022-12-10 18:25:33 <bananicorn> I have written a script to prepare a storage device with Alpine for the Olimex A20-OLinuXino-Lime2 ARM SBC
2022-12-10 18:25:37 <bananicorn> https://codeberg.org/unicorn/alpine-a20-olinuxino-lime2
2022-12-10 18:25:42 <bananicorn> Would be happy about any feedback :)
2022-12-10 18:38:28 <Saklad5> There's a package in APK that isn't compiled with the configuration I want. What's the etiquette for changing that?
2022-12-10 18:38:42 <Saklad5> Should I contact the maintainer? Submit a merge request?
2022-12-10 18:38:51 <Saklad5> Or just leave it be and compile it myself?
2022-12-10 18:40:37 <ikke> You can try 1/2 and if that doesn't work out, go for 3
2022-12-10 18:41:08 <Saklad5> Should it be in that order?
2022-12-10 18:42:48 <Saklad5> I'm probably overthinking this, I'm just going to email the maintainer and wait a few days
2022-12-10 18:52:56 <Saklad5> Actually, I should really just make a GitLab issue.
2022-12-10 18:53:42 <Saklad5> Didn't realize there was an instance for it at first
2022-12-10 18:56:47 <tetsumaki> Hello, there is a very annoying bug in the netavark package. Fixed in Alpine Edge with version 1.4.0 -> https://github.com/containers/netavark/issues/491
2022-12-10 18:56:47 <tetsumaki> But in Alpine 3.17.x, netavark use 1.3.0 version, with a bug.
2022-12-10 18:56:47 <tetsumaki> It's planned to upgrade netavark in 1.4.0 in Alpine 3.17 ?
2022-12-10 18:57:09 <psykose> you should open an issue for it
2022-12-10 18:57:34 <tetsumaki> OK psykose thanks.
2022-12-10 19:47:26 <ikke> How can I make ifupdown-ng not wait for wg0 to be up at boot
2022-12-10 19:48:20 <ikke> it's trying to resolve the dns name, which for some reason it can't (probably because the dns server is waiting for network)
2022-12-10 19:49:42 <psykose> ugh i remember that
2022-12-10 19:49:57 <psykose> i hacked around it by making dnsmasq not depend on networking and parallel startup
2022-12-10 19:50:04 <psykose> which is a terrible solution but
2022-12-10 19:50:32 <psykose> not sure how to really do it aside from separating wg0 out of 'all of networking' somehow
2022-12-10 19:53:21 <ikke> I probably need to move dnsmasq to boot as well
2022-12-10 19:57:05 <ikke> that's better
2022-12-10 19:57:07 <ikke> psykose: thanks
2022-12-10 19:57:26 <ikke> anoying having to wait ~1 minute before it boots and then still need to fix the wg0 iface
2022-12-10 20:00:03 <Ermine> You may try to set wg endpoint as ip address.
2022-12-10 20:00:16 <ikke> yeah, that was my alternative
2022-12-10 20:07:43 <psykose> if only there was somethin better
2022-12-10 20:36:44 <elly> on my system, I don't have networking start at boot because it takes longer than the entire rest of the boot process combined :(
2022-12-10 20:49:29 <bananicorn> elly: I was considering doing this too, does it have negative consequences? How/when do you start it?
2022-12-10 20:50:32 <bananicorn> Especially on my laptop with a rarely connected eth0 dhcp it takes ages
2022-12-10 20:51:08 <bananicorn> I'd love to know whether/how others work around this
2022-12-10 20:55:11 <psykose> you can start things after default with a stacked runlevel so you can log in before anything in it completes
2022-12-10 21:07:51 <bananicorn> psykose: What is a stacked runlevel?
2022-12-10 21:08:41 <psykose> https://ptrcnull.me/posts/openrc-async-services/ https://wiki.gentoo.org/wiki/OpenRC/Stacked_runlevel
2022-12-10 21:16:34 <bananicorn> Thank you!
2022-12-11 00:54:09 <orange3234> Looking for help installing on the AML-S905X-CC with the generic ARM addition. Visited libre.computer and successfully got the Raspbian img to work, so know the board is fine. Presence of the libretech-cc directory in the u-boot folder of generic ARM makes me think this is doable. Followed the instructions at https://wiki.alpinelinux.org/wiki/Alpine_on_ARM and was unsuccessful. Particularly 
2022-12-11 00:54:15 <orange3234> confused by the mention of the SPL when no such file was in the tarball. Any help or advice appreciated.
2022-12-11 02:42:30 <elly> bananicorn: well, the main consequence is having to start networking post boot :P
2022-12-11 06:53:28 <bananicorn> elly: So if I understood correctly, you start it manually?^^
2022-12-11 06:59:15 <bananicorn> orange3234: Often the u-boot binary includes the SPL, though I cannot guarantee it is the same for your board. Which command did you use to flash it to the uSD card? Or are you trying to flash directly to eMMC of the board?
2022-12-11 07:24:03 <bananicorn> In the case of flashing to uSD, it seems your board needs "bs=512 seek=1" when running dd
2022-12-11 07:24:36 <bananicorn> https://github.com/libre-computer-project/libretech-flash-tool
2022-12-11 07:24:50 <bananicorn> Perhaps it will help you to look at this :)
2022-12-11 09:17:48 <uuidNuniq> doas sysctl -a | grep nr_open : fs.nr_open = 1048576 , why is this so small on alpine? while on archlinux it is fs.nr_open = 1073741816
2022-12-11 09:23:41 <uuidNuniq> this is a changelog from systemd 240, https://0x0.st/od5u.txt
2022-12-11 09:25:04 <uuidNuniq> they removed the limit of fs.nr_open and fs.file-max
2022-12-11 09:25:17 <ikke> alpine is not using systemd, so what applies there does not necessarily applies to alpine
2022-12-11 09:25:54 <uuidNuniq> ikke: i know. i am saying, the systemd's idea was good. 
2022-12-11 09:27:38 <uuidNuniq> right now, on alpine, it is not possible to set like : rc_ulimit="-n 2000000 -u unlimited" in rc.conf, because nr_open is limited to like 1048576
2022-12-11 09:29:57 <uuidNuniq> also sysctl starts after openrc's rc.conf?  so setting a higher nr_open in sysctl.conf does not apply to rc.conf 
2022-12-11 09:33:07 <ikke> uuidNuniq: note that it's systemd that is overriding the kernel default
2022-12-11 09:33:23 <ikke> apparently it's hardcoded at 1024*1024
2022-12-11 09:33:58 <uuidNuniq> is there a kernel default? i don't find such option in kernel .config
2022-12-11 09:34:59 <ikke> uuidNuniq: https://github.com/torvalds/linux/blob/master/fs/file.c#L27
2022-12-11 09:36:42 <uuidNuniq> i see, so archlinux (or systemd) overwrite that with 1024 * 1024 * 1024 
2022-12-11 09:37:22 <ikke> I suspect they set it at runtime
2022-12-11 09:38:10 <ikke> as you mentioned, it's in the systemd release notes, and systemd can only set it at runtime
2022-12-11 09:52:56 <dwfreed> a million open files is still a lot
2022-12-11 09:55:50 <psykose> none of the rationale given is relevant as systemd is relying on memcg to do the accounting instead
2022-12-11 09:56:03 <psykose> you can use memcg to do stuff if you want but nothing does by default
2022-12-11 09:58:09 <psykose> there is also nothing stopping `sysctl fs.nr_open=1000000000` at runtime
2022-12-11 09:58:15 <psykose> so i am not exactly sure what the proposal is
2022-12-11 09:58:19 <psykose> you can set it to anything whatsoever
2022-12-11 09:58:33 <ikke> psykose: I think they want to set ulimit higher than nr_open in rc.conf, which is failing
2022-12-11 09:58:58 <psykose> ah, the classic ordering issue
2022-12-11 09:59:16 <psykose> you can still do rc_ulimit in a specific service that runs after sysctl
2022-12-11 09:59:17 <dwfreed> giving everything an insanely high ulimit is a great way to DoS yourself
2022-12-11 09:59:19 <psykose> but not rc.conf itself
2022-12-11 09:59:38 <psykose> i have absolutely zero idea why anyone would want to set that globally but
2022-12-11 09:59:40 <psykose> yeah you can't
2022-12-11 09:59:43 <psykose> unfortunate
2022-12-11 10:00:45 <dwfreed> there's one way to do it, but I think it's a dumb proposition in the first place, so I'm not going to describe it
2022-12-11 10:02:12 <psykose> does it rhyme with $ vmlinuz
2022-12-11 10:02:13 <psykose> :)
2022-12-11 10:02:24 <psykose> sure, i guess that does work
2022-12-11 10:02:25 <psykose> hah
2022-12-11 10:02:32 <dwfreed> no, it just uses openrc
2022-12-11 10:02:43 <psykose> two ways then
2022-12-11 10:02:46 <dwfreed> hint: conf.d overrides rc.conf
2022-12-11 10:03:14 <ikke> dwfreed: so what psykose already mentioned before?
2022-12-11 10:03:25 <dwfreed> no
2022-12-11 10:03:49 <dwfreed> in fact, sort of the opposite of that
2022-12-11 10:03:51 <psykose> nah, it's more subtle
2022-12-11 10:30:08 <uuidNuniq> psykose: i agree. setting for a service that starts after sysctl works. 
2022-12-11 10:33:35 <uuidNuniq> dwfreed: what kind of proposition?
2022-12-11 10:40:46 <psykose> rc_ulimit="-n 2000000 et al in rc.conf
2022-12-11 11:12:58 <jh81> hmz
2022-12-11 11:52:27 <uuidNuniq> that will work but the log will be messed up. because open rc retry until sysctl is set. 
2022-12-11 11:53:00 <uuidNuniq> and longer boot time 
2022-12-11 12:01:59 <jh81> speaking of longer boottimes.. my alpine desktop waits for both wlan0 to get its dhcp, and then chrony to do something, is there a weay i can make it continue while both dhcp and chrony do their thing in the background?
2022-12-11 12:05:37 <uuidNuniq> jh81: yes, try rc.conf to enable paralell
2022-12-11 12:05:54 <jh81> ohh thanks :) im still very new at Alpine and openrc for that matter :)
2022-12-11 12:06:09 <uuidNuniq> rc_parallel="YES"
2022-12-11 12:06:34 <uuidNuniq> however that is not stable. 
2022-12-11 12:07:04 <jh81> ow :)
2022-12-11 12:07:24 <jh81> then i'll just live with it :)
2022-12-11 12:08:22 <uuidNuniq> jh81: it says : WARNING: whilst we have improved parallel, it can still potentially lock the boot process
2022-12-11 12:08:58 <jh81> yeah, i prefer that not to happen on my desktop :D
2022-12-11 12:10:36 <uuidNuniq> jh81: however chrony does not have to depend on net
2022-12-11 12:10:49 <uuidNuniq> it retries i guess?
2022-12-11 12:12:24 <jh81> i'll have to note down what its doing exactly :) 
2022-12-11 12:12:36 <jh81> but alas! i can wait the extra 6 or 7 seconds :)
2022-12-11 12:12:39 <uuidNuniq> the default retries is 2, you may set it higher, like 5, and remove net from depends? 
2022-12-11 12:13:22 <uuidNuniq> that way, your chrony will start without needing net/dhcp. 
2022-12-11 12:13:51 <jh81> i think i can do without chrony all together, not sure of kde depends on it somewhow
2022-12-11 12:13:57 <jh81> will have to do some researching there
2022-12-11 12:14:13 <jh81> ok thanks :) will look into that :)
2022-12-11 16:30:41 <minimal> jh81: regarding slow boot, chrony is doing what it has been configured to do, in /etc/chrony.conf: "initstepslew 10 pool.ntp.org"
2022-12-11 16:31:41 <minimal> from the manpage: "the foreground chronyd process waits for initstepslew to finish before exiting. This is useful to prevent programs started in the boot sequence after chronyd from reading the clock before it has been stepped."
2022-12-11 16:32:41 <minimal> so there's a tradeoff of whether you want other boot services to wait or whether you don't want the wqait and don't care if those services start before the clock is adjusted
2022-12-11 18:10:55 <jh81> ok thanks :)
2022-12-11 20:27:33 <scorpion2185[m]1> how do I chroot into an existing installation?
2022-12-11 20:28:57 <zv> mount the data partition somewhere, mount -B /dev /there/dev, mount -t proc proc /there/proc, mount -t sysfs sysfs /there/sys, chroot /there ?
2022-12-11 20:31:31 <scorpion2185[m]1> are you sure? because you put a ? in the end
2022-12-11 20:31:51 <lopid> more or less. if you've mounted it at /mnt: for s in dev proc sys; do mount -B /"$s" /mnt/"$s"; done; chroot /mnt
2022-12-11 20:32:45 <zv> scorpion2185[m]1: you asked "how" without providing more detail; I cannot offer a copy-paste command, hence the '?'
2022-12-11 20:35:04 <scorpion2185[m]1> i have to mount it yet let's say on mnt
2022-12-11 20:35:51 <ikke> scorpion2185[m]1: first, what exactly do you want to achieve?
2022-12-11 20:36:10 <scorpion2185[m]1> boot it is on /dev/sda1 so do I mount it with mount /dev/sda1 mnt/boot?
2022-12-11 20:36:21 <ikke> sure
2022-12-11 20:36:22 <zv> if you know your partition, maybe 'mount /dev/sda3 /mnt', or you can use 'fdisk -l /dev/sda' to find "Linux filesystem" that is not boot
2022-12-11 20:36:42 <zv> and then you can mount sda1 inside as boot
2022-12-11 20:37:17 <scorpion2185[m]1> I need to install a kernel because I turned alpine into pmos without a device profile package so it doesn't have a kernel
2022-12-11 20:37:46 <ikke> right, then you need to make sure /boot, /sys, /proc and /dev are mounted inside as well
2022-12-11 20:38:36 <scorpion2185[m]1> does anyone have an alpine-chroot script?
2022-12-11 20:38:42 <scorpion2185[m]1> later I'll try it 
2022-12-11 20:39:05 <scorpion2185[m]1> I wrote it down
2022-12-11 20:39:05 <ikke> for d in dev proc sys; do mount --rbind /$d mnt/$d; done
2022-12-11 20:39:27 <ikke> chroot /mnt /bin/sh
2022-12-11 20:39:44 <ikke> (assuming you already mounted /boot
2022-12-11 21:03:31 <dwfreed> uuidNuniq: we're telling you that putting rc_ulimit="-n 2000000" in rc.conf is the dumb proposition
2022-12-11 21:20:00 <Ermine> Is Alpine WSL officially supported or is it third-party?
2022-12-11 21:20:09 <panekj> 3rd party
2022-12-12 10:25:15 <jh81> heh
2022-12-12 11:45:29 <j`ey> 'metadata for package util-linux-dev is too long', any ideas?
2022-12-12 11:47:58 <j`ey> seems like my apk is too old somehow
2022-12-12 11:48:16 <ikke> apk --version
2022-12-12 11:48:35 <j`ey> 2.12.9
2022-12-12 11:48:57 <j`ey> I forgot where the edge/stable version is stored, pretty sure Im on edge
2022-12-12 11:49:11 <j400[m]> hi! mitmproxy is failing to run for me on edge due to something in py3-openssl. can sombody reproduce that?
2022-12-12 11:49:44 <j`ey> and apk upgrade apk-tools tries to upgrade everything
2022-12-12 11:51:15 <j`ey> 'apk add --upgrade apk-tools' also tries to update everything
2022-12-12 11:52:54 <ikke> j`ey: apk upgrade --self-upgrade-only
2022-12-12 11:53:12 <j`ey> trided that
2022-12-12 11:53:24 <j`ey> https://gitlab.alpinelinux.org/alpine/aports/-/issues/14391 looks like I need to do it manually
2022-12-12 11:53:55 <panekj> `apk add apk-tools`
2022-12-12 11:54:15 <j`ey> also tries to update everything
2022-12-12 11:54:17 <ikke> You can find a static build here: https://gitlab.alpinelinux.org/api/v4/projects/5/packages/generic//v2.12.10/x86_64/apk.static
2022-12-12 11:56:51 <j`ey> I got one using apk fetch (not on x86)
2022-12-12 11:59:03 <j`ey> hm https://gitlab.alpinelinux.org/alpine/aports/-/issues/14391#note_276486 this apk add fails for me
2022-12-12 12:00:10 <j`ey> with https://termbin.com/zja5
2022-12-12 12:01:24 <j`ey> maybe I can just run update/upgrade from the one I downloaded..
2022-12-12 12:04:18 <tomzy_0_> #alpine-devel
2022-12-12 12:05:13 <tomzy_0> Hi :) are there any tools in apline linux which would allow me to prepare image that supports dual rootfs partition layout?
2022-12-12 12:05:23 <tomzy_0> the goal here would be to later use some update system with A/B dual image mechanism
2022-12-12 12:05:32 <tomzy_0> e.g. to use this https://rauc.readthedocs.io/en/latest/scenarios.html#symmetric-root-fs-slots
2022-12-12 12:07:43 <j`ey> hm, no, apk upgrade with the fetch'd version.. doesnt upgrade anything :/
2022-12-12 12:07:53 <psykose> j`ey: the other issue is that you have some stray .makedepends
2022-12-12 12:08:28 <psykose> j400[m]: it would be useful to see the actual error
2022-12-12 12:08:36 <j`ey> I have those unknowingly psykose 
2022-12-12 12:08:40 <j`ey> any hints?
2022-12-12 12:09:48 <psykose> apk del openssl-dev
2022-12-12 12:10:03 <j`ey> well.. I think I installed that to build linux or something
2022-12-12 12:10:13 <j400[m]> psykose: https://bin.disroot.org/?75dd26a6b02153f5#FkqDHPk2HLz8QGZJZ5QzpiCac7tKuMX9hbspKnBJz4EY
2022-12-12 12:11:09 <j`ey> psykose: well uh.. something is really weird
2022-12-12 12:11:17 <j`ey> psykose: i did that del, and it installed stuff lol
2022-12-12 12:11:24 <psykose> j400[m]: apk policy py3-openssl
2022-12-12 12:12:05 <psykose> j`ey: del installing stuff isn't weird
2022-12-12 12:12:14 <j`ey> seems weird to me
2022-12-12 12:12:23 <j`ey> anyway, still the same error message
2022-12-12 12:12:57 <psykose> i'm not sure if it fails on the apkindex part or on the specific package
2022-12-12 12:12:57 <psykose> hm
2022-12-12 12:13:30 <psykose> if it's the latter, then apk upgrade --ignore util-linux-dev util-linux
2022-12-12 12:13:43 <j`ey> what does that do?
2022-12-12 12:14:13 <psykose> apk upgrade --help
2022-12-12 12:14:28 <j`ey> well anyway, that gives me the same error
2022-12-12 12:14:47 <psykose> well well
2022-12-12 12:14:58 <psykose> that means you are today's lucky winner of having to manually install the new apk :)
2022-12-12 12:15:10 <j`ey> :|
2022-12-12 12:15:46 <j`ey> as in.. copying the files out of the tarball or?
2022-12-12 12:15:53 <j400[m]> psykose: ```... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/REEbQmppXDRZVAhnsUJixORy>)
2022-12-12 12:17:08 <psykose> j`ey: tar xaf apk-tools-2.12.10-r1.apk -C / --exclude=.SIGN\* --exclude=.PKGINFO
2022-12-12 12:17:19 <j`ey> ..
2022-12-12 12:17:31 <j`ey> really not a fan of doing that 
2022-12-12 12:18:03 <psykose> j400[m]: apk policy py3-cryptography
2022-12-12 12:18:20 <j400[m]> 
ACTION sent a code block: https://matrix.org/_matrix/media/v3/download/matrix.org/cFfIpWoNzbyKyujInTUFqfUy

2022-12-12 12:18:44 <j`ey> cool, now my apk is broken completely :/ since I removed openssl3-dev
2022-12-12 12:18:50 <j`ey> and I cant add it back, says no such package
2022-12-12 12:19:47 <j`ey> psykose: any ideas how to fix this now ^ ?
2022-12-12 12:21:04 <lopid> you may have it in /var/cache/apk
2022-12-12 12:21:08 <j`ey> wtf, now aok upgrade worked..
2022-12-12 12:22:30 <j`ey> I have no clue if it my system is in a good state now :/
2022-12-12 12:24:22 <psykose> j400[m]: could you type `python3` and then import cryptography and `cryptography.__about__.__version__` and paste the whole shell log
2022-12-12 12:25:08 <j`ey> any commands I can run to verify packages or something, on my hard drive
2022-12-12 12:25:22 <j400[m]> 
ACTION sent a sh code block: https://matrix.org/_matrix/media/v3/download/matrix.org/JhtZrSGQMbLjKJnMBjhFObUr

2022-12-12 12:25:24 <psykose> why wouldn't it be
2022-12-12 12:25:38 <psykose> j400[m]: well, it's no 38, so..
2022-12-12 12:25:42 <j`ey> psykose: was that to me?
2022-12-12 12:25:57 <psykose> j400[m]: you probably have some ancient version installed with pip somewhere
2022-12-12 12:26:04 <psykose> j`ey: yes
2022-12-12 12:26:14 <j`ey> cos random stuff happened and it just started working, for no reason
2022-12-12 12:26:40 <psykose> you unpacked the new one and then it worked
2022-12-12 12:26:52 <j`ey> no, I didnt
2022-12-12 12:27:05 <j400[m]> interesting, didn't think of that.
2022-12-12 12:27:05 <j400[m]> yep, that was is it.
2022-12-12 12:27:11 <j400[m]> thanks!
2022-12-12 12:27:16 <psykose> i have no idea what you did, you can post your entire shell log if you want. from what i said that was the last thing i said, so
2022-12-12 12:27:38 <psykose> if you do some other stuff and 'random stuff happened' i can't guess what it is :p
2022-12-12 12:27:47 <j`ey> I removed openssl-dev, it didnt work, then I removed openssl3-dev, then the apk-fetch'd apk stopped working, due to relocation errors
2022-12-12 12:28:01 <j`ey> then my systems apk upgrade started to work
2022-12-12 12:30:09 <psykose> if you post the whole shell log i can guess why or why not something happened, i can't do anything else though
2022-12-12 12:30:35 <j`ey> I can only paste the commands at this point, not the output
2022-12-12 12:30:43 <j`ey> anyway, clearly there was/is some bug with apk here
2022-12-12 12:31:12 <psykose> are you on edge or 3.17
2022-12-12 12:31:44 <j`ey> edge
2022-12-12 12:31:55 <psykose> do you upgrade with -a
2022-12-12 12:32:19 <j`ey> nope
2022-12-12 12:34:09 <psykose> oh right, now i see it
2022-12-12 12:34:16 <psykose> it was indeed openssl3-dev
2022-12-12 12:34:39 <j`ey> so that package doesnt exist anymore?
2022-12-12 12:35:40 <psykose> it doesn't exist and prevented the entire rest of the system from upgrading to openssl3 for the past 2 months
2022-12-12 12:35:40 <psykose> actual openssl3, not 'openssl3 the package that doesn't exist anymore'
2022-12-12 12:35:40 <psykose> then you got the new indexes that fail with the too long record, but you didn't ever get to upgrade apk because it needed openssl3
2022-12-12 12:36:28 <j`ey> well it must have existed at some point for me to have it installed.. so now openssl-dev is openssl3? im not sure I understand what you mean
2022-12-12 12:36:30 <psykose> i guess there's a small workaround for me to add there
2022-12-12 12:36:30 <psykose> thanks
2022-12-12 12:36:50 <j`ey> thanks for the help
2022-12-12 12:39:33 <psykose> the literal 'openssl3' was removed since openssl is 3
2022-12-12 12:40:04 <psykose> so since the package doesn't exist but you have the name in world, and that openssl3-dev also depends on conflicting library names (it depends on the same version= of itself for libssl3, etc), it prevents any upgrade
2022-12-12 12:40:47 <psykose> removing openssl3-dev alone is enough to fix the issue, but at that point you had a too old apk for the other index issue you need a new one for, so
2022-12-12 12:41:11 <j`ey> so two issues at once, always fun
2022-12-12 12:42:37 <psykose> 3b17eff33969bbf2fb6466fb119c573e35f30ba6 is enough for the first part i guess, but i'm not sure if an upgrade for someone behind by like 2 months because of it would work anyway at the same time as the latter issue
2022-12-12 12:42:41 <psykose> not harmful to have though
2022-12-12 12:44:27 <mendelmaleh> hi everyone, i'm having issues running lchsh, it asks for a password and replies "Authentication failed for files.". i tried both the user and root password, what else can it be?
2022-12-12 12:51:37 <psykose> i can't even run that since /etc/default/useradd doesn't exist
2022-12-12 12:52:55 <mendelmaleh> i created the file manually
2022-12-12 12:53:00 <mendelmaleh> https://wiki.alpinelinux.org/wiki/Change_default_shell
2022-12-12 12:58:13 <psykose> use shadow and `chsh` instead haha
2022-12-12 12:58:42 <psykose> or edit /etc/passwd by hand
2022-12-12 12:59:33 <psykose> as for why the password fails, it's probably not related to any password at all and fails for other reasons
2022-12-12 13:09:46 <psykose> edited the page
2022-12-12 13:14:06 <mendelmaleh> awesome, thank you psykose
2022-12-12 13:15:25 <Latrina> good afternoon folks
2022-12-12 13:15:52 <Latrina> has the iptables-nft package been removed from alpine? I am using 3.17
2022-12-12 13:21:46 <psykose> it has never existed
2022-12-12 13:22:09 <psykose> /sbin/iptables-nft is just in iptables
2022-12-12 13:24:22 <Latrina> oh I see
2022-12-12 13:24:27 <Latrina> thanks
2022-12-12 13:27:14 <psykose> (and iptables by default points to xtables-legacy-multi instead of xtables-nft-multi sadly)
2022-12-12 13:28:19 <Latrina> I am getting this error when trying to load the rules
2022-12-12 13:28:20 <Latrina> iptables v1.8.8 (legacy): can't initialize iptables table `nat': Table does not exist (do you need to insmod?)
2022-12-12 13:28:38 <Latrina> looking on internet it looks like I need iptables-ntf 
2022-12-12 13:29:31 <Latrina> I am now wondering if the iptables modules is missing in the kernel
2022-12-12 13:33:16 <Latrina> is kernel/net/netfilter/x_tables.ko the iptables kernel module?
2022-12-12 13:34:59 <psykose> did you initialise all that after upgrading the kernel
2022-12-12 13:35:13 <psykose> you have to reboot if the modules weren't loaded yet as they are not kept after upgrade
2022-12-12 13:35:53 <psykose> afaik it's `modprobe ip_tables` (and it does it for you when loading stuff)
2022-12-12 13:36:27 <psykose> so if you don't see it in lsmod and also upgraded the kernel you have to reboot first
2022-12-12 13:36:40 <Latrina> nevermind
2022-12-12 13:36:45 <Latrina> I had to reboot the thing
2022-12-12 13:36:48 <Latrina> thanks psykose 
2022-12-12 13:37:08 <psykose> yea, sadly no modules kept post-upgrade
2022-12-12 13:37:43 <Latrina> :)
2022-12-12 13:38:08 <mendelmaleh> is there something like bash-completion but for manpages? to install the -doc pkgs for what's installed already
2022-12-12 13:38:24 <psykose> apk add docs
2022-12-12 13:38:50 <mendelmaleh> nice, thanks!
2022-12-12 13:38:54 <psykose> (but this is every -doc subpackage with a valid install_if on the parent (most of them), and not just manpages, of course)
2022-12-12 13:38:57 <Latrina> btw what is x_tables? Does it load ip_tables and ip6_tables together?
2022-12-12 13:39:24 <psykose> modinfo x_tables claims description:    {ip,ip6,arp,eb}_tables backend module
2022-12-12 13:39:48 <Latrina> yes
2022-12-12 13:40:00 <Latrina> I dont understand the meaning of it
2022-12-12 13:40:08 <Latrina> :S
2022-12-12 13:40:42 <psykose> afaict based on that and https://cateee.net/lkddb/web-lkddb/NFT_COMPAT.html, the two mechanisms are x_tables and nf_tables
2022-12-12 13:41:02 <psykose> all of {those} use the x_tables kernel mechanism
2022-12-12 13:41:25 <psykose> nf_tables is separate, the compat module can bridge some x_tables formatted stuff to nf_tables
2022-12-12 13:41:29 <Latrina> oh I see. 
2022-12-12 13:41:33 <Latrina> Thank you man :P
2022-12-12 13:41:52 <psykose> iptables in userspace points to either  xtables-legacy-multi or xtables-nft-multi
2022-12-12 13:42:03 <psykose> the former is just xtables, the latter is that compat interface to actually use nf_tables
2022-12-12 13:42:22 <psykose> and, iptables-nft forcefully always points to the latter and iptables-legacy always to the former
2022-12-12 13:42:39 <psykose> and for nftables there's the entire nft frontend instead as well
2022-12-12 13:42:56 <psykose> and then the 500 other firewall frameworks you already know of that can use any of that above
2022-12-12 13:43:41 <psykose> generally, you cannot mix these, so if anything is using x_tables instead of nf_tables (or nftables compat), you're stuck with that
2022-12-12 13:43:50 <psykose> i think docker still requires legacy? it's been a very long time, not sure
2022-12-12 13:44:03 <psykose> it's the main reason our iptables points to legacy over nft still
2022-12-12 13:44:06 <psykose> the other being awall
2022-12-12 14:00:40 <grawity> you can mix these, just really shouldn't
2022-12-12 15:12:05 <bananicorn> Does alpine have support for the aptX codec for pipewire? I can't seem to find any information about it and couldn't find any relevant codecs packages
2022-12-12 15:13:57 <minimal> bananicorn: isn't aptX a licensed codec?
2022-12-12 15:14:50 <bananicorn> I think so, but some distributions package it
2022-12-12 15:16:15 <bananicorn> And it seems to be working with pulseaudio/pipewire on some distributions
2022-12-12 15:16:18 <psykose> it does support it yes
2022-12-12 15:16:29 <psykose> works for me
2022-12-12 15:16:34 <bananicorn> Do you have a hint which package I can find it in?
2022-12-12 15:16:40 <bananicorn> I am not sure what I am missing
2022-12-12 15:17:02 <psykose> for pipewire? merely installing pipewire-spa-bluez has all the bluez support and depends on it
2022-12-12 15:17:04 <minimal> bananicorn: a bit like some distros not supporting various video related codecs due to legal concerns
2022-12-12 15:17:25 <psykose> unless there's some other use for the codec i don't know of
2022-12-12 15:17:32 <minimal> though I guess the (original) aptX patent(s) have expired by now
2022-12-12 15:17:34 <bananicorn> I only get the options for A2DP with AAC, SBC and SBC-XQ; HSP is also selectable but doesn't work
2022-12-12 15:17:40 <psykose> does your device support it
2022-12-12 15:17:52 <minimal> aptX was invented/patented by a local university-spinoff in the 1980s ;-)
2022-12-12 15:18:02 <psykose> and does the same device show aptx in another distro (if you tested there)
2022-12-12 15:18:15 <psykose> if yes then it would be some other bug than the packaging part
2022-12-12 15:18:16 <psykose> https://img.ayaya.dev/gkL6scZv3z0R
2022-12-12 15:18:44 <bananicorn> I will test with another distro and see if my x220's probably ancient bluetooth even supports it
2022-12-12 15:18:46 <psykose> in general i'm not sure why you would pick aptx over aac but :p
2022-12-12 15:18:52 <bananicorn> The headphones are supposed to support it
2022-12-12 15:18:57 <psykose> what ya got
2022-12-12 15:19:11 <bananicorn> Headphones or bluetooth card?
2022-12-12 15:19:23 <psykose> both i guess, but the latter doesn't really matter here afai,
2022-12-12 15:19:33 <psykose> there's not hardware requirement for the codecs unless there's offload
2022-12-12 15:19:37 <psykose> maybe i'm mistaken
2022-12-12 15:20:39 <minimal> psykose: "not hardware requirement"? you mean headphones don'thave to support the relevant codec? then how would headphones decode audio sent to them using that codec?
2022-12-12 15:20:47 <psykose> card
2022-12-12 15:21:18 <psykose> i have the shittiest $10 bluetooth dongle imaginable, i doubt they have "specific support" for the 10 codecs my headphones have and work with them all lol
2022-12-12 15:21:34 <psykose> but it would be interesting! i'm just guessing
2022-12-12 15:21:43 <bananicorn> I think I am very stupid, my headphones do not seem to support it afterall
2022-12-12 15:21:48 <psykose> the headphones, of course, need to decode it
2022-12-12 15:21:50 <psykose> ah, happens
2022-12-12 15:21:51 <bananicorn> I was so sure that they did, very confused now
2022-12-12 15:22:02 <bananicorn> Sorry about the noise :)
2022-12-12 15:22:05 <psykose> np :)
2022-12-12 15:22:22 <psykose> the specific library is https://github.com/iamthehorker/libfreeaptx
2022-12-12 15:22:36 <psykose> originally distros had libopenaptx
2022-12-12 15:22:41 <minimal> bananicorn: basically "all" BT devices support SBC, for a long time some more expensive devices supported aptX for higher quality
2022-12-12 15:22:43 <psykose> but 0.2.1 changed the license and
2022-12-12 15:22:58 <psykose> sure but sbc sounds like someone drove over your ears with a motorcycle
2022-12-12 15:23:00 <minimal> then Sony came along with LDAC as an alternative, though I think its Sony specific
2022-12-12 15:23:01 <psykose> sbc-xq is alright
2022-12-12 15:23:27 <minimal> and I think, as psykose mentioned, that some devices support AAC as another alternative
2022-12-12 15:23:33 <bananicorn> Very interesting to hear this
2022-12-12 15:23:45 <bananicorn> Between sbc-xq and aac, which one is preferrable?
2022-12-12 15:23:46 <minimal> I'd never heard (pun) of SBC-XQ until now
2022-12-12 15:23:50 <psykose> aptx-hd is a variant of it as well
2022-12-12 15:24:07 <psykose> then there's cvsd(never use this)/msbc for bidirectional with voice
2022-12-12 15:24:23 <psykose> and i think now there is another bidi one in the works and maybe something else for audio too
2022-12-12 15:24:56 <minimal> yes there's several aptX versions in general (as the original aptX was designed for radio stations to use over ISDN)
2022-12-12 15:24:58 <psykose> in general all of these are terrible, as the limitation is usually on receiver decoding
2022-12-12 15:25:31 <psykose> it's not "cheap" to make decoding hardware that can do anything, so they have to prioritise stuff, different standards come up, etc, and you get this mess
2022-12-12 15:25:51 <bananicorn> Yeah it is a mess
2022-12-12 15:25:54 <psykose> if literally every company at once went "ok we make new standard and will absolutely not support anything else and make one good codec" it would be better but
2022-12-12 15:25:59 <psykose> that is impossible in any industry
2022-12-12 15:26:30 <psykose> the intrinsic bluetooth limitations on transmitting enough data impose limitations too, and time-based stuff
2022-12-12 15:26:35 <psykose> not an easy solution for sure :)
2022-12-12 15:27:16 <minimal> its worse than I thought - Samsung has their own codecs now: HD/Scalable/UHQ-BT
2022-12-12 15:27:23 <psykose> :)
2022-12-12 15:27:29 <minimal> and Huawei HWA LHDC
2022-12-12 15:27:38 <bananicorn> Wonderful^^
2022-12-12 15:27:40 <minimal> and Apple's using AAC
2022-12-12 15:28:09 <minimal> now to find a set of headphones that supports all of these ;-)
2022-12-12 15:28:58 <psykose> heh
2022-12-12 15:29:08 <psykose> the encoders are usually all open source
2022-12-12 15:29:22 <psykose> not the hardware decoders though
2022-12-12 15:29:33 <minimal> yeah but "legal" issues may still exist
2022-12-12 15:30:10 <bananicorn> A different question on the same topic, recently my headphones keep switching to HSP automatically, and then I am unable to switch back, sometimes the option disappears entirely (and HSP doesn't work on top of that)
2022-12-12 15:30:46 <bananicorn> Any idea what may be the cause? The automatic switching started happening very recently
2022-12-12 15:30:55 <bananicorn> HSP never worked in Alpine for some reason
2022-12-12 15:31:04 <bananicorn> (Does work in Debian)
2022-12-12 15:31:27 <psykose> are you using pipewire-media-session or wireplumber with pipewire
2022-12-12 15:31:31 <bananicorn> Wireplumber
2022-12-12 15:32:06 <psykose> hm
2022-12-12 15:32:23 <psykose> do you do anything specific the moment it switches?
2022-12-12 15:32:42 <psykose> for me it swaps from a2dp output to the hsp/hfp bidi thing the moment i ""join a call"" for debugging something
2022-12-12 15:33:19 <bananicorn> Usually while playing youtube videos in firefox, today I noticed it also happens with a telegram call
2022-12-12 15:33:36 <psykose> in the middle of one or the moment you enter it
2022-12-12 15:33:51 <bananicorn> I then have to dis/reconnect to get the option to switch back to A2DP
2022-12-12 15:34:05 <bananicorn> It varies, funnily enough
2022-12-12 15:34:18 <bananicorn> With the call, it was at the start
2022-12-12 15:34:40 <bananicorn> With the videos, I have not figured out the pattern
2022-12-12 15:35:25 <bananicorn> I have seen that there is some kind of automatic switching that was implemented (a year ago though?), but I haven't yet found how to turn it off
2022-12-12 15:35:36 <bananicorn> I suspect that this is causing it perhaps
2022-12-12 15:36:42 <psykose> you can probably use the wireplumber options to disable either autoswitching or hsp at all
2022-12-12 15:36:43 <psykose> hm
2022-12-12 15:37:41 <bananicorn> I will look into wireplumber :)
2022-12-12 15:39:14 <psykose> cp /usr/share/wireplumber/bluetooth.lua.d/50-bluez-config.lua to .config/wireplumber/bluetooth.lua.d
2022-12-12 15:39:19 <psykose> then, you should see the relevant part
2022-12-12 15:40:57 <psykose> prolly just ["bluez5.headset-roles"] = "[hsp_ag]" unless there's something more specific
2022-12-12 15:43:21 <psykose> the actual switching doesn't seem to have a clear toggle
2022-12-12 15:47:26 <bananicorn> Thank you!
2022-12-12 16:11:52 <bananicorn> I have found the culprit of auto switching
2022-12-12 16:12:49 <bananicorn> In /usr/share/wireplumber/policy.lua.d/10-default-policy.lua there is "media-role.use-headset-profile" which is true by default
2022-12-12 16:14:09 <bananicorn> Thank you for pointing me in the right direction :)
2022-12-12 16:15:13 <bananicorn> As usual, the arch wiki has all the details for every imaginable problem
2022-12-12 16:55:08 <jh81> Maybe you can add it to the Alpine wiki as well? ;-)
2022-12-12 17:12:59 <scorpion2185[m]1> usr/sbin/grub-probe: error: failed to get canonical path of /dev/sda4'.`
2022-12-12 17:17:17 <minimal> scorpion2185[m]: perhaps you should provide some context relating to that error?
2022-12-12 17:19:35 <scorpion2185[m]1> i am chrooting into an alpine installation ; i added pmos repo
2022-12-12 17:20:01 <scorpion2185[m]1> can I use alpine kernel (mainline)?
2022-12-12 17:20:20 <psykose> how would we know
2022-12-12 17:20:35 <scorpion2185[m]1> pmos has device package that provides a kernel
2022-12-12 17:21:04 <psykose> does your hardware work with the mainline kernel?
2022-12-12 17:21:06 <psykose> yes? yes
2022-12-12 17:21:07 <psykose> no? no
2022-12-12 17:21:25 <scorpion2185[m]1> yes it works with mainline kk
2022-12-12 17:23:13 <minimal> have you mounted /proc, /sys, /dev appropriately inside your chroot directory?
2022-12-12 17:24:11 <scorpion2185[m]1> I mounted the partition on mnt
2022-12-12 17:24:26 <scorpion2185[m]1> is this correct mount -B mnt/dev/ mnt/dev?
2022-12-12 17:25:42 <psykose> that mounts the folder to itself?
2022-12-12 17:25:44 <minimal> why are you trying to mount mnt/dev onto itself?
2022-12-12 17:25:54 <psykose> i'm not sure if that even works without erroring
2022-12-12 17:26:11 <scorpion2185[m]1> IDK no errors
2022-12-12 17:26:49 <scorpion2185[m]1> mount -B dev mnt/dev  ?
2022-12-12 17:27:14 <minimal> so in other words it seems you have not appropriately mounted /proc, /sys, and /dev inside the chroot directory
2022-12-12 17:38:34 <psykose> https://wiki.gentoo.org/wiki/Chroot#Configuration
2022-12-12 18:40:56 <scorpion2185[m]1> With that no errors but it doesn't boot
2022-12-12 18:41:48 <scorpion2185[m]1> I am asking pmos to how use alpine kernel. I think that I can mask pmos repo
2022-12-12 18:42:41 <Ermine> I don't think it's a good idea to mix packages of two distros.
2022-12-12 18:43:11 <scorpion2185[m]1> it is a topping on alpine
2022-12-12 18:52:35 <scorpion2185[m]1> how do I reinstall grub?
2022-12-12 18:52:52 <scorpion2185[m]1> i found a package that should alpine lts
2022-12-12 19:00:08 <scorpion2185[m]1> I removed and reinstalled grub-efi. no luck
2022-12-12 19:00:57 <Ermine> you may want to use grub-install and grub-mkconfig.
2022-12-12 19:03:05 <scorpion2185[m]1> grub-install /dev/sda1 ?  sda1 it is the efi partition
2022-12-12 19:03:29 <scorpion2185[m]1> boot it is in /.
2022-12-12 19:04:19 <Ermine> Wdym? where did you mount /dev/sda1 ? And where does your /boot reside?
2022-12-12 19:05:08 <Ermine> anyway, grub-install --efi-directory=<wherever your /dev/sda1 is mounted>
2022-12-12 19:05:28 <Ermine> Optionally --bootloader-id=Alpine .
2022-12-12 19:06:17 <scorpion2185[m]1> on fstab /boot/efi
2022-12-12 19:20:58 <scorpion2185[m]1> "no errors  reported" stuck with a black screen
2022-12-12 19:21:22 <Ermine> press enter
2022-12-12 19:24:43 <psykose>  
2022-12-12 19:25:11 <Ermine> psykose: :D
2022-12-12 19:25:27 <scorpion2185[m]1> i tried... now i also tried to swith tty
2022-12-12 19:25:48 <scorpion2185[m]1> nothing. can you flash alpine from another os?
2022-12-12 19:28:08 <psykose> usually yes
2022-12-12 19:29:18 <scorpion2185[m]1> is there a guide?
2022-12-12 19:29:30 <scorpion2185[m]1> with LUKS also
2022-12-12 19:37:57 <scorpion2185[m]1> generally is there a complete guide for lvm + luks?
2022-12-12 19:41:19 <elly> there's a page on the wiki somewhere, yep
2022-12-12 19:46:22 <minimal> scorpion2185[m]: you haven't indicated what type of device you are trying to create an image for (x86? z86_64? aarch64?), whether you have ensured that the initramfs contains the appropriate kernel modules, whether you've set grub's configuration appropriately for the cmdline settings needed and for LUKS settings, etc
2022-12-12 19:47:04 <minimal> what partitions you have created etc
2022-12-12 19:48:21 <scorpion2185[m]1> it is a x86_64 device it was working fine. I added pmos no more bootable
2022-12-12 19:48:33 <scorpion2185[m]1> no LUKS right now 
2022-12-12 19:50:28 <scorpion2185[m]1> elly: >You might also need to manually tweak $MNT/etc/fstab. 
2022-12-12 19:50:47 <minimal> again you're only responding with partial info - no one can help you if they don't know what you've done or not done
2022-12-12 19:50:47 <scorpion2185[m]1> >Todo: Does adding the --update option to extlinux ... suffice to make /boot/extlinux.conf be regenerated? Or do we need to manually tweak that file, or run update-extlinux, as well?
2022-12-12 19:51:21 <scorpion2185[m]1> looks a pain 
2022-12-12 20:21:00 <bananicorn> I just saw that apk has the -s option to simulate without making changes
2022-12-12 20:21:44 <bananicorn> So I tried "apk -s upgrade", it showed 33 packages, but then "apk upgrade" upgraded 104 packages
2022-12-12 20:21:50 <bananicorn> Am I misunderstanding how it works?
2022-12-12 20:22:09 <bananicorn> Or is this a bug?
2022-12-12 20:22:40 <ikke> bananicorn: can you show the output of both operations?
2022-12-12 20:23:10 <bananicorn> I just closed the window like the smart person that I am
2022-12-12 20:23:15 <bananicorn> I'll try on another machine
2022-12-12 20:24:43 <c7s> most likely because apk upgrade in the second command also does apk update, right?
2022-12-12 20:25:03 <bananicorn> I made sure to apk update before I did any of this
2022-12-12 20:25:09 <c7s> hm ok
2022-12-12 20:25:29 <bananicorn> I'll try to replicate it and then get back to you^^
2022-12-12 20:26:58 <bananicorn> Ah I ran apk update in a different terminal that was SSH'd into a different machine, so most likely it was what you said c7s
2022-12-12 20:27:10 <bananicorn> My brain is all kinds of funny today
2022-12-12 20:34:33 <bleb> what kernel version is alpine 3.17?
2022-12-12 20:34:45 <ikke> 5.5.x
2022-12-12 20:34:56 <ikke> 5.15.x*
2022-12-12 20:34:58 <bleb> and wouldn't you expect it to says so here https://alpinelinux.org/posts/Alpine-3.17.0-released.html
2022-12-12 20:35:15 <ikke> It has not changed since 3.16
2022-12-12 20:35:41 <Ermine> 5.15.82: https://pkgs.alpinelinux.org/packages?name=linux-lts&branch=v3.17&repo=&arch=&maintainer=
2022-12-12 20:35:48 <minimal> bleb: well you asked about 3.17, not about 3.17.0, so why would the 3.17.0 release notes cover 3.17? ;-)
2022-12-12 20:35:48 <ikke> Or even since 3.15
2022-12-12 20:38:57 <bleb> ok thanks
2022-12-12 20:39:44 <ikke> The next lts release will most likely be 6.1, so 3.18 will most likely include that
2022-12-12 20:41:54 <Ermine> ikke: is it going to be declared LTS post factum?
2022-12-12 20:42:29 <ikke> Ermine: that's how it usually works
2022-12-12 20:43:00 <ikke> GKH will pick a release (apparently the last release of each year) and declare that LTS
2022-12-12 20:45:31 <ikke> https://lore.kernel.org/lkml/CAPDLWs-Z8pYkwQ13dEgHXqSCjiq4xVnjuAXTy26H3=8NZCpV_g@mail.gmail.com/
2022-12-12 20:46:29 <ikke> `I usually pick the "last kernel of the year", and based on the normal release cycle, yes, 6.1 will be that kernel.  But I can't promise anything until it is released, for obvious reasons.`
2022-12-12 20:47:45 <Ermine> Ok, thank you!
2022-12-12 20:49:26 <Apachez> both 6.1 and 6.2 have some cool performance features
2022-12-12 20:49:36 <Apachez> also rumours has it that 6.2 will be the new LTS kernel
2022-12-12 20:54:11 <psykose> i googled 'lts kernel 6.2' and found https://9to5linux.com/linux-lite-6-2-is-slated-for-release-on-november-1st-based-on-ubuntu-22-04-lts
2022-12-12 20:54:22 <psykose> this is the worst name anyone could come up with
2022-12-12 20:54:31 <psykose> xD
2022-12-12 20:55:28 <ikke> linux lite? :D
2022-12-12 20:56:24 <psykose> :D
2022-12-12 21:01:08 <scorpion2185[m]1> >Tango-dialog-warning.png
2022-12-12 21:01:08 <scorpion2185[m]1> Warning: This will delete an existing partition table and make your data very hard to recover. If you want to dual boot, stop here and ask an expert.
2022-12-12 21:01:13 <scorpion2185[m]1> https://wiki.alpinelinux.org/wiki/LVM_on_LUKS
2022-12-12 21:58:16 <bleb> hmm setup-alpine on the 3.17 iso is not giving a diskless option
2022-12-12 21:58:19 <bleb> no more diskless?
2022-12-12 23:28:16 <uniqNuuid> anyone online? can you do a ripgrep test on your SoC?
2022-12-12 23:28:34 <uniqNuuid> i got kernel Oops 
2022-12-12 23:28:52 <panekj> is grep fine?
2022-12-12 23:29:15 <uniqNuuid> grep is fine panekj 
2022-12-12 23:29:21 <uniqNuuid> single threaded 
2022-12-12 23:29:35 <dwfreed> what's the backtrace on the oops
2022-12-12 23:29:37 <panekj> anyway, rg works fine for me
2022-12-12 23:29:45 <psykose> if you get an oops from running ripgrep that is quite the kernel issue
2022-12-12 23:29:55 <panekj> but also you didn't ask what kind of SoC
2022-12-12 23:30:01 <uniqNuuid> rg is too poweroff that it panic my kernel. this happends on my SoC only. not on arm vm not on x86 intel 
2022-12-12 23:30:23 <uniqNuuid> amlogic s905d
2022-12-12 23:31:00 <uniqNuuid> here is the log : https://0x0.st/onNy.txt
2022-12-12 23:31:01 <psykose> what kernel are you running
2022-12-12 23:31:05 <psykose> thanks
2022-12-12 23:31:12 <psykose> well
2022-12-12 23:31:19 <psykose> probably caused by 6.1
2022-12-12 23:31:20 <psykose> shrug
2022-12-12 23:31:29 <psykose> no idea go report it to the maintainers on arch
2022-12-12 23:31:59 <uniqNuuid> i also tested on 5.19.8 linux-aarch64, same result 
2022-12-12 23:32:10 <uniqNuuid> which is the kernel from alarm 
2022-12-12 23:32:22 <psykose> when did the change occur
2022-12-12 23:33:23 <dwfreed> 
ACTION notes that both of these sound like archlinux, not alpine

2022-12-12 23:34:00 <uniqNuuid> if it is a preempt kernel, it panic right after rg command... if it is preempt none, it panics after any new ssh connection  or after some minutes after doing nothing 
2022-12-12 23:34:19 <uniqNuuid> ohh... am i on alpine?
2022-12-12 23:34:21 <panekj> i mean, it's right in the 2nd line: "Linux version 6.1.0-1-phicomm-n1-jst (linux-phicomm-n1@archlinux)"
2022-12-12 23:34:42 <dwfreed> this is #alpine-linux 
2022-12-12 23:34:54 <ptrc> uniqNuuid: yeah, not only wrong channel, but wrong network altogether :p
2022-12-12 23:35:07 <ptrc> (as in, iirc the actual #archlinux is on libera)
2022-12-12 23:35:10 <dwfreed> but your logs seem to be from archlinux, and 'alarm' usually means archlinux on arm
2022-12-12 23:35:22 <uniqNuuid> sorry, i am totally blinded... i am debuging almos sleepy 
2022-12-12 23:35:22 <panekj> #archine-linux
2022-12-12 23:35:30 <panekj> perhaps, go sleep
2022-12-12 23:35:36 <panekj> food for thought
2022-12-12 23:37:19 <uniqNuuid> well, however, the kernel config is borrowed from alpine linux-lts config
2022-12-12 23:37:42 <uniqNuuid> with just a small modification by enabling amlogic
2022-12-12 23:40:59 <uniqNuuid> i think it is a hardware bug. i'll sleep for now, and i'll make a new kernel when i wakeup. 
2022-12-13 00:00:59 <uniqNuuid> oh my, i just found linux-edge has meson support. that's amazing. 
2022-12-13 00:07:50 <elly> meson support?
2022-12-13 00:09:02 <uniqNuuid> yeah, that's amlogic SoC
2022-12-13 00:10:32 <uniqNuuid> i'll switch to alpine :P
2022-12-13 00:11:25 <uniqNuuid> but linux-lts does not have it. i was looking at lts all the time, forgot to check linux-edge 
2022-12-13 00:13:47 <psykose> huh?
2022-12-13 00:14:14 <psykose> the kernel has 'meson support'?
2022-12-13 00:14:23 <uniqNuuid> CONFIG_MESON
2022-12-13 00:15:50 <psykose> that doesn't exist?
2022-12-13 00:17:03 <psykose> ah
2022-12-13 00:17:09 <psykose> meson soc
2022-12-13 00:17:20 <psykose> yeah it does
2022-12-13 00:17:24 <uniqNuuid> CONFIG_ARCH_MESON
2022-12-13 00:17:33 <uniqNuuid> but lts does not have it 
2022-12-13 00:18:09 <ptrc> so it looks like i wasn't the only one surprised by the kernel supposedly having something else than a pile of makefiles :p
2022-12-13 00:18:33 <bleb> what if I want to install alpine on a disk, but make it run from ram and only write to the disk when I lbu commit?
2022-12-13 00:19:07 <bleb> would that be a sys install?
2022-12-13 00:19:50 <ptrc> bleb: that sounds like installing it as sys, then adding some kernel parameters
2022-12-13 00:20:00 <ptrc> like overlaytmpfs 
2022-12-13 00:20:30 <ptrc> and probably apkovl= to point to the other disk?
2022-12-13 00:20:43 <ptrc> or just no `root=`, i suppose
2022-12-13 00:23:31 <minimal> that just sounds like the usual run-from-ram version expect you install it onto a disk rather than a SDcard / USB stick
2022-12-13 00:27:11 <bleb> in diskless/run-from-ram mode how would you update the kernel / other os files?
2022-12-13 00:32:55 <bleb> nvm I see the wiki article
2022-12-13 00:44:47 <bleb> I am following this guide to install alpine on a preexisting partition: https://wiki.alpinelinux.org/wiki/Dualbooting#In_Alpine_2.2.3_or_newer
2022-12-13 00:45:26 <bleb> when I run setup-disk -m sys /mnt, I get grub-install: error: failed to get canonical path of `/mnt/boot/efi'.
2022-12-13 00:45:51 <bleb> install: can't stat '/mnt/boot/efi/EFI/alpine/grubx64.efi': No such file or directory
2022-12-13 00:46:41 <bleb> there is a blurb that says "If you use extlinux (default) and didn't create a separate boot partition... You might need to adjust the paths to the boot files in /boot/extlinux.conf..."
2022-12-13 00:47:03 <bleb> but I have no /boot/extlinux.conf
2022-12-13 00:47:09 <bleb> and no /mnt/boot/extlinux.conf
2022-12-13 00:49:23 <bleb> if I run setup-disk -m sys /mnt again, the grub-install error now says "/mnt/boot/efi doesn't look like an EFI partition."
2022-12-13 00:58:15 <bleb> I guess it really wants a boot partition?
2022-12-13 01:23:48 <elibrokeit> psykose: and here I was wondering why the kernel devs felt the need to implement Meson as a kernelspace module. Surely Meson isn't *that* slow.
2022-12-13 01:24:07 <psykose> but just think of all the performance gains
2022-12-13 01:42:33 <bleb> not able to boot, so I probably need grub-install to work without errors
2022-12-13 01:46:03 <bleb> there is no extlinux.conf anywhere under /boot, and no file that resembles it, e.g. no LINUX /boot/vmlinuz-lts line
2022-12-13 01:54:54 <bleb> maybe I just need to redo everything with gpt instead of msdos partitions
2022-12-13 01:55:44 <bleb> the grub-install command it was running is: grub-install '--target=x86_64-efi' '--efi-directory=/mnt/boot/efi' '--bootloader-id=alpine' '--boot-directory=/mnt/boot' --no-nvram
2022-12-13 01:56:21 <bleb> it said "failed to get canonical path" on first running and now says "/mnt/boot/efi doesn't look like an EFI partition."
2022-12-13 08:51:44 <jh81> good morning, ikke can i bother you in a PM?
2022-12-13 11:17:05 <uuidNuniq> from this: https://git.alpinelinux.org/aports/plain/community/linux-edge/APKBUILD?h=3.17-stable
2022-12-13 11:17:05 <ikke> jh81: I prefer to chat in channels, but if you think it's better suited for a PM, feel ree
2022-12-13 11:17:23 <uuidNuniq> i don't see make Image or make dtbs. 
2022-12-13 11:18:49 <uuidNuniq> so, it is just a make without target, is that vmlinuz compressed or uncompressed? 
2022-12-13 11:21:30 <uuidNuniq> the question is: what does make without targets do? 
2022-12-13 11:21:57 <ikke> It will execute the first target in the makefile
2022-12-13 11:22:24 <ikke> https://stackoverflow.com/a/2057716/20261
2022-12-13 11:27:07 <uuidNuniq> i see https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Makefile?h=v6.1#n17
2022-12-13 11:27:38 <uuidNuniq> so in kernel case, it is PHONY := __all and then __all: 
2022-12-13 11:29:16 <uuidNuniq> i hope the produced vmlinuz is uncompressed, because my uboot does not boot a compressed kernel. 
2022-12-13 11:29:29 <dwfreed> file will tell you
2022-12-13 11:29:42 <uuidNuniq> however it boots a compressed initrd
2022-12-13 11:30:10 <uuidNuniq> dwfreed: ok, i'll install the linux-edge on my vm and file it to see 
2022-12-13 11:31:21 <dwfreed> uboot does not care about initrd at all, other than putting it in the right place and telling the kernel it's there
2022-12-13 11:31:48 <dwfreed> initrd is otherwise handled entirely by the kernel
2022-12-13 11:32:42 <dwfreed> also bootloaders don't care about kernel compression either, because the kernel entry point will be the decompressor
2022-12-13 11:32:49 <uuidNuniq> gonna give it a try on my amlogic soc, tired of building/upgrading archlinux arm kernel myself. 
2022-12-13 11:33:19 <dwfreed> generally the issue there is the placement point for the original kernel image and where the decompressor writes to can overlap in bad ways
2022-12-13 11:34:57 <uuidNuniq> dwfreed: you're saying zImage? a self-extracting kernel 
2022-12-13 11:35:19 <dwfreed> all compressed Linux kernels are self-extracting
2022-12-13 11:35:43 <uuidNuniq> not really. my uboot can't boot image.gz
2022-12-13 11:35:51 <uuidNuniq> make Image.gz 
2022-12-13 11:36:07 <uuidNuniq> however, make Image boots fine 
2022-12-13 11:36:13 <dwfreed> "can't boot" is not an error message
2022-12-13 11:36:25 <dwfreed> or even a useful description
2022-12-13 11:36:55 <uuidNuniq> yeah, i know, but i can't see it. it is headless, i can only judge from ping
2022-12-13 11:37:18 <ikke> so maybe it's booted but does not have a network connection?
2022-12-13 11:37:18 <dwfreed> you really need a serial console
2022-12-13 11:37:57 <uuidNuniq> ikke: everything is the same, the only difference is make Image vs make Image.gz when compiling the kernel 
2022-12-13 11:42:27 <uuidNuniq> i still can't believe rg makes kernel oops. 
2022-12-13 11:42:30 <dwfreed> there's a difference between Image.gz and a compressed kernel
2022-12-13 11:42:54 <dwfreed> Image.gz runs gzip as the very last step
2022-12-13 11:43:04 <dwfreed> which is not going to be self-extracting
2022-12-13 11:44:16 <uuidNuniq> dwfreed: ah, i see, thanks!
2022-12-13 11:48:09 <uuidNuniq> i am an archlinux user for a long time, however the aarch64 target is not supported officially and the alarm project is kinda poorly maintained. 
2022-12-13 11:48:33 <Piraty> like arch :p
2022-12-13 11:49:01 <Piraty> aur at least is a nightmare
2022-12-13 11:49:45 <uuidNuniq> i am gonna install alpine edge in a vm and then rsync over to a mounted usb drive, setup uboot config and fstab etc. and boot my box from there. 
2022-12-13 11:50:46 <uuidNuniq> Piraty: i don't use AUR even on x86_64. i need only a small amount of packages as i run linux only on headless. 
2022-12-13 11:51:47 <Piraty> dwfreed is right, you need a serial console for that. how else will you fix kernel issues
2022-12-13 11:52:48 <Piraty> best scenario i came up with is two rpi, each one connected to serial of the other and only run upgrade on one of them at a time
2022-12-13 11:53:00 <uuidNuniq> the box has HDMI 2.0 that handles 4k resolution. but when it does not boot, the screen is just black after the logo. 
2022-12-13 12:58:36 <uuidNuniq> /boot/vmlinuz-edge: gzip compressed data, max compression, from Unix, original size modulo 2^32 20869632
2022-12-13 13:05:02 <uuidNuniq> is there a hook after kernel installation/upgrade?
2022-12-13 13:05:57 <uuidNuniq> zcat /boot/vmlinuz-edge > Image , this is able to uncompress the file, result in 20M from 7M
2022-12-13 13:07:14 <minimal> uuidNuniq: a hook for what?
2022-12-13 13:08:07 <uuidNuniq> Image: Linux kernel ARM64 boot executable Image, little-endian, 4K pages
2022-12-13 13:08:34 <uuidNuniq> minimal: i'd like to gunzip the vmlinuz-edge after a kernel install/upgrade. 
2022-12-13 13:08:50 <uuidNuniq> a hook for apk or for the kernel package?
2022-12-13 13:10:04 <uuidNuniq> stuff like pacman hooks, or /etc/kernel/postinst.d/ on debian. 
2022-12-13 13:10:42 <minimal> create your own package that sets a trigger like mkinitfs package does? (it does that to rebuild initramfs after kernel update)
2022-12-13 13:11:40 <uuidNuniq> yeah, i don't think i'll create my own package... for the kernel. that's a lot of work. i am moving away from alarm because i am tired of building kernels 
2022-12-13 13:12:07 <minimal> uuidNuniq: I didn't say to build your own kernel package
2022-12-13 13:12:40 <minimal> I aid to build your own package that has a trigger activated by an update to the kernel package
2022-12-13 13:12:48 <minimal> s/aid/said/
2022-12-13 13:12:49 <alpine-meetbot> minimal meant to say: I said to build your own package that has a trigger activated by an update to the kernel package
2022-12-13 13:13:16 <uuidNuniq> minimal: interesting, is there document for that purpose?
2022-12-13 13:14:20 <minimal> https://wiki.alpinelinux.org/wiki/APKBUILD_Reference
2022-12-13 13:16:23 <minimal> also why do you want to decompress the kernel? didn't you say uboot couldn't boot image.gz? didn't dwfreed earlier say that the vmlinuz-lts is not the same thing as a image.gz file?
2022-12-13 13:18:39 <uuidNuniq> yeah, but as i found that the vmlinuz-edge is compressed by gz 
2022-12-13 13:19:04 <uuidNuniq> see above :  /boot/vmlinuz-edge: gzip compressed data, ...
2022-12-13 13:20:41 <uuidNuniq> anyway, i'll try it first. 
2022-12-13 13:20:46 <minimal> uuidNuniq: you seem to be missing the point, dwfreed earlier said that the vmlinuz-lts is not the same thing as a image.gz file
2022-12-13 13:21:06 <minimal> indeed a kernel can be compressed using various compression algorithms
2022-12-13 13:21:16 <uuidNuniq> it is the samething. just with or without a suffix
2022-12-13 13:21:48 <uuidNuniq> grub has gzip module, so grub is fine with "vmlinuz-edge: gzip compressed data"
2022-12-13 13:22:07 <minimal> uuidNuniq: the kernel is self-decompressing, which is different from the bootloader having to do decompression
2022-12-13 13:22:28 <minimal> one thing you haven't indicated regarding your SBC, are you using armv7 or aarch64?
2022-12-13 13:22:46 <uuidNuniq> anyway, i'll try vmlinuz-edge first, then try Image (uncompressed from vmlinuz-edge)
2022-12-13 13:23:20 <uuidNuniq> aarch64 amlogic. only linux-edge has it supported 
2022-12-13 13:23:59 <minimal> ah, now that is an issue - AFAIK aarch64 kernels do not support self-decompression
2022-12-13 13:24:09 <minimal> nothing to do with whichever bootloader is used
2022-12-13 13:25:34 <minimal> so yes for aarch64 either you'll need to provide the kernel compressed in a way that the bootloader supports decompressing, or supply an uncompressed kernel to it
2022-12-13 13:30:18 <uuidNuniq> yeah, minimal i read the documents, looks like i can create a package, which has a triggers="/boot" and a shell script. when the /boot changes, the script is run. thank you for the idea 
2022-12-13 13:30:57 <minimal> uuidNuniq: which u-boot are you using? I don't see an Alpine package for amlogic
2022-12-13 13:32:09 <uuidNuniq> linux-edge has amlogic support in .config. but linux-lts does not. CONFIG_ARCH_MESON is for amlogic. i am using a chainloaded uboot from armbian. 
2022-12-13 13:32:45 <uuidNuniq> the board has a stock uboot in emmc, and then i chainload a new uboot in order to boot a mainline kernel. 
2022-12-13 13:32:45 <minimal> uuidNuniq: I didn't ask about the kernel package, I asked about the u-boot package
2022-12-13 13:34:11 <minimal> where did the "chainloaded uboot" come from?
2022-12-13 13:36:07 <uuidNuniq> from armbian. it is a binary file. 635KB in size
2022-12-13 13:36:55 <minimal> uuidNuniq: so did it have CONFIG_GZIP set when it was built? if not then it won't be able to handle gzip compressed kernel files
2022-12-13 13:37:21 <uuidNuniq> if fatload usb 1 0x1000000 u-boot.ext; then go 0x1000000; fi;
2022-12-13 13:37:32 <uuidNuniq> something like this to chainload 
2022-12-13 13:38:31 <uuidNuniq> less /lib/apk/db/triggers
2022-12-13 13:38:57 <minimal> uuidNuniq: so did it have CONFIG_GZIP set when it was built? if not then it won't be able to handle gzip compressed kernel file
2022-12-13 13:39:20 <uuidNuniq> i don't know how it was built :P i just know it works to boot Linus kernel :)
2022-12-13 13:43:11 <uuidNuniq> strings u-boot.ext |& grep -E 'gzip|gunzip' , looks like it has gzip mentioned 
2022-12-13 13:45:47 <minimal> uuidNuniq: as you are trying this headless then you can't tell whether a failure to boot is a u-boot issue or a Alpine initramfs' init related issue
2022-12-13 13:46:10 <minimal> e.g. do you have necessary kernel modules in the initramfs being used?
2022-12-13 13:46:38 <minimal> are you passing the necessary cmdline options via the bootloader?
2022-12-13 14:17:20 <uuidNuniq> can i ignore linux-firmware-* when i install linux-edge?
2022-12-13 14:17:40 <uuidNuniq> i really don't need those firmwares 
2022-12-13 14:18:02 <panekj> apk add linux-firmware-none
2022-12-13 14:18:25 <uuidNuniq> thanks!
2022-12-13 15:48:15 <bleb> I'm trying to install grub via alpine installation media
2022-12-13 15:48:17 <uuidNuniq> if i want add addtional modules to initramfs, can i create similar files in  /etc/mkinitfs/features.d/ ?
2022-12-13 15:48:56 <bleb> I have these partitions for my installation: http://ix.io/4iwj
2022-12-13 15:49:05 <uuidNuniq> and then include the file name in /etc/mkinitfs/mkinitfs.conf? 
2022-12-13 15:49:11 <bleb> using gpt partitions and efi
2022-12-13 15:49:28 <bleb> if i run "grub-install --target=x86_64-efi --efi-directory=/mnt/boot" I get grub-install: error: failed to get canonical path of `tmpfs'
2022-12-13 15:49:56 <bleb> any ideas what I might try?  setup-alpine did not work because I am using a custom partition scheme
2022-12-13 16:00:48 <minimal> uuidNuniq: yes, you would then need to run mkinitfs (perhaps in a chroot) to regenerate the initramfs file
2022-12-13 16:05:41 <minimal> bleb: you haven't told grub-install where to put grub files
2022-12-13 16:06:06 <minimal> so it will try to put them in /boot/grub
2022-12-13 16:06:21 <minimal> which is not on the NVME drive
2022-12-13 16:09:10 <uuidNuniq> minimal: i added dwc2 dwc3 usb modules and alpine boots! 
2022-12-13 16:09:28 <uuidNuniq> and i am now connected with a monitor
2022-12-13 16:10:04 <uuidNuniq> however, eth0 can't up, dmesh says can't attach to a phy 
2022-12-13 16:10:25 <uuidNuniq> i am comparing modules and see what phy module i need to modprobe
2022-12-13 16:13:01 <minimal> uuidNuniq: perhaps you should check the datasheet/specifications for the SBC you are using?
2022-12-13 16:14:54 <bleb> minimal: ok so I added --boot-directory=/mnt/boot, and now I'm getting grub-install: error: efibootmgr: not found
2022-12-13 16:15:14 <minimal> bleb: ok, and?
2022-12-13 16:16:20 <bleb> no errors after I installed efibootmgr
2022-12-13 16:16:23 <bleb> let's see if it boots
2022-12-13 16:16:42 <Ermine> bleb: either install efibootmgr so grub could add a boot entry, or pass a --no-nvram flag.
2022-12-13 16:17:11 <bleb> what does the --no-nvram flag do?
2022-12-13 16:17:23 <bleb> I noticed that was in the install scripts
2022-12-13 16:19:08 <bleb> ok well it's booting now
2022-12-13 16:19:14 <bleb> minimal: tyvm
2022-12-13 16:20:37 <bleb> hmm no man page for grub-install
2022-12-13 16:21:47 <minimal> there is if you install the grub-doc package...
2022-12-13 16:22:24 <bleb> I did add docs
2022-12-13 16:22:29 <bleb> then I added grub-doc
2022-12-13 16:22:32 <bleb> still no...
2022-12-13 16:23:06 <minimal> ah, it's only providing info file
2022-12-13 16:25:26 <bleb> docs metapackage doesn't provide info
2022-12-13 16:26:18 <uuidNuniq> minimal: if i add the phy to initramfs , do i have to add it for kernel runtime?
2022-12-13 16:26:52 <bleb> and with info installed, info: No menu item 'grub-install' in node '(dir)Top'
2022-12-13 16:27:02 <uuidNuniq> like /etc/modprobe.d
2022-12-13 16:27:45 <bleb> well I can read the grub-install man page elsewhere
2022-12-13 16:28:45 <minimal> Grub doesn't seems to provide manpage, so "apk add texinfo" and then do "info grub"
2022-12-13 16:29:40 <bleb> now that I have seen the man page I can rephrase my question from before
2022-12-13 16:30:10 <minimal> uuidNuniq: you don't need to add the network/phy module to initramfs unless you are netbooting, adding it to /etc/modules or in /etc/modprobe.d/ will load it once Alpine rootfs starts
2022-12-13 16:30:36 <bleb> why does the installer pass --no-nvram to grub-install rather than using efibootmgr?
2022-12-13 16:31:19 <bleb> the efibootmgr method seemed to have the same effect
2022-12-13 16:31:44 <bleb> --no-nvram says "don't update the `boot-device'/`Boot*' NVRAM variables"
2022-12-13 16:31:58 <bleb> so presumably efibootmgr updates the `boot-device'/`Boot*' NVRAM variables
2022-12-13 16:32:48 <bleb> but how does the installer create a grub menu entry without updating those NVRAM variables?
2022-12-13 16:33:47 <minimal> bleb: that has nothing to do with Grub menu entries
2022-12-13 16:34:08 <minimal> it is to do with UEFI boot entries
2022-12-13 16:34:43 <bleb> ok, and?
2022-12-13 16:35:27 <minimal> and what exactly?
2022-12-13 16:36:18 <bleb> how does the installer make the installation bootable without adding a boot entry
2022-12-13 16:36:50 <bleb> or does it add a boot entry using some other command that doesn't depend on efibootmgr?
2022-12-13 16:38:56 <minimal> bleb: the installer is only design with Alpine-only boot in mind, it does not handle multi-OS booting
2022-12-13 16:39:06 <minimal> s/design/designed/
2022-12-13 16:39:06 <alpine-meetbot> minimal meant to say: bleb: the installer is only designed with Alpine-only boot in mind, it does not handle multi-OS booting
2022-12-13 16:39:26 <bleb> so single-os booting does not require boot entries?
2022-12-13 16:40:15 <minimal> not if UEFI fallback is setup
2022-12-13 16:41:02 <bleb> so the installer sets it up somewhere?
2022-12-13 16:41:11 <bleb> any command I could grep for in the scripts?
2022-12-13 16:43:09 <wakyct> bleb, when I installed multi-boot recently I just used efibootmgr to set up boot after manual partitioning
2022-12-13 16:43:29 <wakyct> it seemed the most straightforward at the time, this was my first multi install though
2022-12-13 16:43:30 <minimal> bleb: but you're doing a multi-OS install so it's not relevant then
2022-12-13 16:43:42 <bleb> minimal: my install is only single OS currently
2022-12-13 16:44:24 <bleb> so I'd like to understand why the setup scripts didn't work for me
2022-12-13 16:45:52 <minimal> bleb: I'm confused, I thought you were manually setting things up as you were multi-booting
2022-12-13 16:46:04 <minimal> if not, then use setup-alpine or setup-disk
2022-12-13 16:46:21 <bleb> I want to be able to multi-boot in the future
2022-12-13 16:46:44 <bleb> but alpine is the only OS currently
2022-12-13 16:47:12 <minimal> so use setup-alpine/setup-disk now?
2022-12-13 16:47:38 <bleb> well it's installed now
2022-12-13 16:47:44 <bleb> but I was trying that for hours yesterday and today
2022-12-13 16:48:15 <bleb> a normal sys install takes over the entire disk
2022-12-13 16:48:30 <bleb> so I did custom partitioning
2022-12-13 16:48:48 <minimal> bleb: you partition manually and then run setup-disk telling it to use the existing partitioning
2022-12-13 16:48:57 <bleb> yeah
2022-12-13 16:49:02 <bleb> that's what I tried
2022-12-13 16:49:31 <minimal> well did you read info on how to do that? at the very least "setup-disk -h" output?
2022-12-13 16:49:45 <bleb> only the wiki page on dualbooting
2022-12-13 16:50:42 <minimal> many wiki pages are out of date or incorrect
2022-12-13 16:51:05 <bleb> which is why I'd like to understand what went wrong so I can update it
2022-12-13 16:52:26 <bleb> I guess UEFI fallback only works if the boot and OS partitions are the only ones?
2022-12-13 16:53:21 <bleb> I had made an exfat partition for storage, I guess that means UEFI fallback won't work, so you need efibootmgr?
2022-12-13 16:55:13 <minimal> number of partitions has no relationship to fallback
2022-12-13 16:55:59 <bleb> hm
2022-12-13 16:56:02 <bleb> any ideas then
2022-12-13 16:56:35 <minimal> you haven't provided sufficient information to know what happened
2022-12-13 16:57:11 <bleb> I should have tried grub-install with --no-nvram instead of installing efibootmgr
2022-12-13 16:57:15 <bleb> maybe it would have worked
2022-12-13 16:58:36 <minimal> what evidence indicates that?
2022-12-13 16:58:59 <bleb> (a) I didn't try it (b) alpine was the only OS on the disk
2022-12-13 17:01:31 <minimal> that's not evidence, that's just guesswork
2022-12-13 17:02:53 <bleb> what's your point?
2022-12-13 17:03:10 <bleb> grub-install with --no-nvram definitely wouldn't have worked?
2022-12-13 17:03:46 <minimal> my point was partition the disk and then run setup-disk with the relevant options
2022-12-13 17:05:13 <bleb> yesterday I created partitions for boot, os, and storage, and formatted them to fat32, ext4, and exfat
2022-12-13 17:05:44 <bleb> I mounted the OS partition on /mnt and the boot partition on /mnt/boot
2022-12-13 17:05:48 <minimal> you mean ESP, os, and storage
2022-12-13 17:06:40 <bleb> good to know
2022-12-13 17:07:03 <minimal> the vfat partition is a ESP partition as UEFI requires one to boot
2022-12-13 17:07:22 <bleb> I suppose we should s/boot/esp/ the variable names in setup-disk then
2022-12-13 17:07:23 <minimal> it can also be used for other purposes (such as for /boot)
2022-12-13 17:08:25 <bleb> find_partitions() { ... case "$type" in boot) ...
2022-12-13 17:08:36 <minimal> so you actually have a reason for having /boot in the ESP partition rather than as part of rootfs?
2022-12-13 17:08:43 <bleb> that mislead me into thinking it would be called a boot partition
2022-12-13 17:10:15 <bleb> minimal: I did it that way because setup-alpine does it that way
2022-12-13 17:11:22 <bleb> minimal: would you recommend something else?
2022-12-13 17:12:26 <minimal> bleb: I don't believe setup-disk uses the ESP partition for /boot (unless you are using encryption or perhaps one or 2 other scenarios)
2022-12-13 17:16:10 <bleb> the setup-disk help dialog says for sys: "The following partitions will be created on the disk: /boot, / (filesystem root) and swap."
2022-12-13 17:16:40 <minimal> basically for a UEFI system the ESP partition is mounted by setup-disk as /mnt/boot/efi, but if encryption is used then it is mounted as /mnt/boot instead (and therefore it contains both UEFI files and Alpine boot files)
2022-12-13 17:17:23 <bleb> so it's lies?
2022-12-13 17:18:14 <minimal> lies? no, it seems incorrect however, that text doesn't make a distinction between partitioning for BIOS booting and UEFI booting
2022-12-13 17:19:06 <minimal> calling it lies would mean that someone purposely wrote something they knew was incorrect
2022-12-13 17:19:52 <minimal> also you can control whether or not a swap partition is created
2022-12-13 17:21:18 <invoked> are you using btrfs (sorry, not scrolling back)
2022-12-13 17:22:03 <bleb> no
2022-12-13 17:22:14 <minimal> from memory doing "setup-disk -s 0 ...." will prevent a swap partition being created
2022-12-13 17:22:25 <invoked> point being, if you use btrfs you can add a swapfile subvolume
2022-12-13 17:22:32 <invoked> not that i'm advocating btrfs necessarily
2022-12-13 17:22:48 <invoked> but it's more convenient for a full disk encryption scenario
2022-12-13 17:23:09 <bleb> I'm not doing encryption anyway
2022-12-13 17:23:23 <bleb> and I don't need swap I have 32GB ram
2022-12-13 17:23:48 <invoked> you'll need it for suspend to disk, maybe.
2022-12-13 17:24:27 <bleb> hm good point
2022-12-13 17:24:57 <bleb> minimal: can I tell setup-disk to use an existing partition for swap?
2022-12-13 17:40:11 <uuidNuniq> can't fingure out the phy issue
2022-12-13 17:40:35 <uuidNuniq> looks like linux-edge does not have the phy enabled. 
2022-12-13 17:48:47 <minimal> uuidNuniq: do you know what type of phy it actually uses?
2022-12-13 17:50:28 <uuidNuniq> https://0x0.st/onBk.txt
2022-12-13 17:50:32 <uuidNuniq> minimal: 
2022-12-13 17:54:28 <minimal> bleb: I think it finds any pre-existing swap partition
2022-12-13 18:00:07 <uuidNuniq> https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c?h=v6.0.12#n3796
2022-12-13 18:04:05 <uuidNuniq> maybe use udev to replace mdev? i don't know if that has something to do with device auto discovery
2022-12-13 18:05:20 <minimal> uuidNuniq: thaat will make no difference if the driveris not enabled in kernel
2022-12-13 18:05:28 <minimal> so figure that out first
2022-12-13 18:06:12 <uuidNuniq> i think it is enabled if it is realtek phy 
2022-12-13 18:07:25 <uuidNuniq> i have a working archlinux. how the heck do i know what phy is eth0 using? 
2022-12-13 18:09:58 <psykose> lspci -k if it's on pci
2022-12-13 18:10:26 <uuidNuniq> my box does not use pci 
2022-12-13 18:10:53 <uuidNuniq> it uses stuff like mdio gpio... 
2022-12-13 18:15:49 <psykose> ls -l /sys/class/net/eth0/device/driver
2022-12-13 18:17:13 <uuidNuniq> on arch linux : meson8b-dwmac c9410000.ethernet eth0: PHY [stmmac-0:00] driver [Generic PHY] (irq=POLL)
2022-12-13 18:17:29 <uuidNuniq> eth0: configuring for phy/rgmii link mode
2022-12-13 18:17:51 <uuidNuniq> so it is generic_phy
2022-12-13 18:18:02 <psykose> that does not look like ls -l output
2022-12-13 18:18:30 <uuidNuniq> that's dmesg
2022-12-13 18:18:36 <uuidNuniq> here is ls : /sys/class/net/eth0/device/driver -> ../../../../bus/platform/drivers/meson8b-dwmac
2022-12-13 18:20:40 <psykose> sigh, fuckin busybox
2022-12-13 18:20:42 <psykose> add a / on the end
2022-12-13 18:21:21 <bleb> minimal: ok redoing my installation
2022-12-13 18:21:33 <uuidNuniq> https://0x0.st/onMS.txt
2022-12-13 18:21:53 <minimal> bleb: if you are intending to do multi-os later you may want to think about ESP partition size
2022-12-13 18:22:18 <bleb> minimal: the one I made is 512 MiB
2022-12-13 18:22:19 <psykose> dwmac_meson8b is the module
2022-12-13 18:22:21 <bleb> think that will be enough?
2022-12-13 18:22:59 <uuidNuniq> yeah, i know, ip link shows eth0 there , it just it can be brought up 
2022-12-13 18:23:00 <psykose> which is DWMAC_MESON
2022-12-13 18:23:12 <minimal> bleb: it depends on various factors, such as what use any other OS may make of the ESP partition
2022-12-13 18:23:35 <bleb> how should I think about it then
2022-12-13 18:23:54 <bleb> is there a rough upper bound for how much space an OS is likely to use?
2022-12-13 18:23:57 <uuidNuniq> iv modprobe all of them already 
2022-12-13 18:24:43 <ikke> bleb: not really, /usr and /lib may grow depening on what packages / applications you install
2022-12-13 18:24:47 <psykose> it's not enabled in any alpine kernel to begin with
2022-12-13 18:25:09 <ikke> /var may grow due to data being written by applications
2022-12-13 18:25:38 <minimal> ikke: I was referring to use of the ESP partition
2022-12-13 18:26:41 <uuidNuniq> psykose:  CONFIG_DWMAC_MESON=m in https://git.alpinelinux.org/aports/plain/community/linux-edge/config-edge.aarch64?h=3.17-stable
2022-12-13 18:26:48 <minimal> i.e. if you install Windows I don't know how many Mb of stuff it puts in ESP
2022-12-13 18:26:48 <psykose> ah right
2022-12-13 18:26:53 <psykose> looked in the wrong place
2022-12-13 18:27:28 <uuidNuniq> ip link could see lo and eth0 (DOWN)
2022-12-13 18:27:41 <psykose> and what does it do if you make it up
2022-12-13 18:27:41 <uuidNuniq> ip link set eth0 up , then  ... fail. 
2022-12-13 18:28:10 <bleb> so maybe it's safer to make it 1 GiB
2022-12-13 18:28:17 <uuidNuniq> it then say this : https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c?h=v6.0.12#n3796
2022-12-13 18:28:40 <uuidNuniq> i have to look into the source code... to find some solution 
2022-12-13 18:28:46 <psykose> fails to initialise
2022-12-13 18:28:49 <psykose> what's in dmesg
2022-12-13 18:29:08 <uuidNuniq> clear'
2022-12-13 18:29:33 <minimal> uuidNuniq: you said it had a Relatek phy, is the module for that loaded?
2022-12-13 18:29:41 <psykose> no clue then :)
2022-12-13 18:30:44 <uuidNuniq> loaded
2022-12-13 18:30:46 <minimal> uuidNuniq: look at the various modules load on Arch (where you said network interface works)
2022-12-13 18:31:33 <uuidNuniq> the dmesg log: https://0x0.st/onM1.txt
2022-12-13 18:31:34 <minimal> basically are any dependant modules missing
2022-12-13 18:32:41 <uuidNuniq> minimal: arch is custom build kernel. i have many options Y. 
2022-12-13 18:32:56 <uuidNuniq> here is the lsmod on arch : https://0x0.st/onMj.txt
2022-12-13 18:33:12 <uuidNuniq> sorted 
2022-12-13 18:35:36 <minimal> you chopped out the field that show inter-module deps...
2022-12-13 18:37:22 <uuidNuniq> minimal: https://0x0.st/onMp.txt
2022-12-13 18:40:17 <uuidNuniq> pcs_xpcs?
2022-12-13 18:42:40 <uuidNuniq> no, that's loaded too. anything *stm* *dwm* are loaded on alpine. 
2022-12-13 18:44:51 <minimal> uuidNuniq: phylink, stmmac, pcs_xpcs, stmmac_platform, dwmac_meson8b, dwmac_generic all loaded on Alpine? 
2022-12-13 18:50:15 <uuidNuniq> yeah. all loaded minimal 
2022-12-13 18:51:45 <minimal> and what about the DTB you're using? is it correct/up-to-date for that kernel?
2022-12-13 18:52:00 <uuidNuniq> almost there .... but then this non-sense phy issue. 
2022-12-13 18:52:03 <minimal> did it come with the kernel or did you get it somewhere else?
2022-12-13 18:52:20 <uuidNuniq> the dtb is build against 6.0.12
2022-12-13 18:52:29 <uuidNuniq> with symbols enabled 
2022-12-13 18:52:52 <uuidNuniq> `\ttps://0x0.st/onME.sh
2022-12-13 18:52:57 <uuidNuniq> https://0x0.st/onME.sh
2022-12-13 18:53:05 <minimal> uuidNuniq: it came with kernel 6.0.12 or from somewhere else?
2022-12-13 18:53:19 <uuidNuniq> i build it 
2022-12-13 18:53:27 <uuidNuniq> with the above script 
2022-12-13 18:53:49 <uuidNuniq> let me try the official dtb 
2022-12-13 18:53:52 <uuidNuniq> wait a moment 
2022-12-13 18:57:27 <jh81> ikke alright then :) is this of any use? https://pastebin.com/6RJb7YkQ
2022-12-13 18:57:55 <bleb> OK I made my esp partition 4 GiB
2022-12-13 18:58:02 <uuidNuniq> same thing, with /boot/dtbs-edge/amlogic/meson-gxl-s905d-phicomm-n1.dtb
2022-12-13 18:59:31 <bleb> minimal: now would you have me mount the OS partition on /mnt, then mkdir -p /mnt/boot/efi and mount the ESP partition on /mnt/boot/efi, then run setup-alpine?
2022-12-13 18:59:54 <bleb> then answer none for the disk questions, and run setup-disk -m sys /mnt?
2022-12-13 19:02:39 <uuidNuniq> with or without dts symbols, samething. 
2022-12-13 19:05:18 <minimal> bleb: BTW is this MBR or GPT partitioned?
2022-12-13 19:05:22 <bleb> GPT
2022-12-13 19:05:49 <minimal> ok, so that should in theory work
2022-12-13 19:06:09 <minimal> uuidNuniq: is Arch using the same kernel version?
2022-12-13 19:06:22 <bleb> tangent: any idea why setup-interfaces would show wlan0 when run on the installed OS, but not on the installation media?
2022-12-13 19:06:57 <uuidNuniq> minimal: arch works on all kernel version from 5.15.x to 6.1 with my dtb built from my script 
2022-12-13 19:07:25 <uuidNuniq> build.sh mainline 5.15.80 or 6.1 
2022-12-13 19:07:37 <minimal> uuidNuniq: and you've checked that Arch are not applying any patches to the kernels they build?
2022-12-13 19:08:03 <uuidNuniq> i made the script because i don't want to download the kernel tree. with the script, only few KB files needed 
2022-12-13 19:08:41 <uuidNuniq> minimal: no patch, just the patch-kernel_ver.xz 
2022-12-13 19:09:41 <minimal> is there firmware for this SBC? perhapsthe versions diff between Arch and Alpine
2022-12-13 19:10:12 <uuidNuniq> no firmware. i don't install any linux-firmware. 
2022-12-13 19:10:42 <uuidNuniq> just a stable kernel from dear Linus 
2022-12-13 19:10:53 <minimal> no idea, compare the Arch and Alpine kernel configs...
2022-12-13 19:11:23 <uuidNuniq> yeah, i'll do that next. a lot of diff. 
2022-12-13 19:11:45 <uuidNuniq> because i stripped lots of options. 
2022-12-13 19:11:55 <uuidNuniq> stuff i don't need are N
2022-12-13 19:12:17 <bleb> minimal: it did work!
2022-12-13 19:12:22 <bleb> thanks for all your help
2022-12-13 19:12:35 <bleb> no need for efibootmgr
2022-12-13 19:12:42 <bleb> for now at least
2022-12-13 19:13:17 <ikke> jh81: can you remind me of the context?
2022-12-13 19:13:36 <jh81> yes, i was trying to figure out why my work monitor wouldnt work over usb-c
2022-12-13 19:13:56 <bleb> upon booting to the installation, setup-interfaces shows wlan0.  I unplugged the ethernet cable before rebooting, so it couldn't have installed anything that wasn't already there
2022-12-13 19:14:13 <bleb> so bit of a mystery why wlan0 didn't show up in the installer
2022-12-13 19:14:23 <ikke> jh81: don't recalling discussing that specifically
2022-12-13 19:15:07 <jh81> crap! wasnt you, it was minimal who said i shoul run insmod on arch when connected ;-)
2022-12-13 19:15:10 <jh81> sorry, my bad
2022-12-13 19:15:13 <ikke> np
2022-12-13 19:15:17 <minimal> yes it was me
2022-12-13 19:15:46 <jh81> i spend the night tossing and turning over a not functioning certificate authority server.. im not clear headed today it seems
2022-12-13 19:17:01 <jh81> in any case, this pastebin is what lsmod did on arch :)
2022-12-13 19:17:47 <minimal> so it's AMD graphics based
2022-12-13 19:17:59 <jh81> yupyup, amd radeon noir
2022-12-13 19:18:18 <jh81> i have installed the amd drivers for alpine per the wiki
2022-12-13 19:20:07 <jh81> in a few mins i can boot into alpine, see what lsmod says there
2022-12-13 19:23:20 <minimal> so there are modules loaded related to the USB-C: ucsi_acpi, typec_ucsi, typec, roles
2022-12-13 19:23:36 <minimal> jh81: does any useful appear in "dmesg" regarding the usb-c display?
2022-12-13 19:29:06 <jh81> nothing :| 
2022-12-13 19:29:20 <jh81> i see a lot of amdgpu initialisation messages, and thats it
2022-12-13 19:29:56 <jh81> nothing that says "usb connected/disconnected" or anything that could lead me to the monitor
2022-12-13 19:32:28 <jh81> i can see messages when i plug in the monitor, that it finds the integrated camera and usb hub
2022-12-13 19:33:17 <jh81> this is the only one that got me thinking its the monitor: [64810.451412] usb 1-1: new high-speed USB device number 4 using xhci_hcd
2022-12-13 19:33:49 <minimal> jh81: what about your desktop config tools (like Monitor Preferences)? I assume (in Arch) they show the monitor as well as the laptop screen?
2022-12-13 19:34:21 <jh81> it does
2022-12-13 19:34:35 <minimal> as DP-1 or similar?
2022-12-13 19:34:42 <jh81> but that just displays the monitor model number
2022-12-13 19:35:48 <jh81> no, just the manifactoring and model
2022-12-13 19:36:07 <minimal> so no indication if its HDMI / DisplayPort etc?
2022-12-13 19:36:18 <jh81> i'll first see if i can import those modules you mentioned above
2022-12-13 19:36:21 <jh81> nope
2022-12-13 19:36:37 <minimal> Mate's Monitor Preferences indicates a laptop built-in screen as eDP-1 for example
2022-12-13 19:37:22 <minimal> any when Alpine is booted does the same desktop tool show the external display?
2022-12-13 19:37:40 <jh81> well, i suppose i could boot into a livecd :) but that will have to wait tyill after xmas im afraid
2022-12-13 23:47:11 <uuidNuniq> continue the work: i tried to use udev as devd, black screen  
2022-12-13 23:47:43 <uuidNuniq> is there configuration needed?
2022-12-13 23:47:55 <psykose> did you add yourself to the groups
2022-12-13 23:59:24 <uuidNuniq> setup-devd udev 
2022-12-13 23:59:57 <uuidNuniq> after that, i checked with rc-update, there are 4 udev* services in the list. none mdev* hwdrivers
2022-12-14 00:02:44 <uuidNuniq> with mdev or mdev i am able to see the screen via HDMI. i tried udev to see if it is possible to bring eth0 up. turns out it's black screen. 
2022-12-14 00:02:53 <uuidNuniq> mdev or mdevd
2022-12-14 00:05:42 <uuidNuniq> some GOOD NEWS, ripgrep on alpine search against /sys does not have kernel oops, smooth and stable. 
2022-12-14 00:06:21 <uuidNuniq> while ripgrep on archlinux/debian bookworm got kernel oops very fast. 
2022-12-14 00:08:49 <psykose> did you add yourself to the video/input groups tho
2022-12-14 00:09:22 <uuidNuniq> what.... the... do i have to add myself to that group? i did not know that 
2022-12-14 00:09:42 <uuidNuniq> i am on my own group, the same name as my user name 
2022-12-14 00:15:32 <uuidNuniq> psykose: i am not using desktop. headless only. do i have to add myself to video/input groups?
2022-12-14 00:16:06 <psykose> how is it headless if you see a black screen
2022-12-14 00:16:31 <psykose> unless you mean something else
2022-12-14 00:17:04 <psykose> but no, only if you actually want to start a wm
2022-12-14 00:17:59 <uuidNuniq> command line interface, i mean 
2022-12-14 00:18:08 <uuidNuniq> or console 
2022-12-14 00:25:48 <minimal> uuidNuniq: why would you expect switching from mdev to eudev to have any effect on the ethernet drivers?
2022-12-14 00:45:48 <uuidNuniq> i dont know. just try this and that
2022-12-14 00:46:17 <uuidNuniq> next try: zcat initramfs  | cpio -idmv
2022-12-14 00:48:54 <minimal> uuidNuniq: randomly trying things isn't a realistic way to solve a problem
2022-12-14 00:50:20 <uuidNuniq> oh no it has nothing to do with initrd
2022-12-14 00:55:06 <minimal> uuidNuniq: your graphics chip/driver is lima, right?
2022-12-14 00:56:55 <minimal> or is it meson? your Arch lsmod showed both lima and meson_drm loaded
2022-12-14 00:57:39 <uuidNuniq> that's video
2022-12-14 00:57:44 <minimal> as linux-edge doesn't appear to have the lima DRM module built...
2022-12-14 00:57:56 <minimal> well you said you're getting black screen - that's video related
2022-12-14 00:58:18 <uuidNuniq> black on udev only
2022-12-14 00:58:41 <uuidNuniq> mdev{,d} fine
2022-12-14 00:59:17 <minimal> the drm module is generally loaded by initramfs' init, before the rootfs is mounted (and therefore before eudev would be run)
2022-12-14 01:00:00 <minimal> by CONFIG_DRM_LIMA is not set in the linux-edge aarch64 config...
2022-12-14 01:00:08 <minimal> but CONFIG_DRM_MESON=m
2022-12-14 01:00:20 <minimal> which is why I asked which should it be using for DRM? meson or lima
2022-12-14 01:01:21 <minimal> as long as the relevant module is in the initramfs then its init should load it
2022-12-14 01:04:07 <minimal> uuidNuniq: let me guess, you have "quiet" in your u-boot cmdline setting? try removing it
2022-12-14 08:59:01 <uuidNuniq> no quiet
2022-12-14 08:59:09 <uuidNuniq> why is there a boot in /boot?
2022-12-14 08:59:20 <uuidNuniq> can i unlink it?
2022-12-14 09:24:38 <dwfreed> if /boot is a separate partition, having a boot -> . symlink inside it allows paths to remain the same regardless of whether you're in Linux or the bootloader
2022-12-14 11:18:03 <uuidNuniq> thanks!
2022-12-14 11:18:18 <uuidNuniq> tried all methods, still can't bring eth0 up. 
2022-12-14 14:30:14 <nakedpotato> Hello!
2022-12-14 14:30:48 <nakedpotato> What would be the best way to wait for an interface?
2022-12-14 14:30:56 <nakedpotato> in an init.d file I mean
2022-12-14 14:31:31 <nakedpotato> I got a service that I want to expose on the docker0 interface
2022-12-14 14:31:52 <nakedpotato> adding a dependency on docker is not enough, the interface takes some time to appear
2022-12-14 14:32:02 <ncopa> do you want the service to stop with docker0 disappears?
2022-12-14 14:32:18 <ncopa> you can add rc_depend=docker in /etc/conf.d/your-service
2022-12-14 14:32:50 <nakedpotato> ncopa: not enough unfortunately
2022-12-14 14:34:40 <minimal> nakedpototo: you want a non-docker service to listen on the docker0 interface?
2022-12-14 14:34:51 <nakedpotato> indeed
2022-12-14 14:35:28 <minimal> nakedpototo: strange! however perhaps bcnm-waitif from the bcnm package is what you need?
2022-12-14 14:35:37 <nakedpotato> I'm running a SMTP server on the host and I also want it to be available inside the containers
2022-12-14 14:36:26 <nakedpotato> though
2022-12-14 14:36:36 <nakedpotato> I guess I could add an MX record
2022-12-14 14:37:12 <nakedpotato> oh no, that would expose the server on the network which is what I was trying to avoid
2022-12-14 14:37:21 <nakedpotato> I want it as a local relay
2022-12-14 14:37:52 <minimal> nakedpotato: did you read my last comment?
2022-12-14 14:38:01 <nakedpotato> yes, I'm trying it
2022-12-14 14:38:19 <nakedpotato> sorry for all the rambling :P
2022-12-14 14:38:52 <Habbie> does the smtp server bind to any other interfaces besides docker0?
2022-12-14 14:39:05 <nakedpotato> just localhost
2022-12-14 14:39:40 <Habbie> ok
2022-12-14 14:39:48 <Habbie> because anything non-localhost would likely also work from inside docker
2022-12-14 14:40:18 <Habbie> you could even add a non-127 alias to the lo interface for that purpose
2022-12-14 14:40:56 <nakedpotato> what do you mean?
2022-12-14 14:40:57 <minimal> Habbie: doesn't that depend on the docker network mode being used?
2022-12-14 14:41:26 <Habbie> minimal, right, i've only used host and the default, bridge i think? in those two cases it works
2022-12-14 14:41:30 <Habbie> in host mode 127.0.0.1 just works anyway
2022-12-14 14:42:03 <Habbie> nakedpotato, pick an IP, let's say 192.0.2.1 (but don't use that one). add it to the lo interface. bind SMTP server to it. tell docker containers to point their mail there.
2022-12-14 14:43:12 <ncopa> <nakedpotato> ncopa: not enough unfortunately
2022-12-14 14:43:15 <ncopa> why not enough?
2022-12-14 14:43:30 <ncopa> does docker not start the docker0 interface quick enough?
2022-12-14 14:43:47 <nakedpotato> looks like it doesn't unfortunately
2022-12-14 14:44:44 <ncopa> I havent used bcnm-waitif, but it does sound like it could be used here
2022-12-14 14:46:38 <nakedpotato> indeed
2022-12-14 14:46:52 <ncopa> might be handy with a waitif "service" that simply blocks til the configured iface shows up
2022-12-14 14:46:57 <nakedpotato> it's not usable as a dependency though
2022-12-14 14:46:59 <nakedpotato> yeah
2022-12-14 14:47:19 <ncopa> symlink waitif.docker0  -> waitif
2022-12-14 14:48:03 <ncopa> I mean, that does not exist yet but would be nice
2022-12-14 14:48:16 <ncopa> and shouldnt be too difficult to implement
2022-12-14 14:50:16 <ncopa> but as Habbie commented, you could also add an ip address to a dummy0 interface for example
2022-12-14 15:07:07 <nakedpotato> I wrote a little init.d script that sleeps till /sys/class/net/docker0 exists
2022-12-14 15:07:11 <nakedpotato> seems to work fine for now
2022-12-14 15:07:56 <minimal> nakedpotato: bcnm-waitif didn't work for you?
2022-12-14 15:09:08 <nakedpotato> minimal: that would have also worked fine I think but I dislike dependencies :P
2022-12-14 15:10:12 <minimal> that would give you wait options, i.e. whether when the interface exists, whether when it is up etc
2022-12-14 15:10:49 <invoked> if you don't want that to be brittle you'll want to add a test that the interface is actually working. maybe bcnm-waitif does that for you
2022-12-14 15:20:02 <jh81> 5 restarts, and all 5 audio worked :D
2022-12-14 16:05:05 <jh81> i managed to get rid of the "software renderer in use" notification by installing a bunch of ATI and MESA and VULKAN packages :| now i need to figure out what exactly i did
2022-12-14 17:05:54 <scorpion2185[m]1> can I install alpine from another os? and make a dual boot
2022-12-14 17:14:09 <jhjacobs[m]> Yes :)
2022-12-14 17:48:57 <uuidNuniq> i checked /etc/init.d/modules, the modprobe form is : modprobe name args, what if there a more than one args? 
2022-12-14 17:49:29 <uuidNuniq> modprobe stmmac phyaddr=0,debug=16?
2022-12-14 17:52:16 <psykose> it's not one argument
2022-12-14 17:52:21 <psykose> it's all of them
2022-12-14 17:55:14 <uuidNuniq> psykose: while read module args ; do modprobe -q $module $args
2022-12-14 17:55:22 <psykose> yeah and
2022-12-14 17:56:04 <uuidNuniq> so, how should i write /etc/modules?  with args?
2022-12-14 17:56:18 <psykose> module x=1 y=2
2022-12-14 17:56:32 <uuidNuniq> e.g. i want to load stmmac with 2 args : phyaddr=0 debug=16 
2022-12-14 17:56:39 <psykose> then pass phyaddr=0 debug=16
2022-12-14 17:57:27 <uuidNuniq> as i understand , while read module args, will only read the first 2 space separated item?
2022-12-14 17:57:37 <psykose> have you considered
2022-12-14 17:57:40 <psykose> running the literal shell code
2022-12-14 17:57:43 <psykose> and seeing what it does>
2022-12-14 17:57:46 <psykose> serious question
2022-12-14 17:58:24 <psykose> (no, it passes all the args)
2022-12-14 18:00:46 <uuidNuniq>  the first word is assigned to the first name,  the second  word  to the  second  name, and so on.
2022-12-14 18:01:15 <uuidNuniq> i don't need to run the code, because modprobe does nothing verbose. 
2022-12-14 18:01:26 <uuidNuniq> plus, the script use modprobe -q 
2022-12-14 18:01:38 <psykose> ?
2022-12-14 18:02:20 <psykose> https://img.ayaya.dev/UVPoy8gIKd2a
2022-12-14 18:05:56 <uuidNuniq> man bash: If there are more words than names, the remaining words and their interveing delimiters are assigned to the last name
2022-12-14 18:06:08 <uuidNuniq> psykose: my fault, i did not read the man carefully. i am sorry 
2022-12-14 18:13:13 <uuidNuniq> i am trying to assign a address to stmmac 
2022-12-14 18:13:37 <uuidNuniq> hope this time it is able to bring eth0 up
2022-12-14 18:15:31 <macmpi> Hi, I need to run a program upon udev rule. That program may be long-enouh for udev to try to kill it... So I guess I'd need to run as an OpenRC service. Is there a trick to pass a parameter to OpenRC service (checked the agetty trick, but needs to handle symlink to fake instance, etc...)?
2022-12-14 18:18:20 <macmpi> (sth like the @ in systemd)
2022-12-14 18:26:48 <jh81> so, on my alpine install, i have wifi.. the KDE systray shows a network icon as not connected, how to fix this?
2022-12-14 18:31:42 <scorpion2185[m]1> how do i install it from another os?
2022-12-14 18:34:29 <uuidNuniq> no luck. 
2022-12-14 18:36:42 <ikke> scorpion2185[m]1: We don't have some kind of installer that you can run directly in another OS
2022-12-14 18:37:03 <ikke> scorpion2185[m]1: we only have a live iso that you can boot into
2022-12-14 18:37:39 <scorpion2185[m]1> somebody said yes before when i asked.
2022-12-14 18:38:14 <scorpion2185[m]1> the wiki for lvm/luks states ask an expert for dual boot
2022-12-14 18:38:45 <scorpion2185[m]1> can i prepare the partitions from anotheros right?
2022-12-14 18:40:16 <psykose> only if the other os can mount and partition the drives correctly
2022-12-14 18:40:29 <psykose> i am not sure if that is possible on windows
2022-12-14 18:41:03 <psykose> same for macos i guess
2022-12-14 18:41:18 <psykose> so unless the other os is some other linux distro or a bsd that can do it, not in that way
2022-12-14 18:41:19 <scorpion2185[m]1> linux.
2022-12-14 18:41:23 <psykose> shrug
2022-12-14 18:41:30 <psykose> you never give any info at all about anything :p
2022-12-14 18:41:42 <scorpion2185[m]1> any calamares installer?
2022-12-14 18:41:49 <psykose> nope
2022-12-14 19:02:35 <jh81> scorpion2185[m]1 you asked if it was possible to install alpine from another os, you didnt specify what os :) so thats why i said yes.
2022-12-14 19:03:03 <jh81> And you didnt specify you wanted a graphical gui either :) 
2022-12-14 19:03:50 <jh81>  I installed Alpine from a Debian os and an Arch os. :) that qualifies as another os
2022-12-14 19:03:57 <Ermine> As one can notice, Alpine images do not contain any graphic programs.
2022-12-14 19:04:16 <jh81> No, but it would be nice if it could :) 
2022-12-14 19:04:35 <jh81> there’s definatly a place for Alpine Desktop in the world :) 
2022-12-14 19:04:50 <minimal> Ermine: but do graphic images contain any Alpine programs? ;-)
2022-12-14 19:06:14 <scorpion2185[m]1> jh81: can I install it from Debian ? but to another partition
2022-12-14 19:06:30 <jh81> Yes :) i did that too
2022-12-14 19:06:48 <jh81> The Alpine wiki has articles on how to do that as well :)
2022-12-14 19:06:52 <Ermine> minimal: maybe this Alps image from the alpinelinux.org contain something?
2022-12-14 19:07:18 <scorpion2185[m]1> what wiki? I looked some
2022-12-14 19:07:37 <jh81> Hang on.. 
2022-12-14 19:07:45 <jh81> i’m traveling, so a bit slow
2022-12-14 19:10:08 <jh81> But what are you trying to do?
2022-12-14 19:18:09 <jh81> My apologies, it seems it isnt on the wiki either. There are several articles you can read to figure out how to do it though, especially this one: https://wiki.alpinelinux.org/wiki/Install_Alpine_on_Amazon_EC2 and this one: https://wiki.alpinelinux.org/wiki/Setting_up_disks_manually
2022-12-14 19:18:45 <jh81> basicly its creating a partition, copying the minirootfs to it, chrooting it for setup, then modify the durrent bootload to include an entry, and reboot
2022-12-14 19:20:41 <jh81> Alternative you could use dd to write the normal iso to the partition, then boot from that (after modifying the bootloader) and use the setup routine
2022-12-14 19:21:00 <jh81> but ive only did that once :)
2022-12-14 19:21:56 <jh81> I should write an article about that :|
2022-12-14 19:21:58 <scorpion2185[m]1> I am trying to dual boot , luks+lvm
2022-12-14 19:22:27 <scorpion2185[m]1> can I safely encrypt /boot with luks2?
2022-12-14 19:22:35 <scorpion2185[m]1> "Note: The method is still experimental and you may lose your access to you OS at the next OS update"
2022-12-14 19:22:40 <jh81> That i dont know.. 
2022-12-14 19:22:44 <minimal> scorpion2185[m]: no
2022-12-14 19:23:10 <minimal> grub does not work with LUKS v2 encrypted /boot
2022-12-14 19:24:00 <minimal> it is "supposed" to work as of Grub 2.06 (more than 1 year ago) but in reality there are issues that prevent it working, some (but not all) of which are fixed in Grub's git repo
2022-12-14 19:25:22 <jh81> i know that refind can do it, but i havent done anything with encryption yet at all :(
2022-12-14 19:25:36 <scorpion2185[m]1> jh81: are you going to make a wiki?
2022-12-14 19:26:18 <jh81> Yes, but not today :)
2022-12-14 19:26:37 <minimal> it would be a long and complex Wiki article to try and cover all aspects of dual/multi-booting, e.g. for BIOS booting, for UEFI booting, etc
2022-12-14 19:26:53 <jh81> and there’s that
2022-12-14 19:27:23 <scorpion2185[m]1> nowadays BIOS is not very common
2022-12-14 19:27:39 <minimal> it's common if you have BIOS machines
2022-12-14 19:27:49 <jh81> actually, Alpine is a perfect os for such old systems ;-)
2022-12-14 19:27:55 <minimal> or are creating BIOS-based VMs
2022-12-14 19:28:01 <scorpion2185[m]1> I was looking this https://wiki.alpinelinux.org/wiki/LVM_on_LUKS. I can live with /boot not on luks
2022-12-14 19:29:08 <jh81> Still, i could write a wiki article for what i did on our systems :) anything is better then nothing
2022-12-14 19:30:52 <scorpion2185[m]1> yes then using other wikis one can install it from another os
2022-12-14 19:31:00 <minimal> on a single boot machine installing Alpine doesn't carry any risks, however for dual/multi-boot there's always a risk that during Alpine installation something could happen that could impact on one (or more) of the already installed/working OSes
2022-12-14 19:31:32 <jh81> True, but thats why we create wiki articles with big fat warnings on top :) 
2022-12-14 19:32:31 <minimal> jh81: wouldn't necessarily stop someone later complaining that "Alpine destroyed my Windows files!" etc unfortunately
2022-12-14 19:33:13 <jh81> Yes, but with that onlook you better stop Alpine all together. Because there will *always* be someone who complains
2022-12-14 19:34:13 <jh81> Look at me complaining about there not beeing a graphical livecd/setup :)
2022-12-14 19:34:29 <minimal> jh81: nope, the ISO Alpine-based installer is designed for Alpine to be the only OS installed so there's no issues with that
2022-12-14 19:35:30 <jh81> Yeah well.. i have long since learned that whatever your intended use case is. People will always find other ways to use it. 
2022-12-14 19:35:39 <minimal> its more the complication of multi-OS setups I'm referring to - resizing existing partitions and filesystems, modifying bootloader configuration etc that carries a greater risk
2022-12-14 19:36:22 <jh81> Besides, i would think anyone who wants to tey that (or honestly, Alpine in general) has at least the knowledge to make backuos. 
2022-12-14 19:36:48 <jh81> *try and * backups
2022-12-14 19:37:08 <jh81> Alpine is not like Ubuntu, its not for first time linux users anyway
2022-12-14 19:37:27 <scorpion2185[m]1> can you ping when you do the wiki?
2022-12-14 19:37:43 <minimal> jh81: agreed. The problem can be that if you help someone part of the way (e.g. via a Wiki article) then you might get blamed for what happens
2022-12-14 19:37:52 <jh81> sure, remind me this weekend if i havent gotten to it by then
2022-12-14 19:38:16 <jh81> oh, i get blamed for lots of things :) 
2022-12-14 19:39:34 <jh81> and still im not afraid to expand the horizons of Alpine’s use cases :) 
2022-12-14 19:40:09 <jh81> but yeah, i hear you :)
2022-12-14 19:42:39 <minimal> we live in a world where McDonalds had to put "danger! hot" on the sides of their takeaway coffee cups after court cases where they were sued by people for scalds (from memory they'd put the cup between their legs at the drive through then drove off and it spilled)
2022-12-14 19:43:08 <jh81> yeah, its why dryers contain warnings not to put children in them
2022-12-14 19:44:05 <minimal> Hazelnut bar packaging shock: "may contain nuts!"
2022-12-14 19:44:33 <jh81> but like i said, Alpine isnt for first time linux users anyway. So its not so unreasonable to expect a certain level of knowledge
2022-12-14 19:45:36 <Ermine> minimal: Alpine licenses contain no warranty clause, so Alpine cannot be sued :)
2022-12-14 19:45:47 <jh81> and there will always be a Karen who wants to speak to the manager. Regarding of how many warnings or articles you make
2022-12-14 19:50:16 <CompanionCube> minimal: i mean, if you look at the details they deserved to lose
2022-12-14 19:51:32 <minimal> CompanionCube: yeah its been a while since I read about it - something to do with the actual temperature of the drink
2022-12-14 19:52:32 <jh81> And in the end, we’ll die anyway. So i choose to worry less these days :)
2022-12-14 19:54:01 <Ermine> Y O L O
2022-12-14 19:56:31 <jh81> Well, there’s that :) and the fact that ive been a caregiver for my parents since i was 12. I learned very you g that you can think of every possible outcome, and plan for every possible outcome, and it may still go wrong. Such is the nature of working with humans. That doesnt mean you should give up and leave it there :)
2022-12-14 19:58:36 <scorpion2185[m]1> what folders do you have on EFI , in /boot/efi/EFI/? If you run only alpine
2022-12-14 19:59:14 <jh81> You made backups, right? ;-)
2022-12-14 19:59:56 <jh81> You understand that whatever we say, and whatever you do, we can not be sued or otherwise held responsible for whatever you type on your commandline? ;-) scorpion2185[m]1
2022-12-14 20:01:21 <scorpion2185[m]1> jh81: dm if you want
2022-12-14 20:01:25 <jh81> Look here: https://wiki.alpinelinux.org/wiki/Bootloaders
2022-12-14 20:02:09 <scorpion2185[m]1> I have that wiki opened, but I am cleaning the EFI part since I am reinstalling it
2022-12-14 20:02:41 <jh81> Whats in that dir depends, amongst other things, what your bootloader is / will be
2022-12-14 20:02:42 <scorpion2185[m]1> it seems that there is a problem with dm here on gomuks
2022-12-14 20:03:34 <jh81> I have found refind to be the more easier bootloader out there
2022-12-14 20:03:35 <scorpion2185[m]1> grub , I hve this folders: alpine/ boot/   debian/ grub/  
2022-12-14 20:03:42 <scorpion2185[m]1> *these
2022-12-14 20:04:48 <jh81> Then look at the grub section of the aforementioned wiki page to set it up
2022-12-14 20:05:02 <jh81> atm i wouldnt worry about cleaning it up just yet
2022-12-14 20:09:44 <jh81> But if you are in Debian, then it might be easier to just setup refind
2022-12-14 20:10:21 <scorpion2185[m]1> doesn't it need a EFI stub?
2022-12-14 20:10:29 <invoked> opinions being what they are, efistub has been my first choice for the last few years
2022-12-14 20:10:51 <jh81> Well, it seems invoked can help you with efi stub ;-)
2022-12-14 20:11:04 <scorpion2185[m]1> can you send the PM again?
2022-12-14 20:11:43 <invoked> without uefi i would go with syslinux but it's all going uefi, so...
2022-12-14 20:12:26 <jh81> I’m more a graphical man myself :) Apple user since 1743, so im a bit spoiled i suppose
2022-12-14 20:12:27 <invoked> (that includes risc-v going uefi)
2022-12-14 20:12:49 <jh81> risc-v is the next project on my list :)
2022-12-14 20:17:30 <jh81> scorpion2185[m]1 your client is broken, or you have other issues :) 
2022-12-14 20:17:59 <jh81> but since i got home, i’m away for a while :) 
2022-12-14 20:18:12 <jh81> just ping me when you got a descent client
2022-12-14 20:28:09 <invoked> i find that pretty boot managers are a mess (imo) when you look at the code
2022-12-14 20:29:14 <invoked> uefi itself is enough complexity it just seems like it makes sense to cut out the intermediate bootloader.  unless the motherboard's uefi is buggy.
2022-12-14 20:31:57 <invoked> but doing efistub + luks boot is a whole other thing
2022-12-14 20:34:49 <invoked> s/luks boot/secure boot/
2022-12-14 20:34:49 <alpine-meetbot> invoked meant to say: but doing efistub + secure boot is a whole other thing
2022-12-14 20:38:25 <Ermine> Imagine editing EFI variables to edit kernel command line. And trying to invoke boot menu if you want to boot something else.
2022-12-14 20:42:19 <jh81> people who use pretty bootmanagers usually dont use them for the code ;-)
2022-12-14 22:47:46 <uuidNuniq> minimal: psykose : i think i found the reason, i did a diff of kernel config. the alpine's linux-edge aarch64 config is missing : CONFIG_MDIO_BUS_MUX_MMIOREG . then the phy reg can't be passed to the eth0 
2022-12-14 22:48:26 <uuidNuniq> and i can see the module mdio_mux_mmioreg in archlinux's lsmod 
2022-12-14 22:53:29 <uuidNuniq> archlinux: https://0x0.st/onWn.txt , alpine: https://0x0.st/onW5.txt . Note: lines start with # are removed. 
2022-12-14 22:54:54 <uuidNuniq> if the dev could add the option : CONFIG_MDIO_BUS_MUX_MMIOREG , i'd be glad to test, thank you, 
2022-12-14 22:56:24 <uuidNuniq> last commit was Milan P. Stanić, wonder if he/she is here? 
2022-12-14 22:57:05 <Saijin_Naib> That is @mps
2022-12-14 22:57:52 <uuidNuniq> by the way, the issue was: https://0x0.st/onM1.txt 
2022-12-14 23:51:03 <uuidNuniq> btw, i fixed the ripgrep kernel oops issue while doing diff, turns out the arch kernel has lots of DEBUG option on. i turned off all of them. now rg is fine, kernel no crash. 
2022-12-15 03:41:56 <kailanqq[m]> ✌️👍🙏DiscJockey is my name and I'm an admin with my own store link and group link https://t.me/+b7jYjbVzzOVkYjU8... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/hiScyWexabgGLTbPNWiUXsDH>)
2022-12-15 09:07:23 <michalkotyla> Hi, I want to create my custom alpine .iso and add some packages like docker or awall. I tried with https://wiki.alpinelinux.org/wiki/How_to_make_a_custom_ISO_image_with_mkimage#Configuration - I created similar config to that. Is that ok? packages from "apks" variable should be placed in rootfs after installation?
2022-12-15 09:07:48 <michalkotyla> I do not have then in installer image and after installation either
2022-12-15 11:05:30 <liske> michalkotyla: the apkgs are just put into the .iso - they won't be installed automatically
2022-12-15 11:06:40 <michalkotyla> liske: oh, so I missunderstand that guide. Is any other way to create image with an additional preinstalled packages?
2022-12-15 11:06:45 <liske> for diskless setups those apks are also copied by setup-bootable
2022-12-15 11:11:52 <liske> michalkotyla: I don't know, sry
2022-12-15 11:12:10 <michalkotyla> liske: k, thanks for the info
2022-12-15 11:12:43 <liske> but it's an interesting question ;)
2022-12-15 11:16:42 <liske> michalkotyla: are you aware of this section https://wiki.alpinelinux.org/wiki/How_to_make_a_custom_ISO_image_with_mkimage#Making_packages_available_on_boot ?
2022-12-15 11:16:49 <ikke> you'd use an apkovl to have packages installed upon boot
2022-12-15 11:50:36 <michalkotyla> I saw it, but the main goal is to have it installed after alpine installation, after `setup-alpine` - when I boot from system disk, not installer medium
2022-12-15 11:51:46 <ikke> The trick is that packages that /etc/apk/world should be copied over from the live system to the installed system
2022-12-15 11:53:12 <ikke> https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-disk.in#L668
2022-12-15 11:54:09 <michalkotyla> hmm, thanks - it may be something which I need
2022-12-15 11:54:15 <michalkotyla> I will try with that
2022-12-15 11:54:43 <ikke> So you either add a setup script that installs them
2022-12-15 11:54:52 <ikke> or you have an apkovl that makes sure they are installed
2022-12-15 18:32:15 <scorpion2185[m]1> how do i chroot specifying a shell?
2022-12-15 18:32:28 <ikke> chroot path/to/new/root /bin/sh
2022-12-15 19:12:44 <scorpion2185[m]1> how do i make /etc/conf.d/dmcrypt?
2022-12-15 19:13:05 <scorpion2185[m]1> if i have to
2022-12-15 19:13:26 <scorpion2185[m]1> installing lvm on luks using the miniroot
2022-12-15 19:24:31 <scorpion2185[m]1> does someone with luks have that file?
2022-12-15 19:29:51 <scorpion2185[m]1> how can i install the complete base system ?
2022-12-15 19:33:46 <jh81> How about focussing on a simple install first before you go medling with LUKS and all?
2022-12-15 19:40:56 <scorpion2185[m]1> i already did that but with the iso
2022-12-15 22:50:42 <ymcrp> It appears Alpine somehow breaks 'compress-force' mount option for Btrfs so it falls back to just 'compress'. I've checked ratios with the same files for both options on Debian. Alpine behaves like Debian with 'compress' in both cases. But if one manually start 'btrfs filesystem defragment' with specified compression algorithm (so it won't chech mount options) then file size ratio will become just like with 'compress-force' on Debian. But of course 
2022-12-15 22:50:42 <ymcrp> that is not a solution so can anyone file a bug report? I can't do it right now
2022-12-15 22:58:32 <psykose> what is the output of `mount`
2022-12-15 22:58:49 <psykose> pastebin it
2022-12-15 22:59:27 <psykose> (of both variants, one with compress and one with compress-force in fstab, assuming that's where you pass it)
2022-12-15 23:12:17 <ymcrp> https://paste.mozilla.org/jpWPk1Me
2022-12-15 23:13:10 <ymcrp> I don't think that mount options itself is a problem
2022-12-15 23:13:40 <dwfreed> more likely this is an upstream kernel change
2022-12-15 23:13:52 <dwfreed> what kernel are you running on alpine?
2022-12-15 23:15:10 <ymcrp> I don't know, just used 3.17 iso to install and try
2022-12-15 23:18:03 <psykose> ok, it's actually set on the fs
2022-12-15 23:18:07 <psykose> everything past that layer is the kernel so
2022-12-15 23:18:48 <psykose> i was about to ask what kernel were you running on debian but it probably has 90 million patches on it and the version would be a little useless
2022-12-15 23:20:59 <dwfreed> psykose: I mean, you can look at the patches :P
2022-12-15 23:22:40 <psykose> i think looking at debian patchsets should be strictly in the getting-paid-for-it category
2022-12-15 23:22:41 <psykose> yeesh
2022-12-15 23:23:08 <ymcrp> 5.10.0-19-amd64 on Debian and 5.15.82-0-lts on Alpine
2022-12-15 23:25:18 <dwfreed> what does dmesg say when mounting? btrfs logs compression settings used at info level
2022-12-15 23:26:32 <psykose> (dmesg | grep BTRFS)
2022-12-15 23:29:29 <ymcrp> just one message with 'device fsid ... scanned by btrfs'
2022-12-15 23:30:38 <psykose> it's all caps
2022-12-15 23:30:56 <psykose> add -i if you want, but if you don't see any other messages then the dmesg is incomplete for some reason
2022-12-15 23:31:21 <ymcrp> nono there is more you're right
2022-12-15 23:31:38 <ymcrp> force zstd compression, level 2
2022-12-15 23:32:05 <psykose> then indeed it's forced there as well
2022-12-15 23:33:03 <psykose> how do you compare the file size ratio between force/non-force?
2022-12-15 23:33:35 <ymcrp> copy files from flash and compsize directory
2022-12-15 23:38:19 <psykose> hmm
2022-12-15 23:38:38 <psykose> should work then
2022-12-15 23:39:53 <ymcrp> dont-ban-me-plz I'm going to test if Void have a problem with compression and report here I suppose
2022-12-15 23:40:04 <ymcrp> and then you can ban me
2022-12-15 23:40:20 <ymcrp> for herecy I guess
2022-12-15 23:41:45 <psykose> i would test it but it would require me to get a usb out and go through the dance and that sounds like effort :)
2022-12-15 23:41:51 <panekj> for spelling
2022-12-15 23:51:13 <ymcrp> panekj, that's fair, I've started to speak in English just several months before, so I'm truly sorry
2022-12-15 23:51:32 <panekj> *it was a joke*
2022-12-15 23:54:10 <ymcrp> And I'm still sorry, most of the time I just use phrases that sounds better with literally zero knowledge :^)
2022-12-16 00:30:59 <invoked> your english looks good to me.
2022-12-16 00:48:22 <upstart> hello
2022-12-16 00:49:01 <upstart> i am having issues booting the installation image
2022-12-16 00:49:58 <upstart> after OpenRC starts, it hangs on "loading hardware drivers" or something similar
2022-12-16 00:50:45 <upstart> i can't even ctrl+alt+f6 into the shell to find out what's going on in the journal
2022-12-16 00:51:03 <upstart> any help?
2022-12-16 00:51:58 <Saijin_Naib> What version and platform? What hardware are you working on, and have you seen if another distro of similar kernel version boots okay? 
2022-12-16 00:53:10 <upstart> hello, yes. 3.17 x86_64, dell xps with i7 & nvidia
2022-12-16 00:53:35 <upstart> i am currently running arch, but I wanted to try Alpine
2022-12-16 00:54:18 <upstart> Right now, I'm using kernel 6.0.12
2022-12-16 00:54:24 <ymcrp> invoked, thanks. I suppose it's the contrast with my native tongue skills makes me feel bad
2022-12-16 00:54:41 <ymcrp> psykose, moving back to the Btrfs compression. I'm quite not sure if that's a good news or bad but Void suffers from the exact same problem!
2022-12-16 00:55:35 <psykose> what version of btrfs-progs do you have on debian
2022-12-16 00:56:16 <ymcrp> 5.10.1-2
2022-12-16 00:57:55 <psykose> hmm, no, wouldn't matter
2022-12-16 00:58:09 <psykose> i guess you could manually boot the debian kernel and see if that alone fixes it
2022-12-16 00:58:15 <ymcrp> Don't know if that matters but I've tried musl build of Void
2022-12-16 00:59:32 <ymcrp> 'ok google how to manually boot a kernel'
2022-12-16 01:02:12 <ymcrp> Maybe I'll try it later but now I have to revert to Debian because laptop which I was hoping to switch to Alpine already was and always will be in desperate need of compression
2022-12-16 01:03:17 <ymcrp> I clearly didn't expect that kind of problem
2022-12-16 01:10:41 <minimal> upstart: kernel 6.0.12? Alpine 3.17 does not come with that kernel
2022-12-16 01:11:38 <minimal> upstart: also do you have the linux-firmware-nvidia package installed?
2022-12-16 01:12:32 <minimal> upstart: oh, the hang is while booting the Standard ISO?
2022-12-16 01:15:21 <psykose> ymcrp: i mean, i can't reproduce it https://img.ayaya.dev/LICgJ0Mpo6Q0
2022-12-16 01:15:39 <psykose> not the same with that option
2022-12-16 01:15:50 <psykose> so either you did something else wrong, or it's fixed in a kernel newer than 5.15
2022-12-16 01:19:39 <psykose> for the latter you could upgrade to alpine edge where linux-lts is 6.1 i guess
2022-12-16 01:45:43 <ymcrp> psykose, you're moving files to flash mounted with compress options am I correct? I didn't test it that way because my goal is the root partition with btrfs compression so I did quite the opposite
2022-12-16 01:46:02 <psykose> it's not the opposite because the function is identical
2022-12-16 01:46:03 <psykose> and yes
2022-12-16 01:50:42 <psykose> (i say it's identical because there is no practical difference to doing that vs moving files from flash to a folder, and remounting your root with two arguments as the comparison)
2022-12-16 01:50:55 <psykose> it's going to be the same single compression option on one filesystem
2022-12-16 02:21:03 <ymcrp> psykose, I understand and so will check everything once more and try newer kernels. Thank you for your time
2022-12-16 02:48:00 <bleb> does someone have a list of font packages that provide console fonts?
2022-12-16 02:48:39 <bleb> the wiki gives terminus-font as an example, ideally I would like the default vga font blown up 2x or 3x
2022-12-16 03:45:08 <anderson-crypto1[m]> I'll help the community how earn $30k within 3 days and hours but you will reimburse me 10% of your dividend when you collect it. Note: only interested people should involve. Whatsapp  +1 (559) 666‑3967 immediately.
2022-12-16 03:45:08 <anderson-crypto1[m]> https://t.me/Andersonwatts1
2022-12-16 03:49:09 <upstart> hello
2022-12-16 03:56:55 <upstart> i'm trying to get the installation image to boot, but it hangs once OpenRC tries to load hardware drivers
2022-12-16 03:57:51 <upstart> i can install and use arch with the 6.0.12 kernel on my dell xps laptop
2022-12-16 04:02:01 <upstart> i've tried standard and extended images
2022-12-16 04:47:53 <uuidNuniq> linux-edge is 6.0.12
2022-12-16 05:20:36 <Saijin_Naib> And confusingly, linux-lts on Edge channel is 6.1.0, as minimal said earlier
2022-12-16 05:20:57 <Saijin_Naib> And the linux-lts kernel config might have more things in it you might want/need
2022-12-16 05:42:11 <uuidNuniq> no, linux-lts does not have more things. 
2022-12-16 05:43:40 <psykose> it does
2022-12-16 05:46:50 <psykose> bleb: probably https://pkgs.alpinelinux.org/contents?file=*.pcf*&path=&name=&branch=edge&arch=x86_64
2022-12-16 05:49:13 <SymbioticFemale> i need to install alpine by serial. at grub, how do i make it output to serial?
2022-12-16 05:49:28 <ikke> console=<serial-device>
2022-12-16 05:49:40 <ikke> add that to the kernel command line
2022-12-16 05:50:06 <psykose> it's probably ttyS0 or something but it depends
2022-12-16 05:50:34 <SymbioticFemale> ok so same as debian. thanks
2022-12-16 05:50:52 <ikke> it's a kernel parameter, so yes it would be the same
2022-12-16 05:52:36 <uuidNuniq> linux-lts does not have amlogic :(
2022-12-16 05:54:38 <psykose> which kernel module would that be
2022-12-16 05:56:25 <uuidNuniq> a lot of meson/amlogic/stmmac/dwmac related 
2022-12-16 05:57:46 <psykose> that doesn't look like a list of kernel modules you could ask to enable
2022-12-16 06:59:06 <Symbioti1Female> yeah its not working
2022-12-16 06:59:42 <Symbioti1Female> i've been trying for an hour, still no output after "EFI stub: Entering in SVC mode with MMU enabled"
2022-12-16 08:26:24 <hoel> Hi, most php module packages are available as both php81-module-name and php-module-name. However, php81-pecl-apcu and php81-pecl-imagick have no versionless counterparts. Is this an overlook from the maintainers (and if so, should I open a bug report?) or is this intentional? Thanks in advance.
2022-12-16 08:27:54 <psykose> a lot of them are missing the provides yeah
2022-12-16 08:27:57 <psykose> you could open an issue
2022-12-16 08:31:21 <hoel> On https://gitlab.alpinelinux.org/alpine/aports, correct?
2022-12-16 08:37:46 <psykose> yes
2022-12-16 08:37:55 <hoel> Thanks, will do.
2022-12-16 09:49:55 <hoel> Done: https://gitlab.alpinelinux.org/alpine/aports/-/issues/14468 I’ve included a list of all php modules missing a versionless package. :)
2022-12-16 09:52:49 <psykose> thanks!
2022-12-16 15:16:09 <bleb> <psykose> bleb: probably https://pkgs.alpinelinux.org/contents?file=*.pcf*&path=&name=&branch=edge&arch=x86_64
2022-12-16 15:16:20 <bleb> psykose: any way to get that list on the command line so I can pipe it to uniq?
2022-12-16 15:16:29 <psykose> nope
2022-12-16 15:16:39 <psykose> well kinda, but not really
2022-12-16 15:16:47 <bleb> classico
2022-12-16 15:19:56 <bleb> feels like it's doing the search again with every new page of results
2022-12-16 15:20:36 <psykose> what could a decent search implementation cost, five dollars?
2022-12-16 15:22:09 <bleb> alternatively, how much does a bad search implementation cost in wasted CPU time
2022-12-16 15:23:33 <hoel> Depends on the use case, obviously.
2022-12-16 15:24:32 <psykose> i did actually start rewriting it once but i can't focus long enough for such things :p
2022-12-16 15:24:36 <psykose> got abandoned as fast as it started
2022-12-16 17:06:18 <scorpion2185[m]1> how do I build a package using the git head?
2022-12-16 17:07:31 <psykose> generally you don't
2022-12-16 17:08:39 <scorpion2185[m]1> i have this
2022-12-16 17:08:44 <scorpion2185[m]1> `source="$pkgname-$pkgver.tar.gz::https://codeberg.org/jackc/smlock/archive/$pkgver.tar.gz"`
2022-12-16 17:09:11 <scorpion2185[m]1> it's for testing/opersonal use
2022-12-16 17:09:52 <psykose> make the source whatever source you want by fetching the tarball of a commit
2022-12-16 17:43:06 <scorpion2185[m]1> how can you use nvidia-driver on alpine?
2022-12-16 17:43:27 <panekj> you most likely can't 
2022-12-16 17:44:53 <scorpion2185[m]1> i kinda suspected it and  if you need them?
2022-12-16 17:45:51 <minimal> if you can't use it then you can't use it, regardless of whether you need it
2022-12-16 17:46:49 <minimal> the nvidia drivers are supplied as binaries, likely compiled using glibc, Alpine doesn't use glibc, it uses musl as its C library
2022-12-16 17:47:10 <minimal> try nouveau instead?
2022-12-16 17:47:34 <scorpion2185[m]1> I didn't know for sure that it wasn't possible
2022-12-16 17:48:31 <minimal> so try nouveau?
2022-12-16 17:48:58 <scorpion2185[m]1> nouveau are so pathetic
2022-12-16 17:49:52 <scorpion2185[m]1> no way... I can smash my gpu directly
2022-12-16 17:50:03 <scorpion2185[m]1> is this because of dkms also?
2022-12-16 17:50:13 <q66> no
2022-12-16 17:52:45 <heat> would it not be possible to just use the gcompat stuff?
2022-12-16 17:52:59 <psykose> the nvidia driver userspace component doesn't work because it's reliant on glibc
2022-12-16 17:53:34 <q66> heat: no
2022-12-16 17:54:04 <scorpion2185[m]1> can you add glibc on alpine?
2022-12-16 17:54:09 <q66> no
2022-12-16 17:54:12 <psykose> it further is even more shit because we don't compile mesa with glvnd
2022-12-16 17:54:24 <heat> q66, why not?
2022-12-16 17:54:35 <q66> because gcompat is effectively useless
2022-12-16 17:54:43 <psykose> it works pretty well for a lot of things
2022-12-16 17:54:45 <q66> also glvns
2022-12-16 17:54:47 <q66> glvnd
2022-12-16 17:54:57 <scorpion2185[m]1> is this the same for Amd cards?
2022-12-16 17:55:00 <psykose> no
2022-12-16 17:55:20 <q66> it is if you want to use proprietary drivers 
2022-12-16 17:55:41 <psykose> heat: https://twitter.com/ptrcnull/status/1575597520574906369 xd
2022-12-16 17:56:35 <heat> psykose, there you go
2022-12-16 17:56:36 <heat> Fixed(tm)
2022-12-16 17:56:57 <q66> did you read the thread 
2022-12-16 17:56:59 <q66> it does not work 
2022-12-16 17:57:10 <psykose> they don't learn to read in portugal
2022-12-16 17:57:35 <psykose> (someone from portugal once told me that was true)
2022-12-16 17:57:40 <psykose> ((also called it poortugal))
2022-12-16 17:57:41 <heat> that is correct
2022-12-16 17:57:42 <psykose> confirm??
2022-12-16 17:57:44 <heat> confirmed
2022-12-16 17:58:15 <psykose> but more seriously gcompat working at all is usually luck
2022-12-16 17:58:16 <heat> q66, it doesn't work but it definely looks like the right path
2022-12-16 17:58:24 <psykose> it definitely works for a lot of stuff
2022-12-16 17:58:29 <psykose> random apps linked against glibc etc
2022-12-16 17:58:44 <psykose> you can probably copy random cli apps off arch and they would work for instance (unless they used systemd in some way)
2022-12-16 17:58:58 <heat> because nvidia isn't going to have a change of heart and suddenly release a musl-compatible version (or an open-source one lol)
2022-12-16 17:59:01 <psykose> but for things like this you kinda have to reverse engineer the driver to find out why the specific stuff doesn't work it's garbage
2022-12-16 17:59:23 <psykose> so basically it would be expending a lot of resources for no reason
2022-12-16 17:59:29 <q66> heat: the right path is not using novideo
2022-12-16 17:59:32 <psykose> perhaps nvk will make it better in 2 years :)
2022-12-16 17:59:55 <psykose> the people with the talent to even figure this out also usually are able to just buy themselves a gpu that works instead
2022-12-16 18:00:06 <psykose> as i'm sure you know yourself :p
2022-12-16 18:02:00 <singham> Will I have to do lbu-commit every time I shutdown on lime2?
2022-12-16 18:02:32 <singham> If I remember correctly, last I used alpine on lime2, I had to do lbu-commit to save changes.
2022-12-16 18:05:44 <Apachez> depends on how your alpine is installed
2022-12-16 18:06:11 <singham> means?
2022-12-16 18:07:47 <singham> How do I install so I don't have to lbu-commit ?
2022-12-16 18:08:06 <singham> And what makes the user use lbu-commit
2022-12-16 18:08:18 <minimal> do a Sys-mode (run-from-disk) installation, rather than run-from-ram
2022-12-16 18:20:01 <singham> What would be different while installing?
2022-12-16 18:20:48 <singham> Generic armv7 tarball contains dtb, apk directory and stuff. Will that install on disk?
2022-12-16 18:23:46 <minimal> singham: it would install Alpine onto the disk/SSD/SDcard/whatever in a form that boots and runs directly from the storage device. Whereas run-from-ram loads Alpine components into memory and the rootfs is in memory and runs from there (whic his why lbu is needed to write any changes to the in-memory rootfs back to the storage device)
2022-12-16 18:25:11 <singham> That is very informative. By default, what's the state of configuration?
2022-12-16 18:26:45 <minimal> singham: well which type of install did you choose?
2022-12-16 18:29:12 <singham> I have downloaded generic arm
2022-12-16 18:29:16 <singham> tarball
2022-12-16 18:29:34 <minimal> you ran setup-alpine/setup-disk ?
2022-12-16 18:29:43 <singham> No, not yet
2022-12-16 18:29:51 <minimal> so how did you install it then?
2022-12-16 18:29:56 <singham> So I have to use setup-disk?
2022-12-16 18:30:22 <singham> I gotta install first, was first learning about it
2022-12-16 18:30:37 <minimal> ok, how were you intending to install Alpine?
2022-12-16 18:31:29 <scorpion2185[m]1> is this the same for Amd cards?
2022-12-16 18:31:40 <scorpion2185[m]1> proprietary driver
2022-12-16 18:31:47 <psykose> no
2022-12-16 18:31:53 <psykose> ah, for proprietary yes
2022-12-16 18:32:05 <psykose> i don't know of anyone that uses the proprietary amd drivers though
2022-12-16 18:32:22 <minimal> scorpion2185[m]: you were already told yes & no: yes same issue for AMD proprietary driver, no for opensource amd/radeon drivers
2022-12-16 18:32:51 <scorpion2185[m]1> so no way to use a gpu marvellous
2022-12-16 18:33:09 <scorpion2185[m]1> minimal: i wasn't sure if somebody answered me
2022-12-16 18:33:38 <singham> minimal: By extracting tarball and flashing uboot
2022-12-16 18:34:23 <minimal> singham: so then you're installed to a storage device, not running from ram so lbu is irrelevant
2022-12-16 18:34:48 <scorpion2185[m]1> and no noveau in the repo ? but no use for that anyway just checking it
2022-12-16 18:35:46 <scorpion2185[m]1> is possible that alpine devs contact nvidia and at least try to get them?
2022-12-16 18:35:46 <minimal> scorpion2185[m]: its a kernel module and so would be in the kernel package
2022-12-16 18:36:34 <uuidNuniq> singham: uboot? lol. setup-alpine then chroot, apk del grub syslinux grub-efi , setup your uboot
2022-12-16 18:36:54 <scorpion2185[m]1> ah, could somebody try to contact nvidia or something?
2022-12-16 18:37:35 <ikke> Do you think nvidia cares about it?
2022-12-16 18:37:37 <uuidNuniq> scorpion2185[m]1: contact nvidia? 
2022-12-16 18:37:49 <uuidNuniq> let Linus do it
2022-12-16 18:38:28 <scorpion2185[m]1> uuidNuniq: yes, to have the driver or somehow find a way
2022-12-16 18:38:40 <scorpion2185[m]1> ikke: no but somebody could at least try
2022-12-16 18:38:53 <uuidNuniq> the ISO does not boot? 
2022-12-16 18:38:53 <ikke> scorpion2185[m]1: do you think nobody tried?
2022-12-16 18:38:55 <ikke> https://www.youtube.com/watch?v=IVpOyKCNZYw
2022-12-16 18:39:03 <psykose> scorpion2185[m]1: you should message nvidia
2022-12-16 18:41:40 <scorpion2185[m]1> so somebody tried it nice
2022-12-16 18:41:43 <pj[m]> psykose: alpine dev pls cntct novideo
2022-12-16 18:42:24 <psykose> sorry but i have an nvidia gpu which means i can't see my screen currently
2022-12-16 18:42:34 <psykose> please purchase me the latest gpu so i can see my screen
2022-12-16 18:42:36 <psykose> then i can get to it
2022-12-16 18:42:42 <pj[m]> understandable have a nice day
2022-12-16 18:42:49 <psykose> :)
2022-12-16 18:42:51 <scorpion2185[m]1> same for amd then?
2022-12-16 18:43:11 <pj[m]> you asked that 2 times already and both times it was answered
2022-12-16 18:43:34 <pj[m]> amd proprietary driver will not work, amd open source driver will work
2022-12-16 18:43:35 <singham> 
ACTION thanks you minimal for the nice explanation

2022-12-16 18:44:16 <singham> 
ACTION wishes you great folks a nice day! :)

2022-12-16 18:44:22 <psykose> have an even better one
2022-12-16 18:44:42 <scorpion2185[m]1> pj: now i mean if someone ask them
2022-12-16 18:44:51 <pj[m]> ???
2022-12-16 18:44:54 <scorpion2185[m]1> *asked them
2022-12-16 18:44:57 <pj[m]> they have open source driver
2022-12-16 18:45:12 <pj[m]> why would someone ask for open sourcing the proprietary driver
2022-12-16 18:45:15 <heat> ask who? nvidia doesn't give a shit
2022-12-16 18:45:39 <psykose> why would someone want to open source the slower amd proprietary driver
2022-12-16 18:45:40 <psykose> xD
2022-12-16 18:45:59 <scorpion2185[m]1> of course I mean the other driver
2022-12-16 18:46:12 <pj[m]> but why
2022-12-16 18:46:14 <pj[m]> for what reason
2022-12-16 18:46:19 <pj[m]> what do you meeeeeaaaaaan
2022-12-16 18:46:41 <heat> i think they mean the nvidia driver
2022-12-16 18:46:45 <scorpion2185[m]1> and I mean to get it on alpine somehow better to have it open source
2022-12-16 18:47:04 <psykose> they already have an open source driver
2022-12-16 18:47:07 <heat> yeah, it would be better, but you don't have it open source and that most likely will never happen
2022-12-16 18:47:41 <pj[m]> there is already an open source driver that works, why would someone ask for proprietary driver to be open sourced
2022-12-16 18:47:49 <pj[m]> doesn't make sense
2022-12-16 18:47:54 <heat> no there isnt?
2022-12-16 18:47:59 <pj[m]> heat stop
2022-12-16 18:48:06 <heat> ok
2022-12-16 18:48:11 <pj[m]> we already moved on to amd
2022-12-16 18:48:29 <heat> i'm not sure we did
2022-12-16 18:48:44 <scorpion2185[m]1> the open source driver work very bad
2022-12-16 18:48:54 <heat> yeah we're still in nvidia
2022-12-16 18:49:10 <scorpion2185[m]1> hence did somebody ask amd to get the proprietary driver on alpine?
2022-12-16 18:49:13 <psykose> what
2022-12-16 18:49:15 <heat> what
2022-12-16 18:49:19 <Ermine> what
2022-12-16 18:49:24 <psykose> the amd open source driver is better than the proprietary one
2022-12-16 18:49:26 <psykose> like literally
2022-12-16 18:49:38 <psykose> maybe it's a bit worse in GL and then it's faster at everything in vulkan
2022-12-16 18:49:40 <psykose> what are you on
2022-12-16 18:49:54 <pj[m]> heat: "yeah, we're still in nvidia"
2022-12-16 18:50:03 <scorpion2185[m]1> is that so?ok then
2022-12-16 18:50:10 <psykose> we have repeated that like 8 times
2022-12-16 18:50:32 <heat> intel open good, amd open good, nvidia open bad
2022-12-16 18:50:52 <psykose> sorry can't read that message, got stuck in an intel gpu hang
2022-12-16 18:50:57 <heat> intel no closed, amd closed but not as good as open, nvidia closed good (and used to be best of them all)
2022-12-16 18:51:13 <psykose> intel does have closed now doesn't it
2022-12-16 18:51:15 <Ermine> amd closed source driver supports AMF.
2022-12-16 18:51:22 <heat> does it?
2022-12-16 18:51:32 <heat> I don't think they do
2022-12-16 18:51:37 <psykose> https://www.intel.com/content/www/us/en/download/747008/intel-arc-graphics-driver-ubuntu.html
2022-12-16 18:51:39 <psykose> whatever the fuck this is
2022-12-16 18:52:07 <psykose> unless it's just repackaging
2022-12-16 18:52:21 <Ermine> Ubuntu 22.04 is at 5.15, and Arc Graphics support was merged in later versions afaik.
2022-12-16 18:52:33 <psykose> they backported it
2022-12-16 18:52:50 <psykose> it's actually funny because it's backported to that but not even merged in edge for some stuff
2022-12-16 18:53:01 <scorpion2185[m]1> how do I make sure to load lvm things upon boot?
2022-12-16 18:56:44 <scorpion2185[m]1> I get error no such device: 37...uuid
2022-12-16 18:56:49 <scorpion2185[m]1> error: no server is specified
2022-12-16 18:57:34 <Ermine> I guess you need to add lvm service to boot runlevel?
2022-12-16 18:58:10 <scorpion2185[m]1> error: you need to load the kernel first
2022-12-16 19:32:36 <jh81> What bootloader do you have?
2022-12-16 19:35:10 <jh81> Grub? https://bbs.archlinux.org/viewtopic.php?id=214315 look here, you need to add lvm modules to boot from lvm
2022-12-16 19:36:29 <jh81> Never mind, that page has no solution
2022-12-16 19:37:44 <scorpion2185[m]1> grub i checking the wiki again, I added lvm to boot runlevel but as before
2022-12-16 19:39:50 <jh81> I think you need to load the lvm mosule in grub too
2022-12-16 19:40:12 <jh81> The arch wiki has a good article on grub
2022-12-16 19:40:16 <jh81> https://wiki.archlinux.org/title/GRUB
2022-12-16 19:44:29 <scorpion2185[m]1> how do I use mknitfs inside a chroot?
2022-12-16 19:47:34 <scorpion2185[m]1> for grub do you edit /etc/default/grub right?
2022-12-16 19:48:29 <scorpion2185[m]1> I having this error: flag provided but not defined: -c
2022-12-16 19:48:48 <scorpion2185[m]1> when  mkinitfs -c /mnt/etc/mkinitfs/mkinitfs.conf -b /mnt/ $(ls /mnt/lib/modules/)
2022-12-16 20:03:19 <scorpion2185[m]1> how should I make the line in /etc/default/grub)
2022-12-16 20:03:58 <scorpion2185[m]1> I have 1 luks partition, with lvm on it. and /boot on another partition , not LUKS
2022-12-16 20:05:47 <scorpion2185[m]1> I don't have GRUB_CMDLINE_LINUX_DEFAULT
2022-12-16 20:19:07 <uuidNuniq> scorpion2185[m]1: looks like those commands are run outside of chroot. 
2022-12-16 20:19:45 <uuidNuniq> inside chroot, you'll just run  mkinitfs -f /etc/fstab KERNEL_VERSION
2022-12-16 20:30:28 <scorpion2185[m]1> can i do it inside the chroot?
2022-12-16 20:31:19 <scorpion2185[m]1> will mknitfs -f /etc.. do the same but inside it right?
2022-12-16 20:33:27 <scorpion2185[m]1> uuidNuniq: ^
2022-12-16 20:35:24 <scorpion2185[m]1> flag provided but not defined: -f
2022-12-16 20:49:57 <uuidNuniq> scorpion2185[m]1: how is that possible
2022-12-16 21:05:17 <pitsch[m]> hey just installed alpine and updated via the edge repo on aarch64. i see a lot of segmentation fault errors. ufw, docker and some zfs tooling just gives Segmentation fault and exists
2022-12-16 21:05:46 <pitsch[m]> not that familiar with the busybox/ash so hard to debug on the machine.
2022-12-16 21:06:11 <uuidNuniq> pitsch[m]: paste your /etc/apk/world
2022-12-16 21:07:12 <pitsch[m]> https://termbin.com/fe05
2022-12-16 21:07:34 <psykose> what is your device
2022-12-16 21:08:04 <pitsch[m]> rpi4
2022-12-16 21:08:14 <psykose> hmm
2022-12-16 21:08:20 <psykose> should be fine then
2022-12-16 21:08:25 <psykose> post /etc/apk/repositories
2022-12-16 21:08:33 <psykose> i assume you just type `docker` and it segfaults?
2022-12-16 21:13:12 <uuidNuniq> where is the apk-tools
2022-12-16 21:13:36 <uuidNuniq> does "apk fix" help?
2022-12-16 21:19:30 <scorpion2185[m]1> https://developer.nvidia.com/blog/nvidia-releases-open-source-gpu-kernel-modules/  https://github.com/NVIDIA/open-gpu-kernel-modules
2022-12-16 21:22:59 <pitsch[m]> psykose no matter what i try some docker up FOO or just plain docker. hangs a while then segfault
2022-12-16 21:25:39 <uuidNuniq> pitsch[m]: post /etc/apk/repositories
2022-12-16 21:26:48 <pitsch[m]> relevant line http://ftp.halifax.rwth-aachen.de/alpine/edge/main & community
2022-12-16 21:27:03 <pitsch[m]> that was filled by setup-alpine by checking connection speeds
2022-12-16 21:27:31 <uuidNuniq> all lines
2022-12-16 21:27:49 <pitsch[m]> everything else is just commented out.
2022-12-16 21:28:17 <uuidNuniq> apk upgrade -aU
2022-12-16 21:28:50 <psykose> could you replace that with http://dl-cdn.alpinelinux.org/alpine/ for both, then apk upgrade -Ua, then reboot, and then get a list of commands that segfault
2022-12-16 21:29:26 <pitsch[m]> should i abort the command and change repos or just let rwth handle it?
2022-12-16 21:29:44 <pitsch[m]> already started reinstall
2022-12-16 21:30:04 <pitsch[m]> but with the rwth aachen repo
2022-12-16 21:30:33 <pitsch[m]> not sure if cancelling this pricess is the best idea
2022-12-16 21:30:45 <uuidNuniq> just let it finish 
2022-12-16 21:31:02 <uuidNuniq> what is/are reinstalling? 
2022-12-16 22:38:05 <pitsch[m]> shoudve started it in a tmux...
2022-12-16 22:46:11 <lopid> hmm
2022-12-17 01:14:53 <rtcm> yo sup
2022-12-17 01:15:22 <rtcm> hows it going
2022-12-17 01:15:53 <rtcm> I have some issue booting with xen
2022-12-17 01:16:20 <rtcm> any1 else ever had an awesome blackscreen?
2022-12-17 06:39:40 <uuidNuniq> continue the amlogic s905d chip: i got fedora 37 installed with the help of dracut to create a initramfs. this is all the modules that needed for this board : https://0x0.st/ondR.txt 
2022-12-17 06:40:36 <uuidNuniq> the one is missing from alpine linux-edge is : mdio_mux_mmioreg
2022-12-17 06:41:34 <uuidNuniq> and alpine kernel also missing everything about *scpi*
2022-12-17 06:58:24 <uuidNuniq> psykose, mps , can you add those modules to https://git.alpinelinux.org/aports/tree/community/linux-edge/config-edge.aarch64 ? thank you 
2022-12-17 06:58:37 <uuidNuniq> all modules are here: https://0x0.st/ondR.txt
2022-12-17 07:00:15 <uuidNuniq> it would be wonderful if all the above modules can be enabled. right now about 70% are enabled already. 
2022-12-17 07:00:59 <uuidNuniq> alpine is really efficient for this device. a perfect headless distro for this one. 
2022-12-17 07:01:26 <ikke> uuidNuniq: best to open up a gitlab issue
2022-12-17 07:01:33 <ikke> mps maintains linux-edge
2022-12-17 07:03:25 <uuidNuniq> ok. i'll create a gitlab account, and i'll do a diff and then list the kernel CONFIG options that needs to be enabled. 
2022-12-17 09:51:30 <rtcm> yo sup
2022-12-17 15:28:09 <scorpion2185[m]1> If I have a luks partition with what name will it be unlocked?
2022-12-17 15:29:33 <lopid> does alpine set up luks for you?
2022-12-17 15:30:10 <scorpion2185[m]1> no i am setting it
2022-12-17 15:30:25 <lopid> then you can choose the name when you open it
2022-12-17 15:31:06 <scorpion2185[m]1> how do I set `GRUB_CMDLINE_LINUX_DEFAULT="` ?
2022-12-17 15:31:40 <ikke> in /etc/default/grub
2022-12-17 15:32:17 <scorpion2185[m]1> `cryptroot=UUID=<UUID> cryptdm=<name>`
2022-12-17 15:32:47 <scorpion2185[m]1> doesn'it need in cryptdm the name of the partition?
2022-12-17 15:33:30 <ikke> I have cryptdm=root
2022-12-17 15:34:30 <ikke> It defines the name it will give the device
2022-12-17 15:34:40 <ikke> it will be exposed as /dev/mapper/<name>
2022-12-17 15:36:26 <ikke> So it does not matter that much what you name it
2022-12-17 15:37:17 <scorpion2185[m]1> are these modules fine? gpt uefi
2022-12-17 15:37:17 <scorpion2185[m]1> `GRUB_PRELOAD_MODULES="luks cryptodisk part_gpt lvm"`
2022-12-17 15:37:57 <ikke> I don't have that
2022-12-17 15:38:25 <ikke> I do have lvm and cryptsetup in /etc/mkinitfs/mkinitfs.conf
2022-12-17 15:38:36 <ikke> features="ata base ide scsi usb virtio ext4 lvm cryptsetup"
2022-12-17 15:38:47 <ikke> scorpion2185[m]1: but maybe you are doing FDE?
2022-12-17 15:39:01 <ikke> I don't have my bootloader encrypted
2022-12-17 15:41:41 <scorpion2185[m]1> in  /boot/grub/grub.cfg it shows the lvm uuid
2022-12-17 15:42:08 <scorpion2185[m]1> lvm on luks but /boot not encrypted
2022-12-17 15:42:37 <ikke> For me it's the encrypted partition (sda2)
2022-12-17 15:43:30 <ikke> /dev/sda2 -> /dev/mapper/root
2022-12-17 15:43:37 <ikke> That has lvm on it
2022-12-17 15:43:44 <ikke> and exposes the root and other partitions
2022-12-17 15:44:02 <ikke> (lvm volums)
2022-12-17 15:44:40 <ikke> https://tpaste.us/yvmy 
2022-12-17 15:45:52 <scorpion2185[m]1> in BEGIN /etc/grub.d/00_header  i have the lvm uuid
2022-12-17 15:46:27 <scorpion2185[m]1> it doesn't find that because it is in the luks
2022-12-17 15:46:49 <ikke> It should not be the lvm uuid
2022-12-17 15:55:04 <scorpion2185[m]1> i regenarate grub some times
2022-12-17 15:55:13 <scorpion2185[m]1> why does it use that uuid?
2022-12-17 15:55:43 <scorpion2185[m]1> maybe chroot breaks something
2022-12-17 15:58:12 <ikke> Is that uuid in /etc/default/grub.conf?
2022-12-17 15:58:28 <ikke> err, /etc/default/grub
2022-12-17 15:58:48 <scorpion2185[m]1> no where the heck is it taking it?
2022-12-17 15:58:58 <ikke> Not sure
2022-12-17 16:06:00 <scorpion2185[m]1> can you share your fstab?
2022-12-17 16:07:04 <scorpion2185[m]1> do you need to edit etc/conf.d/dmcrypt ?
2022-12-17 16:07:14 <ikke> https://tpaste.us/YrZg
2022-12-17 16:07:50 <ikke> I have not made any changes there
2022-12-17 16:08:05 <ikke> fyi, I just used alpines installer to set it up
2022-12-17 16:14:12 <scorpion2185[m]1> is vg0 the name of the VG right?
2022-12-17 16:16:23 <ikke> yes
2022-12-17 16:22:24 <minimal> has anyone noticed any poweroff problems since Edge switch to 6.1 kernel? I built VMs using 6.1 and after "poweroff" everything shuts down (i.e. init.d services) but the VM does not actually then (ACPI) power off. I'm using tiny-power-button kernel module rather than button module and acpid. This has worked fine for many months.
2022-12-17 16:22:41 <minimal> Just about to test button/acpid to see if that is also affected
2022-12-17 17:14:07 <scorpion2185[m]1> can you have /boot as ext4 and as a separate partition?
2022-12-17 17:26:49 <scorpion2185[m]1> and a fat32 efi partition mounted at /boot/efi/ right?
2022-12-17 17:27:21 <scorpion2185[m]1> (for efi)
2022-12-17 17:31:39 <scorpion2185[m]1> how should I mount it on fstab? I mean /boot
2022-12-17 18:29:08 <Saklad5> I've been installing the virtual image of Alpine onto cloud servers, and I think I'm in love
2022-12-17 18:29:43 <Saklad5> It's so much easier than having to navigate preloaded things like cloud-init
2022-12-17 18:30:20 <Saklad5> My understanding is that it's basically standard but without the usual hardware drivers. Is that about right?
2022-12-17 18:30:34 <Saklad5> If so, I have to ask: what's mini-root?
2022-12-17 18:30:45 <ikke> That's just a filesystem without kernel / init system
2022-12-17 18:30:57 <ikke> Mostly geared towards docker-style containers
2022-12-17 18:31:35 <Saklad5> That's the description, sure, but what does that mean?
2022-12-17 18:31:45 <Saklad5> What's actually in it, then?
2022-12-17 18:31:56 <ikke> A filesystem, busybox
2022-12-17 18:32:14 <ikke> apk-tools ofcourse
2022-12-17 18:32:42 <Saklad5> Couldn't you theoretically drop apk-tools for a container?
2022-12-17 18:32:55 <ikke> You typically want to be able to install packages
2022-12-17 18:33:08 <Saklad5> You could do that from outside the container, right?
2022-12-17 18:33:29 <ikke> Not with docker
2022-12-17 18:33:53 <scorpion2185[m]1> if you install openrc and the kernel can you boot in a physical machine ? wiht minirootfs
2022-12-17 18:34:07 <Saklad5> I get the impression that literally can't boot in a physical machine
2022-12-17 18:34:22 <Saklad5> Neither can virtual, really: it only has support for the sort of interfaces a hypervisor gives
2022-12-17 18:34:38 <ikke> If you add back everything that is missing, sure
2022-12-17 18:34:50 <Saklad5> So is it just the absolute bare minimum to achieve POSIX compliance?
2022-12-17 18:35:01 <ikke> Meaning a kernel, modules, firmware, bootloader, openrc + service setup
2022-12-17 18:35:28 <ikke> Saklad5: It's used to build the alpine docker image
2022-12-17 18:35:39 <Saklad5> I know
2022-12-17 18:36:10 <ikke> So it contains only the bare mininum that is necessary to make a useful container
2022-12-17 18:36:16 <ikke> Everything else can be installed through apk
2022-12-17 18:36:46 <scorpion2185[m]1> how do you install the modules and service setup?
2022-12-17 18:37:08 <ikke> scorpion2185[m]1: Just use the standard iso
2022-12-17 18:37:11 <Saklad5> Yeah
2022-12-17 18:37:30 <Saklad5> Virtual is pretty much perfect for cloud servers
2022-12-17 18:37:41 <ikke> scorpion2185[m]1: if you want to do it yourself, start with a standard setup and study it
2022-12-17 18:38:20 <Saklad5> But virtual is just standard with less drivers, so if you're not running on bare metal you shouldn't use standard
2022-12-17 18:38:32 <Saklad5> That's my impression having used both, anyway
2022-12-17 18:39:53 <Saklad5> Does the Alpine team accept donations?
2022-12-17 18:40:21 <ikke> Saklad5: We don't have an entity that can accept donations, individual members might accept donations
2022-12-17 18:40:23 <scorpion2185[m]1> how do I install setup-xorg-base?
2022-12-17 18:40:37 <ikke> scorpion2185[m]1: https://pkgs.alpinelinux.org/contents?file=setup-xorg-base&path=&name=&branch=edge
2022-12-17 18:41:01 <Saklad5> I wouldn't use edge for an unattended install unless you have a reason, though
2022-12-17 18:41:28 <scorpion2185[m]1> it is on stable too
2022-12-17 18:46:23 <minimal> Saklad5: what's your concern with cloud-init (speaking as the Alpine cloud-init package maintainer)
2022-12-17 18:46:50 <Saklad5> Nothing: I just have a DigitalOcean Ubuntu image and configuration often feels like I'm fighting it
2022-12-17 18:47:08 <minimal> saklad5: have you tried cloud-init on Alpine?
2022-12-17 18:47:12 <Saklad5> Nope
2022-12-17 18:47:58 <Saklad5> I think it's just a matter of personal preference rather than a legitimate problem
2022-12-17 18:48:00 <minimal> saklad5: perhaps this might interest you: https://github.com/dermotbradley/create-alpine-disk-image
2022-12-17 18:48:34 <Saklad5> I think I'll stick with the wonderfully straightforward ISO installation flow
2022-12-17 18:49:05 <Saklad5> I like not running into problems I don't understand
2022-12-17 18:49:07 <minimal> ok, though that sort of defeats the "cloud" idea
2022-12-17 18:49:54 <Saklad5> Personally, the point of the cloud idea for me is having things in multiple locations and not having to deal with an ISP
2022-12-17 18:50:07 <Saklad5> I'm not really automating anything here
2022-12-17 18:50:13 <ikke> There is also https://gitlab.alpinelinux.org/alpine/cloud/tiny-cloud
2022-12-17 18:52:18 <Saklad5> My setup routine is simple: install OpenSSH, configure it to only use ed25519 host keys, put that host key in an SSHFP record, set up root to only accept login with one of my PGP authentication subkeys, set up Tor to run a single onion service for SSH, set up a daily cron job to upgrade packages, pin relevant software to non-breaking versions
2022-12-17 18:52:43 <Saklad5> And probably set it up as another secondary Knot nameserver, because I might as well
2022-12-17 18:52:51 <Saklad5> This takes like five minutes
2022-12-17 18:55:31 <Saklad5> Most of the manual networking setup is stuff I'd have to concern myself with for DNS anyway
2022-12-17 18:56:35 <Saklad5> I also tweak OpenRC, ever since I learned how to add custom dependency relations
2022-12-17 18:56:52 <Saklad5> I feel that could stand to be a little more prominent
2022-12-17 18:57:40 <Saklad5> In particular, setting "rc_want" in a conf.d file can make everything much cleaner'
2022-12-17 18:59:46 <Saklad5> That reminds me: has anyone else locked themselves out of a server by following the instructions for adding an interface dependency to OpenSSH?
2022-12-17 19:00:49 <minimal> nope, I've just used ListenAddress to restrict the IP listened on, been meaning to look at VRFs however
2022-12-17 19:00:57 <Saklad5> VRF?
2022-12-17 19:01:20 <ikke> virtual routing and forwarding
2022-12-17 19:01:35 <ikke> separated routing tables
2022-12-17 19:02:08 <minimal> e.g.: https://ariadne.space/2021/09/13/efficient-service-isolation-on-alpine-with-vrfs/
2022-12-17 19:02:44 <Saklad5> I think isolation can lend a false sense of security, relative to just hardening everything
2022-12-17 19:03:17 <minimal> why not do both?
2022-12-17 19:03:31 <Saklad5> It adds complexity
2022-12-17 19:04:38 <Saklad5> I try to avoid using both the wildcard and loopback interfaces whenever I can: Unix domain sockets for things that are meant to be strictly local, the external address for everything else
2022-12-17 19:05:28 <Saklad5> Which is what causes the OpenSSH service to complain
2022-12-17 19:06:03 <Saklad5> If you follow the instructions to the letter, you'll end up adding something like "rc_need=net.eth0" to the configuration
2022-12-17 19:06:19 <Saklad5> If you do that on a stock installation, that will prevent OpenSSH from running
2022-12-17 19:06:29 <minimal> yes I'd looked at that some time ago
2022-12-17 19:06:56 <Saklad5> It turns out the correct thing is to ignore the instructions, set up net-online (which I'm guessing was added later) to check the interface in question, and rely on that
2022-12-17 19:07:46 <Saklad5> And in my case, also add "rc_want=tor"
2022-12-17 19:07:46 <minimal> again I'd looked at that also but thought there was an issue with it (my recollection from about 12 months ago is sketchy)
2022-12-17 19:08:17 <Saklad5> If it was a no-op, I wouldn't know, but otherwise it seems to work fine
2022-12-17 19:10:03 <Saklad5> Anyway, since OpenSSH doesn't support Unix domain sockets, I have the Tor onion service direct traffic to the external address of the server, which is what OpenSSH listens on
2022-12-17 19:16:43 <Saklad5> Anyone here have recommendations for running a Matrix server?
2022-12-17 19:18:33 <ikke> No idea how well it works, but I'd go for something like dendrite
2022-12-17 19:18:49 <Saklad5> That's the one I'm currently planning to go with
2022-12-17 19:19:17 <ikke> But it's still in beta
2022-12-17 19:19:35 <Saklad5> The README does say it is ready for early adopters, though
2022-12-17 20:12:18 <minimal> The "powerdown" issue I mentioned earlier with the 6.1 kernel appears to be a Virtualbox 6.x issue - doesn't happen with Virtualbox 7.x or with Qemu
2022-12-17 20:13:18 <ikke> aha ok
2022-12-17 20:15:19 <minimal> basically the VM output shows the message about it "entering powerdown state" (i.e. ACPI S5) but the VM window stays and the VM has to be killed
2022-12-17 20:16:26 <minimal> VB 6.x obviously doesn't like some change between 5.15.x and 6.1 kernels
2022-12-17 20:17:29 <Ermine> minimal: should virtualbox-guest-additions be rebuilt against new kernel headers?
2022-12-17 20:28:08 <minimal> Ermine: I don't think that is a factor, the problem occurs long after OpenRC has shut all services down including the virtualbox ones
2022-12-18 00:40:22 <rtcm> yo sup guys
2022-12-18 00:41:20 <rtcm> why do you think alpine-xen iso doesn't boot
2022-12-18 00:41:52 <rtcm> just some quiet bootloader lines and then blackscreen
2022-12-18 00:42:53 <rtcm> any1 ever had the same issue?
2022-12-18 12:31:29 <satmd> not familiar with alpine-xen, but I'd assume it would open some sort of serial console
2022-12-18 14:38:37 <invoked> satmd: they're gone. but my first guess would be no iommu
2022-12-18 15:09:08 <dwfreed> xen doesn't need iommu
2022-12-18 15:09:27 <dwfreed> kvm doesn't either
2022-12-18 16:53:32 <invoked> ah. i thought xen at some point required iommu but i might just be thinking about qubes.
2022-12-18 17:03:51 <Hello71> qubes doesn't absolutely need iommu either
2022-12-18 17:09:02 <invoked> Hello71: maybe not but it's listed as a requirement
2022-12-18 17:51:51 <jh81> so, building packages for Alpine seems much easier then i thought!
2022-12-18 19:45:55 <scorpion2185[m]1> I can't start x (UI). fbdev error in xorg log
2022-12-18 19:47:22 <ikke> scorpion2185[m]1: what graphics card do you have?
2022-12-18 19:47:50 <ikke> "(EE) Failed to load module "fbdev" (module does not exist, 0)"
2022-12-18 19:47:52 <ikke> That's normal
2022-12-18 19:48:08 <ikke> Same for (EE) Failed to load module "vesa" (module does not exist, 0)
2022-12-18 19:49:40 <scorpion2185[m]1> hello
2022-12-18 19:50:47 <scorpion2185[m]1> intel gemini lake N4100
2022-12-18 19:51:14 <scorpion2185[m]1> I installed xf65-video-intel or whater and setup-xorg-base
2022-12-18 19:51:51 <ikke> scorpion2185[m]1: are you member of the video group?
2022-12-18 19:52:14 <scorpion2185[m]1> yes i using root now
2022-12-18 19:52:45 <scorpion2185[m]1> *I am, EE open /dev/dri/card0: no such file or directory
2022-12-18 19:53:20 <ikke> what does /dev/dri contain?
2022-12-18 19:54:33 <scorpion2185[m]1> by-path card1 renderD120
2022-12-18 19:55:37 <ikke> I have the same error, and I have /dev/dri/card1
2022-12-18 19:57:00 <ikke> Can you perhaps post your xorg.log?
2022-12-18 20:00:27 <scorpion2185[m]1> how can I post from the cli?
2022-12-18 20:00:40 <ikke> apk add tpaste && tpaste <Xorg.log
2022-12-18 20:01:50 <scorpion2185[m]1> pn9W, i am using gomuks on the tty now
2022-12-18 20:02:03 <scorpion2185[m]1> tpaste.us/pn9W
2022-12-18 20:03:18 <scorpion2185[m]1> did I it right?
2022-12-18 20:03:28 <ikke> es
2022-12-18 20:03:30 <ikke> yes
2022-12-18 20:05:40 <Ermine> lsmod | grep i915
2022-12-18 20:06:25 <Ermine> nvmd, however
2022-12-18 20:08:46 <Ermine> Is udev started ?
2022-12-18 20:17:22 <scorpion2185[m]1> any hint?
2022-12-18 20:18:05 <scorpion2185[m]1> yes
2022-12-18 20:18:53 <scorpion2185[m]1> I couldn't start urandom when I followed the wiki
2022-12-18 20:19:01 <Ermine> (rc-service udev status)
2022-12-18 20:19:43 <scorpion2185[m]1> service doesn't exist, also now no idea if it is needed
2022-12-18 20:20:20 <scorpion2185[m]1> I checked with that command
2022-12-18 20:21:00 <scorpion2185[m]1> be back later or tomorrow
2022-12-18 21:16:06 <jh81> hes running startx as root :|
2022-12-18 21:16:28 <jh81> i didnt think startx would even start if you are logged in as root
2022-12-18 21:21:17 <scorpion2185[m]1> yes I said that already here.  startx work for root too
2022-12-18 21:21:21 <scorpion2185[m]1> *works
2022-12-18 21:49:57 <Ermine> s/udev/udevd
2022-12-18 21:49:57 <alpine-meetbot> Ermine meant to say: (rc-service udevd status)
2022-12-18 22:37:14 <bleb> trying to make my console font larger
2022-12-18 22:37:29 <bleb> 32x32 is still too small so I want to set a lower resolution
2022-12-18 22:38:12 <grawity> boot with "video=1024x768" as a kernel option
2022-12-18 22:38:14 <bleb> I tried adding nomodeset video=1680x1050 to the kernel command line, got messed up extreme horizontal stretch
2022-12-18 22:38:19 <bleb> ok I'll try that
2022-12-18 22:38:26 <grawity> don't add "nomodeset", then
2022-12-18 22:39:00 <grawity> is it a physical console or a VM?
2022-12-18 22:40:10 <bleb> yeah still getting the horizontal lines
2022-12-18 22:40:22 <bleb> it's a physical machine with HDMI output
2022-12-18 22:41:44 <bleb> maybe it's a problem with my monitor
2022-12-18 22:41:46 <grawity> I wonder if it's specific to the resolution you're using, or the refresh rate, or bit depth (no idea what's available on your system but video=1680x1050@60 is how you'd specify the refresh rate)
2022-12-18 22:46:24 <bleb> it works on my TV but not on my monitor
2022-12-18 22:46:39 <bleb> it's a weird e-ink monitor so I guess it just doesn't support lower resolutions
2022-12-19 06:00:43 <bananicorn> Hello! I am running Alpine on an ARM SBC and noticed that I am not able to connect to it via serial over micro USB as I could with another distro
2022-12-19 06:01:08 <bananicorn> Is there a package I might be missing for this?
2022-12-19 06:07:30 <aron> bananicorn: this requires kernel drivers/parameters and proper decivetree settings, please check them
2022-12-19 06:10:56 <bananicorn> aron: I see, I will do some testing and research with that
2022-12-19 06:11:25 <bananicorn> I compiled u-boot and the dtb from source so perhaps I missed something there
2022-12-19 06:12:33 <bananicorn> Regarding kernel parameters, I used Alpine's Generic ARM archive and its included kernel, do you know whether it supports this?
2022-12-19 06:14:30 <aron> you'll need g_serial (CONFIG_USB_G_SERIAL)
2022-12-19 06:14:41 <aron> or g_cdc (CONFIG_USB_CDC_COMPOSITE) 
2022-12-19 06:14:53 <aron> I don't know their state in the stock kernel
2022-12-19 06:16:20 <aron> nah, it seems these are not enabled
2022-12-19 06:20:05 <ikke> you also need to make sure console=<serial-device> is set on the kernel cmdline
2022-12-19 06:20:44 <bananicorn> ikke: aron: Thank you both, this will help a lot!
2022-12-19 12:44:06 <Saklad5> Is this a good place to ask about OpenRC?
2022-12-19 12:45:10 <Saklad5> I'm trying to figure out why a "want" dependency isn't getting started consistently.
2022-12-19 12:45:52 <Saklad5> My understanding was that a service with a "want" dependency would be guaranteed to *attempt* starting the service, even if it wouldn't wait for it to start or fail if it isn't installed or doesn't work.
2022-12-19 12:46:56 <Saklad5> But my server keeps ending up in a state where the dependency isn't started, even though starting it manually at that point works fine.
2022-12-19 12:47:28 <Ermine> Maybe it fails to start at boot time?
2022-12-19 12:48:05 <Ermine> What happens if you replace want with need?
2022-12-19 12:48:09 <Saklad5> It works
2022-12-19 12:48:25 <Saklad5> Same with using "rc_before" and "rc_want"
2022-12-19 12:48:55 <Saklad5> The two services are dendrite and nginx, the latter of which is abstracted using "rc_provide=http-server"
2022-12-19 12:49:35 <Ermine> you can try to use "provide http-server"
2022-12-19 12:49:35 <Saklad5> Neither need the other active to function, so a "need" dependency is unnecessary and will just slow down booting
2022-12-19 12:50:13 <Saklad5> Isn't "rc_provide=http-server" the same thing as "provide http-server"? Just without touching the init.d profile?
2022-12-19 12:51:46 <Ermine> Idk, I'd just try. According to docs, this is what you should use.
2022-12-19 12:52:16 <Saklad5> Yes, if I'm *writing* the init.d profile. Which I'm not.
2022-12-19 12:52:35 <Saklad5> OpenRC's documentation could stand to emphasize certain things more.
2022-12-19 12:53:28 <Ermine> Idk then. There's also #openrc chat at libera.chat
2022-12-19 12:54:20 <Saklad5> Thanks
2022-12-19 12:54:28 <Saklad5> For some reason this information is only documented in the comments of /etc/rc.conf, which doesn't even use them
2022-12-19 12:55:13 <Saklad5> I don't think "rc_provide" is the issue, since "rc_need=http-server" works fine
2022-12-19 12:55:20 <Saklad5> I'll go ask on that channel, either way
2022-12-19 12:56:20 <Saklad5> I'm using IRC an awful lot for someone trying to make a Matrix homeserver, but there's something to be said for extreme simplicity
2022-12-19 14:32:39 <eletrotupi_> heya, I'm hitting a 'ERROR: Failed to create boot/System.map-lts: No space left on device', the boot partition was created by the default installation script 
2022-12-19 14:33:23 <eletrotupi_> but I feel that it might be too smol, might be something to take into account on the script?
2022-12-19 14:33:29 <jh81> ahh matrix.. im looking to docker for that
2022-12-19 14:33:32 <eletrotupi_> there's no other kernel there, btw
2022-12-19 14:37:41 <minimal> eletrotupi: setup-disk was changed recently to deal with this issue by creating a bigger partition
2022-12-19 14:37:54 <eletrotupi_> oh, right
2022-12-19 14:38:18 <minimal> resolved in Alpine 3.17 release (from memory)
2022-12-19 15:07:27 <scorpion2185[m]1> How can I fix no UI?
2022-12-19 15:07:37 <scorpion2185[m]1> what does setup-alpine?
2022-12-19 15:08:08 <scorpion2185[m]1> UDH graphics 600 intel gemini lake
2022-12-19 15:17:32 <scorpion2185[m]1> That is the installer script right?
2022-12-19 15:22:53 <eletrotupi> scorpion2185[m]1: yes, setup-alpine is the "installer" script
2022-12-19 15:37:49 <scorpion2185[m]1> AIGLX unable to load driver i965
2022-12-19 15:39:15 <scorpion2185[m]1> error loading /usr/lib/xorg/modules/dri/i965_dri.so no such file
2022-12-19 15:42:03 <scorpion2185[m]1> is that the problem?
2022-12-19 15:58:24 <scorpion2185[m]1> if I can't use the touchscreen what am I missing?
2022-12-19 16:03:41 <scorpion2185[m]1> xf86-input-vmmouse?
2022-12-19 16:09:10 <scorpion2185[m]1> I can start river but not x
2022-12-19 16:12:11 <scorpion2185[m]1> does xf86 need reboot?
2022-12-19 17:09:26 <ham5urg> Is the information here https://wiki.alpinelinux.org/wiki/Alpine_Package_Keeper#Listing_installed_packages still uptodate in regards of listing manually installed packages? Is there a shorter way?
2022-12-19 18:11:58 <ikke> ham5urg: cat /etc/apk/world
2022-12-19 18:37:54 <scorpion2185[m]1> * i CAN'T use the touchscreen on river, it was working on sxmo (x11) installed wiht pmos repo
2022-12-19 18:38:09 <scorpion2185[m]1> test
2022-12-19 18:38:17 <scorpion2185[m]1> I can't use the touchscreen on river
2022-12-19 18:38:30 <scorpion2185[m]1> it was working on sxmo (x11) installed with pmos repo
2022-12-19 18:38:55 <scorpion2185[m]1> I installed all xf85-input packages. I cannot use X.
2022-12-19 18:39:42 <scorpion2185[m]1> ikke is this a problem , from xorg0 log : error loading /usr/lib/xorg/modules/dri/i965_dri.so no such file
2022-12-19 18:45:15 <scorpion2185[m]1> how can I copy from ssh ? I can copy but it is not available on other windows
2022-12-19 18:45:40 <lopid> copy how?
2022-12-19 18:47:59 <scorpion2185[m]1> I copied the text on the ssh terminal window, i can copy it only there
2022-12-19 18:53:37 <lopid> did you select it with the mouse whilst holding shift?
2022-12-19 18:55:58 <scorpion2185[m]1> no
2022-12-19 19:21:33 <thresh> hi.  I'm having a hard time understanding apk behaviour in 3.17 when I have a package with virtual provides (nothing else provides that string too).  It used to work in 3.16.  Here is the description: https://gist.github.com/thresheek/dce4504f0e3c39ad6a3fc1fb69ecc7fd
2022-12-19 19:22:05 <thresh> Seems to be this change: https://gitlab.alpinelinux.org/alpine/apk-tools/-/commit/3b013f458225c2ad8a0d96ec3eb3dde2533e0312
2022-12-19 19:22:33 <thresh> However I'm not sure how to fix that in my packaging then, as I already have an explicit dependancy on a single version 
2022-12-19 19:22:47 <thresh> I'd appreciate any ideas
2022-12-19 19:38:08 <scorpion2185[m]1> how can you set gtk themes on river?
2022-12-19 22:06:56 <ikke> thresh: apk basically does not have an automatic way to pick an alternative for virtual packages
2022-12-19 22:07:16 <ikke> briefly there was an exception if there was only one provider, but that apparently lead to issues, so that was removed again
2022-12-19 22:07:36 <ikke> ok, more then just briefly :P
2022-12-19 22:07:58 <ikke> thresh: it means you have to always explicitly pick a provider
2022-12-19 22:17:53 <vkrishn> if I compile kernel on mobile itself(not a good idea, I know), will I still need extra vendor related patches ?
2022-12-19 22:18:20 <dwfreed> if they're needed to run the device, yes
2022-12-19 22:18:29 <dwfreed> where you compile has no effect on that
2022-12-19 22:18:35 <vkrishn> duh
2022-12-19 22:18:50 <vkrishn> ok, thanks
2022-12-19 22:20:40 <Exagone313> vkrishn: patches are made for hardware compatibility, so any kernel you boot will need those patches (probably).  how and where you build your kernel doesn't change that fact
2022-12-19 22:22:33 <vkrishn> I thought once I add corrent dtb and flash, it would boot to some basic fuctionality
2022-12-19 22:22:41 <vkrishn> correct^
2022-12-19 22:23:00 <vkrishn> enough to run in headless mode
2022-12-19 22:24:33 <thresh> ikke, yeah I think the workaround in my case is to explicitely depend on a non-versioned real package name.  So depend on both nginx-plus-module-ndk and nginx-plus-module-ndk-r28.
2022-12-19 22:25:27 <thresh> ikke, and yeah I think it was working like that for a few years - at least since Alpine 3.8 (which is the first version we supported with NGINX).  Thanks!
2022-12-19 22:26:36 <dwfreed> vkrishn: dtb just describes the hardware; the kernel still has to support it correctly
2022-12-19 22:28:08 <vkrishn> ok
2022-12-19 23:10:47 <bleb> my display turns off after 10-15 minutes of inactivity
2022-12-19 23:11:08 <bleb> or rather the HDMI port stops sending a signal
2022-12-19 23:11:18 <bleb> how would I make it stay on indefinitely?
2022-12-19 23:11:32 <bleb> I tried xset -dpms and no change
2022-12-19 23:19:41 <js_al> I have homes directory mounted on nfs and I'm having flaky intermittent lockups
2022-12-19 23:20:46 <js_al> generally speaking, I'm told to set the debug flags using /proc/sys/sunrpc/nfs_debug, but I don't see those on my system
2022-12-19 23:21:07 <js_al> Is Alpine weird in some way I am not aware of?
2022-12-19 23:22:27 <js_al> The mounts are active, but I'm seeing lots of OPEN ond LOOKUP errors in mountstats, but READ and WRITE look fine
2022-12-20 03:15:38 <bleb> I installed the mupdf and atril packages
2022-12-20 03:15:45 <bleb> mupdf silently exits with status 1, with or without a pdf as an argument
2022-12-20 03:16:05 <bleb> with no argument, atril segfaults
2022-12-20 03:16:10 <bleb> with a pdf as an argument, it silently exits with status 0
2022-12-20 03:17:07 <bleb> any idea what could be wrong?
2022-12-20 03:19:57 <psykose> hmm
2022-12-20 03:19:59 <psykose> both work fine for me
2022-12-20 03:24:34 <bleb> I'm on 3.17
2022-12-20 03:25:35 <bleb> I installed them at first on 3.14 then upgraded to 3.17 following https://wiki.alpinelinux.org/wiki/Upgrading_from_older_versions
2022-12-20 03:25:52 <bleb> since then I have rebooted and run apk update && apk upgrade
2022-12-20 03:26:04 <bleb> and I manually uninstalled and reinstalled mupdf and atril
2022-12-20 03:26:07 <bleb> still same
2022-12-20 03:28:39 <psykose> you have to upgrade with -a, but i don't think that's the issue
2022-12-20 03:28:42 <zcrayfish> dunno about your issue, but I would try the -a and --prune flags with your upgrade; especially considering the big version jump.
2022-12-20 03:28:51 <psykose> the only way i can reproduce it is without a display
2022-12-20 03:38:16 <bleb> I used -a in the main upgrade step
2022-12-20 03:41:53 <bleb> maybe it has to do with my display being eink and having a weird 40fps refresh rate
2022-12-20 04:52:56 <invoked> oh nice. what eink display do you have?
2022-12-20 09:37:15 <jee> hi I've an issue with alpine 3.17, `setcap` isn't anymore installed by default. Which package should I install to get it back ? Thank you
2022-12-20 09:37:59 <psykose> libcap-utils
2022-12-20 09:38:36 <jee> thank you @psykose _||_ 
2022-12-20 09:38:53 <jh81> i was googling that :D
2022-12-20 09:39:00 <jh81> one learns something every day
2022-12-20 09:39:18 <jh81> what does setcap do?
2022-12-20 09:40:40 <jee> i use ddg,  don't find this answer sorry 
2022-12-20 09:41:13 <jh81> its a capabilities thingy :) never knew
2022-12-20 09:41:22 <jee> setting files capabilities 
2022-12-20 09:41:36 <jee> https://linux.die.net/man/8/setcap]
2022-12-20 09:41:53 <jh81> yeah, i figured there's a man page for it
2022-12-20 09:42:23 <jh81> it still doesnt explain what those capability's are :|
2022-12-20 10:14:02 <scorpion2185[m]1> fatal: not a git repository (or any of the parent directories): .git
2022-12-20 10:15:06 <scorpion2185[m]1> i am trying to build a the head of a package 
2022-12-20 10:15:57 <Piraty> jee: https://pkgs.alpinelinux.org/contents
2022-12-20 10:22:12 <jee> Han I didn't know that thank's Piraty very useful 
2022-12-20 10:23:11 <Piraty> so you don't have to query alpine's package content bot anymore
2022-12-20 10:25:11 <jee> do you mean that psykose is a bot ? :D
2022-12-20 10:29:08 <jh81> i should bookmark that too :)
2022-12-20 10:30:24 <Shiz> even sicker protip: if all you need is the command:
2022-12-20 10:30:27 <Shiz> `apk add cmd:setcap`
2022-12-20 10:30:29 <Shiz> :p
2022-12-20 10:30:43 <jee> wtf
2022-12-20 10:30:52 <jh81> nice
2022-12-20 10:31:19 <jee> i fell like an idiot now 
2022-12-20 10:32:02 <Shiz> this also works for shared libraries with e.g. `apk add so:libncurses.so.6`
2022-12-20 10:32:26 <Shiz> (although that's more relevant if you're working with non-recompilable proprietary binaries that link against stuff, else you should just use the normal packages)
2022-12-20 10:32:55 <jee> apk is so powerfull 
2022-12-20 10:33:20 <jh81> dont feel like an idiot :) its not on the wiki page for apk, i didnt know those commands either :P
2022-12-20 10:35:31 <jee> haha i feel like an linux user then :D
2022-12-20 10:37:02 <Shiz> jh81: that's because technically it's not an apk feature, but an abuild one :p
2022-12-20 10:37:16 <jh81> that explains :)
2022-12-20 10:37:29 <Shiz> the packages just have virtual provides= for those cmd:* and so:* names, that are automatically generated by abuild by scanning the built binaries
2022-12-20 10:37:50 <jh81> is this for all opackages? or does the maintainer decide this?
2022-12-20 10:37:58 <Shiz> it's for all packages, abuild does it automatically
2022-12-20 10:38:25 <jh81> nice :) im looking to learn about building packages :) mostly for KDE :)
2022-12-20 10:38:28 <Shiz> or more specifically: for all packages that are built using abuild, which for Alpine is all packages
2022-12-20 10:38:40 <Shiz> but Chimera Linux also uses apk, but doesn't abuild; it probably doesn't work there
2022-12-20 10:38:49 <jh81> ahh okay
2022-12-20 10:38:57 <Shiz> nice, good luck on your journey :)
2022-12-20 10:40:34 <jh81> i will succeed ;-)
2022-12-20 10:45:37 <Shiz> i have complete confidence :0
2022-12-20 15:42:56 <scorpion2185[m]1> how can I insrall older verion of apps_
2022-12-20 15:54:07 <scorpion2185[m]1> the packagge is on testing only
2022-12-20 16:04:14 <scorpion2185[m]1> https://build.alpinelinux.org/buildlogs/build-edge-x86_64/testing/zig/zig-0.9.1-r0.log
2022-12-20 16:04:18 <scorpion2185[m]1> &quit
2022-12-20 16:05:10 <scorpion2185[m]1> how can I fecth that APKB. ?
2022-12-20 16:44:38 <bleb> <invoked> oh nice. what eink display do you have?
2022-12-20 16:44:44 <bleb> invoked: dasung paperlike 253
2022-12-20 16:44:57 <bleb> only got it because employer bought it for me :)
2022-12-20 16:47:01 <invoked> excellent. how do you like it?
2022-12-20 17:36:09 <bleb> invoked: it's great, I want to get a music stand light with a dimmer so I can make it as dim as I want if I'm using it before bed
2022-12-20 17:36:53 <invoked> i want one. they've always been so expensive, though
2022-12-20 17:37:13 <invoked> i have pdf manuals and such that i'd like to have on that display, if i had one
2022-12-20 17:37:16 <bleb> it has had a couple annoying bugs, like it doesn't work with lower resolution inputs, and raspberry pis don't work ootb I think for the same reason
2022-12-20 17:37:32 <bleb> and there's this weird pdf viewer bug which may or may not be due to the monitor
2022-12-20 17:38:40 <invoked> do those screens suffer from burn-in like some of the kindles do?
2022-12-20 17:39:00 <invoked> i would suspect for what they cost, probably not.
2022-12-20 17:44:06 <bleb> there is temporary ghosting which you can remove by pressing a button that flashes the screen black and white
2022-12-20 17:44:37 <bleb> there might be burn-in over time but it had the "no signal" screen for many weeks and no burn in from that
2022-12-20 17:50:48 <scorpion2185[m]1> how can I fetch the patch of that release https://build.alpinelinux.org/buildlogs/build-edge-x86_64/testing/zig/zig-0.9.1-r0.log  ?
2022-12-20 17:58:36 <invoked> bleb: there are people who have hacked kindles to be 2nd displays and more. see also: https://github.com/NiLuJe/FBInk
2022-12-20 17:58:50 <invoked> that's a bit too adventurous for my lifestyle at the moment, though
2022-12-20 18:37:27 <bleb> invoked: cool thanks for the link
2022-12-20 20:00:24 <scorpion2185[m]1> nvm I used built binary
2022-12-20 20:01:53 <scorpion2185[m]1> how do you close/restart a tty? from another tty
2022-12-20 20:16:34 <f00860> I want to replace the iptables backend from legacy to nf_tables.
2022-12-20 20:17:17 <f00860> Is there a way I can do this so it is permanent (sure, I can replace the /sbin/iptables symlink, but after updating iptables, this change is reverted)?
2022-12-20 20:18:04 <f00860> The only somewhat clean way I can think of is creating a dummy package with a trigger which replaces the symlink every time the iptables package is updated
2022-12-20 20:18:24 <f00860> Is there a cleaner way like the update-alternative system in other distributions?
2022-12-20 22:06:56 <psykose> Piraty: lmao
2022-12-20 22:07:43 <psykose> f00860: no, you'd have to do that trigger as you described
2022-12-20 22:07:45 <psykose> but it should work
2022-12-20 22:09:57 <f00860> hm, ok :/
2022-12-21 02:30:38 <felipe> hey guys
2022-12-21 02:31:46 <felipe> I want to apply the custom patches from Nobara Project to Alpine Linux, because I have a Surface Book and it requires some patches that are available on this kernel
2022-12-21 02:32:01 <felipe> I was wondering if it would be possible to use the default Alpine kernel + the custom patches
2022-12-21 02:32:50 <felipe> or even how can I substitute the default Alpine Linux kernel to the Nobara one, that's here: https://copr-dist-git.fedorainfracloud.org/cgit/gloriouseggroll/nobara/kernel.git
2022-12-21 03:12:24 <psykose> if you don't know how then it's not possible
2022-12-21 03:12:59 <psykose> but tl;dr you have to build your own kernel with the patches you need from there (i assume the surface corrupted DMI one)
2022-12-21 07:32:21 <jh81> HAH!
2022-12-21 07:32:34 <jh81> i got my ultra-wide screen at the office working over usb-c
2022-12-21 07:32:53 <jh81> inc keyboard and mouse that are attached
2022-12-21 07:33:26 <jh81> does anyone happen to know what i need to do to get alpine working with wifi and KDE? 
2022-12-21 07:33:39 <jh81> i have a network icon in my tray.. but its not for wifi :|
2022-12-21 07:37:24 <psykose> assuming you have all the networkmanager stuff you need installed i'd assume it should work
2022-12-21 07:45:59 <jh81> thats what i thought :)
2022-12-21 07:46:18 <jh81> i have manually connected to wifi using wpa_passphrase command
2022-12-21 07:46:44 <jh81> so now im on Alpine.. at the office.. showing off my new monitor, connected through USB-C, WITH audio!
2022-12-21 07:46:53 <jh81> its like.. 
2022-12-21 07:47:37 <jh81> well no, cant say it
2022-12-21 07:47:45 <PureTryOut> Yeah you need networkmanager running with something like wpa_supplicant or preferably iwd
2022-12-21 07:49:57 <jh81> i think i go the iwd way.. but i thought that was "bad" ? i clearly remember someone saying that wasnt the good way to go
2022-12-21 07:50:14 <PureTryOut> without more context that statement is useless
2022-12-21 07:50:29 <PureTryOut> iwd is "the future" 😉
2022-12-21 07:52:36 <psykose> iwd does crypto in kernel space which makes it use less cpu for basically anything
2022-12-21 07:52:39 <psykose> amongst other improvements
2022-12-21 07:53:03 <psykose> it's more modern and preferred, but of course there's always a case where it might not work :)
2022-12-21 07:53:38 <psykose> the iwctl cli is quite bad (the way it draws to the screen/refreshes is quite broken?) but aside from that it's always worked for me
2022-12-21 07:53:58 <psykose> it also has the ability to make the iface and do dhcp but i never got that working :p
2022-12-21 07:54:02 <psykose> manage*
2022-12-21 07:54:34 <psykose> i've also gotten that iwctl interface to segfault a bunch of times but not in a reliable way
2022-12-21 07:54:42 <psykose> probably just usual readline mess
2022-12-21 07:54:43 <Misthios> and it connects way faster than wpa_supplicant
2022-12-21 07:54:56 <psykose> yeah the roaming is subsecond a lot of the time
2022-12-21 07:55:06 <Misthios> it makes eduroam actually usable
2022-12-21 07:58:00 <dwfreed> psykose: wpa_supplicant doesn't do all the crypto in userspace either
2022-12-21 07:59:08 <grawity> does it even matter though? the crypto that iwd or wpa_supplicant do is only relevant to handshakes
2022-12-21 08:00:29 <jh81> well.. iwd failed to connect to my network though :| "operation not supported" 
2022-12-21 08:00:38 <jh81> so i'll dive into that later :)
2022-12-21 08:00:55 <jh81> for now i'm just happily going to show off Alpine to my collegues :)
2022-12-21 08:18:48 <jh81> to top it off.. i added kscreen and now KDE shows full resolution! 
2022-12-21 10:35:41 <ham5urg> Is there an Ansible module to handle the repositories?
2022-12-21 10:43:09 <psykose> https://docs.ansible.com/ansible/latest/collections/community/general/apk_module.html ?
2022-12-21 11:04:44 <ham5urg> psykose, the repository parameter of the apk module is taking a specific address. Is there a way just to make the alpine-community-repo available, without specifying a version? Because at next dist-upgrade I would need to edit all scripts to adapt.
2022-12-21 11:06:03 <psykose> http://dl-cdn.alpinelinux.org/alpine/latest-stable/community
2022-12-21 11:06:53 <psykose> but note that when it does change (i.e. v3.16->v3.17) you're meant to use -a
2022-12-21 11:07:23 <psykose> so you probably want to set available: true always if you want to also set that and not have to tweak anything later
2022-12-21 12:20:28 <michalkotyla_> Hello everyone! You helped me a few days ago with configuring gateway/firewall on Alpine. I tried with this same on different installation and got problem: laptop connected to alpine over ethernet, I do not have any response when I try to ping 8.8.8.8. Communication between laptop and alpine is visible in tcpdump
2022-12-21 12:21:03 <michalkotyla_> any tips where can I look to found the problem? I have enabled forwarding, and firewall rules also should be ok - it works for me on another device
2022-12-21 12:24:33 <lopid> "ip route"
2022-12-21 12:35:37 <michalkotyla_> lopid: It return me two values: 169.254.0.0/16 and 192.168.1.0 where 192.168.1.1 is my firewall IP
2022-12-21 12:59:57 <lopid> no default route?
2022-12-21 13:02:36 <michalkotyla_> lopid: ah yes, sorry - I also have default route via 192.168.1.1, its my firewall IP and I can ping it 
2022-12-21 13:05:51 <lopid> does "ip route get 8.8.8.8" confirm it takes that route?
2022-12-21 13:07:33 <michalkotyla_> lopid: yes, i got `8.8.8.8 via 192.168.1.1 dev enp031f6 src 192.168.1.100 uid 999`
2022-12-21 13:07:56 <michalkotyla_> .100 is IP assigned to my laptop, received from dhcp working on alpine
2022-12-21 13:08:03 <lopid> where does "traceroute 8.8.8.8" stop?
2022-12-21 13:11:10 <michalkotyla_> I do not have installed traceroute on that offline laptop (ubuntu), but i see there is similar tool `tracepath` - it return me ` 1?: [LOCALHOST]`
2022-12-21 13:11:22 <michalkotyla_> next steps are `???` and "no reply"
2022-12-21 13:12:14 <lopid> is it only 8.8.8.8 that doesn't work?
2022-12-21 13:12:56 <michalkotyla_> it same situation when i try with 9.9.9.9
2022-12-21 13:13:19 <lopid> is your firewall blocking it?
2022-12-21 13:15:19 <michalkotyla_> lopid: I think it was unblocked by: `iptables -A FORWARD -i eth1 -j ACCEPT` and `iptables -t nat -A POSTROUTING -o eth0 -j MASQUERADE`
2022-12-21 13:15:52 <michalkotyla_> and this same iptables commands works for me a few days ago on another alpine device :\
2022-12-21 13:16:22 <lopid> on the firewall?
2022-12-21 13:16:40 <michalkotyla_> yes, I executed these commands on firewall, on alpine
2022-12-21 13:40:02 <jh81> Well that escalated quickly
2022-12-21 13:44:10 <lopid> can the firewall reach it?
2022-12-21 17:14:33 <scorpion2185[m]1> how do I config hibernation?
2022-12-21 17:14:55 <scorpion2185[m]1> loginctl?
2022-12-21 17:15:16 <ikke> echo mem >/sys/power/state
2022-12-21 17:17:05 <scorpion2185[m]1> I have freeze mem disk there already
2022-12-21 17:22:08 <scorpion2185[m]1> ikke: it did a sort of suspend but no hibernation
2022-12-21 17:23:53 <psykose> think you need a bunch more stuff for hibernation
2022-12-21 17:28:22 <scorpion2185[m]1>  I installed the pkg that provide loginctl and I can suped with that
2022-12-21 19:04:18 <scorpion2185[m]1> what more do I need then?
2022-12-21 19:06:06 <ikke> your bootloader needs to support it as well
2022-12-21 19:07:25 <scorpion2185[m]1> i use grub
2022-12-21 19:07:37 <scorpion2185[m]1> does grub need some config?
2022-12-21 19:08:18 <ikke> "You will need to point the kernel to your swap using the resume= kernel parameter, which is configured via the boot loader."
2022-12-21 19:08:35 <ikke> FDE complicates this
2022-12-21 19:13:13 <scorpion2185[m]1> After finishing your edits, run sudo grub-mkconfig -o /boot/grub/grub?
2022-12-21 19:23:17 <scorpion2185[m]1> do I need add resume to a resume hook and rebuild the initialram disk?
2022-12-21 19:30:11 <scorpion2185[m]1> ikke: besides grub do I need to config something else?
2022-12-21 19:41:17 <scorpion2185[m]1> https://tpaste.us/zQNw
2022-12-21 19:41:18 <scorpion2185[m]1>  wireplumber needed logind library and now
2022-12-21 19:42:07 <scorpion2185[m]1> is this a bug of logind?
2022-12-21 19:43:14 <bananicorn> just fyi wireplumber can also work without logind, info about this is on the wiki :)
2022-12-21 19:44:01 <bananicorn> In the pipewire article
2022-12-21 19:49:08 <scorpion2185[m]1> I see no dbus only
2022-12-21 19:50:50 <scorpion2185[m]1> it complains about systemd
2022-12-21 19:51:53 <scorpion2185[m]1> it seems that elogind is missing something?
2022-12-21 20:05:05 <psykose> i remember that error from wireplumber unless you forbid logind
2022-12-21 20:05:07 <psykose> i forget what caused it
2022-12-21 20:14:45 <scorpion2185[m]1> I started elogind and It complained about dbus, I disabled dbus and it started
2022-12-21 20:19:14 <scorpion2185[m]1> waht command do you use to hibernate?
2022-12-21 20:27:32 <psykose> i never use hibernation
2022-12-21 20:27:57 <panekj> imagine hibernating linux
2022-12-21 20:28:45 <ikke> Low level it would be writing disk to /sys/power/sleep 
2022-12-21 20:30:23 <psykose> i'm sure it's useful if you keep a lot of state and also travel and so don't want to merely sleep or something
2022-12-21 20:30:44 <psykose> but a) i don't keep any window layout or any state of my current session and b) i don't go outside
2022-12-21 20:31:11 <psykose> so if i'm not at my computer i just turn it off
2022-12-21 21:05:58 <scorpion2185[m]1> GRUB_CMDLINE_LINUX_DEFAULT="resume=UUID=<UUID_of_partition>    ,, where that is the swap? and then do I use mkinitfs?
2022-12-21 21:06:06 <scorpion2185[m]1> and rebuild grub config
2022-12-22 10:46:56 <daedric[m]> Greetings
2022-12-22 11:01:24 <jh81> heh
2022-12-22 12:13:02 <WhyNotHugo_> I need linux-firmware-other on my installation media for network to work. Where can I download the package into a thumb drive? How do I tell apk "install this package from this path"?
2022-12-22 12:13:12 <WhyNotHugo_> I don't see a link to the package tarball itself here https://pkgs.alpinelinux.org/package/v3.17/main/x86_64/linux-firmware-other
2022-12-22 12:13:56 <ikke> You can do apk fetch linux-firmware-other to get apk to download the apk
2022-12-22 12:14:06 <ikke> They're available on the mirrors: https://dl-cdn.alpinelinux.org/alpine/
2022-12-22 12:16:04 <ncopa> https://dl-cdn.alpinelinux.org/alpine/v3.17/main/x86_64/linux-firmware-other-20221214-r0.apk
2022-12-22 12:16:27 <WhyNotHugo_> Ah, right, that should have been obvious, thanks!
2022-12-22 12:17:00 <ncopa> mount /dev/sdX1 /mnt ; apk add /mnt/linux-firmware-other*; umount /mnt
2022-12-22 12:17:56 <WhyNotHugo_> Ah, thanks!
2022-12-22 12:50:32 <WhyNotHugo_> So the alpine-extended image doesn't include iw or wpa_supplicant. I can download the packages and their whole dependency trees into a thumb drive, but that sounds like a huge chore.
2022-12-22 12:50:37 <WhyNotHugo_> Am I missing something here?
2022-12-22 12:50:47 <WhyNotHugo_> setup-interfaces won't work due to them being absent.
2022-12-22 13:13:42 <Guest28> Yeah I was unpleasantly surprised by that too. Extended image is like 700 mb but still doesn't include stuff like rfkill or man pages. I wonder what it actually does include lol
2022-12-22 13:30:47 <satmd> rfkill is just an interface to some (simple) files in /sys
2022-12-22 13:49:04 <frag> is there an overview of how to config with /sys and similar files?
2022-12-22 14:02:00 <Fxzx> Can I completely block a package from updating?
2022-12-22 14:03:41 <aron> Fxzx: I do this: apk add pkg=ver
2022-12-22 14:04:35 <Fxzxmic> is this working?
2022-12-22 14:06:28 <Fxzxmic> I use the locally compiled busybox, so I cannot update it first and then go back to the origin version.
2022-12-22 14:07:34 <Fxzxmic> I want to exclude it completely from the update.
2022-12-22 14:07:53 <Guest28> Isn't busybox just one single statically compiled file? Couldn't you just have multiple versions and just place the busybox executable you want in the necessary location ?
2022-12-22 14:13:08 <Fxzxmic> I use bash to log in, but there is no bash in the alpine version, which will cause my login failure.
2022-12-22 15:34:55 <elly> you have bash as part of busybox?
2022-12-22 15:38:31 <lopid> ash
2022-12-22 15:40:35 <ikke> fx
2022-12-22 15:40:41 <ikke> They're gone 
2022-12-22 15:46:28 <minimal> ikke: well no-one replied to him within the statutory 30 seconds so what do you expect ;-)
2022-12-22 16:02:34 <ikke> minimal: we obviously need to up our game
2022-12-22 16:22:36 <scorpion2185[m]1> echo disk >/sys/power/state
2022-12-22 16:22:36 <scorpion2185[m]1> ash: write error: No such device
2022-12-22 16:28:47 <lopid> is it turned on?
2022-12-22 16:45:56 <scorpion2185[m]1> what?
2022-12-22 16:46:25 <scorpion2185[m]1> it shows 100% usage of swap all the time
2022-12-22 16:46:46 <scorpion2185[m]1> how do i set a  keybaor dlayout on bspwm? i can start it from lightdm
2022-12-22 16:48:09 <scorpion2185[m]1> kbd layout it is fine on the ttyas
2022-12-22 16:48:14 <scorpion2185[m]1> ttys
2022-12-22 16:52:50 <icarus> I can't seem to modprobe wireguard with linux-rpi 5.15.55-0-rpi aarch64 after a recent upgrade and it seems that the DKMS modules are absent from the repositories, do I need to switch over to linux-lts?
2022-12-22 16:53:52 <ikke> icarus: should be available: https://pkgs.alpinelinux.org/contents?file=*wireguard*&path=%2Flib%2Fmodules%2F*&name=linux-rpi&branch=edge
2022-12-22 16:54:00 <ikke> icarus: did you reboot after upgrading the kernel?>
2022-12-22 16:55:06 <icarus> yes, modprobe only failed after the reboot
2022-12-22 16:55:25 <icarus> the module version isn't matching the output from uname, will try to upgrade linux-rpi
2022-12-22 17:02:27 <icarus> it looks like it's running off of an older version of the kernel, 5.15.55 is running while 5.15.83-r0 is installed, I think it might be a quirk of my raspi install
2022-12-22 17:02:53 <minimal> are you using a run-from-ram or Sys-mode install?
2022-12-22 17:03:12 <icarus> sys-mode
2022-12-22 17:05:02 <scorpion2185[m]1> does alpine use .xsession or .xinitrc?
2022-12-22 17:09:41 <scorpion2185[m]1> when I log in from lightdm , how can I autostart things for X? and in general
2022-12-22 17:10:09 <ikke> I do it in .xinitrc
2022-12-22 17:10:41 <minimal> icarus: so I don't see how when you last upgraded your linux-rpi (or linux-rpi4) package and rebooted that the wireguard module could be a different version than the running kernel as the module comes with the linux-rpi package.
2022-12-22 17:11:18 <scorpion2185[m]1> do I need a exec  in the end?
2022-12-22 17:11:31 <ikke> scorpion2185[m]1: yes
2022-12-22 17:11:33 <minimal> icarus: what Alpine version are you running? I don't see a 5.15.55-0-rpi in Alpine Edge, 3.17.x, 3.16.x, or 3.15.x
2022-12-22 17:11:36 <icarus> I think it's loading the kernel from /boot on the vfat partition, which doesn't match up with apkinfo
2022-12-22 17:12:35 <icarus> 5.15.55 is outdated, iiirc it was installed on 3.16 and brought up to 3.17
2022-12-22 17:12:38 <ikke> icarus: do you have that mounted?
2022-12-22 17:12:44 <ikke> /boot should be mounted
2022-12-22 17:13:01 <icarus> yep, mounted, present in fstab and visible in df
2022-12-22 17:13:16 <minimal> icarus: so did you do the 3.16 -> 3.17 upgrade correctly? i.e. running "apk -Ua upgrade" ?
2022-12-22 17:13:55 <icarus> hmm iirc it was apk update && apk upgrade, never used the -Ua flags for any of my 3.17 upgrades
2022-12-22 17:14:17 <minimal> icarus: so then your upgrade was likely not performed correctly
2022-12-22 17:15:13 <minimal> icarus~: if you look at the 3.17.0 Release Notes it states to use "--available" for "apk upgrade" ("-a" is the short-form of this)
2022-12-22 17:17:37 <icarus> I see that now, looks like I have quite a few more machines to fix :')
2022-12-22 17:41:14 <Saklad5> There's a specific way to notify an APK maintainer that an update is available, right?
2022-12-22 17:41:28 <Saklad5> I swear i stumbled across it recently, but I can't find it now
2022-12-22 17:43:14 <icarus> from pkgs.alpinelinux.org you can click on the version to flag it as out of date
2022-12-22 17:43:50 <scorpion2185[m]1> why my swap is always 100%? ram is 10-20%
2022-12-22 17:44:19 <Saklad5> @icarus Thanks
2022-12-22 18:02:36 <elly> scorpion2185[m]1: linux aggressively swaps cold pages out so it can have more free RAM either available for hot workloads or for disk cache of hotter pages
2022-12-22 18:03:31 <elly> if you want to change that behavior /proc/sys/vm/swappiness takes a number between 0 and 100
2022-12-22 18:05:02 <elly> where 0 = "swap when OOM only" and 100 = "use swap very freely"
2022-12-22 18:07:30 <Schimon> Using SXMO (edge) I've examined macchanger on Nexus 5:`macchanger -A wlan0` will show a new MAC but won't change to something else.
2022-12-22 18:07:30 <Schimon> Same for `-a` and `-r`.
2022-12-22 18:07:51 <Schimon> May someone confirm this? I've made this test on postmarketOS
2022-12-22 18:39:02 <icarus> I have resolved my comedy of errors - I overwrote my /boot with the new rootfs download, but that got me stuck on 15.79 instead of 15.83, but manually replacing vmlinuz-rpi from the apk into /boot/boot did the trick. Thanks for the help, I'll remember to -a from now on :)
2022-12-22 19:09:44 <scorpion2185[m]1> elly: I have 60
2022-12-22 19:10:25 <scorpion2185[m]1> ERROR: APKINDEX.tar.gz: UNTRUSTED signature
2022-12-22 19:10:25 <scorpion2185[m]1> Updating the apps/x86_64 repository index...
2022-12-22 19:11:08 <scorpion2185[m]1> why is it untrusted? IIRC I didn't do anything on another device
2022-12-22 19:11:23 <ikke> where does this repo come from?
2022-12-22 19:11:31 <ikke> apps is not an alpine linux repo
2022-12-22 19:12:41 <scorpion2185[m]1> it the repo in my home ~/packages/apps
2022-12-22 19:13:02 <ikke> So did you add your public key to /etc/apk/keys?
2022-12-22 19:14:39 <scorpion2185[m]1> no
2022-12-22 19:14:45 <ikke> then that's why
2022-12-22 19:16:12 <scorpion2185[m]1> ~/.abuild/...rsa.pbu?
2022-12-22 19:16:16 <ikke> yes
2022-12-22 19:19:37 <ifreund> is it normal to get useless stack traces from abort() on alpine? https://0x0.st/o5f6.txt
2022-12-22 19:20:04 <ifreund> Compared to a segfault https://0x0.st/o5fl.txt for https://0x0.st/o5fU.c
2022-12-22 19:27:19 <bosco> Hey, I'm trying to set up gnome-keyring on my system, reading the arch wiki it sways that I should edit /etc/pam.d/login, but alpine doesn't seem to have that scheme for PAM, which one is the file I should edit?
2022-12-22 19:33:28 <Ermine> bosco: 1) you are reading ArchWiki and trying to apply this to Alpine; 2) You need to install shadow-login or util-linux-login, since default busybox login does not support PAM
2022-12-22 19:38:50 <bosco> would installing either of those packages be enough or will something else need to be put in place to enable PAM?
2022-12-22 19:39:20 <bosco> I can't find a lot of info on the alpine wiki other than a steam article 
2022-12-22 19:46:25 <Ermine> bosco: at least installing one of those packages would install /etc/pam.d/login
2022-12-22 19:46:44 <bosco> Ok, thanks
2022-12-22 19:47:19 <bosco> going to reboot to see if it took place
2022-12-22 19:52:27 <jh81> its a shame these people dont take the time to actually create an article :)
2022-12-22 19:56:55 <minimal> jh81: the "problem" is not so much writing a wiki article, its doing the prep work for it and also keeping the article accurate as time (and Alpine releases) go by
2022-12-22 19:58:04 <jh81> true, true
2022-12-22 20:02:46 <invoked> also dealing with random changes/spam that people might make. nothing against wikis, but i'd rather have some source format to work with, eg markdown that generates a gitbook type of site.
2022-12-22 20:03:22 <invoked> then you can do a pr/mr type of thing
2022-12-22 20:03:30 <jh81> the horror of it..
2022-12-22 20:04:20 <jh81> people say "use git, its SOOOOO easy!!" and yet.. i always have problems with it. It's a never ending fight
2022-12-22 20:06:09 <Ermine> I mean, if you use thing from other distribution's documentation, you should be careful, since there are differences between the distributions.
2022-12-22 20:07:21 <Ermine> The main problem in making documentation is not format or the tool set. It is about gaining material and understanding the subject you are documenting.
2022-12-22 20:07:25 <jh81> i think the main "problem" (notice the quotes) is that Alpine doesnt see itself as a desktop distro.. but more people are discovering it as such :)
2022-12-22 20:07:33 <jh81> true true
2022-12-22 20:08:00 <invoked> i'd argue it is at least partially about tools and how people iterate on the documentation with those tools.
2022-12-22 20:08:29 <invoked> there is a substantial quality difference i've seen between docs that are generated by some toolchain and the average wiki
2022-12-22 20:08:36 <Ermine> Alpine is not designed as desktop distro (like Ubuntu). The main direction is containers and such I guess
2022-12-22 20:08:46 <invoked> arch wiki is decent but it's had a lot of effort put into it.
2022-12-22 20:09:02 <minimal> Ermine: nope, Alpine predates Docker - people seem to keep repeated that "fiction"
2022-12-22 20:09:09 <Ermine> invoked: you may want to take a look at Alpine Handbook, maybe this is a format what you want
2022-12-22 20:09:16 <Ermine> minimal: ok, my fault
2022-12-22 20:09:34 <minimal> Alpine may have become more popular/generally visible because of Docker
2022-12-22 20:09:38 <jh81> i think a good and up to date handbook is very valued
2022-12-22 20:09:49 <jh81> and special cases can go into a wiki
2022-12-22 20:11:46 <invoked> i think i heard alpine entering the conversation more around when aws picked up
2022-12-22 20:12:31 <invoked> that's anecdotally when i saw it gaining traction
2022-12-22 20:12:45 <jh81> i knew "of" Alpine for quite some time, but only when i started looking into building my own firewall/switch/accespoint software did i actually get familiar with Alpine
2022-12-22 20:12:50 <Ermine> minimal: maybe this is something written my mccay on the wiki that I've read. Double sorry if it was that way.
2022-12-22 20:13:02 <jh81> and then ofcourse the whole security aspect, makes it perfect for a small, secure desktop environment
2022-12-22 20:14:09 <jh81> i've been using debian ever since man figured out how to make fire
2022-12-22 20:14:55 <jh81> then last year i had to choose between alpine and arch, so i figured i'd do arch for desktop and alpine for server
2022-12-22 20:15:15 <jh81> but i really want one single platform for everything :)
2022-12-22 20:40:00 <jh81> i think its a shame Alpine doesnt have a forum :( and yes, there's a reddit, but i'm as old as the devil. 
2022-12-22 20:40:45 <ikke> There was a forum, but it was did, just as the users mailing kist
2022-12-22 20:40:47 <invoked> it has a forum, it's called a mailing list
2022-12-22 20:40:47 <ikke> list*
2022-12-22 20:40:57 <ikke> it was dead*
2022-12-22 20:41:00 <jh81> ;-)
2022-12-22 20:41:39 <invoked> i'm waiting for someone to suggest creating a discord server
2022-12-22 20:41:51 <jh81> eww no
2022-12-22 20:42:18 <jh81> discord is like irc. its good for diurect communication. not something like a forum where you can ask a question, and then later return to it
2022-12-22 20:42:57 <ikke> And it's proprietary, not really a fit for alpine linux
2022-12-22 20:43:16 <jh81> oh! i see alpine is on matodon too :)
2022-12-22 20:43:22 <Ermine> It doesn't run natively on Alpine
2022-12-22 20:43:24 <ikke> :)
2022-12-22 20:43:32 <jh81> mastodon.. pff.. i need to go to bed
2022-12-22 20:44:51 <jh81> ohh.. that malinglist! it looks indeed much like a forum
2022-12-22 21:00:53 <invoked> email, the first social network
2022-12-22 21:03:58 <minimal> invoked: wasn't that "write"/"talk" on Unix-like systems and the VMS equivalent?
2022-12-22 21:06:44 <jh81> email.. if there's anything i almost hate as much as Microsoft, well then its email
2022-12-22 21:10:47 <panekj> invoked: there is Alpine Discord server
2022-12-22 21:26:30 <invoked> roger that
2022-12-22 21:26:45 <invoked> email's fine, it's people that suck
2022-12-22 21:37:59 <panekj> apk policy -X http://dl-cdn.alpinelinux.org/alpine/v3.12/main
2022-12-22 21:37:59 <panekj> WARNING: Ignoring http://dl-cdn.alpinelinux.org/alpine/v3.12/main: No such file or directory
2022-12-22 21:38:05 <panekj> I'm slightly confused
2022-12-22 22:34:03 <iggy> maybe a bad mirror?
2022-12-22 22:46:13 <dwfreed> dl-cdn.alpinelinux.org isn't a mirror, though
2022-12-22 22:47:14 <dwfreed> (I mean isn't a mirror as in it doesn't point to a list of community mirrors)
2022-12-22 22:53:35 <panekj> it is a mirror, though
2022-12-22 23:07:03 <dwfreed> it's a mirror in the sense that it's a copy of alpine linux repos, yes
2022-12-22 23:07:22 <dwfreed> but it's not a community run mirror, and thus can't be "bad" unless something is really broken
2022-12-22 23:07:47 <panekj> what is community "bad"
2022-12-22 23:08:40 <dwfreed> community sysadmin fails to maintain their mirror, or they have a hardware failure, or a bad sync implementation
2022-12-22 23:08:55 <panekj> that seems like a "really broken" mirror
2022-12-22 23:09:55 <dwfreed> my point is that dl-cdn uses alpine's master repo server as source, because it's just a caching CDN, so as long as the master server is fine, it won't ever be bad normally
2022-12-22 23:10:11 <dwfreed> it's not just a DNS name pointing to a bunch of community mirrors
2022-12-22 23:10:29 <panekj> cache invalidation tho.
2022-12-22 23:11:10 <panekj> i mean, it was bad just recently when 3.17 was released
2022-12-22 23:11:12 <dwfreed> for the kind of caching fastly does, that is mostly a solved problem
2022-12-22 23:11:24 <panekj> and it wasn't master server fault
2022-12-23 01:35:05 <yjftsjthsd> Is there a way to start things on boot, but only in the background? Ex. I want ntpd to run, but I don't want it holding up me getting a getty
2022-12-23 01:42:38 <pj[m]> yjftsjthsd: https://ptrcnull.me/posts/openrc-async-services/
2022-12-23 01:51:31 <yjftsjthsd> Nice, thanks. Funny, I'd just looked at the wiki and decided that runlevels didn't look like a nice solution:P But that looks reasonable when it's laid out
2022-12-23 02:16:32 <minimal> yjftsjthsd: generally the idea of waiting for ntpd/chrony to get and set the time is so that the time doesn't not change for other services started before the clock is set
2022-12-23 02:18:20 <pj[m]> but my 4 second boot time
2022-12-23 02:21:26 <minimal> pj[m]~: how do you know it's a 4 second boot time? the clock could have been adjusted to give that impression ;-)
2022-12-23 04:45:03 <yjftsjthsd> Yeah, I get not making it default, but I'm willing to risk it / accept the trade for my system
2022-12-23 12:52:05 <avuton> How can I change the diskless root size during boot (after boot: mount -o remount,size=256M /)?
2022-12-23 12:52:52 <avuton> I thought it would be overlay_size under /etc/conf.d/modloop but that appears to have no effect.
2022-12-23 15:24:34 <brebs> Hi, I received an email that the swi-prolog package has been flagged out-of-date. I've switched swi-prolog to use the standard stable branch, now that major version 9 has been released.
2022-12-23 15:24:34 <brebs> https://release-monitoring.org/project/236197/
2022-12-23 15:24:34 <brebs> Relevant URLs:
2022-12-23 15:24:34 <brebs> https://repology.org/project/swi-prolog/packages
2022-12-23 15:24:34 <brebs> Hopefully Anitya can be pointed at the stable branch, as done with repology?
2022-12-23 15:24:36 <brebs> https://www.swi-prolog.org/download/stable
2022-12-23 15:24:38 <brebs> https://www.swi-prolog.org/download/devel
2022-12-23 15:24:40 <brebs> https://www.swi-prolog.org/ChangeLog?branch=stable
2022-12-23 15:24:42 <brebs> https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/testing/swi-prolog
2022-12-23 15:24:44 <brebs> Other distros:
2022-12-23 15:24:48 <brebs> https://github.com/void-linux/void-packages/tree/master/srcpkgs/swi-prolog
2022-12-23 15:24:50 <brebs> https://archlinux.org/packages/community/x86_64/swi-prolog/
2022-12-23 15:25:54 <psykose> anitya can be edited by anyone
2022-12-23 15:25:55 <ikke> You can change that in anitya (at least, that's where it needs to be changed)
2022-12-23 15:26:56 <brebs> The email said "If the provided information is incorrect, please let us know on IRC or alpine-infra@alpinelinux.org" - I tried subscribing to infra and emailing, but keep getting "500 Rejected. You are not allowed to post to this list."
2022-12-23 15:27:08 <psykose> in this specific case you would probably want to make a new package on anitya
2022-12-23 15:27:24 <psykose> and version filter the 9.1 versions
2022-12-23 15:27:52 <psykose> https://release-monitoring.org/project/236197/ prior as example
2022-12-23 15:27:56 <psykose> login is fedora accounts
2022-12-23 15:28:31 <psykose> then you would have to edit the alpine link there and point it to something that doesn't exist (can't remove mappings lol), then link it in the new one
2022-12-23 15:29:08 <psykose> doesn't make sense to edit that one as 9.1 is indeed the latest version
2022-12-23 15:33:30 <brebs> 9.1.x is "devel", whereas 9.0.x is stable, for swi-prolog. So version-filtering 9.1 sounds good, will try that...
2022-12-23 15:34:19 <psykose> ah
2022-12-23 15:34:25 <psykose> sure then better to prerelease the existing
2022-12-23 15:34:29 <psykose> sigh
2022-12-23 15:34:36 <psykose> damned oddversioners
2022-12-23 15:34:38 <psykose> when will they stop
2022-12-23 15:34:46 <psykose> 
ACTION rants at cloud

2022-12-23 15:34:55 <ikke> heh
2022-12-23 15:35:11 <psykose> edited
2022-12-23 15:35:19 <brebs> Haha. What's the most convenient versioning scheme for you guys, out of interest?
2022-12-23 15:35:25 <ikke> semver
2022-12-23 15:38:04 <ikke> (bonus points for actual backports for serious issues)
2022-12-23 15:40:37 <Piraty> zerover for life
2022-12-23 15:40:43 <brebs> Cool, thanks for your help.
2022-12-23 15:40:48 <Piraty> no problem
2022-12-23 15:40:55 <Piraty> be sure to prefix as many zero as you can
2022-12-23 15:41:11 <Piraty> bonus points for a..z suffixes
2022-12-23 15:41:13 <psykose> decrementing numbers are better
2022-12-23 15:41:38 <Piraty> asymptically approaching 0, great idea!
2022-12-23 15:41:42 <Piraty> inverse 0ver
2022-12-23 15:46:21 <psykose> (my actual serious answer is semver for libraries and more or less literally anything that doesn't have gotchas like oddversioning or weird letters in the version for anything else)
2022-12-23 15:46:58 <psykose> the latter because non-libraries don't really have meaningful abis/api-styled things and without that there is no point to the difference aside from 'big number go up'
2022-12-23 15:47:03 <psykose> everything becomes a breaking change
2022-12-23 15:47:58 <psykose> and the style of random applications doing "semver but 'small breaks' get a pass" is meaningless in the end and all the numbers don't mean shit
2022-12-23 15:48:09 <psykose> so it's just number go up
2022-12-23 15:54:14 <ikke> as long as development releases are explicitly marked as such
2022-12-23 15:55:59 <psykose> yep
2022-12-23 15:59:16 <invoked> imo where it matters is signaling to the user 'ok, pay attention to this, it might mean config file changes or something', but if you can't get everyone to agree, it doesn't mean much
2022-12-23 16:10:37 <scorpion2185[m]1> will hibernate be supported?
2022-12-23 16:12:30 <psykose> what makes it not supported
2022-12-23 16:14:00 <scorpion2185[m]1> I cannot hibernate
2022-12-23 16:14:25 <psykose> well, i think you'll have to patch your dna? could copy the bear implementation..
2022-12-23 16:14:28 <psykose> (sorry, bad joke)
2022-12-23 16:14:34 <psykose> i forget, what did you try so far
2022-12-23 16:15:16 <scorpion2185[m]1> the installer could set it automatically and there is no wiki
2022-12-23 16:15:47 <scorpion2185[m]1> I edited /etc/default/grub and I installed and enabled elogind
2022-12-23 16:17:19 <scorpion2185[m]1> GRUB_CMDLINE_LINUX_DEFAULT="modules=sd-mod,usb-storage,ext4 cryptroot=UUID=333b1db5-8807-4298-ac53-77bfb8f2cf86 cryptdm=root quiet rootfstype=ext4 resume=UUID=1ea336b2-06dd-4800-bbda-f5c6a0bbe107"
2022-12-23 16:17:42 <scorpion2185[m]1> I have lvm on luks
2022-12-23 16:46:53 <scorpion2185[m]1> how can I make a openrc service for autologin?
2022-12-23 16:53:38 <grawity> tty autologin?
2022-12-23 17:18:49 <minimal> avutom: the initramfs' init accepts the "root_size=<value>" cmdline parameter, although this is deprecated and refers to using the "rootflags=size=<value>" cmdline instead
2022-12-23 17:19:01 <minimal> avuton: ^^^
2022-12-23 18:10:36 <Guest28> I'm going to attempt playback of audio on my X220 (yeah, wish me luck). Reading https://wiki.alpinelinux.org/wiki/PipeWire - Isn't it a prerequisite to first go through the steps of configuring ALSA as described here: https://wiki.alpinelinux.org/wiki/PipeWire ?
2022-12-23 18:10:43 <Guest28> Err, I meant https://wiki.alpinelinux.org/wiki/ALSA
2022-12-23 18:12:54 <Ermine> Guest28: no, I don't think so. PipeWire should handle it all itself. Please try without configuring ALSA first and let me know if something goes wrong (or if you succeed).
2022-12-23 18:13:12 <Guest28> Ok, I'm going in!
2022-12-23 18:29:27 <Guest28> It doesn't say anywhere if pipewire-alsa package is required, so I'll regard it as optional for now.
2022-12-23 18:30:45 <Ermine> Guest28: pipewire-alsa is indeed optional, it is needed for clients which use alsa directly otherwise
2022-12-23 18:32:03 <Guest28> Ah I see, so is it safe to assume that alsa-lib (1.2.8-r0) is then how PW actually interacts with ALSA?
2022-12-23 18:36:20 <scorpion2185[m]1> grawity: autologin package, it doesn't have a openrc service
2022-12-23 18:37:27 <Guest28> Hmm, thought I had dbus, but dbus-launch: command not found says otherwise lol
2022-12-23 18:37:49 <Ermine> Guest28: it is in dbus-x11 package.
2022-12-23 18:40:05 <Guest28> Couldn't exec /usr/libexec/pipewire-launch: No such file or directory - Seems that type -a pipewire-launch also doesn't find it.
2022-12-23 18:41:03 <scorpion2185[m]1> https://git.sr.ht/~kennylevinsen/autologin/tree/master/item/autologin.service
2022-12-23 18:41:25 <Ermine> Guest28: pipewire-launch_er_
2022-12-23 18:41:54 <Guest28> Ah, I'll update the wiki.
2022-12-23 18:42:21 <Ermine> Guest28: I can do this for you
2022-12-23 18:42:31 <Guest28> Oh cool, thx!
2022-12-23 18:42:49 <scorpion2185[m]1> I needed to start elogind
2022-12-23 18:42:50 <scorpion2185[m]1> to start wireplumber
2022-12-23 18:43:17 <scorpion2185[m]1> no dbus for wayland?
2022-12-23 18:44:39 <Ermine> scorpion2185[m]1: you don't need elogind for wireplumber
2022-12-23 18:45:00 <Ermine> scorpion2185[m]1: and dbus has nothing to do with wayland.
2022-12-23 18:45:14 <Ermine> dbus-launch will work independently of the windowing system.
2022-12-23 18:45:53 <Ermine> Guest28: if you need some sounds for your test, here are some freely licensed sound files: https://mirrors.edge.kernel.org/pub/linux/kernel/SillySounds/
2022-12-23 18:46:05 <Guest28> Ermine: Not sure if this is good: https://paste.debian.net/plain/1265078/ - But `wpctl status` says `Could not connect to PipeWire`
2022-12-23 18:46:41 <Ermine> Guest28: did you launch dbus system session (rc-service dbus start) ?
2022-12-23 18:46:45 <Guest28> Excellent audio tests, I have some of my own white noise and sine sweeps too :)
2022-12-23 18:47:56 <Guest28> No, just installed that dbus-x11 package and did that dbus-launch command.
2022-12-23 18:48:58 <Ermine> you need system-wide dbus too
2022-12-23 18:49:24 <Ermine> I have a draft article on dbus here, if you need: https://wiki.alpinelinux.org/wiki/User:Ermine/D-Bus
2022-12-23 18:50:22 <Guest28> After doing doas rc-service dbus start, I get this: https://paste.debian.net/plain/1265080/
2022-12-23 18:50:38 <Guest28> Ah thanks, I was looking for a dbus article on the wiki prior to this :)
2022-12-23 18:51:34 <Guest28> I'm guessing I have to logout and log back in again.
2022-12-23 18:52:11 <scorpion2185[m]1> it seems you are missing the elogind module
2022-12-23 18:52:50 <scorpion2185[m]1> wireplumber-0.4/libwireplumber-module-logind.so: No such file or directory
2022-12-23 18:53:29 <Ermine> scorpion2185[m]1: no, this is not a problem
2022-12-23 18:53:41 <Ermine> Guest28: do you have XDG_RUNTIME_DIR set?
2022-12-23 18:53:49 <Guest28> Isn't elogind dependent on systemd though? It would make sense if I'm missing that since Alpine is not shipping with systemd by default.
2022-12-23 18:54:05 <grawity> elogind is specifically the "systemd-free" fork of logind
2022-12-23 18:54:06 <scorpion2185[m]1> I had that error, there is a package wiht that file
2022-12-23 18:54:16 <grawity> but it should be optional
2022-12-23 18:54:25 <grawity> "could not load mandatory module "libpipewire-module-protocol-native":" seems more serious though
2022-12-23 18:54:33 <Guest28> $XDG_RUNTIME_DIR is empty here, but then again I'm not intending to run pactl, do I still need to set it manually?
2022-12-23 18:55:00 <grawity> it's not just for pactl but also for the programs that play audio -- they use the same socket as pactl does
2022-12-23 18:56:34 <Ermine> Guest28: mkdir /run/user/$UID and set XDG_RUNTIME_DIR=/run/user/$UID
2022-12-23 18:56:36 <scorpion2185[m]1> wireplumber-logind
2022-12-23 18:57:31 <Ermine> scorpion2185[m]1: again, missing elogind will not prevent pipewire and wireplumber from starting
2022-12-23 18:59:53 <scorpion2185[m]1> how can you close a tty?
2022-12-23 19:00:04 <scorpion2185[m]1> or restart
2022-12-23 19:00:37 <Ermine> you can kill relevant agetty
2022-12-23 19:02:54 <scorpion2185[m]1> how do I specify for example agetty 1?
2022-12-23 19:03:39 <Ermine> look process table (ps or top) for agetty running on tty1 and kill it
2022-12-23 19:03:57 <scorpion2185[m]1> elogind worked for me so should it be a solution ?
2022-12-23 19:04:32 <Ermine> OTOH I would step back and think why if I was you.
2022-12-23 19:05:06 <Ermine> scorpion2185[m]1: Guest28 tries to run it on X220, no need to pile any extra daemons without real need.
2022-12-23 19:06:10 <scorpion2185[m]1> i am just saying that is a method I wanted to use loginctl anyway
2022-12-23 19:06:17 <scorpion2185[m]1> OTOH?
2022-12-23 19:06:44 <Ermine> On The Other Hand
2022-12-23 19:07:16 <scorpion2185[m]1> IDK I enabled elogind and I disabled dbus on PW
2022-12-23 19:10:07 <scorpion2185[m]1> (it was complaining about dbus after elogind)
2022-12-23 19:11:10 <scorpion2185[m]1> with agettty did you mean /sbin/getty tty3 ?
2022-12-23 19:11:54 <Ermine> Yes.
2022-12-23 19:12:44 <Ermine> Guest28: how does it go?
2022-12-23 19:13:09 <scorpion2185[m]1> I can see 3,4,5,6. I am on 2 and I wanted to kill 1 because there is madnees there
2022-12-23 19:13:41 <Guest28> Currently sidetracked with vim because apparently if you install it via Nix then all hell breaks loose if you have plugins that require Python -_-
2022-12-23 19:14:07 <Ermine> scorpion2185[m]1: what happens there? You should deal with the reason and not with the symptom.
2022-12-23 19:14:48 <scorpion2185[m]1> why can't i see active ttys?
2022-12-23 19:20:58 <Guest28> https://paste.debian.net/plain/1265083/ - I'm guessing I have to logout, login and do export $(dbus-launch) before executing startx ?
2022-12-23 19:21:33 <Saklad5> Does anyone have any experience moving an Alpine Linux installation from a single partitioned disk to separate disks?
2022-12-23 19:22:20 <Saklad5> I used the "dd" command to move each partition to a disk of the same size and filesystem, but it doesn't seem to boot
2022-12-23 19:23:02 <Saklad5> For additional context: I set up the virtual image of Alpine Linux as a custom ISO with Linode, and I'm now trying to get their backup service working. It doesn't support partitioned volumes
2022-12-23 19:24:07 <ikke> You need to update the bootloader to tell it what partition to boot
2022-12-23 19:24:16 <Saklad5> I checked, and it seems to be using a UUID
2022-12-23 19:24:19 <ikke> ok
2022-12-23 19:24:22 <Saklad5> And that UUID is the same on the duplicated volume
2022-12-23 19:24:27 <Saklad5> According to lsblk
2022-12-23 19:24:31 <Saklad5> So it shouldn't matter, right?
2022-12-23 19:24:50 <Saklad5> I'm clearly wrong SOMEWHERE, so stop me if I'm wrong
2022-12-23 19:24:52 <minimal> Saklad5: and are both the original and duplicate FS available at the same time?
2022-12-23 19:24:54 <ikke> Do you get a specific error message?
2022-12-23 19:26:01 <Ermine> Guest28: you can killall dbus-session, pipewire and wireplumber. You need to chown -R 1000:1000 /run/user/1000
2022-12-23 19:26:39 <scorpion2185[m]1> autolgin messed up river session
2022-12-23 19:26:49 <Saklad5> minimal: When they are, it seems to choose a drive to boot non-deterministically
2022-12-23 19:27:00 <Ermine> Guest28: also don't dbus-launch as root
2022-12-23 19:27:26 <Saklad5> @ikke: When I tell linode the "root device" is the boot volume, Lish just says:
2022-12-23 19:27:37 <Saklad5> "Booting from Hard Disk...
2022-12-23 19:27:41 <Saklad5> Boot error
2022-12-23 19:27:41 <minimal> Saklad5:L yes that was what I was getting to - the "UU" of "UUID" standards for Universally *Unique* which of course it isn't if there are 2 fs' with the same UUID ;-)
2022-12-23 19:27:43 <Saklad5> "
2022-12-23 19:27:57 <ikke> Saklad5: you might be missing the bootloader then
2022-12-23 19:28:02 <Guest28> Ermine: Says no processes were killed.
2022-12-23 19:28:05 <Ermine> Guest28: and you may want to run dbus with export $(dbus-launch) so you have only one instance of dbus session buss
2022-12-23 19:28:12 <Ermine> Guest28: that's fine
2022-12-23 19:28:49 <minimal> Saklad5: so if you are boting by UUID and there's 2 fs' with same UUID then how do you expect to control which of the 2 of them is used?
2022-12-23 19:29:37 <ikke> minimal: it seems it does not even get to that part
2022-12-23 19:29:41 <Saklad5> @minimal: I don't expect to: that's just while troubleshooting
2022-12-23 19:29:57 <Saklad5> The issue happens when only the new non-partitioned disks are attached
2022-12-23 19:30:07 <Saklad5> When the old portioned disk is attached, it successfully boots off that
2022-12-23 19:30:24 <Saklad5> partitioned*
2022-12-23 19:30:28 <ikke> Saklad5: the bootloader is located at the first 2048?k of the disk
2022-12-23 19:30:32 <Saklad5> Which one
2022-12-23 19:30:42 <ikke> if you copy the partitions, you probably missed the bootloader
2022-12-23 19:30:46 <Saklad5> …Ah
2022-12-23 19:31:05 <Saklad5> Thanks, that's very helpful
2022-12-23 19:31:05 <minimal> Saklad5: again that makes sense - if you have configured the bootloader to use UUID (rather than PARTUUID) then how do you expect it to boot of a unpartitioned drive?
2022-12-23 19:31:39 <Saklad5> Could you elaborate on the distinction?
2022-12-23 19:31:43 <minimal> Saklad5: doh! forget that, was thinking other way round
2022-12-23 19:31:49 <Saklad5> Yeah, I was wondering
2022-12-23 19:32:09 <minimal> Saklad5: what does "blkid" show for both partitioned and unpartitioned drives?
2022-12-23 19:32:10 <Saklad5> I saw references to PARTUID, assumed that meant "Partition UUID", and made sure it WASN'T like that
2022-12-23 19:32:56 <Saklad5> minimal: One sec, let me a copy of the unpartitioned drive again
2022-12-23 19:33:00 <minimal> Saklad5: that would be output from the full version of blkid, not Busybox's blkid
2022-12-23 19:33:09 <Saklad5> let me attach*
2022-12-23 19:33:10 <minimal> so "apk add blkid"
2022-12-23 19:33:27 <Saklad5> I love a lot of things about Busybox, but good god is it poorly documented
2022-12-23 19:34:07 <minimal> Busybox's blkid is quite limited
2022-12-23 19:34:13 <Saklad5> Understandable
2022-12-23 19:34:15 <ikke> most bb tools are :P
2022-12-23 19:34:25 <Saklad5> They're insanely good at what they do
2022-12-23 19:34:42 <Saklad5> Which is basically "implement POSIX as efficiently as possible"
2022-12-23 19:34:53 <ikke> But I'd start checking the disk you are trying to boot has a bootloader
2022-12-23 19:35:08 <Saklad5> I'm quite confident you're right about this
2022-12-23 19:35:09 <Guest28> Ermine: Seems like it's going forward a bit: https://paste.debian.net/plain/1265084/ - But it can't find `pw-cat` do you have any other suggestions on how I can test audio? ffmpeg maybe/?
2022-12-23 19:35:10 <minimal> Saklad5: that's debatable in some cases lol
2022-12-23 19:35:55 <Ermine> Guest28: pw-cat is in pipewire-tools package.
2022-12-23 19:37:01 <Guest28> Yay! We have audio!
2022-12-23 19:37:20 <minimal> Saklad5: looking at some docs for Linode Backup: "Add disks will be restored as an ext4 file system and their UUIDs will be *different* than the original disks."
2022-12-23 19:37:24 <Ermine> Yay!
2022-12-23 19:37:25 <minimal> s/Add/All
2022-12-23 19:37:25 <alpine-meetbot> minimal meant to say: Saklad5: looking at some docs for Linode Backup: "All disks will be restored as an ext4 file system and their UUIDs will be *different* than the original disks."
2022-12-23 19:37:33 <jh81> hmz
2022-12-23 19:37:44 <Guest28> Ermine: Thanks a lot :)
2022-12-23 19:37:45 <ikke> minimal: right, they seem to do an fs level backup
2022-12-23 19:37:48 <minimal> so you'll have problems booting off a restored disk using UUID
2022-12-23 19:37:49 <ikke> restoring takes quite a long time
2022-12-23 19:38:13 <Saklad5> I was going to review that more carefully before finishing, but thanks for checking
2022-12-23 19:38:14 <Guest28> Ermine: Should I update the wiki page to reflect the dbus stems or do you want to do it?
2022-12-23 19:38:19 <Guest28> *steps
2022-12-23 19:38:32 <Guest28> (and the missing packages etc.)
2022-12-23 19:39:02 <Saklad5> Their instructions for making a custom installation compatible with backups do explicitly say to use device paths, so I'm not surprised by that
2022-12-23 19:39:13 <Saklad5> But I don't think that's the issue I'm specifically trying to solve right now
2022-12-23 19:39:35 <Saklad5> minimal: In the interest of not spamming the channel, could you say which part of "blkid"'s output you want? Or I could contact you on Matrix or something
2022-12-23 19:39:45 <minimal> Saklad5: yeah I wondered about using Labels rather than UUIDs but I suspect their restore also might not setup Labels as before also
2022-12-23 19:39:47 <ikke> Saklad5: you can use a paste service
2022-12-23 19:39:54 <ikke> like https://tpaste.us
2022-12-23 19:40:02 <Saklad5> I could, but you could also just say what you're looking for
2022-12-23 19:40:14 <minimal> Saklad5: the output of the command ;-)
2022-12-23 19:40:49 <Saklad5> 🙄
2022-12-23 19:40:52 <Saklad5> One sec
2022-12-23 19:41:45 <minimal> to have some "hard facts" to look at rather than a theoretical discussion
2022-12-23 19:42:17 <Ermine> Guest28: you can if you want, but I need to figure out the whole XDG_RUNTIME_DIR thing. For now you need to create this and set your XDG_RUNTIME_DIR on every boot.
2022-12-23 19:43:26 <Guest28> Ermine: Does that dir need to be created at every boot? Because I just did `export XDG_RUNTIME_DIR=/run/user/$UID` in my .bashrc
2022-12-23 19:43:44 <Ermine> you need to mkdir -p it
2022-12-23 19:44:02 <ikke> Guest28: yes
2022-12-23 19:44:07 <ikke> /run is tmpfs
2022-12-23 19:44:48 <Saklad5> minimal: https://saklad5.com/temporary/blkid_output
2022-12-23 19:45:05 <Guest28> Ah, I see. Hmm. That's a minor inconvenience at this point for me though.
2022-12-23 19:45:50 <Saklad5> minimal: /dev/sdd is the partitioned drive that that command is running from
2022-12-23 19:46:13 <ikke> so there are duplicate uuids
2022-12-23 19:46:22 <Saklad5> Yes, that's intended right now
2022-12-23 19:46:36 <Saklad5> Or at least expected
2022-12-23 19:46:55 <Saklad5> But when I try to boot without the partitioned drive, it can't use the unpartitioned ones
2022-12-23 19:47:03 <Saklad5> Probably because the bootloader isn't there
2022-12-23 19:47:11 <Saklad5> Is the bootloader outside a partition?
2022-12-23 19:47:15 <ikke> yes
2022-12-23 19:47:19 <Saklad5> Huh
2022-12-23 19:48:03 <Saklad5> Should I give whatever Linode's default bootloader is a shot?
2022-12-23 19:48:30 <minimal> Saklad5: for BIOS booting the MBR is in the 1st disk sector
2022-12-23 19:48:38 <ikke> ^
2022-12-23 19:48:52 <Saklad5> You don't mean actual BIOS do you?
2022-12-23 19:49:14 <minimal> Saklad5: I seem to remember Linode machines booting in a "strange" fashion, do they chainload a bootloader or something?
2022-12-23 19:49:18 <ikke> I don't think linode is using uefi, but not certain
2022-12-23 19:49:51 <Saklad5> Anyway, I currently have the Linode's kernel option set to "Direct Disk"
2022-12-23 19:50:03 <Saklad5> So that's definitely the issue if it's outside a partition
2022-12-23 19:50:11 <Saklad5> The question then is how best to change that
2022-12-23 19:50:12 <grawity> afaik "Direct Disk" *is* actual SeaBIOS looking for MBR bootcode
2022-12-23 19:50:19 <minimal> Saklad5: for booting x86/x86_64 machines its either via BIOS or UEFI and many cloud providers still only support BIOS booting
2022-12-23 19:51:24 <Saklad5> Huh
2022-12-23 19:52:14 <ikke> Saklad5: https://www.linode.com/docs/guides/managing-the-kernel-on-a-linode/
2022-12-23 19:52:53 <ikke> I think for alpine linodes they are already using grub by default
2022-12-23 19:52:55 <Saklad5> Looks like changing the configuration to use a Linode kernel was sufficient to boot sucessfully
2022-12-23 19:53:11 <ikke> yes, because it then does not need a bootloader
2022-12-23 19:53:14 <Saklad5> What are the ramifications of using that instead?
2022-12-23 19:53:29 <ikke> that you get a different kernel from what we offer
2022-12-23 19:53:53 <Saklad5> Let me check how it's configured, one sec
2022-12-23 19:58:26 <minimal> Saklad5: if you look at the link ikke posted it talks about the different Kernel settings
2022-12-23 19:59:39 <Saklad5> I was just checking for specific features. Seems to have what I need
2022-12-23 19:59:43 <minimal> "Direct Disk: Instead of a Linux Kernel, this uses the MBE[sic] Master Boot Record) of the primary disk". Obviously on the unpartitioned sdd drive you don't have a MBR and so that option won't work
2022-12-23 20:00:15 <Saklad5> It worked on the one I set up with an installer, which presumably installed it
2022-12-23 20:00:27 <Saklad5> extlinux, if I understand the documentation correctly
2022-12-23 20:00:50 <minimal> I assume with the "GRUB 2" option then Linode's Grub (NOT any grub on sdd) is loading the kernel/initramfs from sdd
2022-12-23 20:01:27 <Saklad5> So, is /boot just for packages then?
2022-12-23 20:03:03 <minimal> for kernel, initramfs, microcode, grub config
2022-12-23 20:03:07 <Saklad5> But not bootloader
2022-12-23 20:03:52 <Saklad5> I have "linux-firmware-none" and "linux-virt" installed. Are either of those redundant if an externally-supplied kernel is being used?
2022-12-23 20:03:58 <minimal> perhaps ikke can correct me, but my impression of Linode is that it runs its own bootloader rather than (directly) any bootloader on the disk image
2022-12-23 20:04:18 <minimal> so it chainloads
2022-12-23 20:04:46 <ikke> It may be chainloading
2022-12-23 20:05:05 <Saklad5> I know those two packages are a defining aspect of the virtual image of Alpine Linux, but I don't really know what they are
2022-12-23 20:05:45 <minimal> Saklad5: linux-virt is a kernel tailored for VMs (so doesn't have a lot of physical device drivers)
2022-12-23 20:05:50 <ikke> linux-firmware-none just prevents all default firmware packages to be installed
2022-12-23 20:06:13 <Saklad5> @ikke: I figured that was some form of placeholder. So if I remove that, something will install all the default ones?
2022-12-23 20:06:20 <ikke> apk will, yes
2022-12-23 20:06:24 <minimal> as ikke said, for a VM you don't need/want firmware for lots of physical devices installed
2022-12-23 20:06:34 <Saklad5> That's why I used it: this is a VM
2022-12-23 20:06:56 <Saklad5> If I'm using Linode's kernel, can I remove linux-virt entirely?
2022-12-23 20:07:03 <minimal> linux-firmware-none is NEEDED to satisfy a dependancy, remove it and get all the firmware packages instead ;-)
2022-12-23 20:07:12 <ikke> Saklad5: yes, should be possible
2022-12-23 20:07:19 <Saklad5> Well, might as well try it now
2022-12-23 20:07:33 <minimal> but then what happens about initramfs? it is built using kernel modules etc...
2022-12-23 20:08:06 <Saklad5> But is the kernel even getting used?
2022-12-23 20:08:15 <Saklad5> If I'm using Linode's?
2022-12-23 20:08:18 <minimal> Saklad5: is the initramfs getting used?
2022-12-23 20:09:02 <Saklad5> Don't really know how to tell
2022-12-23 20:09:50 <minimal> watch the boot sequence? do a "dmesg" after boot to see which initramfs used?
2022-12-23 20:12:36 <Saklad5> Looks like theirs is being used
2022-12-23 20:12:48 <Saklad5> If I'm wrong, I still have the cloned volumes anyway
2022-12-23 20:14:35 <minimal> I've always thought Linode was weird......so you're not using Alpine's bootloader and not using Alpine's kernel/initramfs, so then it's a hybrid/frankenstein system ;-)
2022-12-23 20:15:32 <ikke> This is from a linode that we use:
2022-12-23 20:15:34 <ikke> Linux version 5.15.85-0-virt (buildozer@build-3-17-x86_64) (gcc (Alpine 12.2.1_git20220924-r4) 12.2.1 20220924, GNU ld (GNU Binutils) 2.39) #1-Alpine SMP Thu, 22 Dec 2022 08:57:12 +0000
2022-12-23 20:16:11 <minimal> ikke: that's using "Direct Disk" then as boot option I assume
2022-12-23 20:16:31 <grawity> linode has 3 options -- a) linode kernel; b) linode grub2 reading your grub.cfg and booting your kernel; c) "direct disk" seabios booting your disk as normal
2022-12-23 20:17:24 <minimal> grawity: yeah, option 2 "Grub 2" obviously requires you to use Grub as the bootloader
2022-12-23 20:17:39 <ikke> minimal: grub 2
2022-12-23 20:19:03 <minimal> As I've never set up an unpartitioned *boot* device I'm just wondering how that works for BIOS booting. When mkfs.ext4 is run does it somehow know to skip the 1st sector so that the MBR can be placed there? Or does it write to the whole device, so not making MBR booting from that device possible?
2022-12-23 20:22:41 <Saklad5> It looks like Linode's snapshots fail if you have *any* unsupported volumes, regardless of the current configuration. So I guess I have to get rid of my crude backups and hope it works
2022-12-23 20:23:24 <minimal> Saklad5: why not use "Grub 2" boot option instead?
2022-12-23 20:23:32 <Saklad5> Because I don't have a reason to do so?
2022-12-23 20:23:39 <Saklad5> Is there a reason to do so?
2022-12-23 20:23:52 <minimal> it would be better using an Alpine-created kernel with Alpine
2022-12-23 20:24:12 <Saklad5> I thought it didn't use Grub 2 normally
2022-12-23 20:24:44 <minimal> according to that Linode page "Grub 2" is the default for most new ~Lindows created after August 2018
2022-12-23 20:24:46 <Saklad5> Also, their kernel might be more tuned to their hardware, judging by the configuration
2022-12-23 20:25:05 <grawity> I think the Linode kernels are mostly a thing from their Xen era
2022-12-23 20:25:26 <grawity> but someone said the recommendation with KVM is to use "direct disk" now
2022-12-23 20:26:21 <Saklad5> It does look like they recommend using an "upstream" kernel
2022-12-23 20:26:38 <Saklad5> In that case… is there a tool for installing a bootloader onto a disk without wiping it?
2022-12-23 20:27:14 <minimal> that's the bit I'm not sure about - wondering what grub-install does when run on an unpartitioned drive
2022-12-23 20:27:15 <ikke> You need to have space for the bootloader, like minimal alluded to
2022-12-23 20:27:26 <Saklad5> How would I do that?
2022-12-23 20:27:41 <Saklad5> I'm fine with making a new disk, adding the bootloader, then copying a partition to it
2022-12-23 20:27:52 <minimal> ikke: but space for the bootloader is generally take care of when you create partitions as the 1st partition starts at 1Mb
2022-12-23 20:28:08 <minimal> if its an unpartitioned drive then I don't know how you "reserve" the start of the disk
2022-12-23 20:28:21 <ikke> I think it's the MBR that takes care of that if I recall correctly
2022-12-23 20:29:16 <minimal> ikke: how? if you're installing Alpine it'll do a "mkfs.ext4 /dev/sda" (for example), rather than sda1, and so mkfs.ext4 has access to the whole disk
2022-12-23 20:30:03 <ikke> /dev/sda1     2048   1050623   1048576   512M EFI System
2022-12-23 20:30:05 <ikke> start is 2048
2022-12-23 20:30:36 <minimal> ikke: that's partitioned though
2022-12-23 20:30:36 <Saklad5> I moved the partition with "dd", it probably starts at 0 right now
2022-12-23 20:30:59 <minimal> the question is regarding an unpartitioned drive
2022-12-23 20:30:59 <ikke> minimal: yes, so that does not work (from what I can tell) when you do not have an MBR
2022-12-23 20:31:26 <Saklad5> I think I can use the "seek" operand to make it start later...
2022-12-23 20:31:40 <Saklad5> OK, I have a stupid plan
2022-12-23 20:31:52 <minimal> right, so the situation seems to be that Linode to do recommend/support partitioned drives, they however do support MBR booting (which would require a partitioned drive) as one of their options
2022-12-23 20:31:58 <Saklad5> How do I install the bootloader at all?
2022-12-23 20:32:04 <minimal> s/to do/do not/
2022-12-23 20:32:04 <alpine-meetbot> minimal meant to say: right, so the situation seems to be that Linode do not recommend/support partitioned drives, they however do support MBR booting (which would require a partitioned drive) as one of their options
2022-12-23 20:32:14 <Saklad5> Linode doesn't fully support partitioned drives: their backup service is incompatible with them
2022-12-23 20:32:15 <ikke> Saklad5: grub-install with the right argument
2022-12-23 20:32:35 <Guest28> Ermine: Would it also work if I set $XDG_RUNTIME_DIR to a location in my homedir?
2022-12-23 20:32:38 <minimal> so it seems like "Grub 2" based booting would be the way to go
2022-12-23 20:32:42 <Saklad5> What about Syslinux?
2022-12-23 20:33:18 <minimal> Saklad5: you'd have to use "Direct Booting" for Syslinux......which requires a MBR which requires a partitioned drive ;-)
2022-12-23 20:33:45 <Saklad5> MBRs require partitioned drives?
2022-12-23 20:33:55 <ikke> They define the partitions
2022-12-23 20:34:02 <minimal> that's what I've been saying yes
2022-12-23 20:34:04 <Saklad5> Right
2022-12-23 20:34:06 <Ermine> Guest28: yeah, but you may want to have it in a tmpfs so leftovers like lock files and socket files are removed on reboot/shutdown
2022-12-23 20:34:16 <Saklad5> It's so weird that Linode won't back partitions up
2022-12-23 20:34:28 <ikke> They have a very crude backup sollution
2022-12-23 20:34:31 <minimal> as otherwise there is no way to separate off/protect the MBR from filesystems
2022-12-23 20:35:02 <Guest28> Ermine: What about just wiping the dir before reboot? This way it doesn't require elevated privs.
2022-12-23 20:35:03 <ikke> I think they create a new linode with the same disk setup, and then rsync (or whatever) all files to the partitions
2022-12-23 20:35:10 <Saklad5> I can't really be bothered to set up database replication right now, so I at least want Linode's backups as a stopgap
2022-12-23 20:35:29 <Ermine> Guest28: this way will probably work
2022-12-23 20:35:41 <Saklad5> @ikke: Wait, seriously?
2022-12-23 20:35:45 <Guest28> Oh cool, great to hear!
2022-12-23 20:35:52 <ikke> Saklad5: I don't know the exact setup they do
2022-12-23 20:35:55 <Saklad5> I know their documentation says they do it at "file-level", but there's got to be a better option
2022-12-23 20:35:56 <ikke> but something along the line
2022-12-23 20:36:04 <minimal> ikke: I'm guessing setup-disk can't handle an unpartitioned disk (even with appropriate options/envvars) as whatever way it calls grub-install would expect to write an MBR
2022-12-23 20:36:26 <Saklad5> I guess I'm using GRUB, then
2022-12-23 20:36:29 <ikke> minimal: setup-disk uses sfdisk to partition it
2022-12-23 20:36:55 <minimal> ikke: I meant in the scenario when you normally manually partition and then run setup-disk
2022-12-23 20:37:06 <ikke> Then yes
2022-12-23 20:37:06 <minimal> i.e. with filesystems mounted already
2022-12-23 20:37:31 <minimal> just looking at grub-install manpage to see if there's any option to stop it trying to write MBR
2022-12-23 20:37:47 <minimal> ah, "--no-bootsector"
2022-12-23 20:38:06 <grawity> hmm do you need to "install" it at all? iirc linode uses its own grub modules as well
2022-12-23 20:38:25 <Saklad5> grawity: My thinking as well
2022-12-23 20:38:27 <minimal> so then "grub-install --no-bootsector /dev/sda" I guess
2022-12-23 20:38:42 <grawity> in which case all you should need is grub-mkconfig
2022-12-23 20:38:49 <Saklad5> I was just about to ask, thanks
2022-12-23 20:39:03 <minimal> grawity: the Linode "Grub 2" bot option reads the OS' grub.cfg so I'd assume you'd need to run grub-install in order for grub-mkconfig to generate a valid grub.cfg file
2022-12-23 20:39:11 <grawity> hmm
2022-12-23 20:39:26 <grawity> iirc no, linode's own docs don't mention that
2022-12-23 20:39:29 <minimal> I don't know if running "grub-mkconfig" directly would work
2022-12-23 20:39:36 <grawity> ...I write my own grub.cfg though
2022-12-23 20:39:47 <minimal> yeah I'm not impressed with their docs
2022-12-23 20:39:51 <Saklad5> I'm planning to at least make sure I understand what the config says
2022-12-23 20:40:34 <minimal> grawity: and when something like a kernel package update triggers grub-mkconfig to be run again?
2022-12-23 20:40:43 <Saklad5> Where is the config supposed to be located?
2022-12-23 20:40:51 <Saklad5> "/etc/default/grub"?
2022-12-23 20:41:03 <minimal> in /boot/grub/grub.cfg
2022-12-23 20:41:19 <ikke> /etc/default/grub is where you configure grub
2022-12-23 20:41:24 <minimal>  /etc/default/grub is used by grub-mkconfig together with template files in /etc/grub.d/ to create grub.cfg
2022-12-23 20:41:31 <ikke> yeah, what minimal said
2022-12-23 20:42:50 <minimal> grawity: the grub package has a trigger on /boot to run grub-mkconfig
2022-12-23 20:43:52 <Saklad5> So what should I do? Just run grub-mkconfig without arguments?
2022-12-23 20:44:05 <minimal> I might experiment in creating my own Alpine disk image for Linode. i'd avoided them so far as they don't support cloud-init plus their weird way of booting things
2022-12-23 20:44:18 <Saklad5> I still can't get reboots to actually happen
2022-12-23 20:44:19 <Saklad5> It's weird
2022-12-23 20:44:37 <ikke> Saklad5: same for us, it just powers down untill lassie boots it again
2022-12-23 20:44:52 <Saklad5> Is that true for the prebuilt Alpine images they offer too?
2022-12-23 20:45:04 <ikke> yes, we only use those
2022-12-23 20:45:19 <minimal> Saklad5: as I'd said earlier I believe you'd need to run "grub-install --no-bootsector /dev/sda"
2022-12-23 20:45:28 <Saklad5> But I'm not installing Grub
2022-12-23 20:45:32 <minimal> and, from memory, grub-install will also run grub-mkconfig
2022-12-23 20:45:55 <minimal> Saklad5: I'm assuming you're going to use the "Grub 2" boot option
2022-12-23 20:46:20 <Saklad5> Doesn't that mean Grub is run from outside my VM?
2022-12-23 20:46:36 <minimal> but it reads the *distros'* grub.cfg file
2022-12-23 20:46:51 <minimal> so you need to have a /boot/grub/grub.cfg file
2022-12-23 20:46:56 <Saklad5> And nothing else
2022-12-23 20:47:03 <ikke> You need a kernel in that case
2022-12-23 20:47:06 <minimal> and I'm not sure if grub-mkconfig will creste that if grub-install has not already been run
2022-12-23 20:47:12 <Saklad5> I had one, and I can just add it back
2022-12-23 20:47:18 <grawity> minimal: could you make a pseudo package that has the same hook but only runs grub-mkconfig?
2022-12-23 20:47:33 <Saklad5> Can I not just run grub-mkconfig now?
2022-12-23 20:47:37 <Saklad5> Then get rid of Grub?
2022-12-23 20:47:49 <ikke> if you install a kernel, it should trigger it
2022-12-23 20:48:37 <Saklad5> So I don't need to do anything beyond adding linux-virt back?
2022-12-23 20:48:40 <minimal> Saklad5: if you get rid of grub then every time the Alpine kernel package is updated it will try and fail to run grub-mkconfig
2022-12-23 20:49:03 <Saklad5> It isn't in the original image
2022-12-23 20:49:09 <Saklad5> Is that because it uses Syslinux?
2022-12-23 20:49:23 <minimal> typically Syslinux  is used for BIOS booting
2022-12-23 20:49:39 <Saklad5> Is it possible to remove Syslinux, if I'm installing Grub?
2022-12-23 20:49:44 <minimal> yes
2022-12-23 20:49:49 <Saklad5> Which package?
2022-12-23 20:49:56 <Saklad5> Wait, stupid question
2022-12-23 20:49:58 <Saklad5> I see it
2022-12-23 21:00:46 <jh81> anyone ever succeeded in getting oracle'
2022-12-23 21:00:53 <jh81> s virtualbox to work on alpine?
2022-12-23 21:02:16 <minimal> jh81: you mean the packages Oracle supply? They'd be compiled for glibc rather than musl
2022-12-23 21:03:13 <Ermine> !17688
2022-12-23 21:03:42 <ikke> The ancient MR
2022-12-23 21:05:13 <jh81> well, that seems complicated
2022-12-23 21:05:21 <Ermine> ptrcnull tried her best
2022-12-23 21:05:38 <ptrc> that was mostly some other person
2022-12-23 21:05:43 <ptrc> i just complained about stuff
2022-12-23 21:05:50 <Ermine> ah
2022-12-23 21:06:55 <jh81> :D
2022-12-23 21:07:03 <jh81> can be a very valuable task too
2022-12-23 21:07:46 <jh81> oh well, i have a proxmox server :) i guess i'll move the virtualbox vm there
2022-12-23 21:07:47 <minimal> jh81: ah, didn't know there was an aport attempt. VB is now at 7.x whereas that was for 6.1.32 - not sure if it might be easier or harder now to try to package 7.x
2022-12-23 21:08:10 <jh81> i'm only just beginning to grasp aport :)
2022-12-23 21:11:27 <ikke> Certainly not the easiest package to start with
2022-12-23 21:11:41 <ikke> Though, maybe void has a musl build for it
2022-12-23 21:12:12 <jh81> https://voidlinux.org/packages/?arch=x86_64&q=virtualbox they do
2022-12-23 21:12:45 <jh81> does that make it easier?
2022-12-23 21:13:13 <ikke> Not sure if that's for musl? void targets both glibc and musl
2022-12-23 21:13:34 <jh81> oh right
2022-12-23 21:13:36 <Saklad5> Grub seems to complain about missing fonts. Would grub-mkfont help with that?
2022-12-23 21:13:40 <jh81> there;s an option to select musl
2022-12-23 21:13:43 <jh81> then there's none
2022-12-23 21:15:30 <minimal> jh81: from looking at the number of musl related patches in the old MR I suspect it will be problematic getting VB working and keeping it working on musl
2022-12-23 21:15:48 <jh81> thought so :)
2022-12-23 21:16:06 <minimal> I don't see many musl related tickets on the VB website
2022-12-23 21:17:37 <minimal> jh81: I run Alpine on VB all the time (Debian host) but on Alpine hosts I use QEMU
2022-12-23 21:17:56 <Ermine> Something tells me that upstream wouldn't accept these patches
2022-12-23 21:18:17 <jh81> but i was hoping i could run VB on Alpine ;-)
2022-12-23 21:18:17 <jh81> yes... i dont have a problem running alpine on VB either :)
2022-12-23 21:20:07 <Ermine> Also there's an extension pack which is non-free, so it isn't going to be available on Alpine
2022-12-23 21:20:11 <jh81> does QEMU have a gui?
2022-12-23 21:20:14 <minimal> whilst its source is available its development is basically only really done by Oracle themselves
2022-12-23 21:20:21 <ikke> jh81: something like virtmanager
2022-12-23 21:20:33 <jh81> hmz, i guess i need to look into that then :)
2022-12-23 21:20:39 <minimal> jh81: there are several, i.e. virt-manager (via libvirtd)
2022-12-23 21:20:44 <ikke> since 7.0, you don't need the extension anymore for extended usb support
2022-12-23 21:21:17 <minimal> ikke: yeah plus it added secure boot support and TPM - haven't had a chance to try those out yet
2022-12-23 21:21:21 <jh81> i always dream of impossible things :|
2022-12-23 21:52:33 <jh81> Well, next project will be Kodi on Alpine :) 
2022-12-23 21:52:44 <jh81> that seems more dowble then
2022-12-23 21:52:52 <jh81> doable the. VB
2022-12-23 21:55:20 <minimal> jh81: isn't libreelec lightweight enough?
2022-12-23 21:55:58 <jh81> Thats what im using now, not sure how it will do
2022-12-23 21:56:22 <jh81> but it still isnt based on Alpine :-)
2022-12-23 22:44:27 <nacelle> so it appears this image: alpine-rpi-3.17.0-armhf.tar.gz is really arm7, where it claims its arm7vl on boot on my pi3b+ and -wont- boot on my pi1b+
2022-12-23 22:44:44 <nacelle> is this a known issue?
2022-12-23 22:45:13 <ikke> haven't heard it
2022-12-23 22:45:34 <nacelle> I'm guessing that theres not a lot of pi1 people considering their age and availability
2022-12-23 22:45:53 <ikke> armhf is armv6 btw, not armv7
2022-12-23 22:46:09 <ikke> arm7vl is aarch64 in 32-bits mode
2022-12-23 22:46:24 <ikke> sorry, that was armv8;
2022-12-23 22:46:29 <ikke> armv8l
2022-12-23 22:46:36 <nacelle> right
2022-12-23 22:46:51 <nacelle> it should be booting up as armv6 or some such I'd imagine
2022-12-23 22:46:56 <ikke> yes
2022-12-23 22:49:32 <ikke> where / how did you notice it's armv7?
2022-12-23 22:51:43 <nacelle> i booted it on a pi3b+
2022-12-23 22:51:51 <nacelle> it wont boot on a pi1b+
2022-12-23 22:51:56 <nacelle> same image
2022-12-23 22:53:15 <ikke> The packages in that image are for armhf 
2022-12-23 22:53:53 <nacelle> yes, they are - which is right for the pi1, according to this: https://wiki.alpinelinux.org/wiki/Raspberry_Pi
2022-12-23 22:54:31 <ikke> So when you try to boot it on a pi1b+, what happens?
2022-12-23 22:59:04 <ikke> The files in the image suggest that it should support the rp1 and the rpi2
2022-12-23 22:59:06 <nacelle> black screen, no video
2022-12-23 22:59:09 <nacelle> right
2022-12-23 23:00:15 <nacelle> basically, this seems like a bug in the built image
2022-12-23 23:00:25 <nacelle> it was built wrong or copied over wrong or something
2022-12-23 23:00:48 <ikke> they are not manually copied over
2022-12-23 23:00:57 <nacelle> right, like a script has a bork in it
2022-12-23 23:01:07 <nacelle> bugs happen
2022-12-23 23:01:21 <nacelle> thats my theory anyways, given what I know
2022-12-23 23:01:33 <ikke> Hard to verify just with the files
2022-12-23 23:01:46 <nacelle> yeah, i'm guessing i might be the first to report this
2022-12-23 23:01:53 <nacelle> i was hoping not
2022-12-23 23:02:01 <nacelle> i have  searched on the bug tracker yet
2022-12-23 23:02:05 <ikke> nacelle: in any case, would you mind opening an issue on gitlab.alpinelinux.org?
2022-12-23 23:02:14 <nacelle> i'll try
2022-12-23 23:02:45 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/issues
2022-12-23 23:03:53 <ikke> I do still have an rpi1 lying around, so maybe I can test it with that, but it would be later
2022-12-23 23:21:35 <nacelle> https://gitlab.alpinelinux.org/alpine/aports/-/issues/14493
2022-12-23 23:52:20 <hiohoihoi> hello
2022-12-23 23:52:56 <hiohoihoi> any way to install virtualbox in alpine linux?
2022-12-23 23:53:50 <minimal> hiohoihoi: this was discussed here about 1 or 2 hours ago - virtualbox is not currently packaged for Alpine due to problems building it
2022-12-23 23:54:07 <hiohoihoi> ok ty
2022-12-23 23:54:11 <minimal> it appears it does not work with the musl C library that Alpine uses
2022-12-23 23:54:27 <hiohoihoi> qemu works good 
2022-12-23 23:55:37 <minimal> you were asking about VirtualBox though
2022-12-23 23:55:50 <hiohoihoi> yes
2022-12-24 00:05:52 <minimal> nacelle: the linux-rpi APKBUILD file uses _carch="arm" for compiling both armhf and armv7 kernels, I think that might be the issue
2022-12-24 00:19:22 <minimal> nacelle: actually it's confusing as the bcmrpi_config file has CONFIG_ARCH_MULTI_V6=y so does that mean the kernel for both linux-rpi armhf and armv7 is compiled as v6 in both cases?
2022-12-24 05:19:32 <psykose> if you were building it yourself gcc would fail to make an armv7 image with the armv6 toolchain
2022-12-24 05:20:13 <psykose> kernel*
2022-12-24 05:20:56 <psykose> probably more the script fetches the wrong thing when you specify that
2022-12-24 07:52:30 <ncopa> re virtualbox: https://gitlab.alpinelinux.org/alpine/aports/-/issues/3937
2022-12-24 08:09:21 <Guest28> Hi, I'm trying to follow this guide: https://wiki.alpinelinux.org/wiki/Flatpak - Installed flatpak and booted, but then: https://paste.debian.net/plain/1265138/
2022-12-24 08:14:23 <psykose> probably need to flatpak repair --system first, i forget
2022-12-24 08:14:32 <psykose> that folder has to present even for user
2022-12-24 08:25:43 <Guest28> flatpak repair --system results in error: Unable to connect to system bus (and I've been advise not to run dbus as root), while as user it succeeds, but still getting those errors: https://paste.debian.net/1265141/
2022-12-24 08:27:12 <psykose> did you forget to run the system dbus service
2022-12-24 08:35:38 <Guest28> Before doing startx I did export $(dbus-launch) ... but I did so as a user.
2022-12-24 08:36:04 <Guest28> Again, I was advised against running dbus as root, is there a way to run a system dbus without root?
2022-12-24 08:36:23 <grawity> you're confusing the per-user bus and the system-wide bus
2022-12-24 08:36:52 <grawity> dbus-launch is for starting a per-user bus (well, a per-session bus) and it has to run under your own account
2022-12-24 08:37:11 <grawity> the system bus is *not* started via dbus-launch, it's started as a system service via openrc
2022-12-24 08:37:37 <grawity> (it shouldn't run as root either, exactly, but the service will take care of dropping privileges there)
2022-12-24 08:41:53 <Guest28> Ah I see, wasn't aware of the privilege handling by the service, I'll try that, thanks!
2022-12-24 11:42:43 <hsv2> hey, if i want to set custom dns but use dhcp for address and routes, what should i config? is it in /etc/network/interfaces? btw where is the documentation of that file syntax?
2022-12-24 11:44:00 <ikke> https://github.com/ifupdown-ng/ifupdown-ng/blob/main/doc/interfaces.scd
2022-12-24 11:54:28 <hsv2> thx
2022-12-24 11:57:04 <hsv2> ok, looks like it's not there… i guess in udhcpd config then, gogo rtfm
2022-12-24 11:59:50 <ikke> RESOLV_CONF=no in /etc/udhcpc/udhcpc.conf
2022-12-24 11:59:57 <ikke> (does not exist by default)
2022-12-24 12:02:39 <hsv2> thx ikke
2022-12-24 12:28:02 <avuton> minimal: Thanks!
2022-12-24 15:28:44 <scorpion2185[m]1> I can't start tor-browser, is this because they compiled it wihtou musl or something compatibility?
2022-12-24 15:30:17 <donoban> if you downloaded binaries there are most sure built with glibc
2022-12-24 15:30:28 <donoban> s/there/they
2022-12-24 15:30:28 <alpine-meetbot> donoban meant to say: if you downloaded binaries they are most sure built with glibc
2022-12-24 15:31:15 <scorpion2185[m]1> yes I download linux64, AFAIK broiwser takes a LOT to build
2022-12-24 15:31:49 <psykose> tor browser doesn't work on musl
2022-12-24 15:31:52 <donoban> its not easy to build it for alpine/musl
2022-12-24 15:32:00 <scorpion2185[m]1> no error when I lauch it
2022-12-24 15:32:04 <psykose> only works in flatpak
2022-12-24 15:44:20 <scorpion2185[m]1> if a package has nothing that build do you use only the package function?
2022-12-24 15:44:30 <ikke> yes
2022-12-24 15:45:23 <ikke> just a heads-up: for aports we do not accept APKBUILDs that just package pre-built binaries
2022-12-24 15:45:29 <scorpion2185[m]1> not sure how is there some example ? https://github.com/spanezz/lightdm-autologin-greeter
2022-12-24 15:46:02 <scorpion2185[m]1> that package has pyhthon scripts so nothing that build
2022-12-24 15:46:12 <ikke> I see
2022-12-24 15:46:43 <psykose> sometimes i wish people didn't package every random github repo with 1 file into aports
2022-12-24 15:46:47 <psykose> but there's no darned policy against it
2022-12-24 15:47:02 <ikke> I can imagine the desire :)
2022-12-24 15:49:52 <scorpion2185[m]1> there are more than 1 file and I found out that when I already started and I'd like to learn to how package arbitrary files
2022-12-24 15:50:45 <scorpion2185[m]1> I find it very useful and so one can easiyl install it
2022-12-24 15:53:19 <psykose> oh yeah my mistake, it's a 20 line script, and a 3 line configuration file
2022-12-24 15:53:28 <psykose> and a desktop file even, guess that makes it three
2022-12-24 15:53:30 <psykose> i got completely owned
2022-12-24 15:53:49 <panekj> fatality
2022-12-24 15:59:01 <charlatan> so... what does everyone use for an "office suite" that supports musl?
2022-12-24 15:59:35 <panekj> libreoffice?
2022-12-24 16:00:00 <panekj> doubt that there is anything else provided
2022-12-24 16:00:10 <charlatan> Oh! Libre works? I really love Softmaker Office, but that might do.
2022-12-24 16:00:17 <panekj> or just use browser with google docs or office 365
2022-12-24 16:01:06 <charlatan> panekj: yea, I've tried that, but the web versions just aren't as powerful
2022-12-24 16:08:29 <scorpion2185[m]1> what's the equivalent for alpine lightdm gir1.2-lightdm-1 python-gobject?
2022-12-24 16:08:49 <scorpion2185[m]1> what's the name of lightdm gir1.2-lightdm-1 python-gobject in alpine? that is for ubuntu
2022-12-24 16:09:57 <psykose> lightdm py3-gobject3
2022-12-24 19:06:32 <scorpion2185[m]1> The author of that made nodm
2022-12-24 19:08:04 <scorpion2185[m]1> how can you reinstall a package?
2022-12-24 19:08:50 <zcrayfish> apk fix I think
2022-12-24 19:10:22 <scorpion2185[m]1> it the package is fine apk won't consider him
2022-12-24 19:11:19 <zcrayfish> s
2022-12-24 19:11:41 <ikke> --reinstall
2022-12-24 19:12:21 <zcrayfish> The documentation for apk fix (apk --help fix) says that --reinstall is the default option, and it that is the way it behaves on my box. YMMV.
2022-12-24 19:13:52 <ikke> Yes, but people want control
2022-12-24 19:14:17 <zcrayfish> In that case, `apk -i fix -r nano' ;) 
2022-12-24 19:14:25 <zcrayfish> (replace nano with whatever package)
2022-12-24 19:16:07 <scorpion2185[m]1> ah yes sorry I thought just apk fix
2022-12-24 19:16:22 <ikke> that's correct
2022-12-24 19:16:45 <scorpion2185[m]1> apk fix
2022-12-24 19:16:52 <zcrayfish> oh I see. :)
2022-12-24 19:16:56 <scorpion2185[m]1> I can't use buffyboard
2022-12-24 19:18:11 <scorpion2185[m]1> shoudl I report upstream ?
2022-12-24 19:18:41 <ikke> what is the issue?
2022-12-24 19:20:15 <scorpion2185[m]1> Could not update current terminal size: invalid argument
2022-12-24 19:20:53 <scorpion2185[m]1> Could not open /dev/uinput: no such file or directory   and it doesn't start
2022-12-24 19:24:59 <ikke> Not sure if it makes sense, but can you try: modprobe uinput
2022-12-24 19:25:18 <psykose> you need to create uinput and the permissions yourself
2022-12-24 19:25:20 <psykose> aside from that it works
2022-12-24 19:27:09 <scorpion2185[m]1> how do I create it? i don't recall the gitlab page mentioning that
2022-12-24 19:30:31 <psykose> modprobe uinput; chgrp input /dev/uinput; chmod 640 /dev/uinput
2022-12-24 19:30:34 <psykose> 660*
2022-12-24 19:30:44 <psykose> whether that lets any application keylogger your system i can't tell you
2022-12-24 19:30:49 <ikke> don't you need to create it first?
2022-12-24 19:30:54 <ikke> it seems to be a character device
2022-12-24 19:31:03 <psykose> gets made by modprobe for me
2022-12-24 19:31:05 <ikke> ok
2022-12-24 19:31:07 <psykose> probably a udev thing
2022-12-24 19:38:09 <scorpion2185[m]1> I opened a tty as root and just modprobing worked
2022-12-24 19:39:47 <scorpion2185[m]1> can I make it start when I open an empty tty?automatically
2022-12-24 19:41:38 <ikke> add a file to /etc/modules-load.d/
2022-12-24 19:42:01 <ikke> Then it gets modprobed at boot
2022-12-24 19:58:00 <scorpion2185[m]1> I edited /etc/elogind/logind.conf
2022-12-24 19:58:35 <scorpion2185[m]1> I added HandlePowerKey=ignore but it turns off when I press it
2022-12-24 20:00:41 <Hello71> did you actually install elogind
2022-12-24 20:01:21 <scorpion2185[m]1> yes and it's enabled
2022-12-24 20:01:41 <Hello71> did you disable acpid
2022-12-24 20:03:39 <scorpion2185[m]1> no, I stopped acpid and now it is fine
2022-12-24 20:05:44 <scorpion2185[m]1> but on wev XF86VolDown/Up don't have any input
2022-12-24 20:06:13 <scorpion2185[m]1> *ouput.  it is like they don't exist
2022-12-24 20:06:36 <scorpion2185[m]1> (wev wayland xev) is this a river bug maybe?
2022-12-24 20:23:55 <Hello71> try evtest
2022-12-24 21:21:02 <scorpion2185[m]1> I see Power Button only
2022-12-24 21:25:47 <scorpion2185[m]1> I tried INtel HID events and Intel HID 5 button array
2022-12-24 21:27:27 <scorpion2185[m]1> * Button only/quit
2022-12-24 21:30:15 <scorpion2185[m]1> they work on X , at least on xev
2022-12-24 21:32:00 <scorpion2185[m]1> * I see Power Button only
2022-12-25 06:39:31 <fz1> I have trouble with pipewire setup by following this guide
2022-12-25 06:39:34 <fz1> https://wiki.alpinelinux.org/wiki/PipeWire#XDG_RUNTIME_DIR
2022-12-25 06:39:55 <fz1> > ensure that your XDG_RUNTIME_DIR is set to a user-writable location
2022-12-25 06:40:14 <fz1> what is this mean? and how to set it to the "user-writable location" ?
2022-12-25 07:43:55 <uniqNuuid> fz1: a place your `id` can write
2022-12-25 07:45:05 <uniqNuuid> are you user 1000? run `id` to check the uid number
2022-12-25 07:51:21 <uniqNuuid> fz1: find / -writable 2>/dev/null 
2022-12-25 07:51:46 <uniqNuuid> this is all the dirs/files you can write. 
2022-12-25 10:36:32 <Ermine> fz1: You need to have a directory to which your user can create and work with files
2022-12-25 10:37:00 <Ermine> fz1: on many distros this is /run/user/$UID
2022-12-25 10:38:42 <Ermine> Merry Christmas everyone!
2022-12-25 10:44:33 <vkrishn> I converted one of 512mb ram, old mobile to SIP(server) phone using hostapd, kinda cool. Guess what's hotspot wifi name ;) - "Earl.Grey"
2022-12-25 10:45:55 <Ermine> vkrishn: I guess you like that kind of tea
2022-12-25 10:46:29 <vkrishn> well the next one would be called "Lemon.Sip"
2022-12-25 10:47:14 <vkrishn> Merry Christmas !
2022-12-25 11:15:25 <donoban> && Happy New Year :D
2022-12-25 11:54:41 <vkrishn> ok, here is final version of "Neighbor's Christmas sip comedy", on t(tea)paste.us, https://tpaste.us/a0pK 
2022-12-25 17:24:59 <bananicorn> On the topic of SIP, is anybody here successfully using Linphone on Alpine? :)
2022-12-25 20:35:04 <ikke> bananicorn: never tried
2022-12-25 20:35:23 <ikke> There is a linphone pacakge in testing
2022-12-25 22:21:42 <bananicorn> ikke: I gave the package a shot but had some trouble, thats why I asked
2022-12-26 00:44:26 <Saijin_Naib> Flatpak an option for you? 
2022-12-26 00:49:34 <aybbtu> I wrote here about btrfs compression issue what I've experienced with Alpine not so long ago. Back then I had a conversation with psykose who couldn't reproduce my issue though I could reproduce it with a clean install of musl Void I assumed there is something wrong with my setup. I didn't test it any further but now I installed stable Debian in VM with the exact same mount
2022-12-26 00:49:35 <aybbtu> options I used with Void and Alpine (compress-force:zstd) and forced compression there works just as intended (and just like before on a bare metal). For example all files installed by apt in Debian effectively being forcefully compressed so when I run 'btrfs fi defragment -r /' it affects nothing but in Alpine defragmentation actually enforcing compression and the values
2022-12-26 00:49:37 <aybbtu> reported by compsize utility is significantly lower after operation. I don't know what's going on here. When I first wrote about it psykose tested compression by copying files on the btrfs formatted flash drive and while it should behave just like copying files to btrfs system it didn't that's why I thought there is something wrong with setup. But now I'm completely lost. Can
2022-12-26 00:49:39 <aybbtu> anyone explain me whats going on and maybe test compression inside Alpine root fs?
2022-12-26 00:54:59 <minimal> aybbtu: so you saw the same behaviour with Alpine and musl Void but not with Debian? The obvious suggestion then is perhaps a musl-related difference in behaviour
2022-12-26 00:55:13 <minimal> aybbtu: did you try glibc Void for comparison?
2022-12-26 00:55:50 <aybbtu> minimal, not yet, just noticed the thing
2022-12-26 00:57:55 <aybbtu> Oh yes and the mount option is compress-force=zstd of course
2022-12-26 00:59:34 <aybbtu> After I've noticed there is something wrong back then I've stopped using compression level since its not a big deal because I'm using level 2 and the default is 3
2022-12-26 01:30:08 <aybbtu> Now testing glibc Void in VM by evaluating libreoffice+firefox+chromium installation size
2022-12-26 01:31:17 <aybbtu> dmesg reports that compression is enforced just like in Alpine and musl Void
2022-12-26 01:38:45 <minimal> aybbtu: but does it actually work as expected?
2022-12-26 01:39:41 <aybbtu> glibc Void is not affected
2022-12-26 01:39:55 <minimal> does that mean it works as expected?
2022-12-26 01:40:03 <aybbtu> just finished defragmentation no difference
2022-12-26 01:40:08 <aybbtu> as expected
2022-12-26 01:40:48 <aybbtu> defragmentation with force compression option of course
2022-12-26 01:40:51 <minimal> so you are seeing a difference in behaviour between musl Void and glibc Void? so then it seems it is somehow musl-related
2022-12-26 01:41:05 <minimal> and so it is not an Alpine issue specifically
2022-12-26 01:41:40 <aybbtu> yes it is and I'm glad I'm not insane
2022-12-26 01:42:45 <aybbtu> But maybe you can point me in the right direction I think it should be reported (?)
2022-12-26 01:43:58 <minimal> well if there is nothing different between musl Void and glibc Void other than the libc (i.e. assuming kernel version is same, btrfs tools are same, etc) then perhaps ask on the musl channel/list?
2022-12-26 01:47:10 <aybbtu> You think it is musl fault and not btrfs? I thought software linked against glibc is the problem and not the other way... Sorry I've just moved to Linux and quite overwhelmed by many things
2022-12-26 01:48:54 <minimal> aybtu: I though you said it worked as expected with glibc but not with musl - so then why would software linked against glibc be the problem?
2022-12-26 01:49:27 <alpine-meetbot> minimal meant to say: aybtu: I thought you said it worked as expected with glibc but not with musl - so then why would software linked against glibc be the problem?
2022-12-26 01:49:27 <minimal> s/though/thought/
2022-12-26 01:51:58 <aybbtu> minimal, because glibc... How should I put it? I think the right word are "not standard conformant"
2022-12-26 01:52:47 <minimal> aybbtu: are you trying to solve a problem or talk about standards compliance?
2022-12-26 01:55:43 <aybbtu> I want to know what to do in such a problem, because these stantards things are quite confusing. I thought if standards are musls desing goal they just turn down bug reports like this isn't it? 
2022-12-26 01:57:40 <minimal> aybbtu: who said standards compliance has anything to do with your problem?
2022-12-26 01:58:00 <aybbtu> My lack of knowledge I guess
2022-12-26 01:58:25 <q66_> nobody also said anything about reporting a problem with musl
2022-12-26 01:58:52 <minimal> I don't know what your problem is, however the one noticable variable between working as expected and not working as expected is musl vs glibc
2022-12-26 01:59:30 <minimal> there may be other variables as well - you should compare the musl Void and glibc Void package versions etc to see if there are other variables
2022-12-26 01:59:58 <q66_> if i had to guess it's btrfs-progs misbehaving on musl for whatever reason
2022-12-26 02:01:29 <minimal> I thought the logic was inside the kernel, with btrfs-progs only being used to set a flag, but I'm not that familiar with btrfs in general
2022-12-26 02:01:47 <q66_> but there is no way for libc to cause a diference in behavior of a kernel
2022-12-26 02:02:25 <q66_> iirc, the way btrfs compression works involves setting xattrs on files
2022-12-26 02:02:39 <q66_> specific ones that most tooling is not aware of
2022-12-26 02:03:01 <aybbtu> Can I somehow completely replace musl with glibc in Alpine system-wide to make a clean test?
2022-12-26 02:03:14 <minimal> q66_: aybbtu: no, Alpine is musl only
2022-12-26 02:05:40 <minimal> aybbtu: why not focus on Void? you have said you see different behaviour there between musl and glibc based installations
2022-12-26 02:12:44 <aybbtu> minimal: I'm not familiar with Void but that doesn't matter just like the versions. It is most certainly musl because there is also working Debian stable which is significantly older
2022-12-26 02:15:11 <vkrishn> theoritically, can a strong wifi(2.4Gz or 5Gz) compromise how eMMC works (specially write) ?
2022-12-26 02:15:46 <minimal> vkrishn: how is that Alpine relevant?
2022-12-26 02:16:17 <vkrishn> not linux related ?
2022-12-26 02:16:46 <minimal> vkrishn: this is an Alpine channel, not a general Linux channel or a hardware channel
2022-12-26 02:17:20 <mendelmaleh> hi, i installed some packages as dependencies for a virtual package (apk add -t), is there a way to export that virtual pkg / store it in my dotfiles? or should i make a real package at this point?
2022-12-26 02:17:55 <vkrishn> minimal: you are not irc AI ?
2022-12-26 02:18:27 <vkrishn> irc AI after me :-)
2022-12-26 02:19:34 <minimal> vkrishn: your question does not appear directly relevant/specific to Alpine
2022-12-26 02:20:36 <zcrayfish> strong omni wifi will not harm the Alpine partitions on your eMMC at all.
2022-12-26 02:21:57 <vkrishn> zcrayfish: thanks, will try to add keyword "alpine" in next queries
2022-12-26 02:22:43 <zcrayfish> #alpine-offtopic might be helpful as well ;)
2022-12-26 02:24:09 <vkrishn> ok, completely forgot about it
2022-12-26 02:30:55 <minimal> aybbtu: why are you reporting a musl bug on the musl channel? you are seeing a difference in behaviour on a musl system vs a glibc system, that however does not mean it is necessarily a musl bug
2022-12-26 02:31:33 <aybbtu> minimal: I reported it in btrfs channel also
2022-12-26 02:32:00 <vkrishn> just in gereral, if there are any AI(bots), put a prefix/suffix 'ai', and pls configure them to not undermine human questiones
2022-12-26 02:32:37 <minimal> aybbtu: ok, but in musl channel you reported it specifically as a musl bug, rather than a problem observed on a musl system
2022-12-26 02:32:50 <minimal> vkrishn: I am not a bot
2022-12-26 02:33:54 <vkrishn> ok, :), afaik remember ncopa generalize this channel linux'y not alpine,ee
2022-12-26 02:34:48 <minimal> vkrishn: the channel topic says "Alpine Linux, the secure lightweight distro"
2022-12-26 02:35:19 <vkrishn> may I should fish that sentence from irclogs
2022-12-26 02:35:25 <vkrishn> maybe^
2022-12-26 02:35:33 <minimal> as zcrayfish pointed out, at the very least it seems something for alpine-offtopic but you question appears hardware related
2022-12-26 02:36:11 <vkrishn> minimal:  yes that should have your first reply, thanks
2022-12-26 02:37:28 <minimal> vkrishn: it was, in the sense that stating "how is that Alpine relevant?" is another way of saying the same thing
2022-12-26 02:37:41 <vkrishn> and if you are channel moderator, do get a prefix-suffix like '@' or similar
2022-12-26 02:37:59 <vkrishn> and a real name (humanise)
2022-12-26 02:38:28 <vkrishn> human-eeis
2022-12-26 05:28:32 <fossy> hey all! i have a slightly strange setup i need some help with. i am PXE booting alpine linux using the netboot images, from iPXE (this is required for various other parts of the setup). i now want this on raspberry pi 4s. boot process is raspberry pi loader -> UEFI -> EFI iPXE -> alpine.
2022-12-26 05:29:38 <fossy> i'm having trouble figuring out what kernel/initramfs/modloop to load for the raspberry pi 4 from UEFI. in the netboot tar for aarch64, the generic -lts version appears to not have required drivers (no interface errors in initramfs) but the -rpi version causes the uefi to hang and appears to never boot the kernel.
2022-12-26 05:30:15 <fossy> i have a suspicion that the -rpi version does not work with efi. would appreciate pointers on the right track
2022-12-26 08:39:18 <uniqNuuid> fossy: https://0x0.st/o5UL.txt
2022-12-26 08:39:24 <uniqNuuid> change accordinglt 
2022-12-26 08:41:35 <uniqNuuid> fossy: reference : kernel ${base-url}/${dir}/vmlinuz-lts ${ipparam} alpine_repo=${repo-url} modules=loop,squashfs modloop=${base-url}/${dir}/modloop-lts quiet nomodeset {{ kernel_params }}
2022-12-26 08:41:48 <uniqNuuid> sorry : https://github.com/netbootxyz/netboot.xyz/blob/eb8d0c510bdffa0db05ca5ee4bf79c4bbd3ab261/roles/netbootxyz/templates/menu/alpinelinux.ipxe.j2#L30
2022-12-26 08:43:37 <uniqNuuid> do not use *-virt, it's for vm. 
2022-12-26 11:10:13 <AlpineNewb> Morning folks - I am trying to install alpine as a Desktop OS with full disk encryption. I saw the message in the LVM_on_Luks wiki page and just wanted to ask if it was still accurate and a concern:
2022-12-26 11:10:14 <AlpineNewb>  Note: The method is still experimental and you may lose your access to you OS at the next OS update
2022-12-26 11:10:44 <AlpineNewb> (This is in reference to using LUKS2... Althought the LUKS project seems to recommend it over LUKS1)
2022-12-26 11:11:08 <AlpineNewb> See: https://wiki.alpinelinux.org/wiki/LVM_on_LUKS#Luks2
2022-12-26 11:14:17 <lopid> still valid, afaik
2022-12-26 11:24:48 <fossy> thanks, uniqNuuid, i'll give it a go
2022-12-26 13:47:37 <AlpineNewb> Sorry - I got disconnected :( ... Is LUKS2 safe to use with Alpine Linux? Is it likely an OS update will nuke my install if I use this? (As per: https://wiki.alpinelinux.org/wiki/LVM_on_LUKS#Luks2)
2022-12-26 16:04:04 <lopid> it'll be safer to stick with luks1 for now
2022-12-26 16:31:08 <smallville7123> how do i get 32-bit clang
2022-12-26 16:32:12 <psykose_> to target 32-bit or
2022-12-26 16:33:11 <smallville7123> yes
2022-12-26 16:33:27 <psykose_> you just do clang --target= of the thing you're targetting
2022-12-26 16:33:29 <smallville7123> specifically so i can get the 32-bit headers
2022-12-26 16:34:10 <smallville7123> eg /usr/incluse/clang-eabi or something
2022-12-26 16:34:17 <psykose_> and then you need to make a sysroot for your target somehow and pass it with --sysroot
2022-12-26 16:34:23 <psykose_> what are you actually doing
2022-12-26 16:34:47 <smallville7123> trying to see what time.h has for 64-bit platforms and 32-bit platforms
2022-12-26 16:35:00 <psykose_> that has nothing to even do with clang it's a libc header
2022-12-26 16:35:30 <smallville7123> how do i obtain it
2022-12-26 16:35:51 <psykose_> https://git.musl-libc.org/cgit/musl/tree/include/time.h
2022-12-26 16:35:54 <psykose_> or from musl-dev
2022-12-26 16:36:40 <smallville7123> apk has a :arch thing when installing packages, right?
2022-12-26 16:36:54 <psykose_> you cannot install packages for a different arch
2022-12-26 16:36:56 <smallville7123> eg      musl:x86 or something
2022-12-26 16:37:03 <psykose_> that header is also the same on all of them
2022-12-26 16:37:05 <psykose_> just to save you some time
2022-12-26 16:37:11 <smallville7123> ok
2022-12-26 16:37:46 <psykose_> to install for another arch you'd have to make a new root first and pass --root to every operation, and then you can keep stuff in there for a sysroot or whatever
2022-12-26 16:37:52 <psykose_> but this sounds completely unrelated to what you want to be doing so far
2022-12-26 16:39:38 <psykose_> (technically i lied and the bits/ include varies per arch so that is different, but i still have no idea what you need it for)
2022-12-26 16:39:43 <smallville7123> where can i find the definition for time_t and timespec
2022-12-26 16:40:18 <artix-live-01> What's up?
2022-12-26 16:42:43 <psykose_> https://git.musl-libc.org/cgit/musl/tree/include/alltypes.h.in
2022-12-26 16:45:11 <smallville7123> hmm TYPEDEF _Int64 time_t;
2022-12-26 16:45:40 <smallville7123> is this true for 32-bit systems as well?
2022-12-26 16:45:44 <psykose_> yes
2022-12-26 16:46:05 <smallville7123> ok
2022-12-26 16:47:02 <smallville7123> STRUCT timespec { time_t tv_sec; int :8*(sizeof(time_t)-sizeof(long))*(__BYTE_ORDER==4321); long tv_nsec; int :8*(sizeof(time_t)-sizeof(long))*(__BYTE_ORDER!=4321); };
2022-12-26 16:47:06 <smallville7123> what does this mean
2022-12-26 16:49:48 <psykose_> invisible struct padding depending on if the host is big-endian
2022-12-26 16:50:31 <smallville7123> in what cases would such be needed?
2022-12-26 16:51:37 <psykose_> the git blame will tell you
2022-12-26 16:53:31 <smallville7123> hmm   "the kernel interfaces use a timespec layout padded to match the representation of a pair of 64-bit values, which requires endian-specific padding."
2022-12-26 16:55:44 <psykose_> on 64-bit it evaluates to time_t long
2022-12-26 16:55:58 <psykose_> on 32-bit it evaluates to either time_t :32 long or time_t long :32 depending on endian
2022-12-26 16:56:54 <smallville7123> ok :)
2022-12-26 16:57:13 <smallville7123> can i assume it is safe to compile with clang ?
2022-12-26 16:57:24 <psykose_> compile what with clang
2022-12-26 16:58:45 <smallville7123> the header
2022-12-26 16:59:01 <psykose_> how do you compile a header
2022-12-26 16:59:08 <Ermine> Let's put it this way: what are you trying to accomplish?
2022-12-26 16:59:08 <pj[m]> 
ACTION pats psykose_

2022-12-26 16:59:36 <psykose_> 
ACTION makes pj[m] time64 safe

2022-12-26 17:00:16 <pj[m]> 
ACTION upgrades psykose_ to musl 1.2.3

2022-12-26 17:00:21 <smallville7123> lol
2022-12-26 17:00:35 <Ermine> 1.2.4 when
2022-12-26 17:00:47 <smallville7123> port timespec to windows
2022-12-26 17:00:50 <psykose_> probably a few weeks to a month
2022-12-26 17:01:51 <psykose_> you're trying to "port timespec to windows"?
2022-12-26 17:01:51 <Ermine> smallville7123: what do you mean? timespec is a structure, just make the same structure in your windows code. What is your final goal?
2022-12-26 17:02:23 <smallville7123> im gonna assume __BYTE_ORDER exists in clang as a define
2022-12-26 17:02:24 <pj[m]> like, what did you try to do before you went all that rabbithole
2022-12-26 17:02:49 <Ermine> Why to assume if you can check docs
2022-12-26 17:03:11 <psykose_> what you're trying to do doesn't even make any sense so far
2022-12-26 17:03:53 <smallville7123> cant be bothered finding if __BYTE_ORDER is a definitiion somewhere in header files or define dby compiler itself
2022-12-26 17:04:17 <Ermine> smallville7123: please answer pj[m]'s question
2022-12-26 17:08:19 <smallville7123> pj[m]: idk
2022-12-26 17:10:36 <pj[m]> aha
2022-12-26 17:11:45 <pj[m]> well, we idk too
2022-12-26 17:41:47 <smallville7123> ok, i now have timespec and timeval correct :)
2022-12-26 19:12:36 <scorpion2185[m]1> my tablet camera is not shown on cheese
2022-12-26 19:12:52 <scorpion2185[m]1> and can I hiberante?
2022-12-26 20:43:56 <Guest483> Hey all, this is my first time using IRC believe it or not. I came here to ask an alpine-setup question but can't see any past chats so not sure if I just fire away or what?
2022-12-26 20:44:30 <Lord> you can ask what you want :-)
2022-12-26 20:44:30 <zcrayfish> hi there, just ask your question and provide a reasonable amount of details if possible... If someone knows the answer they'll pipe up
2022-12-26 20:44:47 <Guest483> Ok sweet! 
2022-12-26 20:44:50 <zcrayfish> A lot of people idle for long periods on IRC, so your answer might take hours. Fair warning so you don't just run off after 5 minutes xD
2022-12-26 20:44:51 <Lord> you won't see what happened when you were not online it's normal with irc
2022-12-26 20:49:51 <Guest483> happen next. I've created the partitions manually at this point and I was wondering if my understanding is correct before I proceed any further.
2022-12-26 20:49:51 <Guest483> I am trying to do a "data" disk mode install and I have what I thought is a common use-case but can't find any info documenting the use-case so I'm thinking I might be wrong about my assumptions. I would like to do a data-disk install but with a /boot partition. My current thinking is that I have to do a manual disk partitioning, but then I'm finding various instructions afterwards and sort of lose my grip on what needs to 
2022-12-26 20:58:41 <ikke> It's indeed not a common setup
2022-12-26 20:59:08 <ikke> You want to run-from-ram system where the bootloader + /var is on the disk?
2022-12-26 21:10:35 <Guest483> Yes, otherwise you have to leave your bootable media plugged in. I would think most people would prefer to have the bootloader be on disk. I can be convinced otherwise though. What is the pro of having it perpetually boot from media?
2022-12-26 22:31:31 <minimal> Guest483: I don't think that can be done, my understanding is that Data mode is exactly that - Data is stored on a disk/SSD/USB stick and everything else comes from the boot media
2022-12-26 22:45:14 <Guest483> Ahhh, i figured it could boot from the harddisk and still load the OS into RAM. I will piddle around with it. I guess I don't mind it booting from the usb. 
2022-12-26 22:54:51 <minimal> Guest483: yes Data mode is where the OS is loaded into RAM, my point is that the only different from normal run-from-ram and Data mode is that in Data mode data is stored on a separate partition rather than loaded via LBU/apkoverlay
2022-12-27 07:34:09 <sicelo> how can one know ahead of time what https://dl-cdn.alpinelinux.org/alpine/v3.17/releases/armv7/alpine-uboot-3.17.0-armv7.tar.gz contains? i.e. package list?
2022-12-27 08:15:09 <clandmeter> sicelo: should be defined in the scripts section in aports
2022-12-27 08:16:29 <clandmeter> https://git.alpinelinux.org/aports/tree/scripts
2022-12-27 08:29:40 <sicelo> thank you
2022-12-27 10:42:19 <chris-alpine> hello, I'm trying a diskless install of apline-3.17.0. but failing to get lbu working. System boots with /dev/sdb mounted on /media/sdb and nothing mounted on /media/usb. I followed the instructions for manual paritioning for diskless but cf disk removes the partition on /dev/sdb and then grub doesn't boot
2022-12-27 10:58:27 <chris_alpine> hello, I'm trying to install alpine 3.17.0 diskless but I'm failing to get lbu to work. After boot /dev/sdb (my usb stick) is mounted on /media/sdb, Any attempt to mount /dev/sdbX results in "Resource Busy" error. I've tried the manual partitioning instructions for diskless, but cfdisk removes the /dev/sdb partition which prevents the system booting
2022-12-27 10:59:50 <Ermine> No need to repeat your message. Wait until somebody comes and helps you
2022-12-27 11:00:54 <chris_alpine> sorry, I'm new to IRC, web client booted me after a few minutes so I installed hexchat and tried again, waiting now :-)
2022-12-27 11:04:07 <ikke> If /dev/sdb is mounted (meaning it's a partition), how can /dev/sdbX exist?
2022-12-27 11:06:18 <chris_alpine> fdisk /dev/sdb shows /dev/sdb1 and /dev/sdb2. cfdisk (from the manual paritioning instructions) shows /dev/sdb /dev/sdb1 /dev/sdb2, but when I create /dev/sdb3 then /dev/sdb is removed
2022-12-27 11:08:17 <chris_alpine> I'm starting with a clean install of alpine-3.17.0 on this usb stick, so if /dev/sdb is mounted as iso9660 at boot, how do I create a writable parition for lbu?
2022-12-27 11:09:19 <ikke> You need to use setup-disk (or manually do it)
2022-12-27 11:10:02 <ikke> iso9660 is read-only, so directly writing that to USB would not work
2022-12-27 11:11:42 <chris_alpine> setup-disk can only see sda which is the HDD of the host machine which I don't really want to overwrite until I'm confident that I can make alpine work
2022-12-27 11:14:18 <chris_alpine> for diskless (or data) installs, where should lbu store config changes so that they can be applied when the system reboots?
2022-12-27 11:34:06 <clandmeter> chris_alpine: what os your boot media?
2022-12-27 11:35:27 <chris_alpine> sandisk 32GB usb stick
2022-12-27 11:35:37 <clandmeter> how did you set it up?
2022-12-27 11:35:45 <clandmeter> just dd iso to it?
2022-12-27 11:36:36 <chris_alpine> I used balena etcher
2022-12-27 11:37:11 <ncopa> chris_alpine: do you boot with efi or legacy bios?
2022-12-27 11:38:18 <chris_alpine> bios claims UEFI:USB
2022-12-27 11:38:46 <ncopa> you can create a GPT partition table, create a single ESP partition, create a FAT filesystem on it, and count it and then: bsdtar -xf /path/to/alpine.iso
2022-12-27 11:39:06 <ncopa> s/count/mount/
2022-12-27 11:39:06 <alpine-meetbot> ncopa meant to say: you can create a GPT partition table, create a single ESP partition, create a FAT filesystem on it, and mount it and then: bsdtar -xf /path/to/alpine.iso
2022-12-27 11:39:29 <chris_alpine> ok, I'll try that
2022-12-27 11:41:17 <ncopa> basically, just extract all the files on the iso to an ESP partition on usb stick
2022-12-27 12:25:51 <chris_alpine> that seems to work much better, many thanks
2022-12-27 12:26:20 <jh81> :-)
2022-12-27 14:25:53 <tonic> Hello. Can you please help and refer me to the required package for curl --negotiate feature (for kerberos auth on proxy)? libcurl/7.86.0 version doesn't support this :(
2022-12-27 14:48:07 <Ermine> What does it say?
2022-12-27 15:45:49 <scorpion2185[m]1> is this the init script usr/share/mkinitfs/initramfs-init
2022-12-27 15:45:49 <scorpion2185[m]1>  ?
2022-12-27 15:46:04 <scorpion2185[m]1> can I hibernate ? lvm on luks
2022-12-27 22:31:05 <lopid> sure
2022-12-28 00:43:30 <thresh> Hi.  I'm looking to add a fuzzy dependancy for between our "modules" and "main" packages.  Looking at https://git.alpinelinux.org/apk-tools/commit/?id=693b4bcdb0f22904a521a7c8ac4f13e697dc4d71, depends="foo~3.14" should be enough.  However in aports.git I see various different ways with `~` in it: "=~", "~", "~=", which one is correct? https://gist.github.com/thresheek/4320340cff53d5eb049c5c201f485f5f
2022-12-28 02:45:32 <bleb> just the other day I was able to successfully compile my latex project
2022-12-28 02:45:42 <bleb> now I'm getting ! LaTeX Error: File `llncs.cls' not found.
2022-12-28 02:46:03 <bleb> any idea what could have gone wrong?
2022-12-28 02:46:24 <bleb> I am poking around in /usr/share and haven't been able to find llncs.cls, but it had no problem just days ago
2022-12-28 02:58:58 <thresh> bleb, maybe reinstalling (or checking) this package will do the trick: https://pkgs.alpinelinux.org/contents?file=llncs.cls&path=&name=&branch=v3.17
2022-12-28 07:36:38 <tonic> @Ermine when i try to use curl with --negotiate it says that my libcurl version doesn't support this
2022-12-28 08:12:49 <ikke> tonic: apparently it would require curl to be compiled against libgss, which at the moment it isn't
2022-12-28 08:17:34 <tonic> @ikke Yes, you're right. I installed this library separately, but nothing happened. Thank you for support and have i nice day 😉 
2022-12-28 15:38:46 <vkrishn> is it possible to use gmail service on non-desktop alpinelinux setup ?
2022-12-28 15:44:38 <ikke> Some imap client should work
2022-12-28 15:57:31 <vkrishn> which ones pls, wanting to do away with thunderbird
2022-12-28 15:58:40 <vkrishn> oh, not pop3 ?
2022-12-28 15:58:52 <ikke> No
2022-12-28 16:04:48 <elly> TIL: while ash has no default rc file, if the environment variable ENV is set at shell startup, it sources the mentioned file
2022-12-28 16:05:07 <ikke> Yes, indeed
2022-12-28 16:05:30 <elly> finally gpg commit signing works properly with gpg-agent for me :)
2022-12-28 18:46:13 <scorpion2185[m]1> can I use osk-sdl on alpine? pmos made it
2022-12-28 18:57:47 <elly> I don't see why you couldn't but it does not appear to be packaged for alpine yet
2022-12-28 18:59:11 <Saklad5> How long does it usually take for a package maintainer to respond to an issue?
2022-12-28 19:01:10 <elly> that hugely depends on the individual person
2022-12-28 19:01:18 <elly> it is also a holiday time in a lot of places right now so probably extra long
2022-12-28 19:01:22 <Saklad5> Fair point
2022-12-28 19:02:16 <Saklad5> I've been waiting weeks, and it'd take me seconds to submit a patch. And probably days to figure out how to test it
2022-12-28 19:03:34 <Saklad5> I've looked into configuring local packages, but I'm not sure how practical it is
2022-12-28 19:04:18 <Saklad5> Especially since the package in question is an authoritative DNS resolver, and for the change to be usable it has to be replicated across multiple machines
2022-12-28 19:05:03 <Saklad5> Does anyone here run their own APK repository?
2022-12-28 19:11:21 <elagost> Saklad5: I do on a home server. I just rsync my abuild directory (~/packages/ by default I think) to a webserver and add that to /etc/apk/repositories on other systems, and copy the build server's key to /etc/apk/keys/.
2022-12-28 19:14:18 <Saklad5> Can APK use client certificates?
2022-12-28 19:14:33 <Saklad5> I'm not asking for a specific reason, just wondering
2022-12-28 19:18:51 <elly> to authenticate requests to the repository server?
2022-12-28 19:20:14 <Saklad5> Yes, on the principle that it is easier to make things public than private after the fact
2022-12-28 19:20:25 <thresh> Saklad5, yes, it can do that
2022-12-28 19:20:50 <thresh> not flexible at all though
2022-12-28 19:21:10 <Saklad5> What do you mean?
2022-12-28 19:22:34 <thresh> I mean you need to put both cert and key to /etc/apk/cert.pem, and there is no way to tell apk for which repos to use those AFAIK
2022-12-28 19:23:34 <thresh> also, not sure what happens if you happen to have two repositories with distinct cert/keys
2022-12-28 19:27:41 <elly> I guess the intended use case is for if you have a single internal repo mirror, and then that mirror is supposed to be the only source you have & contain clones of anything else you want
2022-12-28 19:31:53 <scorpion2185[m]1> any help implement it ? I installed from pmos
2022-12-28 19:40:38 <bleb> <thresh> bleb, maybe reinstalling (or checking) this package will do the trick: https://pkgs.alpinelinux.org/contents?file=llncs.cls&path=&name=&branch=v3.17
2022-12-28 19:40:51 <bleb> thresh: what did you mean by checking?
2022-12-28 19:44:54 <bleb> ahh the problem is I'm still on 3.15
2022-12-28 20:00:49 <Saklad5> I have a copy of Alpine Linux 3.14 running in an iOS app
2022-12-28 20:01:04 <Saklad5> And I'm sort of terrified to update it without waiting for the developer
2022-12-28 20:01:38 <Saklad5> Because I know for a fact that it isn't completely POSIX-compliant
2022-12-28 23:02:25 <paperowl> New to the distro coming from debian, curious if you use alpine for security on a bare/host/VM server, do most veteran users here manually harden kernels or pay for the commercially hardened kernel? If so is there an alpine specific guide or just use generic hardening. Thanks 
2022-12-28 23:02:58 <elly> I am using the stock kernel
2022-12-28 23:03:46 <paperowl> Have you ever looked into the hardening or patches?
2022-12-28 23:04:44 <paperowl> Or at least strip down unnecessary features 
2022-12-28 23:06:00 <elly> I have not, the alpine kernel is pretty small as it is
2022-12-28 23:06:30 <paperowl> https://wiki.alpinelinux.org/wiki/Custom_Kernel is decent starter guide
2022-12-28 23:06:36 <elly> especially if you're using virtual
2022-12-28 23:07:03 <paperowl> Fair enough, just curious if its worth the trouble. Thanks for your reply
2022-12-28 23:09:28 <elly> hmm, I am far from an expert but I would guess that if you are running a stock alpine kernel and keeping your system up to date the attack surface will be quite small
2022-12-28 23:09:58 <Habbie> what "commercially hardened kernel" ?
2022-12-28 23:10:39 <q66> grsecurity probably?
2022-12-28 23:10:58 <Habbie> if you do a custom kernel, is there not a risk of you forgetting to update it?
2022-12-28 23:11:35 <paperowl> Yes thats the one, there are a few guides to harden generically,  the alpine wiki one is incomplete
2022-12-28 23:11:49 <fossy> is it possible to do ntp in initramfs, before apkvol is applied, (for repo certificates), when pxe booting? if not i can just drop https (ie change to http) but wanted to check first
2022-12-28 23:12:13 <paperowl> Yes and also a risk of screwing up, however I admire alpines philosophy of minimalism, so thought kernel tweaks fit that philosophy 
2022-12-28 23:12:28 <fossy> (repo certificates -> ssl certificates)*
2022-12-28 23:51:30 <minimal> fossy: I assume your time "problem" is because you are using an RPI and haven't got a retro-fitted RTC? Those can be bought for approx $5
2022-12-28 23:56:25 <Habbie> (raspberry pi os resets the clock to the time at last shutdown - if you have any kind of local storage, the same trick might be good enough here for certificate validation)
2022-12-28 23:57:01 <dwfreed> enable swclock in openrc
2022-12-28 23:57:05 <dwfreed> it does that
2022-12-28 23:57:12 <Habbie> cool
2022-12-28 23:57:18 <Habbie> is that early enough for fossy's purposes?
2022-12-28 23:57:58 <dwfreed> I'm not sure, I don't use alpine
2022-12-29 00:00:09 <Habbie> same
2022-12-29 00:04:04 <minimal> Habbie: he's network booting. With both Raspios and Alpine the clock is set by the "full" OS during boot, in his case he obviously wants it set before then, i.e. in the initramfs
2022-12-29 00:04:14 <Habbie> yes, this is understood
2022-12-29 00:04:32 <Habbie> i just wasn't sure whether setting openrc flags affects what happens in initramfs
2022-12-29 00:04:34 <Habbie> (in alpine
2022-12-29 00:04:36 <Habbie> )
2022-12-29 00:04:44 <minimal> initramfs doesn't use openrc
2022-12-29 00:04:49 <Habbie> ack
2022-12-29 00:05:52 <minimal> initramfs' init is just a shell script and he wants the clock correct so that when the rootfs is fetched via HTTPS then the clock is correct so that the HTTPS server's cert is not rejects as not yet valid/expired
2022-12-29 00:06:02 <Habbie> also understood
2022-12-29 00:06:33 <minimal> so obviously nothing to do with openrc can be used as that is in the rootfs which can't be fetches until after the clock is corrected
2022-12-29 00:07:39 <minimal> and to store a "date/time of last boot" in the initramfs itself would mean modifying the initramfs upon every shutdown
2022-12-29 00:15:35 <Habbie> indeed
2022-12-29 03:52:28 <elly> in the past when I have had this problem (on a system with no battery RTC) I have had to do the initial https fetch with the client configured to ignore the time for cert validation
2022-12-29 09:32:43 <Alpine_newbie_0> Hi, I'm new to IRC (and a bit to Alpine), I'm lloking for help regarding fsck in alpine
2022-12-29 09:33:11 <Alpine_newbie_0> *looking
2022-12-29 09:38:14 <ikke> Alpine_newbie_0: Welcome here. Please ask your question. If anyone can help you, they'll answer you
2022-12-29 10:08:42 <Alpine_newbie_0> @ikke Thank you. My problem is an Alpine in a container (CT) in Proxmox : each time there's a (micro) power outage (I don't why, a few these times), the filesystem of the alpine container gets corrupted -essentially a docker/jlesage/firefox image that was running-. I dont succeed in programming fsck to repair the fs at boot, so the images dont work anymore. How to script fsck to launch and repair at boot, pls ?
2022-12-29 10:22:40 <Justin[m]> If I might suggest fixing the root of the problem with a UPS.
2022-12-29 10:29:02 <Alpine_newbie_0> @Justin[m] Yes, for sure, currently tracking offers on local ebay (and "improving skills on battery replacement") ... but was thinking improving skills on fsck/alpine/openrc could be good too :)
2022-12-29 10:31:13 <ikke> There is an fsck service, did you try to enable that?
2022-12-29 10:31:32 <ikke> but if the filesystem is already corrupt, that'
2022-12-29 10:31:36 <ikke> that's not helping
2022-12-29 10:32:50 <ncopa> Alpine_newbie_0: what filesystem?
2022-12-29 10:33:43 <Alpine_newbie_0> FS is ext4
2022-12-29 10:34:03 <Alpine_newbie_0> I could add fsck in alpine with apk add fsck
2022-12-29 10:34:23 <Alpine_newbie_0> enabling it as a service would be ... ?
2022-12-29 10:36:41 <Alpine_newbie_0> sorry, adding fsck was with apk add util-linux I think
2022-12-29 10:40:05 <Alpine_newbie_0> fsck not installed as a service, I'm trying to get it that way
2022-12-29 10:50:46 <Alpine_newbie_0> Just typed "rc-update add fsck boot", got "fsck added", rebooted, no check (CT booted in 1 second), docker image remained HS ("Error: unsupported file format")
2022-12-29 10:51:25 <ikke> Hard to tell what needs to happen as I'm not familiar with that setup
2022-12-29 10:52:44 <Alpine_newbie_0> Yes, sure, no pb. Is there a way to script in sth like /etc/init... to provoke fsck cehck at boot, maybe ?
2022-12-29 10:54:17 <ikke> Where does it get in the boot process? If the fs is corrupt, does it even get to starting init / openrc?
2022-12-29 10:57:34 <Alpine_newbie_0> I think fsck did not launch at boot because of /fastboot present, trying to remove it
2022-12-29 13:38:31 <Alpine_newbie_0> @ikke the boot process gets til the end, the error is in the process of launching the firefox image of docker
2022-12-29 13:39:08 <Alpine_newbie_0> I still cannot launch fsck "from the inside" (probably because of /fastboot I currently cannot get rid of)
2022-12-29 13:40:25 <Alpine_newbie_0> I could find a workaround to launch fsck "by the outside", shutting down the alpine container and running "pct fsck CTID" : got "clear" from fsck
2022-12-29 13:41:29 <Alpine_newbie_0> But the docker image seems always corrupted : well, at least it doesn't comes from the FS of the alpine system, ending with fsck search, thanks for all
2022-12-29 14:16:02 <ncopa> actually, I think i saw the fsck error on my macbook, it was quite annoying
2022-12-29 14:32:57 <vkrishn> fsck is being referred to busybox fsck ?
2022-12-29 14:39:18 <ncopa> possibly
2022-12-29 14:39:30 <ncopa> fsck is just a frontend for fsck.ext4
2022-12-29 14:44:36 <vkrishn> which needs e2fsprogs to be installed, but that pkgs is not installed by default
2022-12-29 14:46:40 <vkrishn> anyway one asking is not here, no sweat
2022-12-29 14:57:37 <minimal> for a docker container the backing filesystem (in this case ext4) is part of /var/lib/docker on the host so why would someone run fsck inside the container rather than on the host?
2022-12-29 15:23:51 <vkrishn> I stopped getting more into docker once got going with lxc/qemu and mobiles ;)
2022-12-29 16:12:05 <frag> what browsers work well now? firefox/qutebrowser/(e)links/lynx? any other?
2022-12-29 16:38:40 <pj[m]> chromium
2022-12-29 16:46:11 <frag> pj[m]: oh thanks, will check it out!
2022-12-29 16:46:50 <pj[m]> you can also just get flatpak and install whatever browser
2022-12-29 16:52:17 <frag> tbh i dunno which browser would be good for a slow (rpi4) system, have been using qutebrowser quite successfully but sometimes its quite slow indeed
2022-12-29 16:56:08 <frag> chromium does not seem lighter
2022-12-29 16:58:07 <hsv2> bloatweb isn't for sbc
2022-12-29 16:58:36 <hsv2> if it does'nt look fine in links it shouldn't exist
2022-12-29 16:59:01 <frag> hsv2: sadly that is not the world were living in :(
2022-12-29 17:01:06 <psykose> there's also epiphany
2022-12-29 17:01:10 <psykose> and some other webkit browsers
2022-12-29 17:01:18 <psykose> i have no idea what would be fastest on an sbc itself
2022-12-29 17:01:27 <psykose> you'll have to compare :)
2022-12-29 17:03:23 <donoban> frag: did you disable js on qutebrowser?
2022-12-29 17:03:27 <hsv2> frag: crazy stallman only browses website that runs with their freaky LibreJS bloquer. So I guess it is _possible_ to survive without it :D
2022-12-29 17:04:51 <elly> you can definitely disable JS if you are fine with some websites just not working or partly not working
2022-12-29 17:04:58 <elly> whether you are fine with that is up to you though
2022-12-29 17:06:08 <elly> ftr, chromium runs just fine on an rpi, but some websites are slow - like most browsers the vast bulk of the work is done in rendering the page and running its JS so it highly highly depends what page you want to view
2022-12-29 17:09:43 <frag> psykose: i thikn webkit crashes in like 2sec??? otherwise id like to try vimb..
2022-12-29 17:10:02 <frag> donoban: sometimes, but so many sites need it :/
2022-12-29 17:10:42 <frag> maybe its time to be ... CRAZY ... hehe
2022-12-29 17:13:51 <donoban> I have it disabled by default and enable with 'tsh' when needed
2022-12-29 17:16:12 <frag> yeah i can toggle it with 's', but yeah so many sites need it ... some sites are hard to browse with js enabled tho :p
2022-12-29 18:05:51 <jjuran> Some sites DoS you with animated GIFs only when JS is disabled
2022-12-29 18:23:56 <elagost> newbie question: how does one generate a .patch file like the ones here: https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/main/linux-lts ?
2022-12-29 18:24:46 <ikke> easiest by using git
2022-12-29 18:24:56 <elagost> Using diff is incomplete and it looks like that is generated with git on the actual kernel tree?
2022-12-29 18:24:58 <ikke> git diff generates a valid patch file
2022-12-29 18:25:17 <ikke> yes, you would diff it on the original tree
2022-12-29 18:26:59 <elagost> that won't generate the big section up top though, just from `diff --git` down, right? Just tested on a repo I have already.
2022-12-29 18:27:46 <minimal> jjuran: there is a firefox setting to not loop animated gifs
2022-12-29 18:34:08 <Ermine> I use diff -Nurp
2022-12-29 18:34:54 <Ermine> Can supervise-daemon redirect daemon stdout/stderr to syslog. I skimmed supervise-daemon and cannot find anything relevant
2022-12-29 18:46:28 <jjuran> minimal: Good to know.  There's also uBlock Origin :-)
2022-12-29 18:48:11 <minimal> jjuran: I prefer uMatrix but that's no longer maintained :-( Though I think there was a fork of it
2022-12-29 18:52:37 <pj[m]> elagost: git show and git format-patch will give you mail patch format
2022-12-29 18:59:22 <psykose> Ermine: iirc no
2022-12-29 19:03:44 <scorpion2185[m]1> should I ask in some forum about hibernation?
2022-12-29 19:04:30 <pj[m]> what's a forum
2022-12-29 19:13:16 <minimal> scorpion2185[m]: you just did
2022-12-29 19:22:15 <Ermine> Well, can s-s-d redirect stdin and stdout to the same location then? (Maybe -2 /dev/fd/1 would work?)
2022-12-29 19:22:50 <psykose> s-s-d can use logger -t to go to syslog from stdout/err
2022-12-29 19:23:28 <psykose> via output_logger= error_logger=
2022-12-29 19:23:47 <psykose> supervise-daemon only has output_log= error_log= and it has to be a file so idk
2022-12-29 19:26:27 <Ermine> It's not openrc init.d script, so it wouldn't work this way I guess.
2022-12-29 19:26:43 <psykose> they have argument equivalents
2022-12-29 19:26:46 <Ermine> But I see the direction, thank you
2022-12-29 19:26:54 <psykose> the init.d var= is just declarative form for an equivalent --flag
2022-12-29 19:42:23 <jh81> heh, well that was the last bit of Debian in my house
2022-12-29 19:42:31 <jh81> All is Alpine now :D
2022-12-29 19:43:24 <Ermine> jh81: Congrats
2022-12-29 19:43:47 <jh81> :Y
2022-12-29 19:47:20 <Saijin_Naib> jh81: Worthy goal! I've got two appliances left, both ARM, so not easy for me 
2022-12-29 19:48:32 <jh81> well, i have a libreelec on raspberry running still, now that i think of it
2022-12-29 19:48:41 <jh81> crap
2022-12-29 19:55:12 <hsv2> I have an issue with monero-0.18.1.2-r1 package. It dies without having the time to show anything. I'm no expert but is that a seccomp thing?
2022-12-29 19:55:12 <hsv2> traps: monerod[10378] trap invalid opcode ip:559f2a3e4ebf sp:7ffedd2974b0 error:0 in monerod[559f2a3df000+485000]
2022-12-29 19:57:59 <hsv2> any better idea than rebuilding the pkg with debug info? I (still) don't even how to do it on alpine…
2022-12-29 19:59:21 <ikke> alpine is not using seccomp (unless you run it in docker)
2022-12-29 19:59:42 <ikke> hsv2: what platform / arch are you on?
2022-12-29 20:00:58 <hsv2> x86-64, sys install, edge
2022-12-29 20:01:25 <hsv2> (i didn't know i could just install _some_ package from edge and i did a full edge)
2022-12-29 20:01:51 <ikke> mixing stable with edge is not supported
2022-12-29 20:01:55 <ikke> so that's good
2022-12-29 20:02:03 <hsv2> ah ok
2022-12-29 20:02:30 <Saijin_Naib> <jh81> "well, i have a libreelec on..." <- New project 🤣
2022-12-29 20:02:37 <jh81> indeed
2022-12-29 20:02:51 <jh81> the repo has both kodi and jellyfin, so that should be fun :)
2022-12-29 20:04:08 <hsv2> ikke, idk if that has anything to do but the cpu is a i3 M 330
2022-12-29 20:04:16 <hsv2> old one
2022-12-29 20:04:28 <ikke> hsv2: what does /etc/cpuinfo show for flags?
2022-12-29 20:04:39 <Saijin_Naib> Oh, I had issues on my i5-560m
2022-12-29 20:05:06 <Saijin_Naib> Missing instructions for some packages that are compiled for better architecture baseline
2022-12-29 20:05:11 <hsv2> flags		: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ht tm pbe syscall nx rdtscp lm constant_tsc arch_perfmon pebs bts nopl xtopology nonstop_tsc cpuid aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm sse4_1 sse4_2 popcnt lahf_lm pti tpr_shadow vnmi flexpriority ept vpid dtherm arat
2022-12-29 20:06:11 <hsv2> are CFLAGS adjustable for alpine, hmm pkg build system?
2022-12-29 20:15:15 <psykose> fixed with ce7472ac5ee66b029047ebf5ea17b3672579ffc9
2022-12-29 20:15:43 <psykose> or maybe it's not, can't actually test it
2022-12-29 20:27:49 <hsv2> i'll do it
2022-12-29 21:11:37 <hsv2> psykose, it seems to work
2022-12-30 01:46:20 <fossy> Habbie: yes, one is an raspberry pi :P
2022-12-30 01:46:30 <fossy> thanks for the help!
2022-12-30 01:46:52 <ant64b> ayyy another alpine pi user
2022-12-30 01:46:54 <ant64b> :)
2022-12-30 10:37:39 <Habbie> does anybody happen to know how i can get names and not just numbers out of lsusb?
2022-12-30 10:39:21 <ncopa> maybe you need hwids database, or what it is called nowadays
2022-12-30 10:39:46 <ncopa> hwdata-usb
2022-12-30 10:40:08 <Habbie> that wasn't it
2022-12-30 10:40:12 <Habbie> i see on debian lsusb gets it from /sys/udev
2022-12-30 10:40:23 <ptrc> Habbie: install usbutils
2022-12-30 10:40:31 <ptrc> it has proper lsusb instead of the busybox one
2022-12-30 10:40:35 <Habbie> i thought i had usbutils
2022-12-30 10:40:40 <Habbie> ah i see what went wrong here
2022-12-30 10:40:51 <Habbie> i'm unfamiliar with apk and was inspecting an available package, not an installed one
2022-12-30 10:40:53 <Habbie> thank you, this works
2022-12-30 18:42:03 <scorpion2185[m]1> dbus.exceptions.DBusException: org.freedesktop.DBus.Error.NameHasNoOwner: Could not get owner of name 'org.ofono': no such name
2022-12-30 18:51:53 <scorpion2185[m]1> what's the name of libdbd-sqlite3-perl libgtk3-perl in alpine?
2022-12-30 18:52:20 <scorpion2185[m]1> Can't locate Gtk3.pm in @INC (you may need to install the Gtk3 module) (@INC contains: /usr/local/lib/perl5/site_perl
2022-12-30 19:09:20 <scorpion2185[m]1> how do you tell a package depend this | or this?
2022-12-30 19:10:12 <ikke> apk has no alternatives
2022-12-30 19:10:28 <ikke> at most you can depend on a virtual package that is provided by multiple other (but you need to be careful)
2022-12-30 19:21:54 <scorpion2185[m]1> I made a river-shifttags
2022-12-30 19:23:01 <scorpion2185[m]1> pkg
2022-12-30 19:35:52 <j_v> cut and paste to xterm stopped working, was working yesterday or day before pretty sure
2022-12-30 19:36:27 <j_v> anyone else having similar issue?
2022-12-30 19:37:18 <ikke> What version of Alpine?
2022-12-30 19:37:22 <j_v> edge
2022-12-30 19:37:35 <ikke> Not sure if it's the issue, but libx11 has been downgraded
2022-12-30 19:37:45 <ikke> due to issues
2022-12-30 19:37:52 <ikke> apk version libx11
2022-12-30 19:38:26 <j_v> yep, saw that. i downgraded already to 1.8.2-r2
2022-12-30 19:38:41 <ikke> ok, then I don't know
2022-12-30 19:38:52 <j_v> thanks for suggestion, though... appreciate it
2022-12-30 19:46:40 <j_v> only other change I see in the last 3-4 days that seems like it might have an impact is mesa update
2022-12-30 19:54:46 <psykose> you'll have to be a bit more specific than "stopped working" :)
2022-12-30 19:56:04 <j_v> right, so, highlight something in one xterm (right click, drag) then middle click in another xterm does not paste
2022-12-30 19:57:01 <j_v> i tried xclip to see if the selection was available, it was
2022-12-30 20:02:34 <j_v> downgrading mesa appears to messy
2022-12-30 20:02:52 <j_v> s/to /to be /
2022-12-30 20:02:52 <alpine-meetbot> j_v meant to say: downgrading mesa appears to be messy
2022-12-30 20:05:06 <psykose> xterm doesn't even use gl or vulkan apis
2022-12-30 20:05:28 <j_v> true
2022-12-30 20:05:48 <psykose> also it works for me
2022-12-30 20:05:53 <psykose> (drag middle paste)
2022-12-30 20:06:06 <j_v> ok, thanks for checking.
2022-12-30 20:06:19 <psykose> but it's in xwayland and with nothing else special
2022-12-30 20:06:30 <psykose> all i'd guess is perhaps fancy x settings or some clipboard management stuff
2022-12-30 20:07:19 <j_v> i don't think i changed anything in my x startup and haven't installed anything that is even x related
2022-12-30 20:07:59 <j_v> yeah, thought of that too... created new user, bare .xinitrc with twm only... no work either
2022-12-30 20:08:33 <j_v> must be i changed something system wide, just no seeing it
2022-12-30 20:12:58 <j_v> oh, silly me... sorry for the noise. just tried a different mouse, works fine.
2022-12-30 20:13:07 <j_v> flaky old mouse i guess
2022-12-30 20:13:33 <ikke> heh
2022-12-30 20:14:09 <ikke> I dropped my mouse, and now the middle wheel is very sensitive
2022-12-30 20:14:43 <minimal> ikke: better than your toe being sensitive ;-)
2022-12-30 20:14:53 <ikke> absolutely
2022-12-30 20:16:01 <psykose> j_v: happens, no worries :)
2022-12-30 20:19:48 <j_v> thanks folks, i appreciate the help though, i'll stop blushing from embarassment in a day or two maybe
2022-12-30 20:21:36 <paperowl> Just want to say just set up alpine as my daily driver. I am very impressed with the community and the maintainers. How they managed to organise that many packages and are responsive to questions is incredible 
2022-12-30 20:21:56 <paperowl> Usually when I find a smaller community such as this they are very hard to work with
2022-12-30 20:24:55 <ikke> paperowl: thanks, glad to hear :)
2022-12-30 20:25:03 <psykose> :)
2022-12-30 20:26:48 <paperowl> Is the future of the distro secure? Does it have good governance and plans to ensure continuity under potential difficulties that may arise ?
2022-12-30 20:38:00 <ikke> paperowl: one thing we are still looking at is setting up some legal entity
2022-12-30 20:38:51 <ikke> we do have a governance structure
2022-12-30 20:39:05 <paperowl> I just read the governance page in the wiki
2022-12-30 20:39:19 <paperowl> Seems a looser very Democratic one
2022-12-30 20:40:03 <paperowl> No benevolent dictators or clear  leadership I can see, I might be wrong 
2022-12-30 20:40:13 <ikke> There is the council
2022-12-30 20:40:19 <ikke> which consists of 3 members
2022-12-30 20:40:27 <ikke> which then oversees the technical steering commitee
2022-12-30 20:40:38 <ikke> (wiki may be outdated)
2022-12-30 20:41:29 <psykose> in practice it's whoever has time to do stuff ;)
2022-12-30 20:42:00 <psykose> it's not really a paid distro with corporate interest, though there is sponsorship for hardware
2022-12-30 20:42:13 <paperowl> It does mention the 3 elected members, the base team. So they make the final say 
2022-12-30 20:43:25 <paperowl> I deliberately steer away from the paid ones, hoping to find Foss that is resilient 
2022-12-30 20:43:37 <paperowl> The paid ones can pull the plug at any time 
2022-12-30 20:44:22 <paperowl> At the same time the Foss ones that don't govern well leave you to have to migrate to other platforms and relearn everything 
2022-12-30 20:44:46 <ikke> Alpine has been around for a while
2022-12-30 20:45:23 <ikke> (which is no guarantee for continuation, though)
2022-12-30 20:45:41 <psykose> insinuating something, ikke? :p
2022-12-30 20:45:43 <paperowl> Im hoping at least because its widely used in containers will guarantee a niche and some degree of future proof
2022-12-30 20:48:42 <ikke> psykose: who knows
2022-12-30 20:49:07 <psykose> now i can't tell if you're joking :p
2022-12-30 20:51:01 <ikke> I am
2022-12-30 20:51:29 <psykose> :)
2022-12-30 20:57:12 <invoked> not personally worried about alpine being "freenoded" but if the 3 amigos all take the same airplane or something...
2022-12-30 22:05:29 <jhjacobs[m]> Hmz
2022-12-30 23:38:15 <bl4ckb0ne> i vote psykose for president
2022-12-31 00:34:14 <pj[m]> I vote psykose for bed
2022-12-31 01:32:24 <ktprograms> Hi, how could I debug podman segfaulting on armhf (Raspberry Pi Zero)? I tried getting a core file (with ulimit -c unlimited) but nothing gets created, and in gdb it seems to spawn a bunch of processes then just stop with no message at all (just "application terminated" or similar).
2022-12-31 09:43:05 <scorpion2185[m]1> is hibernation enable dby default in the kernel ?linux-lts https://wiki.alpinelinux.org/wiki/Custom_Kernel#Hibernation_to_prevent_data_loss
2022-12-31 13:00:41 <WhyNotHugo_> wpa_supplicant doesn't detect my home wifi (but does detect SOME networks), but iwd detects all networks just fine.
2022-12-31 13:01:05 <WhyNotHugo_> Any ideas what this could be? It's mostly annoying when using setup-interfaces
2022-12-31 13:21:47 <psykose> pj[m]: i was actually in bed :)
2022-12-31 13:22:21 <psykose> WhyNotHugo_: maybe the networks are using something special that wpa_supplicant is not compiled with? sounds unlikely, but..
2022-12-31 13:22:31 <psykose> aside from that i'd just guess general wpa_supplicant suckiness
2022-12-31 13:29:20 <hsv2> Question regarding ACPI (i guess). I have an Alpine on a laptop, i did a setup-desktop plasma. Now when I close the LID, the laptop enter suspend state. I want to disable that. On alpine wiki, I see there's information about how to _enable_ this feature with busybox acpid (which i am using). They mention to add files in /etc/acpi but I don't have anything there about the LID events. Am I looking in the 
2022-12-31 13:29:20 <hsv2> wrong place? What other service could handle that? I have also upower and elogind but I have no idea if they are related at all with acpi events
2022-12-31 13:29:52 <hsv2> mmmh irc gateway bug? i see only the end of my last message...
2022-12-31 13:30:58 <hsv2> if that wasnt sent here's the tldr: how to disable suspend on lid close? i did a setup-desktop plasma and now i have that behaviour
2022-12-31 13:31:26 <hsv2> i know or i guess its not acpid because i dont see any conf regarding lid events
2022-12-31 13:34:10 <hsv2> i can disable it in kde but that works only if i init my plasma session.
2022-12-31 13:36:19 <Ermine> hsv2: download non-bb acpid
2022-12-31 13:37:03 <Ermine> Or, afaik plasma requires elogind, it would manage all that stuff if so.
2022-12-31 13:37:56 <hsv2> elogind may be the responsible of the handling of the lid event?
2022-12-31 13:38:32 <Ermine> Yes, it handles all acpid duties
2022-12-31 13:39:33 <hsv2> ok... so i can disable bb acpid too. well i'll look elogind doc then. thx Ermine 
2022-12-31 13:42:00 <Ermine> hsv2: you may also take a look if plasma settings has needed knobs
2022-12-31 13:42:09 <WhyNotHugo_> psykose: Any thoughts on tinkering setup-interfaces to use iwd under the hood?
2022-12-31 13:43:07 <hsv2> another question. i didnt dig much into doc, so i may have missed it. is there any automation ready made to build all abuild tree? and is it possible to set some FLAGS? I'd like to try having my alpine built with -Os and -march=native (sorry, gentoo user here)
2022-12-31 13:44:24 <hsv2> Ermine: what do you mean by knobs? anyway I installed plasma just to test it... I'm already back to sway x)
2022-12-31 13:45:13 <frag> hmm using my wired kb, i have to rerun xmodmap ever so often, whats up with that?
2022-12-31 13:51:09 <psykose> WhyNotHugo_: it uses iw now to actually find the networks iirc, i think that's separate to the wpa_supplicant part
2022-12-31 13:51:28 <psykose> unless you mean for it to generate a final config for iwd instead of wpa_supplicant? i guess someone would have to contribute it
2022-12-31 13:51:39 <psykose> it sounds like a good idea to me
2022-12-31 13:52:12 <psykose> that iw part also sounds like why wpa_supplicant can't see them after, the reason someone added iw is because iwlist could not see their wpa3 networks, haha
2022-12-31 13:52:49 <psykose> frag: like in the same session, between reboots/relogs, or
2022-12-31 13:53:22 <frag> psykose: same session, kinda weird, my bt keyboard doesnt have this problem
2022-12-31 13:53:58 <psykose> hsv2: there is lua-aports with buildrepo and the ap scripts, but by default buildrepo will build literally the whole repo and not a part of it
2022-12-31 13:54:07 <psykose> frag: hmm, sadly no idea then
2022-12-31 13:54:18 <psykose> i'd guess something resets it but i can't guess what
2022-12-31 13:54:18 <frag> okok
2022-12-31 14:39:47 <Ermine> psykose: afaik iwd is configured through its dbus interface
2022-12-31 14:40:53 <psykose> you can also create the psk files by hand
2022-12-31 14:41:04 <psykose> and run a separate dhcp client, etc
2022-12-31 15:30:11 <WhyNotHugo_> I'm running steam via `flatpak --user`. When I try and run a game, it fails with: error: While opening repository /var/lib/flatpak/repo: opening repo: opendir(objects): No such file or directory
2022-12-31 15:30:36 <WhyNotHugo_> Steam itself runs fine with `flatpak --user`, I'm not sure why it tries to use the system location when launching games, any ideas?
2022-12-31 15:35:46 <psykose> hmm
2022-12-31 15:35:52 <psykose> works for me if i rm -rf the folder still
2022-12-31 15:36:19 <psykose> i guess i'd try the repair --system just to create that folder even though it's not actually used
2022-12-31 15:46:32 <scorpion2185[m]1> is hibernation enable dby default in the kernel ?linux-lts https://wiki.alpinelinux.org/wiki/Custom_Kernel#Hibernation_to_prevent_data_loss
2022-12-31 15:59:51 <WhyNotHugo_> repair for system fails -- the whole /var/lib/flatpak/ is empty.
2022-12-31 16:00:03 <WhyNotHugo_> Looks like steam tries to run games in a separate container using the system flatpak.
2022-12-31 16:06:06 <psykose> delete the folder and do it :p
2022-12-31 16:06:25 <psykose> scorpion2185[m]1: yeah it is
2022-12-31 17:05:01 <scorpion2185[m]1> echo disk > /sys/power/state
2022-12-31 17:05:01 <scorpion2185[m]1> ash: write error: No such device
2022-12-31 17:05:14 <scorpion2185[m]1> it start suspending and then that error
2022-12-31 19:06:32 <jh81> almost that time of the year! Happy new year everyone, be it near or far!
2022-12-31 21:23:13 <paperowl> scorpion2185[m]1 are you building a custom kernel usj g
2022-12-31 21:23:17 <paperowl> Using tbe w
2022-12-31 21:23:46 <paperowl> Wiki or just using it as a reference. Sorry typos
2022-12-31 21:33:45 <sarpnt> i'm running alpine linux on a single board computer, it mounts mmcblk0p1 at /media/mmcblk0p1 and all the lbu backups are on there
2022-12-31 21:33:54 <sarpnt> but the permissions are wrong
2022-12-31 21:34:06 <sarpnt> any user can read the backups which means anyone can read any files
2022-12-31 21:34:32 <sarpnt> i can't figure out any way to change the permissions (none of the umask, fmask, dmask settings work)
2022-12-31 21:34:43 <grawity> what filesystem is used on mmcblk0p1?
2022-12-31 21:34:46 <sarpnt> vfat
2022-12-31 21:35:17 <grawity> mounting with -o umask=077 should work -- does the option get applied at all? i.e. does it show up in `mount` output afterwards?
2022-12-31 21:36:01 <sarpnt> i can't unmount /media/mmcblk0p1 after booting, it says it's busy
2022-12-31 21:36:08 <sarpnt> i have it in fstab and nothing happens
2022-12-31 21:36:44 <sarpnt> nothing in dmesg about it except later when i was trying to mess around with it
2022-12-31 21:36:49 <sarpnt> so long after boot
2022-12-31 21:39:44 <ikke> sarpnt: you need to stop the modloop service
2022-12-31 21:39:50 <ikke> then you can unmount it
2022-12-31 21:40:19 <sarpnt> what does modloop do?
2022-12-31 21:41:06 <sarpnt> ok after remounting umask did work
2022-12-31 21:41:13 <sarpnt> but it's not working in fstab
2022-12-31 21:41:39 <ikke> modloop contains the firmware and the kernel modules so that they do not need to be loaded in ram
2022-12-31 21:57:47 <sarpnt> /dev/mmcblk0p1 /media/mmcblk0p1 vfat rw,relatime,umask=077,errors=remount-ro 0 1
2022-12-31 21:57:55 <sarpnt> thats the line in my fstab for it
2022-12-31 21:58:06 <sarpnt> is there anything wrong with it
2022-12-31 23:13:01 <ptrc> happy new year, utc+1 alpines :)
2022-12-31 23:13:19 <Solarbaby> you too ptrc
2022-12-31 23:14:42 <psykose> alpiners
2022-12-31 23:15:04 <zv> alpinists?
2022-12-31 23:20:23 <elly> is there a way to donate money towards alpine?
2022-12-31 23:20:47 <psykose> nope
2022-12-31 23:23:00 <Solarbaby> i wish I could do more with alpine linux on my iphone
2022-12-31 23:23:14 <Solarbaby> I gotta check out all I can do with python.. there is probably a lot more then I know
2022-12-31 23:23:55 <ptrc> there's https://ish.app/ fwiw
2022-12-31 23:24:20 <Solarbaby> that's the app I downloaded thank you