2023-07-01 06:09:26 If it is a system installation, it's quite easy to install Alpine to one partition while ignoring other ones. 2023-07-01 06:09:26 What I miss is f2fs as an option for root. It's an interesting choice for flash-based storage devices. 2023-07-01 06:15:37 OdaNobunaga[m]: you can do f2fs if you want. 2023-07-01 06:25:23 Really? The installer said it's not supported. What works is to add f2fs to mkinitfs, then to copy the content of the root partition to another place, then to change the root partition to f2fs with the same UUID as the old one and finally, to copy the root content back to the new f2fs partition. 2023-07-01 06:26:16 And grub.cfg needs three changes from xfs or ext4 to f2fs. 2023-07-01 06:47:37 OdaNobunaga[m]: you can install alpine in a different way 2023-07-01 06:48:22 OdaNobunaga[m]: you can start off with minirootfs 2023-07-01 06:48:48 like this one: https://dl-cdn.alpinelinux.org/alpine/latest-stable/releases/x86_64/alpine-minirootfs-3.18.2-x86_64.tar.gz 2023-07-01 06:52:30 unpack it to your mountpoint, chroot into it and `apk add alpine-base what_ever_you_need` 2023-07-01 06:53:08 Good to know. Thanks! 2023-07-01 06:53:13 and rc-update add SERVICE RUNLEVEL 2023-07-01 06:53:37 list your currently enabled service, note them, and add them in chroot 2023-07-01 06:54:11 and fix /etc/mkinitfs/ , fix fstab, fix grub ... 2023-07-01 06:54:42 and of course setup your network, sshd... 2023-07-01 06:54:53 the whole step does not need setup-alpine. 2023-07-01 06:55:24 this is the most flexible method to install alpine. you decide partition layout, fs type... 2023-07-01 06:56:50 and of course you can do LUKS... 2023-07-01 06:57:04 basically what ever you want do... just in a manual way. 2023-07-01 07:00:23 OdaNobunaga[m]: i wrote a script to automate the whole thing. it takes about 3mins to install to a slow usb drive on a slow arm64 SoC 2023-07-01 07:00:52 if on x86_64 SSD, i guess it takes just 60s... 2023-07-01 09:36:08 If I run setup-alpine with encryption can I after the install boot into live gparted, decrypt the drive with cryptsetup and shrink the partition with gparted? 2023-07-01 09:41:51 how can I write an openrc script from systemd config like this: https://writefreely.org/start#:~:text=/etc/systemd/system/writefreely.service 👀 2023-07-01 09:46:25 By reading openrc doc I guess 2023-07-01 09:46:50 This seems like a reasonably easy service to write 2023-07-01 09:47:40 qaqland[m]: you need supervised daemon 2023-07-01 09:47:56 no you don't 2023-07-01 09:49:10 qaqland[m]: https://0x0.st/H1sG.txt example of caddy 2023-07-01 09:50:23 psykose: you see it is long running daemon, they have a Type=simple and Restart=always 2023-07-01 09:50:45 if you don't use supervised-daemon, then it gets killed or exited due to OOM... 2023-07-01 09:50:50 no auto restart 2023-07-01 09:51:41 qaqland[m]: https://img.ayaya.dev/Rh19FP0RviIM 2023-07-01 09:52:52 that's just not right. or say, not the best. 2023-07-01 10:08:43 thank u! I will try 🤔 2023-07-01 10:34:00 after my fiasco with @minimal yesterday, due to a silent migraine (triggered by irssis blue bars) and basic fatigue, in the context of hwclock not being necessary if the CMOS rtc driver is built into the kernel, and instead the dummy resp. stub function sysclock should be used when e.g. chronyd is running, i think it would be a good idea to polish setup-ntp in this respect, or at least make a 2023-07-01 10:34:06 dedicated entry in alpine's wiki. after i found this here today: https://0x0.st/H1s4.txt, i now understand what was meant yesterday. to save work on setup-ntp, a wiki entry would be nice. if there is no time for it, because there are more important things to do, then no problem. this is just a suggestion, no "i want this to happen". thx. 2023-07-01 10:41:25 on alpine, which window-manager is less resource-intensive dwm (wayland), sway, or river? 2023-07-01 10:48:13 any good alternatives to samba? I am way too dumb to configure it 2023-07-01 10:53:50 @powers i used dwm (x) on void x-years ago for testing purposes, now since 3 years sway on alpine. i see no difference on my old machine reg. mem/cpu/spikes. don't know about river 2023-07-01 11:01:34 because x is deprecated anyway, even of some linux-folks on reddit and youtube don't believe this, i guess i can try sway and check everything for yourself. building dwm stuff instead of using simple configs is at least very ressource-intensive 2023-07-01 11:08:09 meant:"you can try". "suckless" has other "problems" too. as someone who lives near ger, i know the meaning of some photos of some suckless-devs. imho: good software comes from good people. but thats another topic 2023-07-01 11:15:00 oh give me a break 2023-07-01 11:23:40 break from what 2023-07-01 11:25:37 a \ can exceed 79 characters.. 2023-07-01 11:36:18 What's the location for openrc.conf? How can I disable wpa_supplicant permanently on boot? 2023-07-01 11:39:30 rc-update -a del wpa_supplicant 2023-07-01 11:40:43 Hi. I am having an odd issue with ufw when I have ipv6 disabled. It still tries to run ip6tables-restore, which fails 2023-07-01 11:40:43 Thanks 2023-07-01 11:42:22 what is "ipv6 disabled" 2023-07-01 11:43:34 i.e. /etc/default/ufw IPV6=no? 2023-07-01 11:43:38 yes 2023-07-01 11:43:46 and not loading ipv6 kernel module 2023-07-01 11:43:58 are you sure it's ufw calling it at all 2023-07-01 11:44:59 everything in ufw conditionalises it on that 2023-07-01 11:45:11 https://paste.tnonline.net/files/rFuKWRR0SUt0_Screenshotfrom2023-07-0113-44-39.png 2023-07-01 11:46:48 rc.log shows: https://dpaste.com/36RQVXV2K 2023-07-01 11:55:37 I'm trying to compile dwl from. It prints an error message: "/bin/sh/: server-header: not found". Google says nothing. What is this? 2023-07-01 11:59:11 Ahh Wayland not running 2023-07-01 12:00:44 Thin the issue is in "ufw-init-functions" line 142 "if ip6tables -L INPUT -n >/dev/null 2>&1; then" which always is true, even if ipv6 is not enabled in the kernel. 2023-07-01 12:03:46 These lines: https://dpaste.com/CMFMATD29 2023-07-01 12:06:56 why would that be always true? 2023-07-01 12:07:17 it exits nonzero 2023-07-01 12:07:47 or well, perhaps should 2023-07-01 12:07:48 idk 2023-07-01 12:09:19 "ip6tables && echo true" echoes true, so even if ipv6.ko isn't loaded, ip6tables is not complaining. So the test does not work as intended 2023-07-01 12:10:28 yeah, upstream issue 2023-07-01 12:10:31 idk which though 2023-07-01 12:10:39 if iptables documents this is fine then... 2023-07-01 12:11:57 https://dpaste.com/7TDUULHUT 2023-07-01 12:12:14 Apparently, if I'm going to use xorg later, selecting mdevd as the device manager during installation does not make sense. 2023-07-01 12:12:15 yea. ufw or iptables? :D 2023-07-01 12:13:18 ufw is on launchpad and I dont have an account there to file a bug report 2023-07-01 12:39:51 When I run `setup-xorg-base`, I get 2 errors from rc-update: 2023-07-01 12:40:15 service 'hwdrivers' is not in runlevel 'sysinit' 2023-07-01 12:40:31 service 'mdev' is not in runlevel 'sysinit' 2023-07-01 12:40:57 Now, I wonder why it complains about 'mdev' because it just set up udev instead of that. 2023-07-01 12:47:02 those aren't errors 2023-07-01 12:48:15 Okay, so are these mere warnings? 2023-07-01 12:48:54 I mean there's a red insetad of the usual green asterisk at the beginning of these lines. 2023-07-01 12:49:56 they mean literally nothing yes 2023-07-01 12:50:26 You probably mean virtually. ;) 2023-07-01 12:50:39 Well, I can actually start X now. Nice. 2023-07-01 12:51:17 Only need to copy over my Openbox configuration. 2023-07-01 13:01:36 psykose: https://bugs.launchpad.net/ufw/+bug/2025532 2023-07-01 13:02:22 noice 2023-07-01 13:04:17 Hm.. seems launchpad didn't like formatting using backticks. Is it text-only or do they use some other markup? 2023-07-01 13:06:04 looks fine 2023-07-01 13:06:37 i just render it in my head 2023-07-01 13:07:34 something something origin of markdown 2023-07-01 13:13:51 0daNobunga[m]: unless things have changed, if you use f2fs for Alpine root then the problem is it will not be fsck'ed during boot as fsck.f2fs will not check a (readonly) mounted filesystem - the only way around this is to *force* a fsck (rather than just check if one is required) on every boot and that means a complete fsck (whether needed or not) will occur on every boot 2023-07-01 13:15:40 Yeah, if seen this error message. As all 6 Sata slots are busy, I need to run the system from USB. 2023-07-01 13:15:40 Maybe I should really buy an HBA card or a PCI-to-SATA adapter. Then I'll gladly take xfs. 2023-07-01 13:18:18 oleovit: the explanation you found in #11889 was written by me ;-) 2023-07-01 13:20:37 algitbot: rpi does not have hwclock 2023-07-01 13:35:04 if I do `apk add linux-lts` in the miniroot, it pulls in linux-firmware-*, can I somehow exclude the large & useless parts of that? (or at least stop the files from being extracted to the fs) 2023-07-01 13:36:01 ikke: :D 2023-07-01 13:36:51 I guess I assume markdown support these thays. What was really wrong with normal quotation marks and blocks using ### ~~~ etc.. ? =) 2023-07-01 13:37:04 How to add getopt library? Is it in the repos? (Can't find) 2023-07-01 13:40:12 grawity: do "apk add linux-lts linux-firmware-none" 2023-07-01 13:40:40 then you can add whichever specific firmware packages you want 2023-07-01 13:40:46 thanks 2023-07-01 13:41:46 goodbye netronome, now I have a 2023-07-01 13:41:49 ( •_•)>⌐■-■ 2023-07-01 13:41:50 minimal 2023-07-01 13:41:52 installation 2023-07-01 13:42:25 grawity: there's a lot more work you'd need to do to have a minimal installation ;-) 2023-07-01 13:42:46 yeah, like installing a bootloader before I forget again 2023-07-01 13:48:46 oh I need to add openrc too, I thought it was part of the miniroot 2023-07-01 13:49:34 nope 2023-07-01 13:51:17 is it a security risk to use an older alpine version(3.17)? why isnt the default repo "lastet-stable" ? 2023-07-01 13:52:29 grawity: nope, the minirootfs is targeted towards containers which don't have kernels or bootloaders and often/typically don't have an init 2023-07-01 13:53:07 frag: default repo where? 2023-07-01 13:53:26 minimal: /etc/apk/repositories 2023-07-01 13:54:17 frag: eh? that file contains the version of Alpine that was installed 2023-07-01 13:54:52 frag: main on 3.17 is still supported 2023-07-01 13:54:59 if you install 3.17.x then that will contain 3.17, if you install 3.18.x then that will contain 3.18 2023-07-01 13:55:09 as that is the version you have installed 2023-07-01 13:55:57 minimal: id assume most just install the latest version, maybe im wrong .. 2023-07-01 13:56:10 sadly not 2023-07-01 13:56:26 psykose: hmm do you know why not? 2023-07-01 13:57:13 ask the users :D 2023-07-01 13:57:18 frag: and so? if they install 3.18.x and 3.18.x is the latest version then with /etc/apk/repositories containing "3.18" then what's the problem? 2023-07-01 13:57:28 sometimes someone is describing some bug and i go "hmm.. seems familiar.." and then it turns out it was 3.13 2023-07-01 14:00:01 frag: if /etc/apk/repositories was to contain "latest-stable" then that would cause problems, i.e. you install 3.18.x today and repositories contains "latest-stable" and then tomorrow 3.19 is released and latest-stable is changed to point to 3.19 rather than 3.18, then an end-user wouldn't know that the needed to run "apk -Ua upgrade" rather than "apk upgrade" as the latest-stable version had changed 2023-07-01 14:01:43 minimal: ahh.. hmm 2023-07-01 14:07:53 And possibly other breaking changes to account for 2023-07-01 14:36:35 What tool you use fot usb automount? 2023-07-01 15:43:14 malmi70[m]: inotify with mount 2023-07-01 16:13:44 What's a decent system monitoring bar for graphical environments on Alpine? 2023-07-01 16:13:58 I use xmobar on my other system, but it's not available. 2023-07-01 16:14:30 Which is understandable given that it would probably need Haskell. 2023-07-01 16:14:50 msiism: the old solid conky 2023-07-01 16:15:31 Hm… I might give that a try. 2023-07-01 16:17:04 i have btop in a terminal on workspace 10 2023-07-01 16:17:13 still has a bug i haven't reported in a year, maybe one day i'll fix it 2023-07-01 16:17:14 :D 2023-07-01 16:18:15 msiism: there is bottom too 2023-07-01 16:18:23 and there is lm-sensors... 2023-07-01 16:23:13 tmon 2023-07-01 18:08:07 I have burned extended Alpine installer .iso to a USB & reading wiki install guide about the ANSWER_FILE... but where do I put that file? And how? 2023-07-01 18:10:29 setup-alpine offers an option to create an answer file. 2023-07-01 18:10:34 The iso is isofs, which is read-only 2023-07-01 18:10:39 I usally toss the answerfile on a webserver and wget it 2023-07-01 18:10:39 so you cannot ad the answer file there 2023-07-01 18:11:44 you'll need to setup networking/dhcp first before trying to download it 2023-07-01 18:13:01 ah so I could pull it in with wget from another server before I run setup-alpine 2023-07-01 18:13:44 if you use setup-bootable, you'll get a bootable usb disk that you can write to 2023-07-01 18:13:50 yeah. use setup-interfaces and run udhcpc eth0 2023-07-01 18:13:51 but I have to run setup-alpine to set up dhcp? 2023-07-01 18:13:59 ah setup-interfaces k 2023-07-01 18:15:06 setup-alpine should run setup-interfaces, actually. 2023-07-01 18:15:23 once you have the answerfile: setup-alpine -f answerfile 2023-07-01 18:15:24 Okay, I also misunderstood that question. 2023-07-01 18:15:27 msiism: but that defeats the purpose of an answerfile if you run setup-alpine already 2023-07-01 18:15:44 yeah that was my first paradox of 2 looking at this 2023-07-01 18:16:18 but you can manually run setup-interfaces, or just use the `ip` command to give yourself a network connection 2023-07-01 18:17:28 so on my 1st old physical lab server of 3, I could use alpine extended install, run setup-interfaces or ip to get dhcp, grab my answer file of choice, then run full setup-alpine -f 2023-07-01 18:17:39 yes, that would work 2023-07-01 18:19:01 then I could run setup-bootable from an alpine machine to make a 2nd USB? That'd have local writable area for answer files say for lab machines 2 & 3 ? 2023-07-01 18:19:44 ...and thank you all, by the way! 2023-07-01 18:20:02 Oneliner: setup-interfaces; udhcpc eth0; wget http://webserver/answerfile; setup-alpine -f answerfile 2023-07-01 18:20:54 change where needed 2023-07-01 18:21:31 awesome trying now 2023-07-01 18:22:15 it will prompt for root pass unless you add -e to setup-alpine 2023-07-01 18:22:49 if you use setup-interfaces, you don't need to manually run udhcpc 2023-07-01 18:22:53 once rebooted, I'd change the root password since it's blank if you use -e 2023-07-01 18:23:27 I for some reason have to run udhpc 2023-07-01 18:24:04 you might want to run setup-interfaces -r 2023-07-01 18:24:08 not infront of my system at work that I use this on 2023-07-01 18:24:13 to test 2023-07-01 18:24:24 with -r, it will restart the networking service 2023-07-01 18:24:32 ah cool 2023-07-01 18:24:48 but, if you just need dhcp, it should suffice to do: ip l set dev eth0 up; udhcp eth0 2023-07-01 18:24:53 udhcpc eth0* 2023-07-01 18:25:15 side question: I get alpine is great for a lighweight VM, and also for a minimalist distro on bare metal, but is there any reason I shouldn't use it as a VM host OS, say with libvirt/qemu/kvm guests also running alpine? 2023-07-01 18:25:29 ku_: I don't see why not 2023-07-01 18:25:48 ku_: we run all our infra on alpinelinuix 2023-07-01 18:25:50 alpinelinux 2023-07-01 18:26:03 bare metal, lxc containers, docker containers, virtual machines 2023-07-01 18:38:02 ikke: (and others) do you use edge and have rolling updates or decide when to move from say 3.17 to 3.18? 2023-07-01 18:40:44 depends on use case 2023-07-01 18:41:03 ku_: most servers are on some release and we upgrade them at some moment 2023-07-01 18:41:31 ah so case by case when it makes sense for each one 2023-07-01 18:41:38 yeah 2023-07-01 18:41:49 usually start with servers that are less critical 2023-07-01 18:41:56 and then step by step upgrade more critical servers 2023-07-01 18:42:21 My 2nd paradox is re: partitioning. This server used to have 2x300GB drives with a small mdadm RAID1 across sda1 & sdb1 and a larger mdadm RAID1 across sda2/sdb2 say 2023-07-01 18:42:53 but I'd like to try btrfs for data drives and I've added 2 more physical 300GB drives 2023-07-01 18:44:09 Should I stick to ext4 or something for the OS install? Should it be mdadm RAID across a small partition on each of the 4 drives? And them point btrfs at the bigger partitions of the 4 drives? 2023-07-01 18:45:10 Trying not to drift into btrfs questions, and stick to alpine install questions... a lot has changed since I last had my hands on these servers. :) 2023-07-01 18:46:27 (I am re-reading the Partitioning Your Disk guide now.) 2023-07-01 18:48:17 I recently did a test install with btrfs and there was no problem. 2023-07-01 18:48:44 The setup was very simple, though: 3 MBR partitions on one HDD. 2023-07-01 18:49:58 Still reading but how would you size partitions with 4 300GB drives (only 4GB RAM on this old box). 2023-07-01 18:51:51 Depends on what those partitions are. 2023-07-01 18:52:43 I'd probably give swap 4G. Then, I usually use 15G for /. And then the rest goes to /home. 2023-07-01 18:53:02 Given so much space, though, I might be inclined to have, say, /opt on a separate partition. 2023-07-01 18:53:24 do you have so much on /opt? 2023-07-01 18:53:25 I'd probably do a 4-disk btrfs array for / (with /boot being maybe a 2-disk btrfs, GRUB can deal with that), though OTOH btrfs isn't great on HDDs 2023-07-01 18:53:50 (never really saw the point of isolating / from the rest, though) 2023-07-01 18:54:11 ikke: Good question. Probably not. Maybe, I'd rather separate /boot then. 2023-07-01 18:54:13 I have separate partitions for /home, /var/lib/docker and some other 2023-07-01 18:54:55 does anyone have that link handy with the script that sets up default openrc runlevels as part of a normal installation? 2023-07-01 18:55:28 I usually have, like, one 300G HDD to put my stuff on. 2023-07-01 18:55:42 And then some special-purpose external drives. 2023-07-01 18:55:57 ACTION currently installing alpine on a 18G SCSI HDD 2023-07-01 18:56:44 My use case for this 1st box is throw it in a colo for extended family to photo share so would like to maximize storage area -- will also have larger externals for less used files and backups. 2023-07-01 18:57:35 grawity: do you mean this? https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-alpine.in#L188 2023-07-01 18:57:37 ...and learn Alpine in the process of course 2023-07-01 18:59:11 ikke: yeah, but iirc there was a slightly longer list? I'm now looking at the usb stick I booted from, looks like it has stuff like hostname, devfs, mount-ro 2023-07-01 19:00:57 grawity: https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/initramfs-init.in#L675 2023-07-01 19:01:09 ah, that's the one 2023-07-01 19:02:39 if you use LVM then its handy to have /home and other things as LVs which you can grow later 2023-07-01 19:02:59 yeah, but I recall btrfs can do that as well 2023-07-01 19:03:06 with volumes 2023-07-01 19:03:33 btrfs is slightly different, subvolumes have no space reservation at all, it's all one big fs 2023-07-01 19:03:48 are there quotas? 2023-07-01 19:04:32 technically yes but not per-subvolume quotas, iirc, and also they hurt performance 2023-07-01 19:05:11 I use lvm volumes to keep the size of certain directories in check 2023-07-01 19:09:21 Has anyone tried building/packaging OnlyOffice on Alpine? 2023-07-01 19:09:29 I was thinking of LVM versus "typical" partitioning - make the whole disk a PV but only initially size logs, home LVs etc as required and then grow them later as needed whereas with partitioning if your home or logs partitions are not big enough it gets messy resizing various partitions 2023-07-01 19:10:28 is onlyoffice that pile of potatoes with no build system that might as well just be proprietary 2023-07-01 19:10:50 got it done minimal, tried posting all of it up in usetalk no no 2023-07-01 19:12:18 psykose: I haven't even looked at the build system. 2023-07-01 19:15:56 It's there now, must have been a link. 2023-07-01 19:16:21 Thank you all... have to sign off here now but will surely be back with more questions soon enough! 2023-07-01 20:37:30 how do i check if wlan0 is present? (it can be toggled on/off on pinebook pro) 2023-07-01 20:38:16 ip a? 2023-07-01 20:39:24 also rfkill 2023-07-01 20:41:00 hmm, seems unaffected 2023-07-01 22:50:05 hi 2023-07-02 02:54:54 hey guys I ve got a 32 bit netbook do you guys think alpine would work well , I just want tu setup some bots for telegram 2023-07-02 02:56:20 Depends on the packages you add to it, but generally I'd have to go with "yes" 2023-07-02 03:31:57 <[[sroracle]]> I cannot for the life of me figure out what my problem is, but I have a data-install that refuses to use cached APKs when booting. 2023-07-02 03:33:44 <[[sroracle]]> It seems like it's not mounting the disk with the APK cache until the boot is completely over, and nothing I do can convince it to mount it earlier. 2023-07-02 06:08:12 setup-intefaces should be able to connect to a hidden ssid right? 2023-07-02 06:31:36 frag: it will scan for ssids (iw dev wlan0 scan), and you have to pick from what is returned there 2023-07-02 06:31:50 so if `iw dev wlan0 scan` does not list it, it's not possible to select 2023-07-02 11:10:24 Hi there, I was following the wiki on how to install Alpine on a Raspberry Pi (https://wiki.alpinelinux.org/wiki/Raspberry_Pi). I want to create my own overlays that will automatically connect over wifi and setup the software I'd like to run, headlessly. 2023-07-02 11:10:51 It seems like the base 3.18 install comes with wpa_supplicant in the `apks` directory that comes with the installation. Is it possible to somehow add iwd to this repository before actually booting the system? 2023-07-02 11:13:32 I did a naive atempt of adding the apk file downloaded from dl-cdn.alpinelinux.org/ (and dependencies) to the `apks` directory, but then realized there's a signed package index. I was also messing around with running alpine from docker and mounting those directories in various places, but of course the RasPi runs ARM so I wouldn't be able to download the right apk into the repository anyhow. 2023-07-02 11:15:27 As the installation files must have been created somehow, I'm assuming there's tooling that would allow me to add a package to the repository? 2023-07-02 12:38:49 Is Docker's alpine:latest image supposed to not have openrc? 2023-07-02 14:00:36 Is it known issue that audio-related XF86-keys doesn't work in any layout? 2023-07-02 14:17:58 Ermine: why would you run OpenRC in a docker container rather than (a) no init, (b) use tini, or (c) something designed for containers like s6-overlay? 2023-07-02 14:57:58 minimal: I'm newbie regarding docker. I've logged into a container and was surprised that rc-* commands were missing. But I see your point 2023-07-02 14:59:48 -> Ermine what are you running your docker on? 2023-07-02 15:00:38 <- a docker newbie also, setting up a HA server 2023-07-02 15:01:08 I'm running it on Fedora 2023-07-02 15:02:36 OK, I haven't played with Fedora, I'm running alpine on RPI4, LVM over USB RAID1 2023-07-02 15:04:19 I believe docker works fine on Alpine 2023-07-02 15:04:37 Ermine: docker is intended for running "application containers", unlike the likes of LXC which are intended for running "system containers" (including typical init services) 2023-07-02 15:05:59 I'm stress testing my drives for a NAS and hope to store most of the docker apps there. 2023-07-02 15:06:03 Well, docker container is an option suggested by gns3 2023-07-02 15:08:04 -< retired geek, got out of IT/Helpdesk 6 or 7 years ago, back on big learning curve. 2023-07-02 16:08:43 I'm using docker-compose on some VM for services that are too lazy to provide a proper build/installation procedure... Anyway docker works fine on them 2023-07-02 18:07:27 Hi, Im testing on "Alpine Linux" on AWS cloud. Was folloing advise from this page: https://alpinelinux.org/cloud/ (it sortof worked, its running. BUT, I cant connect to i with AWS web thingys, only ssh) 2023-07-02 18:09:05 What todo? do I havto run "setup-alpine", do something with the included cloud-package or what?? 2023-07-02 18:09:35 ncrowl: what is "AWS web thingys"? 2023-07-02 18:11:05 minimal: page title "Connect to instanceInfo", "Connect to your instance i-01e5efda834c2d7c5 (Alpine) using any of these options.. dono, some random forms 2023-07-02 18:12:51 "EC2 Instance Connect" or "AWS Systems Manager Session" or ssh? 2023-07-02 18:14:25 minimal: yes, "EC2 Instance Connect" (ssh is working, for user "alpine" not for root, I have done noo changes 2023-07-02 18:15:14 I believe that the necessary software/config for neither "EC2 Instance Connect" nor "AWS Systems Manager Session" is present in the Alpine cloud images, you'll have to use ssh 2023-07-02 18:17:09 oki minimal. thanks. Anny suggestion on any other cloud -provider to use? (dono if I wana host on any FANG company, are using Digitalocean also)? 2023-07-02 18:17:58 ncrowl: the "official" Alpine Cloud images only presently support AWS/Azure/Google/Oracle 2023-07-02 18:18:20 I have my own script for creating Alpine cloud images for multiple cloud providers 2023-07-02 18:19:57 minimal: oki, u do? I seen old youtube videos, where folks used Digitalocean 2023-07-02 18:21:03 I like the small footprint, is this in line with any *suckless* ideers? 2023-07-02 18:22:21 ncrowl: I DM'ed you a link 2023-07-02 18:31:09 ntpd cant force timezone from commandline? (postmarketos doesnt seem to have much in /etc/zoneinfo 2023-07-02 18:32:03 i don't think it cares about the timezone at all 2023-07-02 18:32:09 ntp does not care about timezones 2023-07-02 18:32:53 what does it care about? it sets the wrong time for me 2023-07-02 18:33:08 it doesn't set the wrong time 2023-07-02 18:33:18 your timezone after the fact is wrong 2023-07-02 18:33:22 not what ntp sets 2023-07-02 18:33:30 does too! 2023-07-02 18:33:33 frag, please show us 2023-07-02 18:33:53 timezone after the fact? 2023-07-02 18:33:55 frag: NTP is based on UTC 2023-07-02 18:34:05 frag, show us the thing that is wrong 2023-07-02 18:34:06 Habbie: how? 2023-07-02 18:34:11 date ; TZ=UTC date 2023-07-02 18:34:14 is a good start 2023-07-02 18:35:06 doas ntpd -n -q -p pool.ntp.org;date # Sun Jul 2 18:34:41 GMT 2023 2023-07-02 18:35:14 that's correct 2023-07-02 18:35:14 its more like 20:34 2023-07-02 18:35:21 no 2023-07-02 18:35:26 in GMT, it's 18:35 2023-07-02 18:35:29 in CEST, it's 20:35 2023-07-02 18:35:38 so the output is correct, but it's not what you want 2023-07-02 18:35:47 you just don't have your timezone set for after 2023-07-02 18:35:48 frag: what timezone are you in? then run setup-timezone as set it to your timezone 2023-07-02 18:36:07 ln -sfv ../usr/share/zoneinfo/Europe/Amsterdam /etc/localtime 2023-07-02 18:36:11 or whatever your location is 2023-07-02 18:36:24 i was going to suggest Amsterdam too, it would at least be correct for frag today 2023-07-02 18:36:26 likely all year 2023-07-02 18:37:38 i think i only have /etc/zoneinfo ... and that dir har one file: GMT 2023-07-02 18:37:46 etc/zoneinfo is not really a thing 2023-07-02 18:37:47 this is postmarketos btw!!!!!!! 2023-07-02 18:37:50 just delete it and forget about it 2023-07-02 18:38:04 then run the ln -sfv i posted 2023-07-02 18:38:04 do you have /usr/share/zoneinfo? 2023-07-02 18:38:05 psykose: huh?????????? 2023-07-02 18:38:22 psykose: from which dir? 2023-07-02 18:38:24 bork the win? 2023-07-02 18:38:29 anywhere? 2023-07-02 18:38:32 i gave you the entire command 2023-07-02 18:39:27 But maybe missing an explanation 2023-07-02 18:40:06 well it still 18:39 2023-07-02 18:40:06 frag, basically time reference should be fixed (ie unshifted time sync), then you apply to your local computer your local time (the zoning difference) 2023-07-02 18:40:10 apk add tzdata 2023-07-02 18:40:28 There are zone files doing that 2023-07-02 18:40:28 then post the output of `ls -l /etc/localtime` and `date` 2023-07-02 18:40:35 You need to tell your system which zone you *want* to use 2023-07-02 18:40:56 Usually that is done by using file /etc/localtime 2023-07-02 18:41:18 It can be a copy of a zone file, but better is just to link it to the one provided by the system and that you want 2023-07-02 18:41:28 They are located in /usr/share/zoneinfo 2023-07-02 18:41:59 /etc/localtime -> ../usr/share/zoneinfo/Europe/Amsterdam 2023-07-02 18:42:04 Now you don't need to copy paste commands and say it doesn't work, you should have enough keys to solve it yourself, but of course help is still available 2023-07-02 18:42:51 still 18:42 tho 2023-07-02 18:43:25 hmm it uses UTC 2023-07-02 18:43:36 can you try: TZ=Europe/Amsterdam date 2023-07-02 18:43:56 you didn't show what date outputs 2023-07-02 18:43:58 Habbie: ahh there we go 2023-07-02 18:44:18 psykose: here u go buddy: Sun Jul 2 18:44:05 UTC 2023 2023-07-02 18:44:25 it said GMT before, that's interesting 2023-07-02 18:44:46 does /usr/share/zoneinfo/Europe/Amsterdam even exist 2023-07-02 18:44:47 (18:xx is correct for both UTC and GMT, to be clear) 2023-07-02 18:45:15 18:38Z do you have /usr/share/zoneinfo? 2023-07-02 18:45:20 this will go better if you answer all questions :) 2023-07-02 18:45:31 psykose: yes, after i installed that package 2023-07-02 18:45:54 Habbie: yes that dir too :] 2023-07-02 18:46:22 env | grep TZ 2023-07-02 18:46:25 Well, rather the content 2023-07-02 18:46:50 s 2023-07-02 18:47:15 But 2023-07-02 18:47:15 20:37:47 frag$ this is postmarketos btw!!!!!!! 2023-07-02 18:47:24 psykose: aha! i tried setting it to CEST earlier ... in a new term i get the result you expected :) 2023-07-02 18:47:27 problem solved 2023-07-02 18:47:27 So personally I don't know how they handle time zoning there, maybe they do some trickery 2023-07-02 18:47:34 ok 2023-07-02 18:48:13 fun fact, TZ=CET works but TZ=CEST does not 2023-07-02 18:48:41 (with tzdata installed. without it, neither work, and become UTC) 2023-07-02 18:48:56 Because CEST doesn't exist 2023-07-02 18:49:06 postmarketos very minimal i guess :S 2023-07-02 18:49:28 quinq, well, it's funny because the output of "TZ=CET date" contains CEST 2023-07-02 18:49:41 to muddy the waters even more, TZ=CEST does work on debian, and i thought those tzdata files were shared between OSes 2023-07-02 18:49:53 What does contains mean in this context? 2023-07-02 18:49:58 do you guys know what could cause this too phaps, when i cryptopen a sd card on my pinebook pro i get this in dmesg: [ 9560.277513] udevd[8040]: conflicting device node '/dev/mapper/lala' found, link to '/dev/dm-1' will not be created 2023-07-02 18:49:59 # TZ=CET date 2023-07-02 18:50:01 Sun Jul 2 20:49:10 CEST 2023 2023-07-02 18:50:14 Yeah, date is correct 2023-07-02 18:50:21 it is 2023-07-02 18:50:33 even when the only file in /dev/mapper before that was 'control' or similar 2023-07-02 18:50:48 this definitely is one reason to prefer Europe/Amsterdam over CET or CEST, in the config 2023-07-02 18:50:53 frag: yes, ignore that dmesg entry 2023-07-02 18:51:00 Habbie, sure 2023-07-02 18:51:29 Habbie: why? CET sounds a bit easier imo, more universal :> 2023-07-02 18:51:59 minimal: hmm ok, had some real problems earlier and thought this might be causing it ... but reformated the sd so ok will try 2023-07-02 18:52:12 (1) if, some day, the promised DST removal in Europe comes through, you will want to be on the right country's time (2) quinq said "CEST does not exist", but you could also argue that in summer, CET does not exist 2023-07-02 18:52:28 frag: will try what? I said to ignore it, it is "normal" 2023-07-02 18:52:49 Exactly 2023-07-02 18:53:16 (not so sure about the removel coming throught though :D) 2023-07-02 18:53:28 (oh me neither :( ) 2023-07-02 18:53:45 minimal: well i just cryptopened it before, now i tried mounting the mapper file and get : 'mount: /mnt/evo: can't read superblock on /dev/mapper/lala.' and more info in dmesg 2023-07-02 18:53:57 so its an unrelated problem? 2023-07-02 18:54:07 EU DST removal has been stuck in political infighting for several years 2023-07-02 18:54:21 minimal, i thought it just basically withered out during covid 2023-07-02 18:54:52 frag: do you want me to say yet again to ignore the particular dmesg entry? 2023-07-02 18:57:48 Habbie: well it needs individual EU countries to push it and some of them aren't interested in doing so 2023-07-02 18:57:59 right 2023-07-02 18:58:56 would cause "fun" for countries with land borders in different timezones, though I guess Portugal/Spain already have that issue 2023-07-02 18:59:27 bring back Railway Time I say! ;-) 2023-07-02 18:59:40 yes - and when DST goes, and we all get to pick our time zone, there's a big chance NL will end up on the wrong time because trade with DE is too important 2023-07-02 19:00:44 i vote to upgrade human brain processing instead 2023-07-02 19:00:50 so people are not confused by 1 hour differences 2023-07-02 19:00:58 i'd love to spend that upgrade on something else 2023-07-02 19:01:16 there's easier ways to get that gym bod 2023-07-02 19:05:22 Is it a good idea to remove those time-saving shifts that were made amongst other things to save energy consumption, in these times? 2023-07-02 19:08:06 should we stop saving energy in these times? no. Should we stop a practice that doesn't really save that much energy and does cause very real heart attacks? i vote yes 2023-07-02 19:08:19 of course, the whole savings from DST is because we use time inflexibly 2023-07-02 19:09:27 DST causing really big amount of heart attacks? 2023-07-02 19:09:45 Is there a study of that? 2023-07-02 19:09:46 it's a statistically significant relation, yes 2023-07-02 19:09:58 i could link you some but i would have just googled them 2023-07-02 19:10:13 (i did a longer read of a couple of them some years ago and it seemed real to me) 2023-07-02 19:11:29 Ah, right 2023-07-02 19:11:39 clearly already vulnerable people, from what i gather 2023-07-02 19:11:44 but there's plenty of those 2023-07-02 19:11:58 It's a factor in increasing risk to people already subject to heart issue that don't take care :/ 2023-07-02 19:12:03 yes 2023-07-02 19:12:09 But that's true to most things 2023-07-02 19:12:15 well, likely at least partially true 2023-07-02 19:12:22 some people have no control over what time they need to get up 2023-07-02 19:12:22 Better solution would be to actually take care of those people, explaining that maybe 2023-07-02 19:12:44 Yeah, that's a job-related issue 2023-07-02 19:13:11 Don't force your employees subject to heart attacks to take risks 2023-07-02 19:14:17 and their partners, and their children.. i'm not warning it's a slippery slope, i'm saying we should ride all the way down to the slope :) 2023-07-02 19:14:32 *down the slope 2023-07-02 19:14:55 Not sure what that means 2023-07-02 19:15:16 basically not "make exceptions for people with specific problems" but "do better in general, for everybody" 2023-07-02 19:15:22 we're getting quite offtopic, though 2023-07-02 19:15:25 do we really have time for this discussion? ;-) 2023-07-02 19:15:47 I do :) 2023-07-02 19:15:53 why would a supranational trading bloc made of sovereign governments already across different timezones need to all agree to end DST? that sounds like a parody 2023-07-02 19:16:03 i'm supposed to be refactoring code so i have all the time 2023-07-02 19:16:13 :D 2023-07-02 19:16:14 Habbie, I like also take care of people with specific problems 2023-07-02 19:16:22 good :) 2023-07-02 19:16:29 Instead of hiding behind other things (like daylight saving time) 2023-07-02 19:17:27 minimal: Countries with one timezone... The Canadian chuckles :-) we even have one with 30 minutes in the difference 2023-07-02 19:17:32 Like *preventing* heart diseases 2023-07-02 19:17:51 timlegge, must bring some fun ^^ 2023-07-02 19:18:06 timlegge: right in most larger countries like that it's up to the sub-national jurisdictions to determine whether they want to use DST or not 2023-07-02 19:18:07 Answering the phone “Hello, what time is it?” 2023-07-02 19:18:30 i have a customer - if they invite me to meetings, they show at xx:07 in my calendar 2023-07-02 19:18:33 twice a year every year its always a surprise... 2023-07-02 19:18:34 i haven't yet found out why 2023-07-02 19:18:38 haha 2023-07-02 19:18:51 yes, I think we have one that does not do DST 2023-07-02 19:19:05 there's about 6 million peoiple globally in my time zone. I never assume people know what time it is here :) 2023-07-02 19:19:17 LACampbell, that's a small time zone! 2023-07-02 19:19:27 timlegge: India has only one timezone, burt it is UTC+5:30 lol 2023-07-02 19:19:36 ^^ 2023-07-02 19:19:37 incidentally my :07 thing comes from India 2023-07-02 19:19:50 Habbie: yeah UTC+12. New Zealand I think a very sparsely populated section of Russias far east 2023-07-02 19:20:06 right 2023-07-02 19:20:28 https://en.wikipedia.org/wiki/UTC%2B12:00#/media/File:Timezones2008_UTC+12_gray.png 2023-07-02 19:20:32 yeah that's not a lot of land area 2023-07-02 19:20:53 and that area is not the most densely populated either, i imagine 2023-07-02 19:20:59 oh I forgot Fiji 2023-07-02 19:21:47 Russia has 11 timezones and I remember when working in mobile telecoms that a particular Russian operator had to bill based on the users' home timezone, not the one they were in, which was "fun" 2023-07-02 19:21:53 yeah 50,000 people in Russia share it lol 2023-07-02 19:22:00 minimal, wow 2023-07-02 19:22:14 minimal, honestly the other option sounds worse? 2023-07-02 19:22:18 That's a vast country 2023-07-02 19:22:58 Habbie: I'm just saying it complicated billing systems 2023-07-02 19:23:17 oh yes 2023-07-02 19:23:56 Computers shouldn't have too much trouble adding or substracting integers (or half ones) to a bill though 2023-07-02 19:24:17 yeah that 'small' slice of russia where 50,000 people live in my timezone is twice the size of germany lol 2023-07-02 19:24:25 quinq, it's finding the right integers that is hard! 2023-07-02 19:24:31 damn maps and their confusing distortions at the edges 2023-07-02 19:24:39 Admitidly it becomes a bit more complex with *moving* live data 2023-07-02 19:24:54 Habbie ^^ 2023-07-02 19:25:01 :) 2023-07-02 19:25:03 yes 2023-07-02 19:25:11 telco billing systems are always a "mess" 2023-07-02 19:25:37 Manager in a meeting “we should fix the billing system” 2023-07-02 19:25:42 fired next week 2023-07-02 19:26:01 built it 50 years ago. Should be fine 2023-07-02 19:26:21 I worked with a pre-paid mobile billing system in 201x that had some code going back to the 1970s lol 2023-07-02 19:26:40 nice :) 2023-07-02 19:26:46 In Fortran? 2023-07-02 19:26:46 basically had layers of code added each decade 2023-07-02 19:27:04 with a HTML interface that looked like it was written in the mid 1990s 2023-07-02 19:27:43 it would take several hours to parse a 100-line XML doc lol 2023-07-02 19:27:53 oh-my 2023-07-02 19:28:57 minimal, but that's because you were moving it accross daylight shaving hour 2023-07-02 19:29:00 s 2023-07-02 19:29:25 quinq: nope, this wasn't in a multi-time zone country lol 2023-07-02 19:31:13 there user to be a company that sold an html solution to scrape mainframe screens and make it a web page. I imagine it is still in use somewhere 2023-07-02 19:31:23 s/user/used/ 2023-07-02 19:32:38 timlegge: yeah, probably owned now by Micro Focus....where old "enterprise" software goes to die ;-) 2023-07-02 20:37:34 did we have now some script to generate apkolv to allow remote installation of linux via ssh? 2023-07-02 20:41:14 there's https://github.com/macmpi/alpine-linux-headless-bootstrap 2023-07-02 20:46:14 ok that is still the only thingie 2023-07-02 20:48:16 not having usb keyboard around and 2 pi and one pc server that would need installing is quite hazzle... just moved and forgot my usb keyborads 120km away =D 2023-07-02 23:46:40 Hi folks, odd question for you. I'm on a fresh alpine install and i'm trying to add kernel params for apparmor. But weirdly enough, there's no /boot directory, so I can't find the config I need to change. WTF? any hunches anyone has would be appreciated 2023-07-02 23:47:21 in / I have bin dev etc home lib media mnt opt proc root run sbin srv sys tmp usr var 2023-07-02 23:50:51 wwwww: which bootloader are you using? 2023-07-02 23:51:33 minimal: the setup script didn't give me a choice if I recall, so whatever the default is 2023-07-02 23:52:26 what device did you install on? 2023-07-02 23:53:08 raspberry pi 2023-07-02 23:53:18 aarch64 2023-07-02 23:54:17 it has a boot partition, the 1st one on the SDcard 2023-07-02 23:54:55 minimal: oh, do i need to mount it? I can't edit kernel params over ssh? 2023-07-02 23:56:09 over ssh? what has ssh got to do with it? 2023-07-02 23:56:28 mount the 1st (FAT) partition and edit the cmdline.txt file, then unmount it 2023-07-02 23:56:43 minimal: cause I'm ssh'd in and there's no /boot dir 2023-07-02 23:56:56 so I assume you're saying i need to mount the sdcard on another machine instead 2023-07-02 23:59:52 wwww: a /boot directory is not required 2023-07-03 00:00:05 no, I said to mount that partition on the RPI 2023-07-03 00:00:19 minimal ah ok 2023-07-03 00:00:37 mount it *anywhere*, /boot is not used on a RPI 2023-07-03 00:01:19 well, more precisely, /boot is not used in order to boot a RPI 2023-07-03 00:02:14 it may however need to be mounted there in order for kernel package updates and initramfs updated to be applied. Assuming you've installed Alpine in Sys mode ("run from SD") rather than run-from-ram mode 2023-07-03 00:03:56 lsblk shows only one partition ;-( 2023-07-03 00:04:15 ive installed alpine in diskless mode I believe. I was never prompted by the setup to choose @minimal 2023-07-03 00:05:04 wwwww: right, I'm thinking of sys-mode, so with diskless everything is in a single partition, so there should be a cmdline.txt file there 2023-07-03 00:05:31 let me run find on the entire partition for that lol 2023-07-03 00:05:35 how did you install? setup-alpine? it should have asked you whether diskless or Sysmode 2023-07-03 00:05:39 look at / 2023-07-03 00:05:56 minimal: yes, setup-alpine. it never asked. and root has no loose files. only dirs 2023-07-03 00:06:10 in / I have bin dev etc home lib media mnt opt proc root run sbin srv sys tmp usr var 2023-07-03 00:06:12 that's it 2023-07-03 00:06:28 i assume setup-alpine worked since my hostname is configured and i can lbu commit and reboot and it works 2023-07-03 00:06:37 that makes no sense, for RPI to boot there needs to be files in the top-level of the FAT partition 2023-07-03 00:06:48 minimal: i agree, hence I am here :D 2023-07-03 00:07:38 minimal: here's lsblk https://pastebin.com/T8xSwBdc 2023-07-03 00:08:09 like config.txt, cmdline.txt, usercfg.txt 2023-07-03 00:09:06 minimal: yeah im running find / -name "cmdline.txt" 2>/dev/null 2023-07-03 00:09:27 and bootcode.bin, fixup.dat, start.elf etc 2023-07-03 00:09:48 uh ok 2023-07-03 00:09:51 something is fucky 2023-07-03 00:09:53 it's at /media/mmcblk0p1/cmdline.txt 2023-07-03 00:09:56 but not at / 2023-07-03 00:09:58 ????!?! 2023-07-03 00:10:05 what in the world 2023-07-03 00:10:34 that's the 1st partition on the SDcard, when I said "/" I was referring to the top-level of the SDcard, not the "run-from-ram" rootfs 2023-07-03 00:10:40 how can I get a boot menu for alpine? 2023-07-03 00:10:42 ohHH 2023-07-03 00:10:43 i see 2023-07-03 00:10:47 minimal: thanks 2023-07-03 00:10:50 diskless is confusing 2023-07-03 00:11:05 tjaguar[m]: Grub can't detect wintoy 2023-07-03 00:11:21 tjaguar[m]: enable it? 2023-07-03 00:11:46 could you put me in the same page ? 2023-07-03 00:11:54 wiki maybe 2023-07-03 00:12:27 hmm wait, files in this dir are readonly and even root can't modify them ? ? minimal 2023-07-03 00:12:59 minimal: how? 2023-07-03 00:13:02 wwwww: is it mounted RO (readonly) or RW (read-write)? 2023-07-03 00:13:05 oh wait i need to mount it rw 2023-07-03 00:13:08 minimal yeah 2023-07-03 00:13:09 ty 2023-07-03 00:13:16 i need coffee 2023-07-03 00:14:12 tjaguar[m]: change setting in /etc/default/grub (GRUB_TIMEOUT from memory, set it to something other than 0) and the regenerate grub.cfg 2023-07-03 00:16:07 Thansk 2023-07-03 00:16:13 minimal: I'm trying it out 2023-07-03 00:19:53 "ttjaguar: change setting in /etc..." <- There aren't any grub directory 2023-07-03 00:20:18 tjaguar[m]: which directory? 2023-07-03 00:21:02 /etc/default/grub is a file, not a directory 2023-07-03 00:22:12 I don't have the file 2023-07-03 00:22:16 Using nano 2023-07-03 00:23:35 so how have you configured Grub? 2023-07-03 00:24:08 you have the grub package installed? 2023-07-03 00:26:23 well my unclea pc was installed with grub 2.06 something 2023-07-03 00:28:43 what does "apk info grub" show? 2023-07-03 00:30:19 ? 2023-07-03 00:30:28 did you run that? 2023-07-03 00:31:01 I think grub was already installed 2023-07-03 00:31:08 did you run that command? 2023-07-03 00:31:15 no 2023-07-03 00:31:18 why not? 2023-07-03 00:31:41 I will run it -a sec 2023-07-03 00:32:15 Grub-2.06 2023-07-03 00:32:31 ? 2023-07-03 00:32:38 minimal: that doesn't say if it's installed 2023-07-03 00:32:38 A sec 2023-07-03 00:34:37 psykose: doh! I was letting the "installed size" output confuse me 2023-07-03 00:34:44 ACTION uploaded an image: (131KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/yONzqBcDoTrmxFKkGnJtPxmc/20230703_083350.jpg > 2023-07-03 00:34:48 Its showing this 2023-07-03 00:34:50 tjaguar[m]: sorry, "apk list -I grub" 2023-07-03 00:36:00 alr 2023-07-03 00:37:46 ACTION uploaded an image: (134KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/wlnNfhEWrCTgEyowNzASivmC/20230703_083728.jpg > 2023-07-03 00:39:01 lol 2023-07-03 00:39:47 you saw nothing 🫡 2023-07-03 00:44:38 "ttjaguar: sorry, "apk list -I..." <- thanks for helping. I will do the stuff by my own now-cya!! 2023-07-03 00:47:21 tjaguar[m]: are you *SURE* you are using Grub installed by Alpine and not Grub installed by another OS/distro on a multiboot system? 2023-07-03 00:47:42 yes 2023-07-03 00:47:54 😅 2023-07-03 00:48:03 yes you are using Grub installed by Alpine? or yes you are using Grub installed by another OS/distro? 2023-07-03 00:48:44 So this was installed by another system ? 2023-07-03 00:48:50 I am asking you 2023-07-03 00:49:04 it is you computer, how would I know? 2023-07-03 00:50:11 I took the computer from a office. It is possible that the grub was installed from another distro 2023-07-03 00:50:42 so then how to configure that Grub would not be an Alpine question 2023-07-03 00:50:57 you "took" it? 2023-07-03 00:51:12 yes 2023-07-03 00:51:27 what does "took" mean? you were given it? 2023-07-03 00:51:40 yes 2023-07-03 00:53:36 you still haven't indicated the output of "apk list -I grub" but I assume it will not show the grub package installed, which would raise the question of how exactly you installed Alpine in the 1st place (if the installer chose Syslinux then that would not explain how Grub is running at boot) 2023-07-03 00:55:19 perhaps Syslinux is running at boot and you incorrectly assumed Grub was being used? 2023-07-03 00:57:37 Let me put sreight im a windows user 2023-07-03 00:57:42 Straight 2023-07-03 00:57:51 He left me like this 2023-07-03 00:58:02 He? 2023-07-03 00:58:10 Was doing some resreach on how to install windows 2023-07-03 00:58:13 But failed 2023-07-03 00:58:49 minimal: The uncle who worked in the office 2023-07-03 00:59:15 He was going somewhere else so i was gifted the lap 2023-07-03 01:00:52 "perhaps Syslinux is running at..." <- ok what should i do now? 2023-07-03 01:01:20 let me send you the pic privately 2023-07-03 01:03:12 you don't need to send me an image, use "apk list -I syslinux" to see if it is installed 2023-07-03 01:03:52 AS WELL AS running "apk list -I grub" 2023-07-03 01:04:52 ok 2023-07-03 01:05:39 Both are installed 2023-07-03 01:09:26 so (a) I'd expect /etc/default/grub file to exist as that is part of the grub package 2023-07-03 01:11:22 wait I will reinstall grub 2023-07-03 01:11:39 however (b) the grub configuration stuff also makes use of Syslinux's /etc/update-extlinux.conf file if syslinux package is installed 2023-07-03 01:12:39 Yes fixed 2023-07-03 01:12:40 in which case you can edit /etc/update-extlinux.conf and change "hidden=1" to "hidden=0" and then regenerate grub.cfg 2023-07-03 01:12:50 minimal: Its fixed 2023-07-03 01:12:59 what is fixed? 2023-07-03 01:13:28 tjaguar[m]: It enabled something called uefi firmware 2023-07-03 01:13:37 Settings 2023-07-03 01:13:44 I clicked it 2023-07-03 01:13:51 Then it shows 2023-07-03 01:14:08 These stuff 2023-07-03 01:14:28 ACTION uploaded an image: (70KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/mJAtudVsttlcahVhaBaFGkBy/20230703_091412_3681294334038923662.jpg > 2023-07-03 01:15:18 now I think i can boot with pendrive 2023-07-03 01:15:31 Let me see 2023-07-03 01:16:35 your original question was "how can I get a boot menu for alpine?" 2023-07-03 01:16:49 Yes 2023-07-03 01:17:03 not "how can I boot Alpine" but rather to get a boot menu 2023-07-03 01:17:23 and then you mentioned Grub 2023-07-03 01:17:40 it does matter now 2023-07-03 01:17:50 I got a problem 2023-07-03 01:17:52 what matters now? 2023-07-03 01:19:29 minimal: installing windows10 2023-07-03 01:20:25 From my guess the ufei boot menu was disabled 2023-07-03 01:20:42 I reinstalled the grub to make it work 2023-07-03 01:20:59 But now it does now detect the pendrive 2023-07-03 01:21:04 What do i do? 2023-07-03 01:22:22 minimal: Thanks for the support on enabling the boot menu 2023-07-03 01:23:29 tjaguar[m]: The lap can't detect the pendrive 2023-07-03 01:23:50 s/now/not/ 2023-07-03 01:27:36 "what matters now?" <- Thanks for the help and patience bro. So bye cya then 😁 2023-07-03 01:28:08 fyi: deleting messages doesn't matter here since the channel is logged on IRC side 2023-07-03 08:43:48 try to net boot alpine 3.18.2 on my x86_64 pc, with amd gpu + igpu + broadcom wifi card, nvme ssd ... it stuck at Loading hardware drivers 2023-07-03 08:47:11 what is the step after /etc/init.d/hwdrivers? 2023-07-03 08:47:15 bootmisc? 2023-07-03 08:50:35 that's usually missing firmware or whatever usually for the graphics driver 2023-07-03 08:52:18 ohh, maybe i need to add firmware to pkgs=? 2023-07-03 08:55:20 maybe this? linux-firmware-amd or linux-firmware-radeon 2023-07-03 08:56:05 or linux-firmware-amdgpu 2023-07-03 08:57:11 i don't know which to add, i have RX 470 2023-07-03 08:57:26 that one 2023-07-03 08:59:25 amdgpu? 2023-07-03 09:03:54 psykose: thank you , adding pkgs=linux-firmware-amdgpu worked. 2023-07-03 09:11:37 looks like intel iGPU does not have/need a separate firmware? 2023-07-03 09:50:17 Hi. Could I ask for the usermode iscsi target tgt, alternatively scst to be added to Alpine? 2023-07-03 09:51:42 how does one add that 2023-07-03 09:52:30 Tgt is simply a software with no kernel module. So just needs packaging 2023-07-03 09:52:55 Might be slightly slower, but won't bring down the kernel if there is a bug 2023-07-03 09:53:41 iSCSI via LIO (targetcli) is really buggy and I am getting fed up :O 2023-07-03 09:54:03 I've tried many distros, and it always causes problems eventually 2023-07-03 09:55:54 scst is https://github.com/SCST-project/scst ? 2023-07-03 09:56:22 ah it's a kernel module thing 2023-07-03 10:02:30 I'm thinking scst is better than LIO with more features 2023-07-03 10:03:02 But tgt is entirely user mode, which is safer if there are iscsi bugs 2023-07-03 10:18:22 sure, added tgt 2023-07-03 10:20:00 Really. That is great 2023-07-03 10:20:06 :) 2023-07-03 10:20:24 test and find out 2023-07-03 10:21:42 I will. Tank you! 2023-07-03 10:22:20 Is it in @testing or should I get from git? 2023-07-03 10:23:19 former 2023-07-03 11:24:05 Installs ok. 2023-07-03 11:24:12 Starts ok 2023-07-03 11:24:34 Won't shutdown ok. Keeps running after rc-service tgtd stop 2023-07-03 11:25:17 And there seems a tool missing tgt-admin (not tgtadm) 2023-07-03 11:25:30 tgt-admin is what reads the config files 2023-07-03 11:28:19 Ahh i needed apk add scsi-tgt-scripts 2023-07-03 11:29:23 does the daemon call it or nah 2023-07-03 11:29:42 But it gives a perl warning when running. Can't located Config/General.pm 2023-07-03 11:30:09 No tgtd is stand alone. It takes commands from tgtadm to add config on the fly 2023-07-03 11:33:24 the default sigterm is installed 2023-07-03 11:33:26 ignored* 2023-07-03 11:33:29 they install an empty handler 2023-07-03 11:33:36 nice dogshit programming 2023-07-03 11:34:20 Heh 2023-07-03 11:34:53 Tried to install perl-conf-libconfig. Didn't help 2023-07-03 11:36:00 fixed both 2023-07-03 11:37:06 Do I reinstall scsi-tgt? 2023-07-03 11:37:32 you just upgrade after it builds 2023-07-03 11:39:10 A cool 2023-07-03 11:56:56 Seems it takes a while for the mirrors to update. Which perl package should I install.? 2023-07-03 11:58:17 hi, is there a man page for busybox sh available on the internet somewhere? 2023-07-03 11:59:21 don't think there is 2023-07-03 11:59:25 Forza: will take a while to build 2023-07-03 12:01:59 Ah it's a new perl module? 2023-07-03 12:02:04 ye 2023-07-03 13:18:25 So, what's a good X screen locker on Alpine. There are several in the repos, as far as I can see. 2023-07-03 13:19:57 What are your judging criterias? 2023-07-03 13:20:18 (I use physlock) 2023-07-03 13:20:30 Ah, what's waht I was gonna try. ;) 2023-07-03 13:21:13 I'm not using an integrated desktop application suite (aka "desktop environment"), so something nicely independent will be nice. 2023-07-03 13:22:07 My only two criteria are really: 1. It should be reasonably secure. 2. It should be able to lock all VTs, also from within X. 2023-07-03 13:25:22 Well, physlock seems alright. Gonna use that then. 2023-07-03 13:32:03 Hello. 2023-07-03 13:32:39 Is there something I could do to get thing rollin on https://gitlab.alpinelinux.org/alpine/aports/-/issues/11573 ? 2023-07-03 13:32:55 Or is this aports not official stuff of alpine? 2023-07-03 13:34:04 get rolling on what 2023-07-03 13:34:18 Get rolling on buildng a alpine package 2023-07-03 13:34:19 the abandoned software or the software with no releases 2023-07-03 13:34:48 You mean alpine linux? 2023-07-03 13:35:15 Looks like there is no interest 2023-07-03 13:35:24 tac_plus+ is already packaged (https://www.pro-bono-publico.de/projects/tac_plus.html) unless you meant something else 2023-07-03 13:35:32 the other one is the abandoned version of it that was forked 2023-07-03 13:36:17 the pkgname is tacacs+ng 2023-07-03 13:36:22 psykose: I have to leave work for today but will be back on Wednesday to continue with the tgt bits :D 2023-07-03 13:36:36 https://pkgs.alpinelinux.org/packages?name=tac&branch=edge&repo=&arch=&maintainer= I can not find it in this list. 2023-07-03 13:37:05 not tac_plus available. Is this package you are refering to ins some other alpine linux ? 2023-07-03 13:37:14 repository? 2023-07-03 13:37:44 https://pkgs.alpinelinux.org/packages?name=tacacs%2Bng&branch=edge&repo=&arch=&maintainer= got it 2023-07-03 13:38:22 thanks psykose 2023-07-03 13:38:23 probably works, give it a try 2023-07-03 13:38:29 Forza: enjoy :) 2023-07-03 13:38:52 msiism, actually I just tried on this laptop, it doesn't work :/ 2023-07-03 13:39:09 (this install is a few weeks new and I had forgotten to install it) 2023-07-03 13:39:16 Works fine on another laptop with other distro though 2023-07-03 13:44:21 Uhhm… works for me here, on "ancient" x86_64 laptop. 2023-07-03 13:44:32 I can lock everything from within X. 2023-07-03 13:45:11 Interesting, mine switches to tty, but doesn't start, only a prompt is blinking, withouth possible input 2023-07-03 13:45:20 Hm… 2023-07-03 13:45:23 I can reset with ctrl+alt+suppr though 2023-07-03 13:45:31 What's your deskop envrionment? 2023-07-03 13:45:41 none 2023-07-03 13:46:14 Come on… ;) 2023-07-03 13:46:20 There's gotta be something. 2023-07-03 13:46:31 Window Manager? 2023-07-03 13:47:02 I use Openbox, Xterm and some other nifty stuff. 2023-07-03 13:47:13 Ah sorry, I thought you really meant desktop environment 2023-07-03 13:47:40 Here it's a window manager (dwm) over Xorg, started with startx (if that means anything specific) 2023-07-03 13:48:56 Might have something to do with the pam module, that isn't present on the other system 2023-07-03 13:49:23 msiism, are you using systemd facilities? 2023-07-03 13:51:31 Ah, also Alpine isn't using the release, but a specific git commit 2023-07-03 13:52:05 looks like physlock wants elogind and pam according to the Git repo 2023-07-03 13:52:34 It's not mandatory 2023-07-03 13:52:50 Not sure, I'm using systemd facilities. I mean, Alpine doesn't use systemd. 2023-07-03 13:53:45 quinq: Are you on the edge branch? 2023-07-03 13:53:45 Some parts of it 2023-07-03 13:53:48 yeah 2023-07-03 13:53:55 I'm on 3.18. 2023-07-03 13:54:07 I'll strace it, bbl 2023-07-03 13:54:27 Okay, nice. 2023-07-03 13:59:24 Erf, in /etc/pam.d/physlock, there's only include system-local-login 2023-07-03 13:59:35 But system-local-login isn't actually something 2023-07-03 14:02:05 Yep, works fine when fixed 2023-07-03 14:02:21 Good to know. 2023-07-03 14:02:49 ACTION is off for actual work. 2023-07-03 14:03:25 I'm making a patch 2023-07-03 14:03:45 cool 2023-07-03 14:04:55 looks to already be done? 2023-07-03 14:05:17 https://gitlab.alpinelinux.org/alpine/aports/-/commit/9b47c65653d9b2541899d2de0b08f98baacb9cb6 2023-07-03 14:05:17 Yeah, just a second 2023-07-03 14:06:06 about 23 minutes ago 2023-07-03 14:06:48 Ah, I see what you mean 2023-07-03 14:06:55 Seems like psykose sneak-patched it 2023-07-03 14:07:00 yep 2023-07-03 14:08:30 is it a sneak if it was visible 2023-07-03 14:09:30 Kind of yeah, as it didn't take part of the discussion ;) 2023-07-03 14:09:38 (i also didn't think it was the actual fix, just unrelated) 2023-07-03 14:09:50 But thanks 2023-07-03 14:13:30 When I am installing with setup-alpine it asks me this... (full message at ) 2023-07-03 14:13:56 The docs are unclear on what i should enter here 2023-07-03 14:17:31 password 2023-07-03 14:18:02 psykose: now it makes sense why it wasnt connecting to network for 3 times 🤦🏼 2023-07-03 14:18:30 idk why it says "pre shared key" since everyone calls that a password so it's confusing :D 2023-07-03 14:18:38 but formally yes a psk is just a password that you entered 2023-07-03 14:18:59 it's a key and it's pre-shared wrt connection state 2023-07-03 14:22:29 psykose: yeah should be fixed. psk refers to the pre-shared key you generate with wpa_passphrase, if you refer to wpa_supplicant which alpine uses by default 2023-07-03 14:25:13 that's not correct, no 2023-07-03 14:26:29 hm, well might be 2023-07-03 14:26:33 i guess it's confusing regardless 2023-07-03 14:26:59 A PSK is a key. 2023-07-03 14:27:09 wpa_supplicant supports both formats 2023-07-03 14:27:27 So the “issue” is regarding whatever script wraps around that 2023-07-03 14:27:59 If that script expects the key *only* in clear-text form, then it might be a good idea to say so 2023-07-03 15:50:33 psykose: continue the chapter : -virt kernel does not need amdgpu, it does not stuck. -lts kernel with / without amdgpu, it stucks. 2023-07-03 15:50:55 i am just amazed -virt kernel boots my hardware... 2023-07-03 15:51:42 yeah cause having no drivers is fine 2023-07-03 15:51:50 but then having them with no firmware is not 2023-07-03 15:52:16 the boot args are the same, except that i replace all -virt to -lts 2023-07-03 15:52:29 then -lts stucks. 2023-07-03 15:52:54 psykose: i have nomodeset 2023-07-03 15:53:01 so 2023-07-03 15:53:13 i don't know why lts kernel stucks... 2023-07-03 15:53:54 i mean, with `nomodeset` it will use the kernel's framebuffer, it does not try to load gpu driver/firmware 2023-07-03 15:56:05 i think it is something else. i tried -lts with pkgs=linux-firmware-amdgpu 2023-07-03 15:56:21 still stucks at Loading hardware drivers 2023-07-03 15:56:27 good luck findin it 2023-07-03 15:56:47 there is no way to find it. it is a netboot... i load everything from iPXE 2023-07-03 15:57:16 there is no debug messages after stuck :P 2023-07-03 15:57:40 anyway, -virt works fine, even without pkgs= 2023-07-03 15:57:53 i'll use that for now. 2023-07-03 16:43:20 Is it true that img2sixel isn't available on Alpine at all? 2023-07-03 16:48:11 apk add cmd:img2sixel 2023-07-03 16:50:14 Thanks. Why cmd? 2023-07-03 16:52:27 `cmd:` is a prefix for a command provider, a package that provides a given command 2023-07-03 16:52:42 you might notice that when you run that, it actually installs libsixel-tools 2023-07-03 18:30:03 Has anyone succsessfully installed the tor _browser_ on musl based alpine? 2023-07-03 18:48:35 scara: It works with flatpak. 2023-07-03 18:49:25 Oh, I thought flatpak wasn't recommended, but whatever gets it to finally run, thanks :) I'll try that. 2023-07-03 18:49:55 (That sentence almost made no sense, but I hope you get the point). 2023-07-03 18:50:00 ptrc: and cmd:XXX ends up in /etc/apk/world? That's pretty cool, it's makes the world file so much clearer. 2023-07-03 18:50:03 For running glibc-only things, you kinda have to 2023-07-03 18:50:09 WhyNotHugo: yes, it does 2023-07-03 18:50:12 scara: I wouldn't say that I recommend it: I just said it works. 2023-07-03 18:50:40 Your alternative is to build from source (likely a pain) or use a chroot with some glib distro. 2023-07-03 18:50:44 *glibc 2023-07-03 18:52:31 I tried the chroot thing. Maby I should use aglibc distro for my graphical machines, kinda like how I kinda have to use gnu utils and can't just rely on busybox, it kinda really sucks but thats just how software is thease days. 2023-07-03 18:53:23 WhyNotHugo: yeah, that's also how dependencies work for dynamically linked stuff ( depending on `so:*` instead of package names ), but you also can install dev packages based on their pkgconfig name (e.g. `pc:adwaita-qt6`) or even python packages (`py3.11:pypi`) 2023-07-03 18:54:36 err, s/pypi/pip/ 2023-07-03 18:55:30 Is using pc:XXX in a APKBUILD discouraged? 2023-07-03 18:55:40 That would have made writing some of them so much easier tbh 2023-07-03 18:57:56 ye 2023-07-03 19:01:27 Are there "devel" packages for libxml2 and libxslt? 2023-07-03 19:01:58 both suffixed with -dev 2023-07-03 19:01:59 leitz: libxml2-dev 2023-07-03 19:02:09 as in, libxml2-dev and libxslt-dev, yeah 2023-07-03 19:02:54 WhyNotHugo, ptrc, thanks! 2023-07-03 21:22:51 i find myself using apk add cmd:... quite often 2023-07-03 21:25:19 did you ever follow up that glfw thing 2023-07-03 21:40:11 in a way: https://github.com/glfw/glfw/pull/2348 but doesn't seem like the maintainers are there 2023-07-03 21:46:16 i suspect i already asked this, but probably i didn't like the answer so i didn't remember it: how are rust projects built in rootbld-like settings as they seem to lack options=net which is needed for cargo? 2023-07-03 21:54:59 you add options=net more or less 2023-07-03 21:58:54 what's the ideal end state for these build systems that like to fetch stuff (like cargo and occasionally cmake)? abuild growing the smarts to figure out what to fetch? 2023-07-03 22:08:54 to allow it in prepare without options=net and that's about it 2023-07-03 22:09:03 so i generally write with that in mind even if that's not the case yet 2023-07-03 22:09:11 or uh, ever probably 2023-07-04 06:55:41 good morning! is it just me or is chromium sort of broken in edge now? 2023-07-04 06:55:49 x86_64 2023-07-04 07:25:48 chromium has been strange lately but a bit random, today was a good day for it... 2023-07-04 09:00:18 I have created an aport for dxx-rebirth, a port of parallax's descent and descent2 games from the 90's. Does any one have any interest in this being in aports? 2023-07-04 11:56:02 Using linux-lts, but missing some firmware, sudo apk add linux-firmware show's 95 installable, but all gives Permission denied 2023-07-04 11:56:24 CPU is Intel Xeon E5-2697 v2 2023-07-04 11:57:50 dmesg show's nothing, my 10gbe not works as expected, a working server using Xeon E5-2680 v4, the driver is bnx2x, but this fw can not be installed in Xeon E5-2697 v2 2023-07-04 11:58:52 Is this caused by cpu features or something ? 2023-07-04 12:03:04 alpine 3.18 2023-07-04 12:13:12 wener: which iso did you use to install Alpine to your server? 2023-07-04 12:48:06 also, is it a diskless install? 2023-07-04 14:02:12 hm, the latest chromium has some rather severe visual bugs with slack and discord both 2023-07-04 14:03:01 rm -r ~/.config/chromium/ShaderCache 2023-07-04 14:03:05 thanks :) 2023-07-04 14:03:21 mostly https://bugs.chromium.org/p/chromium/issues/detail?id=1442633 2023-07-04 14:03:35 ah, no joy there 2023-07-04 14:03:44 I have render tiles flickering and popping in and out 2023-07-04 14:04:07 that does look super similar though, hm 2023-07-04 14:04:20 ah it's Default/GPUCache 2023-07-04 14:05:57 cheers, that fixed it! 2023-07-04 14:06:03 elly: ncopa reported that in #alpine-devel 2023-07-04 14:06:18 ah, I had not read the backlog there yet 2023-07-04 14:07:05 https://gitlab.alpinelinux.org/alpine/aports/-/issues/15069 2023-07-04 14:09:03 "It seems these are cases where Chromium can't invalidate cached shader blobs because the GL_RENDERER string doesn't change across package updates even though the update is clearly a breaking change" 2023-07-04 14:09:37 ahh 2023-07-04 14:10:00 #28 on the chromium bug has a better explanation though 2023-07-04 14:11:16 modern gpu moment 2023-07-04 14:11:28 20 layers of shader languages and caches to render a white background on a webpage 2023-07-04 14:12:04 yep 2023-07-04 14:12:22 an awe-inspiring cathedral of software complexity which ultimately rasters a flat color onto a screen 2023-07-04 14:12:47 the overall impression is somewhat like building the Space Shuttle in order to drive it down to the shops 2023-07-04 14:14:13 but boy is it an optimised drive to the shops 2023-07-04 14:14:25 like no other 2023-07-04 14:15:31 until you have to find parking, at any rate :) 2023-07-04 14:15:53 which I guess in this metaphor is like, icache 2023-07-04 14:22:16 <- thinks google is a virus has changed a source of knowlege into nothing but a source of ads. 2023-07-04 14:22:44 somewhat of a cold take over there, but also not really the topic of discussion :P 2023-07-04 14:24:58 I don't know who to contact here, I have wrote a how to install LMV over RAID1 with SD spare for Raspberry PI 4. 2023-07-04 14:33:04 any viral pathogen experts 2023-07-04 14:33:14 txnintn: create an account on the Alpine Wiki and add your article there? 2023-07-04 14:33:46 I have an account and "talk" 2023-07-04 14:34:19 https://wiki.alpinelinux.org/wiki/User:Txnintn 2023-07-04 14:42:42 I'm tinkering with Raspberry PI install in diskless mode. 2023-07-04 16:42:38 ncopa sys install 2023-07-04 16:42:53 that's sudo ncopa sys install 2023-07-04 16:43:33 Ermine std iso 2023-07-04 16:43:52 Simon Says "sys install" 2023-07-04 16:45:36 Working one 10gbe `ethtool -i eth4` report bnx2x, other's report mlx4_en 2023-07-04 17:05:22 wener: you can try extended iso, it has all possible firmwares included 2023-07-04 17:07:08 I disabled the eth4, reinstall linux-firmware without error, the Permission denied is caused by apk fetch failed, 2023-07-04 17:08:43 but ethtool report `Current message level: 0x00000014 (20) link ifdown Link detected: yes` 2023-07-04 17:21:24 I guess at this point you should configure the interface and bring it up 2023-07-04 17:22:16 Seems Mellanox not works well with dell 2023-07-04 17:22:29 The working one is Broadcom 2023-07-04 17:22:47 Interface is up, mtu 9000, but speed is zero 2023-07-04 17:23:09 did you set an ip address? Or did you get one via dhcp? 2023-07-04 17:23:49 all static 2023-07-04 17:34:34 Try to ping something, does it work? 2023-07-04 18:55:09 Hey folks, networking documentation suggests to use `192.168.1.150/24` since 3.13, but on my 3.18 it doesn't work, I have to specify just IP there and netmask separately. 2023-07-04 18:55:10 Has something changed again or something is wrong with my system? 2023-07-04 18:55:10 https://wiki.alpinelinux.org/wiki/Configure_Networking 2023-07-04 18:56:24 not works if I only have the 10gbe up 2023-07-04 18:56:46 seems I have to setup infiniband 2023-07-04 18:57:44 nazar-pc[m]: you have to add ifupdown-ng iirc 2023-07-04 19:01:41 psykose: Hm, now I see docs mention that Alpine utilizes ifupdown-ng, but my system doesn't have it. 2023-07-04 19:01:41 I was installing Alpine recently on a different machine and saw that a set of default packages seems to be different from what I have, which is somewhat disappointing, would be great if my system after upgrades was more similar to new installation. 2023-07-04 19:02:05 you have to manually add it 2023-07-04 19:03:12 I also noticed adding more interfaces and restarting networking actually breaks networking (but works after reboot), I guess that might also be fixed with ifupdown-ng 2023-07-04 19:07:56 Seems caused by mtu, WTF, https://github.com/geerlingguy/raspberry-pi-pcie-devices/issues/21#issuecomment-1427176397 this comment say the mtu > 4096 may fail 2023-07-04 19:08:18 After change to 1500, iperf3 give the correct speed 2023-07-04 19:09:35 ha 2023-07-04 19:29:04 https://docs.nvidia.com/networking/display/MLNXOFEDv494170/Known+Issues Too much mtu related issues TOT 2023-07-04 19:29:32 If BSDs use efi executables (as does linux), why won't binaries statically compiled on one work on the other? 2023-07-04 19:29:51 different kernel ABIs 2023-07-04 19:29:52 syscall not match 2023-07-04 19:30:52 So potentially one could try a few different calls in the correct order and "guess" which kernel is running? 2023-07-04 19:31:34 WhyNotHugo: https://justine.lol/ape.html 2023-07-04 19:33:54 the way the binary is loaded and launched isn't the same either 2023-07-04 19:35:11 if you were willing to write some clever hand assembly I bet you could make a binary that works on both 2023-07-04 19:35:24 elly: check that APE page 2023-07-04 19:35:38 conceptually cool but overall a quite useless trick 2023-07-04 19:35:53 ikke: yep, I know of APE (although not of any big uses of it) 2023-07-04 19:36:00 a cool piece of sleight of hand 2023-07-04 19:36:02 So this guys bundles a js interpreter which executes a script and the total size is 300kb. 2023-07-04 19:36:06 or, well, not sleight of hand exactly 2023-07-04 19:36:08 ACTION looks at electron 2023-07-04 19:36:34 the js interpreter is a tiny fraction of what electron does, regrettably 2023-07-04 19:36:42 This article is very interesting, answers a lot of the questions I had, thanks! 2023-07-04 19:38:33 electron is what happens when chromium is open source 2023-07-04 19:38:39 if chromium was closed source it wouldn't happen 2023-07-04 19:38:59 Well, it would have been done with something else 2023-07-04 19:39:22 psykose: getting a lot of mileage out of my "THIS WOULDN'T BE HAPPENING IF WE WERE CLOSED SOURCE" t-shirt really 2023-07-04 19:39:30 :D 2023-07-04 19:40:11 I wouldn't have to moderate the bug tracker either actually 2023-07-04 19:40:13 gosh 2023-07-04 19:42:32 how's that these days 2023-07-04 19:42:33 :D 2023-07-04 19:42:50 actually not that bad! 2023-07-04 19:42:55 brb, starting a jxl flamewar while you're busy replying 2023-07-04 19:43:04 joke's on you I am on vacation 2023-07-04 19:43:07 :D 2023-07-04 19:43:15 literally sitting in an airport presently 2023-07-04 19:43:37 i think the vacation starts after you leave the airport 2023-07-04 19:43:40 or does it 2023-07-04 19:43:49 well, me not doing my job started a couple days ago :) 2023-07-04 19:43:55 :D 2023-07-04 20:08:16 WhyNotHugo: I think you are getting EFI binaries (run by UEFI for booting) mixed up with ELF binaries 2023-07-04 21:35:04 I 2023-07-04 21:35:29 I'm struggling to set the locales 2023-07-04 21:35:56 struggling how? 2023-07-04 21:36:03 Tried exporting this in /etc/profile: export MUSL_LOCPATH="/usr/share/i18n/locales/musl" 2023-07-04 21:36:03 export CHARSET=sv_SE.UTF-8 2023-07-04 21:36:03 export LC_ALL=en_US.UTF-8 2023-07-04 21:36:03 export LC_COLLATE=en_US.UTF-8 2023-07-04 21:36:03 export LANG=en_US.UTF-8 2023-07-04 21:36:06 export LC_CTYPE=en_US.UTF-8 2023-07-04 21:36:08 export LC_MONETARY=sv_SE.UTF-8 2023-07-04 21:36:10 export LC_NUMERIC=sv_SE.UTF-8 2023-07-04 21:36:12 export LC_TIME=sv_SE.UTF-8 2023-07-04 21:36:14 2023-07-04 21:36:27 But I only get this from locale: LANG=C 2023-07-04 21:36:27 LC_COLLATE=C 2023-07-04 21:36:27 LC_NUMERIC=C 2023-07-04 21:36:27 LC_CTYPE=C 2023-07-04 21:36:27 LC_TIME=C 2023-07-04 21:36:28 LC_MONETARY=C 2023-07-04 21:36:30 LC_MESSAGES=C 2023-07-04 21:36:32 LC_ALL= 2023-07-04 21:36:32 use a pastebin 2023-07-04 21:37:22 you set them correctly 2023-07-04 21:39:20 Do I have to generate locales, like on debian? 2023-07-04 21:39:58 no 2023-07-04 21:40:24 the `locale` thing from `musl-locales` just prints you what the env vars are set to 2023-07-04 21:40:43 so if you put something in /etc/profile and they're still =C that just means you haven't opened a new session that sources the profile (a new login shell) 2023-07-04 21:40:52 I 2023-07-04 21:40:53 if you do `ash -l` or whatnot they would be updated 2023-07-04 21:41:01 I've tried restarting the computer 2023-07-04 21:41:12 what does `echo $LANG` say 2023-07-04 21:41:23 Only C 2023-07-04 21:41:34 can you pastebin /etc/profile 2023-07-04 21:42:58 https://pastebin.com/wp9pX66T 2023-07-04 21:44:44 if you `echo $MESA_LOADER_DRIVER_OVERRIDE` what does it say 2023-07-04 21:45:49 just blank 2023-07-04 21:46:15 what shell do you use 2023-07-04 21:46:21 maybe bash doesn't parse my profile? 2023-07-04 21:48:27 if it's bash what does `bash -l` and then `echo $MESA_LOADER_DRIVER_OVERRIDE` do 2023-07-04 21:49:59 bash -l does nothing 2023-07-04 21:50:16 'and then' ... 2023-07-04 21:50:21 I mean, it does something, but it is not necessarily obvious 2023-07-04 21:50:29 echo $MESA_LOADER_DRIVER_OVERRIDE gives r600 2023-07-04 21:50:44 how do you normally start bash at all 2023-07-04 21:51:04 in gnome-console in a gnome session, I 2023-07-04 21:51:17 I've just used chsh to change to bash for my user 2023-07-04 21:52:04 I 2023-07-04 21:52:05 you open gnome-console then just type 'bash' or 2023-07-04 21:52:39 I'm gonna restart the session and check now, brb 2023-07-04 21:55:34 So I've used chsh to change my shell to /bin/bash 2023-07-04 21:56:00 so it launches automatically when I open the terminal, as the default shell for my user 2023-07-04 21:57:13 seems like bash -l loads the profile 2023-07-04 21:57:28 Because then the locales are set 2023-07-04 21:59:12 which shell were you using previously? 2023-07-04 21:59:40 the default one, I guess that's ash? 2023-07-04 22:01:23 busybox ash. 2023-07-04 22:01:51 so ash for root and bash for my user 2023-07-04 22:10:28 Maybe I need to add them locally for bash, like in .bashrc? 2023-07-04 22:10:59 or .profile 2023-07-04 22:17:48 setting it in .bashrc seems to do the trick 2023-07-04 22:22:44 this sounds like a gnome session quirk somewhat 2023-07-04 22:22:57 generally, when you log in, whatever shell you logged in with sources /etc/profile 2023-07-04 22:23:05 then that environment is kept to anything you do after 2023-07-04 22:23:32 so for instance, you has /bin/sh as the login shell (ash). then on login on the tty, the env vars are there, and after you run `bash -i` or whatever, they're still there 2023-07-04 22:23:47 the fact you signed in via gnome gdm (i assume) then opened the terminal and they are not means something never sourced it 2023-07-04 22:24:15 iirc gnome session has some quirks where it does sourcing with bash only but idk if it would affect anything related to the actual login shell or how the environment was really updated 2023-07-04 22:25:08 i would have guessed it just uses bash to source the usual stuff and then keep the `env` and inject it into the session which would put them there.. 2023-07-04 22:25:15 maybe that breaks if the user has /bin/sh as login, idk 2023-07-04 22:58:48 Might be a gnome session/gdm quirk, I 2023-07-04 22:58:56 I'll investigate it further 2023-07-04 22:59:00 Thanks a lot! 2023-07-05 00:02:46 'adduser -D -G mygroup -G docker myuser' will add myuser to both groups right? 2023-07-05 00:04:52 no, just the second 2023-07-05 01:16:19 was wondering why I had 5 notifications from here, in hindsight changing my username to ash whilst also being in this channel was not the smartest idea 2023-07-05 01:24:06 from ashes til dust 2023-07-05 01:50:24 being in the 'docker' group and adding the socket volume is all thats needed to use the docker engine inside a container, or am i missing a piece 2023-07-05 01:50:44 still gettting docker: permission denied while trying to connect to the Docker daemon socket at 2023-07-05 01:51:36 you logout and back in after adding your user to the docker group? 2023-07-05 01:51:53 this is from inside a container 2023-07-05 01:52:58 docker in a container? 2023-07-05 01:52:59 and the addgroup/apk calls are done in the Dockerfile but doing 'docker run' at runtime fail 2023-07-05 01:53:33 yeah im trying to have a container that spawns others 2023-07-05 01:55:26 the user inside the container being in some group doesn't mean anything wrt the docker group on the host i think unless you do more mapping stuff 2023-07-05 01:55:36 as root inside the container it probably works with the socket mounted (check that first) 2023-07-05 01:55:54 if yes then it's some group mapping thing 2023-07-05 01:56:00 if not then there's other stuff missing 2023-07-05 01:58:31 it does work as root 2023-07-05 01:59:56 you can do something like -e GID=$(id -g docker) and use the env var inside the container to create a group with the same gid 2023-07-05 02:00:10 i think that's what the entrypoint stuff does to allow that uid/gid mapping that like every linuxserver.io container does 2023-07-05 02:00:22 looks something like https://stackoverflow.com/a/66455612 2023-07-05 02:00:30 but this is a ton of steps to not have to be root to reach docker of course 2023-07-05 02:06:08 If you're running docker as root, just use gvisor too :) 2023-07-05 02:06:57 Or podman 2023-07-05 02:10:10 I wish PodMan was a drop-in replacement for Docker for the stuff I use, it seems really nice and not having to run as root is also a huge bonus 2023-07-05 02:25:48 "was wondering why I had 5..." <- I remember having so much notifications spam since I had my username as me 2023-07-05 02:26:18 Had to turn off contains username entirely 2023-07-05 09:17:13 I'm using an alpine based Python 3.10 docker container, and trying to get Django up and running. So far I can't get the container OS to see pg_config (for postgresql), even when I put pg_config on the container and run ./pg_config. 2023-07-05 09:17:18 Any suggestions? 2023-07-05 09:23:44 how did you 'put it in the container' 2023-07-05 09:49:53 The initial install was wrong: "COPY pg_config /app/bin", but in the container, I moved "bin" to "pg_config", and tried ./pg_config. 2023-07-05 09:52:38 so you copied the pg_config from your host system into the container? 2023-07-05 09:52:41 that's never going to work 2023-07-05 09:53:07 you should install postgresql in the container somehow or get it from some source there 2023-07-05 09:53:13 one method is just using the postgres image 2023-07-05 09:53:29 https://hub.docker.com/_/postgres 2023-07-05 10:08:08 Thanks. 2023-07-05 14:36:06 I'm seeing a strange problem on Edge, running "ldd /usr/bin/python" as root gives the expected output, running as other users is giving "Error loading shared library libpython3.11.so.1.0: Permission denied (needed by /usr/bin/python)". Like "ldd /usr/bin/file" gives "Error loading shared library libmagic.so.1: Permission denied (needed by /usr/bin/file)" 2023-07-05 14:36:34 I don't see anything wrong with permissions etc in /usr/lib/ 2023-07-05 14:36:53 this is on a VM created about 1 hour from Edge repos 2023-07-05 14:38:00 i can't think of anything except a permission error 2023-07-05 14:38:15 that would give eacces on opening the libs when loading 2023-07-05 14:38:24 is / 755? 2023-07-05 14:38:30 indeed, but user, group, and perms, including the dir, all look ok 2023-07-05 14:38:37 the whole tree up to the dir has to be kosher 2023-07-05 14:38:49 if not that then no real clue 2023-07-05 14:40:11 actually its not just /usr/lib - noticed when doing "ls -al /" from a non-user account that the user and group columns were "0" rather than "root" so it can't even read /etc/groups - confirmed by trying to cat /etc/groups 2023-07-05 14:41:30 ah, /etc directory is "o=" for some reason 2023-07-05 14:42:57 fixing that has resolved "ldd" output - wonder how /etc/ ended up as "rwxr-x---" 2023-07-05 17:52:12 psykose: found the cause of the /etc permissions issue, I had some code setting up doas config and I'd stopped using /etc/doas.d/whatever.conf for the config and switched to /etc/doas.conf but the code had a "check" for the directory containing the config file and it was enforcing 0750....which as I'm now using /etc/doas.conf meant it was changing /etc permissions. Doh! 2023-07-05 17:53:19 :D 2023-07-05 17:53:22 happens 2023-07-05 17:53:26 done similar before 2023-07-05 19:59:56 how do I stop x11 shutting down? seems to do it randomly, and in a clean away - nothing weird in the logs 2023-07-05 20:00:48 That doesn't seem right. 2023-07-05 20:03:33 https://gist.github.com/LAC-Tech/a2154fea171368ff424f41b5379c354c 2023-07-05 20:03:59 AFAICT at 129177.130 it starts a clean shutdown 2023-07-05 20:12:10 I've had other linux distros on this machine and only alpine has this issue 2023-07-05 20:16:42 Never seen it 2023-07-05 20:16:54 No oom in dmesg? 2023-07-05 20:16:59 anything in /var/log/messages? 2023-07-05 20:18:05 how are you starting X anyway? is it possible that your client(s) are the issue? Your log looks just like a clean shutdown (as you mentioned) when X has no clients connected. 2023-07-05 20:20:23 starting X with 'exec dbus-run-session openbox-session'. hang on will post my /var/log/messages 2023-07-05 20:21:11 https://gist.github.com/LAC-Tech/94b464da2e4e511f20a7613918e5f16b 2023-07-05 20:25:52 there's nothing in dmesg around the time it shuts down. 129177 is when the shutdown stars, I have a skip between 128944 and 129384 2023-07-05 20:28:54 where's Xorg.conf in alpine? want to get it to use the amdgpu driver, maybe that will help 2023-07-05 20:29:25 There is none by default 2023-07-05 20:30:09 hmm. I'm out of ideas. I installed 'xf86-video-amdgpu'. hope that helps 2023-07-05 20:42:58 swaylock-effects won't install on aarch64/edge because of a bad signature; should I make an issue? 2023-07-05 20:44:17 elagost: what mirror? 2023-07-05 20:45:19 dl-cdn.alpinelinux.org 2023-07-05 20:47:13 elagost: fixed 2023-07-05 20:48:08 thanks, the -doc package has a bad signature too. 2023-07-05 20:48:21 done 2023-07-05 20:49:06 awesome, thank you so much ikke ! 2023-07-05 21:29:17 LACampbell, is it random or when you start a specific program? 2023-07-05 21:30:00 I have this issue on another machine, another distro (musl though), when starting Picard (that's one I can reproduce the issue with) it'll just shut down my X session, can't trace why 2023-07-05 21:35:54 LACampbell, I run sway this way: 2023-07-05 21:35:57 exec dbus-run-session -- sway > .local/session.log 2>&1 2023-07-05 21:36:25 this let's me keep a log of any activity that otherwise would not hit syslog 2023-07-05 21:36:45 try it with openbox-session 2023-07-05 21:57:03 hello? 2023-07-05 21:58:41 greetings 2023-07-05 22:00:43 i use alpine linux now from my Toshiba Portege 7020ct, that very cool, my ram consumption is 45mb 2023-07-05 22:02:27 Nice, that's a decently old one! 2023-07-05 22:03:48 i want to build a kernel to get hardware acceleration and install xorg 2023-07-05 22:04:49 The neofb module for my graphics was removed from lts kernel a year ago, it's sad 2023-07-05 22:05:26 :( 2023-07-05 22:05:53 what about Xorg's neomagic driver? 2023-07-05 22:06:00 i build it 2023-07-05 22:06:11 on my pentium 2 366mhz 2023-07-05 22:06:50 also, now i use tmux, w3m, irssi, htop, mc and wireguard lol 2023-07-06 02:21:16 I'm surprised / pleased it boots in 45mb 2023-07-06 02:21:29 a while ago psykose did some experiments and the stock alpine kernel hung trying to boot with 64 2023-07-06 02:21:46 I guess maybe you have more than 64 available, just only 45 used after boot 2023-07-06 02:30:10 i don't think it will work with 64 MB using the default initramfs because it decompresses into tmpfs 2023-07-06 02:30:29 yep, that was the problem - it ran out of memory decompressing the initramfs 2023-07-06 02:30:39 fixable in theory but I don't think she bothered experimenting further 2023-07-06 02:31:39 it's not really fixable without functionality reduction 2023-07-06 02:33:00 it might work with trimmed mkinitfs features 2023-07-06 02:33:45 e.g. you probably don't need all of cdrom, mmc, nvme, raid, and scsi 2023-07-06 08:15:10 I am new to this and did not know what I was doing bit I had a question about package manager in Alpine Linux. I am using it as a desktop system but even that the add command is blazingly fast I am looking for a more newbee way of software suggestions. 2023-07-06 08:26:39 What desktop do you use? 2023-07-06 08:29:09 Pelle_bom_high: 2023-07-06 08:54:55 Xfce 2023-07-06 12:14:26 Hello! I consider running Alpine as a desktop for a number of reasons (security, musl, openrc, apk). How viable is this idea considering that Alpine seems mostly focused on container workflows or servers? 2023-07-06 12:15:34 alternateved: there are numerous users (including me) that use alpine on the desktop 2023-07-06 12:15:35 I would also rather consider going for Edge branch as software availability seems a bit better and I generally prefer rolling releases. 2023-07-06 12:15:58 Alpine itself won't hold your hand to install it, but it's not too difficult 2023-07-06 12:17:38 ikke: Ah, that is great to hear. Is there anything worth knowing? I don't mind getting my hands dirty and wouldn't mind to learn how to package things (did that already for a couple of distros). 2023-07-06 12:18:07 Things like gnome, kde etc are already packaged 2023-07-06 12:18:39 Learning to package things is certainly worth-wile and also reachable 2023-07-06 12:19:38 I myself use a very minimal desktop setup (x11 + awesome), so not that experienced getting a full desktop setup 2023-07-06 12:19:48 although packaging gnome or kde as a first contribution is... ambitious 2023-07-06 12:19:56 Hello71: certainly 2023-07-06 12:19:58 s/ is/ would be/ 2023-07-06 12:20:08 I plan to go with Sway, mostly. Seems it is a bit better target for Alpine. :) 2023-07-06 12:20:19 Yeah, and more future proof 2023-07-06 12:20:31 I haven't got around to switching to wayland yet 2023-07-06 12:20:35 Yeah, packaging gnome or kde seems like... a lot. 2023-07-06 12:20:46 luckily we have teams for both 2023-07-06 12:21:02 I've used it for some time with Sway and Gnome on different distributions. Not many issues there. 2023-07-06 12:21:19 It's just that I'm quite used to awesemewm 2023-07-06 12:21:29 Haha, I can get that. 2023-07-06 12:23:21 Welp, seems like an interesting adventure ahead of me 2023-07-06 12:23:43 Encryption and btrfs won't be a problem if performed manually I presume? 2023-07-06 12:23:53 No, shouldn't be 2023-07-06 12:27:07 I wonder if I'll notice if GNU coreutils are not there. I know that those could be easy installed, but maybe BusyBox is enough 2023-07-06 12:31:19 In most cases, I just install them 2023-07-06 12:31:29 on a desktop, the size difference is negligable 2023-07-06 12:32:49 True 2023-07-06 13:44:22 alternateved: I tried to leave notes around the wiki as I set up my own desktop with sway. 2023-07-06 13:44:33 So check the wiki. You'll need seatd at a minimum. 2023-07-06 13:58:11 WhyNotHugo: do you have XF86Vol keys? Do they work with Sway? Mine doesn't. 2023-07-06 13:58:50 Oh lol WhyNotHugo 2023-07-06 13:59:13 Today I've read your post which was mentioned on lemmy https://whynothugo.nl/journal/2023/02/18/in-praise-of-alpine-and-apk/ 2023-07-06 13:59:30 If you are the same person 2023-07-06 14:00:26 If yes, then thanks! Your post were very informative and your notes on wiki will surely be appreciated too 2023-07-06 14:07:40 /s/were/was 2023-07-06 14:20:13 ACTION looks down at keyboard 2023-07-06 14:20:35 malmi70[m]: I do have XF86Vol keys, and they work. 2023-07-06 14:20:46 I have them mapped to actually change the volume (I didn't remember this) 2023-07-06 14:20:56 Try using `wev` and see if they register. 2023-07-06 14:21:55 https://git.sr.ht/~whynothugo/dotfiles/tree/1a873636/item/home/.config/sway/apps.conf#L31-40 2023-07-06 14:52:04 $super+XF86AudioRaiseVolume must be a wild ride 2023-07-06 19:13:44 Hmm, it seems that there are some issues with booting with encrypted btrfs root 2023-07-06 19:14:40 alternateved: where are you stuck? 2023-07-06 19:15:06 I boot into grub rescue 2023-07-06 19:15:26 It cannot find device with UUID of unencrypted btrfs root 2023-07-06 19:15:53 And also I get an error with "unknown file system" 2023-07-06 19:16:27 Is that grub that is unable to load it? 2023-07-06 19:17:12 I think so as I land in grub rescue mode 2023-07-06 19:18:02 Hmm. I think the problem is with the encryption somehow - there is an error about UUID of already unencrypted root 2023-07-06 19:18:12 But I did not input any password for decryption 2023-07-06 19:18:54 I don't have experience with full disk encryption 2023-07-06 19:19:23 It's not fde, just root 2023-07-06 19:19:54 Two partitions /boot/efi which is unencrypted and / which is encrypted root 2023-07-06 19:20:04 'full disk encryption' usually actually means 'full partition encryption' 2023-07-06 19:21:00 Isn't true FDE the encrypted /boot and / ? 2023-07-06 19:21:11 yeah 2023-07-06 19:21:23 i do suppose so, but i know FDE is used a lot to actually mean a partition 2023-07-06 19:22:04 Yeah, lots of people understand FDE like so 2023-07-06 19:22:44 True FDE is usually slower. From what I know openSUSE does FDE. Fedora just encrypts / 2023-07-06 19:23:21 it doesn't make any difference 2023-07-06 19:23:38 In what regard? 2023-07-06 19:23:53 it's the same encryption with the same ciphers on the same disk 2023-07-06 19:24:06 how on earth would that be slower 2023-07-06 19:26:18 GRUB doesn't utilize hardware acceleration, apparently 2023-07-06 19:26:40 what grub does has no relevance to the final booted system 2023-07-06 19:26:46 unless by slower you meant slower to boot or something 2023-07-06 19:27:27 in which case probably i guess 2023-07-06 19:27:29 Yeah, slower to boot. Apologies for not being clear 2023-07-06 20:25:05 psykose: he's referring to Grub's luks and luks2 modules that decrypt wherever /boot is in order to be able to load kernel and initramfs 2023-07-06 20:25:44 i'm aware 2023-07-06 20:29:40 Yeah, an FDE-d system does indeed take longer to boot in that sense. 2023-07-06 20:30:14 Btrfs on an encrypted / should not generally be a problme, though. 2023-07-06 20:30:33 I've done atest install with FDE, using btrfs. No problem. 2023-07-06 20:30:53 alternateved: how exactly did you install? via setup-disk? if so, which options did you pass it? I'm assuming it is failing to boot just after installation - can you confirm this? 2023-07-06 20:34:38 I've managed to install it correctly just now 2023-07-06 20:34:51 I've wrongly mounted /boot during my install 2023-07-06 20:35:21 And also used unsupported hash algorithm for luks 2023-07-06 20:50:17 alternateved: so you weren't using setup-disk then? 2023-07-06 20:58:01 I was 2023-07-06 20:58:35 `setup-disk -m sys /mnt` to be exact 2023-07-06 20:58:51 Where /mnt was already partitioned and mounted etc. 2023-07-06 21:27:38 is it possible to specify which windowmanager to use in the commandline 'startx..'? 2023-07-06 21:29:31 .xinitrc `exec wm` at the bottom? 2023-07-06 21:31:29 oh so it cant be specified at the commandline? running diskless and trying to cut back on config files 2023-07-06 21:34:14 nope 2023-07-06 21:34:23 it's basically a wrapper for xinit ..stuff.. 2023-07-06 21:34:27 you can read it 2023-07-06 21:34:38 hmm okok 2023-07-06 21:34:53 and xinit $client is a config file apparently 2023-07-06 21:35:16 or 2023-07-06 21:35:17 hm 2023-07-06 21:35:19 maybe you can? 2023-07-06 21:35:19 idk 2023-07-06 21:35:47 `startx xyz` seems to do client=xyz 2023-07-06 21:35:50 so maybe that works 2023-07-06 21:36:05 try and find out 2023-07-06 21:38:47 will do 2023-07-06 22:09:12 hmmm.. the "shell"? (/bin/sh -> /bin/busybox) Are there any manual for "busybox" verion of the shell?? 2023-07-06 22:10:10 no 2023-07-06 22:10:46 ash is basically dash with a few extras 2023-07-06 22:11:27 psykose: oki, strange.. 2023-07-06 22:17:48 hmm.. "shell" is the biggest subfolder in the busybox sourcecode 2023-07-06 22:21:35 https://pubs.opengroup.org/onlinepubs/9699919799/utilities/V3_chap02.html 2023-07-06 22:52:51 ash is not very thoroughly documented 2023-07-06 22:55:05 both in code and in interface 2023-07-06 23:31:40 busybox is not very thoroughly documented 2023-07-07 08:18:21 psykose: I've probably never used it, it's just for Symmetry with super+XF86AudioLowerVolume 2023-07-07 08:18:39 dangerous game to have it mapped just for an accidental press :p 2023-07-07 08:45:33 I am making a network simulation with alpine linux. I was able to configure my ip successfully in /etc/networ/interfaces but when I type ip route the gateway address I specied in the config does not show up as the default route? 2023-07-07 08:46:23 What am I doing wrong the vm is an aplpine 3.16 and my config is https://bpa.st/35SHO 2023-07-07 08:47:05 */etc/network/interfaces 2023-07-07 08:48:03 they don't need to be netmasks 2023-07-07 08:48:36 er, thingies 2023-07-07 08:48:59 lopid, this is a subnetted network simulation 2023-07-07 08:49:30 so I have /25 or 255.255.128 or a /25 subnet 2023-07-07 08:49:43 ok 2023-07-07 08:49:48 * 255.255.128.0 2023-07-07 08:50:10 aiena: /25 is 255.255.255.128 2023-07-07 08:50:51 right 2023-07-07 08:51:14 yes sorry revising my ancient ccna and goofed up 2023-07-07 08:51:36 my qustion is why the gateway ip is not howing in with ip route command 2023-07-07 08:51:46 *showing 2023-07-07 08:52:22 what does it show? 2023-07-07 08:53:02 only 192.168.0.0/25 dev eth0 scope link src 192.168.0.2 2023-07-07 08:53:18 but not default via 192.168.0.1 2023-07-07 08:53:46 idk, /etc/network/interfaces understand "/25"? 2023-07-07 08:53:56 I have this: netmask 255.255.255.0 2023-07-07 08:54:03 it seems to undertand it 2023-07-07 08:54:04 (for /24 ofc) 2023-07-07 08:54:09 but maybe not for the gateway? 2023-07-07 08:54:16 same thoughts 2023-07-07 08:54:17 ifcong list the right subnet mask 2023-07-07 08:54:26 try to remove it from gateway 2023-07-07 08:54:35 maybe look for other 192.168 occurrences under /etc 2023-07-07 08:55:01 lopid, my config is https://bpa.st/35SHO nothing else is defined 2023-07-07 08:55:07 no dhcp static ips 2023-07-07 08:55:45 wonder if the subnet mask is immaterial for a gateway ill try without the CIDR notation 2023-07-07 08:56:06 can you at least add the gateway manually? 2023-07-07 08:56:22 its added manually in https://bpa.st/35SHO right 2023-07-07 08:56:34 you mean make the gateway accesible 2023-07-07 08:56:40 that is not manual, that is read by some other thing 2023-07-07 08:56:46 as in make it reachable 2023-07-07 08:56:54 ip route add default via … 2023-07-07 08:57:30 ok my gateway isnt live the reason why i used this other method is because the vms ips are persistant after reboot 2023-07-07 08:57:44 aiena: according to manual "gateway" wants an "address" 2023-07-07 08:57:56 and "network" wants "network_address" 2023-07-07 08:58:22 thus, "/25" after gateway address seems invalid 2023-07-07 08:58:51 aron, where is this configuration specifications 2023-07-07 08:58:59 is it for ifconfig 2023-07-07 08:59:03 why not just try it without them? 2023-07-07 08:59:22 aron, yes that fixes it 2023-07-07 08:59:22 yes, you could use the good ol' way 2023-07-07 08:59:33 great! :) 2023-07-07 08:59:49 ok need to lookup where the specs for this config file are 2023-07-07 09:00:09 aiena: I read https://manpages.org/etc-network-interfaces/5 2023-07-07 09:00:13 yep in the examples https://wiki.alpinelinux.org/wiki/Configure_Networking gateway doesnt have the mask 2023-07-07 09:00:36 logical, since the gateway must be in the same network 2023-07-07 09:00:57 so, if it would be outside your network, then, you need a gateway for it 2023-07-07 09:01:00 yes I realised that in guis on widows etc also you specify only an ip 2023-07-07 09:01:21 thanks a lot 2023-07-07 09:01:24 yw 2023-07-07 09:01:30 is there a way like in ifconfig 2023-07-07 09:01:40 to make uip addreses configured with ip command permanent 2023-07-07 09:01:44 don't dismiss the first suggestion next time 2023-07-07 09:02:27 e.g. ip addr add ip/CIDR dev eth0 2023-07-07 09:02:34 works but isnt permanent 2023-07-07 09:03:21 I think /etc/network/interfaces can do this 2023-07-07 09:04:55 yeah I think ubuntu has some other way as there is no ifconfig there by default but alpine is cool :) 2023-07-07 09:05:24 ubuntu have netplan, or what 2023-07-07 09:05:33 totally different 2023-07-07 09:06:31 forget anything based on systemd 2023-07-07 09:11:40 Yeah alpine reminds me of the older simpler buntus 2023-07-07 09:11:47 though its totally different 2023-07-07 09:18:29 thanks for your help seems to working nicely 2023-07-07 10:28:20 alternateved: you issue yesterday with Alpine and LUKS - you didn't install using setup-disk to setup the encryption, you ran "setup-disk -m sys" so you must have setup LUKS manually and that was likely the cause of your problems (i.e. you indicated using unsupported hash algos) 2023-07-07 10:35:22 minimal: I think the problem was that I've mounted my boot partition on /mnt/boot/efi instead of /mnt/boot 2023-07-07 10:35:56 After I did that, then I got an information that whirlpool is unsupported hash algorithm 2023-07-07 10:36:48 And yeah, I did setup LUKS manually. Is it possible now for setup-disk to perform encryption? If so, great! 2023-07-07 10:36:56 your ESP (UEFI) partition? or a separate "boot" partition? Also which options did you use when you ran "cryptsetup luksFormat"? 2023-07-07 10:37:40 yes encryption support was added to setup-disk some time ago (Alpine 3.17 release maybe?) 2023-07-07 10:40:35 alternateved: did you specify "--type luks1" when you ran "cryptsetup luksFormat"? 2023-07-07 10:43:55 My ESP partition. I decided to not have another separate /boot. But I've made the mistake as /boot was then on encrypted / 2023-07-07 10:44:35 minimal: Would I need luks1 since I didn't intended to perform FDE? 2023-07-07 10:45:22 what do you mean? with luksFormat you either create a v1 or a v2 LUKS format, and I believe luks2 is the default if not specified 2023-07-07 10:45:39 s/not specified/none specified/ 2023-07-07 10:46:30 I think luks1 (or luks2 with pbkdf2) should be used with encrypted /boot 2023-07-07 10:46:45 the point that I'm getting to is that luks2 doesn't work with Grub for an encrypted /boot 2023-07-07 10:46:55 Yeah, I wasn't doing that 2023-07-07 10:47:04 it is a Grub issue which will be fixed in the next Grub release (whenever that happens) 2023-07-07 10:47:52 Anyway, that documents my installation process: https://paste.sr.ht/~alternateved/98f30a08514a3b9c260ae2da28d91648a520d466 2023-07-07 10:48:04 well if you had /boot encrypted (which you said you did) and you didn't specify "--type luks1" for luksFormat then you were doing that.....which is likely why it didn't then boot 2023-07-07 10:48:55 Yeah 2023-07-07 10:49:12 I've fixed that now and everything works properly 2023-07-07 10:49:38 Now I'm battling a couple of other small things, but this is mostly seatd/elogind/sway related 2023-07-07 10:50:06 I guess I was tired already yesterday and I did not notice that I was mounting /boot wrongly 2023-07-07 10:51:22 there's nothing necessarily wrong with using Grub and "FDE", you just need to ensure you don't use LUKS v2 - setup-disk ensures v1 is used 2023-07-07 10:52:39 It might be even better from security standpoint (protects from evil maid attacks) but it isn't necessary for my threat model 2023-07-07 11:14:14 qemu (both sdl and gtk ui) seems to do the wrong thing with input for wayland scale>1. looking at https://gitlab.com/qemu-project/qemu/-/blob/master/ui/sdl2.c , it's not "hidpi" aware, but in my understanding sdl2 shouldn't enable that without an explicit flag (see https://wiki.libsdl.org/SDL2/SDL_GetWindowSizeInPixels), so i shouldn't be seeing any ill effects. any considerations on the matter? 2023-07-07 12:58:02 https://pkcaty.cc/invite/i=1534 2023-07-07 13:27:18 don 2023-07-07 13:28:00 don't click it, its content seems to be illegal for about any country, bless curl 2023-07-07 13:29:45 "the point that I'm getting to is..." <- systemd-boot? 2023-07-07 13:30:38 "It might be even better from..." <- Is it even possible to protect from evil maid attacks when the device can just be replaced with one designed to capture your password? 2023-07-07 13:31:54 you can protect from that if you have a non-replaceable device, e.g. when you can create a handshake/protocol that only the original system can answer 2023-07-07 13:32:00 idkrn[m]: what does systemd-boot have to do with this? I was referring to Grub+LUKS2 2023-07-07 13:32:41 minimal: you said it doesn't work with grub 2023-07-07 13:33:13 for a start systemd-boot is not packaged for Alpine (though both gummiboot, it's predecessor, and stubbyboot, derived from systemd-boot, are packaged) 2023-07-07 13:33:49 idkrn[m]: yes I said Grub doesn't currently work for LUKS2 encrypted boot partitions 2023-07-07 13:34:51 satmd: I hope this ends up being simple 2023-07-07 13:36:49 I remember that Grub works with LUKS2 encrypted boot partitions but only with PBKDF2 2023-07-07 13:38:22 idkrn[m]: actually it's not, because it requires a non-clonable entity which is able to verify everything being unmodified. 2023-07-07 13:38:49 alternateved: nope, Grub 2.06 was *supposed* to work with LUKS2 encrypted boot partitions but there were several bugs regarding this - they've been fixed in Grub's git and so the next release (2.11) will work, whenever they get around to releasing it 2023-07-07 13:38:55 a TPM goind measurements is sufficient for many things, but - if someone can clone it - then you have lost 2023-07-07 13:40:29 alternateved: some other distros may have applied patches to fix Grub & LVM2-encrypted-boot 2023-07-07 13:40:35 s/LVM/LUKS/ 2023-07-07 13:40:57 minimal: Oh? I was almost sure that lately I've installed openSUSE with encrypted boot and LUKS2 with PBKDF2 2023-07-07 13:41:22 alternateved: yes I believe SuSE have patches added 2023-07-07 13:41:24 Maybe they did patch it there considering that their setup heavily depends on encrypted /boot 2023-07-07 13:41:40 Good to know 2023-07-07 13:42:07 I'm referring to the standard upstream Grub 2.06 release 2023-07-07 13:42:52 I get it know. I didn't follow what was going on with upstream 2023-07-07 13:43:39 That makes me wonder if there is any good Grub replacement in the works. 2023-07-07 13:43:58 there are alternatives, depending on what you want to do 2023-07-07 13:44:14 I know of systemd-boot (which is kinda out of the question here) and there is also refind, but I've never used it 2023-07-07 13:44:25 syslinux, limine, EFI stub booting, etc 2023-07-07 13:45:01 Any of those support BIOS also? 2023-07-07 13:45:39 syslinux seems so 2023-07-07 13:46:39 syslinux and limine 2023-07-07 13:47:08 whilst syslinux can support UEFI booting I'd strongly recommend not trying to use it for that 2023-07-07 13:47:55 Why so? 2023-07-07 13:48:16 multiple reasons 2023-07-07 13:48:41 been there, done that, won't do it again 2023-07-07 13:48:58 Haha, right. I'll keep it in mind 2023-07-07 13:49:06 What are you using generally? 2023-07-07 13:49:22 syslinux, grub, limine, EFI Stub 2023-07-07 13:50:24 I build Alpine images using all of those 2023-07-07 13:50:26 So it depends on what you need for that particular system 2023-07-07 13:50:57 I create disk images for physical machines, VMs, and Cloud use and test all 4 of those 2023-07-07 13:53:28 How would you rate your experience with limine? 2023-07-07 13:54:20 Luks2 2023-07-07 13:54:35 fine in general, haven't yet tried setting up hybrid booting with it 2023-07-07 17:20:47 "idkrn: actually it's not..." <- Could you do it the same way phones protect their keys from being bruteforced. 2023-07-07 21:13:37 ACF? anybody who use it? its some Alpine version of webmin. I cant see how to cobine it with LAMP 2023-07-07 21:29:57 idkrn[m]: tpms do that, but you need to trust the hardware to be inherently secure. 2023-07-07 21:30:15 it's up to everyone how much you trust in hardware to be bug- and backdoor free 2023-07-07 21:30:36 but I think we're leaving the realms of alpine here 2023-07-07 23:24:32 "idkrn: tpms do that, but you..." <- TPMs aren't as good as what iPhone's and Pixels have from what I understand 2023-07-07 23:24:45 * In reply to @_oftc_satmd:matrix.org 2023-07-07 23:24:45 TPMs aren't as good as what iPhones and Pixels have from what I understand 2023-07-07 23:24:45 idkrn: tpms do that, but you need to trust the hardware to be inherently secure. 2023-07-08 00:22:55 Hi, I have installed the tacacs+ng packet and did testing with the package. 2023-07-08 00:23:37 It works everything out. And it runs. However there is a bug, I tried out against head and do not have that "issue". 2023-07-08 00:24:13 So my question is, should I open a bug report for a community package? Or how to force a update? 2023-07-08 00:24:18 what is the bug 2023-07-08 00:26:32 So, using that particular version I get the promt on the client, but w/o the banner (first issue) then if trying to AUTH then no matter what I do I get this... 2023-07-08 00:26:38 https://dpaste.org/mOTeJ 2023-07-08 00:27:02 I thought there is some protocol change done, but not. read the TACACS RFC 2023-07-08 00:27:17 Well, long story short. 2023-07-08 00:27:36 did gitclone of current head, and build locally 2023-07-08 00:28:03 And well. Using same tacplus config in debug mode on HEAD 2023-07-08 00:28:10 I get the welcome banner 2023-07-08 00:28:22 and actually not that illegal packet message 2023-07-08 00:28:47 Well I can open a bug, it is not a problem 2023-07-08 00:29:11 If you need the data, I have it here. It did cost me some amount of time. 2023-07-08 00:29:45 well but that is not the point, I would like you guys to update that package in the end. 2023-07-08 00:31:26 I am testing using IOS components in GNS3. 2023-07-08 00:31:31 an issue sounds fine 2023-07-08 00:35:28 I'd say it is a bug in the version shipped, but noone ever tested. 2023-07-08 02:11:37 https://gitlab.alpinelinux.org/alpine/aports/-/issues/15103 2023-07-08 02:12:37 IDENTIFY 2023-07-08 02:12:41 HELP 2023-07-08 02:13:43 psykose, is that enough for a bug report? 2023-07-08 02:14:22 and that you update that package soon? 2023-07-08 02:14:39 this would be pretty cool 2023-07-08 02:15:02 is the only difference between standard and extended the packages that are installed on the live system? like will both of them install alpine to disk the same way? 2023-07-08 02:22:34 thanks for taking care of that :) 2023-07-08 02:39:59 psykose, thanks. 2023-07-08 02:43:07 cya o/ 2023-07-08 09:33:30 Hi 2023-07-08 09:35:06 I'm using default busybox syslog, but it seems.ot isn't logging all kernel output visible in dmesg. Is there a way to adjust this? 2023-07-08 09:35:42 probably some klog options 2023-07-08 09:36:20 Hm 2023-07-08 09:36:59 I tried a simple echo h >/proc/sysrq-trigger but it doesn't get logged to messages or dmesg log files 2023-07-08 09:37:07 klog.? 2023-07-08 09:38:22 Syslog has this option: -l N Log only messages more urgent than prio N (1-8) 2023-07-08 09:39:32 I guess 1 is lowest.? 2023-07-08 09:43:17 klogd service thing 2023-07-08 09:47:47 Hm nope. Didn't help 2023-07-08 09:49:46 Oh it did 2023-07-08 09:50:04 :) thanks 2023-07-08 09:50:34 Trying to debug some kernel issues in some vms :) 2023-07-08 12:07:00 What's the advantage of gpep517 over "build" (i.e.: https://github.com/pypa/build ) 2023-07-08 12:07:56 easier to bootstrap 2023-07-08 12:08:16 https://github.com/projg2/gpep517 2023-07-08 12:51:50 Easier in what way? It uses the same build system and doesn't have less dependencies. 2023-07-08 13:40:41 asked in #linux too, do you think usb cabel 7m long (including a unpowered usb hub) is a problem for keyboard/mouse? 2023-07-08 13:43:55 frag: USB 2.0 cable I'm assuming - unless it is an "active" extension cable then that's out-of-spec AFAIK as max is 5m 2023-07-08 13:45:43 hmm.. 2023-07-08 13:48:16 frag: perhaps searching online for "max usb cable length" would have provided that info 2023-07-08 13:49:31 yeah, gods know if its the right info tho ... 2023-07-08 13:50:17 does /etc/crypttab work on alpine? I've tried adding an entry and rebuilding my initramfs, but the device isn't unlocked on boot 2023-07-08 13:50:23 well the 1st result I found gives very comprehensive information on the subject 2023-07-08 13:51:39 eddsalkield: for encrypted rootfs you don't use /etc/crypttab, you pass settings via cmdline 2023-07-08 13:52:19 so does the encrypted device contain your rootfs or not? 2023-07-08 13:53:04 no, it's a non-root partition that I'm trying to mount 2023-07-08 13:53:33 eddsalkield: they it won't be mounted by initramfs, it only mounts rootfs 2023-07-08 13:53:58 instead you setup /etc/crypttab and enable the relevant init.d service so it is done during normal part of boot 2023-07-08 13:54:22 thanks 2023-07-08 13:55:06 /etc/init.d/dmcrypt 2023-07-08 13:58:19 actually that seems to be using /etc/conf.d/dmcrypt rather than /etc/crypttab for details of what to mount 2023-07-08 14:25:39 seems like /etc/init.d/dmcrypt is broken when using cryptsetup in bitlocker compatibility mode 2023-07-08 14:25:41 https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/main/cryptsetup/dmcrypt.initd#L74 2023-07-08 14:26:03 isLuks subcommand fails in this case, even if cryptsetup can open the device 2023-07-08 14:28:27 eddsalkield: then raise an Issue in gitlab for it - I doubt many Alpine users are using bitlocker compat mode 2023-07-08 14:35:40 Something broke in a recent (2 days ago) upgrade on edge: the fonts / unicode rendering in urxvt is broken 2023-07-08 14:42:24 can confirm, happens for me too 2023-07-08 14:49:27 busybox-binsh/ash is 8KB and dash 144KB? 2023-07-08 14:52:03 frag: can confirm, I am 8KB 2023-07-08 14:52:29 (I should really turn off pings on my in this channel) 2023-07-08 14:52:44 reggiere[m]: you are appearing as 'reggiere' here 2023-07-08 14:52:47 s/on my/on my name/ 2023-07-08 14:53:12 ikke: huh, im using matrix bridge, I guess it did not update my name here 2023-07-08 14:53:19 cause its supposed to be "ash" 2023-07-08 14:53:26 I figured 2023-07-08 14:55:43 big difference ... 2023-07-08 15:26:59 can 'apk search' list packages by package name only (not version)? 2023-07-08 15:27:56 Afaik, no 2023-07-08 16:50:13 frag: apk -q search 2023-07-08 16:51:16 or apk search -q 2023-07-08 16:57:43 oh, I confused it with list 2023-07-08 16:57:52 apk list doesn't get affected by -q, but search does 2023-07-08 17:04:32 uuidNuniq: ty :) 2023-07-08 17:05:28 minimal: I've written a patch here https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/48690 2023-07-08 17:18:49 eddsalkield: look good. BTW are you using util-linux blkid rather than busybox's blkid? 2023-07-08 17:19:00 it has just been merged 2023-07-08 18:20:20 WhyNotHugo: it doesn't use py3-packaging->(..) or need anything to build itself 2023-07-08 18:21:03 ikke: thanks! 2023-07-08 18:21:09 also it makes .whl's without compression so it's a little faster 2023-07-08 18:21:19 what's the process for requesting a package get moved from testing to community? 2023-07-08 18:21:26 at the time there were more -build deps too 2023-07-08 18:21:38 eddsalkield: create an issue or merge request 2023-07-08 18:36:38 ikke: in what way is the font rendering broken 2023-07-08 18:43:14 psykose: shows blocks 2023-07-08 18:43:24 i can't reproduce at all for some reason 2023-07-08 18:45:42 https://imgur.com/a/p3PrlJa 2023-07-08 18:46:17 what font do you have configured 2023-07-08 18:46:27 terminus 2023-07-08 18:46:42 like the exact Urxvt.font: line 2023-07-08 18:46:54 Was looking it up 2023-07-08 18:47:18 URxvt.font: xft:Terminus:size=14 works for me 2023-07-08 18:47:29 URxvt.font: -xos4-terminus-medium-r-normal--16-160-72-72-c-80-iso10646-1 2023-07-08 18:48:19 hm 2023-07-08 18:48:22 that doesn't even load for me 2023-07-08 18:49:18 Setting it to what you provided doesn't change it 2023-07-08 18:50:10 strange 2023-07-08 19:07:28 psykose: Oh, right. There's a circular dependency issue there, isn't there? 2023-07-08 19:08:00 yes, packaging->parsing both need a pep517 builder to build 2023-07-08 19:08:11 but if build it the pep517 builder.. 2023-07-08 19:08:21 it's not that hard to "work around", since you can fetch the wheel from pypi 2023-07-08 19:08:30 and then either just -m installer it or just manually unzip it into the correct place 2023-07-08 19:08:34 it's really ugly tho 2023-07-08 19:09:01 so only py3-installer and py3-setuptools get the ugly workarounds 2023-07-08 19:09:12 as small as i could make it 2023-07-08 19:11:12 Makes sense. 2023-07-08 19:12:30 all the pep517 builder stuff is a treasure trove of circles 2023-07-08 19:12:34 Now that everything is self-hosting, I guess you can stop downloading the wheel.. 2023-07-08 19:12:41 well, nmo 2023-07-08 19:12:47 every release is bootstrapped from scratch 2023-07-08 19:12:58 and circular deps break the build ordering 2023-07-08 19:12:58 treasure trove of circles <<-- not unlike Rust? :P 2023-07-08 19:13:23 Oh, didn't know the from scratch thing. So you're fairly certain that Alpine is easy to rebuild completely from source. 2023-07-08 19:13:38 partly that, but even if it was suddenly edge only it wouldn't change anything 2023-07-08 19:13:49 the buildrepo build order doesn't handle circular deps and just gives you the wrong order 2023-07-08 19:13:58 so if you introduce a circle somewhere it just builds things the wrong way around 2023-07-08 19:14:40 one of those impossible graph problems 2023-07-08 19:14:52 it works out okay though 2023-07-08 19:14:59 circular deps are just bad in general 2023-07-08 19:15:17 rust and friends at least have the benefit of being self-contained 2023-07-08 19:15:22 just needing itself 2023-07-08 19:15:24 with setuptools, well 2023-07-08 19:15:29 it has a dep tree of like 20 vendored things 2023-07-08 19:15:35 half of which need setuptools to build 2023-07-08 19:16:01 the other backends (hatchling, poetry-core, ..) all have the same issues too 2023-07-08 19:16:05 How does upstream deal with this? They just built it once and never rebuild from scratch? 2023-07-08 19:16:16 generally that's what self hosting projects do yeah 2023-07-08 19:16:20 you just keep stuff around forever 2023-07-08 19:16:27 for rust there is a c++->rust bootstrap path 2023-07-08 19:16:36 you can build mrustc and build rust 1.52 or something with it 2023-07-08 19:16:48 then 1.53->1.54->... 2023-07-08 19:16:51 one small step at a time 2023-07-08 19:17:23 java is similar (and we do the full from source build, starting at gcc6->gcc-gcj->openjdk7) 2023-07-08 19:17:44 that one is really more trouble than it's worth though and it would be nice if we just did the copying instead 2023-07-08 19:18:30 most lisps are similar, but there is ecl as an easy bootstrap path 2023-07-08 19:18:35 haskell is.. haskell 2023-07-08 19:20:28 I'm going to assume that the haskel situation is awful 2023-07-08 19:21:38 the rust approach where you freeze a specific target version and then guarantee that version N+1 can always be built with version N is probably less annoying for the maintainers than maintaining a bootstrap that works forever, but it seems annoying as the person running it :P 2023-07-08 19:23:01 https://www.joachim-breitner.de/blog/802-More_thoughts_on_a_bootstrappable_GHC 2023-07-08 19:23:35 it's a bunch of random stuff and a million builds 2023-07-08 19:26:28 Do you ever think about talks about all this knowledge that's mostly held by people who've actually bootstrapped languages onto new distros? 2023-07-08 19:30:15 me? i ain't know nothin 2023-07-08 20:10:39 Where is the grub config stored in Alpine? (running on a normal pc) 2023-07-08 20:11:06 There's no /etc/default/grub 2023-07-08 20:11:30 etc/grub.d 2023-07-08 20:12:05 I don't have that either 2023-07-08 20:12:45 oh wait, maybe Alpine selects extlinux 2023-07-08 20:13:00 syslinux 2023-07-08 20:13:33 grub is used only when using EFI or forcing GRUB install 2023-07-08 20:13:57 It seems like I have syslinux, ~2009 PC 2023-07-08 20:14:23 as for checking what bootloader is used, I would recommend actually using package manager instead of looking for configs 2023-07-08 20:15:14 or just looking at /boot 2023-07-08 20:15:29 Thanks for the tip, I was surprised since I've never used anything but grub 2023-07-08 20:23:51 Getting a little dizzy with dozens of tabs open as I try to setup up my 1st Alpine box on an old Atom D510 with 4GB RAM and 4x300GB SAS HDDs... 2023-07-08 20:24:49 Got help here a few days ago and finally back at it, not going to use LUKS or LVM on this one but used parted to split the drives identically into 4 partitions each: 2023-07-08 20:26:18 pangolin: You'll find default entry's in "/etc/update-extlinux.conf" if you need 2023-07-08 20:26:20 1MB for bios_grub, 999MB for boot, 4GB for swap, and the remaining 295GB for a shared btrfs volume 2023-07-08 20:27:05 I've read I don't need to mdadm RAID the 4x4GB swap partitions as linux can stripe them by itself. 2023-07-08 20:27:53 I got the btrfs 4x295GB "pool" (if I'm using that word correctly) set up so it's a 1.07TB drive. 2023-07-08 20:29:14 Before mounting anything or installing Alpine though, I want to make sure this box, which will go into a colo and does not have IPMI or other OOB features, will just keep on working if one of the 4 300GB drives fail. 2023-07-08 20:29:50 even if it would not having other oob features will eventually bite you in a colo for a million reasons 2023-07-08 20:29:54 sometimes stuff just fails a little 2023-07-08 20:30:12 So I think I have to use mdadm to "raid up" the boot partition 2023-07-08 20:30:34 I agree.. it's an old box, that's why it's on my desk now. :) 2023-07-08 20:31:26 But in my 1st rodeo years ago, I set up 2-drive RAID mirroring and was all proud of myself. Didn't realize that w/o a 3rd, everything halts on 1 drive failure. 2023-07-08 20:32:03 Bought many spare drives and never needed them -- er, I mean, **never benefitted from owning them**. 2023-07-08 20:33:05 So this time I want to be 8% smarter and use all my ancient hardware to give this old beast a fighting chance. I want to run it until it can't run no mo. :) 2023-07-08 20:33:25 :D 2023-07-08 20:33:41 . 2023-07-08 20:34:08 I'm stuck at places in guides that say "edit fstab" or mount this or that -- I don't have anything mounted at this point, I can't edit those things yet 2023-07-08 20:35:08 Thanks feuerteufel, that's what I was looking for 2023-07-08 20:35:35 I chose Alpine in search of a reproducible, simpler build and now I've done a whack of manual commands at the terminal and will lose 'em if I reboot w/o a successful install. 2023-07-08 20:35:59 I know my needs are particular to me but I can't seem to avoid building snowflake servers. 2023-07-08 20:36:15 you get a run-from-ram system by default 2023-07-08 20:37:43 To make the partitioning I did with parted "stick", I am "stuck" at https://docs.alpinelinux.org/user-handbook/0.1a/Installing/manual.html#_finalizing 2023-07-08 20:37:53 (or at least scared to continue) 2023-07-08 20:39:27 I guess I don't want to create /mnt/boot until I do this RAID bit or I'd have to do it over? 2023-07-08 20:54:14 are you working from https://wiki.alpinelinux.org/wiki/Setting_up_a_software_RAID_array or some other document for the RAID setup? 2023-07-08 20:59:51 yes that and https://raid.wiki.kernel.org/index.php/A_guide_to_mdadm#Adding_a_drive_to_a_mirror to add the 2 spares -- and I think I've got it... 2023-07-08 21:00:20 I even got the --metadata=0.90 bit just in case 2023-07-08 21:00:50 this is a bios machine not uefi and I used gpt partition tables on all 4 drives up to this point 2023-07-08 21:03:10 So I think I'm almost at the finish line, if I mount this array as /boot and the large btrfs one as my main install target... just checking how to tell setup-alpine that I have these swap partitions, or at least one 2023-07-08 21:06:28 Call for opinions: should I have made a 15GB system partition on the 4 drives and mdadm'd that for redundancy, then have 4x280GB in a btrfs pool for data instead of my current 4x295GB btrfs for os and data? 2023-07-08 21:07:37 I made that choice when I thought it would help me avoid mdadm altogether, then realized I still needed mdadm on boot to hopefully increase my time between data centre visits. :) 2023-07-08 21:07:48 that doesn't seem particularly better or worse to me than what you are already doing 2023-07-08 21:08:25 But btrfs gives me some (all? more?) of redundancy benefits of mdadm without costing me 4x15GB right? 2023-07-08 21:08:31 Whew 2023-07-08 21:08:56 I'm not sure 2023-07-08 21:10:24 I skipped LUKS since this server will be for multisite family photo sharing/backup, files that I expressly want to be as recoverable as possible if either I or the server die. 2023-07-08 21:12:02 So it's my first Alpine practice machine but when I do a "real" server (tomorrow) for work I'll no doubt be back to ask rando q's about that added layer. 2023-07-08 21:13:14 I have a lot of bookmarks to re-read now about the layering and also the use of dropbear to allow remote unlock of LUKS volumes during boot, I'd never heard of that before this week, sounds very useful. 2023-07-08 21:14:52 Getting ahead of myself now, clearly I am stalling because I am unreasonably nervous about the next steps -- breaking now for food and a walk. Thank you all for your help! back on later 2023-07-08 22:10:57 test 2023-07-08 22:13:55 test 2023-07-08 23:10:45 hello :) 2023-07-09 00:23:53 test 2023-07-09 03:02:43 test! 2023-07-09 03:02:48 what are we testing? 2023-07-09 03:06:50 that the tests test testfully 2023-07-09 03:20:29 hehe 2023-07-09 11:24:15 groups qaq and sudo groups qaq show difference 2023-07-09 11:24:29 qaq is my user now 2023-07-09 11:25:43 Is the first command made from a session started after you modified groups? 2023-07-09 11:26:02 no, it is groups and groups qaq diff 2023-07-09 11:26:07 qaq is my current user 2023-07-09 11:26:13 Then you'll have to do that 2023-07-09 11:27:08 exit and login is work same 2023-07-09 11:27:54 Are they effectively different or just sorted differently? 2023-07-09 11:30:35 when exec groups with no argu, it shows that the added group is missing 2023-07-09 11:31:46 groups qaq shows right. i re login both are right 2023-07-09 11:32:27 So problem solved? 2023-07-09 11:32:35 yes 2023-07-09 11:33:11 :) 2023-07-09 11:34:27 maybe there is the cache for current user 2023-07-09 11:35:34 There is no cache, it just keeps the credential from you logged in 2023-07-09 11:41:22 :) 2023-07-09 12:39:44 does the mouse button 6/7 (back/forth) work out of the box for your usb mouse? 2023-07-09 12:40:13 xev registers it but for some reason it doesnt function in qutebrowser nor firefox 2023-07-09 13:00:20 ah there test123 2023-07-09 13:00:58 soo i was wondering if i could get some help in setting up an alpine cluster: 2023-07-09 13:02:01 i want to boot of PXE which requires nfs, but so does the other file sharing i need for the cluster setup, Is this gona cause problems or do i have to make some kind of special setup? 2023-07-09 14:26:33 magnenet: PXE does not require NFS 2023-07-09 14:39:24 After a full apk upgrade, I'm seeing this super annoying behavior. When GTK or Qt application open any modal / filechooser / etc, they "jump" the mouse cursor onto the centre of the new window. 2023-07-09 14:39:41 I don't see that the sway aport has been updated in the last few weeks. 2023-07-09 14:39:46 Any idea what could be producing this change? 2023-07-09 14:44:18 Ah, I've managed to fix the issue with bluetooth mouse. I needed to load uhid kernel module also 2023-07-09 14:53:46 Turns out it was waybar. The latest version reconfigures sway's mouse_warping for some reason. 2023-07-09 16:27:20 Reading https://wiki.alpinelinux.org/wiki/Setting_up_a_software_RAID_array and a similar page on btrfs and they say things like run rc-update so stuff happens at next boot... 2023-07-09 16:28:40 ...but I am still pre-first-reboot, as in, still running Alpine from the USB stick but *after* running setup-alpine ... I haven't done a first reboot yet, haven't hit the problems I've read about with say raid or btrfs 2023-07-09 16:29:56 How to I preempt those problems, avoid them in the first place... I can look into what my system "will be" in /mnt/etc/fstab and it looks okay but I don't think rc-update will affect stuff in /mnt will it? 2023-07-09 16:30:16 I mean it will try to affect my running Alpine instance (the USB stick, local RAM) 2023-07-09 16:31:03 should I be doing some "chroot" step I feel like I saw 100 browser tabs ago but can't find now? 2023-07-09 17:25:34 psykose: You also use waybar? 2023-07-09 17:25:37 nah 2023-07-09 17:25:48 Do you use any bar? 2023-07-09 17:26:05 default swaybar 2023-07-09 17:31:49 Do you use seatd or elogind with Sway? 2023-07-09 17:32:17 former 2023-07-09 17:32:36 And how you manage polkit? 2023-07-09 17:32:51 Do you just the $USER to necessary groups? 2023-07-09 17:32:51 i don't use polkit 2023-07-09 17:33:24 Honestly, I'm having a hard time with the seatd/elogind/polkit thingy 2023-07-09 17:34:09 it's something like `apk add util-linux-login elogind polkit-elogind` and rc-update add of elogind+polkit, then reboot 2023-07-09 17:34:11 that's 90% of it 2023-07-09 17:35:15 The other 10% is the other 90% of the effort 2023-07-09 18:15:32 add seatd, start the service, add your to the seat group 2023-07-09 18:56:16 I'm also using seatd. I don't use polkit, polkit needs either systemd-logind or elogind running. 2023-07-09 18:57:23 I'd love to be able to work around polkit for using a fingerprint reader tho. 2023-07-09 18:57:23 Yeah, seatd seems like an easier solution 2023-07-09 18:57:53 fingerprint reader would work with swaylock, righ? 2023-07-09 18:57:59 s/righ/right 2023-07-09 18:58:08 I hope it would, but I haven't gotten around to making it work. 2023-07-09 18:58:55 psykose: for swaybar you have a single script that stitches all the blocks together into a JSON line? 2023-07-09 19:00:04 i use i3status-rs for the status on the right 2023-07-09 20:10:01 hi 2023-07-09 20:34:51 you can probably patch seatd or swaylock to get the fingerprint reader to work 2023-07-09 20:40:23 you can probably write/find a pam module for fingerprint auth 2023-07-09 20:40:36 no polkit needed 2023-07-09 20:42:19 but what if hackers steal your fingers 2023-07-09 20:43:13 that's why you have 2fa: fingerprint and retina 2023-07-09 20:43:19 then they have to steal your eyes too 2023-07-09 20:43:52 oh noes 2023-07-09 20:44:24 you need a strong passwd 2023-07-09 20:44:25 They could steal your nose too 2023-07-09 20:44:27 like hunter3 2023-07-09 20:47:47 or use your toeprints, they'll never see it coming 2023-07-09 20:49:36 stallman is that you 2023-07-09 20:54:03 but toes are stealable 2023-07-09 21:26:39 use a soulreader because if they steal your soul all is lost anyway ;-) 2023-07-09 21:56:54 now this is getting philosophical 2023-07-10 02:23:42 Hello. Does anyone know why package 'unrar' is not available after Alpine 3.14? 2023-07-10 02:23:52 I have some .rar files and do not know how to extract them now. 2023-07-10 02:24:38 Wouldn't the package be 'rar'? 2023-07-10 02:24:58 https://gitlab.alpinelinux.org/alpine/aports/-/issues/13525 2023-07-10 02:25:10 If you install zip, you should get unzip with it right? 2023-07-10 02:25:26 idkrn[m], rar was weird in this specific way 2023-07-10 02:25:37 try 7z 2023-07-10 02:25:46 ah the url from opcode mentions that 2023-07-10 02:25:51 Yep 2023-07-10 02:26:14 community/py3-rarfile/APKBUILD:checkdepends="py3-pytest unrar libarchive-tools" 2023-07-10 02:26:17 Habbie: Insecure :) 2023-07-10 02:26:18 looks like this might be broken now 2023-07-10 02:26:23 idkrn[m], what? 2023-07-10 02:26:27 Habbie: Very weird 2023-07-10 02:26:37 what 2023-07-10 02:26:41 Habbie: 7zip is scary 2023-07-10 02:27:04 Habbie: my replies might look weird to you on irc 2023-07-10 02:27:14 i am aware of matrix 2023-07-10 02:27:26 I didn't say you weren't 2023-07-10 02:31:00 ~ $ bsdtar xf @en-us_windows_server_2022_updated_june_2023_x64_dvd_11620906.iso.rar 2023-07-10 02:31:03 bsdtar: Declared dictionary size is not supported. 2023-07-10 02:31:20 Looks like not all RAR achives are compatible. 2023-07-10 02:34:55 why did someone make a .rar of a windows iso 2023-07-10 02:35:07 this might sound aggressive but they might need to get their head checked 2023-07-10 02:38:15 These people have nothing on their head other than ms-office, winrar, adobe... etc. 2023-07-10 02:38:40 When I send a simple .tar file to them, they have no clue how to open it, lol. 2023-07-10 02:41:57 even with that i mean the best compression you're really going to get is some lzma2 thing and 7zip can do that fine 2023-07-10 02:42:05 putting it in a rar doesn't have any benefit 2023-07-10 02:42:33 i guess i do still see .r00 in current year though so i think people do it just to be funny 2023-07-10 02:50:57 I tried to use the command '7z x' to extract it, and ended up with the error 'Cannot open the file as archive'. 2023-07-10 02:51:44 I copied the files to a colleague's Windows PC and he uses 7-Zip (Windows version) to extract it just fine. 2023-07-10 02:52:17 So, Linux version of 7-Zip is different from Windows counterpart? 2023-07-10 03:29:06 the rar support of 7zip/p7zip is the same as unrar 2023-07-10 03:29:22 i.e. it's the same sdk just copied in there and shimmed in 2023-07-10 03:29:29 so it's disabled for the same reason unrar doesn't exist 2023-07-10 09:06:15 Anyone who uses apparmor on Alpine? I've set it up and it works, but I'd need a good source of profiles to make it more useful 2023-07-10 09:08:00 There was someone recently who created a bunch of profiles. You can find them here: https://gitlab.alpinelinux.org/users/eex0/snippets 2023-07-10 09:08:04 I have no idea how good they work 2023-07-10 09:11:09 Thanks, but I get "No snippets found"? 2023-07-10 09:15:39 Maybe I should try generating them myself, it's mostly for Firefox 2023-07-10 09:17:31 Oh sorry, they're private 2023-07-10 12:01:09 Hey! 2023-07-10 12:02:15 Can someone help me figure out something with release 3.18.x? 2023-07-10 12:02:28 actuallynotgelb: it works best to just ask your question 2023-07-10 12:02:46 I've installed it on my netbook (I know the hardware is wildly out of date, and bad, but I want to experiment with it), and 3.18 can't see all my memory 2023-07-10 12:02:52 3.17 and other distros don 2023-07-10 12:03:06 don't have that issue* they see all 1.5GB, not just 8 to 900MB 2023-07-10 12:03:34 it's got an Intel Atom N270, so it's i686 only 2023-07-10 12:03:41 What kind of hardware? 2023-07-10 12:04:29 it's one of these, let me grab the model number: https://upload.wikimedia.org/wikipedia/commons/thumb/c/ce/Guang_Hua_Digital_Plaza_Launch_Acer_Aspire_One_WhiteBackground.jpg/800px-Guang_Hua_Digital_Plaza_Launch_Acer_Aspire_One_WhiteBackground.jpg 2023-07-10 12:04:57 Intel Atom N270, 1.5GB of ram and a 160GB HDD, with 945GSE graphic (I believe.. I could be wrong, tho) 2023-07-10 12:04:58 might be related to ea95865daefe76f6fa92806f747e98e829320d32 2023-07-10 12:05:35 perhaps, is there anything I can do about it? 2023-07-10 12:05:51 currently I'm stuck with 3.17, or something like Void glic 2023-07-10 12:05:57 glibc* 2023-07-10 12:07:17 I don't think my hardware can build the kernel very fast, since my Ryzen 7 laptop was sent the warranty and is yet to arrive, so I'm stuck with a 3rd gen i5, but I can try to build the kernel reverting that commit, and then giving it a try on the desired hardware 2023-07-10 12:07:40 I've never build a kernel for Alpine, but should be find - if that's the issue, I can just file a bug report, no? 2023-07-10 12:07:51 yes 2023-07-10 12:08:13 has anyone else complained about this issue, btw? 2023-07-10 12:08:23 I remember one other report 2023-07-10 12:08:43 No issue afaik 2023-07-10 12:11:07 alright, I'll try to get it working, and file a bug report if possible 2023-07-10 12:12:15 abby: All the fprintd stuff depends on polkit (even if you're using it as root). 2023-07-10 12:12:19 And the pam module uses that 2023-07-10 12:12:27 Any recommendations for using Linux on an ancient netbook? Perhaps the switch to Wayland and Pipewire would be beneficial on decrepit hardware, or perhaps something like JWM would be better? 2023-07-10 12:12:45 psykose: Hi. Wanted to feedback that scsi-tgt now works properly! Really good stuff. Thank you very much 2023-07-10 12:12:59 Forza: nice to hear 2023-07-10 12:13:19 I want to make a sort of install script (preferably idempotent and easy to use), to turn Alpine into the ideal plug and play distro for really slow netbooks and machines alike 2023-07-10 12:13:39 I do miss a restore config option. Maybe it could be a separate init.d service 2023-07-10 12:14:11 actuallynotgelb: probably won't be too difficult. It's already rather minimal :) 2023-07-10 12:14:35 I've recently updated my system and somehow urxvt stopped understanding utf8. Anyone encountered this? 2023-07-10 12:14:44 CcxWrk: yes 2023-07-10 12:14:51 I had issues setting audio up, but seems to be an issue with the card I have, and overall I don't know what's the best possible desktop / window manager / wayland compositor 2023-07-10 12:15:06 CcxWrk: it's still unclear why 2023-07-10 12:15:08 `tgt-admin -e -c /etc/tgt/target.conf` is what's needed. 2023-07-10 12:15:12 any performance benefits are appreciated in that case, but a clean and polished look should also be important 2023-07-10 12:15:31 I love one JWM setup I saw on unixporn multiple times, might take his dot files and see where I land 2023-07-10 12:15:41 let me source it 2023-07-10 12:15:43 actuallynotgelb: best is subjective unless you define a criteria 2023-07-10 12:15:45 ikke: *sad noises* is there bugreport to follow at least? 2023-07-10 12:16:09 No, would you mind creating one? 2023-07-10 12:16:11 best in terms of consistency, ease of use, stability, performance, and customizable enough to look half decent 2023-07-10 12:16:27 most of those are controlled by me, however 2023-07-10 12:16:48 but I have to take into consideration what's already available, even if that requires some additional building 2023-07-10 12:17:30 https://i.redd.it/jwm-my-college-setup-just-turned-one-year-old-takes-up-250mb-v0-5wz6l67nw5za1.jpg?s=b93ebc7fada37cc5195bee9db1879b59b873973a 2023-07-10 12:17:37 it's this beauty 2023-07-10 12:17:37 Ugh gitlab. I probably need to register first. 2023-07-10 12:17:53 You can use an existing gitlab or github account if you have one 2023-07-10 12:17:59 https://i.imgur.com/P7OYbiu.jpg imgur if you prefer 2023-07-10 12:18:54 do any of you have tips for setting up a wm? I've done it in the past, but nothing very complete or usable in the real world if I'm honest, I'm always afraid I'll forget to set something important up 2023-07-10 12:19:01 wm or wayland compositor 2023-07-10 12:21:25 No. I do openbox on my servers 2023-07-10 12:21:44 But otherwise gnome desktop which I think uses xwayland 2023-07-10 12:22:23 only on Alpine, learnt that today (why?) 2023-07-10 12:22:34 Uses about 450MiB ram with qbitrorrent running 2023-07-10 12:22:47 Yes on Alpine 2023-07-10 12:22:48 why does my gnome shell depend on xwayland to be running 2023-07-10 12:22:52 only Alpine seems to do that 2023-07-10 12:23:12 Don't know. Smaller, more stable? 2023-07-10 12:23:38 is it? 2023-07-10 12:24:04 Not sure 2023-07-10 12:24:17 Haven't had a problem with it so far 2023-07-10 12:24:24 bcuz I can't kill xwayland without it killing my whole desktop 2023-07-10 12:24:31 I've never seen that, I didn't even think it was possible 2023-07-10 12:54:18 ikke: https://gitlab.alpinelinux.org/alpine/aports/-/issues/15107 2023-07-10 12:54:45 CcxWrk: thanks! 2023-07-10 12:59:49 If you need any details for debugging just ping me (ideally here); though I assume this one is easily reproducible. 2023-07-10 13:00:12 I can reproduce it myself, but I get blocks instead of question marks 2023-07-10 13:01:58 Maybe you can also share what your urxvt.font setting is 2023-07-10 13:06:03 ikke: In the ticket already ;−) 2023-07-10 13:06:10 thanks 2023-07-10 13:07:00 Really the "tofu" or "questionmark" distinction is probably what the unknown character is in the font in question. 2023-07-10 13:08:14 yea 2023-07-10 13:08:48 ikke, psykose : no more kernel BUG: when iscsi clients reboot using tgt. Tarhetcli/LIO was not very stable :( 2023-07-10 13:11:59 I mean the bug happens with LIO, not with tgt 2023-07-10 15:13:32 I wonder if this is an upstream problem. It seems tgtd requires /usr/lib/tgt/backingstore to exists. Or maybe the installation script ought to create it. 2023-07-10 16:14:43 Forza, does your computer lock when it goes to sleep, on GNOME? Mine doesn't, I noticed. 2023-07-10 16:14:55 Oh, I think it's just missing the screensaver, let me check 2023-07-10 16:20:45 oh, it was just a setting 2023-07-10 16:20:49 https://i.imgur.com/muC1vGP.png 2023-07-10 16:22:27 I usually disable it :) 2023-07-10 17:33:30 I´m never using irc for anything again... o-o 2023-07-10 17:34:16 cool 2023-07-10 17:34:44 resphantom: you just did, you used it to make that comment... 2023-07-10 17:35:08 Took way too long to find a working client 2023-07-10 17:35:26 Or maybe I´m just dumb 2023-07-10 20:52:05 irssi and weechat both work fine 2023-07-11 00:00:45 Hi, I'd like to deploy alpine to Digital ocean. Anybody done this? 2023-07-11 00:01:16 (are there any [un]official alpine disk images I can upload to DO?) 2023-07-11 00:09:42 "(are there any [un]official..." <- Where have you looked so far 2023-07-11 00:10:09 I'm chatting with minimal in a PM about it, we might have a solution soon :D 2023-07-11 00:11:44 for context, it appears DO only accepts disk images that have cloud-init set up, and perhaps some other requirements, so the official release media isn't usable for that deployment 2023-07-11 04:39:49 a little bit confused on the setup scripts during installation, am i supposed to run those on the live iso, or on my installed system after running setup-disk? i ran the setup scripts on the live iso and then did setup-disk, but then i had a bunch of services that weren't set up to be started on my installed system 2023-07-11 04:46:29 like for example networking and lvm are both not "added" to openrc, but networking starts anyways somehow, and lvm doesn't but it works fine 2023-07-11 04:52:20 you run `setup-alpine` (or whatever) in the live system, and it asks you where you want to install stuff to 2023-07-11 04:58:45 i ran the setup scripts manually with the semi-automatic install guide, since i wanted my own partition layout. the reason i was thinking i should run setup-alpine on the installed system is because the default MOTD says to do so 2023-07-11 05:00:32 i see why my system works still though even though some of those services aren't enabled, they're under the "needed/wanted" runlevel, which i can only assume means they were required by a different service (there appears to be literally no documentation about this runlevel anywhere) 2023-07-11 05:01:34 not sure how LVM isn't broken though even though the lvm service isn't running 2023-07-11 05:03:23 what's the point of setup-alpine setting a bunch of services to run on boot when it's being run on the live system, so it won't persist? 2023-07-11 06:46:17 hello! I'm having a beef with a colleague who does 'RUN apk add tzdata' in our cross-compiled runtime image; somehow, this works on his Docker Desktop. but on my Linux, this causes a problem because 'apk' is a shell script, and /bin/sh is not compiled for my architecture. I have no idea why his Docker lets him run arm64 executables, but it's either a QEMU thing or a Windows thing. 2023-07-11 06:47:49 so... after searching around, I saw someone do the 'RUN apk add --no-cache tzdata' on the builder image, and then COPY --from=builder /usr/share/zoneinfo /usr/share/zoneinfo 2023-07-11 06:47:56 does anyone have an opinion on the feasibility of that? 2023-07-11 06:49:51 apk is not a shell script 2023-07-11 06:50:06 well, in that case, 'apk' was compiled for arm64. 2023-07-11 06:50:17 RUN is spawning a shell 2023-07-11 06:50:24 yea 2023-07-11 06:50:28 the copy should mostly be ok 2023-07-11 06:50:31 okay, that's why I get the '/bin/sh: apk: exec ... error' :) 2023-07-11 06:51:11 I'm just wondering if copying /usr/share/zoneinfo is enough; I know I could be touching /etc/localtime (either symlink or copy). 2023-07-11 06:51:26 that's separate and nothing makes that unless you do 2023-07-11 06:51:44 that said the docker buildx build stuff should be able to cross build fine relatively transparently 2023-07-11 06:52:28 yeah, cross-build. but running a cross-compiled binary in the target image on the build platform doesn't seem to work for me. I've read there's things you can do with QEMU to make that work. 2023-07-11 06:53:09 maybe you're missing binfmts? 2023-07-11 06:53:24 that's all you generally need yeah 2023-07-11 06:53:41 qemu-$arch of whatever and whatever your distro has that sets up them as a binfmt handler 2023-07-11 06:54:05 o 2023-07-11 06:54:07 https://wiki.alpinelinux.org/wiki/How_to_make_a_cross_architecture_chroot <- something like this? 2023-07-11 06:54:08 oops sorry 2023-07-11 06:54:39 what distro are you using 2023-07-11 06:55:07 my docker images are all alpine 2023-07-11 06:55:12 host 2023-07-11 06:55:18 arch 2023-07-11 06:55:30 https://archlinux.org/packages/extra/x86_64/qemu-user-static-binfmt/ 2023-07-11 06:56:27 https://wiki.archlinux.org/title/QEMU#Chrooting_into_arm/arm64_environment_from_x86_64 2023-07-11 06:56:49 should just be pacman -Syu qemu-user-static qemu-user-static-binfmt 2023-07-11 06:57:06 wow, thanks a lot :) 2023-07-11 06:57:50 ah, and systemctl enable --now systemd-binfmt 2023-07-11 07:00:38 this is super valuable, thanks. I kinda knew it was possible, but thought it was too involved. 2023-07-11 07:01:58 conceptually all it is, is 'an executable' and then some magic written to /proc/sys/fs/binfmt_misc/register to map some start-of-executable-bytes to the thing to run it with 2023-07-11 07:02:04 it's a bit magic but actually very trivial 2023-07-11 07:02:19 arch has a binfmt handler for wine somewhere even for .exe's :) 2023-07-11 09:23:49 Hi, I wanted to know if Alpine Linux kernels are gonna be build with CONFIG_RUST=y. I updated my LTS kernel recently (6.1.38) and it seems that it isn't enabled. 2023-07-11 09:24:51 enabling that doesn't do anything 2023-07-11 09:25:11 It allows you to compile Rust kernel modules 2023-07-11 09:25:18 which don't exist, yes 2023-07-11 09:25:37 They do, though: https://rust-for-linux.com/ 2023-07-11 09:25:51 Still a very new thing, but looks promising 2023-07-11 09:25:56 which kernel module uses that 2023-07-11 09:26:13 (spoiler: none of them) 2023-07-11 09:26:26 once something uses it, then yeah 2023-07-11 09:26:48 There aren't many, as I said. It's a new thing. 2023-07-11 09:27:04 But kernel mode memory safety sure seems to be the way to go 2023-07-11 09:27:12 there are [zero], .. 2023-07-11 09:27:24 all of it is experimental wip stuff that you can't use 2023-07-11 09:27:27 Plus, if distros don't enable it, obviously people aren't gonna write drivers in it (since no one can use them) 2023-07-11 09:27:40 once something uses it it would be enabled 2023-07-11 09:28:00 https://github.com/Rust-for-Linux/rust-out-of-tree-module 2023-07-11 09:28:12 Very simple, fully working, if you have CONFIG_RUST=y 2023-07-11 09:28:29 yes, and that's not an actual module 2023-07-11 09:28:33 i already know all of this 2023-07-11 09:28:44 i am telling you nobody uses it for anything yet 2023-07-11 09:29:17 Yes, that's right 2023-07-11 09:29:47 But what harm would it be if you enabled CONFIG_RUST=y to allow people to use it? 2023-07-11 09:30:35 it's not as simple as making it =y 2023-07-11 09:35:03 psykose: Why? Additional build requirements? 2023-07-11 09:35:31 that and it's broken every upgrade of rust/bindgen 2023-07-11 09:37:26 haha, I like this conversation. 2023-07-11 09:51:50 reason for that is that the kernel effectively uses the unstable nightly interfaces from a specific rustc version 2023-07-11 09:52:00 maybe next kernel lts it will be usable (whatever at the end of year) 2023-07-11 09:52:12 assuming they actually backport all the needed things over time 2023-07-11 09:52:41 currently though it's not enablable in -lts because every rustc version you then have to go and figure out some thousand line patch to fix it after 2023-07-11 09:52:55 for a gain of.. maybe someone will use a hypothetical out of tree module that exists or not 2023-07-11 09:53:13 it's the worst price/reward ratio possible 2023-07-11 10:12:03 https://git.alpinelinux.org/aports/tree/testing/vaultwarden/vaultwarden.confd#n58 2023-07-11 10:12:30 I think line 58-63 can be deleted. 2023-07-11 10:14:30 then delete it :D 2023-07-11 10:16:15 the sample template still has it 2023-07-11 10:19:16 yes, the developer said templates still need to be updated for this. 2023-07-11 13:43:02 I noticed that docker-rootless-extras is only available in testing. Can I still use it in production? What would be a good way? if I add the testing repo last in my repo list will apk only install testing packages if it cant find them in latest-stable? 2023-07-11 13:43:42 generally mixing edge/testing with stable isn't supported, but that package is just a text file, so it should be ok 2023-07-11 13:44:20 that said it's also literally a text file 2023-07-11 13:44:37 you can curl it from the repo yourself into some PATH place 2023-07-11 13:48:19 hmmm thanks for the clarification, I think I prefer a repo for eventual updates. But if it's such a simple package maybe it's worth moving to community? 2023-07-11 14:06:13 hello ppl 2023-07-11 14:06:26 I have an issue with Xorg 2023-07-11 14:06:38 something that should be simple but it's driving me insane 2023-07-11 14:06:57 I need to start a X server + 1 (one) program at startup, as "kiosk" mode 2023-07-11 14:07:27 I managed to configure and run it properly from root with .xinitrc 2023-07-11 14:07:31 and it works 2023-07-11 14:07:45 how can I make it autostart at boot time ? 2023-07-11 14:07:55 tried local.d rc script 2023-07-11 14:08:16 starts X but suddenly closes and I have no idea on what is going wrong with it 2023-07-11 14:08:24 Janik[m]: I wouldn't mind seeing docker-rootless-extras in community tbh 2023-07-11 14:09:24 WhyNotHugo: Is there a place where I should raise make this suggestion? 2023-07-11 14:13:01 Assuming that the package is of acceptable quality for community, you can create and tag me in a merge request moving it from testing to community. 2023-07-11 15:22:08 anyone can explain me why a startx command into local.d script is not working ? 2023-07-11 15:22:16 any way to debug it ? 2023-07-11 16:06:17 Good day! nl80211 stopped working for me, is there any documentation i can read for troubleshooting it? 2023-07-11 16:06:46 `ioctl 0x8933 failed: no such device` for restarting networking 2023-07-11 16:07:32 and for wpa_supplicant `nl80211: Driver does not suppport authentication/association or connect commands` 2023-07-11 16:45:30 hey guys wassup 2023-07-11 16:46:13 does anybody here know if I could run Alpine on orange pi zero LTS ? 2023-07-11 16:47:34 how well do the nvidia-src (open GPU kernel modules) work on Alpine for stuff like gaming? 2023-07-11 16:47:43 they don't 2023-07-11 16:54:29 "how well do the nvidia-src (open..." <- Interesting distro choice for gaming 2023-07-11 16:57:20 wk: there is some mention for orange pi on the wiki, but not sure that specific model 2023-07-11 17:02:03 well, I looked at nouveau (so I could at least play around with Alpine on my desktop) and its feature implementation status leaves a lot to be desired 2023-07-11 17:04:51 alpine doesn't have a non-nouveau nvidia driver 2023-07-11 17:10:56 what is the purpose of the nvidia-src package, then? 2023-07-11 17:14:36 fun and games with the kernel module 2023-07-11 17:14:38 which builds fine 2023-07-11 17:14:47 a driver is two parts though and not just a kernel module 2023-07-11 17:15:07 the actual userspace driver is the same nvidia blob as always 2023-07-11 17:15:31 there's at least three distinct reasons why that one doesn't work or can't be added 2023-07-11 17:15:41 hey guys is it really that alpine will drop 32 bit support ? 2023-07-11 17:15:48 no 2023-07-11 17:17:23 I saw on Telegram that they would drop older I686 and penthium 2023-07-11 17:17:37 https://t.me/alpine_linux/1152 2023-07-11 17:18:08 already have, and "older I686 and penthium" is not "32 bit support" 2023-07-11 17:18:35 What DE is more stable on alpine gnome or kde? 2023-07-11 17:21:16 psykose: thanks for the advice , I ve got a old netbook I was thinking about using Alpine for a microserver 2023-07-11 17:21:34 which old netbook 2023-07-11 18:53:23 it is a intel atom 2023-07-11 19:00:26 that doesn't narrow it down at all 2023-07-11 19:01:39 there's a ton of atoms 2023-07-11 19:02:07 that said i think all of them except 1 have sse2 2023-07-11 19:02:11 could be misremembering 2023-07-11 22:29:02 "But kernel mode memory safety..." <- It would take decades for Linux to make any meaningful progress 2023-07-12 00:08:43 Kernels also must do unsafe things apparently 2023-07-12 08:55:44 "It would take decades for..." <- I agree. But as someone who is learning how to make kernel modules, the option to use Rust so that bad things don't happen in the kernel mode is very appealing 2023-07-12 08:57:45 Maybe there could be an experimental linux-lts-rust package or something? 2023-07-12 08:58:23 if you are developing rust modules, you'd clone the linux kernel repo and enable any option that you need 2023-07-12 08:59:54 I'll probably end up doing that 2023-07-12 09:00:52 note that using rust is not a magic fix to all security issues. 2023-07-12 09:01:26 If I decide to do that, could I make an APK called linux-lts-rust and publish it on the Alpine repos? Or are the Linux builds handled only by the main devs? 2023-07-12 09:01:47 tyler[m]12: we want to avoid too many custom kernels 2023-07-12 09:02:28 ikke: True. But at least it will avoid most common C problems 2023-07-12 09:27:53 "Assuming that the package is..." <- I never used cgit before. How can I create a merge request there? I could'nt figure it out 2023-07-12 09:28:43 Janik[m]: cgit is a read-only interface 2023-07-12 09:28:54 \you craete a merge request on gitlab.alpinelinux.org 2023-07-12 09:30:22 ah alright. Thanks! 2023-07-12 10:00:10 https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/48771 2023-07-12 10:00:10 "Assuming that the package is..." <- I created a merge request. Your username is not the same on gitlab though 2023-07-12 13:29:35 rename suprag 2023-07-12 13:30:02 excuse me 2023-07-12 14:47:17 "note that using rust is not a..." <- Almost is 2023-07-12 15:06:30 chromium renders badly again 2023-07-12 15:07:33 and rm -r .config/chromium/Default/GPUCache solved it again 2023-07-12 16:06:21 funny 2023-07-12 16:28:04 'apk add docs' will always get the doc subpackages when available. Is there a way to do that also for dev subpackages? 2023-07-12 16:30:11 no 2023-07-12 16:30:59 There is no install_if directive for -dev subpackages 2023-07-12 21:11:09 ncopa: same here 2023-07-12 21:11:26 ikke: should there be? 2023-07-12 21:13:01 doesn't sound like a good idea 2023-07-12 21:13:02 It's a very niche use case 2023-07-12 21:15:13 I doubt one would really want the dev packages for *all* their library packages 2023-07-12 21:49:58 Quick question, does alpine have any immutable-version in the works? Offical or otherwise. 2023-07-12 21:51:09 mervin: there is a guide in Alpine Wiki about DIY immutable spin of Alpine https://wiki.alpinelinux.org/wiki/Immutable_root_with_atomic_upgrades 2023-07-12 21:52:24 alternateved: thanks for the response, i saw that earlier and will likely take this route 2023-07-12 21:53:15 I was just curious if there were any bonafide versions before I started building it. As it would be easier to implement 2023-07-12 22:16:27 Alternateved: does that article really describe an immutable setup rather than a snapshotted setup though? 2023-07-12 22:50:17 "and rm -r .config/chromium/..." <- I guess it's happened to multiple distros 2023-07-13 02:37:45 @psykose Should /usr/share/webapps/vaultwarden-web/templates/ be included in the main package instead of the web-vault package? 2023-07-13 02:38:02 don't see why 2023-07-13 02:43:39 there are email templates 2023-07-13 02:46:30 hmm 2023-07-13 02:46:31 right 2023-07-13 03:29:39 if /var/lib/vaultwarden/.env works on the build by alpine? 2023-07-13 03:32:25 probably 2023-07-13 03:34:16 Ok 2023-07-13 05:07:27 the new vaultwarden push service is wonderful 2023-07-13 05:24:40 what's vaultwarden? 2023-07-13 05:25:50 Self-hostable Bitwarden server 2023-07-13 06:12:25 "Self-hostable Bitwarden server" <- https://github.com/bitwarden/server is already self-hostable 2023-07-13 06:14:43 lol 2023-07-13 06:33:35 Of course, but Vaultwarden is an alternative community implementation. Written in Rust I believe 2023-07-13 06:41:13 idkrn[m]: offical server uses Microsoft SQL Server, it is not suitable for personal use. 2023-07-13 06:44:50 you can use postgres or mysql too 2023-07-13 07:09:18 i don't see how MS SQL Server would make it unsuitable for personal use anyway :) 2023-07-13 07:24:49 can't tell if trolling 2023-07-13 07:44:19 dwfreed: i treat my installation as a development workstation. in fact, i wouldn't mind having debug symbols and sources for all installed packages either. 2023-07-13 08:48:16 psykose, i was not trolling 2023-07-13 09:02:52 didn't know microsofts proprietary database was suitable for personal use nowadays 2023-07-13 09:02:58 i hope they support oracle db too 2023-07-13 09:03:01 /sarcasm 2023-07-13 09:09:30 psykose, they even provide a docker image! 2023-07-13 09:09:44 it wouldn't be my choice, but it's possible (and i do run it on occasion for testing) 2023-07-13 09:30:37 :) 2023-07-13 10:25:16 MS SQL Server needs at least 2GB RAM 2023-07-13 10:25:48 who among us doesn't, these days 2023-07-13 10:26:49 A server with such a large memory requires a lot of money, and for individuals, it has excessive performance. 2023-07-13 10:26:49 my brain does fine with 128mb 2023-07-13 10:27:53 I only use it run one password vault. 2023-07-13 10:53:26 Fxzxmic, ah, that's fair 2023-07-13 12:17:29 hmz, the screenflickering is back 2023-07-13 18:20:59 "Of course, but Vaultwarden is an..." <- Yes, but that wasn't what was being discussed 2023-07-13 18:50:36 how stable is edge? 2023-07-13 18:50:43 who can tell me something about that? 2023-07-13 18:50:59 pretty stable? what kind of answer are you looking for here 2023-07-13 18:51:06 I run it on my laptop and it usually works 2023-07-13 18:51:21 hmz 2023-07-13 18:51:43 well im looking not so much for an answer, more so an opinion :) 2023-07-13 18:52:05 i want to run docker as a regular user, for this i need docker-rootless-extras which is only in edge 2023-07-13 18:52:29 as far as i got it, anyway :) 2023-07-13 18:52:56 my opinion is that edge works pretty well 2023-07-13 18:53:22 in your opinion, i trust :) 2023-07-13 18:53:24 edge it is then! 2023-07-13 18:53:33 alright, good luck! 2023-07-13 18:54:02 fyi, it has been moved to community, so it would be part of 3.19 when it's released 2023-07-13 18:54:08 i'm really starting to love Alpine :) 2023-07-13 18:54:20 but I run edge on my laptop as well 2023-07-13 18:54:46 yeah, i think i run it on my own laptop too, now that im thinking about it 2023-07-13 18:55:02 but thats because of firewalld 2023-07-13 18:55:02 i still need to learn nftables 2023-07-13 18:59:03 the only real issue i have with Alpine as my main desktop os is that every so foten, the whole system freezes for up to 5 minutes when i have more then 10 tabs open in chrome 2023-07-13 18:59:18 especially when i run virtmanager alongside 2023-07-13 19:00:50 ence its name Alpine 2023-07-13 19:01:31 its supposed to freez in the Alps 2023-07-13 19:01:45 :D 2023-07-13 19:01:48 true, true 2023-07-13 19:27:02 Hi :) 2023-07-13 19:27:43 Just to feedback some more on the scsi-tgt inclusion in Alpine. Still works great and is very performant too. :) I am very grateful for this inclusion 2023-07-13 19:38:13 Hmmm... do you find it better than built-in Linux one? 2023-07-13 19:38:33 And if so, what features do you lack in TCM 2023-07-13 20:31:44 hello there 2023-07-13 20:31:53 This is a support channel for Alpine? 2023-07-13 20:32:08 yes, just ask your question 2023-07-13 20:32:51 Well, I am trying all Linux Distros, and wanted to know what Alpine was most similar to? 2023-07-13 20:33:47 Alpine is kinda its own thing 2023-07-13 20:35:34 ok 2023-07-13 20:35:42 Which version do you recommend I learn on? Or is there only one version now? 2023-07-13 20:35:51 Alpine has a dedicated package manager, uses musl instead of glibc. uses openrc as srevice manager 2023-07-13 20:35:58 completely different? 2023-07-13 20:36:30 Yippee: Either the latest stable (3.18 atm) or edge 2023-07-13 20:36:30 And busybox instead of coreutils. 2023-07-13 20:36:43 ikke: what is the lightest version? 2023-07-13 20:36:48 no differnece 2023-07-13 20:36:51 difference* 2023-07-13 20:37:03 Okay. I am absolutely floored at Linux. 2023-07-13 20:37:12 How has the entire world not adopted Linux yet...? ;) 2023-07-13 20:37:14 Alpine Linux does not come in all kinds of variants 2023-07-13 20:37:22 you install the OS and then the packages you want 2023-07-13 20:37:26 ok. Can you download GNOME desktop environment? 2023-07-13 20:37:29 yes 2023-07-13 20:37:32 wow 2023-07-13 20:38:06 So a completely new learning curve then? 2023-07-13 20:38:11 I just loaded alpine onto an eeepc and its nice to see only around 40mb ram usage 2023-07-13 20:38:29 anyone managing alpine using ansible? 2023-07-13 20:39:32 Yippee: It's still linux, so whatever you know about linux can transfer, and if you install gnome, gnome will be the same as everywhere 2023-07-13 20:39:40 so it's more like the glue that is different 2023-07-13 20:40:05 And it's a bit more manually 2023-07-13 20:40:24 hmz, i managed to get ansible working by installing python on the alpine client, now i need to figure out how to make it work with doas instead of sudo 2023-07-13 20:41:35 ikke: thank you, do you know if software like Davinci Resolve / Kdenlive is compatible? (Video Editing software) 2023-07-13 20:41:57 whenever I add a package with apk, it sits there for a while before anything happens, which has not been my experience when running in a VM 2023-07-13 20:42:08 Yippee: https://pkgs.alpinelinux.org/package/edge/community/x86_64/kdenlive 2023-07-13 20:42:29 thank you 2023-07-13 20:42:31 I don't know about Davinci Resolve 2023-07-13 20:42:44 thats ok, not even Debian 12 has compatibility with Davinci yet 2023-07-13 20:43:01 i think resolve is an appimage or so? 2023-07-13 20:43:17 Davinci Resolve 2023-07-13 20:43:27 Black Magic 2023-07-13 20:49:10 Can you encrpyt the entire O/S in Alpine? 2023-07-13 20:49:15 encrypt* 2023-07-13 20:53:11 How is Alpine standard only 200MB? Does it download all the updates during installation? 2023-07-13 20:54:19 Yeah, it just contains whatever is necessary for booting and getting a network connection, most of the rest is downloaded on the fly 2023-07-13 20:54:43 I actually prefer that 2023-07-13 20:54:45 Yippee: the installer has an option to have an encrypted root, but /boot is still unencrypted 2023-07-13 20:55:22 ikke: thank you. I suppose I could encrypt a file after installation to keep all my data stored in /HOME folder? 2023-07-13 21:00:50 yeah, that should be possible 2023-07-13 21:02:02 excellent. I am going to try installing this beast 2023-07-13 21:02:07 I'll be back from time to time 2023-07-13 21:02:17 thanks so much for chatting and helping me everyone. I look forward to Alpine 2023-07-13 21:22:05 consus[m]: The issue with TCM/LIO is that it is a kernel module, and when errors occur, it can bring down the kernel or prevent clean shutdowns. 2023-07-13 21:22:28 Yippee: all linux distros? :D 2023-07-13 21:23:25 Yippee: I'd agree with ikke that Alpine is quite unique. If you are familiar with OpenRC (Gentoo and some others use it), then you'd be at home with Alpine :) 2023-07-13 21:25:25 consus[m] https://wiki.tnonline.net/w/Blog/iSCSI_target_in_user-space 2023-07-13 22:02:11 Forza: YEP - I sincerely want to get a broad understanding of ALL Linux Disros - so that I can build my own that is unique and different. 2023-07-13 22:02:24 Custom Tailored for Film Editors. 2023-07-13 22:02:25 :o 2023-07-13 22:02:39 I thought there was one like that before? 2023-07-13 22:02:44 lol there is 2023-07-13 22:02:46 lots I'm sure 2023-07-13 22:02:55 Maybe it was just a Ubuntu flavour 2023-07-13 22:03:14 Amiga 4000 2023-07-13 22:03:15 Q: What is a flavour? Is that like a fancy word for desktop environment/ 2023-07-13 22:03:40 Variant of a base distro 2023-07-13 22:04:03 Like Kubuntu is a variation of Ubuntu 2023-07-13 22:04:45 is there anything different between them when it comes to terminal commands? 2023-07-13 22:05:45 ah yes, video toaster linux 2023-07-13 22:11:18 ffog: what is that> 2023-07-13 22:15:50 just a video editing joke 2023-07-13 22:20:22 lol 2023-07-13 22:20:51 I'm so daft. I don't normally use chat since msn Messenger was a big thing in like early 2000's 2023-07-13 22:23:02 :) 2023-07-13 22:26:04 Yippee: sorry, lost network for a bit. There is no difference in commands between flavours of a distro 2023-07-13 22:26:26 thanks 2023-07-13 22:38:16 JohannesJ[m]: from memory there is a module in Ansible community for doas support 2023-07-14 10:03:20 Does anyone have an insigh into intel-sof audio issues? I got my hands on a Dell XPS 9320 and it does not detect a soundcard, even though the sof-firmware blob is installed. I suspect missing kernel support for Soundwire or Dell specific devices, but did not yet dig into the current .config of the lts kernel 2023-07-14 14:41:23 telmich: I don't suppose you have some other setup where it is working to compare against? 2023-07-14 15:02:46 hi. I'm having trouble installing zfs-rpi4 on 3.18 with tmpfs install. In previous versions I would just `apk add zfs-rpi4` and copy the generated initramfs to the microsd but after the last version that doesn't work anymore, in fact it breaks all kernel modules. I think it's because zfs-rpi4 is for kernel 6.1.37 but the current alpine kernel release is 6.1.34. Any ideas where to go from here? 2023-07-14 15:09:48 zipi: this is Alpine run-from-ram? 2023-07-14 15:09:58 yes 2023-07-14 15:21:59 oh i saw the issue, i need to update modloop to 6.1.37 2023-07-14 15:25:39 zipi: for run-from-ram to check the kernel you need to use "update-kernel" AFAIK but it can be problematic (i.e. not enough ram free). Generally I believe for run-from-ram to change kernel you're basically expected to reinstall Alpine 2023-07-14 15:26:04 Someone else can probably confirm this, I don't use run-from-ram myself 2023-07-14 15:34:06 thank you. The kernel is updated to the new version by installing zfs-rpi4, it creates a dir at /boot which I copy to /media/mmcblk0p1. Then after mkinitfs (forgot this step before asking the question) and reboot I can now load zfs. Now the only issue is modloop doesn't work because it's still the old version, but I thinking maybe I don't need modloop after all 2023-07-14 15:41:09 zipi: look at /sbin/update-kernel 2023-07-14 15:41:40 "Kernel and firmware update script for Alpine installations set up with setup-bootable" 2023-07-14 15:43:06 yep i'm looking now if there's a way to create the modloop without building the entire kernel because the pi will take several hours to compile that 2023-07-14 15:45:31 zipi: *building* the kernel? what do you mean? why would that happen? 2023-07-14 15:45:58 guess i didn't understand what the script does 2023-07-14 15:46:12 if updates the kernel package etc 2023-07-14 15:46:24 s/if/it/ 2023-07-14 15:47:25 oh god i'm stupid 2023-07-14 15:47:37 no space left on device 2023-07-14 15:48:06 i think it will work now, thank you 2023-07-14 15:48:46 space == ramdisk? I did mention about memory issues 2023-07-14 15:49:07 i had prepared an upgrade script to deal with this a couple versions ago and this that step failed and i didn't pay attention to it 2023-07-14 15:49:20 yep i have to specify a temp dir 2023-07-14 16:07:57 that took a while but everything works now 2023-07-14 16:08:12 minimal: thank you very much for your help! 2023-07-14 18:40:55 Hi. Does openrc have a reboot command for services, is it just the standard "stop" that is used? 2023-07-14 18:41:26 reboot? you mean restart I assume 2023-07-14 18:43:09 if [ "$RC_RUNLEVEL" = "shutdown" ]; then 2023-07-14 18:43:14 Yea 2023-07-14 18:43:53 that's a machine runlevel, not a service "state" 2023-07-14 18:44:11 Sorry. I used the wrong terminology 2023-07-14 18:44:37 the runlevels (and which services apply to them) being shown by "rc-status" 2023-07-14 18:44:51 Forza: what would a reboot command for services mean? 2023-07-14 18:45:04 I wanted to modify a service 'tgtd' to do a different thing during reboots than if you simply do rc-service tgtd stop 2023-07-14 18:45:24 Not sure if openrc distinguishes between reboots and shutdowns 2023-07-14 18:45:38 there's no different for a service between a shutdown or a reboot 2023-07-14 18:46:21 Can we not check run level? 2023-07-14 18:46:41 no 2023-07-14 18:46:45 a machine shutdown is typically just a transition from "default" to "shutdown" runlevel 2023-07-14 18:46:49 RC_RUNLEVEN 2023-07-14 18:46:52 RC_RUNLEVEL 2023-07-14 18:46:54 https://manpages.debian.org/testing/openrc/openrc-run.8.en.html 2023-07-14 18:47:06 then the machine reboots and OpenRC is started 2023-07-14 18:47:35 if [ $RC_RUNLEVEL = "shutdown" ]; then ..; fo 2023-07-14 18:48:19 Yea. I'll try 2023-07-14 18:49:17 Psykose i found out that tgtd exists when all targets have been removed from the running config, so we need to modify the init script a little. Was thinking to do that and maybe submit an update, if that's OK? 2023-07-14 18:50:39 You never have to ask permission to create an MR :) 2023-07-14 18:50:50 :) 2023-07-14 18:51:19 I might need some advice along the way though. Haven't really written init scripts much 2023-07-14 18:51:35 sure 2023-07-14 18:51:43 Thanks 2023-07-14 18:51:57 Forza: fyi, this is great start: https://github.com/OpenRC/openrc/blob/master/service-script-guide.md 2023-07-14 18:52:15 Oh great. Thank you. 2023-07-14 18:52:42 My normal fixes is to put everything in local.d.... 🤦‍♂️ 2023-07-14 18:53:07 Forza: and than openrc-run man page contains info about what's available in an init.d script 2023-07-14 18:55:50 Nice. I'll work on it as best I can 2023-07-14 19:19:04 hi, is there a correct way to manually set nameserver during initial setup? 2023-07-14 19:19:20 using dhcp i mean 2023-07-14 19:24:34 ghostbuster: doesn't the DHCP server indicate nameservers to use? 2023-07-14 19:24:51 it's not working for some reason so i want to just manually set namesrver to 1.1.1.1 2023-07-14 19:27:29 ghostbuster: can you try to add `dns=1.1.1.1` to /etc/udhcpc/udhcpc.conf 2023-07-14 19:27:41 (that file nore directory exist by default) 2023-07-14 19:28:02 do i do that before i run setup-alpine? 2023-07-14 19:28:11 yeah 2023-07-14 19:30:03 so just to clarify, i don't edit /etc/udhcpd.conf? 2023-07-14 19:30:29 no 2023-07-14 19:31:02 You are not running a dhcp server 2023-07-14 19:31:15 ah 2023-07-14 19:31:20 confused it with dhcpcd 2023-07-14 19:31:49 The missing c changes a lot 2023-07-14 19:31:54 indeed 2023-07-14 19:32:40 that worked, thank you 2023-07-14 19:32:47 cool 2023-07-14 21:24:38 Is there a way to have indented ebegin sections? 2023-07-14 21:24:49 in init scripts 2023-07-14 21:26:02 can't think of a reason to do that 2023-07-14 21:36:47 Just style 2023-07-14 21:36:59 then don't 2023-07-14 21:37:02 Ok so i did something https://bpa.st/3LMA 2023-07-14 21:37:52 Don't really need all the ebegin bits, but I used it as debug info. 2023-07-14 21:38:31 that doesn't even work because tgt-admin is separate 2023-07-14 21:38:47 Also I am not sure how to handle the shutdown properly. The stop_pre seems hackish as last line effectively closes the daemon 2023-07-14 21:39:02 you could just not close it 2023-07-14 21:39:07 I know it depends on scsi-tgt-scripts 2023-07-14 21:39:45 But I wanted it to handle restarts and reloads better without leaving targets wondering too much 2023-07-14 21:40:20 Is it not better to let tgtd close itself instead of killing it.? 2023-07-14 21:41:09 Is there a reason to have the scripts as a separate package? 2023-07-14 21:41:42 they depend on a bunch more stuff 2023-07-14 21:41:57 If we do that we could have a separate init.d script for tgt-admin? 2023-07-14 21:42:06 Yea true. 2023-07-14 21:43:02 Hm so then we can skip tgt-admin parts in tgtd init and just start the daemon in offline mode and leave it to the user to set it into ready mode after loading config? 2023-07-14 21:43:29 is it possible to use it without tgt-admin 2023-07-14 21:43:39 Yes 2023-07-14 21:43:55 Tgt-admin is a wrapper for tgtadm 2023-07-14 21:44:17 It reads the config file and issues tgtadm commands 2023-07-14 21:44:35 One could do those commands manually or through bash scripts etc 2023-07-14 21:46:12 https://bpa.st/5P2Q 2023-07-14 21:46:16 Like those 2023-07-14 21:48:14 did you have a default /etc/tgt/targets.conf in mind 2023-07-14 21:49:32 Hm. Not really 2023-07-14 21:49:44 Maybe there ought to be an example 2023-07-14 21:50:37 I started one like this https://bpa.st/S3VQ 2023-07-14 21:51:02 shrug 2023-07-14 21:51:05 guess just leave it as nothing 2023-07-14 21:51:17 Works too 2023-07-14 21:51:50 sure, pushed mostly that 2023-07-14 21:51:53 I think a basic example would be alright? They provide one on the github repo I think 2023-07-14 21:52:05 the sigint stops it the same afaict 2023-07-14 21:52:28 this https://github.com/fujita/tgt/blob/master/doc/targets.conf.5.xml ? 2023-07-14 21:52:39 nah that's a manpage 2023-07-14 21:52:41 https://github.com/fujita/tgt/blob/master/conf/examples/targets.conf.example 2023-07-14 21:53:26 Both man page and examples are slightly incomplete though :o 2023-07-14 21:59:45 hmm 2023-07-14 21:59:46 not sure 2023-07-14 22:05:10 Perl is a big dependency 2023-07-14 22:05:44 Shame they didnt write tgt-admin in bash o c or something like it 2023-07-14 22:08:02 To be fair. I have used the init script you made and haven't yet had any issues with clients/initiators. I have to simply manually run tgt-admin -e in local.d. this is why I wanted to modify the init script 2023-07-14 22:09:02 Can the init script check for the presense of tgt-admin and do different things depending on ? 2023-07-14 22:09:11 bad idea 2023-07-14 22:09:30 could just have two theoretically 2023-07-14 22:10:53 a tgt-admin with `need tgt` etc 2023-07-14 22:10:57 dunno tho 2023-07-14 22:10:59 looks ok as is 2023-07-14 22:11:30 Yea. I suppose anyone using it will know what to do? 2023-07-14 22:13:12 even with a sample file they need to know it and fill out most of it from what i can tell 2023-07-14 22:13:22 " a tgt-admin with `need tgt` etc" not a badd idea. Would probably be ok.without tgt-admin ,any configuration is manual, so the user has to know how to do it 2023-07-14 22:13:33 Yes. 2023-07-14 22:13:42 Like a samba share 2023-07-14 22:13:45 pretty much 2023-07-14 22:14:05 You still need to decide what to share and how it should be setup 2023-07-14 22:14:21 anyways 2023-07-14 22:14:29 feel free to test the current and see if it even works or not :D 2023-07-14 22:14:50 openrc has a 'want' now 2023-07-14 22:14:58 it behaves similarly to systemd 2023-07-14 22:15:18 You mean the tgtd init in current package? Yes it is working, but it isn't loading any configuration :) 2023-07-14 22:15:34 no, the new one 2023-07-14 22:15:34 it will start the initscript if it exists, but will not fail the dependency if the initscript fails to start or does not exist 2023-07-14 22:15:46 this is the opposite of a want 2023-07-14 22:16:41 ah, I misread the backlog 2023-07-14 22:16:52 Ah yes I'll test it for a bit. But I understand the dependency on perl is unnecessary for some, so I might split into two init scripts to make it optional 2023-07-14 22:17:38 shrugging emoji 2023-07-14 22:17:55 Haha 2023-07-14 22:17:57 Sorry 2023-07-14 22:18:02 :( 2023-07-14 22:18:31 Didn't mean to burden with all this 2023-07-14 22:20:42 how's https://img.ayaya.dev/dWZgjq6QkzoV 2023-07-14 22:23:49 should be identical just in a separate file 2023-07-14 22:26:33 pushed that 2023-07-14 22:26:50 Oh. 2023-07-14 22:27:02 That's kind of you 2023-07-14 22:27:33 But now I don't get to practice making merge requests ;) 2023-07-14 22:29:05 it's mostly just pressing three buttons 2023-07-14 22:29:10 you wrote the whole thing anyway 2023-07-14 22:30:40 :) 2023-07-14 22:32:11 Running a basic alpine install in a vm with one file exported with tgtd uses about 70-80MB ram only. Quite lightweight. 2023-07-14 22:32:42 one day i'll find out what that even does :D 2023-07-14 22:35:07 Hehe 2023-07-14 22:39:40 Anyway, I am grateful for the inclusion of it. 2023-07-14 22:41:43 How many devs and admins are there on Alpine Linux? 2023-07-14 23:54:44 well, I'm a dev and a sysadmin sort of, but not of alpine Linux. I run it at home on my ppc64 server 2023-07-14 23:55:11 mainly for an AIX vm 2023-07-15 02:11:18 Hi Alpine Linux 2023-07-15 02:14:13 I installed you to my USB stick, I ran you, but you didn't give me anything in return. You booted me, but you provided me with a login / username to which I had no idea what to enter/ 2023-07-15 02:19:17 I'd like to install you Alpine, cause I know you deserve a fair chance to show your true colors ;) 2023-07-15 02:33:00 Installer should be username root with a null/empty password  2023-07-15 06:47:42 where does alpine store the umask? 2023-07-15 06:58:46 It's something that processes inherit 2023-07-15 06:59:10 So alpine does not set any specific default umask, it just uses what linux comes with by default 2023-07-15 07:20:50 "nico: I don't suppose you have..." <- Not yet, but I'll try to get an ubuntu live boot up later today, just need to find the usb-c-to-a adapter for that 2023-07-15 07:39:36 ikke, thanks I just saw that alp uses minimal pam stuff 2023-07-15 07:46:45 No pam by default 2023-07-15 09:05:57 ACTION Good morning 🙏  2023-07-15 09:06:47 rob[m]: aix in a vm? Cool. Never came across anyone using aix. 2023-07-15 09:15:49 <|cos|> ACTION didn't find AIX trivial to gain access to. Someone on reddit claimed it should be possible to launch a free tier instance at cloud.ibm.com, but I never managed to figure out how. I wanna follow along if the AIX conversation moves to where it is on-topic. 2023-07-15 09:29:14 Ok, so sound under ubuntu with an old kernel 2023-07-15 09:30:52 ikke: do you have somewhat experience with the sof stuff? 2023-07-15 09:32:47 on ubuntu with firmware-sof-signed 2.0-1ubuntu4.1 and kernel 5.19.0-32-generic sound works 2023-07-15 09:38:16 And the notebook is significantly less hot under ubuntu, there must be some tuning done that I did not yet find under Alpine 2023-07-15 09:45:50 telmich: no, I have not 2023-07-15 09:45:52 which notebook do you have 2023-07-15 09:47:07 ah 2023-07-15 09:47:08 hmm 2023-07-15 09:47:14 that should just be SND_SOC_SOF_INTEL_TGL and then the firmware 2023-07-15 09:47:49 what's your dmesg 2023-07-15 09:53:58 It's a dell xps plus 9320 2023-07-15 09:54:12 I have it currently booted into ubuntu, but can reboot into -lts or -edge 2023-07-15 09:54:37 ubuntu dmesg is also nice 2023-07-15 09:55:29 and lspci -k of both 2023-07-15 09:56:18 ok, just a sec, ubuntu lost the wifi connection 2023-07-15 10:02:35 uploading to https://www.nico.schottelius.org/temp/xps-9320/ 2023-07-15 10:03:55 ...booting into ubuntu 2023-07-15 10:08:48 it loads the module and firmware fine 2023-07-15 10:09:14 on lts, yes 2023-07-15 10:09:16 on edge, no 2023-07-15 10:09:24 On lts the aplay -l is missing the speaker 2023-07-15 10:09:33 I just uploaded all infos from ubuntu 2023-07-15 10:09:37 including the .config 2023-07-15 10:11:01 all i can think of is the `No SoundWire machine driver found` 2023-07-15 10:11:19 for some reason the ubuntu one detects that fine 2023-07-15 10:13:29 we also have this one on -lts: [ 10.738867] skl_hda_dsp_generic skl_hda_dsp_generic: ASoC: Parent card not yet available, widget card binding deferred 2023-07-15 10:16:01 doesn't look that special 2023-07-15 10:16:08 could be one of the 50 million other SND_SOC modules 2023-07-15 10:16:39 ubuntu has: CONFIG_SND_INTEL_BYT_PREFER_SOF=y 2023-07-15 10:17:00 maybe SND_SOC_RT715_SDW 2023-07-15 10:17:02 Even though dmesg indicates that sof is taken anyway, due to cannonlake arch 2023-07-15 10:19:55 yeah it probably is that 2023-07-15 10:19:57 i'll toggle a few 2023-07-15 10:28:17 https://git.alpinelinux.org/aports/commit/?id=f5340c67131c test that when it builds 2023-07-15 10:28:20 should work from what i read 2023-07-15 10:36:36 ack, rebooting back to Alpine for building 2023-07-15 10:36:41 Hi 2023-07-15 10:36:55 Salut hungry 2023-07-15 10:36:56 it'll already build in a bit if you wait like 40 minutes 2023-07-15 10:36:57 or so 2023-07-15 10:36:58 how are you? 2023-07-15 10:37:14 ok, that is pretty neat 2023-07-15 10:37:45 neato squeato 2023-07-15 10:38:16 how do we go into building Alpine from USB? 2023-07-15 10:38:34 I see the login information component. Is it user root and password emptyy? 2023-07-15 10:39:26 That is correct, you can locally login without a password, hungry 2023-07-15 10:39:44 After that, does Alpine install? 2023-07-15 10:40:30 No, you will need to do that yourself 2023-07-15 10:40:50 lol 2023-07-15 10:40:52 Is it easy/ 2023-07-15 11:22:12 hmm, 6.1.38-2 stays with the same error 2023-07-15 11:22:22 (files already uploaded to https://www.nico.schottelius.org/temp/xps-9320/) 2023-07-15 11:31:12 Should likely be unrelated, but ubuntu has as a module: CONFIG_SOUNDWIRE_QCOM 2023-07-15 11:31:37 And I wonder whether CONFIG_SND_INTEL_BYT_PREFER_SOF is actually required 2023-07-15 11:32:48 And I'd have expected CONFIG_SND_SOC_SOF_CANNONLAKE=m do do the job, but seems it does not 2023-07-15 11:34:21 CONFIG_SOUNDWIRE_QCOM isn't in this laptop pretty sure 2023-07-15 11:36:39 the default doesn't really make a difference, you can try snd_intel_dspcfg.dsp_driver=3 to emulate the same as it being default 2023-07-15 11:43:02 do you have the same userspace audio setup on both ubuntu and here 2023-07-15 11:43:02 psykose: unfortunately you are right, same result 2023-07-15 11:43:36 I did not configure much on alpine, as the missing speaker on the aplay -l output did indicate it cannot work 2023-07-15 11:43:40 I do have alsaucm installed on Alpine 2023-07-15 11:46:08 not necessarily afaik 2023-07-15 11:46:13 overall tho looks like https://github.com/thesofproject/sof/issues/6111 2023-07-15 11:46:26 if you rmmod the snd_sof thing and reprobe it does it work lol 2023-07-15 11:48:29 I did unload everything a few minutes ago, including snd_sof, but unfortunately did not change the situation 2023-07-15 11:49:07 I actually get similar dmesg output as in https://github.com/thesofproject/sof/issues/6111 if I boot into the -edge kernel with timeouts 2023-07-15 11:52:35 but you are right, in the ubuntu dmesg I also only see the HDMI outputs 2023-07-15 11:58:00 I have this error from an init.d script: /lib/rc/sh/openrc-run.sh: /etc/init.d/tgt-admin: line 6: /etc/tgt/targets.conf: Permission denied 2023-07-15 11:58:26 File is world readable : -rw-r--r-- 1 root root 199 Jul 15 00:23 /etc/tgt/targets.conf 2023-07-15 11:59:50 think i typod that 2023-07-15 12:02:01 telmich: enabled a few more sus looking things 2023-07-15 12:02:07 if not that tho then i don't think it's kernel related 2023-07-15 12:02:56 ack, will test as soon as it's uploaded 2023-07-15 12:03:16 What typo? 2023-07-15 12:04:32 Aside from the No SoundWire machine driver found message, the rest of the "errors" I can also locate in the ubuntu dmesg. Firmware wise we are a bit newer than ubuntu, but I hope that should not be a problem 2023-07-15 12:08:33 you could download an old one and just replace some files manually but i doubt it 2023-07-15 12:08:40 Forza: : ${ 2023-07-15 12:10:55 a : in front 2023-07-15 12:19:17 Hm. Should there be an : infront? Dont see how it is wrong? 2023-07-15 12:20:52 Even setting required_files=/etc/tgt/targets.conf doesn't work. Hmm. 2023-07-15 12:23:02 https://bpa.st/JT2A debug output 2023-07-15 12:33:47 Doh..... 2023-07-15 12:33:53 Should be tgtd_conf="${tgtd_conf:=/etc/tgt/targets.conf}" 2023-07-15 12:35:10 you are double assigning that variable 2023-07-15 12:35:24 : "${tgtd_conf:=/etc/tgt/targets.conf}" 2023-07-15 12:35:34 ^ is the correct way to do it; that leading colon is requires 2023-07-15 12:35:38 s/s$/d/ 2023-07-15 12:35:55 Ah hum 2023-07-15 12:36:17 the leading colon is a no-op command, which is being used to do the variable expansion and assignment 2023-07-15 12:36:45 I see. Hadn't seen that before 2023-07-15 12:36:53 Forza: man bash | grep -F '${p' 2023-07-15 12:36:55 It works too 2023-07-15 12:37:08 Thanks 2023-07-15 12:37:31 using grep on a manpage is just dumb 2023-07-15 12:37:40 or man bash and search \$\{p 2023-07-15 12:37:54 Btw. Is "$var" or "${var}" preferred in scripts? 2023-07-15 12:38:01 latter generally 2023-07-15 12:38:14 better to search for "Parameter #xpansion", and then you don't have escape any special characters 2023-07-15 12:38:19 s/#/E/ 2023-07-15 12:38:27 dwfreed: it is not dumb, i am telling what to look for. 2023-07-15 12:38:47 uuidNuniq: yes, but then you can't look around at the context, you can only see the lines you asked for 2023-07-15 12:38:52 which is dumb 2023-07-15 12:38:56 dwfreed: -C10 2023-07-15 12:39:18 and if you want more than 10 lines of context? 2023-07-15 12:39:27 better to just look at in a proper pages 2023-07-15 12:39:28 :) 2023-07-15 12:39:33 s/s$/r/ 2023-07-15 12:39:37 I really can't type today 2023-07-15 12:40:44 Problem is solved :) 2023-07-15 12:41:14 psykose: thanks. I've tested loading and reloading and restarting and it seems to work as intended now 2023-07-15 12:42:06 Quite interesting that running alpine as iscsi target is specifically mentioned in the about page :) 2023-07-15 13:05:55 psykose: Just rebooted into 6.1.38-3, unfortunately no change. I am wondering a) whether the "No SoundWire machine driver found" is significant or b) whether it's possible / needed to fix it in userspace 2023-07-15 13:06:08 might be 2023-07-15 13:06:12 idk what else to change 2023-07-15 13:07:47 I wonder what I can gather from ubuntu, will reboot into it to see if something jumps at me 2023-07-15 13:09:48 Do you know if there are any plans for replacing eudev with systemd-udev as they did on Gentoo? 2023-07-15 13:14:44 Does Alpine have commercial support options? 2023-07-15 13:15:07 telmich: check the list of modules and compare with alpine 2023-07-15 13:15:10 ok, uploaded the alsa info from ubuntu now here directly: http://alsa-project.org/db/?f=0a31b5d9961839f11b2b72996a57cef2a0609e9b 2023-07-15 13:16:21 Forza: no 2023-07-15 13:16:54 !!! 2023-07-15 13:16:56 Seen it! 2023-07-15 13:17:06 ubuntu: snd_soc_sof_sdw (card 0) 2023-07-15 13:17:13 alpine: snd_soc_skl_hda_dsp (card 0) 2023-07-15 13:17:17 both from alsa-info 2023-07-15 13:17:38 psykose: ok thanks. :) 2023-07-15 13:17:38 Now the only questions are why and how to change it 2023-07-15 13:17:49 hmm 2023-07-15 13:18:57 find | grep snd_soc_sof_sdw | wc -l -> 0 2023-07-15 13:19:01 I'd say that explains it 2023-07-15 13:19:31 snd_soc_skl_hda_dsp isn't found either 2023-07-15 13:19:52 ah, the names are more fancy 2023-07-15 13:19:54 ah, - vs _ 2023-07-15 13:19:55 I hate that 2023-07-15 13:20:06 they both exist 2023-07-15 13:20:09 you can try blacklist one 2023-07-15 13:20:41 CONFIG_SND_SOC_INTEL_SOUNDWIRE_SOF_MACH is what gives the sof one 2023-07-15 13:20:50 then the extra codecs are separate 2023-07-15 13:20:54 but it loads the wrong one it seems 2023-07-15 13:21:49 minimal: i tried your alpine generation script, but ran into some issues. Reported them on your github page. 2023-07-15 13:22:10 so ubuntu also has the snd-soc-skl_hda_dsp.ko available 2023-07-15 13:22:27 but not loaded, question is how did they do that 2023-07-15 13:22:36 could just be initrd magic 2023-07-15 13:22:57 or it's CONFIG_SND_INTEL_BYT_PREFER_SOF 2023-07-15 13:23:17 which i forgot to set, but you also tested =3 2023-07-15 13:23:24 which should be the same 2023-07-15 13:24:01 I tried loading before with dsp_driver=3, in theory only that module should need /read it, or am I mistaken? 2023-07-15 13:24:22 I can try adding it as a kernel command line instead of module parameter, but afaik, it should not change anything 2023-07-15 13:24:45 Wasn't there somewhere a map of pci ids to modules? 2023-07-15 13:24:57 it is a kernel command line parameter 2023-07-15 13:25:02 well 2023-07-15 13:25:11 if you mean doing `options` in the module setting 2023-07-15 13:25:13 then it's the same 2023-07-15 13:25:30 I passed it from the modprobe call 2023-07-15 13:25:44 But anyway, let me try a) blacklisting and b) kernel command line adding the parameter 2023-07-15 13:26:29 set it anyway 2023-07-15 13:27:54 what a strange module name it is, in any case: ./6.1.38-3-lts/kernel/sound/soc/intel/boards/snd-soc-skl_hda_dsp.ko.gz 2023-07-15 13:28:04 mixing dashes and underlines, is this to confuse everyone? 2023-07-15 13:28:38 first two dashes are the subsystem part 2023-07-15 13:28:46 underscore part is the name part 2023-07-15 13:29:55 ... but in the lsmod output, dashes are underline converted? 2023-07-15 13:33:55 makes perfect sense! 2023-07-15 13:33:55 not 2023-07-15 13:35:25 I am too stupid to blacklist the module 2023-07-15 13:35:50 modprobe.blacklist=snd-soc-skl_hda_dsp at kernel command line did not do the job 2023-07-15 13:36:40 adding blacklist snd_soc_skl_hda_dsp in /etc/modprobe.d/blacklist.conf also does not work 2023-07-15 13:38:04 Adding blacklist snd-soc-skl_hda_dsp into /etc/modprobe.d/blacklist.conf also does not work 2023-07-15 13:39:17 sun:/lib/modules# mv ./6.1.38-3-lts/kernel/sound/soc/intel/boards/snd-soc-skl_hda_dsp.ko.gz ~ 2023-07-15 13:39:21 that works 2023-07-15 13:39:29 works as in, not loaded anymore 2023-07-15 13:39:29 it's probably dashes only 2023-07-15 13:39:34 lsmod name 2023-07-15 13:39:40 or 2023-07-15 13:39:42 underlines only 2023-07-15 13:39:42 idk 2023-07-15 13:39:44 lsmod name 2023-07-15 13:41:23 so by removing snd-soc-skl_hda_dsp.ko.gz from the modules, I don't see any card anymore 2023-07-15 13:41:40 even though sof_sdw is loaded 2023-07-15 13:41:59 no, wait, it is actually not loaded 2023-07-15 13:42:33 So loading snd_sof_pci_intel_tgl does not trigger loading snd_soc_sof_sdw 2023-07-15 13:43:09 and loading it manually, does not show any card nor produces any line in dmesg 2023-07-15 13:45:10 searching for SOF is a nightmare 2023-07-15 13:45:20 Especially when software matches on sof 2023-07-15 13:53:36 ok, I think I have to stop for today 2023-07-15 13:53:59 But the latest state is: I was able to blacklist the module and now no card is shown anymore 2023-07-15 13:54:50 So either something changed significantly between 5.19 and 6.1.38 (somewhat likely) and the module shifted or we might not have the right suport in the snd_soc_sof_sdw module 2023-07-15 13:56:18 everything is enabled in it 2023-07-15 13:56:19 more or less 2023-07-15 13:56:26 pretty much literally every intel relevant part 2023-07-15 13:57:40 ah 2023-07-15 13:57:47 maybe it's AVS_MACH 2023-07-15 13:58:46 it was new from 5.15->6.1 lts (technically 5.19) and i think a bunch of stuff moved to it 2023-07-15 13:58:50 but it's completely disabled 2023-07-15 14:01:13 ubuntu has: CONFIG_SND_SOC_INTEL_AVS=m 2023-07-15 14:01:31 lts has: CONFIG_SND_SOC_INTEL_AVS=m 2023-07-15 14:01:44 ah, but I see what you mean 2023-07-15 14:01:53 CONFIG_SND_SOC_INTEL_AVS_MACH_* 2023-07-15 14:03:06 sure is a lot of kernel builds 2023-07-15 14:03:15 if this doesn't work then really definitely for sure nothing left that is obvious though 2023-07-15 14:03:20 from the SND_SOC part 2023-07-15 14:03:22 is it possible to reverse search which modules are claiming a specific pci id? and then to traverse back, what we don't have enabled? 2023-07-15 14:03:36 it's not related to pci id, that's already enabled afaik 2023-07-15 14:04:00 What I mean is, whether there is a reason ubuntu selects sof and we are going hda-dsp 2023-07-15 14:05:49 Now re-reading https://thesofproject.github.io/latest/getting_started/intel_debug/introduction.html - this sounds stuff looks much more messy than OSS and that one was not clean either 2023-07-15 14:15:58 now to find out if AVS works or not 2023-07-15 14:16:03 i bet still no :D 2023-07-15 14:20:04 I am with you in that bet :-( 2023-07-15 15:21:45 Just upgraded to 6.1.38-5 and unfortunately it's still the same result, uploaded files again to https://www.nico.schottelius.org/temp/xps-9320/, including the latest alsainfo 2023-07-15 15:23:48 all i can think of is udev fuckiness or the new firmware breaking it 2023-07-15 15:25:28 I'm 90% convinced that if the right module was to claim the device, it would work[tm] 2023-07-15 15:25:59 My question is, why it does not and how to find out the decision tree in the kernel 2023-07-15 15:27:44 if you find out let me know :D 2023-07-15 15:27:52 i have some sleeping to get to 2023-07-15 15:28:51 Go for it 2023-07-15 15:28:55 And thanks for your help 2023-07-17 02:27:27 hello 2023-07-17 10:18:27 Hi, am wondering if a new version of cryptsetup or grub had some significant change on 3.18 as after upgrading and rebooting am not able to boot and drop to grub rescue where I got some error where my passphrase is incorrect and crypto disk is not found but I am sure of my passphrase 2023-07-17 10:28:37 cryptsetup isn't used by grub, grub itself has LUKS support, i'm not using it but maybe someone else does here 2023-07-17 10:42:41 change of keyboard? 2023-07-17 10:49:26 could be keyboard layout indeed 2023-07-17 10:50:42 hello 2023-07-17 10:50:48 does grub support keyfiles for luks? it would be a good way to differentiate between input problem and decryption problems 2023-07-17 10:50:54 akndjd: o/ 2023-07-17 10:50:59 i need some help recovering files using alpine 2023-07-17 10:51:46 i have an old PC, but for some reason the preinstalled os on it (windows 7) refuses to boot, which means i can only access the hard drive using a USB drive wit linux on it 2023-07-17 10:52:16 can i use alpine to somehow copy the files from the HDD to another usb stick witout the need to install alpine itself on the hardware? 2023-07-17 10:52:40 it might be easier to use a windows recovery disk to recover the install 2023-07-17 10:53:48 but if you want to use alpine you can install ntfs-3g in the live session after running setup-apkrepos and setup-networking 2023-07-17 10:54:07 and then you can mount in and access the data assuming it's not bitlocker encrypted 2023-07-17 10:55:08 if you have a bitlocker encrypted media you can use dislocker 2023-07-17 10:55:41 just keep in mind you will likely use filesystem metadata if you copy them elsewhere so you can't copy system data and expect it to work flawlessly 2023-07-17 10:55:51 lose*, not use 2023-07-17 10:56:13 i dont wanna copy system data 2023-07-17 10:56:18 soi guess that should work 2023-07-17 10:56:25 yeah it should be fine then 2023-07-17 10:56:50 so, after i mount into my HDD usinf ntfs-3g i can just cp the files to another thumb drive? 2023-07-17 10:56:57 yeah 2023-07-17 10:57:59 you may have problems with fast shutdown since windows keeps some unflushed metadata, in that case you may use ntfsfix 2023-07-17 10:58:20 however that may result in some data loss so make a backup just to be safe 2023-07-17 10:58:37 i just wanna copy images and videos 2023-07-17 10:59:13 yeah that's fine, but just mentioning that 2023-07-17 10:59:29 since you'll likely get a warning and get a read only mount if that is the case 2023-07-17 10:59:48 but you can still copy stuff FROM it without problems 2023-07-17 11:00:14 yeah thats the goal, just copy for now 2023-07-17 11:00:22 ill format the whole thing if the backup is successful 2023-07-17 11:16:02 caskd Apachez thanks and yes I tried the keyboard and adding a very simple passphrase to avoid any locale issue and the .bin is used to avoid puting the passphrase twice as Iam in a FDE setup 2023-07-17 11:21:01 well, the keyfile is used from the initramfs which i assume is encrypted in /boot 2023-07-17 11:21:17 that is mostly irrelevant right now since grub is the one causing problems 2023-07-17 11:22:51 i was wondering if grub can use a keyfile for /boot to decrypt it to rule out any input problems (duplicate keypresses, skipped presses, input locale etc...) 2023-07-17 11:23:25 as if that works, there's something wrong with the input and if that doesn't work either then grub doesn't support the encryption algorithm used for the disk 2023-07-17 11:23:50 grub should support luks2 nowadays but it used to have problems with it 2023-07-17 11:24:55 hi 2023-07-17 11:25:16 i have some questions regarding getting filles from the HDD 2023-07-17 11:26:04 1. some folders are written in a language that alpine cant render (arabic), which means i cant just cp the files from there to the USB because the unicode is broken 2023-07-17 11:26:38 2. how do i determin the size of the files in GB because the thumb drive i have can only handle 8 GB at a time 2023-07-17 11:26:49 grub does support luks2 but not argon2i kdf caskd 2023-07-17 11:27:29 (assuming latest release 2.06 without any additional functionality patches /glares at arch and debian) 2023-07-17 11:28:51 blt[m]1: could you run cryptsetup luksDump /dev/YOUR_DEVICE_HERE from a live ISO and give us the output via a pastebin? 2023-07-17 11:29:10 make sure you use the /boot for this, not your actual root 2023-07-17 11:29:48 caskd: who? me? 2023-07-17 11:30:12 akndjd: no, this is for the other person i mentioned 2023-07-17 11:30:19 oh ok 2023-07-17 11:31:07 akndjd: you should still be able to copy the files regardless of encoding across the drives but i am not sure if that would break the metadata so double check that 2023-07-17 11:31:25 regarding directory size, use `du -sh ./DIRECTORY` 2023-07-17 11:31:41 it may take a bit depending on how big the directory is 2023-07-17 11:47:05 moum\\hey caskd orry to bother you, but i cant copy anything from the hard drive 2023-07-17 11:47:16 it says cp: read error: i/o error 2023-07-17 11:47:37 do you have any idea what might cause that? 2023-07-17 11:48:11 i mounted the system by running: ntfs-3g /dev/sdX /mnt/windows 2023-07-17 11:48:30 i created the folder "windows" beforehand 2023-07-17 11:56:53 i think your drive might be dying 2023-07-17 11:57:07 which might explain the reason why it doesn't boot 2023-07-17 11:57:49 try installing smartmontools and running smartctl -a /dev/sdX 2023-07-17 12:00:11 if the SMART overall-health section shows anything but PASSED then i'd recommend you to run ddrescue to a drive that has just as much or more data 2023-07-17 12:00:25 more storage available* 2023-07-17 12:49:18 blt[m]1: how was your system originally installed? did you use setup-disk (including for encryption) or did you do part of it manually? 2023-07-17 12:51:16 minimal: I created/followed this tuto : https://wiki.alpinelinux.org/wiki/Full_disk_encryption_secure_boot 2023-07-17 12:51:20 blt[m]1: actually I see you mentioned Grub FDE, so you did the LUKS setup manually. How exactly did you set up LUKS? v1? v2? which options etc 2023-07-17 12:53:12 caskd: luks2 for FDE does not actually work in Grub, there are bugs that prevent it working, it should be fixed in the next upstream release (2.12rc1 recently came out) 2023-07-17 12:55:05 minimal: oh, okay 2023-07-17 12:55:38 blt[m]1: that wiki article looks flawed 2023-07-17 12:56:10 minimal: Could you guide me why/where? 2023-07-17 12:57:09 i.e. it says to create the LUKS via "cryptsetup ... luksFormat" but doesn't specify a luks version - recent versions of cryptsetup default to LUKSv2 2023-07-17 12:57:48 yet later in the article it says a add GRUB_PRELOAD_MODULES="luks ..." - the "luks" module is for LUKSv1, not LUKSv2 2023-07-17 12:59:09 the GRUB_PRELOAD_MODULES normaly should not be needed, however here it appears to be used as a partial workaround for Grub's problems with LUKSv2 FDE 2023-07-17 13:00:42 Yes I have used the GRUB_PRELOAD as a workaround with the current issue with luksv2 FDE 2023-07-17 13:01:09 blt[m]1: but it is not a complete solution 2023-07-17 13:01:22 I would like to use a normal grub-install but we have to wait the fix upstream 2023-07-17 13:01:26 plus that value in the Wiki article is wrong as I already said 2023-07-17 13:02:24 minimal: Which is the correct one 2023-07-17 13:02:42 yet later in the article it says a add GRUB_PRELOAD_MODULES="luks ..." - the "luks" module is for LUKSv1, not LUKSv 2023-07-17 13:03:25 For luksv2 2023-07-17 13:03:37 ? 2023-07-17 13:04:26 blt[m]1: are you the author of that wiki article? 2023-07-17 13:04:59 Yes 2023-07-17 13:05:16 That's what I am using in my current setup 2023-07-17 13:05:29 As a workaround of a fully working luks2 fde 2023-07-17 13:06:17 so how did it work for you previously when the article is incorrect? 2023-07-17 13:09:15 Good question... We are two who are currently using it like it is described in the article.. 2023-07-17 13:09:33 Do you think changing the module for luks v2 could be enough ? 2023-07-17 13:10:09 that only works around the bug in grub-probe 2023-07-17 13:10:38 there were multiple fixes required in upstream Grub 2.12 to fix the problems with Grub 2.06 LUKS2 FDE 2023-07-17 13:11:58 So I may update the article when it will be available... You think Alpine would add 2.12 soon after available in upstream? 2023-07-17 13:13:46 to Edge yes, typically not to other Alpine versions 2023-07-17 13:15:04 I think your article is not a good idea as Grub LUKS2 FDE is currently "problematic" 2023-07-17 13:19:40 Don't know if I can put it back as draft ,I will put a note at the beginning and will update it as soon as 2.12 is available 2023-07-17 13:20:11 blt[m]1: which version of Alpine did you use to originally setup LUKS on? 2023-07-17 13:20:38 I think it was 3.16 or 3.17 2023-07-17 13:20:48 it is possible you used an Alpine release where "cryptsetup luksFormat" still defaulted to LUKSv1, not LUKSv2 2023-07-17 13:21:13 And after an update I think 3.18.2 not able to boot anymore 2023-07-17 13:21:29 But was able to after 3.18 2023-07-17 13:22:02 minimal: And you think after the update recently it is now on luksv2 so my issue 2023-07-17 13:22:19 It start to work around 10 days qgo 2023-07-17 13:23:19 no, there is nothing in Alpine to automatically convert v1 to v2 2023-07-17 13:24:21 basically I do not see how those Wiki instructions would ever work to booting using GRUB LUKS2 FDE 2023-07-17 13:24:46 the GRUB_PRELOAD_MODULES workaround does not resolve all issues 2023-07-17 13:25:10 I am pretty sure the luksv2 is the one used on my partition, will double check when I boot from iso 2023-07-17 13:25:37 and again, the GRUB_PRELOAD_MODULES workaround does not resolve all issue 2023-07-17 13:25:57 so I do not see how it ever worked 2023-07-17 13:26:15 unless you did some additional stuff that is not mentioned in that Wiki article 2023-07-17 13:28:38 Nope I did only what is on 2023-07-17 13:30:02 so I do not understand how it worked as there will be other important stuff missing from the generated grub.cfg 2023-07-17 13:31:38 so I expect when you updated to 3.18.x that the grub.cfg was regenerated and lost some settings you had previously manually added to it 2023-07-17 13:51:38 blt[m]1: can you pastebin (or equivalent) your grub.cfg file? 2023-07-17 13:58:59 minimal: OK when I come back in the evening , thanks for your help 2023-07-17 14:09:30 is it possible to map arrow-keys to altgr+hjkl and have them work with alt combinations? eg having problem making alt+altgr+l work in irssi to switch window 2023-07-17 15:32:34 I have started service pdns 2023-07-17 15:32:54 local-port=53 is in the config 2023-07-17 15:33:25 But netstat -lt returns no process listening on port 53 2023-07-17 15:33:53 netstat -ul 2023-07-17 15:35:12 returns nothing 2023-07-17 15:36:21 Is the service actually running? 2023-07-17 15:37:21 Yes, top shows it, and rc-status too show it to be running 2023-07-17 15:38:01 --guardian=yes . What is that config? Is it causing the problem? 2023-07-17 15:38:15 I'm not too familiar with pdns itself 2023-07-17 15:38:17 Guarding from listening on 53? 2023-07-17 15:42:38 why would that make sense for a dns server? 2023-07-17 15:43:29 "Run pdns_server inside a guardian. This guardian monitors the performance of the inner pdns_server instance. It is also this guardian that pdns_controltalks to." 2023-07-17 15:43:36 I don't know. I'm just asking about guardian keyword 2023-07-17 15:58:59 Can anyone recommend ZFS Boot Menu for Alpine Linux? https://docs.zfsbootmenu.org/en/v2.2.x/guides/alpine.html 2023-07-17 16:03:10 hello 2023-07-17 16:03:33 quick question, i cant instal cerbot on alpine 3.18 even tho its in the repo 2023-07-17 16:03:34 y? 2023-07-17 16:05:32 nor ffmpeg or imagemagick 2023-07-17 16:05:47 how is it failing? 2023-07-17 16:05:51 did u enable the community repository? 2023-07-17 16:06:52 yes 2023-07-17 16:06:55 i think 2023-07-17 16:07:08 i ran this 2023-07-17 16:07:09 awk 'NR==2' /etc/apk/repositories | sed 's/main/community/' | tee -a /etc/apk/repositories 2023-07-17 16:07:49 check the contents 2023-07-17 16:08:04 you might have ran it as a normal user instead of root or whoever can write to it 2023-07-17 16:09:34 Where do I find logs of pdns? 2023-07-17 16:09:48 Where do I find syslog logs? 2023-07-17 16:10:13 yup 2023-07-17 16:10:17 it ws the issue 2023-07-17 16:14:57 Where do I find syslog logs? 2023-07-17 16:16:04 /var/log 2023-07-17 16:16:24 I have pdns on, where do I find it's logs 2023-07-17 16:16:39 There's no entry regarding pdns in /var/log 2023-07-17 16:16:45 is there anything else in /var/log ? 2023-07-17 16:16:49 have you read the man pages of pdns 2023-07-17 16:17:00 singham does not have a pdns problem, yet 2023-07-17 16:17:03 singham has a logging problem first 2023-07-17 16:17:09 the pdns problem cannot be debugged until that is solved 2023-07-17 16:17:16 (singham also asked some questions in #powerdns) 2023-07-17 16:17:22 Yup, found 2023-07-17 16:17:27 Thanks folks 2023-07-17 16:17:31 if pdns keeps its own logs, it should explain this in its man page(s) 2023-07-17 16:17:40 logs are in messages file 2023-07-17 16:18:10 Habbie: Stop beating me up man! 2023-07-17 16:18:20 Go easy on this kid here :P 2023-07-17 16:18:21 i'm not beating you up 2023-07-17 16:18:38 i'm clarifying your problem for the audience 2023-07-17 17:41:21 thxquestion 2023-07-17 17:41:37 can i install alpine x86_64 on 32 bit systems 2023-07-17 17:41:46 like, will they work fine as well? 2023-07-17 17:44:02 x86_64 things cannot run on 32 bit systems 2023-07-17 19:26:02 oes ^y, ^t and ^v not work in ash? 2023-07-17 19:26:10 *does 2023-07-17 19:26:44 nope 2023-07-17 19:28:16 weird to leave those out i guess .. 2023-07-17 19:28:21 oh well .. 2023-07-17 19:28:29 It's missing more 2023-07-17 19:28:44 ^. as well 2023-07-17 19:28:54 sorry, alt+. 2023-07-17 20:25:53 is there a webcam recording tool other than cheese in the repo? it seems to not be working with pipewire 2023-07-17 20:26:54 bl4ckb0ne, i like guvcview 2023-07-17 20:27:29 (but i have no idea about pipewire) 2023-07-17 20:30:33 can it do preview? 2023-07-17 20:31:16 i use it for viewing and sometimes recording 2023-07-17 20:31:20 but perhap you should just try it 2023-07-17 20:31:27 perhaps 2023-07-17 20:31:54 i mean view what you are recording 2023-07-17 20:35:04 oh, i think so 2023-07-17 20:51:47 helo 2023-07-17 20:52:34 idk if this is off topic, but im having trouble making my domain point to my local server 2023-07-17 20:52:45 its running alpine so i decided to ask here 2023-07-17 20:52:53 also im new to sysadmin stuff 2023-07-17 20:53:57 you have a public IP on your local? 2023-07-17 20:54:39 ffog: yes i think so 2023-07-17 20:54:46 its online rn 2023-07-17 20:55:11 are you using ssl? 2023-07-17 20:56:06 supposedly yeah 2023-07-17 20:56:12 how do i check it again? 2023-07-17 20:56:27 what are you tryin to serve, http? 2023-07-17 20:56:31 yeah 2023-07-17 20:56:33 a site 2023-07-17 20:57:12 you might need to open port 80 on your router for the server, something like this 2023-07-17 20:57:41 ok, what then/ 2023-07-17 20:57:59 then point your domain at your local IP 2023-07-17 20:59:18 the local ip? 2023-07-17 20:59:24 i have a bug on my setup where it works fine from the outside but locally i have to use the ip 2023-07-17 20:59:26 like 192.168 2023-07-17 20:59:35 192.168.x.x 2023-07-17 20:59:36 no thats just within your local network 2023-07-17 20:59:41 you can do like.. curl ipv4.icanhazip.com 2023-07-17 20:59:52 oh ok 2023-07-17 21:02:06 well, it opens my router page 2023-07-17 21:02:31 hah woops 2023-07-17 21:03:08 i think its a router config sorta issue, it depends on the router 2023-07-17 21:03:25 is it like a comcast type ISP thing? 2023-07-17 21:03:33 some routers will serve the management page if you visit the outside IP from inside 2023-07-17 21:03:40 but will still map your port correctly for outside visitors 2023-07-17 21:03:47 the setting to change that might be called 'hairpin NAT' 2023-07-17 21:04:33 i dont have any of that 2023-07-17 21:05:00 you can test visiting it from outside your network to see 2023-07-17 21:05:02 and you don't need it 2023-07-17 21:05:04 exactly that 2023-07-17 21:06:24 doesnt workfrom outside 2023-07-17 21:07:22 theres a lot of things to investigate probably, none of which are Alpine specific I don't think 2023-07-17 21:07:30 agreed 2023-07-18 10:35:32 minimal: https://paste.chapril.org/?e35689c67c625935#C48hRRhGs7wQQAvVqswtnnpE4GoVGA5ndUPLTMpVvHuw here we go. I also tried with luks2 as premodule but got the same issue 2023-07-18 12:35:48 minimal : I suspect that it is a result of 3.18.2 update where mkinitfs had been updated to a new version, it was working fine from 3.16.1 until 3.18.1 2023-07-18 12:36:27 the point releases don't mean anything for upgrading 2023-07-18 12:36:44 unless you only upgrade on the same day as them or something 2023-07-18 12:37:21 Correct but as I use to check almost every other day it could match 2023-07-18 12:38:05 But yes I also noticed that cryptsetup was also updated 2023-07-18 12:38:59 there was no upgrades for cryptsetup at any point after release 2023-07-18 12:39:28 did you forget to use -a when changing release repositories? 2023-07-18 12:39:54 What is strange to me is that I would like to confirm it is due to LUKS2 and wanted to convert my LUKS2 to v1 but got a message where the slot is not compatible and I only got one key/slot with pbkdf2 which is supposed to be compatible with luksv1 2023-07-18 12:40:24 psykose: No, I always do --available between major 2023-07-18 12:40:47 what upgrade are you talking about then 2023-07-18 12:40:51 there wasn't one 2023-07-18 12:44:38 I was referring to a conversation I had with minimal yesterday, where my system doesn't boot anymore and this match with an upgrade, I believe 3.18.2 , and I sent him the wiki link I created and he wanted to see my grub.cfg as be didn't understand how it would ever work with the workaround I used for luks2 and grub fde. I know that grub 2.12 rc1 will finally fix a lot of bug and it is the way to go but I try to make my system available 2023-07-18 12:44:38 in the meantime as I am not able to boot anymore. (Just to give you the context) 2023-07-18 12:45:18 i already know the context 2023-07-18 12:45:36 "cryptsetup was also updated" isn't a thing that ever happened on 3.18 so it seems weird 2023-07-18 12:48:39 Maybe I was wrong for that but I can't see how it was working and it doesn't anymore 2023-07-18 12:49:05 mkinitfs could be the reason 2023-07-18 12:50:39 Another friend has the same issue with a fresh install on 3.18 so the wiki we wrote is definitely not OK but I try understand why and I want to prove it is because luks2 but can't convert to luks1 2023-07-18 12:52:21 the only change in mkinitfs since release is https://gitlab.alpinelinux.org/alpine/mkinitfs/-/commit/317c1ec274c84e39e3f777a81e6484aa768e58ef 2023-07-18 12:52:35 which is meaningless 2023-07-18 12:52:58 Indeed... 2023-07-18 12:53:21 blt[m]1: that grub.cfg is missing at least 1 required entry for LUKS (whether v1 or v2) to work - this is bug to one of the problems I have already explained regarding LUKv2 *not* working with Grub FDE 2023-07-18 12:53:27 So am puzzled 2023-07-18 12:54:12 as I said before, I suspect you had a modified grub.cfg and then an upgrade caused grub-mkconfig to regenerate it, so losing those changes 2023-07-18 12:54:19 minimal: Thanks I tried also luks2 instead and same issue, would you recommend me to add something else 2023-07-18 12:54:57 blt[m]1: I never said changing "luks" to "luks2" in the PRELOAD list would fix the issue, only that it work me "more correct" to specify that 2023-07-18 12:55:10 I did *not* expect it to work after you changed luks to luks2 2023-07-18 12:55:25 Ha OK I misunderstood sorry 2023-07-18 12:55:39 Would you see anything else I can try? 2023-07-18 12:56:05 try? switch to LUKSv1 which works 2023-07-18 12:57:52 you say following the steps in the Wiki article works, I disagree 2023-07-18 12:57:56 minimal: Yes, that's exactly what I am trying to do, but converting is not working for me... And I don't understand as I only got one slot remaining on pbkdf2 2023-07-18 12:59:07 I swear I followed it step by step and it worked but now I can't say why 2023-07-18 12:59:36 well said I said there are things missing from that Wiki article so I cannot see how it would ever work 2023-07-18 13:00:39 And I want to convert it luks1 but got an issue where crypsetup is complaining about my slot is not luks1 compatible 2023-07-18 13:08:41 blt[m]1: well that obviously depends on settings chosen (or automatically selected) when LUKS was originally setup 2023-07-18 13:12:15 minimal: I did cryptsetup -v -c aes-xts-plain64 -s 512 --hash sha512 --pbkdf pbkdf2 --iter-time 1000 --use-random luksFormat /dev/nvme0n1p2 2023-07-18 13:24:59 blt[m]1: I have no idea why you cannot convert back to LUKS1 2023-07-18 13:27:56 going back to your original problem, the grub.cfg you provided is missing "insmod gcry_rijndael", "insmod gcry_sha512" 2023-07-18 13:28:09 it is also missing "cryptomount -u " 2023-07-18 13:28:51 and it is also missing "insmod luks2" 2023-07-18 13:29:19 these are all due to the bugs that I have mentioned multiple times that Grub 2.06 has that cause LUKSv2 FDE to not work 2023-07-18 13:30:49 "insmod ext2" also missing (in Grub terms ext2 = ext2/3/4). Your grub.cfg instead has "insmod ext4", I do not know if this is valid. 2023-07-18 13:31:28 minimal: thanks for that ... I will try and see if I have any chance adding them, at least as a temporary solution before 2.12 rc1 available in edge to resinstall everything clean 2023-07-18 13:32:16 correction, "insmod ext2" *is* present further down in the file 2023-07-18 13:32:30 remove "ext4" completely from the PRELOAD definition 2023-07-18 13:32:59 note I am *still* not saying these changes will enable LUKSv2 FDE to work 2023-07-18 13:35:58 OK so removing ext4 only ? (Changing luks to luks2 I guess?) 2023-07-18 13:36:30 I already told you to change "luks" to "luks2" yesterday 2023-07-18 13:36:38 Ok 2023-07-18 13:36:57 Yes it is on my new grub.cfg jus wanted to confirm 2023-07-18 13:37:10 "going back to your original..." <- And I imagine that as well 2023-07-18 13:37:31 blt[m]1: that Wiki article does mention setting "cryptomount -u" *BUT* it says to put it in a /root/grub-pre.cfg file. Eek, it is using that as a "early" config file 2023-07-18 13:38:42 how long has package pinning been a thing for? just discovered it and it's fantastic at helping me install packages only in edge/testing when running latest-stable 2023-07-18 13:39:19 Since forever 2023-07-18 13:39:26 minimal: Yes was about to tell you that I have a separate file for that. So I will add the two insmod you mentioned 2023-07-18 13:39:30 And that is not supported 2023-07-18 13:39:49 eddsalkield: things may break due to upgrades 2023-07-18 13:40:46 the pattern of pinning edge packages on a stable version is demonstrated on the wiki 2023-07-18 13:40:47 https://wiki.alpinelinux.org/wiki/Alpine_Package_Keeper#Repository_pinning 2023-07-18 13:43:01 seems like the policy of preferring pulling in dependencies from non-tagged repositories is designed to partially remediate the issue of things breaking during upgrades 2023-07-18 13:44:09 it doesn't remediate that at all 2023-07-18 13:44:38 blt[m]1: to summarise, you're doing something non-standard in the way you are trying to get LUKS2 FDE working with Grub, so please do not complain if it doesn't work or works for a while and then breaks 2023-07-18 13:45:27 blt[m]1: and I do not think that Wiki article should exist as it encourages other people to do non-standard things as well 2023-07-18 13:47:12 psykose: won't it ensure that if you install bleedingapp@testing then you won't end up pulling in dependencies that break other non-pinned packages on the system? So the only package that could break is bleedingapp@testing 2023-07-18 13:47:38 in that sense sure 2023-07-18 13:48:02 in the "i had to close 40 bug reports in the past year where people did that and it broke something", uhh 2023-07-18 13:48:03 minimal: I got that and I am not complaining at all I am only trying to ask for help to understand why and to temporarily have a working system until 2.12 rc1 would be available. 2023-07-18 13:48:03 Re wiki I will put a disclaimer and will update it as soon as 2.12 rc1 will be available as I told you yesterday 2023-07-18 13:48:57 psykose: that seems sufficient for the use case of installing new software only in testing then - but I can see why we don't want to officially support it 2023-07-18 13:49:19 how much effort would it be to build latest-stable/testing packages? 2023-07-18 13:49:25 clearly there's a demand 2023-07-18 13:50:07 a lot 2023-07-18 13:51:06 eddsalkield: testing is not intended for normal releases, the correct mechanism is for packages in testing to migrate to community and then end up in the next Alpine release 2023-07-18 13:51:12 doing that also doesn't work most of the time 2023-07-18 13:51:21 i.e. for "new software only in testing" 2023-07-18 13:52:00 Isn't part of the problem that packages do not really move from testing that much? For example ibus-anthy is in testing for like forever despite working fine. 2023-07-18 13:52:21 somewhat 2023-07-18 13:53:06 that is a separate issue though, some stuff is intentionally not in stable and edge-only 2023-07-18 13:53:55 and for as long as there is >0 packages doing that, someone is going to try do this dumb thing 2023-07-18 13:54:18 and then email me directly from their work company email account blaming me for "their build script somewhere adding things from edge/testing dont work" 2023-07-18 13:56:28 Cannot speak for others, but for me personally, all packages I port from testing to my repo are in testing for 3+ release cycles. I many people actually need something edge/testing-only. But I feel your pain, I am always surprised how many tutorials and answers just basically say "enable edge/testing and install it from there". 2023-07-18 13:56:52 I wonder how many* 2023-07-18 13:57:22 one could always open a merge request to just move it for most things 2023-07-18 13:57:55 graywolf: unfortunately many people treat Alpine Wiki articles as "official" when any "random" person can write an article which may or may not be correct 2023-07-18 13:59:51 psykose: I tried it few time, but then the mailing-list - gitlab integration broke... and I never properly figured out the gitlab way :/ 2023-07-18 14:00:40 minimal: Not sure there is a reasonable solution for that proble 2023-07-18 14:14:17 hi. me again. maybe o.t. if so, i'm sry, don't answer me. does anyone see a connection to 34.104.35.123 at chromium startup? i have the feeling, it's malicious. even happens in a vm, fresh chr. install, no add-on. for more: https://gist.github.com/llightcb/6878a8e334bc777e18863ebc678d8cc8#file-example_dnscrypt_proxy_blocklist-txt-L6 2023-07-18 14:16:37 The ip address does belong to Google 2023-07-18 14:17:24 well it's for Google customers (according to Whois info) 2023-07-18 14:18:35 123.35.104.34.bc.googleusercontent.com . google cloud. something. look at this: https://www.abuseipdb.com/check/34.104.35.123 2023-07-18 14:20:23 ..and the gist link with full drill stuff. 2023-07-18 14:20:42 "hi. me again. maybe o.t. if so..." <- Why do you think it's malicious 2023-07-18 14:21:39 oleovit: you have a drill for the IP the wrong way around "drill -x 123.35.104.34" 2023-07-18 14:21:43 port 80, https://www.abuseipdb.com/check/34.104.35.123 , no samsung whatever 2023-07-18 14:22:29 that's because a lookup *the wrong way around* shows Samsung 2023-07-18 14:24:27 and it blocks 34.104.35.123 2023-07-18 14:24:46 with *.nic.or.kr 2023-07-18 14:25:38 in dnscrypt. thats why i used the 123.35.104.34 part of 123.35.104.34.bc.googleusercontent.com 2023-07-18 14:26:21 what has *.nic.or.kr got to do with 34.104.35.123 ? 2023-07-18 14:28:33 i dont know. but as i said, i wanted to know what the 123.35.104.34 part of 123.35.104.34.bc.googleusercontent.com is. drill it, see inverse.nic.or.kr, block it in dnscrypt, no connection to 34.104.35.123 at chromium startup 2023-07-18 14:29:24 "i wanted to know what the 123.35.104.34 part of 123.35.104.34.bc.googleusercontent.com is" - it is a string 2023-07-18 14:29:30 so, https://www.abuseipdb.com/check/34.104.35.123 is telling me something. and i wanted to know if anyone else has this connection to some google cloud 2023-07-18 14:30:37 of course i drilled the string too! 2023-07-18 14:30:45 thats not the point 2023-07-18 14:30:50 it is "standard practice" for the reverse lookup of an IP address to contain the IP in reverse (due to the way in-addr.arpa. works) 2023-07-18 14:31:37 you are trying to find a connection between 34.104.25.123 and 123.25.104.34 where there is none 2023-07-18 14:31:51 they already moved past that point 5 messages ago 2023-07-18 14:33:14 vaguely related, is DoH the default in chromium now, as it is in chrome? 2023-07-18 14:34:10 just curious. since the dnscrypt block wouldn't matter if the browser is doing its own lookups 2023-07-18 14:39:56 sry, was away. chromium doh is not default, as far as i remember, and i have flag #enable-async-dns 2023-07-18 14:41:19 disabled. again, i drilled everything, of course the whole string too, and blocking this korean thing is blocking https://www.abuseipdb.com/check/34.104.35.123 . 2023-07-18 14:42:19 but i guess, its really o.t., so sry. 2023-07-18 14:43:46 i think #chromium is on libera, but google's page for it says freenode, so who knows. 2023-07-18 14:45:14 #chromium could be the place. answers from #security in libera and some related "security" channels have only the standard answrs: use firefox blabla. 2023-07-18 14:46:28 they might not be wrong :) 2023-07-18 14:49:09 oh, imho, mozilla is much worse. and has much more "spyware" stuff on-board. i remember a discussion about this some years ago on alpines git. and if madain is right in one thing: ff is privacy-theatre. in terms of security, chromium is the right choice. done with mozilla (used since 2005) 2023-07-18 14:50:11 they're all bad, and browsers are probably too complex for anybody to really know. 2023-07-18 14:52:38 of course, but we dont have that much choices :( 2023-07-18 14:54:05 "a strange game. the only winning move is not to play" - WOPR 2023-07-18 14:58:24 anyway, blocking "korea" works, there is a relation, the original ip seems to be malicious, and it happens on a fresh install. this google cloud connections has nothing to to with chromium itself in the context of all the other connections at startup (which can also be reduced with various flags). i guess, there is no "exit" out of this game, invoked 2023-07-18 14:59:37 if you don't know what it does then you can't know it's malicious 2023-07-18 15:07:29 only stuff happening on port 80 and abuseipdb.com is not to neglect, psykose. anyway, i stop to spam my beloved alpine channel. but anywhere i ask, only typical "use ff" answers. it's frustrating. thx for your work anyone! 2023-07-18 15:08:17 ..a lot of "any". my written en is horrible,sry 2023-07-18 16:01:50 minimal: just for your info I removed ext4 from preload and added the two you suggested but same result, by the way the two gcry_ are in the wiki but in grub-mkimage and not in preload 2023-07-18 16:06:44 (By the way I added the disclaimer on top of the wiki page) 2023-07-18 16:17:16 blt[m]1: I keeping coming back to say that there are issues with Grub FDE LUKS2. I don't know what else you expect me to say 2023-07-18 16:18:32 If you manage to get it to work somehow then congratulations. If you manage to get it to keep working then congratulations. If it somehow works for a while and then breaks......well don't say I didn't warn you. 2023-07-18 16:20:34 minimal: nothing, I was just letting you know as I made some test and as you suggest some things to test, just that I made it and you didn't spent your time for nothing, take it easy man 2023-07-18 19:56:55 current /a 2023-07-19 02:12:39 quick question, I am using cups (print standard) and would like to change the default-directory (.cups) 2023-07-19 02:15:03 i read the man, and don't see a environment-variable to change it. so im asking here to see if anyone has an idea 2023-07-19 02:15:09 something happened on edge, foot randomly crashes periodically. segfaults in libwayland-client.so.0.22.0 2023-07-19 02:16:05 this is when alpine is used in a container and given wayland socket access. 2023-07-19 02:17:04 so far, i've seen 'CUPS_(CACHE|DATA|STATE)DIR'. none for the default-directory, unless one of these apply 2023-07-19 02:25:06 Armote[m]: sounds like an issue with the new foot version 2023-07-19 02:25:32 you'd have to get a core and backtrace it 2023-07-19 02:30:30 Guest6394: this is a fair point. mailing-lists tend to have a long-delay and i took my chances asking here 2023-07-19 02:31:49 im using alpine, and people here partake in all sorts of packages, so i flipped the coin 2023-07-19 02:34:25 mervin: https://github.com/OpenPrinting/libcups/issues/43 2023-07-19 02:34:39 3.0 will.. use ~/.config/cups if it exists i guess, no toggles https://github.com/OpenPrinting/libcups/pull/45 2023-07-19 02:34:41 eventually 2023-07-19 02:40:03 psykose: thanks for the reply. glad to see it being rolled out 2023-07-19 02:44:09 will also check github next time before asking, good idea 2023-07-19 07:03:43 mmmeeeeoooowww 2023-07-19 07:06:11 meow 2023-07-19 07:06:20 purrr... 2023-07-19 07:06:32 Is apline worth checking out? 2023-07-19 07:06:43 Brrmeow! 2023-07-19 07:06:51 justbecauseican: IMO yes 2023-07-19 07:07:07 I tried to install it - but it got stuck on some ancient looking MSDOS screen - and did absolutely nothing 2023-07-19 07:26:13 justbecauseican, that's because you had booted MSDOS instead of Alpine Linux 2023-07-19 07:29:58 lol 2023-07-19 09:50:51 Hi anyone here 2023-07-19 10:34:42 hi 2023-07-19 11:24:52 Hello 2023-07-19 11:41:47 I want to enable dnssec for pdns 2023-07-19 11:41:55 When I try 2023-07-19 11:42:06 pdnsutil secure-zone example.com 2023-07-19 11:42:36 It throws an error that gsqlite3-dnssec backend not here. Try with that flag. 2023-07-19 11:42:50 Not sure it this is important, but on same pages copyright notice is outdated (f.g. https://www.alpinelinux.org/about/ and couple of others) 2023-07-19 11:42:59 Is the default package compiled without that support? 2023-07-19 11:52:11 Any pdns users? 2023-07-19 12:11:00 alternateved: it does not really matter 2023-07-19 12:12:06 Thought so 2023-07-19 12:12:32 Any replies on my query? I was offline for a while 2023-07-19 12:13:04 Hello there. I have a question regarding this bug. https://gitlab.alpinelinux.org/alpine/aports/-/issues/15103 2023-07-19 12:13:30 Is there something I can support you with to get this thing going? 2023-07-19 12:13:58 Or is this matter of building community packages more of "it is some sort of best effort" compared to regular packages? 2023-07-19 12:15:05 Don't get me wrong. I want just to get the bug solved. No hard feelings here. 2023-07-19 13:02:15 Bid put a comment to the bugreport. Hoping that someone will reply if needing help. https://gitlab.alpinelinux.org/alpine/aports/-/issues/15103 2023-07-19 13:02:27 cya 2023-07-19 13:06:04 So, maybe running on Edge wasnt such a good idea afterall 2023-07-19 13:06:14 my external monitors wont work anymore :( when i connect them through USB-C dock 2023-07-19 16:45:56 Hi, I wonder how to format Alpine Linux. Can someone help me ? 2023-07-19 16:46:31 format what? 2023-07-19 16:47:35 Reset like if I just start Alpine for the first time. 2023-07-19 16:48:04 Sorry for my bad English i'm frenche 2023-07-19 16:48:10 french* 2023-07-19 16:49:35 if you still have your issue, I speak French. I'm on vacations right now, but you can pm me if you want. I'm back tomorrow 2023-07-19 16:50:16 lafamilia: je parle français, mais je suis en vacances. Je reviens demain. Tu peux m'envoyer un message si tu as encore ton problème 2023-07-19 16:51:20 No I don't have any issue I just want to know if there was a simple way to reset Alpine. 2023-07-19 16:51:57 There is no simple/trivial way to reset everything again 2023-07-19 16:53:01 I installed Alpine by a Usb key, if I re-insatll that will not reset everything ? 2023-07-19 16:54:05 sure 2023-07-19 16:54:46 Just run the installer again 2023-07-19 16:55:58 There is an integrated installer or I need to do eveything againe with my usb key ? 2023-07-19 16:56:06 again* 2023-07-19 16:56:32 It's not trivial to overwrite a running system, as the partitions are in use 2023-07-19 16:57:14 but alpine-conf is installed by default, so you can run setup-alpine 2023-07-19 16:58:36 Yes I already did that (run setup-alpine) but when I'm asked how I want to install Alpine and I choose sys, it doesn't propose my "main" partition... 2023-07-19 16:59:22 because it is mounted 2023-07-19 17:01:04 so how can I do ? Can I "un-mounted" it ? 2023-07-19 17:02:10 How you supposed to execute setup-alpine then? 2023-07-19 17:02:37 Yes I understand that was a selly question 2023-07-19 17:03:33 So if I do the installation with my usb key, the alpine installed on my disk will be not running, am I right ? 2023-07-19 17:03:39 yes 2023-07-19 17:03:52 Okay so that will be good 2023-07-19 17:04:04 Thank you for your help ikke ! 2023-07-19 17:32:02 openssh security update, how exciting :) 2023-07-19 17:34:23 yeah, edge and 3.18 are already updated 2023-07-19 17:37:12 I saw! thank you 2023-07-19 17:38:22 fastest security updates ever over here 2023-07-19 17:49:36 Does apk have a man page? Or is the user-handbook the only documentation? 2023-07-19 17:51:17 https://pkgs.alpinelinux.org/package/edge/main/x86_64/apk-tools-doc 2023-07-19 17:51:27 https://pkgs.alpinelinux.org/contents?branch=edge&name=apk%2dtools%2ddoc&arch=x86_64&repo=main 2023-07-19 17:52:01 Perfect, thanks 2023-07-19 17:53:32 bahamat: its in apk-tools-doc 2023-07-19 17:57:07 happy that this channel is snappy. i was too late 2023-07-19 17:58:40 lot of ircs take long to answer but alpine is on-point 2023-07-19 18:04:01 Alpine is not snappy. It uses Flatpak if anything. 2023-07-19 18:04:24 I don't think we even use flatpak 2023-07-19 18:04:31 we're more of a "just install it" group 2023-07-19 18:04:49 malware.apk... JUST INSTALL IT 2023-07-19 18:04:51 i use elf 2023-07-19 18:06:06 XFaCiEer: exactly, that's the spirit 2023-07-19 18:06:38 malware? apk can install that. a 9 year old gtk app with 4 open remote code exec bugs? apk can install that. wayland and X at the same time? apk can install that 2023-07-19 18:06:47 anything you want, apk can install :) 2023-07-19 18:12:28 apk add glibc 2023-07-19 18:12:39 okay, that's a bridge too far :P 2023-07-19 18:15:47 musl (-_-) : glibc (OwO) 2023-07-19 18:17:25 proprietary c-lib (*o*) 2023-07-19 20:41:42 What is mdev? 2023-07-19 20:42:41 bahamat: https://wiki.gentoo.org/wiki/Mde 2023-07-19 20:43:00 err, maybe not the best link 2023-07-19 20:43:14 but anyway, it's a replacement to (e)udev from busybox 2023-07-19 20:43:31 Do I need that in a container? or can I omit it? 2023-07-19 20:43:53 also, what mounts /run? 2023-07-19 20:50:56 no 2023-07-19 21:00:58 bahamat: /run can just be a directory, but many distros will mount it as tmpf (in fstab) 2023-07-19 21:01:07 but for containers that's not relevant 2023-07-19 21:24:28 ikke: An lxc container has /run on tmpfs, but a docker container doesn't. 2023-07-19 21:24:35 I'm trying to figure out the difference. 2023-07-19 21:24:57 an lxc container is more or less statefull, while a docker container (typically) is not 2023-07-19 21:25:40 I get that, but I'm trying to figure out what mechanism it is that does the mount. 2023-07-19 21:25:53 the localmount service 2023-07-19 21:25:58 docker container don't run services 2023-07-19 21:26:03 containers* 2023-07-19 21:26:58 Where is the needed/wanted runlevel? 2023-07-19 21:27:10 those are not runlevels 2023-07-19 21:27:47 Does that just mean it's a dependency of something else? 2023-07-19 21:28:39 It has a dependency on something else 2023-07-19 21:29:34 you're asking a lot of questions that don't really make any sense, which suggests your actual issue is something completely unrelated and you just ended up on this step 2023-07-19 21:30:47 I'm creating a container image for a cloud system using the docker image as a base. 2023-07-19 21:30:56 So there are some "missing parts" that I'm filling in. 2023-07-19 21:32:44 bahamat: maybe take a look at https://github.com/lxc/lxc-templates/blob/master/templates/lxc-alpine.in 2023-07-19 21:33:26 but if you want to actually boot it, you need even more 2023-07-19 21:34:47 Like what? 2023-07-19 21:35:01 a kernel for example 2023-07-19 21:35:25 unless the hypervisor specifies the kernel itself 2023-07-19 21:35:28 Oh. Yeah, I've got that covered already. 2023-07-19 21:37:54 The last big thing I need to take care of is /run. 2023-07-19 21:44:38 why would a container image have a kernel? 2023-07-19 21:45:01 minimal: it's not for a container 2023-07-19 21:45:21 they just decided (for whatever reason) that they want to create a vm image from a container image 2023-07-19 21:45:25 ikke: "" I'm creating a container image for a cloud system using the docker image as a base. 2023-07-19 21:45:42 note "a container image" 2023-07-19 21:46:00 It's for illumos running in LX brand zones. 2023-07-19 21:46:12 The illumos kernel does linux system call translation. 2023-07-19 21:46:37 So no kernel or networking configuration needed, that's all handled by the zones system. 2023-07-19 21:48:43 Doing `docker cp $container:. - > $tar` gives us more control over what's in the image than trying to use the lxc images, and we can do the image builds via github actions. 2023-07-19 21:48:59 So it's more convenient than just getting the lxc images. 2023-07-19 21:49:25 But I want it to be a pretty much complete system, so I'm adding in the missing bits. 2023-07-20 00:31:34 Hello. Anybody uses Alpine 3.18 with Xfce for desktop purpose? 2023-07-20 00:32:29 I'm sure there are... did you have an issue? :) 2023-07-20 00:32:39 I do not know why my computer always completely freezes. 2023-07-20 00:33:13 I am using a FirePro W4100. With no xf86 driver installed, it freezes twice a day (I mean per 8 working hours in my company). 2023-07-20 00:33:32 With 'xf86-video-amdgpu' installed, maybe once per 2 days. 2023-07-20 00:34:04 I tried to change the card to a Quadro P1000, install 'xf86-video-nouveau', and twice per day again. 2023-07-20 00:35:05 At first I thought it was hardware issue, but when I install Windows and let some game benchmark run for 2 days (weekend), everything is stable. 2023-07-20 00:40:06 ouch. 2023-07-20 00:40:12 Memory test, disk test... everything passed. Windows has a hell of hardware diagnostic tools for me to play around, and yet no problem was found. 2023-07-20 00:41:45 well while we're waiting for anyone else to come up with better ideas; did you try any of the magic sysrq keys after the freeze happened? 2023-07-20 00:42:50 No. I only tried Ctrl + Alt + F<1-6> key to get out of the tty7 (which Xfce is on), but it did not work. 2023-07-20 00:43:25 I could not get out of the frozen desktop, and had press the Reset button on the PC. 2023-07-20 00:46:47 sounds like a gpu driver bug 2023-07-20 08:58:16 Hello. Can anyone show me how to install and use ntfs3 (format drive, mount) in Alpine? 2023-07-20 08:59:06 I only see 'ntfs-3g' packages on the repository. I want to use ntfs3 from Paragon Software which was merged in Linux 5.15. 2023-07-20 08:59:28 ntfs3 would be included in the kernel already 2023-07-20 08:59:32 mount -t ntfs /dev/something path 2023-07-20 08:59:49 the kernel module is just ntfs3 2023-07-20 08:59:56 you probably don't have to manually load it :) 2023-07-20 09:00:07 unless busybox mount is that stupid 2023-07-20 09:00:12 Which tool do I need to format the drive as NTFS (the equivalent of 'ntfs-3g-utils')? 2023-07-20 09:00:25 ntfs-3g-progs 2023-07-20 09:00:55 mkfs.ntfs in it 2023-07-20 09:01:15 Um, not somethine like 'ntfs3-progs'? 2023-07-20 09:04:51 ntfs-3g-progs 2023-07-20 09:04:54 mkfs.ntfs in it 2023-07-20 09:05:10 mkfs.ntfs innit 2023-07-20 09:06:39 @psykose Actually BusyBox mount is stupid. It cannot detect the filesystem type if the device has not been mounted once before. 2023-07-20 09:07:00 i already said `mount -t ntfs` 2023-07-20 09:07:03 that tells it the type 2023-07-20 09:07:31 Yeah I know. This morning I try: setup-bootable -v "/media/cdrom" "/media/vda1" 2023-07-20 09:08:01 I mean '/dev/vda1', and the script failed because it cannot detect the type of the filesystem. 2023-07-20 09:08:15 I needed to manually mounted it once, then unmount (or use the mountpoint directly). 2023-07-20 09:21:54 /dev/vda1 on /media/vda1 type ntfs (rw,relatime,uid=0,gid=0,fmask=0177,dmask=077,nls=utf8,errors=continue,mft_zone_multiplier=1) 2023-07-20 09:21:57 drwx------ 1 root root 4.0K Jul 20 08:19 vda1 2023-07-20 09:22:16 I am logged in as root. Is there any reason why I cannot write to the mount point? 2023-07-20 09:22:24 what did you type 2023-07-20 09:22:27 what was the error 2023-07-20 09:23:00 touch /media/vda1/test.txt -> touch: /media/vda1/test.txt: Permission denied 2023-07-20 09:23:42 And before that: mkfs.ntfs --fast "/dev/vda1"; mount -t ntfs "/dev/vda1" "/media/vda1"; 2023-07-20 09:23:50 hmm 2023-07-20 09:24:07 that normally should work so i assume this is an ntfs3 bug, or that it's meant to require something more 2023-07-20 09:24:11 can you strace the touch 2023-07-20 09:25:01 and check dmesg 2023-07-20 09:25:55 it might be -t ntfs3 2023-07-20 09:26:27 and if neither work then try -t ntfs-3g with the `apk add ntfs-3g` fuse driver 2023-07-20 09:26:31 it's all a pile of cards :p 2023-07-20 09:27:32 Yeah I tried '-t ntfs3' and it worked! 2023-07-20 09:28:23 :) 2023-07-20 09:28:26 i wonder what -t ntfs does then.. 2023-07-20 09:28:29 hmm 2023-07-20 09:28:33 maybe it's just the old ntfs driver 2023-07-20 09:28:39 yeah 2023-07-20 09:29:46 if you check dmesg it might've printed why it didn't work but indeed it's the wrong driver 2023-07-20 09:29:52 annnoying 2023-07-20 09:32:26 Hi guys, I have reinstalled my system entirely (issue with luks2 grub fde) and now using luks1 before 2.12 rc1 grub would be available. Everything is working fine apart for my yubikey not available for my user, gpg --card-status is giving :... (full message at ) 2023-07-20 09:37:37 it's probably plugdev, but it could just be root:root only 2023-07-20 09:37:40 check the actual device 2023-07-20 09:38:53 did you also start pcscd 2023-07-20 09:39:05 well 2023-07-20 09:39:11 yes thanks it is started 2023-07-20 09:39:15 gpg uses scdaemon instead 2023-07-20 09:39:32 for which there's some interaction https://whynothugo.nl/journal/2023/03/13/using-a-yubikey-for-both-gpg-and-totp/ 2023-07-20 09:39:38 but it might be a separate issue 2023-07-20 09:40:30 i suppose its impossible to go from edge back to 3.18 without reinstalling everything hmmz? 2023-07-20 09:40:53 it's not impossible 2023-07-20 09:41:29 oh? 2023-07-20 10:17:32 psykose: thanks, I dont see scdaemon in init.d but pcscd (which is started), you mentioned plugdev, where you think I should have a look ? 2023-07-20 10:17:53 check the actual device in /dev for the yubikey 2023-07-20 10:18:05 but idk, i neither have a yubikey nor use gpg 2023-07-20 10:20:30 psykose: man 2023-07-20 10:20:42 s/man/many thanks ! your article made the trick!/ 2023-07-20 10:20:49 that's not my article :p 2023-07-20 10:20:51 I just added pcsc-shared to my scdaemon.conf 2023-07-20 10:21:00 hehe yes I know but the one you shared :) 2023-07-20 10:21:11 but yeah seems to be the only thing needed 2023-07-20 10:21:22 thanks WhyNotHugo for your top notch investigative analysis 2023-07-20 10:45:17 JohannesJ[m]: change to repositories back to 3.18, and then apk upgrade --update --available 2023-07-20 14:33:06 Anyone had luck adding flinch and nvidia drivers? Or best to go team red? 2023-07-20 14:33:20 **glibc and nvidia drivers 2023-07-20 14:34:52 that's not possible to do 2023-07-20 14:35:41 "team red" or even "team blue" is the better option 2023-07-20 14:36:20 I'm curious what those teams are 2023-07-20 14:36:27 AMD, Intel 2023-07-20 14:36:59 ah 2023-07-20 14:43:18 team beige ;-) 2023-07-20 15:03:42 elly: and nvidia is the green one 2023-07-20 18:51:49 psykose: I work on the Yocto Project and found https://git.alpinelinux.org/aports/commit/main/iproute2/include.patch?id=bd46efb8a8da54948639cebcfa5b37bd608f1069 . We are having a similar issue with iproute2 builds using musl. Have you submitted this to the upstream project, or plan to? 2023-07-20 18:53:27 i haven't submitted it, no 2023-07-20 18:53:32 my patience for email patches is almost zero 2023-07-20 18:53:34 feel free to steal 2023-07-20 18:54:13 psykose: alright, thank you 2023-07-20 20:28:19 does back/forth mouse button work for you in browsers? if so, what does xev call those buttons? 2023-07-20 20:31:35 frag: I don't recall it not working, but I cannot test atm 2023-07-20 20:32:17 no idea but i get BTN_SIDE/BTN_EXTRA via kernel event 2023-07-20 20:32:56 which is side/extra in whatever the xkeyboard name is or w/e 2023-07-20 20:33:05 also never not worked for me 2023-07-20 20:33:36 psykose: huh, weird eh? 2023-07-20 20:33:44 oh never not worked.. 2023-07-20 20:41:30 my mouse reports button 6 and 7 2023-07-20 20:41:54 for left and right respectively 2023-07-20 20:43:08 abby: and they work in the browser? 2023-07-20 20:46:40 they work for scrolling horizontally 2023-07-20 20:47:56 don't think i've ever scrolled horizontally 2023-07-20 20:48:09 is that like scrolling but canadian 2023-07-20 20:48:57 abby: aha! maybe thats the problem, cause i also get button 6/7 2023-07-20 20:49:30 they worked in this, for example https://codepen.io/jaemskyle/pen/XWJvRL 2023-07-20 20:50:18 they do indeed work for scrolling horizontally, i never do that either (almost) 2023-07-20 20:51:25 scroll to starboard 2023-07-20 21:11:02 horizontal scrolling is most useful if you live your life in spreadsheets 2023-07-20 21:11:18 there was a time where that was the big thing in mice; has mostly died out these days 2023-07-20 21:13:31 also useful when dealing with websites with bad viewport: settings 2023-07-20 21:45:16 blt[m]1, psykose I'm happy to hear that it helped :) 2023-07-20 21:45:37 s/hear/read/ 2023-07-20 21:46:20 WhyNotHugo: yes ! Many thanks again for your article 2023-07-20 22:01:46 Yall keep mentioning me in this chat 2023-07-20 22:02:51 bad luk 2023-07-20 22:04:25 what is it with people picking some common word username and then showing up later saying 'haha i got pinged' 2023-07-20 22:37:25 Ok, I'm hoping someone can explain what I'm seeing... 2023-07-20 22:37:47 several of my init scripts are being replaced with a script that uses /sbin/runscript as the interpreter 2023-07-20 22:38:31 but I don't have /sbin/runscript, and `apk search cmd:runscript` only returns minicom 2023-07-20 22:39:39 nothing does that that i've ever heard of 2023-07-20 22:39:47 what is 'several of my init scripts' 2023-07-20 22:40:34 fsck, hwclock, and a couple more. 2023-07-20 22:41:34 Oh, you know what, it's ksh that's doing it, so this is part of the zone brand. 2023-07-20 22:42:08 i don't think ksh rewrites random files on your disk at random 2023-07-20 22:42:25 unless you told it to 2023-07-20 22:42:29 could be wrong 2023-07-20 22:42:43 Well, it's a script written in ksh. And it's the provisioning system that's doing it. 2023-07-20 22:43:05 that still doesn't make sense 2023-07-20 22:43:28 Doesn't matter. 2023-07-20 22:43:42 what matters is that the problem is on my end, not yours. 2023-07-21 01:23:07 Hi guys. I created a bootable drive using 'setup-bootable', then booted from that drive. 2023-07-21 01:24:18 It is a FAT32 partition. I expected it to be writable. However when I booted into Alpine, I saw that it was automatically mounted as read-only: 2023-07-21 01:24:22 /dev/vda1 /media/vda1 vfat ro,relatime,fmask=0022,dmask=0022,codepage=437,iocharset=utf8,shortname=mixed,errors=remount-ro 0 0 2023-07-21 01:25:01 And I am unable to unmount it: umount /media/vda1 -> umount: can't unmount /media/vda1: Resource busy 2023-07-21 01:25:48 For what reason is it automatically mounted? And how can I save .apkovl files to it? 2023-07-21 01:48:51 livy: you should be able to make it rw using "mount -o remount,rw /dev/vda1" 2023-07-21 01:50:01 setup-bootable creates a "run-from-ram" installation, so when you boot Alpine is read from the drive and copied into RAM and run-from-ram, so the drive is mounted ro rather than rw 2023-07-21 01:51:51 however the "lbu" tool should remount the drive rw when it saves apkovl files AFAIK 2023-07-21 02:06:46 @minimal Thanks, it works. 2023-07-21 02:08:24 Do you know why I cannot unmount it? If all files are copied to RAM and run from RAM, there is no use for the boot media. 2023-07-21 02:08:34 Yet I received the error: umount: can't unmount /media/vda1: Resource busy 2023-07-21 02:17:31 livy: because as the error /media/vda1 is busy - something else is using it 2023-07-21 02:18:03 e.g. is /dev/vda1 also mounted at another mountpoint? 2023-07-21 02:18:31 "mount | grep vda1" 2023-07-21 02:20:12 I am testing the bootable media on a virtual machine. I am the sole user and just booted the vanilla OS. 2023-07-21 02:20:34 did you run that command? 2023-07-21 02:20:47 There are no other users and no other bind mounts. 2023-07-21 02:20:57 Yes, I ran it as root. 2023-07-21 02:21:13 well something is using /media/vda1 2023-07-21 02:21:41 are you inside the /media/vda1 directory? 2023-07-21 02:22:15 Of course no, I am inside '/root'. 2023-07-21 02:22:56 I rebooted the virtual machine and ended up with the same result -- unable to umount /dev/vda1 from /media/vda1. 2023-07-21 02:23:11 well then you need to find out what is using /media/vda1 2023-07-21 02:36:31 I tried to boot from the emulated CD-ROM drive again. This time it is strange: 2023-07-21 02:36:46 /dev/vda1 /media/vda1 vfat ro,relatime,fmask=0022,dmask=0022,codepage=437,iocharset=utf8,shortname=mixed,errors=remount-ro 0 0 2023-07-21 02:36:49 /dev/sr0 /media/cdrom iso9660 ro,relatime,nojoliet,check=s,map=n,blocksize=2048,iocharset=utf8 0 0 2023-07-21 02:37:25 I do not know why '/dev/vda1' is automatically mounted. I can unmount the '/media/cdrom', but cannot do the same with vda1. 2023-07-21 02:38:36 To get rid of that device being automatically mounted, I had to use: dd if="/dev/zero" of="/dev/vda" bs=4M count=5 2023-07-21 02:38:41 I think it is normal for the FAT partition to be mounted with Alpine run-from-ram. However I do not use run-from-ram so am not certain 2023-07-21 02:39:38 you used "dd" to wipe part of the device that you installed Alpine onto? so how do you expect to boot from that drive in future? 2023-07-21 02:40:56 I am booting from CD-ROM. It does not allow me to re-install Alpine on that device. 'setup-alpine' will fail because the device is being used. 2023-07-21 02:41:36 I think some Alpine daemon auto-mounts and use that device for... unknown purpose. 2023-07-21 02:42:26 it is used to store apk cache and apkovl files 2023-07-21 02:47:00 try stopping the modloop service and then unmounting 2023-07-21 02:50:47 Thanks a lot. I stopped the 'modloop' service and was able to unmount /dev/vda1. 2023-07-21 06:29:08 hey, how i can choose my own version of partition table? 2023-07-21 06:40:22 hey, how i can choose my own version of partition table? 2023-07-21 06:44:55 skip this bit? https://docs.alpinelinux.org/user-handbook/0.1a/Installing/manual.html#_partitioning_your_disk 2023-07-21 07:40:10 twitter no longer is supported by yt-dlp and similar? 2023-07-21 07:41:58 there was a bug until the most recent release i think 2023-07-21 07:42:37 i just tried it and it does work with yt-dlp 2023.7.6 from pypi 2023-07-21 07:42:56 abby: oh sweet 2023-07-21 09:58:24 hey 2023-07-21 09:58:36 can i get help here? 2023-07-21 09:59:10 sure. this a support channel. 2023-07-21 10:02:21 is alpine have installation guide, if u want leave a link pls 2023-07-21 10:06:00 is alpine have installation guide? 2023-07-21 10:07:14 cuc: you mean something like this? https://wiki.alpinelinux.org/wiki/Installation 2023-07-21 10:07:28 Or handbook here: https://docs.alpinelinux.org/user-handbook/0.1a/index.html 2023-07-21 10:09:47 "twitter no longer is supported..." <- Would most likely be the other way around 2023-07-21 10:18:24 terminaldweller alternateved, am asking because of this, which one should i use? 2023-07-21 10:24:21 both should be fine. have you found contradictory explanations/guides in the wiki and the handbook? 2023-07-21 10:25:58 idkrn[m]: seems it is supported in the latest version 2023-07-21 10:26:48 cuc: I would say one supplements the other so you should be fine using both. It all depends what you want to achieve 2023-07-21 10:27:59 Official install scripts and general overview how to work with Alpine seems to be in handbook. For anything more detailed or some special use cases, check wiki 2023-07-21 11:01:38 apternateved, I just read a bit of the installation page in the wiki, and didn't really understand how to configure my own partitions 2023-07-21 11:02:29 apternateved, cant see this option in setup-alpine 2023-07-21 11:03:52 i think i need read not wiki but docs 2023-07-21 11:07:21 cuc: if you don't want to use `setup-disk` utility (https://docs.alpinelinux.org/user-handbook/0.1a/Installing/manual.html#_setup_disk) then you would want to use `parted` most likely which is described here: https://docs.alpinelinux.org/user-handbook/0.1a/Installing/manual.html#_parted 2023-07-21 11:10:35 alternateved, i know i can use setup-disk to my partitions, but it strange for me 2023-07-21 11:10:50 i about how use it? 2023-07-21 11:11:24 i need run setup-alpine -q first for this, or how 2023-07-21 11:11:28 ? 2023-07-21 11:13:36 so setup-alpine -q will configure dhcp netowrk and a simple working config without asking any questions 2023-07-21 11:13:43 cuc: You could use `setup-alpine` and skip part related to disks and then you could run `setup-disk` 2023-07-21 11:14:08 you can run it interactively and then when asked about the disk say "none" 2023-07-21 11:14:16 ok 2023-07-21 11:14:31 Yeah, what ncopa said 2023-07-21 11:14:39 then you can manually partition the disk they way you want it and mount it somewhere, lets say on /mnt for example 2023-07-21 11:14:53 then you run: setup-disk /mnt 2023-07-21 11:15:10 alternative is use Semi-Automatic Installation guide? 2023-07-21 11:15:11 which will install the currently running system to a disk mouned at /mnt 2023-07-21 11:15:47 I have a note on how I install Alpine available here: https://paste.sr.ht/~alternateved/a79bc3d94664ee50f0f05f2f0de51361ae9924d5 2023-07-21 11:16:25 cuc: yes, semi-automatic guide 2023-07-21 11:17:02 i actually did a scripted automatic install this week, for qemu 2023-07-21 11:17:34 with alpine 3.18 there is support for tiny-cloud, which is similar to cloud-init 2023-07-21 11:18:23 alternateved, is fdisk available on standart iso? 2023-07-21 11:18:50 i think ye 2023-07-21 11:18:56 cuc: I don't know. Could be. 2023-07-21 11:19:35 cuc: Yeah, it's there. 2023-07-21 11:19:52 just why u use parted not fdisk if it already in 2023-07-21 11:20:05 I prefer it 2023-07-21 11:21:10 also if i need dual boot with win and have no uefi support i need mbr as i know 2023-07-21 11:21:30 and on mbr i dont really need separated /boot 2023-07-21 11:21:35 with efi32 2023-07-21 11:23:38 alternateved, am i right? 2023-07-21 11:28:20 cuc: Sorry, it's been a long time since I've used Windows or mbr so I won't be of help in that regard 2023-07-21 11:30:31 ok nm 2023-07-21 12:17:40 Hello, these days while building some docker containers, I discovered that the builtin alpine's busybox package has netcat with the unsecure `-e`flag enabled. Although busybox doesn't have a flag to disable it, wouldn't be better to disable the entire "netcat binary" in the busybox's build config and let whom who need it to manually install through the package manager? Alpine is likely the best 2023-07-21 12:17:46 security oriented distribution (for ones which can't simply go scratch in containers) and in my opinion keeping netcat (preinstalled), which couldn't be deleted or disabled as it's in busybox builtins it's dangerous as it creates a permanent way to attackers to leverage their access and gain a clean view of the container without additional complexity. 2023-07-21 12:26:44 furthermore, I'm not saying that having netcat or utilities with the same purpose is bad as in first an hypothetical attacker has to break one or more application and find a way to spawn it, but that's it 2023-07-21 12:37:55 I know there are a tons of different ways to achieve that but how you would simply force a dhcp renew when an interface is going up ? (my usecase is my server is in bridge mode with my 4G router, receiving the carrier IP address and if it lose power and came back, my alpine linux server kept the old IP address) would you do a post-up script on /etc/network/interface or something similar or is there an option on dhcp ? 2023-07-21 12:55:49 did your alpine server also lost power? because it should have got a new ip address after reboot 2023-07-21 13:07:28 https://wiki.alpinelinux.org/wiki/Udhcpc actually ifup seems the only way blt[m]1 2023-07-21 13:13:44 duckl1ng: nope it has a battery so I can have a situation where 4G router can lose power but my alpine server can be still on power 2023-07-21 13:18:16 f3derico: thanks, so you would use the /etc/network/if-up.d/ ? or inside the interface block on /etc/network/interfaces? 2023-07-21 13:21:49 if interface is aware (so it goes firstly down and then up) of this situation but dhcp client doesn't renew the lease then if-up.d is the most viable option 2023-07-21 13:27:36 something like: ping -c 4 1.1.1.1 || (ifdown eth0; ifup eth0) in cron every 5 minutes should do the job 2023-07-21 13:36:52 thanks guys ! I was thinking to try a post-up in /etc/network/interface 2023-07-21 13:39:12 maybe something like post-up kill -SIGUSR1 $(cat /run/udhcpc.eth3.pid) 2023-07-21 13:49:38 manually it is working but it doesnt seem it is through post-up, perhaps because it happens too early and the 4G modem needs 2 minutes to initialize after the interface cames physicaly up 2023-07-21 22:28:48 how does the APKBUILD suid option work? im working on a build of my own and having trouble with libfuse and util-linux wanting root 2023-07-21 22:30:28 https://wiki.alpinelinux.org/wiki/APKBUILD_Reference mentions it allowing setuid binaries but not really about implementation 2023-07-21 22:33:36 it doesn't do anything 2023-07-21 22:33:56 there's a check if something is suid and the build errors by default 2023-07-21 22:33:59 it makes it allow it instead 2023-07-21 22:34:04 it doesn't make anything suid 2023-07-21 22:35:58 which is what that entry says too i guess 2023-07-21 22:37:03 if you mean the literal check, https://gitlab.alpinelinux.org/alpine/abuild/-/blob/master/abuild.in#L871 2023-07-21 22:38:23 util-linux isn't a program so i'm not sure what you mean by it wanting root 2023-07-21 22:38:32 fuse mounts generally don't need root but you have to load the kernel module first 2023-07-21 23:21:18 when trying to build util-linux from src i get: chgrp: changing group of '//-util-linux.bin-2.39.1/bin/wall': Invalid argument 2023-07-21 23:22:41 sounds like your build environment is broken 2023-07-21 23:50:46 i run the build program inside `unshare` 2023-07-21 23:52:45 don't think chgrp works in unshare 2023-07-21 23:52:59 for util-linux with autotools you can try CHGRPPROG=true 2023-07-21 23:53:40 not sure why it runs it at all 2023-07-22 00:21:37 install-exec-hook-wall:: 2023-07-22 00:21:37 chgrp tty $(DESTDIR)$(usrbin_execdir)/wall 2023-07-22 00:39:11 makes sense 2023-07-22 00:41:41 generally installs are as root so it's not out of the question 2023-07-22 00:41:58 userns does not permit you to have more than one group in the namespace (what unshare is using iirc) 2023-07-22 00:42:31 the packaging workarounds for such things (sandboxed installs) are generally something like FAKEROOTDONTTRYCHOWN=1 fakeroot .. which changes user/group info but does not actually call the syscall 2023-07-22 00:42:44 then the packaging system can make a tar or whatever with the correct info anyway 2023-07-22 00:43:06 manually doing things though i don't think there's anything you can easily do 2023-07-22 00:43:16 you just can't install things that call chown family calls from unshare 2023-07-22 00:43:55 as long as it sticks to root-only then it works etc, just not this case 2023-07-22 00:44:42 the underlying limit is on the kernel not supporting it, i never explored if the uidmap/gidmap stuff lets someone make another way around it aside from fakeroot 2023-07-22 00:58:52 making my build env a chroot would definitely be more proper, using unshare for network sandboxing was just a simpler route up to this point 2023-07-22 00:59:17 but chroot also offers fs sandboxing so its been on my TODO list regardless 2023-07-22 01:02:33 well, unshare has many modes 2023-07-22 01:02:40 i think network-only doesn't imply the rest of the userns stuff.. 2023-07-22 01:02:43 maybe i'm forgetting 2023-07-22 01:03:13 i think it's only unshare -U that has the userns thing 2023-07-22 01:03:24 buut you can't call chgrp to random groups as not-root anyway 2023-07-22 01:03:37 hence the rest of the stuff for packaging systems and a `make install` 2023-07-22 01:10:49 yeah i was doing `unshare -c -n ...` which infers -U 2023-07-22 01:15:36 :) 2023-07-22 01:15:43 this whole area is always complicated as hell 2023-07-22 05:40:21 Could I install anaconda on alpine linux? 2023-07-22 05:56:40 don't think any of their binaries would run 2023-07-22 06:02:04 "but chroot also offers fs..." <- Chroot isn't really a sandbox 2023-07-22 06:03:43 An LSM would be better suited for the task 2023-07-22 06:49:12 lsm? 2023-07-22 07:03:34 psykose: is there any alternative for python 2023-07-22 07:03:42 just using python? 2023-07-22 07:04:30 qaqland[m]: For what purpose? 2023-07-22 07:04:59 switch different version 2023-07-22 07:05:27 qaqland[m]: ohh, you mean like Debian's update-alternatives? 2023-07-22 07:07:00 Isolate different Python versions 2023-07-22 07:08:11 Many tutorials start by recommending the use of this, but Idk what it's for.() 2023-07-22 10:04:06 @qaqland:mozilla.org did you mean python distro 2023-07-22 10:10:31 python using venv to Isolate 2023-07-22 12:07:14 Hi! root@localhost:/etc/apk$ cat /etc/os-release 2023-07-22 12:07:14 NAME="Alpine Linux" 2023-07-22 12:07:15 ID=alpine 2023-07-22 12:07:15 VERSION_ID=3.12.12 2023-07-22 12:07:15 PRETTY_NAME="Alpine Linux v3.12" 2023-07-22 12:07:16 BUG_REPORT_URL="https://bugs.alpinelinux.org/" 2023-07-22 12:07:16 HOME_URL="https://alpinelinux.org/" 2023-07-22 12:07:18 arch keys/ protected_paths.d/ repositories world 2023-07-22 12:07:18 root@localhost:/etc/apk$ cat /etc/apk/ 2023-07-22 12:07:20 #/media/dm-0/apks 2023-07-22 12:07:20 root@localhost:/etc/apk$ cat /etc/apk/repositories 2023-07-22 12:07:22 http://alpine.sakamoto.pl/alpine/v3.12/main 2023-07-22 12:07:22 http://alpine.sakamoto.pl/alpine/v3.12/community 2023-07-22 12:07:24 #http://alpine.sakamoto.pl/alpine/edge/community 2023-07-22 12:07:24 #http://alpine.sakamoto.pl/alpine/edge/main 2023-07-22 12:07:26 #http://alpine.sakamoto.pl/alpine/edge/testing 2023-07-22 12:07:28 ERROR: unable to select packages: 2023-07-22 12:07:28 root@localhost:/etc/apk$ apk add paxctl 2023-07-22 12:07:30 paxctl (no such package): 2023-07-22 12:07:30 required by: world[paxctl] 2023-07-22 12:08:04 wher is take this package? 2023-07-22 12:17:25 3.12 is end of support. paxctl is available since 3.13 2023-07-22 12:18:14 *is past end of support 2023-07-22 12:20:17 If you wish to upgrade, here's a handy guide: https://alpinelinux.org/releases/ 2023-07-22 12:20:23 ANALGIN__: ^ 2023-07-22 13:15:24 dhruvin: thnx! 2023-07-22 13:17:49 ANALGIN_: why are you using 3.12 rather than 3.18? 2023-07-22 13:31:16 hi there, i am having a problem with remapping scancodes to keycodes using hwdb config files 2023-07-22 13:32:37 so, like, i made the /etc/udev/hwdb.d/61-keyboard-local.hwdb file as the docs told me, i indicated in it the corresponding key codes 2023-07-22 13:32:49 so for example KEYBOARD_KEY_107=playpause 2023-07-22 13:33:19 they even show up when i open udevadm info /dev/input/event8 2023-07-22 13:33:34 but when i run evtest on that same device, they're instead shown as a scancode event with the correct code and key events with an "unknown" code :-( 2023-07-22 13:34:55 i have rebooted several times, i even rebuilt the initramfs to make sure, but nothing changes 2023-07-22 13:42:21 if you modify hwdb you have to run udevadm hwdb --update (and then either trigger udev or something else, i don't remember), rebooting/regenerating initfs don't do anything 2023-07-22 13:42:31 aside from that i'm not sure why hwdb would be the correct place 2023-07-22 13:42:46 i guess it does work 2023-07-22 13:43:37 there's setkeycodes too apparently 2023-07-22 13:44:55 oh wow, the matrix bridge is slow 2023-07-22 13:45:10 but yeah, i ran all of those commands 2023-07-22 14:00:25 tbh that *was* the original purpose of hwdb -- to collect various model-specific quirks and overrides 2023-07-22 14:01:05 (well, and pci.ids too, but keyboard.hwdb was there from almost day one) 2023-07-22 14:01:24 `udevadm trigger ` to re-run the udev rules (which import and apply info from hwdb) 2023-07-22 14:07:35 i have already done that 2023-07-22 14:08:01 when i run udevadm info /dev/input/event8, the KEYBOARD_KEY_107=playpause and all the other options i set are there 2023-07-22 14:08:34 the problem is that they aren't detected when i run evtest or wev or any other program 2023-07-22 14:08:53 evtest, for example, shows 2023-07-22 14:08:56 Event: time 1690034918.370344, type 4 (EV_MSC), code 4 (MSC_SCAN), value 107 2023-07-22 14:09:02 Event: time 1690034918.370344, type 1 (EV_KEY), code 240 (KEY_UNKNOWN), value 1 2023-07-22 14:09:07 Event: time 1690034918.370344, -------------- SYN_REPORT ------------ 2023-07-22 14:09:16 and then shortly the same, but for the key being released 2023-07-22 14:10:41 what does `udevadm info /devicepath` show for the KEYBOARD_KEY_107 you remapped 2023-07-22 14:10:58 E: KEYBOARD_KEY_107=playpause 2023-07-22 14:11:00 hmmm 2023-07-22 14:11:05 so it is remapped but then later is not 2023-07-22 14:11:16 that's weird 2023-07-22 14:14:10 the kicker is that the same keyboard also has other remaps in hwdb that are more unambiguous, like the rfkill key is defined in 60-keyboard.hwdb 2023-07-22 14:14:15 and that one works 2023-07-22 14:24:55 yeah i'm quite puzzled 2023-07-22 14:28:32 it needs to be KEYBOARD_KEY_, not decimal 2023-07-22 14:28:43 it is hex 2023-07-22 14:29:02 the other two keys are 10e and 10f 2023-07-22 14:29:19 as in "code 4 (MSC_SCAN), value 10e"? 2023-07-22 14:29:35 exactly 2023-07-22 14:30:33 lol it really does output hex for MSC_SCAN and decimal for everything else 2023-07-22 14:33:19 can you `strace -e ioctl udevadm test /dev/input/event8` to see if it's applying the mappings? 2023-07-22 14:33:30 let's see 2023-07-22 14:35:16 the only ioctl call is to set the terminal size 2023-07-22 14:35:50 then at the very end it sez "unable to open device '/sys/dev/input/event8'" 2023-07-22 14:37:14 hmm, I guess older versions only accepted /sys paths? 2023-07-22 14:37:23 give it /sys/class/input/event8 then 2023-07-22 14:43:12 okay, that works better, let's see 2023-07-22 14:44:12 it has three ioctl calls 2023-07-22 14:44:20 one is 2023-07-22 14:45:13 iioctl(5, EVIOCGABS(ABS_X), 0x7ffcf74ccc80) = -1 EINVAL (Invalid argument) 2023-07-22 14:45:26 the other is ioctl(5, TIOCGWINSZ, 0x7ffcf74d1648) = -1 ENOTTY (Not a tty) 2023-07-22 14:46:11 it also shows all the rules, both the already-existing ones and the ones i defined 2023-07-22 14:47:13 if y'all want, i'll dpaste the whole output somewhere 2023-07-22 15:16:54 here's the whole output, i think it does have some error messages in it? http://dpaste.com/AHEPZ9FRW.txt 2023-07-22 15:19:41 it's more useful without the -e ioctl, but i can't read it anyway 2023-07-22 15:48:59 i am starting to think that the udev event mapping doesn't matter at all 2023-07-22 15:49:21 there's another Fn key combination that in `evtest /dev/input/event8` returns a valid keycode 2023-07-22 15:49:49 if i press Fn+Insert, it returns a scancode 101 and keycode 364 (KEY_FAVORITES) 2023-07-22 15:49:59 nowhere is that defined in the hwdb 2023-07-22 16:11:31 https://github.com/torvalds/linux/blob/d192f5382581d972c4ae1b4d72e0b59b34cadeb9/drivers/platform/x86/ideapad-laptop.c#L1051C56-L1051C56 2023-07-22 16:11:45 i think this mapping is in the linux source code o_O 2023-07-22 16:34:06 hello! how is an Alpine installation done over UART these days? the wiki assumes but the default setup seems to use grub, looking at /boot 2023-07-22 16:35:51 those are distinct things unrelated from eachother? 2023-07-22 16:51:35 if i use setup-desktop, everything works out fine. but if I just install dwm following https://wiki.alpinelinux.org/wiki/Dwm when I run startx, X starts with dwm but mouse and keyboard don't seem to work 2023-07-22 16:51:52 did you install + enable udev? 2023-07-22 16:51:57 it appears as if frozen but hitting power button will close X and start shutdown 2023-07-22 16:52:05 nope, i figured it was something like that, will try thanks 2023-07-22 16:52:55 setup-xorg-base adds xf86-input-libinput and starts eudev 2023-07-22 16:53:11 hmm well i did setup-xorg-base for sure 2023-07-22 16:53:14 being in the input/video group is also relevant but neither of these do that 2023-07-22 16:53:32 i've tried this on two machine so i dont think its hardware specific 2023-07-22 16:53:36 setup-user does that now 2023-07-22 16:53:45 ah 2023-07-22 16:54:09 Any hints on watching Netflix on Alpine? yt-dlp doesn't do Netflix, and I can't find an equivalent tool for it. 2023-07-22 16:54:38 so udev should already be enabled by setup-xorg-base? 2023-07-22 16:55:02 WhyNotHugo: i don't think it's possible to bypass the drm thing 2023-07-22 16:55:11 in.. linux on general 2023-07-22 16:55:24 on<->in 2023-07-22 16:55:31 aside from the lowest level or w/e 2023-07-22 16:55:50 "lowest level or w/e"? What? 2023-07-22 16:56:26 widewine l1 works inside a flatpak browser but nothing past that afaik 2023-07-22 16:56:32 all that gets you is the 720p from what i remember 2023-07-22 16:56:58 it's been a long time since i checked :D 2023-07-22 16:57:07 the higher levels require much more drm integration with the OS 2023-07-22 16:57:36 some streamers do network probes too, not sure about netflix 2023-07-22 16:57:44 er 2023-07-22 16:57:48 i think the levels are backward 2023-07-22 16:57:59 L3 is lowest and L1 highest 2023-07-22 16:58:07 right 2023-07-22 16:59:18 so for non-hd video you can just use a flatpak browser that works with L3 and the drm module can't run on musl at all 2023-07-22 17:05:33 psykose: not sure if you replied to me, my IRC had an issue and missed my own message 2023-07-22 17:09:01 if i startx under root the mouse and keyboard work, i've noticed. but not under regular user 2023-07-22 17:10:04 are there any usage statistics for packages, e.g. download counts? 2023-07-22 17:10:07 ffog: that's due to what psykose mentioned 2023-07-22 17:10:17 you need to add your used to input and video groups 2023-07-22 17:10:22 crapStone[m]: no 2023-07-22 17:10:25 psykose: ok, got it working :) thanks regardless! 2023-07-22 17:10:28 crapStone[m]: we have no way of measuring that 2023-07-22 17:10:54 ok, thx 2023-07-22 17:15:46 ikke & psykose: that did it, thanks for the help 2023-07-22 19:20:26 how i can setup disk manualy, i dont want to use partitions that alpine-setup (i dont see the point in separared boot and swap, if this is really necessary, explain me why) 2023-07-22 19:21:04 SWAP_SIZE=0 will not make it at all 2023-07-22 19:22:16 @psykose, if i want to make partitions so as not to take up the entire disk? 2023-07-22 19:22:56 I believe you can instruct setup-alpine to do no disk setup 2023-07-22 19:23:02 do it manually and call setup-disk on the mounted partition then 2023-07-22 19:23:05 after it then finished, you can do any manual setup 2023-07-22 19:24:02 but when i do setup-alpine 2023-07-22 19:25:11 you just say no 2023-07-22 19:25:13 manually partition 2023-07-22 19:25:16 , ... 2023-07-22 19:26:16 you mean none? 2023-07-22 19:26:20 yes 2023-07-22 19:27:04 when disk step in setup-alpine? 2023-07-22 19:27:11 yes 2023-07-22 19:28:25 and i can mount only one partition to /mnt, then setup-disk /mnt 2023-07-22 19:29:52 ^and its mean no swap and no separated partition for boot? 2023-07-22 19:32:10 Yes: https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-disk.in#L1595-1604 2023-07-22 20:23:25 Is there a need to add a post-down (or post-up) line in /etc/network/interfaces under an interface block or it is enough to put the script in /etc/network/if-post-down.d (or up.d) to fire the script? In other words , are the scripts in the .d folder are automatically runs or there is a need to add lines as well in /etc/network/interfaces ? Does networking service is managing that ? I made some tests and it doesn't seems to run my script 2023-07-22 20:23:25 if I connect / disconnect an Ethernet cable 2023-07-22 20:54:54 How do I build my own Alpine image (e.g.: like the ones in alpinelinux.org/downloads) ? 2023-07-22 20:58:15 aports/scripts/mkimage.sh seems to be a starting point, but the output of my (likely erroneous) usage isn't clear 2023-07-22 20:58:26 mostly documented by https://wiki.alpinelinux.org/wiki/How_to_make_a_custom_ISO_image_with_mkimage i guess 2023-07-22 20:59:05 there's also https://github.com/alpinelinux/alpine-make-vm-image as an alternative (not the same thing) 2023-07-22 20:59:29 and https://github.com/dermotbradley/create-alpine-disk-image 2023-07-22 21:02:00 I still don't quite understand how to use these. 2023-07-22 21:02:06 ./scripts/mkimg.arm.sh does nothing, since that just defines functions. 2023-07-22 21:02:31 you run just mkimage.sh 2023-07-22 21:02:35 The output of mkimage.sh isn't too clear for me: https://paste.sr.ht/blob/8937d05bf535552ab775bdf033aaf9bcd8f621af 2023-07-22 21:02:35 the other stuff is the profile 2023-07-22 21:02:51 --profile arm would them use the .arm 2023-07-22 21:03:02 though that will then fail since making a cross image doesn't really work 2023-07-22 21:03:05 they're not very good 2023-07-22 21:03:58 Okay, so I can try and do the arm build in a VM. But the output from mkimage.sh is not clear to me. 2023-07-22 21:04:13 you need --repository and --profile 2023-07-22 21:04:44 try --profile standard --repository http://dl-cdn.alpinelinux.org/alpine/edge/main as an example 2023-07-22 21:06:42 why it just does nothing without those passed and doesn't say "please pass this" idk 2023-07-22 21:09:18 Yeah, I'll try and patch it to compain if --repository is not provided. 2023-07-22 21:09:24 It's because the argument can be specified more than once. 2023-07-22 21:09:30 cmd:mksquashfs is also required and not listed there. 2023-07-23 02:33:51 so when will linux become less commie and more normal? 2023-07-23 06:21:38 how make the `setup-alpine` didn't have partitions on the entire disk? also i want only one partition, is it possible? 2023-07-23 06:22:13 like my disk 200gb i want partition with alpine 32gb 2023-07-23 06:22:51 dont want separated partition for /boot 2023-07-23 06:23:06 also dont want partition for swap 2023-07-23 06:32:41 envnoob: I'm not sure anyone else is awake right now, but... setup-alpine mostly just calls other setup-scripts... you could run the other scripts one by one until you get to setup-disk 2023-07-23 06:32:58 then when you get to that setup you can either pass options to setup-disk or do what you want manually. 2023-07-23 06:33:17 There's probably an easier way (answerfile?) but damned if I can remember it. 2023-07-23 06:50:44 mb i need answer `none` when question about disk in `setup-alpine`, then mount needed partitions to /mnt and run `setup-disk -m sys /mnt ? 2023-07-23 07:23:56 help me also, is some preinstalled consolefonts in iso? 2023-07-23 07:28:51 how i can make font bigger in standart iso? 2023-07-23 07:31:27 help pls 2023-07-23 08:03:06 how i can install classical 2023-07-23 08:03:22 how i can install classic fdisk? 2023-07-23 08:03:31 envnoob, https://wiki.alpinelinux.org/wiki/Fonts 2023-07-23 08:03:41 At the end of the page, you have an example 2023-07-23 08:04:33 i already setuped font 2023-07-23 08:04:51 is there only one fdisk? 2023-07-23 08:04:53 fdisk is provided by util-linux-misc, you should have it by default afaik 2023-07-23 08:05:32 but it is different 2023-07-23 08:06:02 What is different from what? 2023-07-23 08:06:18 from fdisk in gentoo or arch 2023-07-23 08:06:34 there some other things 2023-07-23 08:07:14 like this fdisk not start first partition from 1M 2023-07-23 08:07:25 Why would it 2023-07-23 08:08:00 bucause its busybox fdisk 2023-07-23 08:08:27 i think i need not busybox fdisk 2023-07-23 08:08:48 Well, I already told you how to get it 2023-07-23 08:08:54 ok 2023-07-23 08:09:17 But it's true, I tend to forget that alpine uses bbox by default :/ 2023-07-23 08:09:46 bbox is nice 2023-07-23 08:11:36 after using fdisk i go and remove util-linux-misc 2023-07-23 08:16:56 quinq, how make the `setup-alpine` didn't have partitions on the entire disk? also i want only one partition, is it possible? 2023-07-23 08:17:25 like my disk 200gb i want partition with alpine 32gb 2023-07-23 08:17:55 dont want separated partition for /boot, also dont want swap partition 2023-07-23 08:17:57 I don't know how alpine setup scripts work, but yeah Linux works with a single partition 2023-07-23 08:18:24 You don't need a sparate /boot nor a swap (partition) 2023-07-23 08:18:55 i want separate /boot since i uefi, but i on bios 2023-07-23 08:19:05 no uefi support on my pc 2023-07-23 08:19:05 Now, if you use complex boot models like UEFI, they might impose you to provide a FAT partition or something, but I'm not sure, I don't use it 2023-07-23 08:19:45 Then as long as your boot loader can read your root file-system, no need for a separate one 2023-07-23 08:20:36 some people said me answering `none` to disk question in `setup-alpine` script, then run `setup-disk -m sys /mnt` 2023-07-23 08:21:08 i have to mount everything I want in /mnt 2023-07-23 08:21:59 I cannot anwser for other people 2023-07-23 08:23:01 i try it now 2023-07-23 08:23:36 they said is the easiest way to install alpine with manuap partitioning 2023-07-23 08:24:20 I would agree 2023-07-23 08:24:35 though that implies knowing how to do it 2023-07-23 08:35:23 why i cant mount /dev/sdb1 to /mnt 2023-07-23 08:44:40 Have you created a file-system there? 2023-07-23 09:00:24 quinq, some troubles with util-linux-misc fdisk 2023-07-23 09:00:34 but now mounted 2023-07-23 09:00:53 sometimes commands not work 2023-07-23 09:02:32 how i can download os-prober? 2023-07-23 09:04:23 apk add os-prober 2023-07-23 13:32:50 who know what difference between `setup-disk -m sys /mnt` and `setup-disk /mnt` ? 2023-07-23 13:34:30 i didnt see any difference, but mb but maybe there is one here. 2023-07-23 13:34:39 ? 2023-07-23 14:17:43 akashaa: maybe read setup-disk --help? 2023-07-23 14:30:04 https://wiki.alpinelinux.org/wiki/Alpine_setup_scripts 2023-07-23 14:56:40 is there some alpine linux discord server? 2023-07-23 14:57:30 nope 2023-07-23 15:05:47 is therw was ansver on my question, i just i was just going out for a while 2023-07-23 15:05:50 ? 2023-07-23 15:07:04 question was: what difference between `setup-disk -m sys /mnt` and `setup-disk /mnt`? 2023-07-23 15:09:31 somebody knows difference? 2023-07-23 15:10:43 have you looked at setup-disk --help 2023-07-23 15:22:41 I wish this page had a link to download the actual APK: https://pkgs.alpinelinux.org/package/edge/main/aarch64/u-boot-rockchip 2023-07-23 15:23:00 E.g.: a link to http://dl-cdn.alpinelinux.org/alpine/edge/main/aarch64/u-boot-rockchip-2023.07.02-r0.apk 2023-07-23 15:32:21 abby, yes but not see reason why `setup-disk /mnt` and `setup-disk -m sys /mnt` commands are equal 2023-07-23 15:33:08 akashaa: because a "sys" mode installation is the default type for setup-disk 2023-07-23 15:34:26 thanks 2023-07-23 15:34:48 also why this proccess when loading bar is so long? 2023-07-23 15:39:33 also have another question: why after partitioning the disk with fdisk from util-linux-misc command `mkfs.ext4 /dev/sdxX sometimes doesn't work? 2023-07-23 15:40:40 WhyNotHugo: you could add it 2023-07-23 15:40:50 not sure how confusing it would be to see a useless download link for most people though 2023-07-23 15:41:19 adding manual .apk's is not the same as adding from repo so nobody should really ever use it 2023-07-23 15:41:37 it would be a mistake for anyone to click it except this once :D 2023-07-23 15:42:49 Hello 2023-07-23 15:43:18 On a new install, I'm trying to update from 3.18 to edge, so I replaced the entries in /etc/apk/repositories from 3.18 to edge 2023-07-23 15:43:42 When I call apk update, I see for example: fetch http://dl-cdn.alpinelinux.org/alpine/edge/main/x86_64/APKINDEX.tar.gz 2023-07-23 15:43:54 But then v3.18.0-6822-g364bf747c46 [http://dl-cdn.alpinelinux.org/alpine/edge/main] 2023-07-23 15:44:17 Is that expected that edge gives 3.18? 2023-07-23 15:45:17 I'm asking this, because the actual problem I have is that when I call apk search, I get this: WARNING: opening from cache http://dl-cdn.alpinelinux.org/alpine/edge/main: UNTRUSTED signature 2023-07-23 15:45:34 I'm not sure what happened, but more importantly how to fix that 2023-07-23 15:46:45 I tried to re-install alpine-keys, but I can't find how to re-install anything 2023-07-23 15:48:57 quinq: what keys do you have in /etc/apk/keys? 2023-07-23 15:50:41 I have 5: http://ix.io/4Bjk 2023-07-23 15:53:32 the 3.18 is expected yes 2023-07-23 15:54:26 ok 2023-07-23 15:55:25 those keys look fine (comparing with an Edge system) 2023-07-23 15:56:25 Hummm 2023-07-23 15:56:38 Ohhh… 2023-07-23 15:56:58 I get that error as a non-priviledged user, but it works as root 2023-07-23 15:57:18 Indeed I can see now that the permissions are highly suspect 2023-07-23 15:58:30 still, why mkfs.ext4 not working sometimes to crated partition, have no idea really? 2023-07-23 16:00:15 Fixing the perms doesn't fix the issue 2023-07-23 16:01:51 akashaa: mkfs.ext4 does not create a partition, it formats a ext4 filesystem so it will 100% fail every time to create a partition :-) 2023-07-23 16:16:37 I must be tired, I had chmod a+x instead of a+r -_- 2023-07-23 16:16:38 fixed. 2023-07-23 16:24:15 minimal, im about i doing /dev/sdb1 partition with fdisk (util-linux-musc version) and then mkfs.ext4 /dev/sdb1 2023-07-23 16:24:37 and it notbworking sometimes for some reason 2023-07-23 16:24:47 *not working 2023-07-23 16:25:33 we can't see your screen or what it fails with 2023-07-23 16:27:03 i can send all commands i do, and output 2023-07-23 16:30:01 akashaa: you have not specified which command is not working. Is the fdisk to create the partition not working? or is the mkfs.ext4 command to format a partition not working? you need to show the problematic command and the output when it fails 2023-07-23 16:30:56 ok i did `apk add util-linux-misc e2fsprogs` 2023-07-23 16:31:20 and that command failed? 2023-07-23 16:31:31 then i creating one partition on /dev/sdb with fdisk 2023-07-23 16:31:55 if i do blkid /dev/sdb1 exists 2023-07-23 16:32:16 but when i do mkfs.ext4 /dev/sdb1 it not work 2023-07-23 16:33:27 it fixing by deleting and creating fdisk partition and try mkfs ext4 commmand until it work 2023-07-23 16:33:44 for some reason it work sometimes 2023-07-23 16:33:51 "it not work" - you have not indicated what precisely this means 2023-07-23 16:34:24 wait i want send you output of what i call not work 2023-07-23 16:34:28 wait a sec 2023-07-23 16:42:17 minimal, now it worked 2023-07-23 16:42:43 but sometimes not work 2023-07-23 16:43:07 mb i need wait some times after write with fdisk command 2023-07-23 16:43:10 you have not provided any evidence of it not working so no-one can help you until you do 2023-07-23 16:43:54 but i cant trigger error now 2023-07-23 16:44:45 i booted to flash usb and doin same as before and now cant triger its error, all work 2023-07-23 16:45:07 I didn't say "now", the whole time you have been talking about "it not work" but you have so far provided no details of any error etc 2023-07-23 16:45:30 i rebooted pc since error was 2023-07-23 16:46:32 mb i now try call it again, wait 2023-07-23 16:55:17 i think mb i missed some letter before or something 2023-07-23 16:55:26 now its isnt error 2023-07-23 16:55:52 but i have also strange output when other command 2023-07-23 16:56:34 now i created 32G partition and mounted it to /mnt 2023-07-23 16:56:49 invalid argument i have 2023-07-23 16:58:59 minimal, mb i need something like `mount -t ext4 /dev/sdb1 /mnt`? 2023-07-23 16:59:49 what about flashfs or whatever its called? 2023-07-23 17:01:19 yeah you have to put -t 2023-07-23 17:01:54 Apachez: im about `mount /dev/sdb1 /mnt` write to me `mount: mounting /dev/sdb1 on /mnt failed: Invalid argument` 2023-07-23 17:02:31 why there its working https://docs.alpinelinux.org/user-handbook/0.1a/Installing/manual.html#_finalizing 2023-07-23 17:02:48 and they not specify -t ext4 2023-07-23 17:03:06 it works if you already had ext4 loaded or w/e 2023-07-23 17:03:45 but i already format sdb1 with ext4 and still not work 2023-07-23 17:04:53 all done after mkfs.ext4 /dev/sdb1, but mount /dev/sdb1 /mnt returns invalid argument 2023-07-23 17:05:10 i just cant mount it 2023-07-23 17:09:40 @psykose, ok i think i really need do it with -t ext4 2023-07-23 17:16:13 @psykose, ok i mounted it with -t ext4 it work 2023-07-23 17:18:49 now i runned `BOOTLOADER=grub setup-disk /mnt` it seems to have been executed normally, but for some reason it was written after all: You might need fix tge MBR to be able to boot 2023-07-23 17:18:57 what its mean? 2023-07-23 17:21:16 @psykose should i ignore this^ and just reboot? 2023-07-23 17:30:32 nice, i managed to corrupt all my private repos by renaming the packager priv/pubkey 2023-07-23 17:33:15 corrupt in what sense 2023-07-23 17:34:06 clients will scream with untrusted signature 2023-07-23 17:38:46 i was expecting that all pubkeys are enumareted or located by id-s, and not by the filename x.x took me a bit long to realize 2023-07-23 17:39:05 :) 2023-07-23 18:12:47 which installation mode is recommended for a regular desktop PC? I'm new to alpine linux 2023-07-23 18:13:41 the wiki mentions diskless, data disk and system disk 2023-07-23 18:15:08 last one 2023-07-23 18:17:36 but wouldn't it also be possible to use diskless or data disk mode and then save everything with 'lbu commit' before every reboot? I'm just curious 2023-07-23 18:20:21 m_ben: it's your choice which to use 2023-07-23 18:20:25 sure 2023-07-23 18:20:37 prefered if runned from sdcards since they dont like many writes 2023-07-23 18:22:07 and you can mount a specific drive or partition for persistent stuff 2023-07-23 18:22:20 so the "lbu commit" would only be if you change stuff in your system itself 2023-07-23 18:28:31 I'm trying to understand what the advantages and disadvantages are of each mode. I'm considering switching distros after 10 years. it's for a normal workstation, so I'm not using an sdcard 2023-07-23 18:35:22 m_ben: "sys" works like a typical distro. With diskless and data disk modes changes have to be saved or else they will be lost 2023-07-23 19:10:31 is it possible for an ordinary user like me to use the edge branch? 2023-07-23 19:13:08 why wouldn't it be 2023-07-23 19:14:30 can i switch to edge if i installed alpine with 3.18 live iso? 2023-07-23 19:14:51 sure, it's just changing the repos and `apk upgrade -a` 2023-07-23 19:17:19 @psykose, i just need replace v3.18 in /etc/apk/repositories to edge? 2023-07-23 19:17:53 yes 2023-07-23 19:20:32 what command i need for upgrade packages in one command - apk -U upgrade? 2023-07-23 19:23:09 apk upgrade -a 2023-07-23 19:25:18 i can do this without apk update before? 2023-07-23 19:29:07 why don't you just try this basic stuff and find out? 2023-07-23 19:29:12 it's one command away 2023-07-23 19:29:15 you just.. type it 2023-07-23 19:30:34 upgrade runs an update if the repo indexes don't exist (like right after you changed repositories) or if they're like an hour old 2023-07-23 19:30:38 -U makes it a minute instead 2023-07-23 19:32:47 note the -a, which you need 2023-07-23 19:33:26 by the way, there isn't an apk upgrade option that makes it behave like update;cache sync -a;upgrade -a, is there? i find the on-the-fly unpacking apk does to be quite perilous on flaky connections 2023-07-23 19:33:40 apk upgrade --help 2023-07-23 19:33:40 is well written and useful 2023-07-23 19:33:51 ovf: nope, nothing does the cache commands 2023-07-23 19:36:39 bad network handling has been a pet peeve of mine :D 2023-07-23 19:36:40 is that by design or would a patch adding such a flag accepted? the normal upgrade does write to the cache anyway, if one is set up. 2023-07-23 19:37:00 ask in apk-tools 2023-07-23 19:37:05 with an issue 2023-07-23 19:37:45 thanks. might just ask with a patch then. 2023-07-23 19:38:01 https://gitlab.alpinelinux.org/alpine/apk-tools/-/issues/10891 2023-07-23 19:38:12 exact same request :D 2023-07-23 19:43:33 @psykose, if i get issues with edge i can change it back to lates stable? 2023-07-23 19:48:38 psykose: nice find. looks like apk cache download/sync -a is what you want for upgrade 2023-07-23 19:50:22 ascensionwired: it's not a "tested" thing but generally sure, some people have done it 2023-07-23 19:50:25 just always remember the -a 2023-07-23 19:54:10 i dont beed apk update if using -a? 2023-07-23 19:54:40 or i can run apk -U upgrade -a ? 2023-07-23 19:55:03 i already told you in an answer above 2023-07-23 19:55:37 you can always run apk upgrade -Ua if you want, no harm done 2023-07-23 19:56:59 and i can back from edge to latest-stable always? by replacing it back in /etc/apk/ropositories and then apk upgrade -a ? 2023-07-23 20:03:13 on edge version linux lts version? 2023-07-23 20:03:34 @psykose? 2023-07-23 21:31:43 bb 2023-07-23 21:33:31 Exporting LC_ALL or LC_TIME doesn't change how dates are formatted (including in thing like Python applications). 2023-07-23 21:33:46 I think that I need to keep something in mind due to musl? 2023-07-23 21:38:58 'date' (at least the coreutils version) is probably the only tool I've ever seen that uses the locale to determine the actual strftime string 2023-07-23 21:39:59 if a program does not vary its strftime format string depending on locale, then varying the locale is not going to produce different results 2023-07-23 21:40:55 I'm aware that programs that ignore locale won't vary their output based on locale. 2023-07-23 21:41:05 My specific issue is that no programs respect LC_TIME or LC_ALL on alpine. 2023-07-23 21:41:18 Even a simple Python test script. 2023-07-23 21:41:40 what is the python test script? 2023-07-23 21:43:05 Here's one: python -c 'import time; import locale; locale.setlocale(locale.LC_TIME, ""); print(time.strftime("%c"))' 2023-07-23 21:43:27 todoman is another one which isn't respecting locale 2023-07-23 21:43:42 But, TBH, I can't find any applicaiton where the locale works at all 2023-07-23 21:43:59 This also produces bogus results: https://github.com/delthas/go-localeinfo 2023-07-23 21:44:24 https://wiki.musl-libc.org/open-issues.html mentions various limitations 2023-07-23 21:46:22 It mentions that translation based on LC_TIME is unsupported, but nothing regarding date formats. I'll ask in #musl anyway. 2023-07-23 21:46:47 "The bulk of the data should be derived mechanically from the Unicode CLDR where possible, but the CLDR seems to lack certain time format variants corresponding to the ones C/POSIX needs for nl_langinfo/strftime." 2023-07-23 21:47:22 the python implementation of time.strftime is basically just a call to strftime, which would be implemented by musl 2023-07-23 21:49:01 yeah there's no LC_TIME implementations 2023-07-23 21:51:08 the more annoying part of it is the lack of _NL_TIME_WEEK_1STDAY or whatever 2023-07-23 21:51:38 ask in #musl anyway 2023-07-24 09:35:30 Flatpaks are all failing to start recently, with error "bwrap: Can't mkdir /usr/lib/x86_64-linux-gnu/GL: Read-only file system" 2023-07-24 09:35:34 https://paste.sr.ht/blob/0a36832f3d91c22b1641be905b6f1d7245f5b736 2023-07-24 09:36:07 Anyone else seeing this issue? Everything I find seems to be from around 2020 and totally unrelated 2023-07-24 09:38:34 No issues here. which alpine release are you using? 2023-07-24 09:39:56 Edge 2023-07-24 09:41:16 hm 2023-07-24 09:41:37 i don't have this issue 2023-07-24 09:41:41 wonder why.. 2023-07-24 09:48:54 WhyNotHugo, i good ol strace might help finding out why it's trying that 2023-07-24 09:49:06 that or launching flatpak with -v or -vv 2023-07-24 09:55:24 Or both! (in stacking order) 2023-07-24 10:08:07 hello, has anyone had a chance to look at my previous messages regarding the netcat thing? 2023-07-24 10:10:47 f3derico, nothing in the back-log, maybe ask again 2023-07-24 10:11:03 I also missed it. 2023-07-24 10:15:43 oh ok, so to summarise it, i've noticed that the alpine's busybox binary in the official container image has netcat preinstalled. that would be good in mostly common usages but busybox's netcat applet has the `-e` option (deprecated in the openbsd version and many others too) which could help attackers to raise their system overview. as alpine states that's a security distro (and that's certainly 2023-07-24 10:15:49 true) i don't understand why it builds netcat utility in busybox binary 2023-07-24 10:18:08 of course the netcat binary itself is harmless but it's matter of choices and as soon as i discovered it while debugging other things i came with the idea of asking there. thanks :) 2023-07-24 10:18:47 f3derico: I wanted to suggest you to create an issue for it, but you were already gone then 2023-07-24 10:19:12 The manpage for nc on openbsd-current doesn't say anything about that option being deprecated??? 2023-07-24 10:19:31 yes but the busybox one still has support for it 2023-07-24 10:19:38 nc -e doesn't let you do anything you can't already do though 2023-07-24 10:20:27 this doesn't sound like an improvement for anything 2023-07-24 10:20:49 you can replace nc -e with an equivalent random shell loop from the internet and end up in the same state, who cares 2023-07-24 10:21:05 it's not a privesc and it does nothing meaningful 2023-07-24 10:21:24 this sounds like some corporate manager checklist 2023-07-24 10:21:25 is that equivalent of something like nc -z && prog? 2023-07-24 10:21:46 -z closes 2023-07-24 10:22:01 it keeps it open and the remote controls the binary 2023-07-24 10:22:05 ah ok 2023-07-24 10:22:15 effectively it's nc --remote-shell haha 2023-07-24 10:22:37 you can take over a server and nc -e back to your own server, and from your server you type stuff into the shell backwards 2023-07-24 10:22:44 psykose: nah the thing is that neither debian or ubuntu's one as far as i know comes with netcat preinstalled 2023-07-24 10:22:46 ..which you can already do if you can just run nc 2023-07-24 10:22:53 but yes there are various way to replace it 2023-07-24 10:24:06 Again, as psykose points out, if some malicious actor already has access to your machine and they can run programs, nc with or without -e is kind of your least problem 2023-07-24 10:25:22 quinq: ok this explaination came in my mind when thinking about it, just wanted to see if it was the right one :) 2023-07-24 10:27:07 but yes that's just uncommon to have it preinstalled as many others do not 2023-07-24 10:27:34 making it not exist would require removing it entirely from busybox which isn't workable for a lot of reasons 2023-07-24 10:27:41 IT's good to discuss things anyway 2023-07-24 10:28:36 (a symlink to busybox being removed doesn't remove `busybox nc `) 2023-07-24 10:29:05 psykose: the first time f3derico asked this they mentioned that fact 2023-07-24 10:29:07 psykose: uhm, it's literally a build parameter, but yes if someone relies on it, he'll be forced to apk install openbsd version or ncat 2023-07-24 10:29:33 Yeah, it might be used in default dhcp scripts for example, or other things (though I have no idea ^^) 2023-07-24 10:29:42 i'm aware how busybox build configs work :) 2023-07-24 10:29:54 XD 2023-07-24 10:31:10 anyway i was also searching about programmatically disable busybox applets on the way (such as an option for root users only) but nothing :| 2023-07-24 10:31:50 lol 2023-07-24 10:32:07 that would be so great for customized environments to keep least applications as possible 2023-07-24 10:32:45 psykose: bruh actually the only way is recompiling it \_0_/ 2023-07-24 10:32:45 if it's a customised environment you can build your own busybox in like 30 minutes of automation with exactly what you want in it 2023-07-24 10:34:47 that's clear, but one thing is do per-app customizations and another is prepending to them containers one or worst doing custom images which not always are updated like upstreamed ones 2023-07-24 10:35:28 Use patches, not full confs 2023-07-24 10:35:44 This way you keep your upstream and you only modify what you know you want to 2023-07-24 10:36:50 (patch being either an out-of tree patch in your build system, or a git commit on top) 2023-07-24 10:38:21 Though depending on your distribution model, usually clients rely on what *you* give them, not what the upstream does, would be strange to say “sorry upstream broke it and we didn't see” 2023-07-24 10:38:54 shrug 2023-07-24 10:39:34 it sounds like you're shipping something somewhere for some reason, went 'this has nc i dont want it', and instead of fixing the issue in a million different ways you want to just remove nc entirely from the base of alpine 2023-07-24 10:39:41 the easy response is "no" and i'll stick with it :) 2023-07-24 10:40:23 not only because it doesn't make sense distro-wise, but also because there's like 100 other applets you don't want either 2023-07-24 10:40:28 wget is the second thing to go to 2023-07-24 10:40:45 yes! 2023-07-24 10:40:48 what you're thinking of is a "distroless container" or an app container or whatever, and in those you kinda just do your own thing 2023-07-24 10:41:10 something like wolfi 2023-07-24 10:42:19 even for alpine specifically, baselayout and musl doesn't even have a shell by default, and you could just pick any shell and symlink it to /bin/sh to start, then just add your own busybox build on top or w/e 2023-07-24 10:42:28 idk, there's really an infinite amount of ways to design this 2023-07-24 10:44:27 psykose: sure, I just wanted to discuss about it 2023-07-24 10:45:00 discussion is all good :) 2023-07-24 10:49:20 in more fun news i burnt my hand removing a cake from the oven 2023-07-24 10:49:21 gr 2023-07-24 11:06:31 arg :( 2023-07-24 11:06:41 But the question is what kind of cake? :D 2023-07-24 11:07:09 (eating it once it's cooled down will hopefully make the pain irrelevent) 2023-07-24 11:07:30 want cake :< 2023-07-24 13:22:51 why if i follow this step i have not swap on after rebooting? https://docs.alpinelinux.org/user-handbook/0.1a/Installing/manual.html#_finalizing 2023-07-24 13:24:19 because swap is not in your /etc/fstab or whatever 2023-07-24 13:28:13 firstly i runned setup-alpine and answered `none` when i was asked questions about the disks 2023-07-24 13:28:30 then i created swap and root partitions 2023-07-24 13:28:56 then i formatted root partition to ext4 and mkswap to swap partition 2023-07-24 13:29:16 then i mounted root partition to /mnt and did swapon 2023-07-24 13:29:40 then i runned BOOTLOADER=grub setup-disk /mnt 2023-07-24 13:32:04 https://docs.alpinelinux.org/user-handbook/0.1a/Installing/manual.html#_finalizing there they say do this 2023-07-24 13:32:19 but not swap for me no on 2023-07-24 13:32:44 did you check the contents of /mnt/etc/fstab AFTER you ran setup-disk? 2023-07-24 13:36:24 no but why in docs they do as they do 2023-07-24 13:36:45 they just swapon and after setup-disk 2023-07-24 13:48:41 setup-disk will create a fstab 2023-07-24 13:58:23 but why it not add swap on it 2023-07-24 13:58:41 https://docs.alpinelinux.org/user-handbook/0.1a/Installing/manual.html#_finalizing 2023-07-24 13:58:53 i followed this instructions 2023-07-24 14:00:43 i have already said above what steps i have taken, have no idea why setup-disc not detected swap 2023-07-24 14:01:06 *setup-disk 2023-07-24 14:01:27 did you actually chck the contents of /mnt/etc/fstab after you ran setup-disk? 2023-07-24 14:01:40 no 2023-07-24 14:02:02 well then do so, it should have an entry for the swap 2023-07-24 14:11:44 @minimal, why in docs they not specify what you said 2023-07-24 14:12:01 no specify what? 2023-07-24 14:12:20 like manually adding swap entry to fstab after setup-disk 2023-07-24 14:12:35 I never said to manually add a swap entry 2023-07-24 14:12:46 i but what you said 2023-07-24 14:12:48 ? 2023-07-24 14:12:54 when did I say that? 2023-07-24 14:13:23 sorry, no you said check fstab after setup-disk 2023-07-24 14:13:43 but still, why setup-disk not detect swap 2023-07-24 14:15:57 it seems you need to pass SWAP_DEVICES=/dev/sda2 to setup-disk 2023-07-24 14:19:20 oh I hope it works, I have no idea why they didn't mention it, as if setup-disk itself should detect swap which turn on 2023-07-24 14:22:00 the docs didn't mention it as whomever wrote the docs did not realise it was required. setup-disk does not appear to detect swap partitions when you manually partition/format disks 2023-07-24 14:22:43 it expects the SWAP_DEVICES variable to tell it the swap partitions and then it adds entries for fstab for those 2023-07-24 14:24:45 but why did the one who wrote say to on swap, just to swap work only on live iso? 2023-07-24 14:26:07 I already answered this: "the docs didn't mention it as whomever wrote the docs did not realise it was required" 2023-07-24 14:32:02 i also have another small question: are all the steps for these pages give in the end, after the same result? it just seems to me that setup-alpine actually does a little more, i saw exactly what they offer to create a user there, instead semi-automatic installation https://docs.alpinelinux.org/user-handbook/0.1a/Installing/setup_alpine.html 2023-07-24 14:32:07 https://docs.alpinelinux.org/user-handbook/0.1a/Installing/manual.html 2023-07-24 14:35:10 and accordingly, the recommendations after installation may differ slightly for manual and setup-alpine method 2023-07-24 14:36:08 is this page https://docs.alpinelinux.org/user-handbook/0.1a/Working/post-install.html right for me or should i read something else? 2023-07-24 14:51:37 @minimal, mb i misunderstood something, but i feel its post installation recomendations only for after menual installing bot after `setup-alpine` script 2023-07-24 15:14:50 hello 2023-07-24 15:15:03 Can You help me? 2023-07-24 15:15:36 with what? 2023-07-24 15:15:47 where I find kernel module acpi-cpufreq 2023-07-24 15:17:19 honza: https://pkgs.alpinelinux.org/contents?file=acpi-cpufreq*&path=&name=linux-lts&branch=edge 2023-07-24 15:17:44 It's in the -lts or the -edge kernel 2023-07-24 15:18:12 honza: if the module can not be found, you most likely need to reboot 2023-07-24 15:18:16 if I have virt and 4 ? 2023-07-24 15:18:41 virt doesn't have it. 2023-07-24 15:19:01 ok. Thank You 2023-07-24 15:24:38 in alpine only lts kernel available to me! 2023-07-24 15:25:22 ascensio1wired: linux-edge is only in community, make sure you have it enabled 2023-07-24 15:45:30 ikke, how i can download it, and switch to it? 2023-07-24 15:51:40 enable the community repo by editing /etc/apk/repositories 2023-07-24 15:53:41 then install linux-edge same as any other package (apk add linux-edge) 2023-07-24 16:07:40 this replace lts? 2023-07-24 16:08:33 it installs alongside of it. 2023-07-24 16:09:14 you should be able to select one or the other with whatever boot method you are using. 2023-07-24 16:09:44 I personally keep lts around on my box just in case if I screw something up on the edge initrd. hehe. 2023-07-24 16:13:06 zcrayfish, you mean grub? 2023-07-24 16:15:08 https://wiki.alpinelinux.org/wiki/Kernels, there instructions, and they say install kernel before setup-alpine 2023-07-24 16:15:15 grub, syslinux, efibootmg; whatever you are using. 2023-07-24 16:20:12 how i can load edge kernel before setup-alpine? 2023-07-24 16:23:16 you don't 2023-07-24 16:23:44 but there this way https://wiki.alpinelinux.org/wiki/Kernels, 2023-07-24 16:26:09 then if that article is correct then do what it says rather than asking how to do with it says? 2023-07-24 16:26:17 s/with/what/ 2023-07-24 16:29:54 you can do it...but it requires having an alpine system to build a custom alpine iso 2023-07-24 16:30:35 so you have to install using an existing image...then switch that install over to use -edge 2023-07-24 16:31:30 what it's talking about is if you want to run a specific flavor of the kernel rather than the normal one. -edge is not a flavor, it's a distribution 2023-07-24 16:32:32 so if I wanted to run linux-virt kernel rather than the standard linux kernel I'd have to apk add linux-virt before running setup...then setup would install -virt over default. 2023-07-24 16:33:00 for what you want...which is to run a bleeding edge distribution; you have to do that after you install 2023-07-24 16:34:16 dewdude, i just need apk add linux-edge in alpine installation not from iso 2023-07-24 16:34:20 ? 2023-07-24 16:34:23 no 2023-07-24 16:34:40 you have to install whatever version of alpine you have 2023-07-24 16:34:55 then you "upgrade" it to alpine-edge 2023-07-24 16:35:02 there is no "linux-edge" kernel 2023-07-24 16:35:14 you sure? 2023-07-24 16:35:17 you're confusing alpine-edge...which is a distribution version...with a flavor of linux kernel 2023-07-24 16:35:18 yes. 2023-07-24 16:35:28 but what is linux-edge package? 2023-07-24 16:35:58 https://pkgs.org/download/linux-edge 2023-07-24 16:36:19 the latest kernel. 2023-07-24 16:36:25 but it's also a part of the "edge" branch 2023-07-24 16:36:25 dewdude: there is a linux-edge package in 3.15+ and Edge 2023-07-24 16:36:33 yeah... 2023-07-24 16:36:39 dewdude: I use linux-edge 24/7..... 2023-07-24 16:37:15 https://pkgs.alpinelinux.org/package/edge/community/x86_64/linux-edge 2023-07-24 16:37:21 i see that. 2023-07-24 16:37:22 i mean what is this 2023-07-24 16:37:25 dewdude: so how is there no linux-edge kernel? 2023-07-24 16:37:28 is it kernel? 2023-07-24 16:37:38 it's just latest kernel. it's not a "flavor" 2023-07-24 16:37:52 flavor would be like -virt vs -rpi 2023-07-24 16:38:04 where one kernel is built for virtualized guest and the other is built for raspberrypi 2023-07-24 16:38:18 it is a flavour 2023-07-24 16:38:21 i've always gotten this by changing from 3.whatever to edge 2023-07-24 16:38:23 in apk 2023-07-24 16:38:29 https://pkgs.alpinelinux.org/package/edge/community/x86_64/linux-edge 2023-07-24 16:38:39 i think its kernel 2023-07-24 16:38:47 im talking about it 2023-07-24 16:39:21 it does not fit the classic definition of "flavor of linux kernel" 2023-07-24 16:39:33 as it's not anything except a standard kernel...just newer version than -lts 2023-07-24 16:40:20 but what way i can install it and use? 2023-07-24 16:40:20 I'm referring to "setup-disk -k " 2023-07-24 16:40:45 oh ok 2023-07-24 16:41:23 the reason I said "no" earlier to the question of running "apk add" before setup-disk is that AFAIK linux-edge package is not on the standard/extended/virt ISOs 2023-07-24 16:41:47 yeah, pretty sure that one is download only 2023-07-24 16:41:56 minimal, really inalready see 2023-07-24 16:42:23 i can skip disk setupping in setup-alpine 2023-07-24 16:42:46 then install it and run setup-disk -k linux-edge 2023-07-24 16:43:24 no, I just said why you *cannot* do that 2023-07-24 16:43:53 i can skip disk setupping in setup-alpine 2023-07-24 16:44:12 as i said message above 2023-07-24 16:44:15 please read and understand what people are saying rather than just picking a few words out of sentences 2023-07-24 16:45:07 what didn't i understand? 2023-07-24 16:45:36 when I said that linux-edge is NOT on the ISO and so you cannot add it during installation 2023-07-24 16:46:25 i said i can add any packages if i skip disk-setup in setup-alpine by answer none 2023-07-24 16:46:36 after i can install anything 2023-07-24 16:47:07 after adding what i need i can setup-disk 2023-07-24 16:47:25 or i wrong? 2023-07-24 16:49:47 minimal, am i wrong? 2023-07-24 17:07:52 dewdude: linux-edge definitely is a flavor, it has different settings as lts 2023-07-24 17:07:58 help 2023-07-24 17:08:00 dewdude: it's a different kernel, a different package 2023-07-24 17:08:40 dewdude: it's also maintained by different persons 2023-07-24 17:09:08 you need to enable the community repository before you can install linux-edge 2023-07-24 17:09:27 so no, linux-edge is _not_ linux-lts on the edge repo 2023-07-24 17:21:08 explain how i can install edge this way, or i cant https://wiki.alpinelinux.org/wiki/Kernels 2023-07-24 17:23:59 You might be able to if you add the community repo before you run that command 2023-07-24 17:26:21 but i cant apk add before setup-alpine 2023-07-24 17:26:35 mb before i need diskless install 2023-07-24 17:26:43 1) you don't need to apk add it necessarily. 2) you can manually setup the network 2023-07-24 17:27:01 and 2023-07-24 17:27:34 ? 2023-07-24 17:27:50 add community to /etc/apk/repositories, then run setup-alpine -k edge 2023-07-24 17:29:53 oh 2023-07-24 17:29:56 ok 2023-07-24 17:46:33 ikke i cant run setup-alpine with -k 2023-07-24 17:47:03 right, that's an argument to setup-disk 2023-07-24 17:47:13 FLAVOR=edge setup-alpine 2023-07-24 17:47:32 i can skip disk step by answer none on questions about disk in setup-alpine script 2023-07-24 17:48:06 then i can install with disk-setup -k edge 2023-07-24 17:48:57 yes 2023-07-24 17:49:43 i think its a best choice, because all i need before setup-disk is change already added repos in /etc/apk/repositories file 2023-07-24 17:55:38 where i can find all env vars that setup-* usin? 2023-07-24 18:03:30 point a text editor at /sbin/setup* 2023-07-24 18:03:44 Isn't it a shell script you can read 2023-07-24 18:03:47 zcrayfish: Yeah 2023-07-24 18:27:59 like from where i can get know about variables like FLAVOR or SWAP_DEVICES? 2023-07-24 18:29:06 ascensio1wired: They are not explicitly documented, just used in the script: https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-disk.in#L715 2023-07-24 18:29:12 and it's KERNEL_FLAVOR, not FLAVOR 2023-07-24 19:30:34 Hey everyone. I created an Alpine Docker container ($ docker run -it alpine /bin/sh), nothing special really. However, the lbu command does not seem to be installed: "/bin/sh: lbu: not found". I did not find any mentions of someone else having this problem on the internet. Surely it's just me in some way... but this seems really weird. Any ideas? 2023-07-24 19:31:18 danielg: why do you expect lbu to be present in a docker image? 2023-07-24 19:31:47 Why wouldn't it? 2023-07-24 19:31:49 You can get it by installing alpine-conf, but it does not really make sense to have it present by default 2023-07-24 19:32:03 Because it's useless in a container 2023-07-24 19:32:56 I see. Thanks. However wouldn't it be useful when using persistent Docker volumes? 2023-07-24 19:33:05 No 2023-07-24 19:34:26 I guess I'll have to read more on lbu when I have more time to understand why 2023-07-24 19:34:35 there would be nothing to restore the apkovl for starters 2023-07-24 19:34:56 but it's also really not how you typically use docker anyway 2023-07-24 19:35:47 Yeah I actually just wanted to test out Alpine commands and setup procedures since I'm migrating vagrant provisioners to Alpine, 2023-07-24 19:35:55 so no typical use case for Docker containers 2023-07-24 19:36:12 right, so you can install it if you want, it's just not there by default 2023-07-24 19:36:31 The docker image only contains the bare minimum 2023-07-24 19:37:27 Thx 2023-07-24 20:35:44 hey guys - trying to set the chickenbit to mitigate zenbleed since we don't have the latest ucode yet (i think?), however rdmsr is failing since /dev/cpu/0/msr doesn't exist. I have /dev/cpu/0 and I did modprobe cpuid -- there's no mention of msr in dmesg, although not sure if there should have been. Any idea why /dev/cpu/0/msr is not available? 2023-07-24 20:37:51 https://ipv6.social/@w8emv@hachyderm.io/110770834666992055 2023-07-24 20:38:26 modprobe msr 2023-07-24 20:39:03 thanks, worked! o/ 2023-07-24 20:39:35 ed helping ed :P 2023-07-24 20:40:01 eds all the way down heh 2023-07-24 21:15:32 hey. a friend and I are trying to figure out how to get a diskless/bootable install of alpine on a flash drive, but we can't seem to figure it out. docs are a little scattered surrounding it and there doesn't seem to be a clear section that provides steps to get it working 2023-07-24 21:16:19 we'd appreciate a quick overview and potentially any further support with the steps necessary. I have the alpine live cd running here and a device to write to, I'm just not sure exactly what I need to do 2023-07-24 21:17:23 Are you wanting the exact same functionality from your flash drive as you get from the live CD? 2023-07-24 21:17:43 pretty much 2023-07-24 21:17:47 quinq, zcrayfish: This is the output of strace and 'flatpak -vv', but I don't see any clues: https://paste.sr.ht/~whynothugo/74cd19691ed8ac482d3f45a0c3a20dafbc259c05 2023-07-24 21:17:52 Maybe I should report this upstream? 2023-07-24 21:17:53 sounds like you want to install the flash drive in the start 2023-07-24 21:18:06 If others can't repro, it's probably not an Alpine issue 2023-07-24 21:18:08 encrypted though 2023-07-24 21:19:03 something I noticed: using setup-alpine and selecting the flash drive and then crypt/data leads to there being a 4G swap(??) partition and the luks partition, but nothing to boot from. so it seems like that would need to come first (setup-bootable?), and then manually stepping through what setup-alpine would usually do on it I guess? 2023-07-24 21:22:29 you can avoid the swap partition by doing this first: export SWAP_SIZE=0 2023-07-24 21:22:32 not sure about the rest tho 2023-07-24 21:33:04 grumn: data mode uses the device for data storage only, it does not install Alpine onto the device - that's why it is called data mode 2023-07-24 21:35:14 here's my end goal: a bootable flash drive with encrypted alpine, which can use lbu to persist some packages, and then the flash drive can be unplugged and the system used from there 2023-07-24 21:35:34 then don't select data mode ;-) 2023-07-24 21:35:54 then which is appropriate? 2023-07-24 21:36:03 is diskless a valid mode that just isn't listed? 2023-07-24 21:36:25 what does "encrypted alpine" mean exactly? you mean a LUKS-encrypted rootfs? 2023-07-24 21:36:34 yes 2023-07-24 21:36:47 goal is to make sure it can't be tampered with 2023-07-24 21:37:08 then use one of the sys modes - which are *not* diskless 2023-07-24 21:37:33 diskless boots off media and creates a rootfs in *ram* 2023-07-24 21:37:58 sounds like what I want 2023-07-24 21:38:33 boot, unlock, load everything into ram, (potentially install packages and use lbu), then unplug and use alpine from ram 2023-07-24 21:39:00 diskless does *not* have LUKS encryption 2023-07-24 21:39:15 ah so that's just not possible then 2023-07-24 21:39:19 hmmmm 2023-07-24 21:40:49 WhyNotHugo: I am not quite sure, but I'll think about it... In the mean time, can you make sure that your user is able to write to /run/user/$(id -u)/ ? It's my best guess so far. 2023-07-24 21:41:38 is it feasible to do manually? I'd like to be able to do it in the future, effectively so the flash drive can't be tampered with by anything (malware). I can definitely use sys for now 2023-07-24 21:42:22 setup-bootable basically just copied files across from the install media into a partition on the boot device 2023-07-24 21:43:11 grumn: boot UEFI or boot BIOS? how do you prevent tampering? even with LUKS someone could alter the Grub stuff to log the unlock passphrase you type ;-) 2023-07-24 21:43:42 (or the initramfs contents if you're not using Grub's "FDE") 2023-07-24 21:43:52 sheesh I guess that's also something to be concerned about lmao 2023-07-24 21:43:56 more to explore 2023-07-24 21:44:09 you didn't answer whether UEFI or BIOS 2023-07-24 21:44:34 uefi. although I think even I'd be pretty capable of an attack via modifying some binary in /etc. less complex 2023-07-24 21:45:02 wel with UEFI you need to have a ESP FAT-formatted partition which cannot be encrypted as the UEFI does not support encryption 2023-07-24 21:45:37 so you'd want to use Secure UEFI to protect against changes to ESP boot contents 2023-07-24 21:48:53 how does that work with flash drives? does something get stored in UEFI on the machine that prevents modified ESP from booting? 2023-07-24 21:49:45 my questions may get way off base, I'm not used to dealing with much on the level of UEFI/BIOS 2023-07-24 21:54:50 on all UEFI machines there is a key/certificate store on the machine, typically containing a Microsoft cert. You can add additional ones. With secure UEFI then the bootloader and other components such as kernel and initramfs may ( be signed and 2023-07-24 21:55:05 oops, may (or may not) be signed 2023-07-24 21:56:20 Is there anything a user can install to get alpine 'debug' packages so valgrind and gdb resolve? 2023-07-24 21:57:04 so, set up the flash drive, get it signed in UEFI on whatever machines it might be used on (while it can still be verified to be safe), and good to go 2023-07-24 22:02:19 I guess with that it'd be safe enough to use the live cd so it's diskless, just make sure initramfs is signed I think. the only worry then is someone having access to both and signing their own malicious live cd! lmao 2023-07-24 22:32:51 what its mean? https://github.com/alpinelinux/alpine-conf/blob/master/setup-disk.in#L1600 2023-07-25 02:39:24 ed: should be patched back to 3.17 2023-07-25 03:45:05 hello everyone. 2023-07-25 03:46:55 greetings. 2023-07-25 03:53:27 i'm looking for wake on lan package. which apk is best for alpine linux on ish app on iphone? 2023-07-25 03:55:33 you're using the iPhone to wake the Alpine machine? 2023-07-25 03:57:33 yes. 2023-07-25 04:00:23 dang 2023-07-25 04:02:46 wol 2023-07-25 04:03:08 ethtool should be able to setup the card to receive too 2023-07-25 11:14:37 hey i downloaded irssi package on alpine why this prints to me `unknown command SCRIPT` when i try load scripts with `/SCRIPT LOAD <...>`? 2023-07-25 11:15:27 sounds like an irssi question 2023-07-25 11:18:39 works4me 2023-07-25 11:28:29 maybe try installing irssi-perl as well 2023-07-25 11:30:23 yeah, needs that 2023-07-25 11:38:22 yeah 2023-07-25 13:29:43 How to install mongodb in AL3.17 or 3.18? 2023-07-25 13:32:10 you don't 2023-07-25 13:35:53 But what if i build from binary? 2023-07-25 13:57:07 you don't 2023-07-25 13:59:01 I'm curious. I believe it's not in the repos due to its license. But about building it from source, is it due to something related to glibc? 2023-07-25 14:05:08 you can probably built it yourself 2023-07-25 14:05:59 indeed, the build instructions and deps don't look too bad 2023-07-25 14:06:04 but it'll probably take a fair while to build 2023-07-25 14:08:14 I would be kinda surprised if it works with musl though 2023-07-25 14:09:10 ah so it falls in "can, but probably shouldn't" category :) 2023-07-25 14:09:27 generally big software packages like that don't Just Work without someone first dealing with the various glibc assumptions they make 2023-07-25 14:09:51 and if it isn't packaged for alpine, the odds that someone has already done that are not favorable 2023-07-25 14:14:10 https://github.com/mongodb/mongo/pull/1061 I remember alpine used to have mongodb. A very old version. But I don't think it will be any easier to upgrade it to latest release. 2023-07-25 14:15:13 ya 2023-07-25 14:15:38 also it would be hard to have confidence in the result, since a lot of the musl-vs-glibc problems crop up at runtime 2023-07-25 14:16:40 ANALGIN_: you can try digging it up from aports repo and maybe build it from source. As elly said, YMMV. So that, or maybe a container! 2023-07-25 14:17:03 your mileage may vary explosively, which is not what I would want out of a database :P 2023-07-25 14:23:13 it was removed because it's nonfree 2023-07-25 14:23:20 not related to it building or not 2023-07-25 14:23:24 building it isn't that hard 2023-07-25 14:32:52 ...and it no longer supports x86 architecture. In the end you won ;( 2023-07-25 14:42:57 who is running mongo on 32-bit x86 2023-07-25 15:02:40 https://wiki.alpinelinux.org/wiki/Kernels , i still don't understand why this way is being considered here, but not the launch setup-alpine and choosing diskless, and then run disk-setup -k ? 2023-07-25 15:08:33 ascensionwired: because people writing those pages don't always know that it's possible in another way (or deem the way they suggest simpler / better) 2023-07-25 16:55:13 disconnect 2023-07-25 17:14:28 alpine isn't webscale? D: 2023-07-25 17:15:45 why not 2023-07-25 17:15:54 no mongodb 2023-07-25 17:30:43 abby: ;-) 2023-07-25 17:35:37 what is more minimal and lightweight: useradd or adduser? 2023-07-25 17:36:12 not having users 2023-07-25 17:36:30 i always forget which is which, wish there was only one 2023-07-25 17:37:13 useradd is from shadow or wherever, adduser is the perl script 2023-07-25 17:37:31 maybe it's from bbox on alpine 2023-07-25 17:44:55 abby, i think im asking not about bbox adduser 2023-07-25 17:45:16 are you thinking about thought? 2023-07-25 17:53:31 adduser comes from Busybox 2023-07-25 17:57:11 minimal, in ubuntu adduser from busybox too? 2023-07-25 17:58:44 ubuntu doesn't use busybox afaik 2023-07-25 18:01:23 it doesn't 2023-07-25 18:02:21 now i am thinking about thought - and i can't stop! 2023-07-25 18:02:42 ascensionwired: why does it matter, this is the Alpine channel, not the Ubuntu channel 2023-07-25 18:03:48 ok sorry minimal 2023-07-25 18:40:41 what difference between util-linux-misc qnd util-linux-dev? 2023-07-25 18:41:12 the latter is development files for the libraries (libmount, libuuid, ..) 2023-07-25 18:41:25 `apk info -P` is a great way to answer such questions btw 2023-07-25 18:41:27 the former is all of the tools from util-linux 2023-07-25 18:41:32 it will show you what commands/libraries/etc a package would provide 2023-07-25 18:43:00 what is cmd: and pc: ? 2023-07-25 18:43:43 installed runnable binaries in path and pkgconfig files 2023-07-25 18:44:26 in util-linux-dev ony few pc: 2023-07-25 18:44:33 and no cmd: 2023-07-25 18:45:13 in util-linux-misc many cmd: and no pc: 2023-07-25 18:45:39 yes 2023-07-25 18:46:38 what package i need if i want util-linux programs? 2023-07-25 18:46:48 util-linux-misc 2023-07-25 18:49:14 ok 2023-07-25 18:58:05 setup-disk -k rpi equal to FLAVOR=rpi setup-disk? 2023-07-25 18:59:30 you ask a lot of questions that you can answer for yourself by just using the computer instead of needing people to type into it for you 2023-07-25 19:07:36 @psykose, yaah, thank you, that was very sound advice, right now 2023-07-25 19:11:36 nice post on Alpine 2023-07-25 19:11:39 https://drewdevault.com/2023/07/25/Alpine-does-not-make-news.html 2023-07-25 19:15:16 is there zen kernel on alpine? 2023-07-25 19:16:22 no 2023-07-25 19:26:28 why `udhcpc: broadcasting discover` displays around 4-5 times before `udhcpc: broadcasting select ...` and it takea around 10-20 seconds? 2023-07-25 19:31:44 because the dhcp broadcast was performed 4-5 times and it took 10-20 seconds to get a dhcp lease 2023-07-25 19:32:34 why it takes 10-20 seconds to get a dhcp lease depends 2023-07-25 19:32:36 for me it takes like 4 2023-07-25 19:32:51 network interface, quality of connection, quality of dhcp server on the router or whatever, .... 2023-07-25 19:36:50 ok i rebooted again and yes its takes 5-6 seconds not 10 2023-07-25 19:37:20 can i install cowsay on alpine? 2023-07-25 19:40:21 only the shit one 2023-07-25 19:40:25 and it's in edge/testing 2023-07-25 19:44:30 hey all 2023-07-25 19:46:18 @psykose, am think, i can compile it, will i succeed with musl? (also i noticed neovim pulls for itself libgcc in dependencies, and neofetch pulls bash as one of dependancies for some reason) 2023-07-25 19:46:34 neofetch is a bash script 2023-07-25 19:46:50 ok im noob 2023-07-25 19:46:58 libgcc is the gcc runtime library which is a pretty normal dependency 2023-07-25 19:47:31 and what about compiling cowsay with musl, is it possible? 2023-07-25 19:48:31 cowsay is in aports/testing 2023-07-25 19:49:00 psykose: Use pfetch 2023-07-25 19:49:03 Lol 2023-07-25 19:49:23 i don't use anything 2023-07-25 19:49:49 I've been retooling my application "stack", as in converting my containers to Alpine. In doing so, I've updated a few of the alpine linux wiki articles (for instance nzbget, transmission, etc). 2023-07-25 19:50:27 and I'd like to keep at it. However, I'm running into an issue with an rc service not starting in its default config. 2023-07-25 19:51:23 "Exec format error" Any pointers to run this down? I might even submit a change to the init file if I can figure it out. 2023-07-25 19:51:49 what rc service 2023-07-25 19:51:53 how did you run it 2023-07-25 19:51:56 smokeping 2023-07-25 19:52:22 idkrn[m], i asked not psykos, am inadequate noob as you see 2023-07-25 19:52:22 its init file is invoked by rc service 2023-07-25 19:52:39 are you by chance running the wrong architecture 2023-07-25 19:53:23 @psykose, I believe you corrected my entry to the Unifi controller wiki entry lolol 2023-07-25 19:53:25 thank you 2023-07-25 19:54:40 was that the issue? (exec format error is usually just running a wrong-architecture binary) 2023-07-25 19:56:32 I started by examining the init file for wrong paths, shell, etc, but it seems to check out. I can certainly look at the lib dependencies of the smokeping package to see what gets installed, maybe there's a clue there? 2023-07-25 19:56:58 no, none of that is relevant 2023-07-25 19:57:02 what machine are you running it on 2023-07-25 19:57:21 can you also paste your whole console of you starting it and the error etc 2023-07-25 19:59:33 ok, will do. for reference, this is an LXC container running AL 3.18.2 2023-07-25 20:00:56 # rc-service smokeping start 2023-07-25 20:01:00 * Starting smokeping ... 2023-07-25 20:01:04 * start-stop-daemon: failed to exec `/usr/sbin/smokeping': Exec format error 2023-07-25 20:01:08 * start-stop-daemon: failed to start `/usr/sbin/smokeping' [ !! ] 2023-07-25 20:01:12 * ERROR: smokeping failed to start 2023-07-25 20:01:28 sounds like the binary is the wrong arch 2023-07-25 20:03:43 smokeping is perl isn't it 2023-07-25 20:03:48 yes 2023-07-25 20:04:50 perl binary runs ok? 2023-07-25 20:06:19 i thought it was, but now I see no output form perl -v... 2023-07-25 20:06:45 what is /etc/apk/arch 2023-07-25 20:06:48 and what is your computer 2023-07-25 20:06:54 what does `file /usr/sbin/smokeping` return? 2023-07-25 20:07:10 x86_64 2023-07-25 20:07:25 and does usr/sbin/smokeping start with the perl shebang instead of random corrupted data 2023-07-25 20:07:26 and can you clarify what is "my computer"? 2023-07-25 20:07:55 what are you running it oin 2023-07-25 20:07:57 on* 2023-07-25 20:08:24 alpine linux container with ash as shell 2023-07-25 20:08:39 no 2023-07-25 20:08:41 the computer 2023-07-25 20:08:41 cpu 2023-07-25 20:08:43 architecture 2023-07-25 20:08:44 hardware 2023-07-25 20:09:10 it's a dell r430 running KVM and LXC 2023-07-25 20:09:20 xeon 64 bit 2023-07-25 20:09:34 alright 2023-07-25 20:09:35 jdusablon: and did you check the smokeping file like psykose suggested? 2023-07-25 20:10:05 are you sure running `perl -v` just returns nothing? 2023-07-25 20:10:12 yep 2023-07-25 20:10:14 if that's the case then idk why it's broken 2023-07-25 20:11:30 (it should not be nothing) 2023-07-25 20:11:51 @ikke are you referring to the exec invocation in the init file? 2023-07-25 20:12:10 no, /usr/sbin/smokeping itself 2023-07-25 20:13:07 nothing in the init file matters 2023-07-25 20:13:28 if you just run `/usr/sbin/smokeping` yourself you should get the same error, technically 2023-07-25 20:13:33 well 2023-07-25 20:13:49 more or less 2023-07-25 20:13:58 perl should run, anyway 2023-07-25 20:14:08 # /usr/sbin/smokeping 2023-07-25 20:14:31 "/usr/sbin/smokeping: line 4: use: not found" 2023-07-25 20:14:44 "/usr/sbin/smokeping: line 7: syntax error: unexpected "(" " 2023-07-25 20:14:51 so there is no shebang line? 2023-07-25 20:14:52 huh 2023-07-25 20:14:56 please show us: head -1 /usr/sbin/smokeping 2023-07-25 20:14:58 and it's interpreted as a shell script 2023-07-25 20:15:28 #!/usr/bin/perl 2023-07-25 20:15:47 ls -l /usr/bin/perl 2023-07-25 20:15:51 file /usr/bin/perl 2023-07-25 20:16:16 -rwxr-xr-x 2 root root 0 Jul 25 18:33 /usr/bin/perl 2023-07-25 20:16:21 welp 2023-07-25 20:16:24 it's there... 2023-07-25 20:16:34 size 0? 2023-07-25 20:16:37 0 byte file, your perl is broke 2023-07-25 20:17:08 apk fix perl 2023-07-25 20:17:17 lol 2023-07-25 20:17:41 oh what 2023-07-25 20:18:01 that had an effect, now perl shows version 2023-07-25 20:18:17 yes, it helps if it's the actual perl binary 2023-07-25 20:18:22 instead of a 0-byte file 2023-07-25 20:18:33 and now the service starts. 2023-07-25 20:18:49 ¯\_(ツ)_/¯ 2023-07-25 20:19:32 maybe go through your history and see how that happened ;) 2023-07-25 20:20:11 it broke today as you can see from the file timestamp 2023-07-25 20:21:07 yeah, I've been wrestling with the install this morning 2023-07-25 20:21:44 thank you all for the help. I now have to troubleshoot http paths, but at least it's working at the os level 2023-07-25 20:32:11 is the mouse pointer position recorded somewhere in /proc or something? 2023-07-25 20:33:48 The mouse is implemented by X or wayland, don't think that it's reflected in /proc 2023-07-25 20:42:39 it is in the tty layer as well, by whatever is in termcap 2023-07-25 20:44:25 the terminal cursor is not the mouse though 2023-07-25 20:44:42 It's just events, right, like a button press 2023-07-25 20:45:30 well yeah, like how tmux and screen handle mouse clicks and etc 2023-07-25 20:47:58 maybe i need to use xdotool 2023-07-25 21:16:30 pretty strong warnings on the edge wiki page ... many use it as their main system right? 2023-07-25 21:17:30 ye 2023-07-25 21:18:55 it's probably not significantly different philosophically from running debian sid 2023-07-25 21:21:25 mhmhm 2023-07-25 21:21:46 cool people use edge 2023-07-25 21:35:28 now im cool too! 2023-07-25 21:37:08 psykose: smh microsoft shill 2023-07-25 21:37:31 @psykose, you mean kernel or repo version? 2023-07-25 21:39:14 vyivel: O.o 2023-07-25 22:21:10 Alpine Edge, 60% of the time it works all the time (Anchorman misquote) ;-) 2023-07-25 22:22:10 ascensionwired: cool people use both at the same time. 2023-07-25 22:22:11 ;) 2023-07-25 22:26:20 what mouse does cool ppl use? 2023-07-25 22:26:31 i have some random zowie mouse 2023-07-25 22:26:34 and i'm not cool 2023-07-25 22:26:35 so not that one 2023-07-25 22:30:06 razer? 2023-07-25 22:30:18 i see we have razercfg in repo 2023-07-25 22:30:23 ACTION stares at his M590 in shame 2023-07-25 22:31:48 the best mouse is the one you're happy with 2023-07-25 22:32:06 the mouse i'd like the most doesn't even exist 2023-07-25 22:32:08 grumble 2023-07-25 22:32:18 i have logitech rx250, pretty good tbh, but lacks back/forth, i guess i should to without tho ... 2023-07-25 22:32:35 I wish I could waterproof mine... I have kids. :| 2023-07-25 22:32:55 the rx250 from 15 years ago? 2023-07-25 22:33:32 it looks so cute 2023-07-25 22:34:25 ehh, its been 15y already? i found it in a drawer 2023-07-25 22:43:57 i don't actually know 2023-07-25 22:44:01 it just looks like that kind of design 2023-07-25 22:44:03 xp era 2023-07-25 22:44:18 its prob right :/ 2023-07-25 22:44:30 it's cute 2023-07-25 22:47:23 cute 15yo mouse jk.. 2023-07-25 22:51:49 :3 2023-07-25 22:52:01 zcrayfish: i feel like half the people in this channel have kids lol 2023-07-25 22:53:08 I hope the rest of them have not experienced the joy of having their wireless mouse disappear into the toilet or coffee cup. xD 2023-07-25 22:56:56 :D 2023-07-25 23:02:56 maybe attach some kind of wire to it :> 2023-07-25 23:05:35 I'm actually thinking about taking my current one apart and spraying most of the PCB with waterproofing (cosmoline?)... hehe 2023-07-25 23:13:12 "idkrn, i asked not psykos, am..." <- I was just pointing out an alternative :) 2023-07-26 01:30:33 ACTION initiates boredom. 2023-07-26 01:30:56 Anyone on 3.18.2 or edge brave enough to see if doublescan modes in Xorg actually work? =) 2023-07-26 01:31:36 heh, I meant to send that to -offtopic. oh well 2023-07-26 01:31:59 :) 2023-07-26 01:33:04 xorg in 2023 2023-07-26 01:33:26 lxqt isn't quite there yet on wayland. 2023-07-26 01:33:38 getting close though 2023-07-26 01:33:53 TIL people use lxqt on alpine 2023-07-26 01:34:12 hehe 2023-07-26 01:37:53 It works very well on X11. I mean assuming your not trying to toy with something stupid like I mentioned above. 2023-07-26 01:38:05 lxqt-panel is a real pain on wayland though 2023-07-26 06:02:32 how would a mouse disappear into the toilet ? 2023-07-26 06:02:54 i mean, it's a weird place to bring a mouse 2023-07-26 06:20:29 i used both from start zcrayfish 2023-07-26 06:20:52 excellent. 2023-07-26 09:38:00 alpine-uboot-3.18.2-aarch64.tar.gz is missing u-boot/pinebook-pro (because support for this board was merged AFTER the release). Is there an equivalent tarball for edge? 2023-07-26 09:40:20 Oh, I guess I can just download the apk with that file from edge 2023-07-26 10:43:48 Trying to boot Alpint on a pinebook-pro drops me into GRUB. 2023-07-26 10:43:52 Boot log is here: https://paste.sr.ht/~whynothugo/9c37cb196e7d7c596797b4202495f7025c71ce89 2023-07-26 10:44:14 I looks like u-boot manages to mount the partition (since it can launch GRUB). 2023-07-26 10:44:39 But GRUB itself doesn't boot successfully. And I've no idea how to figure out what's wrong with GRUB. 2023-07-26 10:49:12 i don't recall the commands, but i believe you can ask grub, on that command line, what devices it sees, what files are on there, etc. 2023-07-26 10:55:05 grub has autocomplete with tab 2023-07-26 10:55:18 ls ( 2023-07-26 11:08:04 'devicetree' just prints empty 2023-07-26 11:10:10 The dtb is in sdcard:/boot/dtbs-lts/rockchip/rk3399-pinebook-pro.dtb 2023-07-26 11:10:24 And I guess u-boot picks it up or it wouldn't be able to find grub 2023-07-26 11:13:39 Why do ARM images even have GRUB? u-boot can boot directly into Linux. 2023-07-26 11:14:43 does u-boot have a minecraft theme? https://github.com/Lxtharia/minegrub-theme 2023-07-26 11:15:07 lol 2023-07-26 11:15:14 No, but it does detect my hardware properly 2023-07-26 11:15:28 that's a hard tradeoff. 2023-07-26 11:16:06 I find that my Arch failing to detect my GPU without me installing packages improves my battery life significantly. :-P 2023-07-26 11:18:09 Thanks, I needed a good laugh 2023-07-26 11:27:47 Reformatted the sdcard as fat32 and that worked -- the wiki says that grub works on ext4, but that doesn't seem to be correct. 2023-07-26 11:50:13 I wouldn't have bet on that. 2023-07-26 11:50:29 ACTION has run ext2 for /boot since forever. :P 2023-07-26 11:51:02 no, my current /boot is actually vfat, too. 2023-07-26 11:51:09 /boot != /boot/efi 2023-07-26 11:51:32 abby, I think the reason I chose vfat this time was so that I could have my grub boot and my efi boot on the same partition. 2023-07-26 11:57:59 red color of package version mean out of date? https://pkgs.alpinelinux.org/packages 2023-07-26 12:00:47 Reportedly out of date 2023-07-26 12:02:24 ikke, is it for the maintainer to update it? or does it not carry significant information? 2023-07-26 12:03:27 The maintainers would have received an email as well 2023-07-26 12:04:06 But there is also a lot of spam/non-sense flags 2023-07-26 12:04:26 i see 2023-07-26 12:04:39 its reason why am asking 2023-07-26 12:10:10 Would be nice if titles in the wiki were numbered (e.g.: 1.1, 1.2, 1.3, 2, etc). Makes nested titles easier to read. 2023-07-26 12:22:32 I've dumped all my findings in https://wiki.alpinelinux.org/wiki/Alpine_on_ARM, and have gotten as far as the kernel starting -- and the freezing forever. 2023-07-26 12:24:07 need wcal in repo 2023-07-26 12:33:13 https://github.com/leahneukirchen/wcal 2023-07-26 13:48:22 just for the weekly view? 2023-07-26 13:50:03 i use taskwarrior's calendar mostly, which is nice. but it doesn't have weekly views except for burndown, i think. but otoh it wouldn't be all that useful to me 2023-07-26 13:57:26 What is the status of snapper in repos? I tried to build and compile the one from testing, but it fails. 2023-07-26 13:58:06 replacement for 'cal' invoked, monthly/3 months/year 2023-07-26 13:58:33 busybox cal doesnt show which day it is 2023-07-26 14:08:02 task is probably way more than you need, but maybe check it out. one thing i can't live with now (that task calendar provides, and util-linux cal does also) is week numbers 2023-07-26 14:09:09 it took me a while to build habits around taskwarrior but i quite like it. 2023-07-26 14:10:21 is on alpine not busybox dd? 2023-07-26 14:16:33 ascensionwired: did you try it? 2023-07-26 14:18:14 invoked: almost 2 MB :o looks ok tho :] 2023-07-26 15:09:53 ikke, i think its in coreutils 2023-07-26 15:10:05 *(package) 2023-07-26 15:12:11 https://tpaste.us/1Xj4 2023-07-26 15:15:47 i mean by default alpine have busybox dd, but coreutils dd mb better 2023-07-26 15:16:48 depends on what you need 2023-07-26 15:18:16 bbox is good 2023-07-26 15:18:32 i just asked nothing more 2023-07-26 15:19:09 You made a statement / question whether busybox dd was available or not on alpine 2023-07-26 15:29:52 no, i asked for other dd, and apready finded it 2023-07-26 15:30:11 Well, that was not clear from what you said :) 2023-07-26 15:36:42 ive already found where it is, and i don't need it anymore, in most cases i use bbox things 2023-07-26 15:42:38 what was the procedure if I have 4 disks and want them to be used on / + swap, /var, /home, /usr ? 2023-07-26 15:43:34 artok, 4 partitions? 2023-07-26 15:44:20 *5 2023-07-26 15:45:30 yeah 2023-07-26 15:46:38 partitioning them first and then mounting root one to /mnt , var to /mnt/var ... and swap? 2023-07-26 15:47:30 And setup fstab 2023-07-26 15:47:40 i think you need setup-alpine and answer `none` on questions about disks, storing configs and apk cache dir 2023-07-26 15:48:13 artok, then you need partitioning and formatting if you need 2023-07-26 15:48:49 (mkswap ofc) 2023-07-26 15:50:31 then you need mount root partition to /mnt var partition to /mnt/var, home partition to /mnt/home, usr partition to /mnt/usr 2023-07-26 15:51:38 artok, i don't really know if you need some special mount flags for this 2023-07-26 15:54:33 after this i think you need run `SWAP_DEVICES= setup-disk /mnt` 2023-07-26 15:55:46 this should run sys mode installation to you 2023-07-26 15:57:21 idk if any flags are needed for mounting to make everything work for artok? 2023-07-26 16:12:31 yah thanks 2023-07-26 19:39:46 i need `-a` flag always if `apk upgrade`? 2023-07-26 19:40:06 ascensionwired: no, but you want to when you upgrade to a different alpine version 2023-07-26 19:40:39 ok 2023-07-26 19:40:45 thanks 2023-07-26 19:41:14 ascensionwired: providing -a (--available) would also discard any version pinning you have done 2023-07-26 19:44:09 discard? 2023-07-26 19:45:22 it will upgrade (or downgrade) to the version available in the repositories 2023-07-26 19:46:07 ok. also pinned == tagged? 2023-07-26 19:48:51 afaik, no 2023-07-26 22:03:29 is setup-interdaces uses wpa_supplicant, no iwd? 2023-07-26 22:08:02 https://github.com/alpinelinux/alpine-conf/blob/master/setup-interfaces.in see for yourself :) 2023-07-26 22:09:00 see only wpa_supplicant 2023-07-26 22:09:10 but why not iwd? 2023-07-26 22:09:22 because nobody's contributed a patch to use iwd? 2023-07-26 22:09:26 patches welcome 2023-07-26 22:10:01 I believe we discussed iwd support a long time ago and concluded that it wasn't better enough to be worth including in the base installer 2023-07-26 22:10:13 given that we'd have to keep shipping wpa_supplicant anyway, to support older cards 2023-07-26 22:10:41 at this point I think every driver in Linux is either cfg80211 or dead 2023-07-26 22:10:51 oh, now i understand 2023-07-27 03:03:54 no, that's not what happened 2023-07-27 03:04:09 some people decided to still keep shipping wpa_supplicant because of the iwd dbus boogeyman 2023-07-27 03:04:15 it had nothing to do with old cards :p 2023-07-27 03:04:42 there's an iwd dbus bogeyman? 2023-07-27 03:04:58 iwd/iwctl requires running dbus to use 2023-07-27 03:05:59 eiwd just removed iwctl entirely to drop that dependency, but that's like wpa_supplicant with no wpa_cli 2023-07-27 03:06:11 theoretically usable, realistically who in their right mind would want to ship that as the default 2023-07-27 03:15:48 ahh 2023-07-27 03:58:24 why on earth does busybox nc print "punt!" to stderr when you ctrl-c it 2023-07-27 04:03:38 elly: because the message exists in the original netcat-traditional 2023-07-27 04:05:14 https://salsa.debian.org/debian/netcat/-/blob/master/netcat.c#L225 2023-07-27 04:05:56 indeed, but why did that remain in busybox over all these years 2023-07-27 04:05:57 most distros patch out the message from netcat-traditional; eg: https://salsa.debian.org/debian/netcat/-/blob/master/debian/patches/no-sleep-punt.patch 2023-07-27 04:06:15 why doesn't every nc print punt 2023-07-27 04:06:18 probably because nobody cared? 2023-07-27 04:06:39 people should put more dumb messages in things 2023-07-27 04:07:07 psykose: this is only the full nc 1.10 compat version, which I suspect many busybox users don't enable, because they want a very minimal busybox 2023-07-27 04:07:34 and these days a lot of distros give preference to the openbsd netcat, which probably also doesn't have this message 2023-07-27 04:07:37 - fprintf(stderr, "punt!\n"); 2023-07-27 04:07:38 + fprintf(stderr, "psykose said this was a good idea\n"); 2023-07-27 04:07:40 patch on the way 2023-07-27 04:07:45 let's go 2023-07-27 04:08:30 hm, we might want to default to busybox nc instead of the combat nc 2023-07-27 04:08:33 compat nc 2023-07-27 04:08:43 combat netcat 2023-07-27 04:08:46 with big claws 2023-07-27 04:08:52 frightening 2023-07-27 05:44:11 any xfce4 users? the clock applets on the panel display all non-local timezones as UTC. it's really weird 2023-07-27 05:44:27 wondering if it's some thing xfce4 assumes should be installed but it isn't so it panics and defaults to UTC 2023-07-27 05:46:26 to clarify I can definitely display tiems from other timezones. like polybar worked fine. 2023-07-27 05:48:30 my pragmatic solution - just use conky which handles it fine 2023-07-27 06:51:58 ikke, so what is tagged and what is pinned? 2023-07-27 07:15:53 @LACampbell the xfce4-panel contains the clock applet, which relies on what provides time to the system. Does the xfce4 clock applet timezone conflict with tzdata? 2023-07-27 07:58:28 jdusablon_: I wonder if that's it. like it shows me the full list of timezonesa. but no matter which one I select, unless it's blank or local time, it defaults to UTC 2023-07-27 08:01:23 does apk del remove config files, /var/lib/ or cache ? 2023-07-27 08:02:32 How can I do to remove all about one package at all? 2023-07-27 08:02:56 apk del --purge 2023-07-27 08:10:27 got it 2023-07-27 08:20:16 Alpine Linux does not make the news 2023-07-27 08:20:16 https://drewdevault.com/2023/07/25/Alpine-does-not-make-news.html 2023-07-27 08:20:16 Alpine is boring because it just works and without scandals. 2023-07-27 08:23:21 well drew ruined it by writing a blog about it! 2023-07-27 08:32:00 Can an ISP provide IPv4 and IPv6 over a single wire? 2023-07-27 08:32:12 lan cat6 ? 2023-07-27 08:32:33 the two things have nothing to do with eachother 2023-07-27 08:32:35 so yes 2023-07-27 08:32:39 So, I have heard there are ipv4 and ipv6 networks 2023-07-27 08:33:06 "ipv4 ipv6" notrelatedto "single wire" not related to "cat6 wire standard" 2023-07-27 08:33:11 The ones which i know, v4 doesn't give v6 and v6 doesn't give v4 2023-07-27 08:34:01 Yes, the way I'm visualizing is there would be 2 lan ports on server 2023-07-27 08:34:13 then your visualisation is wrong 2023-07-27 08:34:14 one with v4 and other with v6 2023-07-27 08:34:18 Nope 2023-07-27 08:34:28 the things have nothing to do with eachother 2023-07-27 08:34:30 They all go over the same wire 2023-07-27 08:34:31 it's separate layers 2023-07-27 08:34:34 i mean 2023-07-27 08:34:38 nothing stops you from doing that 2023-07-27 08:34:46 you can run a v4 network on one eth and a v6 on another 2023-07-27 08:34:48 And both can be provided by ISP? 2023-07-27 08:34:51 if you really want to for some reason 2023-07-27 08:34:52 Yes 2023-07-27 08:34:54 ye 2023-07-27 08:35:03 The ones that don't are lazy 2023-07-27 08:35:09 Thanks folks for resolution of my doubt 2023-07-27 08:35:15 :P 2023-07-27 08:35:17 Lol 2023-07-27 08:35:19 i've never heard of an isp that has separate connections for v4/v6 2023-07-27 08:35:28 usually they just don't have v6 at all or whatever 2023-07-27 08:35:36 conceptually it does not make sense 2023-07-27 08:35:42 Here, I'm fighting hard for a static ipv4 2023-07-27 08:35:44 why would they have two separate cables that just connect to the same thing 2023-07-27 08:36:39 yeah, my bad 2023-07-27 08:36:46 psykose: ^ 2023-07-27 08:38:12 static ipv4 may be difficult, because they are becoming more expensive for ISPs 2023-07-27 08:39:19 i don't find static ipv4 to be very useful unless you have an actual business to run 2023-07-27 08:39:45 You mean useless? 2023-07-27 08:40:02 > don't find 2023-07-27 08:40:11 even the easiest of dyndns scripting to update a dns mapping every N time in case your ip changed took me like an hour 2023-07-27 08:40:23 at worst that gives like 10 minutes of downtime when it rolls over or whatever 2023-07-27 08:40:42 if you don't have some million-users website to serve then you don't even notice 2023-07-27 08:42:07 The thing is if an ISP has v6, they will by default surely not provide v4 2023-07-27 08:42:29 not necessarily 2023-07-27 08:42:41 So if you want to make your service global, you must buy static ipv4 2023-07-27 08:42:55 that's not true at all 2023-07-27 08:43:19 All features are accessible by v6 because most users don't host 2023-07-27 08:43:40 Take a look at my service: https://humaaraartha.in 2023-07-27 08:43:49 accessible only with ipv6 2023-07-27 08:43:57 ipv4 cannot reach it 2023-07-27 08:44:17 I also host my own DNS 2023-07-27 08:44:57 On superb alpine 2023-07-27 08:45:15 With super minimal mini_httpd 2023-07-27 08:45:32 super slow you mean 2023-07-27 08:46:03 What super slow? 2023-07-27 08:46:10 It is super fast! 2023-07-27 08:47:25 It's faster than a hypersonic missile 2023-07-27 08:47:37 mini_httpd is one of the slowest webservers out there 2023-07-27 08:47:50 i guess if you have like three requests an hour it doesn't make a difference 2023-07-27 08:47:54 The whole world is requesting me to buy this technology to upgrade their military might 2023-07-27 08:49:50 Come on, at high loads, it gives an issue 2023-07-27 08:50:08 I'm thinking about shifting to openBSD's httpd 2023-07-27 08:50:18 But tisn't available for alpine 2023-07-27 08:50:40 FastCGI, Openhttpd and alpine, perfection! 2023-07-27 08:52:22 well yeah, that's how you measure the difference 2023-07-27 08:52:42 all of them are the same for 'one request', how many you can get in N time is the difference 2023-07-27 08:53:16 what is alternative to pacman -Rs, apk del -r ? 2023-07-27 08:54:12 nothing, -r is the default behaviour 2023-07-27 08:54:20 er 2023-07-27 08:54:20 -s 2023-07-27 08:54:25 del -r is -ss 2023-07-27 08:54:51 -r is by default? 2023-07-27 08:55:21 read it again until it makes sense 2023-07-27 08:56:11 though i guess -r is not like -ss, slight difference between the two 2023-07-27 08:56:34 i know apk delete all dependencies by default 2023-07-27 08:56:51 what do you think -Rs does 2023-07-27 08:57:29 remove a package and its dependencies which are not required by any other installed package 2023-07-27 08:57:39 so what do you think is the difference 2023-07-27 08:57:52 its not default pacman behavior 2023-07-27 08:58:05 and it is for apk 2023-07-27 08:58:17 you sure? 2023-07-27 08:58:35 i see apk delete all dependancies automatically 2023-07-27 08:58:46 if just apk del 2023-07-27 08:58:53 which means it's the default behaviour? 2023-07-27 08:59:06 but pacman if pacman -R delete only 1 package 2023-07-27 08:59:10 indeed 2023-07-27 09:02:04 in pacman i need everytime do pacman -Rs package or even -Rsu, but in apk i just need only apk del and its remove all dependancies which not required by other installed pkgs 2023-07-27 09:08:24 Yes, that's due to how apk works 2023-07-27 09:27:47 https://wiki.alpinelinux.org/wiki/Dualbooting, why is formatting and mounting shown here before the basic setup, if in most cases apk add ... probably won't work, I think if swap the "Format and mount the HDD partition" and "Basic setup" items on this page it will be clearer, don't think that something will get worse from this, on the contrary it will be possible to run apk add e2fsprogs in 2023-07-27 09:27:53 most cases 2023-07-27 09:29:40 im sorry if im wrong, correct me if 2023-07-27 11:41:13 it's okay that i sometimes do apk upgrade -a instead just apk upgrade? 2023-07-27 11:41:22 it does not harm anything 2023-07-27 11:44:29 oh, good 2023-07-27 11:47:32 how i can add testing packages? 2023-07-27 11:48:14 My Alpine install is missing the execinfo.h. I tried installing musl-dev, but that didn't fix the issue. What should I do? 2023-07-27 11:48:42 there is no execinfo.h 2023-07-27 11:49:00 ascensionwired: you have to be on edge and then you just add the repository 2023-07-27 11:49:54 tyler[m]12: you have to patch execinfo out of program in question 2023-07-27 11:49:59 mean not edge but testing 2023-07-27 11:50:17 you have to be on edge and then you add the testing repository 2023-07-27 11:50:54 I don't even know what that file is, I just need it for compiling a program. Wasn't there a libexecinfo-dev package (I found a reference to that in a GitHub issue)? What happened to it? 2023-07-27 11:51:28 It was there, but it was removed 2023-07-27 11:52:10 which program 2023-07-27 11:52:41 psykoze, what is `@testing http://dl-cdn.alpinelinux.org/alpine/edge/testing`? 2023-07-27 12:01:46 Something you should avoid unless you are on edge 2023-07-27 12:08:05 challenge accepted! 2023-07-27 12:12:21 "tyler: you have to patch..." <- I removed `execinfo.h` from the source code of the program (plus some other patches) and now it compiled, thx 2023-07-27 12:17:49 i could make a stub for that i guess 2023-07-27 12:50:05 what i need to compile with musl? 2023-07-27 12:57:49 musl-dev 2023-07-27 13:17:29 how compile/build something with musl on alpine? 2023-07-27 13:18:23 musl is the libc, so nothing special 2023-07-27 13:19:58 echo 'int main() { printf("hello\n"); }' >a.c; gcc a.c 2023-07-27 13:20:21 gcc will do everything necessary 2023-07-27 13:21:38 google.com/?q=how+to+build+stuff+on+linux 2023-07-27 13:22:50 yeah, that one too 2023-07-27 13:22:59 apk add build-base for the basic c compiler setup, then whatever else from there, but it's not alpine specific 2023-07-27 13:25:28 gcc package is musl? 2023-07-27 13:26:11 no? 2023-07-27 13:26:17 gcc is a compiler; musl is a libc 2023-07-27 13:26:44 what do you actually want to achieve? 2023-07-27 13:45:05 want use cmake to compile with musl 2023-07-27 13:49:27 you compile with a compiler, musl is not a compiler 2023-07-27 13:50:40 i think what you mean to say is that you want to compile something against musl instead of glibc 2023-07-27 13:51:05 which you can only do if the thing is not doing glibc-specific stuff. 2023-07-27 13:56:30 more exactly "which you can always do by doing nothing and just compiling it", because that's just the default 2023-07-27 13:56:42 you just.. build the project 2023-07-27 13:56:47 there's nothing special 2023-07-27 13:57:01 there's no magic secret sauce to """"compile with musl"""" 2023-07-27 14:10:53 just asking bucause some musl-gcc etc packages in other distros 2023-07-27 14:11:52 Because other distros are glibc-based and gcc will link against glibc by default 2023-07-27 14:21:20 in alpine no glibc? 2023-07-27 14:21:26 no 2023-07-27 14:32:42 at best you can install a glibc shim on a musl system that may enable you to run a binary built on a glibc system. see the 'gcompat' package 2023-07-27 14:35:59 ascensionwired: musl is one of alpine Linux defining features 2023-07-27 14:52:39 invoked, if i dont want gcompat, i can compile something by myself? 2023-07-27 14:56:55 Yes, if that's possible, that's the preferred way 2023-07-27 15:18:46 is there a vi clone/like editor in the repo (except vim/vis)? 2023-07-27 15:19:05 also i know and like busybox vi, but its a bit limited eh 2023-07-27 15:19:20 busybox vi works suprisingly well tho... 2023-07-27 15:20:09 neovim is in community 2023-07-27 15:21:59 was thinking of something less bloaty than vim 2023-07-27 15:22:11 nano 2023-07-27 15:22:24 openvi is in testing, apparently 2023-07-27 15:22:34 i'm surprised to see no nvi 2023-07-27 15:23:21 Habbie: oh interesting 2023-07-27 15:26:17 frag: there's also sam(1) from plan9port 2023-07-27 15:28:09 (* definitely not a vi clone but not a major leap either) 2023-07-27 15:34:10 invoked: dont think sam is in repo? 2023-07-27 15:35:39 plan9port is. 2023-07-27 15:36:09 aha 2023-07-27 15:36:38 to get started you probably want to read the manpage and run it 'sam -d' 2023-07-27 15:37:18 once you understand what makes it different you might like it a lot, i dunno. just try it 2023-07-27 15:50:07 there is no llvm on alpine? 2023-07-27 15:50:47 did you look? 2023-07-27 15:50:48 (clang) 2023-07-27 15:51:38 oh all exist 2023-07-27 15:56:32 do you even try things before typing into the chat? 2023-07-27 16:24:55 vieb looks good, works with my banks, except chat in one bank (seems only to work with firefox) 2023-07-27 17:23:59 to run make for build c programs i need build-base package? 2023-07-27 17:25:34 that's the easiest way to get a bunch of packages that you'll most likely need 2023-07-27 17:30:03 why don't you just try? 2023-07-27 17:35:07 if i want build with clang i dont need this package? 2023-07-27 17:35:31 https://pkgs.alpinelinux.org/package/edge/main/x86_64/build-base 2023-07-27 17:35:36 check the depends section 2023-07-27 17:35:44 that's what's installed when you install build-baase 2023-07-27 17:40:18 if i want build with clang i dont need this package? 2023-07-27 17:40:26 find out? 2023-07-27 18:17:30 how i can remove packages, like remove all configs and etc related to its? 2023-07-27 18:18:02 that's the default if you didn't modify any configs 2023-07-27 18:18:06 but if you did, apk del --purge 2023-07-27 18:18:51 ok 2023-07-27 18:50:44 Hey everyone. I'm trying to set up an OpenLDAP server in an LXC using Alpine following this: https://wiki.alpinelinux.org/wiki/Configure_OpenLDAP. Following all the steps (correcting for my domain) I get the following error: ldap_bind: Invalid credentials (49). Am I missing something? 2023-07-27 18:51:10 I didn't change the pw so it's still using "secret". I left that just to make sure. 2023-07-27 18:52:05 manwixh: you saw that error in response to which particular command you ran? 2023-07-27 18:52:26 Adding my org.ldif using "ldapadd -x -D "cn=Manager,dc=home" -w secret -f org.ldif" 2023-07-27 18:55:19 manwich: and you edited slapd.ldif so that olcSuffix and olcRootDN used "dc=home"? 2023-07-27 18:57:48 @minimal: I did change olcSuffix but didn't change olcRootDN. I'll try that. Thanks. 2023-07-27 18:58:30 manwich: well both were documented in the "Customising Configuration for Your Domain" section of that article... 2023-07-27 18:59:37 Yeah, I see that now. I've ran through this multiple times and missed it this time. The previous times, I've changed that and still ended up with the same error. Trying again on a clean install now. 2023-07-27 19:00:47 Wait, there are two spots that have olcRootDN. I *did* change the first one to my domain. The second one just has olcRootDN: cn=config. Do I need to change that? 2023-07-27 19:02:57 I don't have a copy of slapd.ldif in front of me, but no I don't believe you need to change that line 2023-07-27 19:03:03 did you check the openldap logs? 2023-07-27 19:06:23 Looking at the logs, I can see the connection being made, the BIND, then it logs this: "RESULT tag=97 err=49 qtime=0.000027 etime=0.000193 text=". I've edited the command using dc=home and also dc=$mydomain. Neither works. 2023-07-27 19:07:53 manwich: have you tried running slapcat to view the LDAP contents to verify that the changes were made ok? 2023-07-27 19:09:18 Yeah and it looks fine. I *think* I've found my issue though. I'm pretty sure I'm missing my *complete* domain. I've been using "dc=domain" but probably need to be using "dc=domain,dc=tld". 2023-07-27 19:09:58 Sometimes something just needs another set of eyes on it......... 2023-07-27 19:10:26 well the DN in LDAP obviously needs to match the DN you specify for commands like ldapadd 2023-07-27 23:38:30 Can the default vi editor copy-paste? None of the vi/vim instructions I find are working - all I get is "not implemented" errors 2023-07-27 23:40:23 no 2023-07-27 23:40:34 `yy` works, that's about it 2023-07-27 23:41:00 as well as `dd` in memory i guess 2023-07-27 23:49:36 yes it can yank and paste to and from your registers 2023-07-27 23:50:12 even has marks 2023-07-27 23:51:02 :) 2023-07-27 23:51:15 actual clipboard copypaste is much more integration 2023-07-27 23:51:56 “actual” clipboard 2023-07-27 23:52:27 for that, you can simply copy and paste using your terminal emulator 2023-07-27 23:52:43 the pasting you do in insert mode, of course 2023-07-27 23:55:02 busybox vi is missing the copy command t 2023-07-28 00:03:23 Was it mentioned before that busybox ed and vi are terrible? 2023-07-28 00:03:45 Just in case: busybox ed and vi are terrible. 2023-07-28 00:03:45 That's why you can `apk add nano' ;) 2023-07-28 03:38:15 Hi. I've read in the Wiki this is the appropriate place for questions, IIUC. 2023-07-28 03:39:41 greetings pelito, what's your question? 2023-07-28 03:39:51 I'm having a weird issue. I just replace my Archlinux32 setup in an old netbook CQ10-mini with an Atom N270 and 2GB of RAM with Alpine 3.18, using the LTS kernel. Inside the OS both /proc/meminfo and free report it has 846MB total instead. This didn't happen before and the BIOS still reports 2GB and the memtest passes. Anybody had a similar experience? Couldn't find anything in the bug tracker. 2023-07-28 03:46:46 On a side-note, I'm setting it up for some hobby self-hosting for me and my friends and it's super cool that WriteFreely is packaged in the repos. I was a pleasant surprise to find it there. 2023-07-28 03:48:05 For a little bit more context, since the Atom N270 is 32 bits the kernel image is i686. In case that's relevant. 2023-07-28 03:50:40 that's right. I know some older Atoms like that had insanely low memory limits (sometimes due to the chipset used), but AFAIK yours should be fine with up-to 2GB. 2023-07-28 03:51:01 so not sure why you are experiencing that... hopefully someone else can think of something 2023-07-28 04:13:21 pelito: when you get a chance, you could pastebin your dmesg output? 2023-07-28 04:14:55 Yes, I'm pretty sure it's not a chipset issue. As I said, I had it running Arch before and it detected the 2GB, and the BIOS setup screen reports the same. I tend to think it's a software issue. 2023-07-28 04:15:53 I'll try to paste the output tomorrow. I think there were a few messages about conflicting ranges for IO mappings or something like that, that may be related I guess. 2023-07-28 04:16:11 I'm in Argentina and should be sleeping. 2023-07-28 04:17:52 I don't think the Alpine kernel on i686 supports PAE; so you might be running into an issue there... Good chance dmesg has something about it 2023-07-28 04:18:14 Get some rest and we'll check it out tomorrow! 2023-07-28 04:19:24 I thought it might have been something PAE related but couldn't find any references there. Also, I thought PAE was only required for 4+GB. Thanks for the help! 2023-07-28 04:21:23 I think the kernel package included the config so I could also check that there. 2023-07-28 05:37:34 Weee... catfish finally works without patching in Alpine! 2023-07-28 09:07:26 I installed xorg-server and openbox. When I type xinit openbox, I see black screen but can't right click to bring up openbox menu. How do I fix that? 2023-07-28 09:10:06 Have you also ran setup-xorg-base? 2023-07-28 11:50:55 Mmm, and added your user to video/audio/input groups?  2023-07-28 11:51:30 that would be next if (s)he had stuck around long enough 2023-07-28 11:52:23 Ah, whoops. Well, maybe they will read the backlog later 🤷 2023-07-28 11:54:20 seems usb cabel 9m works for kb/mouse 2023-07-28 12:54:25 https://gitlab.alpinelinux.org/Saijin-Naib/aports/-/jobs/1071028 2023-07-28 12:54:34 This is builder test failure something I'm doing wrong? 2023-07-28 13:03:14 Wow, I am not awake yet. Is this builder test failure something I'm doing wrong? Sorry. 2023-07-28 13:34:42 No, that's something that broke due to a security fix in golang. Thought I had fixed it. 2023-07-28 13:54:21 Ahhh, okay.  2023-07-28 15:54:46 hey, why when im running `cmake ..` before builfing https://github.com/fastfetch-cli/fastfetch i have `-- Library:issong: LIBPCI` but i installed pciutils and if i `ls /usr/lib/libpci*` i get `/usr/lib/libpci.so.3 /usr/lib/libpci.so.3.10.0` output? 2023-07-28 15:55:25 *(`-- Library: missing: LIBPCI`) 2023-07-28 15:57:48 ascensionwired: you need pciutils-dev 2023-07-28 15:57:52 https://pkgs.alpinelinux.org/contents?branch=edge&name=pciutils%2ddev&arch=x86_64&repo=main 2023-07-28 15:58:51 For compiling, you need the *.so file 2023-07-28 15:59:44 that's not a hard dependency though. 2023-07-28 17:16:55 i asked here than i exitit, sorry 2023-07-28 17:17:12 *(exited) 2023-07-28 17:19:22 ascensionwired: so you didn't see my reply? 2023-07-28 17:24:33 yea 2023-07-28 17:24:36 sorry 2023-07-28 17:25:02 If you want to build against libpci, you need to install pciutils-dev, which provides libpci.so 2023-07-28 17:26:05 https://pkgs.alpinelinux.org/contents?branch=edge&name=pciutils%2ddev&arch=x86_64&repo=main 2023-07-28 17:34:36 ikke, ok, but as i said before 2023-07-28 17:34:57 libpci.so in /usr/lib/ already 2023-07-28 17:35:22 and pciutils-dev is dependancy of pciutils 2023-07-28 17:35:24 You mentioned /usr/lib/libpci.so.3 and /usr/lib/libpci.so.3.10.0 2023-07-28 17:35:45 it isnt it? 2023-07-28 17:36:48 ok, sorry its different things 2023-07-28 17:37:12 anyone else have chromium 115.0.5790.110 crashing on startup? 2023-07-28 17:37:26 I have: [1:1:0728/103632.009475:FATAL:credentials.cc(111)] Check failed: . : Invalid argument (22) 2023-07-28 17:37:43 rather annoyingly it's crashing in the (privileged) zygote process so I can't strace it 2023-07-28 17:40:27 doesn't seem to crash for me 2023-07-28 17:42:47 can i run neofetch when motd? 2023-07-28 17:43:29 afaik, motd is just static files 2023-07-28 17:43:41 you could have a cronjob update the motd 2023-07-28 17:44:10 i cant run script as motd? 2023-07-28 17:45:19 maybe with PAM 2023-07-28 17:46:41 in alpine no /etc/update-motd.d/ ? 2023-07-28 17:47:02 no 2023-07-28 17:49:28 ok i think have wrong idea to run script when motd, of i can run script or program from profile 2023-07-28 17:50:01 You can run it in ~/.profile 2023-07-28 17:50:10 yeah 2023-07-28 17:51:04 how i can install packages via `apk add package@testing` ? 2023-07-28 17:53:24 are you runnign a stable branch or edge? 2023-07-28 18:22:56 you mean kernel or repo? 2023-07-28 18:23:16 edge kernel and edge version of alpine 2023-07-28 18:24:03 edge repos 2023-07-28 18:24:11 if you are on edge, you can just add te testing repository 2023-07-28 18:24:22 You _can_ tag it if you want, but that's not required 2023-07-28 18:24:25 i on edge 2023-07-28 18:24:38 i already tagged and installed 2023-07-28 18:25:08 what is pinned and what difference with tagged? 2023-07-28 18:25:30 They can be synonums 2023-07-28 20:42:15 hi. what is the default login for the Mini Root Filesystem tarballs? https://dl-cdn.alpinelinux.org/alpine/v3.18/releases/armv7/alpine-minirootfs-3.18.2-armv7.tar.gz to be specific? 2023-07-28 20:46:08 root with no password didn't work when i tried 2023-07-28 20:48:49 sicelo: might not have one set at all 2023-07-28 20:49:40 meaning? 2023-07-28 20:49:49 how do i login? 2023-07-28 20:50:51 set password with passwd 2023-07-28 20:51:06 you'd need to set one before your can login 2023-07-28 20:51:26 chroot into it or something 2023-07-28 20:52:32 heh, ok 2023-07-28 20:52:34 the minirootfs is mostly used for docker containers 2023-07-28 20:52:37 rootfs is usually not a proper bootable system, it's used as container, chroot or other way which allows just dropping into it 2023-07-28 20:53:59 which one should i choose if i want to boot on armv7? i have a kernel that i can provide. 2023-07-28 20:54:17 Generic ARM contains raw apks ... no idea how it's meant to be used 2023-07-28 20:55:30 Netboot and Standard - according to descriptions requires networking. I won't have immediate access to networking 2023-07-28 20:59:03 guess i'll try the chroot option, with the Mini Root tarball (it does boot fine - just had the login problem outlined above) 2023-07-28 21:44:11 It seems like my kernel does have PAE enabled if I believe the config in /boot. It was added and then reverted, so it makes sense. I'm using 3.18. 2023-07-28 21:46:15 I guess the netiquette for pasting logs is using pastebin, right? 2023-07-28 21:46:40 yes 2023-07-28 21:46:50 not necessarily pastebin.com 2023-07-28 22:11:00 Here's my /proc/kmsg, I guess it's more useful than dmesg here (it lost some messages as well). https://paste.debian.net/1287256/ 2023-07-28 22:18:45 you can believe the config in /boot; you absolutely have PAE enabled 2023-07-28 22:18:58 otherwise you wouldn't have NX bit, afaik 2023-07-28 23:59:28 note that PAE does not mean a single app can use more than 4 GB of RAM (and usually not more than 3 GB, because 1 GB is reserved for the kernel), it just means the system can have more than 4 GB of RAM, and the total usage among the kernel and all applications can exceed 4 GB (assuming available physical RAM) 2023-07-29 00:34:27 dwfreed: his system only has 2GB of RAM anyway. It's only reporting 846MB in /proc/meminfo and free 2023-07-29 00:36:21 sounds like 2 1 GB DIMMs and one's not working 2023-07-29 04:07:07 thanks as ever, psykose 2023-07-29 04:07:31 Catfish I think was the first aport I tried to make. Happy to finally have it at least in testing and working very nicely 2023-07-29 08:04:43 dwfreed no, that's not the case. There's a single SODIMM, it's a netbook. The BIOS detects 2GiB. The memtest passes on those 2GiB. It was correctly detected two days ago from Archlinux32 before I reformatted to install Alpine. I'm 99.9% certain it's not a hardware issue. 2023-07-29 08:05:58 If PAE is not the culprit, the only other sign I see that something's wrong are the ACPI region warnings, but I have no idea if any of that maps to RAM and may affect that measurement. 2023-07-29 08:10:50 it could be pae 2023-07-29 08:10:55 in that pae is enabled and breaks it 2023-07-29 08:11:05 if you boot the 5.15 kernel from 3.17 does it detect 2gb 2023-07-29 08:22:29 pelito: good chance acpi_enforce_resources=lax will reduce or even eliminate those warnings in dmesg. Probably won't do anything for your RAM issue though. 2023-07-29 08:46:38 I'll try those later today and come back, thanks! 2023-07-29 10:00:16 hey, can i use sway on alpine, and hoe? 2023-07-29 10:01:34 yes, sway is in the repos. 2023-07-29 10:01:47 ACTION is unfamiliar with hoe 2023-07-29 10:04:11 zcrayfish, is it work out of box? or i need do something 2023-07-29 10:04:13 ? 2023-07-29 10:05:57 I dunno if it requires configuration or not. I'm a lxqt (labwc) user myself. 2023-07-29 10:06:40 you need to install and configure it yourself 2023-07-29 10:13:22 https://wiki.alpinelinux.org/wiki/Wayland https://wiki.alpinelinux.org/wiki/Sway something there like XDG_RUNTIME_DIR and eudev 2023-07-29 10:14:41 i don't know for sure if this is necessary 2023-07-29 10:16:01 I just use elogind personally. 2023-07-29 10:16:26 and udev as well. 2023-07-29 10:18:58 Me too (when I was using Alpine on desktop) 2023-07-29 10:20:37 is pam needed for it too? 2023-07-29 10:22:07 you don't need pam_rundir if you'r eusing elogind 2023-07-29 10:24:11 i also can setup xdg_runtime_dir manually, without elogind or pam_rundir 2023-07-29 10:26:30 ? 2023-07-29 10:30:28 yes you can 2023-07-29 10:30:41 you do need either elogind or seatd though 2023-07-29 10:37:48 seatd and elogind are needed for different purposes as far as i understand 2023-07-29 10:40:56 yes 2023-07-29 10:43:28 and i dont need elogind if config XDG_RUNTIME_DIR manually, as on wiki page i messaged bellow 2023-07-29 10:46:03 https://wiki.alpinelinux.org/wiki/Sway 2023-07-29 10:46:40 need i also eudev, or i can with mdev too? 2023-07-29 10:56:03 you do need elogind or seatd no matter how you set XDG_RUNTIME_DIR; you don’t need udev or mdev or whatever 2023-07-29 11:10:36 from https://wiki.alpinelinux.org/wiki/Sway i need both seatd and elogind when "Configure XDG_RUNTIME_DIR" 2023-07-29 11:11:15 but they say i can configure xdg_runtime_dir witjout elogind 2023-07-29 11:13:33 export XDG_RUNTIME_DIR=path 2023-07-29 12:02:01 somewhy cant run sway 2023-07-29 12:03:12 error: wl_display_roundtrip failedialize.ED=1 to suppress this check 2023-07-29 12:07:56 eudev needed 2023-07-29 12:08:14 sway not support mdev 2023-07-29 12:13:00 or mb i need do something 2023-07-29 12:21:04 hi, I cant find "tinycc" in Alpine? (why not?) 2023-07-29 12:21:50 ncrowl: because it is not there 2023-07-29 12:23:18 minimal: I noticed, but it has existed (a websearch returned something) ((why not)) 2023-07-29 12:23:42 ncrowl: actually I'm wrong, it is there as package tcc 2023-07-29 12:24:07 it helps if you search for the right thing :-) 2023-07-29 12:24:51 minimal: nope, apk search tcc, returns something else "distcc" (dono what that is) 2023-07-29 12:25:27 https://pkgs.alpinelinux.org/package/edge/testing/x86_64/tcc 2023-07-29 12:26:20 are you running Edge? if not then that explains it - tcc is in testing, so Edge only 2023-07-29 12:28:12 edge? (dono, I running the default, on cloud VM, and just tested yesterday native on a stationary computer,, Was thinking of moving away from Debian (havto run "unstable" to have Nvidia drivers,,, 2023-07-29 12:29:53 ncrowl: default what? I'm referring to the version of Alpine - Edge is the "development" version of Alpinr which has a testing repo, whereas releases, like 3.18.x, do not have the testing repo 2023-07-29 12:30:47 minimal: oki, I havto look it up (im new to this distro) 2023-07-29 12:31:20 also you want to use the official Nvidia driver? they're binary-only and so won't work on Alpine AFAIK, whereas the likes of nouveau will 2023-07-29 12:34:22 ncrowl: ^^^ 2023-07-29 12:35:54 minimal: oki, 2023-07-29 12:37:22 Debian often just blocks none-free software(for whatever reson), but Alpine blocks "Nvidia" also? 2023-07-29 12:38:21 ncrowl: no. Alpine uses a different C library (musl) than most Linux distros (glibc). Nvidia built their drivers for glibc and AFAIK do not provide a version built against musl 2023-07-29 12:39:07 minimal: oki, must have source-code (none binary) 2023-07-29 12:39:54 nouveau is the "3rd party" developed drivers for Nvidia cards 2023-07-29 13:34:56 is it possible to sway work with mdev? 2023-07-29 13:35:20 or xorg and wayland support only eudev 2023-07-29 14:02:33 ask in #sway 2023-07-29 14:06:40 No need for that 2023-07-29 14:07:02 it most likely requires eudev 2023-07-29 14:07:07 look at the build recipe: https://git.alpinelinux.org/aports/tree/community/sway/APKBUILD 2023-07-29 14:19:08 > A group of White Men hold a private event on private property and the media and dregs of society go into a spin. https://file.silentsilas.com/LlS.jpg 2023-07-29 14:21:17 frag: what has this to do with Alpine? 2023-07-29 14:21:46 that's even offtopic for the -offtopic channel. jeeze. 2023-07-29 14:23:01 minimal: oh sorry, wrong window :/ 2023-07-29 14:47:27 Howto set the $PATH variable, in busybox version of a shell? (that Alpine use) 2023-07-29 14:48:30 ncrowl: if you just want it for one user, ~/.profile will work. 2023-07-29 14:52:17 zcrayfish: oki, ehe, I wanna use countless alias, and pywall -themes. Il better switch back to Bash,, 2023-07-29 16:08:29 local ".xinitrc" and ".Xresources" are skipped (when running Dwm -window manager, but not XFCE) any hint why?? 2023-07-29 16:23:45 i decided to be lazy and do a raid5 on 5x10TB drives in a USB3 enclosure using mdadm and alpine because....i broke my budget on hard drives, lol 2023-07-29 16:38:25 ncrowl: usually dwm would be invoked by xinitrc and Xresources would be loaded by xinitrc before doing that... are you using a desktop manager? 2023-07-29 16:40:38 as an example my .xinitrc looks like this: https://bpa.st/AFXQ 2023-07-29 16:40:43 and then I start X with `startx` 2023-07-29 16:51:01 can i say setup-alpine change device manager? 2023-07-29 16:53:07 i need DEVOPTS= setup-alpine? 2023-07-29 16:56:53 Like setup-udev? 2023-07-29 16:57:47 humm: you mean setup-devd 2023-07-29 16:58:03 perhaps 2023-07-29 16:58:22 that replaced setup-udev several releases ago 2023-07-29 16:58:50 Alpine 3.16+ 2023-07-29 16:58:52 DEVDOPTS= one of mdev, mdevd, or udev 2023-07-29 16:58:58 before you run setup-alpine, yeah 2023-07-29 16:59:34 https://github.com/alpinelinux/alpine-conf/blob/master/setup-devd.in#L20 2023-07-29 17:31:32 sad that sway dont work with mdev 2023-07-29 17:35:21 hmm, how does it not work? 2023-07-29 17:36:53 dont know why, but it not work 2023-07-29 17:37:07 what happens when you try to start it? 2023-07-29 17:39:57 ACTION notes that eudev is a build depend there. dunno about runtime 2023-07-29 17:42:46 ahh 2023-07-29 17:42:59 I wonder if it depends on udev rules to get device access or somesuch 2023-07-29 17:43:56 elly: thx for the answer. I used lightdm, tested also "startx dwm" (will lauch frozen, -hadto press reset on the computer) dodo 2023-07-29 17:44:47 (root user wasn't frozen, but but.. 2023-07-29 17:44:59 dontzap strikes again 2023-07-29 17:46:19 having zap enabled is good :P 2023-07-29 17:47:36 I even have magic sysrq enabled.... for what good it does. My laptop keyboard doesn't have that key. 2023-07-29 17:48:55 elly, i think xorg needs udev too 2023-07-29 17:49:20 mb all wayland needs udev 2023-07-29 17:49:26 to work 2023-07-29 17:49:47 i dont know, mb exist way to run sway with mdev 2023-07-29 17:51:20 yeah, that's why I was asking 2023-07-29 17:51:45 but i cant run it with mdev only if install udev 2023-07-29 17:52:04 dont know why its not dependancy 2023-07-29 17:54:13 setup-xorg-base install udev instead mdev too 2023-07-29 17:56:26 as i understand xorg definitely needs udev, and it won't work without it 2023-07-29 17:56:39 mb am wrong, fix me 2023-07-29 17:56:59 I wonder if libinput has something to do with it 2023-07-29 17:57:02 ACTION is too lazy to check 2023-07-29 17:58:42 yes wlroots needs libinput 2023-07-29 17:59:34 mdev suck in this? 2023-07-29 17:59:57 or mdev not like replacing udev 2023-07-29 18:00:15 ACTION shrugs 2023-07-29 18:01:09 https://gitlab.freedesktop.org/wlroots/wlroots/-/issues/2257 2023-07-29 18:01:42 xorg does not need udev, afaik 2023-07-29 18:01:56 our package in alpine may depend on udev because if you don't have udev you have to write an xorg.conf by hand (I think) 2023-07-29 18:02:49 sounds right 2023-07-29 18:11:00 ok elly, i meant, xorg in alpine needs udev 2023-07-29 23:31:40 Going through Firefox' settings I realized that Alpine ships the browser with "data collection" switched on. Wouldn't it be better to have that off by default? Debian does it that way, as far as I remember. 2023-07-29 23:34:06 ACTION wonders if vendor-prefs.js controls that 2023-07-29 23:39:26 datareporting.healthreport.uploadEnabled & app.shield.optoutstudies.enabled would be the relevant settings if so 2023-07-29 23:39:49 Good to know, thanks. 2023-07-29 23:47:18 no problem; and I just confirmed that adjusting vendor-prefs.js with the above toggles would be enough to disable everything under "Firefox Data Collection and Use" 2023-07-29 23:49:31 The maintainer of the firefox package quit today :( 2023-07-29 23:49:36 Oh… 2023-07-29 23:49:52 I was just wondering why they're not in the channel. 2023-07-29 23:50:08 Beyond what I just told you, that's all I know. :( :( 2023-07-29 23:50:20 That's a pity. 2023-07-29 23:50:33 indeed. I'll still try to put a PR in against Firefox though 2023-07-29 23:52:06 Cool. 2023-07-29 23:55:07 https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/49297 2023-07-30 00:00:55 That was quick. Thanks and goodbye. 2023-07-30 00:01:42 hehe. hope she or he knows that just because a PR was put in doesn't guarantee 100% that it will happen. 2023-07-30 00:01:46 =) 2023-07-30 00:11:12 "hehe. hope she or he knows..." <- I know that pain 🤣 2023-07-30 00:11:28 :) 2023-07-30 00:11:52 I mean, I would not even be the slightest bit bothered if the PR got denied because it's a game of whack-a-mole honestly 2023-07-30 00:12:09 every time FF changes a setting name or comes up with some new convoluted way to upset privacy advocates 2023-07-30 00:16:31 BTW, I'll be glad when your 49293 gets merged, since I find myself using Mousepad a lot. 2023-07-30 00:20:34 Sweet! I'm taking a poke around to see what other fun build time things I can ask to enable for extra features  2023-07-30 00:20:51 Also, Catfish is in testing, with both plocate and zeitgeist, so it is really fast 2023-07-30 00:22:13 I'd appreciate more testers 2023-07-30 00:24:57 I'll check it out in a bit!! 2023-07-30 00:36:39 Seems to work pretty well. I tested it looking for files in ~/.steam ; lots of symlinks in there! 2023-07-30 00:39:27 zcrayfish: there are several 3rd Firefox config file repos for more secure/private configs 2023-07-30 00:39:36 s/3rd/3rd party/ 2023-07-30 00:57:29 "Seems to work pretty well. I..." <- Sweet! 2023-07-30 00:57:29 Can you see zeitgeist running in your Task Manager?  2023-07-30 00:57:29   2023-07-30 00:57:43 Does searching within archives work for you?  2023-07-30 01:14:04 I see zeitgeist. Not having much luck searching inside zip files though 2023-07-30 01:23:44 It's an option that needs to be turned on from the preferences, if you've not dug in there yet 2023-07-30 04:05:38 Saijin_Naib[m]: yes I had that on 2023-07-30 04:08:18 also I realised that I forgot to increment pkgrel. fun. 2023-07-30 05:00:03 zcrayfish: Hmm... It is working for me with zip at least. Not sure what other compression formats it supports. Can you launch catfish from the terminal and capture any log messages that might help? 2023-07-30 07:42:24 https://git.alpinelinux.org/aports/commit/?id=9e88698cc296a8492a8f70938c58f5755f39609d : who is maintaining chromium now. 2023-07-30 07:44:10 oleovit, google evil robots 2023-07-30 07:45:16 for real? can anyone remember my last appearance here? 2023-07-30 07:45:33 and now, google evil. 2023-07-30 07:45:39 Maybe you're not in the dimensension you think you are 2023-07-30 07:45:59 mozilla evil much more 2023-07-30 07:46:39 maybe everyone is overreacting to this proposal,?? 2023-07-30 07:46:56 What proposal? 2023-07-30 07:47:02 ... 2023-07-30 07:47:37 Sorry oleovit if you maid a proposal some time before, I didn't see it 2023-07-30 07:47:53 Maybe propose it again 2023-07-30 07:48:20 what? _the proposal_ 2023-07-30 07:51:47 WEI. my last appearnce regarding this ip at startup - adressed @google, fixed @google 2023-07-30 07:52:13 we need a f.. browser. mozilla is not the answer 2023-07-30 07:53:13 so please, find a new maintainer for chromium or compile ungoogled-chromium or whatever. and dont over-react to this wei thing 2023-07-30 07:56:07 heh, people already complaining and it has been less than 24-hours. 2023-07-30 07:57:22 What happened 2023-07-30 07:57:37 No sé :) 2023-07-30 07:59:57 psykose quit :( 2023-07-30 08:02:08 wat 2023-07-30 08:02:10 why 2023-07-30 08:03:46 idk. I'm sure there was a good reason. I wish her well. 2023-07-30 08:08:09 quinq: "the proposal" is web environment integrity but our friend here could not be bothered to actually spell it out for some reason 2023-07-30 08:10:52 oh ok, thank you pj, I'll look for that 2023-07-30 08:15:13 Holy crap alpine is doomed 2023-07-30 08:17:06 Alpine will get through this. 2023-07-30 08:17:19 consus[m], breath in a paper bag :) 2023-07-30 08:17:39 Im eating the bag right now 2023-07-30 08:18:08 Of course it will, but still, a minute of appreciation 2023-07-30 08:18:17 Well that's one way to get your daily fibre budget... 2023-07-30 08:50:18 ps: ok. https://github.com/chromium/chromium/commit/6f47a22906b2899412e79a2727355efa9cc8f5bd . anyway: firefox can't be forced to the user. there must be "choice". and time has to go by (eu-laws, et cetera). please find a new maintainer (like psykose before) who cares about chromiums functionality, patches, etc. . have a nice day and thx for your work (first time neing amgry of blind activism, 2023-07-30 08:50:24 because i still remember not being taken seriously by anyone when it came to this little damn ip (except google, which makes the whole thing out of this world). 2023-07-30 08:52:41 Must be fun doing those constant hit and runs. 2023-07-30 08:58:25 I'd read this child story, This Little Damn IP 2023-07-30 09:28:53 Sorry if I miss the train: are the reasons why psykose left public? Are they definitely out of the alpine project? 2023-07-30 09:37:49 All I know is she abandoned maintainership and contributorship of all projects on the gitlab, removed her privs on the wiki, and left all #alpine-* channels. 2023-07-30 10:06:36 Maybe psykose will explain in their own time 2023-07-30 10:14:36 Hi, i run latest i686 build and I'm trying to mount some old hard drive. Filesystem is FAT and is using UTF8 encoding. ``mount: mounting /dev/sda1 on /hd1 failed: Invalid argument`` 2023-07-30 10:15:05 what does it say at the end of dmesg 2023-07-30 10:15:41 the dmesg output is: FAT-fs (sda1): utf is not a recomended IO charset for FAT filesystems, filesystem will be case sensitive! 2023-07-30 10:16:01 that is just a warning and can be ignored. 2023-07-30 10:16:04 anything else? 2023-07-30 10:16:38 just this 2023-07-30 10:16:48 And the disk won't mount. 2023-07-30 10:21:03 zcrayfish: how else they can protect themselves from reasoning, can't change opinion if you can't see messages from the other side 2023-07-30 10:21:33 right 2023-07-30 10:21:55 although there is a better platform to just spew bs which no one wants to listen to :> 2023-07-30 10:23:28 I would shit my bricks if I would receive messages "from the other side" 2023-07-30 10:23:44 hahaha 2023-07-30 10:24:03 probably next up in the congress, now when they acknowledged ufos and aliens or "aup" as they call it today - next they will admit "we have contact with ghosts" :D 2023-07-30 10:24:35 this looks like fun ;) https://www.lightningmaps.org/#m=oss;t=4;s=201;o=0;b=;ts=0;ts24=1;z=10;y=59.3836;x=17.7779;d=2;dl=2;dc=0; 2023-07-30 10:24:38 I'm more interested in how do we get those ghosts to maintain chromium 2023-07-30 10:25:30 feed them ectoplasma? Havent you learned anything from the documentary ghostbusters and ghostbusters II ? ;-) 2023-07-30 10:26:19 I don't remember if ghostbusters made ghosts do useful stuff 2023-07-30 10:26:51 I might have seen ghostbusters with different meaning of the word "bust" 2023-07-30 10:27:21 sure, that green blob was helpful even if he ate anything he could find 2023-07-30 10:27:32 like a ectoplasma edition of a dog :P 2023-07-30 10:27:37 https://yewtu.be/watch?v=0tdyU_gW6WE 2023-07-30 10:29:44 Apachez, I can remember some ghosts driving taxis 2023-07-30 10:30:19 And opening doors to other dimensions from where world destroyers would come from 2023-07-30 10:30:50 I wouldn't say that's unuseful 2023-07-30 10:30:56 Maybe counter-productive though 2023-07-30 10:57:25 RhinoFlip: what is full command you used to mount? 2023-07-30 11:17:38 When setting up WiFi, I noticed that some of the commands I ran to do that would not work through `doas`, but require me to use `su` for apparently more "serious" root access. Why is that? 2023-07-30 11:18:02 msiism: there is no difference 2023-07-30 11:18:17 there is no root-light and root-full 2023-07-30 11:18:18 Examples are `wpa_supplicant` and `iwlist`. 2023-07-30 11:18:28 Yeah, that's what I thought. 2023-07-30 11:18:43 msiism: so can you show what commands you run and what errors you got? 2023-07-30 11:19:14 Yeah. Just gotta get a cable connection real quick… 2023-07-30 11:20:26 i've noticed that trying to run pipelined commands that redirect stdout don't work with doas. can't remember if they worked with sudo 2023-07-30 11:21:17 j_v: it's because the redirection is done by your sheel 2023-07-30 11:21:19 shell 2023-07-30 11:21:21 which runs as your user 2023-07-30 11:21:44 they only elevate the direct command it is executing 2023-07-30 11:21:52 yep, figured it was something like that 2023-07-30 11:21:53 (sudo foo) >to file 2023-07-30 11:22:15 ah, subshell, should've thought of that 2023-07-30 11:22:22 no, that's not the solution 2023-07-30 11:22:26 more like showing what is happening 2023-07-30 11:22:54 a command like tee could be used to write as root 2023-07-30 11:23:02 usually just use `su -c 'cmd -args | less'` 2023-07-30 11:23:05 Okay, `wpa_supplicant -i wlan0 -c /etc/wpa_supplicant/wpa_supplicant.conf` "now" works with `doas`. It probably did before then. 2023-07-30 11:23:06 doas foo | doas tee file 2023-07-30 11:23:23 there is also sponge if you need to write to the same file 2023-07-30 11:23:47 cat myfile | sed 's/foo/bar/' | sponge myfile 2023-07-30 11:23:51 cat myfile | sed 's/foo/bar/' | doas sponge myfile 2023-07-30 11:24:20 sponge first reads stdin before it overwrites the file 2023-07-30 11:24:22 yeah, i avoid reading and writing same file... botched stuff doing that before 2023-07-30 11:24:48 > truncates the file before any command is able to read it 2023-07-30 11:26:19 Okay, `iwlist scan` works through `doas` just as well. What was I doing…? 2023-07-30 11:26:46 I don't know, hard to tell without any error message 2023-07-30 11:27:12 ACTION tries some more. 2023-07-30 11:28:39 Okay, it was about output redirection. 2023-07-30 11:29:20 $ doas wpa_passphrase "NotMyWLAN" >> /etc/wpa_supplicant/wpa_supplicant.conf 2023-07-30 11:29:28 bash: /etc/wpa_supplicant/wpa_supplicant.conf: Permission denied 2023-07-30 11:29:33 yeah, that's what j_v mentioned 2023-07-30 11:29:37 Right. 2023-07-30 11:29:43 Okay, problem solved then. 2023-07-30 11:30:54 Hm… `iwlist scan` even works as a normal user. 2023-07-30 11:31:58 wpa_passphrase "NoMyWLAN" | doas tee -a /etc/wpa_supplicant/wpa_supplicant.conf 2023-07-30 11:32:22 Oh, that's why you'd use `tee`. I see. 2023-07-30 11:33:59 so, trying to pipe dmesg to pager as normal user... `doas dmesg | less` obviously fails unless persist is in affect 2023-07-30 11:34:37 Why would that fail? 2023-07-30 11:34:58 Hoe laat en waar spreken we af? 2023-07-30 11:35:16 password prompt lost 2023-07-30 11:35:46 Wrong channel 😑 2023-07-30 11:37:25 the pipe hides the password prompt 2023-07-30 11:38:30 whereas `doas sh -c 'dmesg | more'` works fine 2023-07-30 11:38:46 might as well use su -c at that point 2023-07-30 11:39:17 Hm… strangely using `tee` this way doesn't work. Maybe because wpa_passphrase is reading from stdin: https://paste.debian.net/plainh/8776db50 2023-07-30 11:49:33 @jv ``mount /dev/sda1 /hd1`` 2023-07-30 11:51:34 Try with mount -v 2023-07-30 11:51:52 This might show a bit more about what the mount wrapper actually does 2023-07-30 11:52:01 RhinoFlip: might want to check if kernel module is loaded for fat filesystem 2023-07-30 11:56:52 ikke: installed sudo just to compare this redirection scenario... `sudo dmesg | more` works as expected, 2023-07-30 11:58:14 but concidering that doas is about 1/5th the size of sudo, I am fine with having to get more creative with stuff like pipelined commands 2023-07-30 12:04:46 tahnk you j_v , I'll check that out 2023-07-30 12:05:11 RhinoFlip: also might try adding the '-t fstype' 2023-07-30 12:05:38 RhinoFlip: mount -t vfat /dev/sda1 /hda1 2023-07-30 12:08:11 RhinoFlip: lsmod | grep fat 2023-07-30 12:18:05 have runit on alpine? 2023-07-30 12:18:18 hi everyone 2023-07-30 12:20:41 https://pkgs.alpinelinux.org/packages 2023-07-30 12:21:18 can search easily there, allows wildcards also 2023-07-30 12:22:12 but, yes, runit packages exist for alpinelinux 2023-07-30 12:34:06 j_v, dmesg output is ``vfat 24576 0 2023-07-30 12:34:06 fat 73728 1 vfat 2023-07-30 12:34:06 `` 2023-07-30 12:34:23 sorry, don't know ho wto format it properly 2023-07-30 12:34:29 how to* 2023-07-30 12:39:34 filip_: which probably means that the module is loaded, but just to be sure should run `lsmod | grep fat` 2023-07-30 12:43:07 sorry, this is output of ``lsmod | grep fat`` not dmesg 2023-07-30 12:44:04 filip_: ok so module is loaded (could be issue with drive, so want verbose output), suggest try adding -v option to mount command, like quinq suggested 2023-07-30 12:44:28 and i said this is kernel ``i686`` which is not, i confuesed another lw distro 2023-07-30 12:44:53 i already tried ``-v`` and there is no additional info 2023-07-30 12:45:51 filip_: no sure then, maybe need dosfstools installed 2023-07-30 12:46:14 j_v, sure, im on it 2023-07-30 12:46:27 ```pendrive:/# mount -v -t fat /dev/sda1 /hd1 2023-07-30 12:46:27 ``` 2023-07-30 12:46:27 mount: mounting /dev/sda1 on /hd1 failed: No such device 2023-07-30 12:46:27 pendrive:/# 2023-07-30 12:47:32 Do you have a /dev//sda1? 2023-07-30 12:47:37 so, there is no /dev/sda1... what is output of `fdisk -l /dev/sda`" 2023-07-30 12:47:57 quinq, you got this, i will back off 2023-07-30 12:48:13 ```Disk /dev/sda: 37 GB, 40020664320 bytes, 78165360 sectors 2023-07-30 12:48:13 4865 cylinders, 255 heads, 63 sectors/track 2023-07-30 12:48:13 Device Boot StartCHS EndCHS StartLBA EndLBA Sectors Size Id Type 2023-07-30 12:48:13 Units: sectors of 1 * 512 = 512 bytes 2023-07-30 12:48:14 ``` 2023-07-30 12:48:39 ``/dev/sda1 * 0,1,1 1023,254,63 63 78140159 78140097 37.2G 7 HPFS/NTFS 2023-07-30 12:48:39 `` 2023-07-30 12:49:05 ahhh, not vfs 2023-07-30 12:49:13 mean vfat 2023-07-30 12:49:18 ntfs 2023-07-30 12:49:22 ok 2023-07-30 12:49:36 so it will be than ``mount -t ntfs /...`` ? 2023-07-30 12:50:05 try it 2023-07-30 12:50:11 Yeah, though you shouldn't need to force the fs type 2023-07-30 12:50:12 thank you 2023-07-30 12:50:18 this works :) 2023-07-30 12:50:18 yep 2023-07-30 12:51:04 could be using bb mount... can't remember is that auto loads fs modules like util-linux does 2023-07-30 12:51:16 right 2023-07-30 12:51:22 s/remember is/remember if/ 2023-07-30 12:52:13 proly not 2023-07-30 13:03:58 zcrayfish https://gitlab.xfce.org/apps/catfish/-/issues/98 2023-07-30 13:04:33 It looks like Catfish only supports Deflate, bzip2, and LZMA right now due to using inbuilt Python zipfile function 2023-07-30 13:05:15 Your archives maybe are not compliant with what this can understand? 2023-07-30 13:05:15 Most all of mine are 7z, so nothing in them gets searched, but I made zips to test with so missed that, haha 2023-07-30 13:05:15   2023-07-30 13:34:02 If I want to modify a service script in /etc/init.d, should I just modify it in place or is there a better way to do that, like maybe with some overlay file? 2023-07-30 13:35:33 If I do modify it in place, what happens when I upgrade the package for that script? Will it be overwritten? 2023-07-30 13:40:50 afaik, updated files in /etc are moved to .apk-new 2023-07-30 13:40:57 I mean $file.apk-new 2023-07-30 13:43:43 Though it might be for *any* file that was changed, not only in /etc 2023-07-30 14:54:10 after a few hours since my chromium-topic, one question: has something happend with psykose? i see "deleted user" and then "ghost user" - is she/he still on board? 2023-07-30 14:55:12 She left 2023-07-30 14:55:27 that's .... 2023-07-30 14:56:20 i'm shocked. hurts. have to quit 2023-07-30 14:57:14 I don't think it's that big of a deal but ok 2023-07-30 14:59:05 hum 2023-07-30 14:59:30 I can take on maintainership of the chromium package if someone who knows more about alpine packaging than I do is willing to consult with me on it 2023-07-30 14:59:48 What do you need? 2023-07-30 15:00:06 right this second, nothing - I'm already familiar with chromium and the APKBUILD for it 2023-07-30 15:00:14 I had been working with psykose to upstream the alpine patches anyway 2023-07-30 15:00:47 but in general I am going to need someone who knows about Alpine Things to keep me from making foolish mistakes 2023-07-30 15:01:25 well, ping me if you will need help, I'm not an active anymore in alpine but I know quite a bit about abuild and around it 2023-07-30 15:01:48 cheers 2023-07-30 15:01:55 I may also just ask in #alpine-devel if I need help 2023-07-30 15:02:06 I wonder who will take over main/rust 2023-07-30 15:02:14 not it :P 2023-07-30 15:06:07 is it proper to just make a PR to set the maintainer to me? or do I need to be nominated by someone else? 2023-07-30 15:07:27 It's perfectly fine 2023-07-30 15:08:50 great 2023-07-30 15:16:07 gitlab is just different enough from both github and gerrit that I keep struggling with it 2023-07-30 15:16:10 perhaps some day 2023-07-30 15:17:47 are you using the new navigation? 2023-07-30 15:17:54 I don't know 2023-07-30 15:19:15 well, on gitlab.a.o new navigation is opt-in in right top corner after clicking your profile picture 2023-07-30 15:19:32 aha! 2023-07-30 15:19:53 I've enabled that, let's see if that is a bit less confusing 2023-07-30 15:19:54 thanks 2023-07-30 15:20:12 it's definitely even more different than other git forges 2023-07-30 15:20:38 but I've been using that interface at work and I think it's slightly better than previous one 2023-07-30 15:21:05 I'll see how I go with it 2023-07-30 15:21:18 we use gerrit at work, which has its own idiosyncrasies 2023-07-30 15:22:21 You should also bump pkgrel when taking maintainership of package 2023-07-30 15:22:27 so it updates that information in package itself 2023-07-30 15:27:26 ah okay 2023-07-30 15:32:48 the main difference between gerrit and gitlab/github is that gerrit is designed around the idea that you are a contributor to a large project that you don't have your own "fork" of, so it doesn't have this merge request flow or similar 2023-07-30 15:34:00 but I think I correctly made a merge request 2023-07-30 15:35:47 I'm vaguely familiar with gerrit from other Google projects 2023-07-30 15:40:48 do I need to actually send the MR to someone or does it get reviewed without my doing so? 2023-07-30 15:43:32 Someone will review it at some point 2023-07-30 15:44:09 great 2023-07-30 15:44:59 elly: thanks for doing that. 2023-07-30 15:45:50 I'm always more than happy to help you and as you noticed, others as well 2023-07-30 15:46:21 sure thing 2023-07-30 15:46:39 it looks like firefox will need a maintainer as well 2023-07-30 15:46:57 Yes. Maybe ncopa will take that one back 2023-07-30 15:47:03 cool 2023-07-30 15:47:30 Then there's rust and llvm as big things left 2023-07-30 15:47:54 yep 2023-07-30 15:48:03 What is rust actually necessary for? 2023-07-30 15:48:15 I have a little experience with the llvm build system but not enough to confidently maintain it 2023-07-30 15:48:17 Currently nothing 2023-07-30 15:48:26 But it will be required for building Linux drivers 2023-07-30 15:48:32 quinq: Linux might need it soon 2023-07-30 15:48:40 Hopefully only unuseful drivers 2023-07-30 15:48:51 The Apple DRM driver will be written in Rust 2023-07-30 15:48:52 Asahi GPU driver is written in rust 2023-07-30 15:49:19 it will also soon be a build dep for chromium, if it isn't already, and I think already is for firefox (?) 2023-07-30 15:49:27 I think mesa also needed rust for some new stuff 2023-07-30 15:49:38 yeah, rusticl 2023-07-30 15:49:38 Why ;_; 2023-07-30 15:49:46 there are things already needing rust 2023-07-30 15:49:49 They couldn't wait for rust to have a mature build environment 2023-07-30 15:49:51 elly: yes, but this is specifically about rust in main 2023-07-30 15:49:56 ahh 2023-07-30 15:50:33 I could probably try to take on rust but not llvm 2023-07-30 15:51:26 mesa requires rust in main 2023-07-30 15:52:52 yeah, as Newbyte mentioned, rustic 2023-07-30 15:53:19 s/rustic/rusticl/ 2023-07-30 15:54:38 https://docs.mesa3d.org/rusticl.html#rust-update-policy hmm, looks like mesa is pretty conservative about what versions of rust they depend on 2023-07-30 16:07:12 I do recall pycrypto requiring rust, but I don't see it as a dependency 2023-07-30 16:07:40 ah, py3-setuptools-rust 2023-07-30 16:08:33 but that's in community 2023-07-30 17:18:11 Anyone have a public-facing mpd server I can test xfce4-mpc-plugin against? 2023-07-30 17:18:30 I am having a hell of a time packaging owntone, and I can't seem to get mpd to work with my music on my mounted NAS folder 2023-07-30 17:18:46 Or conversely, want my apk to test on your machine with? 2023-07-30 17:44:11 Saijin_Naib[m]: I don't have mpd running, but perhaps you can share what issues you are having? 2023-07-30 17:58:25 That's the thing, nothing I can tell is wrong. Install package, configure mpd to look at mounted NAS volume, start service and... Nothing  2023-07-30 17:58:51 No error, but it doesn't start  2023-07-30 17:59:19 NAS is mounted to /media/samba/NAS/username 2023-07-30 17:59:32 root owns down to username, then username owns that and below 2023-07-30 17:59:45 Does mpd user need to be in another group  2023-07-30 17:59:45 ?  2023-07-30 18:00:51 I wouldn't know myself 2023-07-30 18:01:30 Yeah, it's been frustrating, hence trying to package owntone which is supposed to be much easier to manage, as well as supporting tons of service types for output  2023-07-30 18:02:23 Well, xfce4-mpc-plugin compiled and runs fine, no console errors... I guess this is why the testing repo exists, right? 🤣 2023-07-30 18:02:40 Someone can yell at me if it actually doesn't work  2023-07-30 20:21:26 Hi all :-) 2023-07-30 20:22:56 o/ 2023-07-31 01:18:30 How can I enable plymouth on Alpine? I installed and set the Chicago95 theme as default 2023-07-31 01:21:40 there is no plymouth support afaik 2023-07-31 01:22:27 well, to be more specific, the standard initramfs (from mkinitfs), does not support 2023-07-31 01:22:58 if you were to use dracut or possibly booster then they might support it, not sure 2023-07-31 01:23:14 dracut supports it 2023-07-31 01:23:24 more importantly does dracut support alpine? 2023-07-31 01:23:59 dracut is packages for Alpine, not sure how well it works however 2023-07-31 01:24:04 I've been meaning to test it 2023-07-31 01:24:36 but the standard/official initramfs for Alpine is provided by mkinitfs 2023-07-31 01:24:41 guess its time for both of us to test that then 2023-07-31 01:24:58 yeah I saw that default one, I tried to add plymouth as a feature but no luck 2023-07-31 01:25:02 ACTION prefers to just used deferred framebuffer takeover and set an image in the BIOS/UEFI. 2023-07-31 01:25:12 I started looking at both booster and dracut a while ago but didn't have much time to test them 2023-07-31 01:27:41 Yeah the other thing I want to get working is lightdm-webkit-greeter 2023-07-31 01:27:56 Because Chicago95 has a really awesome theme for it 2023-07-31 01:28:18 I got it to build and install although I havent been able to enable it so far 2023-07-31 02:41:08 What would be needed for Plymouth to work in mkinitfs?  2023-07-31 02:41:23 It's just shiny, but it is a nice nice 2023-07-31 02:42:36 I think it'd be fairly intrusive, you'd need to modify the various tasks the initramfs does to tell plymouth about their progress and such 2023-07-31 02:43:31 what is plymouth? 2023-07-31 02:47:15 it's a boot progress display, kinda 2023-07-31 02:48:20 It's like deferred framebuffer takeover... except with the potential for animation. Far too much effort for little gain. 2023-07-31 02:48:51 oh. not my cup of tea, but oh well :) 2023-07-31 02:49:06 Am I conflating things, but wasn't it also supposed to improve or speed up the boot time? 2023-07-31 02:49:18 I thought I remember Ubuntu/Canonical talking that up when they added it by default 2023-07-31 02:49:30 Or that could have been when they changed something in their init/supervisor system 2023-07-31 02:52:08 It's supposed to provide a flicker-free graphical boot process. 2023-07-31 02:53:12 I don't see how it would speed up boot, though 2023-07-31 02:53:42 it shouldn't make much difference. 2023-07-31 02:54:43 Yeah, I think it was around the same time they changed init or supervisor 2023-07-31 02:54:48 Like I said, I might be conflating the two 2023-07-31 02:54:54 You print less characters to slow terminals! 2023-07-31 02:54:55 This was what, like 15 years ago, haha 2023-07-31 02:55:04 LOL quinq 2023-07-31 02:55:44 I've often seen claims that init systems that compete with openrc are faster... but... Most of the time they do a lot more shit and end up losing their supposed speed advantage. 2023-07-31 02:56:39 rc_parallel (disabled by default because it doesn't always work!) is typically faster than them regardless. YMMV. 2023-07-31 02:57:00 I've been afraid to enable it due to the warning in the conf, haha 2023-07-31 02:57:40 I've had it enabled for years; and the only side affect I have seen... on a few of my boxes the system tries to set ipv6 related sysctls before the ipv6 module is loaded. LOL 2023-07-31 02:58:04 but like I said, ymmv :) 2023-07-31 03:01:51 hmm 2023-07-31 03:01:58 perhaps I'll try 2023-07-31 03:35:09 @Saijin_Naib[m], Its mainly to try and get the Chicago95 theme working to its fullest although its probably best to compromise and not use it for the boot process 2023-07-31 03:35:27 Yeah, fair enough 2023-07-31 03:35:39 Are you trying to package Chicago95? That thing is a bear, no? 2023-07-31 03:46:34 Ah, poop. Forgot to put contrib/maintainer in my xfce4-mpc-plugin MR 2023-07-31 03:46:37 Uuugh 2023-07-31 03:50:16 Does anyone have handy where the accepted license types are for APKBUILD? 2023-07-31 04:45:47 Saijin_Naib[m]: To my understanding its SPDX licenses: https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package#license 2023-07-31 05:08:00 Ah, awesome, thanks  2023-07-31 05:08:11 Into the folder that goes 2023-07-31 05:43:22 re me as maintainer for firefox, I use chromium mostly nowadays 2023-07-31 06:10:20 ncopa: when you get a chance could you fix the issue template on the wiki, that way there's a chance it could be used again? 2023-07-31 06:10:23 see https://wiki.alpinelinux.org/wiki/Template_talk:Issue 2023-07-31 06:11:04 ncopa: ok, then we need to find someone else. I can try to keep it up to date, but not sure if I'm up to more challenging issues 2023-07-31 06:12:47 apk does not store pkg files of the installed packages, right? 2023-07-31 06:13:14 so: download, install, and unlink 2023-07-31 06:13:23 zcrayfish: thanks for the ping! is this ok? https://wiki.alpinelinux.org/w/index.php?title=Template%3AIssue&diff=24015&oldid=13473 2023-07-31 06:13:32 aron: if you run setup-apkcache, it does 2023-07-31 06:13:48 I didn't 2023-07-31 06:14:04 ncopa: Looks like you fixed it, thanks! =) 2023-07-31 06:14:09 Then, no, by default it doesn't 2023-07-31 06:14:34 bummer, I have to rebuild my package, but sadly it takes a few days 2023-07-31 06:15:42 I messed up my script, and the pkg file was deleted :( 2023-07-31 06:18:02 aron: aw! 2023-07-31 07:11:39 omg, I found a backup from that pkg! 2023-07-31 07:11:45 pfew 2023-07-31 07:11:47 for the win! 2023-07-31 07:12:18 at least half of it 2023-07-31 07:12:26 (it's a split pkg9 2023-07-31 08:30:37 Linux-lts can run with rockchip aarch64 (rk3566)? I have been not able to boot with uboot or efi 2023-07-31 08:35:45 If I'm running into a bug that probably only occurs in an Alpine docker container, can I still raise that on GitLab or do I need to raise it against alpine-docker? 2023-07-31 09:02:05 You can report it on gitlab 2023-07-31 10:49:37 ikke: d'oh, sorry for the silly issue, thanks for the extra info. 2023-07-31 10:50:13 voltagex: no worry 2023-07-31 11:41:44 https://www.phoronix.com/news/Alpine-Linux-Prolific-Packager 2023-07-31 11:41:57 I'm sorry to hear psykose is leaving Alpine 2023-07-31 11:48:49 me too 2023-07-31 11:52:59 what drama did I miss this time? 2023-07-31 11:56:50 read up on phoronix article, too bad... hopefully she will return in future and hopefully other maintainers will pick up the now unmaintained packages 2023-07-31 11:57:08 alpine is too good to go under 2023-07-31 11:57:52 it won't. 2023-07-31 12:13:32 o/ 2023-07-31 12:13:48 she never gave any reason why she left 2023-07-31 12:14:02 it just happened all of a sudden 2023-07-31 12:14:12 its unknown what triggered it 2023-07-31 12:14:41 I just came to say goodbye and thanks, just read the news. I've been using Alpine for like a week only and she already helped me with issues. Sorry to come too late. 2023-07-31 12:15:06 > She hasn't decided what to do next but likely "catch up on a year of sleep." 2023-07-31 12:15:49 Yeah, sounds pretty much like burnout to me. It's rather common in FOSS maintainers. But I think it's more respectful to avoid speculation about why people leave. 2023-07-31 12:16:00 sleep is nice. 2023-07-31 12:16:07 Sleep is VERY nice. 2023-07-31 12:17:01 never underestimate a good sleep 2023-07-31 12:17:26 I'm sorry to hear that psykose left 2023-07-31 12:18:19 While I am saddened to see her leave Alpine Linux, I can't wait to see what project she does next 2023-07-31 13:15:21 well, we carry on 2023-07-31 13:16:54 I'll miss her 2023-07-31 13:29:44 Same. They gave me an inordinate amount of help with learning packaging  2023-07-31 13:45:38 i may have an idea 2023-07-31 14:09:08 I find quite inappropriate to write an article on Phoronix about they leaving. 2023-07-31 14:09:15 ACTION thanks psykose for her service 2023-07-31 14:10:23 which now Hacker News has picked up on 2023-07-31 14:10:59 speaking from my own experience working on distributions a long time ago, life can make decisions for you suddenly 2023-07-31 14:11:33 took me a while to feel like i didn't let a bunch of people down 2023-07-31 14:14:03 raspbeguy: is it supposed to be a secret? 2023-07-31 14:15:42 zcrayfish: there's a lot of ground between "secret" and "tell the world" 2023-07-31 14:18:11 ACTION notes that the git commits deleting maintainership+contributorship, wiki logs denoting removal from admin group, and the IRC logs showing her leaving this channel are all public. 2023-07-31 14:20:35 Not a secret. But I think it's quite indecent 2023-07-31 14:23:48 I'm curious if you read it. 2023-07-31 14:25:58 it's probably best to not speculate 2023-07-31 14:26:05 ACTION clones aports 2023-07-31 14:28:11 lol. 2023-07-31 15:52:26 getting posted on HN gives me second hand stress just thinking about it 2023-07-31 16:10:21 "what i wonder is, how can she do..." <- being smart and working with other distro maintainers 2023-07-31 16:33:10 well, team effort :) 2023-07-31 16:35:33 if each person here took on one of psykose' old packages we would be in very good shape! 2023-07-31 16:54:09 elly: indeed :) 2023-07-31 16:54:26 How big is the core Alpine team? 2023-07-31 17:12:21 Forza: Not that large, but Alpine is mostly a community effort 2023-07-31 17:17:14 Let's hope for a continued good future for Alpine :) 2023-07-31 17:18:06 Forza: you're welcome to take part in that future 2023-07-31 17:20:59 Ermine: thanks. I was thinking to see if I could take over packaging for scsi-tgt that psykose helped create recently 2023-07-31 17:21:14 I bet you can :) 2023-07-31 17:21:42 Is there a requirement to use my full name and email and stuff? 2023-07-31 17:21:52 no, and psykose herself never did 2023-07-31 17:22:08 you should use an email address that works, because people will contact you about your packages :) 2023-07-31 17:22:40 yeah, we don't require real names 2023-07-31 17:22:43 Sure. That'd work. I'll think about it. What would be required to get going/signed up? 2023-07-31 17:22:50 there's no signup process, afaik 2023-07-31 17:22:54 Forza: not a lot, indeed 2023-07-31 17:23:15 Real name might be OK, just don't want it publicly posted 2023-07-31 17:23:17 just create an MR, put your name / pseudonum in the maintainer field, bump pkgrel 2023-07-31 17:23:26 Forza: it would be part of pkgs.a.o 2023-07-31 17:23:49 MR = Merge Request? Same as PR on Github? 2023-07-31 17:23:54 yues 2023-07-31 17:24:10 so you'd fork aports in gitlab 2023-07-31 17:24:12 create a branch 2023-07-31 17:24:20 commit the change, push it, and then create a merge request 2023-07-31 17:24:29 Forza: all of this stuff is just stored in the aports git repo so it is 100% public 2023-07-31 17:24:35 but, you can use a pseudonym 2023-07-31 17:25:11 Do you have some special formats and such? Like adding specific signed-by lines, comments, etc? 2023-07-31 17:25:43 Forza: its really simple, a look at git log will show you most of it 2023-07-31 17:25:44 in the aports repo there is a short guide document about how your commits should look 2023-07-31 17:26:35 https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/COMMITSTYLE.md 2023-07-31 17:26:41 beat me to it :) 2023-07-31 17:26:58 Forza: we also appreciate if maintainer responds to issues and similar for the package, but usually those are few (depending on which package it is) 2023-07-31 17:28:29 I think it's more about taking responsibility than do the actual work. keep an eye on it, fix issues, update it, or ask for help with that if needed 2023-07-31 17:30:05 ask for help can be done in #alpine-devel 2023-07-31 17:30:24 Seems doable. Probably not very many issues, since at the moment I don't think there are any other users of scsi-tgt :D 2023-07-31 17:30:52 yeah, it seems like relatively easy package 2023-07-31 17:53:51 zcrayfish Mousepad with plug-ins is out 😎 2023-07-31 17:54:12 I created an account on the gitlab.alpinelinux.org - first steps 2023-07-31 17:54:29 "Forza: we also appreciate if..." <- If I maintain something, am I automatically tagged for issues on that package?  2023-07-31 17:54:59 Saijin_Naib[m]: I assign maintainers to issues, yes 2023-07-31 17:55:05 so not automatically, but semi-automatically 2023-07-31 17:55:17 ikke-matically 2023-07-31 17:55:20 Good enough for me  2023-07-31 17:55:58 ;) 2023-07-31 17:56:19 I think longer term I want to graduate my stuff from testing, but I will need lots of help and handholding, especially around back ports, finding vulns, etc 2023-07-31 17:56:26 sure 2023-07-31 17:56:36 feel free to hop by on #alpine-devel 2023-07-31 17:56:50 oh yeah - how *does* stuff graduate from testing? 2023-07-31 17:56:53 But that's for devs 😬 2023-07-31 17:57:04 Saijin_Naib[m]: you are one when you maintain packages 2023-07-31 17:57:13 I think we ask, and then you need to do 6mo of security updates 2023-07-31 17:57:30 ikke: Where's my sticker 🤔 2023-07-31 17:57:30 elly: it's not a formal proces 2023-07-31 17:57:32 I ought to read https://wiki.alpinelinux.org/wiki/Alpine_Linux:Contribute as well 2023-07-31 17:57:45 elly: testing was introduced mostly to give people a chance to test their packages 2023-07-31 17:58:17 Some people like to have it in testing for some time, others create MRs after they verified it works for them 2023-07-31 17:59:07 I guess it depends on the size of the project/app too. A smallish thing is easier to assess as working compared to a monster app 2023-07-31 17:59:09 Saijin_Naib[m]: sweet, will check it out shortly (currently feeding zcrayfish 2.0) 2023-07-31 17:59:33 That's very true... Also harder to get different folks hands on to test 2023-07-31 18:00:07 zcrayfish: My pet sperm is sleeping, but soon I'll need to figure out what it wants to throw off the table for lunch 2023-07-31 18:01:07 ikke: ah, okay 2023-07-31 18:03:37 elly: others put package in testing and never come back 2023-07-31 18:04:24 Hi all.. new alpine user here. I have a thinkpad x1c6 and run i3. It worked fine for years on ubuntu and debian but with alpine, when i3 starts the keyboard/mouse are not recognized. I have to ssh in and kill i3 to get back to a shell. So far my searches haven't found a solution. Any ideas? 2023-07-31 18:06:06 pjs: i don't know. But maybe there is a specific driver for the keyboard/mouse. I've read that there are some i2c or other special input devices needed on some laptops 2023-07-31 18:06:11 pjs: I have one: you may not have the right X input module 2023-07-31 18:06:14 I ran into this problem before 2023-07-31 18:06:33 You could check lsmod and lspci output to see if there's a difference between alpine and Ubuntu 2023-07-31 18:07:10 pjs: how did you install i3? did you run setup-xorg-base? 2023-07-31 18:07:25 pjs: you might try using synaptics instead of evdev for your mouse, and see if that helps 2023-07-31 18:07:55 did install synaptics package but let me try that 2023-07-31 18:08:08 zcrayfish: yes, I ran setup-xorg-base and also installed i3 2023-07-31 18:08:24 pjs: even if you install it you then have to tell X to use it 2023-07-31 18:08:37 make sure you are in the input group 2023-07-31 18:08:55 oh! that's a good call, you may not have access to the raw devices 2023-07-31 18:10:29 I am in the input group 2023-07-31 18:11:43 you may want to check /var/log/Xorg.0.log for relevant errors (and then warnings) too 2023-07-31 18:11:45 pjs: you may need to configure X explicitly then, like: Section "InputClass" Identifier "mouseN" Driver "synaptics" EndSection 2023-07-31 18:12:06 if it's any help, I wrote about my own X + i3 config here: https://elly.town/d/blog/2022-10-20-alpine-framework.txt although this was on a Framework, not a Thinkpad 2023-07-31 18:12:58 elly: thanks! let me try this 2023-07-31 18:13:19 zcrayfish: yea, I'm going through the logs... not super useful but some stuff I'm trying to follow 2023-07-31 18:29:55 ok, I got it working.. not toally sure what did it though haha 2023-07-31 18:30:14 that might be good to know for the future, in case it breaks later 2023-07-31 18:40:26 also if you do figure it out please write about it and publish it :) 2023-07-31 18:40:46 elly: I definitely will. I'm troubleshooting it now to see which was the fix 2023-07-31 18:41:16 yay 2023-07-31 18:58:40 I've got gnome-keyring auto unlock working with sway on arch linux, but with a similar setup, it doesn't work on alpine. I'm using emptty. Keyring unlocks when logging into xfce, but not on sway. Is there anything I'm missing? 2023-07-31 18:59:50 I added the 2 lines described here https://wiki.archlinux.org/title/GNOME/Keyring into a few pam.d files to no avail. 2023-07-31 21:48:58 "Hi all.. new alpine user here. I..." <- Is this change recent? I run Alpine with dwm on my ppc64le server and it was ok 2023-07-31 21:49:36 it's been about a month or two that I didn't boot Alpine 2023-07-31 22:55:17 rob[m]: yea, just installed today heh 2023-07-31 22:55:29 but I've got most of the issues worked out.