2019-06-01 02:57:24 so, I see Nix package manager is on the Wishlist on the wiki. What does that mean exactly? 2019-06-01 02:57:43 Someone wants the nix package manager available on Alpine 2019-06-01 02:58:22 integrated or what? 2019-06-01 02:58:35 because it's already in the aports? 2019-06-01 02:59:43 Just available i guess 2019-06-01 02:59:56 Fair enough, thanks. 2019-06-01 03:08:53 The gitian distribution method seems quite cool 2019-06-01 03:59:50 mps: my irc client lived in my vps's tmux session, so its 24 hours here :D 2019-06-01 04:00:21 and i will login every work day unless the great firewall is high enought to prevent me from the outside world 2019-06-01 07:05:55 Hello... The timestamps in my /var/log/messages are one hour off from localtime (it's 09:xx AM here, new logs show 08:xx). I setup the time zone with 'setup-timezone' (Europe/Ljubljana), which created the /etc/zoneinfo/Europe/Ljubljana file and I also manually edited /etc/TZ, so that it contains that string. 2019-06-01 07:06:15 I also have the -Z option for syslogd... 2019-06-01 07:07:16 Because of DST our local time is currently 2 hours off UTC... 2019-06-01 07:07:47 <_ikke_> What it the time of the server? 2019-06-01 07:08:24 'date' returns the correct local time: Sat Jun 1 09:08:02 CEST 2019 2019-06-01 07:09:22 <_ikke_> No clue myself then 2019-06-01 07:33:31 It's even weirder. :) If I run 'date | logger', it's one hour early. 2019-06-01 07:33:51 Jun 1 08:31:36 elgo-fw daemon.info dhcpd: DHCPACK on 172.16.10.56 to 84:38:38:da:b8:ac (android-257fd420d150235d) via eth0 2019-06-01 07:33:51 Jun 1 10:33:17 elgo-fw user.notice root: Sat Jun 1 09:33:17 CEST 2019 2019-06-01 07:33:51 2019-06-01 07:34:55 And now my kvirc just died on me. :D 2019-06-01 09:53:48 is there a way to list unused packages? 2019-06-01 10:01:40 cim209: define unused 2019-06-01 10:01:56 no revdeps and not in world? 2019-06-01 12:21:46 dngray: made the wiki page BTW: https://wiki.alpinelinux.org/wiki/Setting_up_ZFS_with_native_encryption 2019-06-01 12:25:05 yunfan: also my irssi is under tmux on server, so I'm 'always online' even when I am not :) 2019-06-01 12:25:38 /away 2019-06-01 12:26:10 no, don't like to bother other with my status 2019-06-01 12:28:31 yunfan: Off topic, yes, but walls are not bad things. I would like to have them more to block a lot of trash which come unsolicited :) 2019-06-01 13:02:30 mps: bother how? does irssi announce away to channels? 2019-06-01 13:04:20 tcely: yes, but I 'filtered' out all 'state' msgs 2019-06-01 13:05:16 it is really annoying to see a lot of 'join/leave/away' messages 2019-06-01 13:05:57 IRC sends these, and it is on the client will it show them or not 2019-06-01 13:07:08 join/part/quit messages are filtered in my client too. I just have never seen an away message the client didn't generate. 2019-06-01 13:07:41 try some 'raw' clients :) 2019-06-01 13:09:25 that sounds painful ;-) 2019-06-01 13:10:07 here is first line in my irssi ignore list, i.e. ignore *: JOINS PARTS QUITS NICKS 2019-06-01 13:13:00 i was aiming at the possibility of your irc client responding via PM or something else when you're mentioned/PMed 2019-06-01 13:15:53 understand. but I don't like too much automation in irc clients, fear of possible security issues. 2019-06-01 13:17:28 away functionality is built into most of the common clients 2019-06-01 13:18:45 yes, it is. but don't mind to use it, I'm always online even when I sleep 2019-06-01 13:18:55 but you're not present 2019-06-01 13:19:15 I'm not even sure it's always the client responding. Some servers may handle that bit. 2019-06-01 13:19:50 i haven't come across badly mannered irc servers in a while now, i guess it's because i haven't been exploring much 2019-06-01 13:20:28 yes, but you can send me message even when I sleep and you will not feel bad about that :) because you think I'm behind keyboard 2019-06-01 13:21:11 https://tools.ietf.org/html/rfc1459#section-5.1 2019-06-01 13:24:36 actually I prefer mail communication, and use irc only on some support services, and alpine because looks like Alpinists don't use mail much 2019-06-01 20:08:46 Hi 2019-06-01 20:08:57 Having an issue with ufw on edge testing: ModuleNotFoundError: No module named 'ufw' 2019-06-01 20:09:02 This happens when running ufw 2019-06-01 20:16:43 Maybe some py2/py3 confusion 2019-06-01 20:22:40 <_ikke_> Works for me in a new edge docker container 2019-06-01 20:22:50 <_ikke_> / # ufw 2019-06-01 20:22:52 <_ikke_> ERROR: not enough args 2019-06-01 20:27:09 Cogitri: that's my assumption too, perhaps they changed the `python` symlink or something 2019-06-02 10:16:29 mps: well, that effects were completely made by decision. what do you think those trash come from? :D 2019-06-02 10:38:39 yunfan: most trash come from public media, imo 2019-06-02 10:39:36 I started to build linux-vanilla kernel for sunxi devices, enabling different option one by one 2019-06-02 10:40:39 looked also to arch linux alarm kernel config, it will be useful to compare with alpine config 2019-06-02 10:43:21 don't know did you noticed that alpine u-boot need DOS partition table type, doesn't work with GPT 2019-06-02 10:43:56 I think this is the limit of stock u-boot 2019-06-02 11:06:46 mps: why, but the uboot support ext2/3/4 2019-06-02 11:07:04 aha, you mean gpt, sorry 2019-06-02 11:08:46 looks like u-boot image is too big for GPT 2019-06-02 11:10:49 well, i dont know the details on this, but, is that the reason ? 2019-06-02 11:11:13 the problem is the kernel had boot in and hang there, so uboot is ok i think 2019-06-02 11:11:54 u-boot is ok, but have to be flashed on card with DOS partition table 2019-06-02 11:13:51 arch linux kernel boot fine, as kernels I built from upstream 2019-06-02 11:14:26 only have to find which parameters must be enabled or changed for alpine kernel to boot 2019-06-02 11:31:28 so how many options were different between arch and alpine? 2019-06-02 11:31:40 hundreds or thourands? or more? 2019-06-02 14:10:21 exit 2019-06-02 14:10:23 oops 2019-06-02 15:23:14 hi. I'm having a strange issue at boot. Alpine OS is on a XFS partition on LVM. Initramfs loads but then it fails with the error "mounting /dev/mapper/myVG-myLV /sysroot: no such file or directory" and i have an initram shell. In that shell if i just type: mount -t xfs /dev/mapper/myVG-myLV /sysroot; exit; it works and alpine boots. I guess i missed something when generating the initramfs. Any idea ? 2019-06-02 15:24:26 You've enabled `xfs lvm` in mkinitfs.conf, yes? 2019-06-02 15:25:22 yes 2019-06-02 15:26:00 features="ata base cdrom ext4 xfs scsi keymap kms mmc lvm raid usb virtio" 2019-06-02 15:28:24 does order matter ? 2019-06-02 15:35:06 No 2019-06-02 15:35:33 Hm, sorry, my knowledge about mkinitfs is pretty limited :/ 2019-06-02 15:41:17 <_ikke_> Cogitri: Did you ask ncopa to take over gvfs? 2019-06-02 15:41:52 i'm trying to add 'set -x' at the beginning of the initramfs-init script to see what's going on. 2019-06-02 15:42:45 _ikke_: He said that he was glad for me to take over anything GNOME when I started working on GNOME, so I haven't asked him for gvfs specifically because he seems pretty busy right now. If you want to I can ask him for all the pkgs again 2019-06-02 15:43:03 <_ikke_> Cogitri: Nah, I think it's good 2019-06-02 15:43:31 ok, got it.. the script does mount but it misses the '-t xfs' option 2019-06-02 15:44:01 fridim: I'm pretty sure that mount should do that automagically though 2019-06-02 15:44:13 Actually, not sure with busybox mount 2019-06-02 15:44:35 yep, it's busybox 2019-06-02 15:45:01 maybe i just add the other mount binary into my initramfs. Or i patch that initramfs-init 2019-06-02 15:47:33 a bit tedious if you just want to boot alpine from XFS, no ? 2019-06-02 15:49:10 Yup, it should "just work". I guess it'd be a good idea to open a bug report about this 2019-06-02 15:50:41 sure. 2019-06-02 15:50:45 trying something else first 2019-06-02 15:54:20 i added an entry in /usr/share/mkinitfs/fstab that matches, it doesn't help 2019-06-02 16:07:23 after reading that script, i added 'rootfstype=xfs' to the 'linux' line and it works now ;) 2019-06-02 16:07:43 (that script = initramfs-init) 2019-06-02 16:08:03 <_ikke_> fridim: nice 2019-06-02 18:48:16 Does anyone have QXL working on Alpine Linux in a Qemu/KVM VM? 2019-06-02 18:48:38 My VM display just bugs out with X 2019-06-02 18:48:39 ACTION uploaded an image: Screenshot_20190602_204818.png (20KB) < https://matrix.org/_matrix/media/v1/download/fam-ribbers.com/iMIUccNqhJUiBTWwDAitWhMt > 2019-06-02 18:58:39 PureTryOut[m]: it worked for me last time (about two months ago) when tried it 2019-06-02 19:01:33 How did you do it? What did you install in the VM? 2019-06-02 19:01:45 I'm trying Virtio as well but that doesn't work either 2019-06-02 19:01:55 `dmesg | grep drm` doesn't show anything 2019-06-02 19:02:08 uhm, not sure, I deleted VM. It was for testing 2019-06-02 19:02:29 I'm planning to use it for testing as well but no luck so far lol 2019-06-02 19:03:05 I only remember (have written) qemu parameter '-vga qxl -serial stdio -soundhw hda' 2019-06-02 19:04:59 two years ago it also worked when I used Alpine in qemu to test it before full switch 2019-06-02 19:05:45 I played with xfce and awesome about two months, and firefox, audio etc. all worked quite fine 2019-06-02 19:05:58 with qxl, I mean to say 2019-06-02 19:06:00 I'm using it through libvirt, but Qxl is definitely enabled and picked up by Xorg according to the logs. It just... does this weird shit 2019-06-02 19:07:21 Even modesetting doesn't work, it doesn't detect screens (when using the vga adapter) 2019-06-02 19:09:27 X just seems to hang with Qxl. `xrandr --listmonitors` never completes 2019-06-02 19:10:21 sdl or gtk works? 2019-06-02 19:12:11 I'm not using Qemu through regular sdl or gtk, but through libvirt which I control with virt-manager. I it uses Spice to control and show the display 2019-06-02 19:13:02 PureTryOut[m]: few months ago I installed debian under qemu on alpine for my daughter and I enabled there qxl also 2019-06-02 19:13:18 But Debian is not Alpnie 2019-06-02 19:13:19 *Alpine 2019-06-02 19:13:45 true, but maybe could help to see what is needed to be installed 2019-06-02 19:14:20 The required stuff is installed already though 2019-06-02 19:14:27 I've done this with other distros before 2019-06-02 19:14:39 Really the only thing the client needs is `xf86-video-qxl`, which _is_ installed 2019-06-02 19:14:46 Xorg runs, but completely hangs with cpu spike to 100% 2019-06-02 19:15:13 maybe it need udev and dbus, not sure, I forgot 2019-06-02 19:15:27 but know that it worked well 2019-06-02 19:15:43 Both are already installed 2019-06-02 19:16:02 also, `setup-xorg` doesn't work as `pciutils` conflicts with `hwids-pci` when installing 2019-06-02 19:16:09 ohm, then I don't have any idea more 2019-06-02 19:18:37 I guess I'll install Alpine on a usb on my laptop instead to test my Plasma stuff... 2019-06-02 19:25:17 where the scripts/mkimages.sh should be run to build arm tarball? on installed system or from aports tree? 2019-06-02 19:52:58 Ugh and now wifi fails to connect with wpa_supplicant... "p2p-dev-wlan0: Failed to initialize driver interface" 2019-06-02 19:57:10 Ok that was just networkmanager being annoying and running at the same time. Then it couldn't get an ip address with dhcpcd, but rebooting fixed that 2019-06-02 21:05:11 yolloo.. anyone alive? 2019-06-02 21:05:11 or dead room? 2019-06-02 21:05:27 no, all life ended 5 minutes ago 2019-06-02 21:06:08 i should make a bot that looks for common variations of that and links them to the "don't ask to ask questions" post 2019-06-02 22:03:54 I was wondering about FRR package. It would be nice to have it available for networking labs. 2019-06-02 22:03:54 Is there any way to package it to 3.9? 2019-06-02 22:04:21 <_ikke_> elico: not very likely 2019-06-02 22:05:25 _ikke_: is there any specific reason that you might know? 2019-06-02 22:06:01 <_ikke_> It's not common for new packages to be added to an existing release 2019-06-02 22:07:08 The package is in the edge tree but is no up to date.. 2019-06-02 22:07:08 Would it be possible to pack it manually for 3.9? can you redirect me towards a specific guide? 2019-06-02 22:07:22 <_ikke_> elico: yes, that should be possible 2019-06-02 22:07:23 looks like frr APKBUILD needs some cleanup 2019-06-02 22:07:40 mps: called ? 2019-06-02 22:07:52 <_ikke_> elico: https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package 2019-06-02 22:07:52 north1: hehe :) 2019-06-02 22:09:27 source field needs change, and prepare() should be added for bootstrap.sh 2019-06-02 22:10:08 _ikke_: pretty long for this hour at night. 2019-06-02 22:10:08 I might try to look at it later. How would I be able to contact the current maintainer "Arthur Jones" from https://pkgs.alpinelinux.org/package/edge/testing/x86_64/frr 2019-06-02 22:10:08 ? 2019-06-02 22:10:30 <_ikke_> elico: their e-mail address is in the APKBUILD 2019-06-02 22:10:39 _ikke_: thanks! 2019-06-02 22:12:03 (building and testing Alpine arm kernels for ARM take a lot of time) 2019-06-02 22:12:30 meh, arm for ARM :) 2019-06-02 22:14:23 ncopa: here? 2019-06-02 22:14:41 <_ikke_> http://tpaste.us/KgvP 2019-06-02 22:15:00 <_ikke_> elico: he's mostly active during weekdays 2019-06-02 22:16:21 _ikke_: ok, thanks. I will try to contact him later. Squid-Cache 4.7 was published a month ago it's work upgrade the edge fro 4.6 to 4.7. 2019-06-02 22:18:44 elico: I also noticed that, but didn't posted patch thinking that that will do someone with access to main in aports 2019-06-02 22:19:00 and not only that one 2019-06-02 22:19:56 clamav is outdated but new version doesn't build on armv7 2019-06-02 22:22:20 mps: Are you active here? It sounds like a nice room. 2019-06-02 22:25:06 elico: when I have time, yes. channel is more active at workdays 2019-06-02 22:25:21 workdays ie work hours? 2019-06-02 22:25:48 workdays mostly 2019-06-02 22:26:27 over weekend it is somewhat quiet 2019-06-02 22:26:37 which are? I'm working from sunday till friday including.. 2019-06-02 22:27:02 monday to friday, mostly 2019-06-02 22:27:22 and European work time 2019-06-02 22:27:28 mps: thanks! 2019-06-02 22:27:51 most developers are in Europe, but not all 2019-06-02 22:28:02 so, it depends at the end 2019-06-02 22:28:04 that makes things a bit more clear. 2019-06-03 02:54:41 Is there a linux-stable kernel vs linux-vanilla (lts) that's pre-built with alpine? 2019-06-03 02:55:42 i.e. 5.1.6 at the moment? 2019-06-03 02:56:17 if so, no, not in official repos 2019-06-03 02:57:12 was there some discussion / reasoning behind sticking with longterm and not stable releases? 2019-06-03 02:57:42 @danieli looking for 5.1.6 or 5.0.20 2019-06-03 02:58:13 I'm trying to use iwd, and it depends on features from 4.20+ kernels, and using linux-vanilla won't cut it 2019-06-03 03:01:03 I'm hoping to locate where discussion on alpine's linux kernel choices happened so I can follow thought process when linux-hardened was dropped and linux-vanilla standardized on longterm release at expense of stable. 2019-06-03 03:08:29 superduck: LTS kernel is default linux-vanilla kernel 2019-06-03 03:08:55 @mps I know that, but I'd like to know why 2019-06-03 03:10:31 well, to have LTS kernel. Alpine stable releases maintained some years and LTS is best option for that 2019-06-03 03:11:23 why not linux-stable and linux-lts? Hence I'm looking for where that discussion occurred, to understand the reasoning behind not splitting packages and going with a single linux-vanilla packaging based on lts. 2019-06-03 03:13:06 hard to tell, somewhere in irc logs 2019-06-03 03:14:24 probably ML 2019-06-03 03:15:08 I would also like to have newer kernels, because of iwd 2019-06-03 03:15:39 if you don't need eap, 4.19 works ok for iwd 2019-06-03 03:16:11 iwd says its missing symbols, so I built a new version and it complained I needed a newer kernel 2019-06-03 03:16:24 I'm running on an old 32bit atom netbook 2019-06-03 03:16:30 acer aspire one 2019-06-03 03:16:36 even searched for pkcs8 patch for 4.19 but didn't found it 2019-06-03 03:17:20 this error msg from iwd is harmless, if you do not use eap 2019-06-03 03:17:59 for me it works quite fine on 4.4.179 kernel on aarch64 and armv7 2019-06-03 03:18:08 iwd wouldn't run, it was more than error messages, it was missing library symbols 2019-06-03 03:18:33 then it something else issue and not kernel 2019-06-03 03:19:35 which iwd and ell versions is installed on your box 2019-06-03 03:20:07 I powered it off, I'll have to check tomorrow. 2019-06-03 03:20:40 I pulled ell and iwd from edge so current versions there for x86 2019-06-03 03:21:16 ell should be 0.20 and iwd 0.18 2019-06-03 03:22:17 I don't know how well it works on x86, though. no one tested it on this arch afaik 2019-06-03 03:23:49 superduck: do you use eap mode? 2019-06-03 03:24:37 no psk 2019-06-03 03:25:12 just home wifi, trying to deprecate wpa_supplicant and get connman working in xfce4 so I can go with xorg or wayland for my desktop 2019-06-03 03:25:15 psk works fine from the beginning iwd life in alpine 2019-06-03 03:25:33 Not on x86 from my experience. 2019-06-03 03:25:53 skip connman, it is little behind 2019-06-03 03:26:38 I don't want python or perl on my system, I'd also like to dump dbus, but that was a bad desktop usability decision I can't escape practically 2019-06-03 03:26:42 hmm, please fill a bug report on bugs.a.o with some details, I will look at it. 2019-06-03 03:26:48 so connman gets me away from networkmanager 2019-06-03 03:26:58 @mps thanks, I'll post it tomorrow 2019-06-03 03:27:24 btw, I'm maintainer of ell and iwd in alpine 2019-06-03 03:27:41 sweet 2019-06-03 03:27:48 what's ell? Its unfamiliar to me. 2019-06-03 03:28:06 embeded linux library 2019-06-03 03:28:17 what do you recommend for a gui applet for controlling wifi with iwd? 2019-06-03 03:28:35 I'd go with wicd, except for the python dependency and unsure if it works with iwd. 2019-06-03 03:28:52 only know that the networkmanager works, although I don't use it 2019-06-03 03:28:57 I'm not wanting to manage iwd via cli, so that was why I was going with connman, though I didn't see a connman gui. 2019-06-03 03:29:17 I think systemd is better than networkmanager. 2019-06-03 03:29:39 I tried to package connman-gtk but it is a full of bugs 2019-06-03 03:29:57 as is most software 2019-06-03 03:30:28 I use iwctl cli, or iwctl in scripts 2019-06-03 03:31:04 ok, you've convinced me to pull out the laptop and get you some details 2019-06-03 03:31:37 when I learn to program system dbus maybe I will write text ui for iwd 2019-06-03 03:32:43 dbus is another tool that should die 2019-06-03 03:33:07 good idea, terrible implementation 2019-06-03 03:33:11 long live kdbus :) 2019-06-03 03:33:17 ACTION mumbles something about Poettering 2019-06-03 03:34:10 agree with you both 2019-06-03 03:35:29 I'll extend my griping to anything using XML encoding for the backbone of the protocol. 2019-06-03 03:36:05 I'll give javascript and ini data formats a pass for being reasonably terse and readable 2019-06-03 03:36:13 well json not javascript 2019-06-03 03:38:39 mhm, I did a lot of SOAP programming earlier :( 2019-06-03 03:39:03 I'm befuddled why squashfs isn't included in the default mkinitfs configuration if modloop usage depends on it 2019-06-03 03:39:48 Alpine trying to be minimal 2019-06-03 03:40:08 then why reinvent the purpose of initramfs with modloop in the first place? 2019-06-03 03:40:52 iirc, squashfs have much better compression 2019-06-03 03:41:38 perhaps, but that time probably would've been better spent working on zram rootfs with zstd 2019-06-03 03:43:55 Error relocating /usr/libexec/iwd: l_key_generate_dh_private: symbol not found 2019-06-03 03:44:04 I have about 30 of those symbol not found errors 2019-06-03 03:44:23 if it is that much then problably a library failed to load ? ell ? 2019-06-03 03:44:43 how do I check if ell is installed? 2019-06-03 03:44:59 apk info -e ell 2019-06-03 03:45:01 apk info ell tells me about the package, but how do I show status of package? 2019-06-03 03:45:23 apk version ell 2019-06-03 03:45:33 not installed 2019-06-03 03:45:48 'apk add ell' :) 2019-06-03 03:46:03 i sometimes do `apk info | grep ` 2019-06-03 03:46:18 @mps why isn't that a dependency of iwd to pull in? 2019-06-03 03:46:32 did `apk add ell` and now have 256 symbol not found items 2019-06-03 03:46:40 it should 2019-06-03 03:46:43 well no, just do apk list -I | grep '{ell}' 2019-06-03 03:46:47 do I need to install and run dbus first? 2019-06-03 03:47:30 'apk info -R iwd' will tell deps 2019-06-03 03:47:59 iwd from edge/community didn't pull in ell for me, had to manually install it 2019-06-03 03:48:09 you can try without dbus but I suspect that the connman will work without dbus 2019-06-03 03:48:12 @north1 I like that it's showing licenses in your output 2019-06-03 03:48:21 i'm not running connman 2019-06-03 03:48:27 superduck: apk list -I will show all installed packages 2019-06-03 03:48:33 I'm running /usr/libexec/iwd to get these errors 2019-06-03 03:48:51 superduck: tell me what 'apk version iwd' says 2019-06-03 03:49:13 iwd-0.18-r0 @edge 2019-06-03 03:49:21 ell-0.20-r0 ? 2019-06-03 03:49:28 right 2019-06-03 03:50:35 ldd /usr/libexec/iwd spews the symbol not found errors too 2019-06-03 03:51:08 any other basic packages I might be missing? 2019-06-03 03:51:19 apk fix -r iwd 2019-06-03 03:51:37 nope 2019-06-03 03:52:04 ooh, wait ell wasn't happy, did apk fix and it reinstalled 2019-06-03 03:52:29 yay! I've got iwd working!!! 2019-06-03 03:52:43 iwd didn't pull in ell and ell didn't install cleanly for me 2019-06-03 03:53:01 I pulled ell from edge/main 2019-06-03 03:53:14 yes, it is there 2019-06-03 03:53:45 right, but `apk add iwd` didn't magically depend on ell and pull it in cleanly as a dependency 2019-06-03 03:54:25 iwd is showing as depending on libc.musl, libell.so, and libreadline.so 2019-06-03 03:54:27 hmm, something is broken, I worked wherever I tested it 2019-06-03 03:55:01 'apk info -r ell' 2019-06-03 03:55:15 ell is required by iwd 2019-06-03 03:55:22 so seems sane 2019-06-03 03:55:24 version? 2019-06-03 03:55:39 ell-0.20-r0 iis required by: iwd-0.18-r0 2019-06-03 03:55:48 i.e. 'ell-0.20-r0 is required by:' 2019-06-03 03:56:04 ^^ 2019-06-03 03:56:05 that's as it should be 2019-06-03 03:56:10 agreed 2019-06-03 03:56:15 but my environment didn't do that 2019-06-03 03:56:25 maybe broken repository setup 2019-06-03 03:56:40 anyone have a decent newbie tutorial for iwd? 2019-06-03 03:57:33 by coincidence I wrote something but not finished and polished it yet :) 2019-06-03 03:57:47 got it online? 2019-06-03 03:58:00 not yet 2019-06-03 03:58:27 also, I could use a link to how to sign a local APKINDEX.tar.gz so that it makes the initial /media/usb/apks repo on my system happy 2019-06-03 03:58:48 I need to add iwd instead of wpa_supplicant 2019-06-03 03:59:49 @mps when will you update connman so it can depend on a virtual package instead of wpa_supplicant? 2019-06-03 04:01:15 I don't have plan to work with connman for now, it is 'undeveloped' now 2019-06-03 04:01:45 btw, connman is made by same people who made iwd 2019-06-03 04:01:56 @mps intel's a big company 2019-06-03 04:02:26 well, I wrote people and not company ;) 2019-06-03 04:03:02 The laws of the USA equate companies to people...so ambiguous statement? 2019-06-03 04:04:51 no, in this case the same people 2019-06-03 04:05:20 here is short intro for iwd and iwctl http://arvanta.net/mps/iwd-howto.md 2019-06-03 04:06:41 it is not updated to current repo, forget @testing there 2019-06-03 04:06:58 was reviewing and was going to ask 2019-06-03 04:08:07 does iwd need wireless-tools? 2019-06-03 04:08:58 no 2019-06-03 04:09:58 @mps mlan0 or wlan0? 2019-06-03 04:10:23 your wifi device 2019-06-03 04:10:35 doesn't that default to wlan0 for most devices? 2019-06-03 04:10:44 in my case it is mlan0 on box where I wrote this 2019-06-03 04:10:57 can iwctl set wlan0 interface to up so that it can scan? 2019-06-03 04:11:17 yes, read to the end 2019-06-03 04:11:44 iwctl station mlan0 scan 2019-06-03 04:12:34 right, but the interface is down for me so it won't scan 2019-06-03 04:12:52 uhm, I didn't slept this night, going to make coffee 2019-06-03 04:13:27 adapter phy0 set-property Powered on 2019-06-03 04:13:50 btw, this document is very useful, just nit-picking for my simple `ip link set wlan0 up && iwctl station ap ESSID pw` type of usage 2019-06-03 04:14:23 I wrote it as personal notes, not a real guide 2019-06-03 04:14:52 I don't know english well, especially to write something serious 2019-06-03 04:14:53 those are the best guides 2019-06-03 04:15:04 it doesn't show 2019-06-03 04:16:16 I'm self taught in english and never have been in english speaking country 2019-06-03 04:16:39 You're english is impressive. 2019-06-03 04:16:56 scanning doesn't seem to work for me with iwd 2019-06-03 04:17:05 meh, thanks for nice words :D 2019-06-03 04:17:10 though `station wlan0 connect ESSID` worked 2019-06-03 04:17:43 scan will not show much 2019-06-03 04:17:58 tried station wlan0 scan ; station wlan0 show 2019-06-03 04:18:01 got zip 2019-06-03 04:18:14 i'm not found of iwctl repainting the whole screen 2019-06-03 04:18:28 it's a terrible curses app 2019-06-03 04:19:28 for anyone that cares, setting ESSID via console is a broken feature as ESSID's are octet based strings not ascii/utf-8 or other strings 2019-06-03 04:19:57 so it needs to be an integer selection of visible essid or else provide a hexedit mode 2019-06-03 04:20:46 iwctl station mlan0 connect network-name 2019-06-03 04:21:29 but, I'm going to drink coffee now, see you later 2019-06-03 04:21:46 @mps thank you for your gracious assistance, it's been fantastic 2019-06-03 04:22:27 you are welcome :) now I know that iwd works on x86 2019-06-03 04:25:16 I'll let you know how pm-suspend plays with iwd too ;) 2019-06-03 04:36:39 superduck: suspend-to-ram and resume works on my box 2019-06-03 05:59:37 yunfan: I have some success with booting armv7 Alpine on sunxi board 2019-06-03 12:53:16 hi all :-) 2019-06-03 12:55:09 Hello there 2019-06-03 18:55:44 <_ikke_> I have one 3.8 vm and one 3.10 vm, both nearly identical in setup (started the same way under qemu). The 3.8 has /dev/dri/card0, the 3.10 vm does not. What am I missing? 2019-06-03 18:56:28 <_ikke_> both have xf86-video-modesetting and xf86-video-qxl isntalled 2019-06-03 18:56:46 <_ikke_> lspci shows a vga controller is available 2019-06-03 19:25:26 <_ikke_> Not sure what I'm missing 2019-06-03 19:29:53 Do you run udev stuff on boot? 2019-06-03 19:30:34 Not quite sure what the setup would look like with mdev, but I've encountered that when I forgot to add udev-trigger udev-settle and udev-postmount to my sysinit 2019-06-03 19:33:54 <_ikke_> yay for long irc history 2019-06-03 19:33:59 <_ikke_> nomodeset kernel parameter 2019-06-03 19:34:06 <_ikke_> was causing it 2019-06-03 19:35:19 <_ikke_> Shiz: thanks :P 2019-06-03 19:36:01 <_ikke_> now it works, yay 2019-06-03 19:36:43 Oh :D 2019-06-03 19:37:05 Nice 2019-06-03 19:42:46 <_ikke_> funny is that the 3.8 vm has nomodeset and just works fine 2019-06-03 19:42:59 <_ikke_> probably something in the kernel 2019-06-03 19:46:07 hello, i have found this suspect line in dovecot edge build "./configure: line 27716: rpcgen: not found" and rpcgen is available here : https://pkgs.alpinelinux.org/package/edge/main/x86_64/rpcgen -> log build : https://build.alpinelinux.org/buildlogs/build-edge-x86_64/main/dovecot/dovecot-2.3.6-r1.log 2019-06-03 19:47:14 same with all version (3.9, 3.8, 3.7, not tested after) 2019-06-03 19:48:01 <_ikke_> Tetsumaki: Are you facing an issue because of this? 2019-06-03 19:51:08 no, never, but i say it in case you find it relevant 2019-06-03 19:51:39 <_ikke_> Ok, I think the fix is simple, just wondering if it caused any issue 2019-06-03 19:52:58 Yes 2019-06-03 22:53:01 is there a way for me to check which exactly release my alpine continer is running? 2019-06-03 22:53:54 cat /etc/os-release 2019-06-03 22:54:09 `/etc/alpine-release` is a likely candidate for that 2019-06-03 22:54:34 or, as Cogitri wrote, faster than me ;) 2019-06-04 07:17:33 it won't always be spot on though 2019-06-04 08:56:05 anyone got jitsi meet server working on alpine linux? 2019-06-04 17:05:06 Reminder. If you are maintaining a package please check for it in: https://github.com/alpinelinux/aports/projects/1#column-5386375 2019-06-04 17:23:32 Maybe we could have a both which pings the maintainer like NixOs does? 2019-06-04 17:25:05 would be nice 2019-06-04 17:25:13 Cogitri: pinged on irc or via email? 2019-06-04 17:25:20 i'm not familiar with how nixos does it 2019-06-04 17:26:49 They just do @githubusername on GitHub 2019-06-04 17:28:01 <_ikke_> how do they determine the maintainer? 2019-06-04 17:28:04 Cogitri: did you mean bot? 2019-06-04 17:28:26 Ah, yes 2019-06-04 17:28:30 There are OWNERS and CODEOWNERS files that try for that kind of notification 2019-06-04 17:28:39 https://github.com/NixOS/nixpkgs/pull/19315 2019-06-04 17:28:48 We have an issue mapping email addresses to GH usernames 2019-06-04 17:28:54 Yup, they do it by analysing the history, we could use CODEOWNERs 2019-06-04 17:28:59 <_ikke_> That would need to be maintained though 2019-06-04 17:29:22 That can be automatically generated, can't it? 2019-06-04 17:30:22 Our issue with CODEOWNERS is that maintainers are not all in the project 2019-06-04 17:30:42 We could make a bot that traces git history, but that seems like a lot of work 2019-06-04 17:31:04 that'd be a bit more work but not impossible 2019-06-04 17:31:30 danieli: to be useful it'd need to trace through moves 2019-06-04 17:31:31 git commits are published on mqtt together with author and hash and stuff, so a bot could listen for git/# messages sent via msg.a.o and fetch commit info from cgit based on that 2019-06-04 17:31:52 Maybe we can just reuse what Nix has 2019-06-04 17:32:06 if it's open source, i don't see why not 2019-06-04 17:33:09 https://github.com/facebookarchive/mention-bot 2019-06-04 17:33:21 Not maintained anymore though, I'll search for something else in a bit I guess 2019-06-04 18:57:56 May, somebody here can confirm or disprove that the nginx packages is build without ipv6 support? 2019-06-04 19:06:56 <_ikke_> We use nginx with ipv6 2019-06-04 19:14:00 hm was a bit precarious because of the --with-ipv6 flag,so probably deprecated? 2019-06-04 21:28:03 Does Apache in edge support http2 proxy? 2019-06-04 21:28:36 parazyd: i'd check that against the version number if i were you 2019-06-04 21:28:48 or does it have to explicitly be compiled with http2 support? 2019-06-04 21:28:52 Mmm, seems not. 2019-06-04 21:28:54 --enable-proxy-http \ 2019-06-04 21:28:59 Only this is in the APKBUILD 2019-06-04 21:29:36 try it, it might just need recompilation to support it 2019-06-04 21:29:44 Yep, will do :) 2019-06-04 21:46:01 danieli: Indeed! I'll send the maintainer a patch. 2019-06-04 21:51:04 Sent! 2019-06-05 00:01:35 Give me ONE reason not to install alpine on this x220 tablet RIGHT NOW 2019-06-05 00:12:07 periffis: x220 tablet? 2019-06-05 00:15:09 mps: Aye. 2019-06-05 00:15:56 it's intel based, so I presume it is not hard to install 2019-06-05 00:17:23 I have alpine on samsung chromebook one plus (arm64) which also can be in tablet mode 2019-06-05 01:08:23 is it possible to install and start sshd from a new install on a rpi so you can ssh to it and not need a monitor at all? 2019-06-05 03:41:05 m0rr1gan: you'll need a monitor for installation. After that you can just ssh in 2019-06-05 06:35:25 does the rpi image not obtain an IP via DHCP and all that jazz? 2019-06-05 07:22:37 is it possible to install and start sshd from a new install on a rpi so you can ssh to it and not need a monitor at all? : I think it's do-able 2019-06-05 07:55:22 Hello Alpine community! 2019-06-05 07:58:09 I have two questions. #1: does Alpine Linux support Linux integrity measurement architecture(IMA)? if not, is there any plan for supporting it in the future? #2: is it possible to deploy Alpine Linux on bare metal machines using Ubuntu MaaS? 2019-06-05 07:58:10 Hello there 2019-06-05 08:29:09 yunfan: found which driver missing for serial console for allwinner/sunxi board 2019-06-05 08:30:44 python: name 'httpd' Not defined. anyone know where python modules are kept? python{2,3}-dev? 2019-06-05 08:40:14 hmm actually this might be a bug with fish-tools... 2019-06-05 08:43:09 xcko: there isn't a built-in module 'httpd' 2019-06-05 08:43:16 there's http.server in python 3, or SimpleHTTPServer in python 2 2019-06-05 08:43:45 thanks, I'm not familiar with python. I just poked through some of the code and saw that too 2019-06-05 08:47:41 I think it may be because I don't have ipv6 built into my kernel. 2019-06-05 08:49:18 I doubt it 2019-06-05 08:52:52 well, wanna see if you can replicate? install fish and fish-tools, run fish, then run fish_config 2019-06-05 08:54:22 here is a paste of the traceback https://0x0.st/zusv.txt 2019-06-05 08:58:50 Guest50634: the kernel supports it but the distro itself doesn't really have anything specific to support it. And if you ask me, it's quite unlikely there will be distro support as IMA is, let's put it nicely, challenging to set up 2019-06-05 08:59:11 Guest50634: actually, I think the Alpine kernel doesn't have everything IMA-related enabled 2019-06-05 09:01:44 xcko: i found these: https://github.com/fish-shell/fish-shell/issues/3857 https://github.com/fish-shell/fish-shell/issues/3530 2019-06-05 09:07:13 interesting thanks danieli 2019-06-05 09:38:15 well added ipv6 to my kernel and now fish_config works, thanks 2019-06-05 09:55:48 tangentially, it seems fish-tools works with python3 yet python2 was installed as a runtime dep. Is that something to email the maintainer about or should I just submit a patch to aports? 2019-06-05 10:26:57 <_ikke_> xcko: submitting a PR will do 2019-06-05 10:27:14 <_ikke_> If it's a small and obvious fix, it can just be applied 2019-06-05 10:30:23 in the fish APKBUILD it I see python listed in the depends, so maybe that's fine and someday alpine will switch the default python to python3? 2019-06-05 10:31:11 it doesn't appear to explicitly ask for python2, so I'm not sure there is anything to change 2019-06-05 10:31:15 default python should be 3 already 2019-06-05 10:31:31 <_ikke_> xcko: the python package has been replaced with python2 2019-06-05 10:31:54 <_ikke_> so either remove it, or replace it for python3 if it's not present 2019-06-05 10:32:53 <_ikke_> https://git.alpinelinux.org/aports/tree/main/python2/APKBUILD#n13 2019-06-05 10:33:04 <_ikke_> so if you install python, you get python2 2019-06-05 10:33:32 <_ikke_> (it's the provides= one line above that causes this) 2019-06-05 10:35:50 xcko: I'm looking on your patch https://patchwork.alpinelinux.org/patch/4883/ and thinking to apply it, and would like to hear your rationale for it and maybe if someone have objections 2019-06-05 10:36:49 rationale for usecase or rationale for writing the APKBUILD the way I did? 2019-06-05 10:37:29 usecase, sorry I wasn't clear 2019-06-05 10:37:56 mps: remove empty variables on root scope of the APKBUILD, default value builddir also remove. 2019-06-05 10:38:48 my main question which I asking self is: does anyone still use POP3, aren't we all moved to IMAP 2019-06-05 10:39:22 north1: I had these in mind 2019-06-05 10:39:32 mps: Greg (kernel dev) said he used pop3 to apply lots of patches when ddevault showed his aerc terminal client 2019-06-05 10:39:38 north1: thanks, forgot to add 2019-06-05 10:40:42 ah no confused myself, he said mbox instead of Maildir 2019-06-05 10:41:06 ah I use pop3 because I'm used to it. I don't want to leave anything on my mail provider, just download everything and delete it. 2019-06-05 10:41:22 mpop is written by the same person that wrote msmtp, which is in main 2019-06-05 10:42:14 honestly though, if you have an easy tool for getting mail over imap with easy socks5/tor support, I'd consider switching 2019-06-05 10:43:32 xcko: np, you (and north1) convinced me. I'll test build and if it pass apply 2019-06-05 10:43:47 my nickname appears as north1 ? 2019-06-05 10:43:52 it should be maxice8 2019-06-05 10:44:19 <_ikke_> where? 2019-06-05 10:44:22 north1: I see 'north1' 2019-06-05 10:44:25 <_ikke_> Yes 2019-06-05 10:44:31 <_ikke_> it's north1 on irc 2019-06-05 10:45:26 <_ikke_> Now it's good 2019-06-05 10:45:34 ah nice 2019-06-05 10:46:47 maxice8: so, from now you are not north1? 2019-06-05 10:47:17 i used north1 to register my IRC account and my voidlinux forums (RIP) account 2019-06-05 10:47:28 better move this to -offtopic too 2019-06-05 10:47:44 in the future though, remove empty vars and default builddir for pkg builds right? 2019-06-05 10:48:05 <_ikke_> Certainly for new packages 2019-06-05 10:48:27 xcko: you can use apkbuild-lint from maxice8/atools@GitHub to check for when you should do these things 2019-06-05 10:48:32 it is available as atools in the repo 2019-06-05 10:49:15 xcko: for vars yes, but builddir should be left in, especially if pkg could be moved to main. we discussed this yesterday on #alpine-devel 2019-06-05 10:49:28 mps: cd "$builddir" you mean 2019-06-05 10:49:47 yes 2019-06-05 10:50:01 unless version 2.29.0 of abuild is still around in old alpine 2019-06-05 10:50:48 <_ikke_> for new packages it shouldn't matter though. Even if they are moved to main, there is nothing to backport to older versions 2019-06-05 10:51:00 cd "$builddir" will be fair game when abuild version 3.3.0 is no longer around 2019-06-05 10:51:43 _ikke_: well, maybe. we are not sure if something wouldn't be backported 2019-06-05 10:52:04 3.3.0? Damn, only 3.9 and newer has that... 2019-06-05 10:52:20 yes, when backporting these lines could be added, I know 2019-06-05 10:52:23 <_ikke_> mps: It's rare (but it does happen) that a new package is backported 2019-06-05 10:53:24 ok, I will not nitpick about this. do whatever you think is appropriate 2019-06-05 11:11:32 maxice8: it's been north1 on irc all along :P 2019-06-05 11:14:23 herp derp, https://pkgs.alpinelinux.org/package/edge/community/x86/pdns should have GPL-2.0-only as the license 2019-06-05 11:14:29 people are blind 2019-06-05 11:18:19 it was last updated by tcely 2019-06-05 11:24:48 _ikke_: http://ix.io/1KYP one line change, not tested yet though. I can test it tomorrow and mail to aports 2019-06-05 11:25:54 <_ikke_> Feel free to do so 2019-06-05 11:28:55 danieli: I created a bug since i figure that's the best way to get attention :P 2019-06-05 11:29:12 danieli: only noticed since I'm making OpenWRT packages for it 2019-06-05 11:29:13 xcko: Also need to bump pkgrel 2019-06-05 11:30:12 and -that- is a minefield 2019-06-05 11:30:55 <_ikke_> tcely: ^ 2019-06-05 11:53:43 xcko: I cleaned mpop APKBUILD and removed unneeded commit messages. I passed all builders :) 2019-06-05 16:44:08 0/ 2019-06-05 16:44:16 <_ikke_> hello 2019-06-05 16:45:49 :) 2019-06-05 16:46:32 didnt know dropbear became so popular it's offered as a choice when installing ssh 2019-06-05 16:46:52 <_ikke_> I think that's already quite a long time the case 2019-06-05 16:47:11 it's quite popular on embedded devices for one 2019-06-05 16:47:18 Does anyone know what the store config question in alpine is for? is that like a kickstart config? 2019-06-05 21:07:37 Hi all. We are trying to build postmarketOS for my device but we are getting error about colord-dev package. I checked from packages website of you but i couldn't see armv7 build for that package. May i ask is there any ETA for building armv7? Thanks in advance. 2019-06-05 21:12:34 PureTryOut[m]: perhaps you know? ^ 2019-06-05 21:12:47 oh nevermind, looks like they found #alpine-devel 2019-06-05 21:12:50 <_ikke_> revolutionary also askes this in #alpine-devel 2019-06-05 21:12:52 <_ikke_> ye 2019-06-05 21:13:02 yes i tried my chance 2019-06-05 21:13:13 is this crime? 2019-06-05 21:14:14 Well, asking in two channels at the same time usually means wasting effort :) 2019-06-05 21:14:22 If no one answered after some time it's reasonable to ask elsewhere, but maybe don't do that immediately :) 2019-06-05 21:14:51 sensible :) 2019-06-05 21:17:00 it's not a crime, it's just confusing 2019-06-05 21:17:18 thanks. sorry for the noise 2019-06-05 21:18:14 no worries 2019-06-05 22:30:48 My capistrano seems to fail with a tar command: https://gist.github.com/nuheluxulu/d7d8edfe54a9084ca6112ddce1e261d1 Anyone know what the cause is? 2019-06-05 22:40:53 nvm, figured it out :) 2019-06-05 22:41:27 I'm an idiot. its git related 2019-06-05 23:57:28 thanks terror, it looked that way from the installation guide also thanks tmhoang, unless I built a custom image, not possible it doesn't look like but yes, doable potentially 2019-06-06 00:21:31 well switching fish to use python3 would require changing the shebangs to /usr/bin/env python3 instead of /usr/bin/env python. I can patch it, or wait for alpine to change the default python in time, the latter seems to be the better option. 2019-06-06 03:54:05 mps: sorry , my proxy service had been baned the last days 2019-06-06 08:49:19 yunfan: no problem, have you seen my messages that I found which drivers need to be enabled in kernel to boot allwinner/sunxi board 2019-06-06 08:49:36 and, good morning/day :) 2019-06-06 09:47:03 Hey, my iso came with no mirrors besides the CD mirror 2019-06-06 09:47:05 what do I do now 2019-06-06 09:47:11 is there some way I can generate a mirrors list 2019-06-06 09:47:22 online and easilly copy and paste it like with debian? 2019-06-06 09:47:41 I noticed it sometimes happens 2019-06-06 09:48:00 kubast2: setup-apkrepos iirc 2019-06-06 09:48:13 it doesn't come with repos pre-setup 2019-06-06 09:48:31 i'm pretty sure that's in the installation guide(s?) 2019-06-06 09:49:01 I need to verify first that I gotten the internet connection right 2019-06-06 09:49:12 I'm doing a manual install on a vps with static ip 2019-06-06 09:49:40 yep it doesn't ping 8.8.8.8 so hmm maybe I put a wrong ip mask? 2019-06-06 09:50:47 Yeahhhh 2019-06-06 09:50:54 rip 2019-06-06 09:51:12 I thought I didn't do that but since running alpine-setup again I have probablly put gateway into ip addres 2019-06-06 09:51:32 okay I still get wget bad address 2019-06-06 09:52:15 try setup-apkrepos when you've fixed your internet connection 2019-06-06 10:02:47 danieli, my inet gets set to gateway 2019-06-06 10:02:48 and /8 2019-06-06 10:02:51 instead of /27 2019-06-06 10:03:15 now I have culprit 2019-06-06 10:03:29 the alpine-setup 2019-06-06 10:04:25 it doesn't magically break it, it gives you three options 2019-06-06 10:04:29 no network, static, or dhcp 2019-06-06 10:06:32 Yeah I used static 2019-06-06 10:06:40 and typed in my ip address, my network mask etc. 2019-06-06 10:09:25 /8 is mask 255.0.0.0, /27 is 255.255.255.224 2019-06-06 10:09:27 seems strange 2019-06-06 10:18:54 yeah I rebooted the iso and set everything as it was, I can't ping 1.1.1.1 dns server which was the one I was told to use, and the one which worked in my debian install 2019-06-06 10:19:09 I can't resolve domains I guess is the pboelm, but what I see now 2019-06-06 10:19:19 is that my debian install had a non-tandard broadcast address? 2019-06-06 10:19:52 or it just tells me about the broadcast since it isn't 2019-06-06 10:19:56 on /8 /16 /24 2019-06-06 10:20:17 morning sirs 2019-06-06 10:21:00 on /8 /16 /24 2019-06-06 10:21:01 fuck 2019-06-06 10:21:10 broadcast != netmask 2019-06-06 10:21:30 Yeah IK 2019-06-06 10:21:38 but broadcast depends on netmask 2019-06-06 10:22:00 hmm I think I just can't ping anything outside of gateway/local network 2019-06-06 10:22:02 on alpine 2019-06-06 10:22:36 I set ip address, netmask, gateway now everything is inline, even the dns address with current debian install 2019-06-06 10:22:47 but I can't ping anything outside of a local network realm 2019-06-06 10:23:29 and it shows the correct setup in `ip link` + correct default route in `ip route`? 2019-06-06 10:23:45 er, you might have to use ifconfig/route 2019-06-06 10:24:10 yep 2019-06-06 10:24:36 ah I also have to set a broadcast address 2019-06-06 10:25:01 bcast = mask | ip 2019-06-06 10:25:06 it should figure it out automaticall 2019-06-06 10:25:08 +y 2019-06-06 10:25:32 well it didn't for me 2019-06-06 10:25:54 what's in your arp table? 2019-06-06 10:26:56 try adding a default route, ip route add default via $IP_ADDRESS 2019-06-06 10:28:00 duplicate I guess 2019-06-06 10:28:07 and it also says 2019-06-06 10:28:28 WAit "default via my ipaddress dev eth0" 2019-06-06 10:29:02 should be via gateway 2019-06-06 10:29:17 er, gateway ip 2019-06-06 10:29:20 "ip: either "to" is duplicate or "gateway ip/27" is garbage 2019-06-06 10:29:40 uh? 2019-06-06 10:29:54 you're supposed to replace it with the gateway's IP address, not literally the string "gateway ip" 2019-06-06 10:29:54 I deleted all ip route interfaces 2019-06-06 10:30:13 without mask? I didn't use quotes 2019-06-06 10:30:14 should not touch loopback btw 2019-06-06 10:30:34 okay I'm a dumbass 2019-06-06 10:30:53 ip route change 2019-06-06 10:31:04 looool 2019-06-06 10:31:09 yeah 2019-06-06 10:31:21 ip route add default via 192.168.1.1 (if your router is 192.168.1.1, sorry) 2019-06-06 10:32:00 either "to" is duplicate or "89.XXX.XXX.XXX" is garbage 2019-06-06 10:32:35 to configure your nic with an address, just do "ip addr add 192.168.1.10/24 brd + dev eth0" 2019-06-06 10:32:43 say your IP = 10.0.0.5, gateway = 10.0.0.1, block is /27, netmask = 255.255.255.224, bcast = 10.0.0.31 2019-06-06 10:32:48 n0a110w: i'm not sure alpine comes with iproute2 by default 2019-06-06 10:32:55 ooooo 2019-06-06 10:34:19 it does not 2019-06-06 10:34:46 only network related packages the default standard install includes is busybox and network-extras (metapackage for various things but not iproute2) 2019-06-06 10:36:21 Okay I have now ip address+netmask+broadcast 2019-06-06 10:36:48 my b 2019-06-06 10:37:28 no worries :) ifconfig and friends *are* deprecated afterall 2019-06-06 10:37:39 in linux anyway 2019-06-06 10:38:14 well I need to set a gateway/route now 2019-06-06 10:39:57 if you have the `ip` command, run `ip route add default via $GATEWAY_IP dev $INTERFACE` 2019-06-06 10:40:31 yeah I finally can ping outside of gateway/via gateway 2019-06-06 10:49:19 danieli, is there some tutorial on /etc/network/interfaces 2019-06-06 10:49:43 I never used those and when I did I often used pregenerates 2019-06-06 10:51:31 or used dhcp all together 2019-06-06 10:51:57 yeah okay it works just like that now 2019-06-06 10:55:11 almost there, I can't ping my vps from my network 2019-06-06 11:03:23 and I broke it again 2019-06-06 11:03:47 so yeah putting gateway into network interfaces is a no no 2019-06-06 11:04:34 setting default manually and setting a wrong one in etc network interfaces works fine tho 2019-06-06 11:06:34 or I'm jsut stupid and tired because I went to sleep at 4 am 2019-06-06 11:06:55 probably a good idea to ask for support from your hosting provider 2019-06-06 11:07:04 danieli, they had issues with it too so 2019-06-06 11:07:08 he just shrugged of at some point 2019-06-06 11:07:18 You can manage via console? Okay, if need something, please let me know. I will close ticket for this for now okay? 2019-06-06 11:07:37 yeah I got it this time 2019-06-06 11:07:53 idk why the wrong gateway is shown under the correct one in default 2019-06-06 11:07:55 but it works 2019-06-06 11:07:58 and idc for now 2019-06-06 11:08:37 I mean at least now I ahve a promox interface/account 2019-06-06 11:08:45 so I can finally reboot the vps and access vnc 2019-06-06 11:09:57 and I guess now I have backups and snapshots and access to firewall settings 2019-06-06 11:19:50 Hello! Please, can anyone help me? I can`t find apk of versions 1.x 2019-06-06 11:20:21 <_ikke_> PHPoenX: Welcome. Can you please clarify what you mean, versions 1.x of what? 2019-06-06 11:21:47 I think he means the 1.x tree of apk? We're at 2.x right now, aren't we? 2019-06-06 11:22:23 _ikke_: apk is alpine package keeper 2019-06-06 11:22:31 we know that, but we don't know what you mean 2019-06-06 11:24:59 Cogitri: yes, where is 1.x tree of apk? I found only 2.x 2019-06-06 11:27:32 <_ikke_> Is there a reason you need version 1.x of apk? 2019-06-06 11:27:41 <_ikke_> (fyi, I don't know where it is) 2019-06-06 11:32:53 Yes, I have reason 2019-06-06 11:34:47 not willing to share..? 2019-06-06 11:37:37 Apk 1.x is shell script, I want see how it works and use some things from it 2019-06-06 11:39:09 wherever it is, it's old, obsolete, and probably insecure 2019-06-06 11:54:23 danieli: I'm not going to use the apk of this version. I want to look at the code and take some developments from it for another project 2019-06-06 15:53:42 yunfan: are you here? I want to add your board to Alpine u-boot if you can test it 2019-06-06 15:54:00 what is model of your board 2019-06-06 19:21:22 Hello people o/ Trying to pxeboot alpine in raspberry 3b+. TFTP is working great, vmlinuz-rpi and initramfs-rpi are loaded... But /sbin/init don't start. I will try to use mkinitfs now. There is any doc already done about pxe and raspberry? Thank you! 2019-06-06 20:03:05 Is there a way to list explicitly installed packages with apk? I didn't see anything about it on the Wiki. 2019-06-06 20:03:43 yebl: cat /etc/apk/world 2019-06-06 20:04:06 mps: Awesome, thanks! 2019-06-07 07:58:26 Hi, I'm trying to bootstrap into a directory, and have a 2 custom packages, one is of type 'all' (thus armhf in my case in the APKINDEX file), the other is just a few shell scripts, so is of the type 'noarch' (also in the generated APKINDEX file). However, when I do a 'apk add --arch armhf --root /tmpdir --initdb --update-cache busybox pkg1 pkg2' apk fails with ERROR: pkg1: package mentioned in index not found (try 2019-06-07 07:58:26 'apk update'). If i change the noarch to all however, it all works fine. Where am I going wrong? 2019-06-07 08:00:17 oliv3r[m], generate index with "--rewrite-arch armhf" or similar. currently noarch is not supported in indexes. 2019-06-07 09:05:48 how does the online repo do this? all packages in any subrepo are listed as arch $arch? 2019-06-07 09:15:24 in any case, yes, this did the trick :) 2019-06-07 09:15:27 thanks fabled 2019-06-07 09:25:00 I'm running setup-alpine right now, and notice that all of the mirrors it is detecting right now are testing over http. What's up with that? Couldn't I get MITM'd? 2019-06-07 09:25:40 mirrors always use a trustless server model anyone can host it and the packages are signed 2019-06-07 09:26:21 you cant have a signed domain and allow other people to attatch to the network without exposing the private key so it doesnt make any difference 2019-06-07 09:27:12 ahhh, i see 2019-06-07 09:27:18 thanks, spare 2019-06-07 09:34:19 Whoops. man not found. 2019-06-07 09:34:24 What should I be using instead of man? 2019-06-07 09:34:35 mdocml 2019-06-07 09:34:41 <_ikke_> apk add docs 2019-06-07 09:34:55 Also, upon reboot, it's not connecting to the network. 2019-06-07 09:36:13 maybe we should remove man 2019-06-07 09:36:56 I can't tell--there are too many vagueries on irc anymore--are you just implying I wouldn't be trying to use alpine if man didn't exist? 2019-06-07 09:37:03 Or is this a joke about removing humanity 2019-06-07 09:37:37 mps: ah, you woke up 2019-06-07 09:38:20 maxice8: I'm trying :) 2019-06-07 09:38:34 Is there a recommended/built-in window manager or desktop environment in alpine? 2019-06-07 09:38:56 periffis: I've used i3 in the past 2019-06-07 09:39:10 On a laptop? 2019-06-07 09:39:12 periffis: mdocml is better supported in Alpine than man (not hu-man) 2019-06-07 09:39:21 periffis: yes 2019-06-07 09:39:33 That's neat. If I could only connect to the network...sec 2019-06-07 09:40:02 aha! i was using /etc/wpa_supplicant.conf instead of its actual directory 2019-06-07 09:40:16 alright, now that I used wpa supplicant to connect, I can't just ctrl-c out of it, right? 2019-06-07 09:42:05 Seems interesting that the docs aren't already included--if I'm not on the network, I'd argue that it would be nice to get the docs to find out how to properly connect to the network 2019-06-07 09:42:20 but perhaps i'm being too hopeful about someone like me getting the docs and learning without just failing back to another connection 2019-06-07 09:42:55 alpine tries to be small (use less space) 2019-06-07 09:43:57 for nearly every package there is corresponding -doc package 2019-06-07 09:44:51 you can install all of them for installed packages by 'apk add docs' as _ikke_ wrote 2019-06-07 09:46:07 mps: i finally got on the network by putting wpa_supplicant in the background with & (is that how i'm supposed to do it?), and then running udhcpd manually to get an ip address. didn't work on boot for some reason 2019-06-07 09:46:16 worked now, though. finally got it. 2019-06-07 09:47:36 periffis: never tried install by wifi so don't know, but hear from some people that it work 'out of box' 2019-06-07 09:47:57 heh, interesting 2019-06-07 09:49:01 add wpa_supplicant to openrc boot stage 2019-06-07 09:49:14 i think it was--let me double check 2019-06-07 09:49:19 # rc-update add wpa_supplicant boot 2019-06-07 09:50:25 isn't it added by default if install is done over wifi 2019-06-07 09:52:55 maxice8: will that overwrite any other config? and yeah, i did install over wifi. also, i tried following this: https://wiki.alpinelinux.org/wiki/Awesome(wm)_Setup and i'm getting "unsatisfiable constraints" with aterm and awesome. do i need to update something? 2019-06-07 09:53:30 no, but it won't respect custom commandline options you gave to it 2019-06-07 09:53:37 it will use /etc/wpa_supplicant/wpa_supplicant.conf by default 2019-06-07 09:53:51 awesome pkg is in community/ so you will need to enable it 2019-06-07 09:54:37 periffis: oh, nice. there are three of us who use awesome wm :) 2019-06-07 09:55:05 mps: lmao 2019-06-07 09:55:28 also, if i do "apk upgrade" and it just says "OK: 694 MiB in 167 packages," is that telling me it just upgraded instantaneously? 2019-06-07 09:55:53 <_ikke_> periffis: no, that's what on your system 2019-06-07 09:55:57 it tells you there is nothing to upgrade 2019-06-07 09:56:05 usually that means all is upgraded already 2019-06-07 09:56:17 ah, sweet. 2019-06-07 09:56:36 apk version will tell if something is ready for upgrade 2019-06-07 09:56:42 you want to do apk update first or run apk upgrade -U 2019-06-07 09:56:53 apk tries to be quiet 2019-06-07 09:57:26 makes sense. 2019-06-07 09:57:53 i still can't install awesome. aterm and awesome are missing themselves. does this mean my particular mirror for my repo doesn't have it? 2019-06-07 09:58:18 periffis: you need to add community repo 2019-06-07 09:58:22 periffis: do you have the community/ repo enabled ? awesome is in it 2019-06-07 09:58:25 ahhh 2019-06-07 09:58:28 aterm is not packaged 2019-06-07 09:58:35 what's aterm? 2019-06-07 09:58:44 in /etc/apk/repositories 2019-06-07 09:58:55 wow, i can't believe i completely skimmed over the community repository bit on the wiki 2019-06-07 09:58:59 bad habits on reading wikis 2019-06-07 10:00:13 you can try setup-xorg-base 2019-06-07 10:02:01 dang. does that just detect a bunch of system settings and go with defaults? that's kinda neat 2019-06-07 10:02:13 that worked, mps 2019-06-07 10:02:15 you're a genus 2019-06-07 10:02:27 although i had also enabled the community repo manually before 2019-06-07 10:03:01 oh, no, I'm like "Alice in wonderland" ;) 2019-06-07 10:04:41 "Experience is most brutal teacher, but you will learn, my dear you will learn" Lewis Carrol - Alice in wonderland 2019-06-07 10:15:17 mps: heh 2019-06-07 10:16:16 does apk not come with visudo? 2019-06-07 10:17:18 no, visudo is part of `sudo` 2019-06-07 10:17:20 periffis: visudo is in sudo apk 2019-06-07 10:19:37 interesting. 2019-06-07 10:19:57 is there a better way to do things than using sudo and a lower-privileged user? 2019-06-07 10:22:13 i just use doas which is basically sudo but smaller 2019-06-07 10:22:54 periffis: sup 2019-06-07 10:23:26 but only for some definition of "better" 2019-06-07 10:24:25 what the hell is doas, oh boy 2019-06-07 10:24:27 more to learn 2019-06-07 10:24:37 KatolaZ: is sup literally a command, lol 2019-06-07 10:24:50 OpenBSD's sudo, basically 2019-06-07 10:24:59 doas is simpler, if you just need one user to get a root shell quickly 2019-06-07 10:25:15 You could use polkit for more fine grained access I guess (due to using JS for rules) 2019-06-07 10:25:18 periffis: sup is sudo but from suckless. 2019-06-07 10:25:23 lmao 2019-06-07 10:25:33 i'm lolling at the idea of asking my terminal "sup" 2019-06-07 10:25:49 do all of these tools use the same backend concepts? 2019-06-07 10:25:57 no periffis 2019-06-07 10:26:02 sup has no config file 2019-06-07 10:26:07 all rules are compiled in 2019-06-07 10:26:26 this is interesting 2019-06-07 10:28:31 i didn't know about sup. but i love things that suck less. so im glad i caught that 2019-06-07 10:28:54 what makes suckless and its tools better, other than the obvious built-in "marketing" of the name "suck less" 2019-06-07 10:29:13 is suckless related to cat-v? 2019-06-07 10:29:43 I'm using only st, heavily patched and tweaked 2019-06-07 10:29:51 better is not the word , because that's an opinion . read suckless's philosophy 2019-06-07 10:30:13 i use dwm, st, dunst, dmenu . my own fork of dwm and st as is necessary.. 2019-06-07 10:31:15 ohhh. and slock. almost forgot about slock. 2019-06-07 10:31:18 some time also used dunst but now I even forgot for it 2019-06-07 10:40:02 what's st? 2019-06-07 10:42:21 periffis: suckless terminal 2019-06-07 10:42:36 ah, neat. terminal application? is it barebones? 2019-06-07 10:43:28 yes, but can be extended by patches to be very 'featurefull' 2019-06-07 10:44:16 <_ikke_> How is st compared to urxvt? 2019-06-07 10:48:10 _ikke_: well, urxvt was my terminal for two or three decades and it is quite good 2019-06-07 10:49:11 but about year or two ago I switched to st because it have more useful features for my use case 2019-06-07 10:49:58 try out my fork of st: https://github.com/jsav0/st 2019-06-07 10:50:25 <_ikke_> can st be set to have no graphical interface? 2019-06-07 10:50:48 font increase/decrease works very well, scrollback with keys, copy/paste delimiters, and small size small memory use 2019-06-07 10:51:12 it has all the 256colors you need :) 2019-06-07 10:51:18 and, I think st is best on color handling 2019-06-07 10:51:45 n0a110w: I have xst fork also 2019-06-07 10:51:57 share :) 2019-06-07 10:52:38 Should I be using aterm? I'm on kernel 4.19. 2019-06-07 10:52:51 https://github.com/gnotclub/xst.git 2019-06-07 10:53:23 it have xresource patch 2019-06-07 10:53:40 And that would go in rc.lua? 2019-06-07 10:53:47 For some reason, startx isn't working 2019-06-07 10:53:57 Also, the error goes off of the screen, lol 2019-06-07 10:54:35 n0a110w: although last update is year old 2019-06-07 10:54:47 Hello, again! What about apk of 1.x versions? 2019-06-07 10:54:53 you can debug startx with Xorg log or debug your .xinitrc 2019-06-07 10:56:33 n0a110w: we have st-xrdb in testing. I thought to extend it with some other patches 2019-06-07 10:58:37 Ah, it needed dbus. 2019-06-07 10:58:51 ahh neat! i may eventually lean on my .Xresources , maybe soon 2019-06-07 10:59:12 i like it 2019-06-07 11:00:34 it is easier to tweak by changing values in .Xresource than changing source and rebuilding 2019-06-07 11:05:05 you are right indeed. if we can get keybindings incorporated into .Xresources, i'm all in! 2019-06-07 11:06:28 yes, that is missing part 2019-06-07 11:07:29 but, by luck I don't change key often, it is good as it is. but will be useful for distro packaging 2019-06-07 11:09:07 Why is firefox in the edge repo? 2019-06-07 11:09:53 firefox is on testing/ which is built for edge only 2019-06-07 11:09:54 because firefox-esr is officially supported in alpine 2019-06-07 11:10:40 true. keybindings don't change much once you got em where you like em 2019-06-07 11:12:03 mps: i assume that if i want the most recent features in firefox, i'll have to enable the edge repo? 2019-06-07 11:21:36 periffis: mixing stable and edge could be problematic. If you want it then is better to change all /etc/apk/repository lines to edge 2019-06-07 11:35:09 mps: Can I easily upgrade my packages to their edge equivalents by doing so? 2019-06-07 11:46:48 periffis: usually, yes. apk update && apk upgrade 2019-06-07 11:47:00 ah, good. 2019-06-07 11:47:07 but anyway, read wiki about upgrading 2019-06-07 11:47:10 mps: do you use a display manager? 2019-06-07 11:47:52 <_ikke_> apk upgrade -U for short 2019-06-07 11:48:07 yes, just for fun and test, I use slim 2019-06-07 11:48:22 mps: how do i add slim to work on boot? 2019-06-07 11:48:40 rc-upgrade add slim 2019-06-07 11:48:50 upgrade or update? 2019-06-07 11:49:33 <_ikke_> update 2019-06-07 11:49:35 update download index files and upgrade do upgrade 2019-06-07 11:49:47 <_ikke_> he's talking about rc-update 2019-06-07 11:49:56 ah, so 2019-06-07 11:50:22 rc-update, of course, sorry 2019-06-07 11:50:35 hmm. wpa_supplicant running on boot seems to halt my boot process by quite a bit 2019-06-07 11:50:58 maybe not, actually. it just stops there pretty early 2019-06-07 11:51:15 <_ikke_> periffis: possible entropy issue? 2019-06-07 11:51:33 do i need to just shake the mouse on boot? loller 2019-06-07 11:52:03 why the hell would slim use enter instead of tab for me to--oh, right, because it's just wrapping around the process, i presume 2019-06-07 11:52:04 hmm 2019-06-07 11:58:12 .2 2019-06-07 12:01:20 is there a webpage I can check programmatically to get the current release version? I've been a bad admin and would like to make myself a reminder when I log into systems that are not running the current release. 2019-06-07 12:01:39 or better, is there a package with such a tool in it that only updates the cache like once a week or what have you. 2019-06-07 12:02:35 tw: for instance this path on any alpine mirror: http://uk.alpinelinux.org/alpine/latest-stable/releases/x86_64/latest-releases.yaml 2019-06-07 12:03:06 yank the 'version' field in the first part 2019-06-07 12:03:18 I see it, that'll be perfect. 2019-06-07 12:04:02 actually 2019-06-07 12:04:13 I think what I want is the branch. I have it check for updates already, which I think will keep with the point releases. 2019-06-07 12:04:21 it's possible to pull and parse the version from https://alpinelinux.org/downloads/ 2019-06-07 12:04:32 you can see the current version in bold, together wit the release date 2019-06-07 12:05:07 Less handy, imo. yaml is easier to process from cli tools. xmlstarlet can be finicky at times. 2019-06-07 12:06:48 agreed 2019-06-07 12:07:27 Hmmm. I'm going to try building hcxtools--it looks like it uses make and gcc, though. Is that likely to cause problems? 2019-06-07 12:08:22 try it. most likely not. 2019-06-07 12:08:46 glibc assumptions are what cause me problems building stuff. 2019-06-07 12:10:28 practically everything uses make and gcc 2019-06-07 12:10:45 weird, releases file is the only one that isn't signed. Not that I'm going to use it for anything implicitly trusted, but still. 2019-06-07 12:11:24 latest-releases.yaml? 2019-06-07 12:11:31 i'm not sure it's actually used for anything by us or by alpine 2019-06-07 12:11:44 Oh, so it could go away? 2019-06-07 12:12:00 Aha. I just needed build-base, curl-dev, and openssl-dev. 2019-06-07 12:12:39 When I'm in the wild and not just googling stuff, what should I actually do to understand where I can get stuff like libcurl? How would I know, without google, that libcurl was in the package curl-dev? 2019-06-07 12:13:05 you could also install the alpine-sdk metapackage for some commonly needed software 2019-06-07 12:13:28 periffis: for instance https://pkgs.alpinelinux.org/packages?name=*curl*&branch=edge&arch=x86_64 2019-06-07 12:13:50 aha. neat. 2019-06-07 12:13:50 or `apk update` -> `apk search curl` 2019-06-07 12:15:04 or even go to the curl package @ pkgs.a.o and look at sub-packages 2019-06-07 12:16:23 naja, he wants to build something and he knows he needs pkg x as a dep, so he needs x-dev to install for building against that dep. <- general rule of thumb 2019-06-07 12:19:55 ah, so it's conventional. that makes sense. 2019-06-07 12:20:58 yeah, that is a rule of thumb 2019-06-07 12:21:20 debian usually does something like lib-dev or something like that 2019-06-07 12:21:23 fedora -devel 2019-06-07 12:22:56 debian policy has every shared library have its own package 2019-06-07 12:52:08 holy crap, they really do, even for shared libraries that have explicit deps on each other, built from the same source. 2019-06-07 12:54:13 libsystemd0 2019-06-07 13:02:51 Anyone used aports/scripts/bootstrap recently? I am having issues bootstrapping armhf, complains that musl and musl-dev dependency cant be met, the host system musl-dev is installed and the musl and musl-dev armhf have been built successfully 2019-06-07 13:08:41 Seems to kick it back off again I need to manually apk add the armhf packages to the sysroot 2019-06-07 15:19:33 I thought up this cool pun--both ganglion and a lion in a gang--and I won't be able to even register an email address for it. I guess I could run email through gangli.online or some garbage for a billion a month. 2019-06-07 15:20:13 wrong channel? 2019-06-07 15:20:20 Aaaaanyway, I'm getting some interesting results in firefox. No numbers are showing up. Do I need to install some fonts? Is this up to firefox, up to awesomewm, or up to alpine? 2019-06-07 15:20:50 Also, is ash used specifically because it is lightweight? 2019-06-07 15:21:14 it's used because it's part of busybox 2019-06-07 15:21:24 Ah! Neat. 2019-06-07 15:21:25 ganglion: everything works fine with firefox and awesomewm 2019-06-07 15:21:54 mps: It must be me, then! I'm currently in a weird font that literally won't render numbers. 2019-06-07 15:22:03 adding additional fonts doesn't hurt 2019-06-07 15:22:11 you're free to install tcsh, csh, dash, bash, zsh, ksh (loksh) or any other shell, you don't have to stick to busybox ash 2019-06-07 15:22:34 and hm, i'd install some base font and run fc-cache -fv, then restart firefox 2019-06-07 15:24:41 what base fonts do you recommend, danieli? 2019-06-07 15:26:47 a monospace font like ttf-dejavu and font-noto/font-noto-emoji 2019-06-07 15:31:21 ohhh. that's way better. thanks, danieli 2019-06-07 17:10:26 Hey guys. 2 questions regarding alpine. 1: Why does the netboot default to 3.9.1, when 3.9.4 has been released and does in fact have netboot files on the mirrors?. 2: Am I doing something wrong, or is it correct the the zfs package, but not kernel modules, are included in the netboot release? 2019-06-07 19:28:50 how come less in alpine doesn't have a +G option 2019-06-07 19:30:06 <_ikke_> cim209: it's busybox less 2019-06-07 19:30:15 <_ikke_> use apk add less to get gnu less 2019-06-07 19:32:07 no wonder 2019-06-07 20:40:46 when i see "installing bash-doc," for example, or any -doc package, where are those docs going? 2019-06-07 20:41:08 more appropriately, how do i access them? is this listed in the pkg info? 2019-06-07 20:41:12 <_ikke_> https://pkgs.alpinelinux.org/contents?branch=edge&name=bash-doc&arch=x86_64&repo=main 2019-06-07 20:41:22 apk info -L bash-doc 2019-06-07 20:41:24 <_ikke_> or apk info -L bash-doc 2019-06-07 20:41:26 <_ikke_> right 2019-06-07 20:41:40 effin a 2019-06-07 20:41:54 okay, so i see that it goes to usr/share/doc, but what's the tool to read that stuff? 2019-06-07 20:42:30 doc bash isn't it. i know you know that, but if anyone else was guessing, they're wrong with me. 2019-06-07 20:42:58 <_ikke_> man 2019-06-07 20:43:01 <_ikke_> man bash 2019-06-07 20:43:21 -doc packages usually supply man pages 2019-06-07 20:43:26 that's a de-facto standard 2019-06-07 20:44:13 ah, i see. i was told yesterday that people were using mdocml instead of man, but i think i misinterpreted that to mean "mdocml has the same ui as man, just use that" which obviously isn't the case 2019-06-07 20:44:27 <_ikke_ "man"> Eh? Wouldn't that be /usr/share/man then? 2019-06-07 20:44:48 `/usr/share/doc` usually is HTML doc (e.g. for API) 2019-06-07 20:44:57 <_ikke_> `/usr/share/man/man1/bash.1.gz 2019-06-07 20:45:14 so when the doc package contains man pages, i assume that those man pages will be just as useful as the non-man expansions in the /doc/ folder? 2019-06-07 20:45:29 Cogitri: ah, interesting. 2019-06-07 20:45:31 <_ikke_> man pages are usually meant as references 2019-06-07 20:46:04 <_ikke_> Most of the time they are not complete guides 2019-06-07 20:46:44 ganglion: if you do 'apk info -L mdocml' you will see 'usr/bin/man' 2019-06-07 20:46:48 which is sad to me, because a complete guide should arguably be called a manual 2019-06-07 20:46:48 but then you could always say that the root of manual is "hand-ual" and therefore more akin to "handbook" anyway 2019-06-07 20:46:51 lmao 2019-06-07 20:46:53 interesting. 2019-06-07 20:47:24 do i edit my default shell by editing the shadow file? 2019-06-07 20:47:49 shadow file only contains password hashes 2019-06-07 20:48:01 yep, wrong one. i'm trying to recall, now. 2019-06-07 20:48:10 passwd. 2019-06-07 20:48:18 for that there's either chsh or manually editing /etc/passwd 2019-06-07 20:48:46 <_ikke_> chsh is not installed by default 2019-06-07 20:50:08 got it changed over. so, now, my curiosity becomes what i can easily do in order to not interrupt any of my programs running in the terminal, but log out and log back in 2019-06-07 20:50:13 is that what people use nohup for? 2019-06-07 20:50:39 <_ikke_> either that, or a multiplexer like screen / tmux 2019-06-07 20:50:57 ah, so tmux runs outside of my user? or how do linux processes care about the user running them? 2019-06-07 20:51:15 i see that my username's running tmux as per ps aux 2019-06-07 20:51:20 <_ikke_> it does run as your user, it just makes sure it's not tied to the current shell session 2019-06-07 20:51:31 that's a bit confusing 2019-06-07 20:51:34 <_ikke_> tmux starts a server in the background 2019-06-07 20:51:48 it's like a terminal you can connect to and disconnect from at will 2019-06-07 20:52:02 danieli: i'm interested more in the deeper aspect of it 2019-06-07 20:52:12 perhaps you'd care to elaborate then? 2019-06-07 20:52:16 <_ikke_> ganglion: it does a double fork 2019-06-07 20:52:39 <_ikke_> https://stackoverflow.com/a/10932710 2019-06-07 20:53:58 <_ikke_> https://stackoverflow.com/questions/881388/what-is-the-reason-for-performing-a-double-fork-when-creating-a-daemon 2019-06-07 20:54:17 right, the technical aspect 2019-06-07 20:55:03 <_ikke_> that way, the server is no longer a child of the shell 2019-06-07 20:55:07 now i'm trying to figure out how to copy from tmux 2019-06-07 20:55:09 lmao 2019-06-07 20:56:49 ganglion: https://unix.stackexchange.com/questions/58763/copy-text-from-one-tmux-pane-to-another-using-vim 2019-06-07 20:56:53 i worked around it by just remembering the important part of the url 2019-06-07 20:57:05 tcely: i gotta copy it to firefox, though. i'll read that link 2019-06-07 20:58:10 <_ikke_> ganglion: if tmux is running remotely, it cannot copy to your local clipboard for you without doing some trick 2019-06-07 20:58:26 <_ikke_> You'd need to copy from your terminal 2019-06-07 20:58:49 <_ikke_> I use this setting 'bind-key -T copy-mode-vi y send-keys -X copy-pipe-and-cancel 'xsel -i -p -b' 2019-06-07 20:58:59 <_ikke_> to be able to copy locally with tmux 2019-06-07 20:59:16 _ikke_: you could paste into tpaste or another such site ;-) 2019-06-07 20:59:49 <_ikke_> right, but that's not into your local clibpoard 2019-06-07 21:03:05 that's interesting. why does my terminal application not have the ability to easily hoist text out of its contents? 2019-06-07 21:03:24 <_ikke_> sorry, what do you mean with that? 2019-06-07 21:03:54 It DOES! That's wild. 2019-06-07 21:04:23 ganglion: Copying with your terminal app gets harder when you split tmux into multiple areas. 2019-06-07 21:04:26 _ikke_: I can just use ctrl+shift+c to use st to copy the text out. I see, I see. Obviously, if the text isn't ON the screen at the moment, that'd be where tmux makes it harder. 2019-06-07 21:04:51 tcely: i assume when i split things in the vertical axis, then i'm going to get worse copying 2019-06-07 21:04:59 <_ikke_> with ctrl+b [ you go into copy mode 2019-06-07 21:05:37 <_ikke_> with ctrl+b z, you would 'zoom' into a pane to make if full screen 2019-06-07 21:05:48 <_ikke_> (I do everything in tmux) 2019-06-07 21:05:48 that's neat. 2019-06-07 21:05:50 ^^ that's a good trick 2019-06-07 21:06:00 tmux is nuts. thanks for the help, _ikke_ 2019-06-07 21:06:05 and tcely, and everyone. 2019-06-07 21:07:01 do i need to log out in order to have my next st instance use bash instead of ash? 2019-06-07 21:07:02 ACTION has to dig up the tmux config from my previous work laptop at some point 2019-06-07 21:07:15 <_ikke_> ganglion: yes 2019-06-07 21:07:20 alright, brb 2019-06-07 21:08:12 haha, i win 2019-06-07 21:08:40 or do i? does that mean that tmux is running "behind" or "within" ash in its session? 2019-06-07 21:08:54 <_ikke_> You have two components 2019-06-07 21:09:04 <_ikke_> the server is running behind the shell (ash / bash) 2019-06-07 21:09:14 <_ikke_> the client is running within the context of your shell 2019-06-07 21:09:52 ohhh. okay. so this server was launched from ash, and then i switched to bash, but now my client's on bash? so what does that eventually mean for tmux's server? 2019-06-07 21:10:02 <_ikke_> nothing 2019-06-07 21:10:13 i can't even remember what i just barely switched to bash for. neat 2019-06-07 21:10:15 <_ikke_> though, the existing sessions still use your previous shell 2019-06-07 21:10:28 is there a way to throw them into a new shell? 2019-06-07 21:10:34 <_ikke_> tmux starts a new login shell for each pane 2019-06-07 21:10:34 just for curiosity's sake 2019-06-07 21:10:42 <_ikke_> exec bash 2019-06-07 21:10:50 <_ikke_> or exec bash -l 2019-06-07 21:11:56 ah, that's something i was doing. maybe it wasn't bettercap not liking bash, but not liking st. hmm. also, " and % _split_ a pane, right? so that's why i'm getting the same ash shell when i try to spawn a new one with " 2019-06-07 21:12:02 ?** 2019-06-07 21:15:20 I'm using tmux from first beta, can't imagine work without it 2019-06-07 21:16:05 <_ikke_> THe only thing i still need to find a proper solution for is how to combine remote and local tmux sessions 2019-06-07 21:16:28 mps: abduco + dvtm is much lighter and equally funcional ;P 2019-06-07 21:16:38 abduco? 2019-06-07 21:16:39 s/ncio/nctio/ 2019-06-07 21:16:45 _ikke_: that would be better than bread :) 2019-06-07 21:16:54 <_ikke_> mps: ha! 2019-06-07 21:17:05 _ikke_: could you just run a meta-tmux session 2019-06-07 21:17:22 despite me seeing somewhere ages ago that this isn't encouraged and disabled by default, iirc 2019-06-07 21:17:51 <_ikke_> tmux has some kind of control protocol that could be used for thi 2019-06-07 21:17:53 <_ikke_> this 2019-06-07 21:18:01 <_ikke_> but tmux cannot use i itself yet 2019-06-07 21:18:24 by the way, repasting an old t220 really, really improves temperatures 2019-06-07 21:18:28 it can't even overheat anymore 2019-06-07 21:18:51 it would be nice to copy on remote and paste on local, or even second remote 2019-06-07 21:19:09 <_ikke_> indeed 2019-06-07 21:19:32 seems like something demanded by everyone--why isn't it implemented by default? 2019-06-07 21:19:43 <_ikke_> ganglion: probably hard to get right 2019-06-07 21:19:47 makes sense 2019-06-07 21:20:23 _ikke_: say i have some nmap output i want to scroll up to read, but it's in a little baby tmux pane i've made 2019-06-07 21:20:45 what's the appropriate way to do it? i've abused meta-[ quite a lot for it, but i'd like to be able to use the pgup keys 2019-06-07 21:21:00 <_ikke_> well, combine those 2019-06-07 21:21:15 <_ikke_> ctrl+a [, then pg+up / pg+dwn 2019-06-07 21:21:25 ah, pageup works, i'm just using a t220 that has weird ass "next page" buttons down where i want the pageup/down keys to be. 2019-06-07 21:21:47 <_ikke_> yay 2019-06-07 21:21:56 that freezes output to that window, right? if i really wanted to shoot myself in the foot/be my own neckpain, is there a way to see a live output while still trying to copy? 2019-06-07 21:23:02 <_ikke_> not that I'm aware of 2019-06-07 21:23:52 alrighty. 2019-06-07 21:24:00 at what point do you start charging, _ikke_ 2019-06-07 21:26:58 <_ikke_> When you want personal dedicated help I guess 2019-06-07 21:29:31 the default tmux meta keybind is ctrl+b 2019-06-07 21:29:38 indeed 2019-06-07 21:29:47 do any of you change that default? 2019-06-07 21:29:53 many do 2019-06-07 21:29:54 <_ikke_> danieli: You are right, but I was used to screen's ctrl+a, so I changed that 2019-06-07 21:30:02 :) that's one common reason 2019-06-07 21:30:19 was ctrl-b easier on mit keyboards than on modern ones? 2019-06-07 21:31:07 also, damn, metasploit isn't a package? how will i ever be a skid 2019-06-07 21:31:40 ctrl-a are neighbor on all my (remaped) keyboards 2019-06-07 21:31:41 a curiosity of mine: how would i go about cleaning up a previously installed program that i had done "make install" to install if a package did indeed come out? 2019-06-07 21:31:57 mps: you map capslock to ctrl, then? 2019-06-07 21:32:05 right 2019-06-07 21:32:09 i have tons of security tools in a private aports repo (and mirror built from it) 2019-06-07 21:32:11 <_ikke_> ganglion: hope that make uninstall is defined 2019-06-07 21:32:22 it'll take a lot of effort to get it into upstream alpine 2019-06-07 21:32:35 where does make uninstall usually end up? in the source files? 2019-06-07 21:32:36 but i decided to just stay on arch, debian, and fedora instead 2019-06-07 21:32:51 <_ikke_> ganglion: yes 2019-06-07 21:33:12 ah, so if i delete those after installation, whoops. 2019-06-07 21:33:28 <_ikke_> You can try to force install the new package 2019-06-07 21:33:51 some sources have 'make uninstall' 2019-06-07 21:34:31 danieli: for security tooling, you mean? or do you just not use alpine 2019-06-07 21:34:49 in general on my daily driver machines 2019-06-07 21:35:09 interesting 2019-06-07 21:35:16 what are you using alpine for, usually? 2019-06-07 21:35:24 lots of different stuff 2019-06-07 21:35:31 also, what makes the difference between whether you install arch, debian, and fedora for a particulra workstation? 2019-06-07 21:37:07 that's nobody else's business - stick to what you know and like 2019-06-07 21:37:45 sticking to what i know would mean sticking to nothing. 2019-06-07 21:38:03 corollary: you can't like what you don't already know. 2019-06-07 21:38:42 then throw a dart at a map and go with that 2019-06-07 21:39:26 makes sense 2019-06-07 21:39:46 i'd recommend ubuntu for a beginner 2019-06-07 21:39:49 though i'd rather be able to use the knowledge and experience of others to get ahead, to not stumble through the same mire 2019-06-07 21:40:20 <_ikke_> Using Archlinux did learn me a lot about linux though 2019-06-07 21:41:24 why so? 2019-06-07 21:41:30 Gentoo and Exherbo also taught me a lot (more than Arch for me personally) 2019-06-07 21:41:41 Well, especially the building part has been forced down my throat by them :) 2019-06-07 21:41:48 <_ikke_> ganglion: because archlinu does very little out of the box 2019-06-07 21:41:54 i grew up with bsd and linux, my first linux distro was slackware 2019-06-07 21:42:06 <_ikke_> ganglion: and the wiki is quite good 2019-06-07 21:42:14 it's nice to get the basics in through ubuntu because it 'just works' and includes commonly used software 2019-06-07 21:42:22 arch doesn't hold your hand 2019-06-07 21:42:28 <_ikke_> correct 2019-06-07 21:47:45 shoutout to crux too for another great distro w/o hand holding 2019-06-07 21:49:16 the concept of a distro not holding your hand is a little weird to me--the documentation still exists, right, so it's more like holding your hand than anything else, right? 2019-06-07 21:49:58 people seem to look at the works of other human beings as being detached from their progenitor, but what the progenitors of linux have in mind, almost always, is a human being using the software they create, whether it's themselves or another 2019-06-07 21:50:15 ubuntu comes with an installer where you can click a button and look at a pretty screen while it installs itself for you 2019-06-07 21:50:32 with arch, it just boots into a live environment and you have to manually partition the disk and install it 2019-06-07 21:50:41 gentoo et al. takes it a step further, you compile everything yourself in gentoo 2019-06-07 21:50:42 danieli: but can you still configure everything necessary? that doesn't sound like it's teaching you anything. 2019-06-07 21:50:49 yes, but it's not done automagically 2019-06-07 21:50:51 ganglion: yeah it's the difference between needing to read the docs and understand them to not 2019-06-07 21:51:00 it sounds more so like you end up using other people, whether live or in the form of text they already wrote, to hold your hand for you. 2019-06-07 21:51:18 i disagree 2019-06-07 21:51:38 <_ikke_> you need to show a lot more initiative 2019-06-07 21:51:40 well, a similar arguement is that you rely on others for the code they wrote 2019-06-07 21:52:00 _ikke_: you as in me, or you as in the royal you in the context of arch vs ubuntu? 2019-06-07 21:52:10 xcko: absolutely. there's always reliance. 2019-06-07 21:52:21 <_ikke_> you as the user 2019-06-07 21:52:44 the fact that there's code to write, that there's a processor to write code for, that there's a goddamned compuper--that's all a reliance that involves a hand being held. the fact that we speak fucking english is a testament to this. 2019-06-07 21:53:31 that's way out of the context of what we meant 2019-06-07 21:53:40 i'm not critical of the idea of learning, of being initiated: i'm critical of the absolute egotism involved in claiming that someone is self-taught, for example, which usually follows the expression of "doing it yourself," which you're never actually doing 2019-06-07 21:53:55 i understand, danieli, but it'd be nice if other people ever felt the same way when i expressed this. 2019-06-07 21:54:36 i understand what you mean, but i don't really have a response - it got a bit too philosophical for me 2019-06-07 21:54:39 ganglion: do you have another term besides "hand holding" that we used for differntiating distros? 2019-06-07 21:54:54 unfortunately, it's basically baked-in to the experience of doing something without automation: this search for the INDIVIDUAL to dominate, despite this requiring the delusion of the individual superimposing its will...in the field grown by its collective 2019-06-07 21:55:07 xcko: i don't see it as really relevant. you can still configure ubuntu as terribly as you want. 2019-06-07 21:55:36 <_ikke_> but a lot more has already been done for you out of the box 2019-06-07 21:55:57 i think this is another irritating aspect: it would be cool if you were able to use multiple tools to get the end result that you want, but you really can't. there's a commonly accepted convention for almost every method of setting up a system for a specific purpose, and it's usually not included in the docs. 2019-06-07 21:56:01 the point was that you learn a lot from having to do things yourself rather than having for instance an installer do it for you 2019-06-07 21:56:31 _ikke_: indeed, but compared to the whole mountain of "fabbing a processor, creating the market system, establishing english, compiling the distribution..." it's really a tiny drop in the bucket. 2019-06-07 21:56:51 danieli: but do you really? I usually understand what an installer is doing by reading the output of the commands it's automatically running. 2019-06-07 21:57:00 many don't 2019-06-07 21:57:05 i think it's a matter of how much information is hidden from the user behind that pretty screen, not a matter of it doing it for you. 2019-06-07 21:57:13 ubuntu's doesn't by default 2019-06-07 21:57:29 they show a pretty screen or let you enter some usernames and passwords while it's working in the background 2019-06-07 21:58:06 it'd be neat if they just worked the more technical stuff into the side, because as far as i remember with most ubuntu setups i've done, they don't ever try to bridge gui and cli in a meaningful way 2019-06-07 21:58:15 many users just don't care about the commands or learning, they just want a working system 2019-06-07 21:58:33 they don't parse output from commands that they're running behind buttons, for example, so every time you hit "more details" you just get a deluge of terminal output 2019-06-07 21:59:05 danieli: i wouldn't care about the commands if they were as obtuse when i was trying to get something done, either. most of those users, however, will absorb information in the right context. 2019-06-07 21:59:17 they're not static, after all. humanity would be rather boring if they were 2019-06-07 22:01:43 be back soon; thanks for your help, everyone. 2019-06-07 23:29:57 I'm getting linker errors about getcontext() and swapcontext but #include is there, are there extra linker flags I need? 2019-06-07 23:36:19 I have musl-dev installed which owns the /usr/include/ucontext.h header 2019-06-07 23:38:41 maybe libucontext hmmm 2019-06-08 10:31:21 musl doesn't have netax25, ax25 library. 2019-06-08 12:40:25 hi! 2019-06-08 12:40:45 Hi 2019-06-08 12:41:15 i'm trying to "apk add wireguard-tools@edge" after already having added linux-virt@edge and rebooted, but the package doesn't exist. am i doing something wrong or looking in the wrong repo? 2019-06-08 12:44:03 wireguard-tools is in edge/community, not edge/main 2019-06-08 12:44:23 it also (in general) isn't a good idea to mix stable and edge repositories unless you can unbreak it when it breaks 2019-06-08 12:45:18 danieli: that VM is only going to run wireguard and nothing else. ever. there's not even space left for anything else. if it breaks i guess i'd just shrug and reinstall. 2019-06-08 12:45:30 fair enough 2019-06-08 12:45:36 it was just a heads up :) 2019-06-08 12:45:53 thank you :) 2019-06-08 12:47:16 danieli: it installed fine now! one thing i'm wondering though: https://wiki.alpinelinux.org/wiki/Configure_a_Wireguard_interface_(wg) says to install "wireguard-vanilla" or "wireguard-virt" and https://www.wireguard.com/install/ says to install "wireguard-tools" (only). is either one of these outdated? 2019-06-08 12:48:03 it's tools to configure an interface you don't have are kinda useless, no? 2019-06-08 12:48:32 sure, but it didn't pull in either of the kernel-specific packages. 2019-06-08 12:49:11 i guess some superusers could need the tools standalone, but i don't see most people needing tools without the module 2019-06-08 12:49:30 hmm in that case https://www.wireguard.com/install/ is simply incomplete 2019-06-08 12:49:46 (especially since it specifically says "module & tools") 2019-06-08 12:49:52 something might have changed on our side, there were a lot of changes to wireguard recently 2019-06-08 12:50:06 especially in edge, which is to be expected 2019-06-08 12:50:25 okay 2019-06-08 12:50:27 ACTION chills 2019-06-08 12:52:06 ryouba: I have custom made upstream kernel with wireguard module, and I don't like wireguard-tools pulls alpine kernel 2019-06-08 12:52:31 good point mps 2019-06-08 12:53:13 Alpine tries to be simple and use deps only when it really necessary 2019-06-08 12:55:17 well, we do have some massive packages, but they're split up into subpackages wherever it's feasible 2019-06-08 12:57:31 hmm, gimp having dbus in makedepends is mistake, imo 2019-06-08 13:02:54 It's only going to become more dependent on it, GIMP devs apparently plan on moving to D-Bus for plugin IPC 2019-06-08 14:08:54 mps: when you add "DNS" to the [Interface] section of a WireGuard config file, does that config still start for you? 2019-06-08 14:09:44 ryouba: didn't tried wg with dns, so don't know 2019-06-08 14:10:18 hmm ... i guess i should first fix the problem of not being able to ping each other 2019-06-08 14:11:02 wireguard doesn't have a problem listening and connection on/to port 443, does it? 2019-06-08 14:16:11 if you run it as root or with cap_net_bind_service=+ep, sure 2019-06-08 15:04:55 So, I've been having a fun time trying to figure out a sane way to get teamviewer working on Alpine (Specifically, a headless one). I'm using a small box running alpine as a nas for my mom's work, and I want some way to fix small problems without having to worry about port forwarding. 2019-06-08 15:05:35 teamviewer is proprietary software though? and as far as I remember it's just a wine wrapper on linux 2019-06-08 15:05:47 I know I can get it working somewhat in a chroot, but I haven't really messed with making it a service (or rather I have no idea how well that would work). 2019-06-08 15:05:55 if they do have binaries available for linux, i'm 95% sure it won't work with musl libc 2019-06-08 15:06:01 <_ikke_> How would teamviewer work in a headless setting? 2019-06-08 15:06:08 it uses the framebuffer console 2019-06-08 15:06:22 log in, it just shows you the console as if you plugged in a monitor 2019-06-08 15:06:25 <_ikke_> It already requires wine to work 2019-06-08 15:07:07 as far as I can tell they changed that: the only stuff it seems to require now is dbus, glibc, and some qt5 2019-06-08 15:07:19 glibc being the problem 2019-06-08 15:07:21 it's possible to run wine apps headless by essentially piping the X stuff into a fake framebuffer 2019-06-08 15:07:40 <_ikke_> ShaRose: What I ran into with Archlinux for newer versions is that it requires a DM to work 2019-06-08 15:07:44 er, fake X server 2019-06-08 15:07:52 it does graphical operations but doesn't give any graphical output 2019-06-08 15:09:30 I was able to get it to install by adding bash coreutils file xz dbus-libs qt5-qtdeclarative and https://github.com/sgerrand/alpine-pkg-glibc/releases/download/2.29-r0/glibc-2.29-r0.apk 2019-06-08 15:09:36 but teamviewer setup segfaults 2019-06-08 15:09:44 daemon apparently runs though, so there's that 2019-06-08 15:12:15 if nobody has any ideas, I know it works 100% with a debian chroot (it even shows the alpine console properly if you bind dev!) but I'm not sure how I'd add that as a service 2019-06-08 15:27:57 ShaRose: Wouldn't SSH with reverse port forwarding work and be way easier at that? 2019-06-08 15:34:15 Cogitri not my router: and the local ISP just loves resetting the config whenever there's a ghost of an issue 2019-06-08 15:34:45 i've used software like autossh for that before 2019-06-08 15:35:34 I've never heard of that, but I'll look into it 2019-06-08 15:37:13 wow that sounds so much nicer. (Man I wish I heard of that when I was asking around for options..) 2019-06-08 15:37:36 I even specifically asked for reverse shells (not here clearly) but nobody had any idea. 2019-06-08 15:39:54 the whole idea of teamviewer et al. is that they do NAT hole punching - you can connect to the machine behind a stateful firewall because you're piggybacking on the connection that was made out to the internet, not vice versa 2019-06-08 15:40:47 oh, I know: I even knew about reverse shells, but wasn't aware of any software to do it that wasn't a hack designed to be hard to notice (like using ncat and such) 2019-06-08 15:42:05 I use reverse shells a lot, but mostly not for ergonomic or practical reasons.. :) 2019-06-08 15:42:19 more so for professional and hobby security work 2019-06-08 15:43:03 mmm. In my case when they reset the router, I was 4 hours away: they couldn't really work without access to the nas, so I had to pack up and leave in like 30 minutes including a shower :P 2019-06-08 15:43:06 it wasn't fun 2019-06-08 15:44:19 bonus points because I was only there because the night before my sister's internet went out (the contractors rebuilding a fence noticed a coax cable was in the way and cut it without telling anyone) and then when she rebooted her phone to fix usb tethering to her computer bell had pushed an update blocking magisk, so her phone locked up 2019-06-08 15:44:28 was an excellent shitstorm 2019-06-08 15:44:53 every time i hear war stories like those, i feel spoiled for having a good internet connection and home network 2019-06-08 15:45:25 I've been on oil platforms though, the 4G (if it's close enough to mainland norway you can even get it) is crap 2019-06-08 15:56:37 danieli the business in question uses a modem so shit it doesn't have a way to list devices that it gave DHCP leases to 2019-06-08 15:56:39 like at all 2019-06-08 15:56:53 it probably does but just doesn't expose it 2019-06-08 15:57:03 if it hasn't contacted the router to get in it's arp table (which it DOES list oddly enough) there's no way to see if it exists 2019-06-08 15:57:44 there isn't a secret tech mode that lists it either, my dad WORKS for the ISP and he couldn't find it either 2019-06-08 15:58:19 he could have logged in but I forgot that on that box it's just root: I thought I set up a standard user for local access but nope 2019-06-08 15:58:51 either way, autossh looks like it'll work great, so I'll go set that up now :P 2019-06-08 16:24:55 the manufacturer might just have something, but it's not something you'll get into super easily 2019-06-08 17:33:47 shame. seems like wireguard is incapable of punching through the same firewall openvpn punches through just fine running on port 443. 2019-06-08 19:27:33 what's the meaning of this? https://gist.github.com/sixtyfive/5564c3063c451466eab775bc19caab10 ... seemed like such an innocent command :-/ 2019-06-08 19:31:41 > Can't open PFXFILE for reading, No such file or directory 2019-06-08 19:32:00 Sounds pretty clear what openssl complains about 2019-06-08 19:57:18 Cogitri: oh! took a while but it seems i need a thing called ACF and that'll generate such a file. 2019-06-08 19:59:54 oh wow this is neat 2019-06-08 20:17:53 am looking for php-5xcache as listed in https://wiki.alpinelinux.org/wiki/Nginx_with_PHP#Nginx_with_PHP5 can someone tell me best way to get it? 2019-06-08 20:20:12 we don't have php5 anymore 2019-06-08 20:20:23 ah 2019-06-08 20:20:27 it's no longer supported by php either 2019-06-08 20:20:34 end of life for 5.6 was in december 2019-06-08 20:20:46 i strongly recommend you upgrade whenever possiblew 2019-06-08 20:24:56 ah how do I tell if a packaege is installed? 2019-06-08 20:34:29 I ma trying to remve all the php5 packages .. I did "apk add package-name package-name " and the installed apk del with all the packages on one line does no seem to work tho 2019-06-08 20:36:52 hi. what is the name of screen keyboard package? 2019-06-08 20:37:47 For GNOME it's Caribou 2019-06-08 20:37:57 for Mate? 2019-06-08 20:39:39 last time i knew, mate didn't have one 2019-06-08 20:40:20 don't take my word for it though 2019-06-08 20:42:54 actually i am coming from pmOS. i installed postmarketos-ui-mate and it doesn't have embedded keyboard stuff. when phone is locked i can't unlock that. It does have Mate Keyboard but it is not opening automatically. I am bit confused. 2019-06-08 20:42:54 I think Caribou should work for Mate too, but don't quote me on that 2019-06-08 20:44:00 It's best to ask that in a pmOS related channel, I guesd 2019-06-08 20:44:34 This here is more about Alpine desktop/server I suppose 2019-06-08 20:45:31 yes i know. i just wanted to ask. 2019-06-08 20:45:46 sorry for the noise 2019-06-08 20:45:59 to be fair, it was a fairly generic question, and not pmOS-specific 2019-06-08 20:46:13 but yeah, we have no clue about postmarket's own packages 2019-06-08 20:50:18 Yup, I don't mind the fairly generic questions, but once it started with pmOS specific stuff it got a little off track (and the chances of someone helping him in a pmOS related channel are way higher I imagine) 2019-06-08 21:44:48 hmm, somebody could point me out from where came last two lines in that config? (I never added these 2 last at all, it happened on edge and on 3.9 is fine): http://sprunge.us/HGkChy 2019-06-08 21:50:48 which exact file is that? 2019-06-08 21:50:52 absolute path 2019-06-08 21:51:14 "/etc/default/grub" 2019-06-08 21:58:08 danieli, ah ye I see from where: https://git.alpinelinux.org/aports/tree/main/grub/grub.post-upgrade 2019-06-08 21:58:26 oh sorry, i was a little preoccupied 2019-06-08 21:58:31 yes, that's what i was thinking too 2019-06-08 21:58:31 :) 2019-06-08 21:59:15 still weird that it just added at all and I dont see any reason to even use GRUB_CMDLINE_LINUX_DEFAULT 2019-06-08 22:02:40 there is "-z" if string is null but it doesnt check if string exist at all and after doing some crazy additions :P that file should be untouchable imho or I should just add that _LINUX_DEFAULT and not bother :) 2019-06-08 22:51:01 check if variable exist at all* 2019-06-08 23:42:47 How do I set up locales on Alpine? 2019-06-08 23:42:57 I've tried searching, but I get... a lot of different answers, none of which make a lot of sense. 2019-06-08 23:46:08 koz_: musl doesn't support locales, yet 2019-06-08 23:46:21 mps: So I gotta do everything in a C locale? 2019-06-08 23:46:36 well, mostly 2019-06-08 23:47:27 about year ago I built coreutils with locale support and it worked for some things but not for all 2019-06-08 23:48:41 Alrighty, I'll give it a whirl and see. 2019-06-09 00:06:31 Pretty sure C.UTF-8 actually is the default for musl 2019-06-09 00:06:44 But we don't set LANG (for some reason?) so some applications don't recognize that 2019-06-09 00:08:17 Cogitri: yes, C.UTF-8, I'm tired and didn't looked carrefully at question 2019-06-09 00:09:06 fixing patches, and answering at late night doesn't go well together 2019-06-09 00:09:38 :> 2019-06-09 00:14:17 Me doing anything complex at late night doesn't go well :P 2019-06-09 00:14:52 I'm bootstrapping some ada stuff on alpine 2019-06-09 00:15:18 ada is still used? 2019-06-09 00:15:54 you'd be surprised 2019-06-09 00:16:00 trying to fix something few hours in a row at night is very useless when can go sleep and after wake up fix it in just a minute :D 2019-06-09 00:16:29 well, no, that was rhetorical question :) 2019-06-09 00:17:48 few years ago I meet a man who was to much occupied with COBOL programming 2019-06-09 00:19:38 too* never will learn difference between 'to' and 'too' 2019-06-09 00:23:30 that is why I love irc... always can learn something new... COBOL, thanks mps :D 2019-06-09 00:25:14 MY_R: I will follow your advice and try to sleep. good night 2019-06-09 00:26:35 ye, good idea! too many times I wasted my time cus I was too tired :) 2019-06-09 00:27:38 sleep well mps! 2019-06-09 00:31:43 thanks, will try :) 2019-06-09 00:34:08 "Stop trying to hit me and hit me!" :D 2019-06-09 00:38:37 When running alpine in data mode, are you able to install software? 2019-06-09 00:38:47 or does that need to be done everytime? 2019-06-09 00:39:36 I was running on diskless mode I believe. rebooting my vm after a few days and just noticed the different modes. Was wondering how it partitioned my disk, but didnt bother to check 2019-06-09 01:11:51 hi all. is there any way to unlock pc from Terminal with SSH? P.S: Pc locked with a password. 2019-06-09 01:18:01 revolutionary: what's locking it? couldn't you just end the process? 2019-06-09 01:18:08 or does the locking process do something special, here 2019-06-09 01:18:22 You have root over ssh, revolutionary? 2019-06-09 01:19:10 yes priontology 2019-06-09 01:19:20 what process is doing the locking? 2019-06-09 01:19:39 i mean lockscreen 2019-06-09 01:36:44 revolutionary: right--what are you using for your lockscreen? 2019-06-09 01:43:00 default alpine stuff priontology. i am using Mate desktop environment 2019-06-09 01:59:03 revolutionary: hmm. mine didn't come with a lockscreen. i'm unsure. 2019-06-09 01:59:45 hmm. thanks. 2019-06-09 02:27:55 revolutionary: why are you trying to unlock it? 2019-06-09 02:28:24 there is no specific reason. 2019-06-09 02:28:50 i'd say it depends on what desktop environment/OS it is and the configuration 2019-06-09 02:29:23 mate 2019-06-09 09:48:53 danieli: Have you ever used metasploit pro? 2019-06-09 11:00:39 there is a pro version of metasploit? 2019-06-09 12:12:00 Does rc-service no longer exist? https://wiki.alpinelinux.org/wiki/Alpine_Linux:FAQ#My_cron_jobs_don.27t_run.3F 2019-06-09 12:14:10 `apk search cmd:rc-service` suggests it's still a thing 2019-06-09 12:51:11 @Cogitri ah, it needs to be installed. Am looking for a way to add a cron in my alpine. I was thinking of creating a file in /etc/periodic/daily but I keep reading how people touch a file, run crond on it and then delete it 2019-06-09 12:52:10 You can use `crontab -e` if you don't want to deal with files 2019-06-09 12:52:58 Cogitri: I'm trying to script it :) 2019-06-09 12:53:32 scripting crontab -e seems like more trouble 2019-06-09 12:54:15 Ah :) 2019-06-09 12:55:55 Cogitri: but do you know if echo "certbot renew" > /etc/periodic/daily/certbot.cron would do the trick? 2019-06-09 12:56:39 I'm doing a echo "certbot renew" > crontab.tmp && crontab crontab.tmp && rm crontab.tmp 2019-06-09 12:56:51 which seems so ugly 2019-06-09 13:04:26 I've always used `crontab`, so dunno, sorry 2019-06-09 13:04:36 And acme.sh installed itself automatically 2019-06-09 17:43:17 Cogitri: yeah. I'm playing with too many technologies at the same time 2019-06-09 17:43:27 dockers + capistrano + alpine all at once 2019-06-09 18:09:07 priontology: I might have why are you wondering? 2019-06-09 19:02:40 Anyone here.....? 2019-06-09 19:02:58 maybe 2019-06-09 19:03:15 Maaaaaybe 2019-06-09 19:03:21 wow 2019-06-09 19:04:05 Well I am using Android on desktop and sometimes I do network tethering 2019-06-09 19:04:51 When I connect my Android during boot... udhcpc detects it and assign ip address etc... 2019-06-09 19:05:46 But if I hotplug it... I dont see the device working 2019-06-09 19:05:51 Why so? 2019-06-09 19:13:18 "android on desktop"? 2019-06-09 19:14:03 I mean Alpine xD 2019-06-09 19:14:12 autocorrect abusement 2019-06-09 19:14:31 i'd check the syslog to see how the kernel treats the phone 2019-06-09 19:14:33 Ah, that makes more sense :P 2019-06-09 19:14:36 ^ 2019-06-09 19:15:01 Understood... will check tomorrow. 2019-06-09 19:53:34 <_ikke_> SpaceToast: with salt on alpine, do you get this error? "Error loading grains, unexpected linux_gpu_data output, check that you have a valid shell configured" 2019-06-09 19:53:49 <_ikke_> (well, it's more like a warning) 2019-06-09 20:22:19 _ikke_: yes, it's a warning 2019-06-09 20:22:24 you can install pciutils to get rid of it 2019-06-09 20:25:35 <_ikke_> SpaceToast: alright, thanks 2019-06-09 21:56:45 Hey, how is NVIDIA support with Alpine? 2019-06-09 21:57:38 Well, you can't use the proprietary driver, I suppose 2019-06-09 21:58:07 That’s what I suspected. 2019-06-09 21:58:19 So either use cards that are of the GTX 9xx series (without clocking support) or something older 2019-06-09 21:58:24 Or just buy AMD :P 2019-06-09 21:59:27 My Boss bought me an Dell Precision Workstation Laptop with a Xeon and a Quadro... not much to change here. 2019-06-09 21:59:44 Oof 2019-06-09 22:01:54 :/ 2019-06-09 22:02:24 What was the reason to switch back to OpenSSL? 2019-06-09 22:03:40 LibreSSL having some shortcomings on Linux (e.g. no support for certs with validity until >=2038 on 32-bit arches) and patching for it being annoying 2019-06-09 22:03:57 And OpenSSL has become a lot better in recent days 2019-06-09 22:05:38 Sounds good 2019-06-09 22:08:31 also needed 1.1.1 for some software requiring it for tls 1.3, such as newer python 2019-06-09 22:08:38 not sure if libressl has even caught up yet 2019-06-09 22:10:25 Not really, at least Nextcloudclient still needs OpenSSL 2019-06-09 22:11:14 I was about to install NextCloud, at the end we decided to use the Synology stuff. 2019-06-09 22:11:56 Nextcloud is pretty neat, I've been using it ever since it's a thing 2019-06-09 22:13:55 Don’t try to put a index.html/php in the cloud synched directories... server hangs. At least on the raspberry pie a friend of me says. 2019-06-09 22:14:09 that was last week. 2019-06-09 22:28:17 danieli: I was wondering if there was an easy way to get it without attaching my student identity to it 2019-06-09 22:28:36 i'm not a student, so i have no clue how it would work for students 2019-06-09 22:36:05 Ah. I'll change my question, then. Does it attach anything to your identity, danieli? 2019-06-09 22:37:33 yes. 2019-06-09 22:38:42 i know some people though, so i managed to avoid a few of the hoops - but the general answer is yes 2019-06-09 22:38:46 Where at? That's hyperunfortunate. 2019-06-10 11:55:26 Hello good people. I am testing Alpine on my machine, I want to use it for my personal work and so far I enjoy it for the simplicity. One point that I am curious about is that randomly it becomes slow with some simple terminal commands like `ls`. I am using DWM and when it boots the terminal, I type my user and [enter] and it shows the password field only about 2 seconds after I pressed enter. 2019-06-10 11:55:46 In the terminal (st) it is also randomly slow with commands. 2019-06-10 11:57:14 gerep: on slow GPU (without driver) st can be very slow 2019-06-10 11:57:26 urxvt is a lot faster 2019-06-10 11:58:06 mps: but the weird part is that it is random, sometimes the `ls` is "normal" but sometimes it is slow. 2019-06-10 11:58:33 mps: I will get the information about the laptop and share here. 2019-06-10 11:58:52 disk powersaving does that 2019-06-10 11:59:39 dmesg (especially long lasting) is good test 2019-06-10 12:05:39 TBB: this option is "on" even when I am not running it on my battery? 2019-06-10 12:06:28 I just checked, it is a Pentium P6000 1.87GHz with 2.62Gb RAM. 2019-06-10 12:06:32 A good machine. 2019-06-10 12:06:47 2 cores. 2019-06-10 12:11:03 gerep, check "dmesg" and be sure that you using good GPU driver, on old integrated intels "modesetting" driver can work much better from "intel" one 2019-06-10 12:11:50 MY_R: I will. 2019-06-10 12:15:21 MY_R: lspci | grep ' VGA ' 2019-06-10 12:15:28 Is that a good way to get that information? 2019-06-10 12:15:54 gerep, yep 2019-06-10 12:16:38 MY_R: 00:02.0 VGA compatible controller: Intel Corporation Core Processor Integrated Graphics Controler (rev 02) 2019-06-10 12:17:03 well :D 2019-06-10 12:17:19 Is that good or bad? haha 2019-06-10 12:18:14 very generic 2019-06-10 12:18:22 hahaha 2019-06-10 12:18:43 I will look in the Wiki and see if I can install something more specific. 2019-06-10 12:22:39 grep dmesg for some gpu related stuff and be sure you have: mesa-dri-intel mesa-gles mesa-egl mesa-gl 2019-06-10 12:26:11 and if you have HDD then check SMART status, that usually (bad blocks) caused for me some weird random delays and of course look at: /var/log/Xorg.0.log 2019-06-10 13:58:57 MY_R: I have install the intel video following this: https://wiki.alpinelinux.org/wiki/Intel_Video I also have installed those mesa packages. The output of lspci is still returning the same output (I am not sure if it should be something different) 2019-06-10 13:59:40 MY_R: I will check for those, possibly, bad blocks, I am using HDD. What exactly should I look for on Xorg.0.log file? 2019-06-10 14:01:25 gerep, some of the wiki pages are outdated, like this one: This page was last edited on 19 September 2017, at 23:36. 2019-06-10 14:01:51 there is no need to add anything to modules file 2019-06-10 14:03:03 MY_R: can I remove those two modules from file? 2019-06-10 14:03:10 ye 2019-06-10 14:06:22 gerep, you can force xorg to use any driver you want just by creating file like this: /etc/X11/xorg.conf.d/20-gpu.conf and inside something like that: http://sprunge.us/bxOsGb 2019-06-10 14:07:37 gerep, in Xorg something with missing/failed words or just with prefix EE and WW for warnings 2019-06-10 14:12:25 MY_R: In the log file: (ww) Open ACPI failes (/var/run/acpid.socket) (No suck file or directory) and (ee) /usr/share/libinput: failed to find data files 2019-06-10 14:12:58 Zero warnings. 2019-06-10 14:13:19 acpi you can ignore and would be good to install libinput stuff 2019-06-10 14:13:26 About the badblocks, I do not have the executable and I cannot find it anywhere in the available packages. 2019-06-10 14:13:36 MY_R: I will. 2019-06-10 14:13:53 xf86-input-libinput libinput 2019-06-10 14:15:00 gerep: badblocks is in e2fsprogs-extra 2019-06-10 14:15:39 or just smartmontools and check by "smartctl" command 2019-06-10 14:23:51 MY_R: running smartctl on my /dev/sda3 returns no errors 2019-06-10 14:26:04 mps: in that case, how can I install it from e2fsprogs-extra? I have updated the /etc/apk/repositories file but that is probably not enough :) 2019-06-10 14:26:32 gerep: apk add e2fsprogs-extra 2019-06-10 14:27:05 did you run an actual scan, as in smartctl -t long /dev/sda ? 2019-06-10 14:27:07 mps: Thank you. Obvious haha I still need to read about `apk`. 2019-06-10 14:27:37 TBB: just "smarctl --all /dev/sda3" Will do as you suggested. 2019-06-10 14:27:44 apk is a lot simpler than many other package managers 2019-06-10 14:28:14 gerep: apk -U add runs the update command before add 2019-06-10 14:28:24 gerep: that's not the same at all, that one shows you the stats but if you haven't run a self test that report is not up to date 2019-06-10 14:29:02 TBB: Got it. Now it is running the long test, it asked me to wait for 79 minutes =) 2019-06-10 14:29:22 yeh, it's time consuming, and I suggest you wait for at least 90 if it says 79 :D 2019-06-10 14:31:56 heisenbug? 2019-06-10 14:32:06 oops, wrong channel, gdi 2019-06-10 14:32:46 TBB: I wlll =) 2019-06-10 14:33:21 about that possible powersaving option, hdparm would probably be the tool to tell you 2019-06-10 14:33:33 It is really weird this random slowness and with time it starts to bother haha 2019-06-10 14:33:42 TBB: Great, will check that one. 2019-06-10 14:33:56 TBB: for now it is best to wait for the test to finish? 2019-06-10 14:35:05 yup. SMART tests tend to get interrupted if you either run other smartctl commands or do something else low-level to your drive; normal filesystem use doesn't interrupt them though 2019-06-10 14:43:04 "hdparm -B255 /dev/sda" :D to avoid that crap disk power save, first command which doing before/after install... 2019-06-10 15:54:51 Hello! 2019-06-10 16:02:22 Can someone explain / redirect me to some resource on how to run alpine from grub completely from RAM using an initrd as main root-fs? this is an old argument (https://bugs.alpinelinux.org/issues/5384) but I think that after 3 years there are better solutions. Thank you. 2019-06-10 16:04:34 anyone know how to enable blinking cursor in st term, or patch for it 2019-06-10 16:32:09 TBB: came back from lunch and stopped the execution of smartctl, where do I get the results? 2019-06-10 16:55:19 mps: I think it has a configuration for that: https://git.suckless.org/st/file/config.def.h.html#l48 2019-06-10 16:55:25 I didn't try it though =) 2019-06-10 16:57:59 gerep: thanks, but this don't work for me 2019-06-10 16:59:25 xst (patched fork of st) have blinking cursor but will need some work to take it from xst and apply to st 2019-06-10 17:00:36 mps: yes :( tried it now and also didn't work, not even the thickness configuration 2019-06-10 18:49:51 Hi guys. Is there any API to know the latest release of a package? 2019-06-10 18:50:04 <_ikke_> franklinyu: not at this moment 2019-06-10 18:50:39 i suppose it would be possible to look at http://dl-cdn.alpinelinux.org/alpine/latest-stable/releases/x86_64/latest-releases.yaml if it's updated? 2019-06-10 18:50:40 franklinyu: repology.org could help to some degree 2019-06-10 18:50:56 OH, a package, nevermind! 2019-06-10 18:51:56 we don't expose any API for that, but i believe this parses output from pkgs.alpinelinux.org: https://github.com/genuinetools/apk-file 2019-06-10 18:55:37 _ikke_: Thanks, good to know. 2019-06-10 18:56:25 mps: Thanks! Repology seems interesting. I would look into its documentation to see how it works out. 2019-06-10 18:58:23 danieli: I'm still reading the repository, but how is this different from using APK directly? 2019-06-10 18:59:11 Or maybe this tool can run outside Alpine Linux, for example under Ubuntu? 2019-06-10 18:59:20 it's always up-to-date, doesn't require you to pull the apkindexes from a mirror (apk update) first, and it allows you to search packages for files inside them 2019-06-10 18:59:38 no idea, i'm assuming it's intended for alpine 2019-06-10 19:41:42 any gotchas i need to know about for a shared /usr mount? 2019-06-10 19:42:34 guessing i would just have to refrain from using apk on the systems mounting it readonly 2019-06-10 19:43:29 hm yeah i guess thats it; could be a good excuse to handle my configs in /etc more reasonably too 2019-06-10 20:30:48 danieli: Thank you! Good to know. 2019-06-11 00:43:51 hi all. i can't set screen resolution on Mate. Is there any additional package for that for install? 2019-06-11 00:45:20 for edge 2019-06-11 05:28:22 anyone know what happened to the sbcl packages for alpine? they were available a couple weeks ago but aren't now 2019-06-11 05:31:03 m18e: which architecture? 2019-06-11 05:31:09 x64 2019-06-11 05:31:13 i use sbcl but not on alpine, so i haven't paid attention 2019-06-11 05:31:14 but i'll check 2019-06-11 05:31:19 I was pulling it from edge/testing 2019-06-11 05:31:35 (I think) 2019-06-11 05:31:57 it was moved to unmaintained 2019-06-11 05:32:02 Date: Sun May 12 18:26:31 2019 -0300 2019-06-11 05:32:20 oh! 2019-06-11 05:32:25 you can build it yourself using this if you'd like: https://git.alpinelinux.org/aports/tree/unmaintained/sbcl 2019-06-11 05:32:40 its clisp dependency is unmaintained, so sbcl was moved there too, according to the git log 2019-06-11 05:32:56 that makes sense 2019-06-11 05:33:12 unfortunate, but makes sense 2019-06-11 05:33:25 not enough hands :( 2019-06-11 05:34:49 appreciate the archaeology though, glad to know I'm not just going crazy :) 2019-06-11 05:35:35 no worries :) nice to meet another lisp user 2019-06-11 05:35:55 well, clojure and emacs lisp are fairly common, but oh well 2019-06-11 05:37:26 I don't do as much lisp development as I used to, unfortunately it's just required as part of our build pipeline by a tool 2019-06-11 05:38:02 I play around with it and still mess around w/ my emacs config, but not too much production code being written in CL on my part 2019-06-11 05:38:14 mostly just LFE and elisp now 2019-06-11 05:39:03 and hm - you're welcome to maintain clisp + sbcl if you'd like, it doesn't look like it would take too much effort. another alternative would be to build it yourselves and host it on an internal mirror 2019-06-11 05:39:10 fwiw I found this project to build a docker image with sbcl: https://github.com/daewok/docker-sbcl which seems maintained 2019-06-11 05:40:02 looks like it, but alpine:3.8 is not current stable, 3.9 is (and soon 3.10) 2019-06-11 05:40:29 it's based on 3.9 2019-06-11 05:40:36 i didn't check the dockerfile, just the readme 2019-06-11 05:40:47 yeah, looks like the readme is wrong, the dockerfile is using 3.9 2019-06-11 05:40:59 welp, someone's getting a pull request for christmas 2019-06-11 05:42:06 all i want for christmas is a... pull request? 2019-06-11 06:26:37 fwiw that docker-sbcl image seems to have unblocked me, I know it's not alpine, but I'm running again :) 2019-06-11 06:26:45 thanks for all your help, danieli! 2019-06-11 06:36:52 wpa_supplicant and udhcpc seem to be starting before networking is even started. 2019-06-11 06:37:04 How do I change that with openrc? Why would openrc do this to me? ;-; 2019-06-11 06:37:19 I changed things to be launched in parallel, to my understanding. 2019-06-11 06:39:02 did you change the depend directive in either of those? 2019-06-11 06:41:52 and which runlevel are they set to run under? boot or default? 2019-06-11 06:42:09 those are the first things I can think of. 2019-06-11 06:52:59 zcrayfish: I'll check momentarily. What does the runlevel do to make a difference? 2019-06-11 06:53:33 well presumably the boot runlevel runs before the default runlevel, so that might have something to do with it. 2019-06-11 06:56:22 Ah, I see. So the system enters these states, boot, nonetwork, default, etc., etc., and when it enters these states, it triggers openrc to launch specific applications? 2019-06-11 06:56:37 With this in mind, what creates the entering and exiting of these states? What determines runlevel after booting? 2019-06-11 07:01:56 you can see a few of the triggers in /etc/inittab. aside from those, idk 2019-06-11 07:03:36 I think I need to set urandom earlier. 2019-06-11 07:06:01 runlevels have been a thing since probably the 80s 2019-06-11 07:06:21 It seems to be in boot by default... if you put it in sysinit, isn't there a chance /var/lib/misc/ won't be mounted by the time urandom runs? 2019-06-11 07:07:04 zcrayfish: I'm unsure. It seems that I have the bug where I need to move my mouse occasionally on boot because there's "not enuf endorpy" 2019-06-11 07:07:15 it's possible to specify dependencies in openrc service files 2019-06-11 07:07:30 priontology: if that really is the case, see if installing and running haveged helps it 2019-06-11 07:07:57 Does haveged have to run every time on boot? 2019-06-11 07:08:00 I only ran into that once or twice on a very freshly installed system... presumably because the random seed had yet to be saved. Fortunately. 2019-06-11 07:09:07 Right now, I'm essentially trying to make boot time as fast as possible while still being able to boot what I need. Maybe I should keep wireless services off (and the device itself turned off) until I actually need it, like when I click a widget for connection, or something like that. 2019-06-11 07:09:12 How do you guys do it? 2019-06-11 07:10:21 I noticed that udhcpc is running on boot and sending discover over an interface that literally has no connection (eth0). What's up with that? I assume it's just running in general as part of networking? Why not just make it run only if there has been a detection that the ethernet cable has been plugged in, etc? 2019-06-11 07:11:52 Also, for some reason, whenever I run wpa_cli after wpa_supplicant has been run as a background service, wpa_cli just can't connect to wpa_supplicant. I haven't been able to use wpa_cli, and instead have had to deal with no real feedback from wpa_supplicant (dmesg, I guess, is some sort of feedback) 2019-06-11 07:17:54 priontology: alpine is not user friendly distro, you have to set configs, fix setup, i.e. make your fingers 'dirty' 2019-06-11 07:18:30 you have to setup wpa_sp config file with proper permissions and path 2019-06-11 07:19:51 For wpa_cli to connect to it? 2019-06-11 07:19:53 Take a look at the ctrl_interface directive in wpa_supplicant.conf; and make sure wpa_cli knows where and can read both your wpa_supplicant.conf and whatever is in the ctrl_interface directive. 2019-06-11 07:20:56 Ah, also, if I'm using st from an installed package, how do I patch it? Was it you that was using st, mps? 2019-06-11 07:21:02 zcrayfish: right, that parameters need to be set properly 2019-06-11 07:22:17 Inside of wpa_supplicant.conf, I only have network={...} "directives." 2019-06-11 07:22:45 priontology: yes, I made my private patched st apk, didn't put it in aports because I tweaked it for keyboards without keypads, no insert, home, end pgup/pgdown keys 2019-06-11 07:22:59 Interesting, mps 2019-06-11 07:23:15 What's the process to do that? Also, where should I normally set ctrl_interface? 2019-06-11 07:23:37 priontology: here is first line from my wpa_sp.conf: ctrl_interface=DIR=/run/wpa_supplicant GROUP=netdev 2019-06-11 07:24:07 you have to add yourself to correct group, netdev in my case 2019-06-11 07:24:29 Yourself as in the user you plan to run, I suppose? 2019-06-11 07:24:47 whichever user whom will be running the wpa_cli command. 2019-06-11 07:25:10 Perfect. Sorry to ask trivial questions, but I want to be hyperclear. 2019-06-11 07:25:50 np, it is ok to ask clear questions 2019-06-11 07:26:38 priontology: we now have iwd which can be used instead of wpa_sp 2019-06-11 07:27:55 Written by intel, though. D= 2019-06-11 07:28:19 Do you use it on your current machine? 2019-06-11 07:28:37 true, but it is fully open source 2019-06-11 07:29:13 heh, to endorse myself, I packaged iwd for alpine ;) 2019-06-11 07:30:24 and yes, I use it in AP and station 2019-06-11 07:30:31 that's really interesting. 2019-06-11 07:30:44 i guess i could just apk remove wpa_supplicant and then apk add iwd? 2019-06-11 07:30:56 how do i trust that your package isn't just some ROOTKIT, huh, mps? ;D 2019-06-11 07:31:07 Do you have any issues using it? 2019-06-11 07:31:29 then any alpine pkg can be rootkit 2019-06-11 07:32:38 it passed alpine entering and building process, you can build it yourself if you do not trust alpine 2019-06-11 07:32:42 heh, true 2019-06-11 07:32:59 mps: Did you have to jump through any hoops to get it to build on alpine yourself? 2019-06-11 07:33:16 Also, if I apk remove wpa_supplicant, will that remove it from openrc? 2019-06-11 07:33:50 priontology: at the end you right, also I'm not sure if there are some pkg's which can be rootkit's 2019-06-11 07:34:15 the deps should remove it from openrc.... watch and see if wpa_supplicant-openrc is removed too 2019-06-11 07:34:20 priontology: iwd is official alpine pkg 2019-06-11 07:35:57 I don't see uw-alpine (email client) in the repos. So sad. ;) 2019-06-11 07:35:59 i just realized if i remove wpa-supplicant, i won't be able to install iwd, since i'm on wifi 2019-06-11 07:36:04 what happens if i install iwd FIRST? 2019-06-11 07:36:52 inb4 I do this and some stupid antisecurity scripts i have don't work anymore because i don't use wpa_supp 2019-06-11 07:38:26 if you're adding a package in apk, you could always pass the -s (simulate) flag to it to see exactly what apk will do before actually doing it. 2019-06-11 07:40:54 that's neat. 2019-06-11 07:40:56 EFFIN a. 2019-06-11 07:47:49 mps: I've successfully installed your rootkit! ;D 2019-06-11 07:49:07 ACTION in Mr. Burns' voice: "Excellent!" 2019-06-11 07:50:00 looool 2019-06-11 07:53:46 here is short notes or guide I wrote about using iwd on alpine http://arvanta.net/mps/iwd-howto.md 2019-06-11 07:54:43 it is incomplete and some things are superseded. I wrote it when introduced first version in alpine, as notes for myself 2019-06-11 07:57:22 btw, you can ask Cogitri here how to use iwd with networkmanager 2019-06-11 08:24:44 Hi all :-) 2019-06-11 08:34:24 Do you use firefox ? 2019-06-11 08:35:53 I've issues with firefox and firefox-esr 2019-06-11 08:36:02 It seems to be https://bugs.alpinelinux.org/issues/7454 2019-06-11 08:36:44 But this issue has been closed recently 2019-06-11 08:45:09 Hum, what Alpine version are you on? 2019-06-11 08:45:13 At least on Edge it works just fine for me 2019-06-11 08:45:37 Cogitri: I'm on edge 2019-06-11 08:47:19 It is reproducible: open (https://www.geocaching.com), click on 'Log in' then on the password field 2019-06-11 08:47:34 and the tab will crash 2019-06-11 08:48:27 I haven't applied any workaround, just a clear new firefox installation 2019-06-11 08:50:09 wfm 2019-06-11 08:50:52 hum ... 2019-06-11 08:51:34 Cogitri: have you changed something like security.sandbox.content.syscall_whitelist ? or did you set MOZ_DISABLE_CONTENT_SANDBOX env variable ? 2019-06-11 08:53:52 No, I haven't changed any of that 2019-06-11 08:54:37 That's weird 2019-06-11 08:58:27 What's your cpu arch ? x86_64 ? 2019-06-11 08:59:10 Yup. I don't think we have FF on any other arch yet due to Rust 2019-06-11 13:04:25 Hey guys, I've just done my first alpine install via chroot. It only installed 'alpine-base', which doesn't include a kernel. I see 'linux-vanilla' in the repo, is this the default kernel used by alpine? Also, where can I find out what other packages are normally installed in a standard (not extended) alpine install? 2019-06-11 13:05:07 <_ikke_> linux-vanilla is indeed the default 2019-06-11 13:05:19 <_ikke_> But why do you need a kernel installed for a chroot? 2019-06-11 13:06:33 I want to boot it bare-metal, I just used chroot as a way to install, kind of like Gentoo or Void would do it 2019-06-11 13:06:45 <_ikke_> Ok, right 2019-06-11 13:07:31 FWIW, I did the same install, chroot install it pretty comfy 2019-06-11 13:07:40 Be sure to enable the necessary services via rc-update though, pltrz 2019-06-11 13:07:40 Pretty sure when doing a chroot install no services are enabled 2019-06-11 13:08:14 I've booted a couple of systems forgetting to enable services, it's pretty bare bones but boots... 2019-06-11 13:08:20 Yes, I followed the instructions here 2019-06-11 13:08:21 https://wiki.alpinelinux.org/wiki/Installing_Alpine_Linux_in_a_chroot 2019-06-11 13:09:26 Used a statically compiled apk, bootstrapped the system, and enabled all the services that page suggested 2019-06-11 13:09:50 Good 👍 2019-06-11 13:10:03 Maybe also use udev and enable its services if this is a desktop system 2019-06-11 13:10:48 Oh, neat, thought it wouldn't boot at all without some essentials like modules 2019-06-11 13:11:28 <_ikke_> linux-firmware is something you'll probably need 2019-06-11 13:12:15 Is that a metapackage for all the linux-firmware-*? 2019-06-11 13:12:36 <_ikke_> yes 2019-06-11 13:12:50 <_ikke_> If you know exactly what you need, you can just install those 2019-06-11 13:13:02 Ok cool. 2019-06-11 13:13:50 That's installed with linux-vanilla iirc 2019-06-11 13:13:56 <_ikke_> Cogitri: you are right 2019-06-11 13:14:02 Yeah I'm on an intel (integrated graphics) system, so I know I'd need linux-firmware-{intel,i915} but not sure what else. Any way of knowing? 2019-06-11 13:14:05 <_ikke_> there is linux-firmware-none you can install though 2019-06-11 13:14:21 ^ oh i see 2019-06-11 13:15:08 meh I guess I'll just stick with all of them for now, no reason to overcomplicate my first alpine install 2019-06-11 13:18:26 Yup, best to first make sure it boots :) 2019-06-11 13:18:54 <_ikke_> I recall someone posting a shell oneliner to figure that out 2019-06-11 13:19:49 <_ikke_> Found it: modinfo `lsmod | grep -v Module | awk -e '{ print $1}'` | grep firmware 2019-06-11 13:21:46 oh nice! will try it once I get the system going 2019-06-11 13:31:11 hi 2019-06-11 13:32:03 current docker image 3.9.4 has security issue with busybox package. any eta to release patched version of alpine docker image? 2019-06-11 13:32:33 <_ikke_> linuxguy: Any link/ 2019-06-11 13:35:13 CVE-2019-5747 vulnerabilities in busybox 2019-06-11 13:35:58 <_ikke_> hmm, 1.30.0 2019-06-11 13:36:50 alright, rebooting to try my install! brb 2019-06-11 13:37:01 https://cxsecurity.com/cveshow/CVE-2019-5747 2019-06-11 13:37:31 <_ikke_> busybox needs to be patched first 2019-06-11 13:37:36 <_ikke_> on 3.9 2019-06-11 13:37:58 yeah alpine edge version didn't have this vulnerability 2019-06-11 13:38:20 <_ikke_> correct, that's already on 1.30.1 2019-06-11 13:39:53 most of our CI/CD build pipeline we use latest alpine image, now pipeline build failing with this security issue. 2019-06-11 13:40:13 I am sure others also facing same kind of problem 2019-06-11 14:08:35 _ikkie_ busybox 1.30.1-r1 already patched for this issue. 2019-06-11 14:10:22 <_ikke_> for 3.9, we need to fix 1.28 though 2019-06-11 14:10:36 <_ikke_> sorry, 1.29.3 2019-06-11 14:11:02 oh ok 2019-06-11 14:11:27 <_ikke_> hmm, it mentions that 1.29.3-r10 was fixed as well 2019-06-11 14:12:14 <_ikke_> Yes, there as well 2019-06-11 14:12:23 <_ikke_> so it should be fixed already 2019-06-11 14:13:03 <_ikke_> linuxguy: So I think it's a false positive you are getting in your CI 2019-06-11 14:13:10 so only need to update this package and create the docker image 2019-06-11 14:13:24 <_ikke_> What version of busybox are you getting? 2019-06-11 14:13:37 <_ikke_> You should get 1.29.3-r10 2019-06-11 14:14:53 imm. as per JFrog Xray scan, it patched only in busybox version 1.30.0 2019-06-11 14:15:05 not in 1.29.3.-r10 2019-06-11 14:15:16 <_ikke_> linuxguy: By busybox itself, but we have patches busybox 1.29.3 ourselfs 2019-06-11 14:15:35 <_ikke_> https://git.alpinelinux.org/aports/tree/main/busybox/CVE-2019-5747.patch?h=3.9-stable 2019-06-11 14:15:58 <_ikke_> https://git.alpinelinux.org/aports/tree/main/busybox/APKBUILD?h=3.9-stable#n55 2019-06-11 14:18:23 I think most of the scanning software go by busybox package number 2019-06-11 14:18:46 <_ikke_> yes, which can cause false positives 2019-06-11 14:19:29 <_ikke_> If upstream doesn't backports it fixes, we will for our stable branches 2019-06-11 14:21:07 good to know stable release has this fix. Thanks for the information. 2019-06-11 14:28:00 ok, so the boot failed, and it's probably because the root fs is an lvm volume. I did install 'lvm2' and 'lvm2-*' just to be safe, but still no luck. Is there some kernel module I need? 2019-06-11 14:30:51 try adding lvm to your mkinitfs FEATURES in /etc/mkinitfs/mkinitfs.conf 2019-06-11 14:31:15 and don't for get to rerun mkinitfs 2019-06-11 14:35:33 thanks for the tips. didn't know about /etc/mkinitfs/, rebuilding my initfs and gonna try again 2019-06-11 15:32:07 so I added 'lvm' to /etc/mkinitfs/mkinitfs.conf, rebuilt with mkinitfs, still no luck. THe lvm volumes are not available when in the initramfs 2019-06-11 16:36:49 and if you do `lvm vgchange -ay` they become available? 2019-06-11 16:36:58 in the initram? 2019-06-11 17:46:55 hello 2019-06-11 17:47:08 hi revolutionary 2019-06-11 17:47:58 i can't search package names in pkgs.alpinelinux.org website because i have to know certain package names and i don't know the exact name. for example when i wrote mate in the package name field it can't find that. 2019-06-11 17:48:10 <_ikke_> *mate* 2019-06-11 17:48:25 hmm let me try 2019-06-11 17:48:31 <_ikke_> https://pkgs.alpinelinux.org/packages?name=*mate*&branch=edge&arch=x86_64 2019-06-11 17:48:38 i thought the glob matching was % or am i thinking crt.sh 2019-06-11 17:49:34 <_ikke_> It used to be % in the beginning, but it switched to * 2019-06-11 17:49:55 ah okay cool 2019-06-11 17:52:35 thank you _ikke_ 2019-06-11 17:58:21 i can't set screen scale. is there any additional tool or something like that for Mate? 2019-06-11 18:00:59 <_ikke_> I don't know anything specific for Mate, but perhaps xrandr (cli tool)? 2019-06-11 18:01:16 You're on Xorg, right? 2019-06-11 18:01:18 Then xrandr should work 2019-06-11 18:43:29 % is essentially the SQL wildcard if you use `LIKE' 2019-06-11 18:43:32 so that would be a bad sign 2019-06-11 20:05:58 is there a way to remove unneeded linux-firmware-X ? 2019-06-11 20:08:21 Install any of the linux-firmware-* packages 2019-06-11 20:08:29 And the others will be purged 2019-06-11 20:08:49 is there a way to discover which one are really needed ? 2019-06-11 20:09:58 <_ikke_ "Found it: modinfo `lsmod | grep "> ^ 2019-06-11 20:10:23 <_ikke_> Found it: modinfo `lsmod | grep -v Module | awk -e '{ print $1}'` | grep firmware 2019-06-11 20:12:18 hi 2019-06-11 20:13:25 it returns nothing 2019-06-11 20:13:30 so maybe i don't need anything ? 2019-06-11 20:13:43 in debian one used to be able to specify "dns-nameservers" in /etc/network/interfaces, iirr for both static or dhcp entries. how can you specify that with Alpine? 2019-06-11 20:18:27 nice there is linux-firmware-none :-D 2019-06-11 20:20:49 there is no cloud. so it's probably for running in the cloud. 2019-06-11 20:21:27 thanks Cogitri MY_R ;-) 2019-06-11 20:25:58 at least there must be documentation for all the things that can go into /e/n/i... 2019-06-11 20:26:15 (i did try with google and in the wiki!) 2019-06-11 20:28:37 ryouba, maybe that: https://wiki.alpinelinux.org/wiki/Udhcpc 2019-06-11 20:29:30 oh hmm so this is done in a completely different way than it was back then in debian 2019-06-11 20:29:34 that helps, thank you MY_R! 2019-06-11 20:29:50 gl! 2019-06-11 21:03:38 rambleed.com 2019-06-11 21:03:40 omfg 2019-06-11 21:03:50 don't wanna be a cloud provider of any kind now 2019-06-11 21:43:30 MY_R: is there a way to configure the system so that it only tries to bring eth0 up if it's already plugged in? 2019-06-11 21:48:42 mesa 19.1 came out today with lima drivers, are there any ongoing plans to add them to alpine? 2019-06-11 21:50:24 oh llvm8 got added to main in aports, Cogitri does that mean rust will be coming to other archs soon? 2019-06-11 21:50:47 Yup, am working on that 2019-06-11 21:51:01 I have my last exam tomorrow though, so I haven't been able to work on it today :) 2019-06-11 21:51:03 \o/ 2019-06-11 21:51:51 woo! good luck on exams :) 2019-06-11 21:56:01 Thank you :) 2019-06-11 22:00:47 priontology, do you want get rid of some boot delay when your cable isn't plugged in? 2019-06-11 22:01:14 MY_R: absolutely. 2019-06-11 22:01:41 i have this recurrent wet dream where my system turns on in five seconds. 2019-06-11 22:02:34 heh 2019-06-11 22:03:42 never try mess with it but you could try lower some timeout setting in udhcpc, here is how to add extra options: https://wiki.alpinelinux.org/wiki/Configure_Networking#IPv4_DHCP_Configuration 2019-06-11 22:03:59 and check link under with: https://busybox.net/downloads/BusyBox.html#udhcpc 2019-06-11 22:04:17 I never tried mess* 2019-06-11 22:54:02 MY_R: ah, neat. 2019-06-12 03:25:20 can someone tell me the mp3 players in alpine? 2019-06-12 03:25:36 i installed VLC but no luck, didn't work. 2019-06-12 03:26:30 mpg123 comes to mind. Unsure if there are others. 2019-06-12 03:27:22 oh, audacious is a good old one too 2019-06-12 03:27:52 the latter is in the community repo 2019-06-12 03:43:48 mpg123 didn't work. 2019-06-12 03:43:57 i will try audacious 2019-06-12 03:53:51 is there any way to increase volume from Terminal from mute to 70? 2019-06-12 03:55:44 dunno offhand. alsamixer? 2019-06-12 05:58:52 mps: sorry i had been blocked for quite a loooong time 2019-06-12 05:59:13 mps the board name is pcduino3 nano 2019-06-12 07:07:37 yunfan: new release 3.10.0_rc3 have support for sunxi serial and some ethernet cards 'out of the box' 2019-06-12 07:08:27 and have few new u-boot for some sunxi boards 2019-06-12 10:17:17 hi, any volunteer to help test 3.10 rc3 on rpi? 2019-06-12 10:17:40 specifically, i wonder if this is still an issue: https://bugs.alpinelinux.org/issues/10155 2019-06-12 10:35:07 Greetings! 2019-06-12 10:47:05 ncopa: I am willing to give it a try, but I have a RPi 1 available. Is this version interesting for you? 2019-06-12 10:48:24 would be nice to know if it works yes 2019-06-12 10:50:22 ncopa: Good. Could you please guide me about where to download it and where to report the test? 2019-06-12 10:51:36 http://dl-cdn.alpinelinux.org/alpine/v3.10/releases/armhf/ 2019-06-12 10:51:48 you can report here on IRC 2019-06-12 10:51:52 ok 2019-06-12 10:52:08 alternatively, if you hit in to specific issues, file a bug on bugs.alpinelinux.org 2019-06-12 10:53:01 I suppose I should try rc3, right? 2019-06-12 10:53:09 <_ikke_> yes 2019-06-12 10:54:09 yup 2019-06-12 10:56:48 Ok, I'll try it. It ccould take a while though. I will say something here when its done. 2019-06-12 12:31:31 ncopa: Image tested on RP1 (256MB RAM version). It boots fine. I can login as root. Are other tests required? 2019-06-12 12:31:50 can you install it too? 2019-06-12 12:32:04 do you mean by running alpine-setup? 2019-06-12 12:32:07 yes 2019-06-12 12:32:10 sure 2019-06-12 12:36:57 ncopa: Everything seems fine. I runned setup-alpine and followed instructions. Network works. Only problem was at the end with the warning: "Ignoring APKINDEX... No such file or directory" but I think it's normal, it's not the first time I see it. 2019-06-12 12:38:40 aggaz: I think so. same happened to me yesterday on some sunxi/allwinner SBC's 2019-06-12 12:40:42 Is there anything else worth testing? 2019-06-12 12:41:14 im happy with that aggaz. thanks alot! 2019-06-12 12:42:20 ncopa: my pleasure. Well, that was easy enough. I also have an RP12 and a RPI3 (not plus) currently used as servers, but I think I can turn off one for a moment and test the same sd card if you think it will be useful. 2019-06-12 12:43:16 aggaz: for RPi3 you will need aarch64 version, I think 2019-06-12 12:43:35 i think you can use both armv7, armhf and aarch64 for rpi3 2019-06-12 12:43:40 but is ok to try with current you have 2019-06-12 12:43:54 Am I correct in assuming that by not running "lbu commit" no changes will be written in the SD card? Or should I re-copy everything. 2019-06-12 12:43:55 would be nice to test yes 2019-06-12 12:44:25 lbu commit will create a .apkovl.tar.gz file 2019-06-12 12:44:27 I mean: will the current SD card behave as a "not installed" version? 2019-06-12 12:44:33 yes 2019-06-12 12:44:42 even if I runned setup-alpine? 2019-06-12 12:44:53 yes i think so 2019-06-12 12:45:08 you could check though, if there are any .apkovl.tar.gz files 2019-06-12 12:48:04 would be nice if someone could test install alpine in hyper-v as well 2019-06-12 12:48:48 I see no apkvol files, I suppose everything is pristine 2019-06-12 12:48:57 no need to re-copy 2019-06-12 12:49:02 correct 2019-06-12 12:49:41 But I see a "cache" directory that was not there before 2019-06-12 12:49:56 its the downloaded apk cache 2019-06-12 12:49:58 maybe that is the source of the warning? 2019-06-12 12:49:59 you can delete it 2019-06-12 12:50:17 the installer was looking for a file in that directory 2019-06-12 12:50:21 that was not there 2019-06-12 12:50:23 I think 2019-06-12 12:50:39 may be setup-cache didnt create it 2019-06-12 12:50:40 not sure 2019-06-12 12:51:03 well, I will keep the SD card as is. Let's see if on the RPi3 I wil see the warning again... 2019-06-12 12:57:43 RPi3 (not 3+) works. I still see the warning despite the presence of the "cache" directory. Can confirm that the sistem was pristine since /etc/hostname was not modified. 2019-06-12 12:58:03 It boots and it runned setup-alpine correctly 2019-06-12 13:03:24 RPi2 works fine too. 2019-06-12 13:04:26 nice 2019-06-12 13:10:10 I also have an Orange Pi One if you are interested 2019-06-12 13:10:38 By the way, I remember that I tryied 3.8 or 3.9 some time ago on it and I had problems 2019-06-12 13:10:51 I think 3.8 2019-06-12 13:10:54 Orange Pi, is it also allwinner? 2019-06-12 13:10:58 yes, H3 2019-06-12 13:11:51 I'm not sure that we have u-boot for it 2019-06-12 13:12:53 I see. I don't rememeber how I did it at the time... probably I compiled and installed uboot manually... 2019-06-12 13:13:33 probably, that is how I do install on SBC's, i.e. build u-boot myself 2019-06-12 13:15:23 it would be to much for distro to build all u-boot for different boards, 1374 now 2019-06-12 13:15:39 s/be to/be too/ 2019-06-12 13:16:09 yes of course 2019-06-12 13:16:47 it is unfortunate that the arm boot situation is so messy 2019-06-12 13:18:11 price for diversity, don't have strong opinion is that bad or good 2019-06-12 13:27:46 aggaz: you can test on Orange Pi keeping your u-boot and use alpine-uboot-3.10.0_rc3-armv7.tar.gz 2019-06-12 13:29:03 we enabled serial console driver for some allwinner boards on latest kernels and some ethernet drivers 2019-06-12 13:29:16 mps: good, I will give it a try 2019-06-12 13:29:43 I had problems with serial console in the pat (3.7 I believe) and if I remember correctly network in 3.8 2019-06-12 13:29:53 s/pat/past/g 2019-06-12 13:29:55 thanks, would be pleasure to work with you to improve alpine on arm's 2019-06-12 13:30:17 cool 2019-06-12 13:30:27 aggaz: yes, that was I reason we enabled serial console driver for some boards 2019-06-12 13:31:27 I didn't dare to enable serial console drivers for all boards because I couldn't test if they break something 2019-06-12 13:31:45 I enabled it only for boards where I can test 2019-06-12 13:32:04 I understand 2019-06-12 13:34:56 btw, I presume you know that you need to change extlinux.conf for these boards, here is my change 'APPEND modules=loop,squashfs,sd-mod,usb-storage,ext4,sunxi-mmc console=${console}' 2019-06-12 13:35:19 mps: yes I did that 2019-06-12 13:35:32 nice :) 2019-06-12 13:35:44 funny thing was that it worked on 3.7 but not on 3.8 2019-06-12 13:35:49 (if I remember correctly) 2019-06-12 13:37:44 also I can't remember now, but yes 3.8 and 3.9 didn't worked 2019-06-12 13:45:50 in default sh, how can I write a script where it executes default commands "foo 'bar hello' world". But if user supplied arguments, it executes those arguments? 2019-06-12 13:46:10 ./script.sh by default runs ./foo 'bar hello' world 2019-06-12 13:46:19 saml: look at variable expansion 2019-06-12 13:46:32 ARG=${1:-"default"} 2019-06-12 13:46:34 ./script.sh ./foo -v will execute ./foo -v 2019-06-12 13:46:48 problem is arguments contain whitespaces, and it seems to confuse the shell 2019-06-12 13:46:57 saml: read above 2019-06-12 13:47:08 ARG="${1:-'default'}" 2019-06-12 13:47:43 that just means first argument has default. If user passes multiple arguments, I want to execute them as is 2019-06-12 13:47:58 ./script.sh ./foo a b c # same as ./foo a b c 2019-06-12 13:48:34 saml: see $@ 2019-06-12 13:49:29 or $*, depending on the cases 2019-06-12 13:49:53 $@ almost without exception, rule of thumb 2019-06-12 13:49:59 namely 2019-06-12 13:50:21 there are some rare cases in which you want $*, but this does not seem one of those 2019-06-12 13:54:03 saml: you want something like: 2019-06-12 13:54:05 ARG="${@:-foo bar}" 2019-06-12 13:54:06 echo $ARG 2019-06-12 13:55:27 hello, I want to make a tiny alpine for rpi that will just boot take a picture/video. where do i start stripping? 2019-06-12 13:55:35 for example, remove networking 2019-06-12 13:56:06 https://gist.github.com/saml/333ccaae450a4d482d49eaafb6d5d6bf problem is this 2019-06-12 13:56:22 once I capture cmd="$@", I don't know how to invoke it 2019-06-12 13:57:00 cmd="$@"; $cmd doesn't seem to be the same as "$@" 2019-06-12 13:58:15 saml: ever heard of "eval"? :) 2019-06-12 14:01:52 oh 2019-06-12 14:07:21 kayterina[m]: basically disable the relevant openrc services 2019-06-12 14:58:43 mps: why did you add "ext4" and "sunxi-mmc" on your "APPEND" line? 2019-06-12 15:08:05 aggaz: ext4 driver is built as a module, and I want to install in sys mode on ext4 FS 2019-06-12 15:19:50 mps: I just installed 3.10 on OPi One. Serial works and setup-alpine is now ongoing 2019-06-12 15:20:10 and ethernet also? 2019-06-12 15:20:10 network works too 2019-06-12 15:20:13 yes 2019-06-12 15:20:16 it seems fine to me 2019-06-12 15:20:22 very good 2019-06-12 15:20:25 indeed 2019-06-12 15:20:36 manual u-boot only was needed 2019-06-12 15:21:01 do you think it would be possible to add this particular u-boot to the release files? 2019-06-12 15:21:41 please tell me exact code name of your board, I'm thinking to make list of arm boards which are confirmed to work 2019-06-12 15:22:17 I am not sure about a code name... I know this board as "Orange Pi One" 2019-06-12 15:22:26 I thought to add more u-boot targets but not sure how much of them and which ones we should add 2019-06-12 15:23:37 https://www.aliexpress.com/store/product/Orange-Pi-One-ubuntu-linux-and-android-mini-PC-Beyond-and-Compatible-with-Raspberry-Pi-2/1553371_32603308880.html 2019-06-12 15:23:41 this is the board 2019-06-12 15:23:57 one of these, https://linux-sunxi.org/Xunlong_Orange_Pi_One_%26_Lite 2019-06-12 15:25:05 yes the one with the ethernet 2019-06-12 15:25:56 aggaz: big thanks for your work! 2019-06-12 15:26:57 mps: my pleasure 2019-06-12 15:27:08 nfortunatly it seems that usb is not working 2019-06-12 15:27:18 I noticed that 2019-06-12 15:27:27 i can not see an usb key or an usb drive on dmesg/dev 2019-06-12 15:27:36 that is first thing for next update 2019-06-12 15:27:57 ok 2019-06-12 15:28:37 just FYI, currently I am using a custom image with debian testing on this particular board, and everything works fine. I suppose is something kernel related 2019-06-12 15:28:58 I didn't want to introduce big changes in for rc3 release, fearing I can break something 2019-06-12 15:29:11 yes I understand 2019-06-12 15:29:43 I use alpine on all boards but build kernels/drivers myself 2019-06-12 15:30:01 anyway I would love to see alpine working on this particular board... lets keep this conversation open for future changes :) 2019-06-12 15:30:47 yes, I hope we will continue to fix rest issue and make it better later 2019-06-12 15:34:13 I need someone to fix mkimage.sh for armv7 uboot target by which will be easier for me to test changes and new drivers 2019-06-12 15:34:41 what is mkimage.sh? is it something alpine specific? 2019-06-12 15:35:44 there is in alpine aports main git repo dir scripts with scripts to build distro images, iso and tar.gz for now 2019-06-12 15:37:15 https://git.alpinelinux.org/aports/tree/scripts 2019-06-12 15:37:30 ok I see it 2019-06-12 15:39:11 now I'm going to lunch and some private duties, see you later :) 2019-06-12 15:39:17 bye! 2019-06-12 16:31:12 is there a way to change the way apk works? like theres this file: https://git.alpinelinux.org/aports/tree/main/postgresql/APKBUILD?id=40437ec074e6bde2cf18c969099795b0079141dc -- can I somehow change it to use 9.5.17 ? 2019-06-12 16:38:35 ash_worksi: you can contribute changes to APKBUILD files, yes 2019-06-12 16:39:07 but we (upstream) will not accept changes like a downgrade without a very good reason 2019-06-12 16:39:29 i suggest you download the directory the APKBUILD is in, edit it, and build it with `abuild' 2019-06-12 16:40:17 danieli: right, I wouldn't want to try to change the build file. I just want to do like an in-place install 2019-06-12 16:40:30 danieli: my real goal is to install plperl with pg 9.5 2019-06-12 16:40:52 an in-place install of only pl/perl? 2019-06-12 16:41:03 well, no, with PG 9.5 2019-06-12 16:41:41 and potentially other extensions... I am trying to get a starting migration going for sqitch 2019-06-12 16:41:46 we don't build pl/perl separately, so you'd have to either find an apkbuild for pg 9.5 (in the history or elsewhere) or adapt the build file and build it locally 2019-06-12 16:45:02 ahills: I did this for pg 9.4, found commit in aports and rebuild on current stable. everything works fine 2019-06-12 16:47:26 I need a hand with mkinitfs ... I am running it in a docker container hosted on an ubuntu box. 2019-06-12 16:47:55 when adding the apk it seems to invoke just fine via the install trigger 2019-06-12 16:48:09 but on my interactive sh ... I am always getting: 2019-06-12 16:48:28 / # mkinitfs 2019-06-12 16:48:28 /lib/modules/4.15.0-50-generic does not exist or is not a directory 2019-06-12 16:49:12 danieli: who would I do the later? 2019-06-12 16:49:27 danieli: ie: where is the build file on the system? 2019-06-12 16:49:45 <_ikke_> ash_worksi: they are not by default on the filesystem 2019-06-12 16:49:52 it's not included by default on your system, you'll have to download it using `git clone' or manually from git web browsers 2019-06-12 16:50:10 then what? (note: I am running alpine from a container) 2019-06-12 16:50:30 you'd have to use git for that 2019-06-12 16:50:35 ash_worksi: I wrote wrongly to ahills_ 2019-06-12 16:50:38 or download the files individually through some git browser online 2019-06-12 16:51:02 ash_worksi: I did this for pg 9.4, found commit in aports and rebuild on current stable. everything works fine 2019-06-12 16:51:44 no I mean, what is the process. Assume I have the necessary aports file and changed it locally... I would assume there's an apk command option which instructs apk to use that file? 2019-06-12 16:51:59 <_ikke_> ash_worksi: You need abuild for that 2019-06-12 16:52:13 apk doesn't use those files directly, they're just an instruction to create prebuilt packages, which apk in turn uses 2019-06-12 16:52:27 I see 2019-06-12 16:52:56 <_ikke_> https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package#Setup_your_system_and_account 2019-06-12 16:54:49 thanks _ikke_ , danieli 2019-06-12 16:56:31 any tips for running mkinitfs in a container? 2019-06-12 17:01:42 JjOtt: why would you do that though? You don't need a kernel/initramfs in a container 2019-06-12 17:04:07 I am building my own startup process - so I've got alpine running in a container as part of my build 2019-06-12 17:04:28 and one of the things I need is to make my own initramfs that has my own modules and init scripot 2019-06-12 17:05:38 What's going on is that it's choosing the wrong kernel, its referencing the docker host's kernel when I really want the one found in the container ... just need the right options and environment vars it seems 2019-06-12 17:12:46 mkinitfs should work in that case 2019-06-12 17:12:54 just figured this out 2019-06-12 17:13:12 In the trigger for mkinitfs it's called there 2019-06-12 17:13:26 cat /usr/share/kernel/vanilla/kernel.release 2019-06-12 17:13:34 mkinitfs `cat /usr/share/kernel/vanilla/kernel.release` 2019-06-12 17:31:25 If I want to run my own custom install code at startup - where would the best place to hook-in be? I am considering either an init.d script or changing the end of the initramfs init to hand over to my code vs the regular init stuff ... any suggestions? 2019-06-12 17:31:37 This is for a fully automated installer btw 2019-06-12 18:09:27 I'm trying to host a static website with TLS. I am kinda hopping around right now trying to find the best solution. I basically want a minimal web server and distribution to just host the static site since there's no PHP, etc. Right now I'm using OpenBSD with httpd. However, I'm not that familiar with BSD and would maybe rather use Linux. Would nginx be the smallest web server on Alpine? I'm new to Alpine as well. 2019-06-12 18:10:51 d9a: darkhttpd is small and simple 2019-06-12 18:12:42 actually, busybox have simple httpd server 2019-06-12 18:13:11 https://busybox.net/downloads/BusyBox.html#httpd 2019-06-12 18:32:30 Thanks. I'll look into those. I also just realized that busybox might be just as familiar as the OpenBSD coreutils. However, I guess updating Alpine is probably more like other distros. With OpenBSD, I have to run two commands to update everything, and I'm still confused as to if binary packages get security updates. 2019-06-12 18:43:05 busybox does not support https tho? 2019-06-12 18:47:03 Don't think so, no 2019-06-12 18:48:06 d9a if you're familiar with nginx I'd stick with that 2019-06-12 19:11:41 JjOtt: Except some Apache experience like 10 years ago that I don't remember, I am not familiar with anything. I guess now the server I'm most familiar with is now OpenBSD httpd. Might stick with that then. haha 2019-06-12 19:36:10 d9a: a little bit of googling uncovers that darkhttpd does officially 'suck less' as shown on www.suckless.org so worth a look! 2019-06-12 19:36:48 'suckless' can usually be taken with a pinch of salt 2019-06-12 19:37:32 JjOtt: I'm looking into that. Thank you. I do think it's cool that the communities I hang out in know what I want without me even having to explicitly say. haha. Only reason I'm not using the "official" suckless web server is the fact that it doesn't support https 2019-06-12 19:37:59 Yeah. There are some things that can't be made simpler so they will suck regardless in terms of "suckless" definitions. haha 2019-06-12 19:43:06 Indeed - having a look at the suckless list ... not the panacea I was hoping for 2019-06-12 19:44:47 I know the answer is yes, so I'm asking for a bit more detail, but do any of you use Alpine for desktop use? From what I've seen it's more specialized for containers and servers. Just curious as to why you do if so. 2019-06-12 19:45:16 JjOtt: I still think it's super cool that I didn't have to mention suckless, and you knew exactly what I wanted. haha 2019-06-12 20:01:02 d9a, working just great but in my case I'm using "edge", firefox, i3, mpv, telegram and I don't need more :) 2019-06-12 20:03:15 and all on single full encrypted LUKS partition with LVM 2019-06-12 20:05:34 MY_R: Cool. My use case is similar. I just use dwm, st, surf, and a few CLI applications. I was just curious. 2019-06-12 20:11:27 d9a, with such a sane configuration you should be happy with Alpine, grab some useless disk or backup and just try it out :) 2019-06-12 21:51:22 Alright, perhaps I'm a fool, here. 2019-06-12 21:52:45 I'd like to configure st and vim with solarized themes, but I can't seem to get either working for different reasons. st requires me to patch the source, which I've done, but then I can't make it (x11/xatom is required, couldn't find it in the repo), and vim is installed and working, but busybox also provides vi, and so I want to know how I can shadow busybox's vi with the vim I pulled from apk. 2019-06-12 21:52:49 Well, from edge, I guess. 2019-06-12 21:56:46 priontology: alias vi=vim 2019-06-12 22:09:50 Hey guys, I'm back trying to install alpine (sys mode) in an LVM logical volume. First, I tried a chroot install, added the `lvm2` and `linux-vanilla` packages, added `lvm` to /etc/mkinitfs/mkinitfs.conf, rebuilt the initfs, no luck. Next, I tried booting the standard and extended ISOs, but no luck with either of them. The issue is that my LVM partitions are not being recognized, and I'm not sure why. 2019-06-12 22:15:04 priontology: to build st, install fontconfig-dev, freetype-dev, libx11-dev, libxext-dev, libxft-dev (plus ncurses-terminfo) 2019-06-12 22:16:01 pltrz: Can you manually mount the thing in the emergency shell of the initramfs? 2019-06-12 22:20:26 Cogitri: no, /dev/mapper/ does not exist 2019-06-12 22:34:12 pltrz: is your hard disk / partition recognized at all? 2019-06-12 22:39:32 pltrz: can you do 'lvm vgchange -ay' in the initram shell and do you have /dev/mapper afterwards? 2019-06-12 22:42:10 Does anyone run skype on their alpine install? 2019-06-12 22:42:21 Actually, bad question. I'll see what I can run out there. 2019-06-12 22:44:48 p4Wv1qn095FW: I'll test that now 2019-06-12 22:49:47 p4Wv1qn095FW: what was the command? I didn’t get that irc history while on mobile lol 2019-06-12 22:54:39 pltrz_: 'lvm vgchange -ay' 2019-06-12 22:54:51 Thanks 2019-06-12 23:02:11 p4Wv1qn095FW: magic! 2019-06-12 23:03:23 So the volume group was not active then? 2019-06-12 23:03:42 Wow I feel stupid 2019-06-12 23:22:16 xcko: Thanks for the recommendation. 2019-06-12 23:22:32 xcko: Now, if I have the st package installed, do I need to remove that with apk? 2019-06-12 23:22:43 Before I make/make install, etc 2019-06-12 23:25:37 priontology: depends on your prefix. typically st installs into /usr/local, which is before the /usr paths by default (/etc/profile) so no, you don't have to. 2019-06-12 23:26:32 That's so weird. So how does apk know something has changed, if I end up doing an apk del later? 2019-06-12 23:26:39 Or are you saying that apk doesn't install somewhere else? 2019-06-12 23:27:02 Also, when do I know when to make install or sudo make install? 2019-06-12 23:27:31 apk installs into /usr, /usr/local is for you. if you overwrite files in /usr, odds are you will conflict with apk. 2019-06-12 23:28:02 If you need elevated permissions to put files in /usr/local or whatever your prefix is, then yes you need sudo. Otherwise, you do not. 2019-06-12 23:28:40 ie, try without sudo and if it fails, use sudo 2019-06-12 23:29:14 Ah, okay. 2019-06-12 23:29:29 Huh. 2019-06-12 23:29:35 What the hell is tic? 2019-06-12 23:30:22 terminfo compiler. idk if alpine packages that. I just comment out that line in the Makefile 2019-06-12 23:30:27 lmao 2019-06-12 23:30:36 what does it change in the end result? 2019-06-12 23:30:57 Unless you're developing a new terminal, tic is, frankly, useless. 2019-06-12 23:31:42 Aight, so now I need to...alias st to be /usr/local/st? 2019-06-12 23:32:18 st comes with it's own st.info so users can change tab indent lengths, which would need tic to make changes take effect. 2019-06-12 23:32:24 Er, usr/local/bin/st. 2019-06-12 23:32:30 priontology: try `which st` 2019-06-12 23:32:31 This is exciting. I love solarized themes. 2019-06-12 23:32:42 Oh, it already did it for me. Hell yeah? 2019-06-12 23:32:57 yeeeeee 2019-06-12 23:33:10 Now I just need a solarized theme for awesome. We're almost in business. 2019-06-12 23:33:25 see `echo $PATH` to view your default path where binaries are searched. 2019-06-12 23:33:27 Is it just me, or does solarized remind people of the color choices in plan 9? 2019-06-12 23:34:00 /usr/local is in front of /usr, which is why when you run `st` it goes to /usr/local/bin/st 2019-06-12 23:34:18 Ah, so it's greedy, then. 2019-06-12 23:34:49 I think the solarized light theme is similar to plan 9, but plan9 had no syntax highlighting ;) 2019-06-12 23:34:55 lmao 2019-06-12 23:35:03 and i'm glad that you agree a bit. 2019-06-12 23:35:05 neat! 2019-06-12 23:35:18 the light theme looks a lot like paper to me. 2019-06-12 23:36:44 I think the patch to switch color schemes for st is pretty nifty when using solarized. Just hit F6 to switch between light and dark during run time. 2019-06-12 23:38:59 dammit. my box is surely dying: 2019-06-12 23:39:01 kernel:[182410.008034] Uhhuh. NMI received for unknown reason 3c on CPU 0. 2019-06-12 23:44:02 xcko: Yeah, that was clever. I wish it was more pluggable, but I guess the idea would be that this would increase code complexity, or something. 2019-06-12 23:44:26 pluggable? 2019-06-12 23:44:39 Configurable through means other than recompilation. 2019-06-12 23:45:16 There is the xst fork of st with Xresources support. I've never used it myself 2019-06-12 23:45:42 Ah, true. Who is the original dev? 2019-06-12 23:48:08 The original developer of st? I'd check the git log. __20h__ probably. or are you asking which fork is the original? st was first, xst is a fork of st 2019-06-12 23:48:32 Yeah, asking the original dev. Interesting stuff. 2019-06-12 23:48:36 Back soon! 2019-06-13 01:03:22 Anyone know how to resolve so:libproj.so.13 as an unresolvable constraint of the postgis in edge? searching the packages I only seem to find libproj.so.15 2019-06-13 01:04:37 oninoshio: I'm guessing postgis has to be rebuilt against the latest proj4, I remember seeing someone working on that the other day - I guess they forgot the rebuild. 2019-06-13 01:06:09 hm, "testing/postgis: upgrade to 2.5.2" 2019-06-13 01:06:51 ah nevermind, that PR was closed 2019-06-13 01:07:44 testing/postgis 2.5.1-r1 is what apk is trying to install 2019-06-13 01:09:06 yeah, someone probably updated proj4 recently and forgot to rebuild postgis 2019-06-13 01:10:51 is there someone/somewhere I can ask to do that? I mean, I suppose I could just build it myself, but I'm lazy. 2019-06-13 01:11:37 I relayed it to the channel #alpine-devel 2019-06-13 01:11:49 okay, thanks. 2019-06-13 01:11:51 many of the alpine devs are in europe, and it is currently 3.11 am here in CET 2019-06-13 01:17:26 i guess they have some issues on armv7 edge 2019-06-13 07:59:18 Hi 2019-06-13 07:59:57 Does anyone know if udhcpc can support IPv6 or how to make dhclient work on Alpine Linux ? 2019-06-13 08:11:54 guycaprio: try udhcpc6 2019-06-13 08:13:14 Wow ... I feel stupid hehe 2019-06-13 08:13:36 Thanks ! 2019-06-13 08:14:05 haha, don't feel stupid, nobody uses udhcpc6 (thanks to SLAAC) 2019-06-13 08:14:07 :) 2019-06-13 08:30:50 <_ikke_> zcrayfish: enterprises are more keen to use dhcpv6 2019-06-13 08:31:06 any idea if apk list -q can have better/parsable output? Like pip list --format=json for example ?:) 2019-06-13 08:31:22 I specifically meant udhcpc6, not dhcpv6. 2019-06-13 08:31:52 But I was also being facetious :) 2019-06-13 08:34:15 muhaha: it does not 2019-06-13 08:34:45 I've implemented it together with some other WIP stuff, but it's not ready for GA 2019-06-13 10:44:06 Hello ! I would like to mount /proc with hidepid=2 but I can't find where is mounted /proc on Alpine ? My /etc/fstab only contains HDD 2019-06-13 11:36:40 it's a procfs mount that is done already in the initramfs 2019-06-13 11:38:10 you might be able to add a line to fstab with that option plus remount 2019-06-13 12:43:12 I get a list of initcalls from dmesg that I want to remove.How do I find from which script they are called? 2019-06-13 13:44:16 hi 2019-06-13 13:44:47 how can I add the RFCOMM_CONFIG 2019-06-13 13:44:59 and rebuild the kernle 2019-06-13 13:45:02 kernel 2019-06-13 13:45:14 you can recompile the kernel using abuild and the APKBUILD ('recipe') file in our git 2019-06-13 13:45:27 https://git.alpinelinux.org/aports/tree/main/linux-vanilla 2019-06-13 13:45:40 thank you @danieli 2019-06-13 13:45:47 you're welcome :) 2019-06-13 13:45:54 i will return after trying 2019-06-13 13:46:11 there's some stuff about setting up an abuild environment @ wiki.alpinelinux.org, should be easy to find 2019-06-13 14:18:36 TBB: it works, thanks 2019-06-13 15:35:01 i lost my notes on the nss_wrapper debugging -- can anyone remind me who was involved in that and links for the gdb dumps? 2019-06-13 16:05:36 hello, could someone tell me, doesn't cron in alpine support variables define in it? 2019-06-13 16:36:37 FreeBDSM: which cron? busybox or some other cron? 2019-06-13 16:53:03 p4Wv1qn095FW: I don't know, how would I check? 2019-06-13 16:53:32 p4Wv1qn095FW: just figured out: busybox 2019-06-13 17:02:16 is there a -doc package for the basic busybox tools? can't seem to find anything via search 2019-06-13 17:56:31 kvik: no 2019-06-13 20:04:47 I saw a message on Alpine's website "Alpine Linux is a security-oriented, lightweight Linux distribution based on musl libc and busybox." I am bit privacy an security addicted person although i am not an expert on Linux. If possible could you please share with me hardened security/privacy tips about Alpine Linux. I'd want to read articles, PDFs and whatever about Alpine. Thanks! 2019-06-13 20:12:38 revolutionary, well... the bit that was most often noted was the grsecurity-like kernel. other than that, it's the minimalism, the clean libc (musl) and... that's about it. 2019-06-13 20:14:14 TBB: respectfully i didn't get what do you mean... 2019-06-13 20:15:37 that's basically the feature set that earned Alpine the reputation as secure, and those are all pretty good things to have, even though there's no grsec-offshoot kernel any longer 2019-06-13 20:16:13 <_ikke_> There are some compiler protections as well 2019-06-13 20:16:21 as i know from the past grsecurity stuff was fine, correct? 2019-06-13 20:16:22 ah, true 2019-06-13 20:16:44 the most important thing i am looking for security/privacy hardened tips on Alpine. 2019-06-13 20:17:46 grsec is not in Alpine because upstream changed license and distribution model 2019-06-13 20:17:54 revolutionary, yeah, although, to be honest grsec wasn't utilized to its full potential from the distro side (because maintaining things like mandatory access control lists is a ridiculous amount of work - only Canonical and Red Hat have the resources for it) 2019-06-13 20:17:59 it is not free anymore 2019-06-13 20:19:16 mps: yes, sadly 2019-06-13 20:19:35 You mean i don't need any extra privacy/security tweaks on Alpine, correct? 2019-06-13 20:20:27 imo, security is not product but a process (and work) 2019-06-13 20:20:35 ^ 2019-06-13 20:21:44 so? 2019-06-13 20:22:52 if someone need secure system then a lot of work is needed, and at the end no one can be sure is the system is really secure, just could have _impression_ that it is secure 2019-06-13 20:23:31 imho Alpine is lightweight and by that can easier understand what is going on under "hood" and adjust to own needs 2019-06-13 20:23:40 mps: i agree. may i ask what does mean "security oriented" exactly? 2019-06-13 20:24:16 MY_R: yes, less moving part means easier to understand and fix 2019-06-13 20:24:42 I imagine it means security implications are part of the thinking process behind producing packages and releases 2019-06-13 20:24:55 maybe 2019-06-13 20:25:08 TBB: or it was at the time ;) 2019-06-13 20:26:04 I'll say this as not a contributor or a member of the dev team, however I've participated in getting some small details of Alpine security improved 2019-06-13 20:26:07 as i understand and according to your talking there is no security/privacy hardened tips about Alpine yet. 2019-06-13 20:26:35 <_ikke_> There is no whitepaper or PDF detailing everything that's being done, no 2019-06-13 20:27:11 n00bs like me unlucky in this step. 2019-06-13 20:27:14 revolutionary: as TBB wrote, there is efforts to keep it secure to reasonable degree 2019-06-13 20:27:57 mps: i believe Alpine devs or volunteers'opinions 2019-06-13 20:28:01 <_ikke_> revolutionary: there is some documentation effort going on 2019-06-13 20:28:11 <_ikke_> But it takes time 2019-06-13 20:28:30 _ikke_ thank you for information about that 2019-06-13 20:29:31 personally I do whatever my knowledge and time allow to keep it secure, but I cannot assert it is secure 2019-06-13 20:30:23 personally if you share what you did on your system we (n00bs and experts) can read those and we can apply those on our systems if we see those as acceptable. 2019-06-13 20:30:46 i mean i need your and other experts/volunteers'ideas, experiences. 2019-06-13 20:30:50 <_ikke_> And you as a user have a large impact on that as well 2019-06-13 20:30:52 <_ikke_> The more things you install or more complex things you install, the less secure your system is going to be 2019-06-13 20:31:49 i don't think to install complex things. I am just trying to tweak system with hardened tips like modify sysctl, etc. 2019-06-13 20:32:25 having JS enabled browser in distro (or machine) and thinking about security is pointless, imo 2019-06-13 20:32:32 maybe i want too much things as a user :| 2019-06-13 20:33:18 by using musl alpine by default have smaller surface for attack, for example 2019-06-13 20:33:22 ordinary user 2019-06-13 20:33:27 we still need to remember not only about software but hardware and that is the rabbit hole :\ 2019-06-13 20:34:36 MY_R: right, hardware is becoming more and more important in 'security' thinkering 2019-06-13 20:35:08 :( 2019-06-13 20:35:20 <_ikke_> Apparently you need some hardened ddr4 ram nowadays 2019-06-13 20:35:40 i have ddr2 2019-06-13 20:35:48 my PC is 14 years old :| 2019-06-13 20:35:50 hehe :D 2019-06-13 20:35:55 :) 2019-06-13 20:36:04 the best i could do 2019-06-13 20:38:33 I have got Alpine on two machines, one with c2d and another c2q and still have no problem with "speed" even if both cpus dont have any hw AES for disk encryption :) 2019-06-13 20:39:01 i just wanted to see some security superiority on Alpine because when i saw that message (security oriented) i thought "Yes, if they wrote that on main page they absolutely have different stuff from the others". For example when you visited this (https://www.gnu.org/distros/free-distros.html) website you will see PureOS's description text (PureOS, a GNU distribution based on Debian with a focus on privacy, security, and 2019-06-13 20:39:01 convenience.) Imo this means "we have really good stuff about security". Am i thinking wrong? 2019-06-13 20:39:31 <_ikke_> revolutionary: The problem is that it's a moving target 2019-06-13 20:39:38 <_ikke_> it's not something constant 2019-06-13 20:39:54 <_ikke_> Like we used to have grsec patches, but not anymore 2019-06-13 20:39:56 hmm 2019-06-13 20:40:02 well, CVEs are fixed in stable releases for long enough time 2019-06-13 20:41:32 basically like in most Linux distributions and that making them secure enough :) 2019-06-13 20:42:36 'security' and 'privacy' are good words for marketing ;) 2019-06-13 20:42:36 My laptop's showing the right time through date, but not through awesome. 2019-06-13 20:42:54 mps, you forgot about 'stable' :D 2019-06-13 20:43:05 :D 2019-06-13 20:43:05 _ikke_: what happened to grsec? 2019-06-13 20:43:15 grsec is dead 2019-06-13 20:43:35 Was it mainstreamed? 2019-06-13 20:43:36 revolutionary: if you want security use OpenBSD. That's their ethos, their main value. 2019-06-13 20:44:14 priontology, try import your time zone using TZ env 2019-06-13 20:44:16 <_ikke_> It went behind a paywall 2019-06-13 20:44:17 revolutionary: grsec is not dead, but if you want it you have to go to Spender's site and buy licencse 2019-06-13 20:44:31 i mean it is not free anymore 2019-06-13 20:44:36 and i don't like that 2019-06-13 20:45:01 MY_R: into awesome? It originally started as one hour off, and correct with the date function, and after I used the timezone tools, it got all weird 2019-06-13 20:45:01 priontology: https://pax.grsecurity.net/ 2019-06-13 20:45:01 mps: not saying it is right, but for all intents and purposes, it is dead. 2019-06-13 20:45:54 terror: theoretically, and you are right 2019-06-13 20:46:03 terror: OpenBSD is bit far to me for now :p 2019-06-13 20:46:12 Wish someone with deep pockets buy them and upstream their work. They've been treated unfairly and it would be peanuts for the Googles and IBMs of this world. 2019-06-13 20:46:56 hmmmm, you think google and ibm wants end user security 2019-06-13 20:47:15 revolutionary: it is important to understand the values of the projects/communities you join. Linux is going to err on the side of "not breaking stuff" vs "security" every time. 2019-06-13 20:47:21 Those are Linus' values 2019-06-13 20:47:25 priontology, to have peace I made /etc/profile.d/timezone.sh with "export TZ='Europe/Amsterdam'" inside :P but dunno what awesome using to get it 2019-06-13 20:47:44 mps: it was just an example, plenty of companies have the resources to upstream GRSEC if they wanted to. 2019-06-13 20:47:59 my awesome works out of the box with time 2019-06-13 20:48:07 Can't someone rebuild the sources? 2019-06-13 20:48:20 They are bit old 2019-06-13 20:48:39 <_ikke_> Someone it the alpine team tried 2019-06-13 20:48:46 <_ikke_> to forward port the patches 2019-06-13 20:48:47 revolutionary: is not about the sources, is about the know how of everything is hang together. 2019-06-13 20:49:24 That requires spender and the PaX team to help upstreaming and they should get paid to do that. 2019-06-13 20:49:58 They won't do it for free (and I suspect they wont do it for money either unless the offer is outrageous) 2019-06-13 20:50:08 i guess PaX team won't help to the people anymore 2019-06-13 20:50:19 They have burned too many bridges with the kernel devs. 2019-06-13 20:51:05 i see 2019-06-13 20:51:57 now you are saying "just install and use Alpine" 2019-06-13 20:52:43 revolutionary: use whatever you feel comfortable with :) 2019-06-13 20:53:44 terror: which OS/distro are you using if it's not top-secret? 2019-06-13 20:57:09 I've been using Alpine as my daily driver since 2014 2019-06-13 20:57:26 But I don't use it as a 'desktop'. 2019-06-13 20:57:37 what do you mean that? 2019-06-13 20:57:55 which purpose for are you using? 2019-06-13 20:57:58 No GUI 2019-06-13 20:58:01 No browser 2019-06-13 20:58:04 ahh, ok 2019-06-13 20:58:09 Like OpenBSD 2019-06-13 20:58:10 Just cli 2019-06-13 20:58:12 yes 2019-06-13 20:58:19 Then you are using Lynx 2019-06-13 20:58:44 No, I'm using a mac 2019-06-13 20:59:15 For anything that's UI. Linux is just a server box where I run stuff. 2019-06-13 20:59:55 then you are not using Android 2019-06-13 20:59:58 or IOS 2019-06-13 21:00:10 Nokia 3310? 2019-06-13 21:00:18 What does Android or IOS have to do with anything? 2019-06-13 21:01:13 well, for me OpenBSD is/was very usable as a desktop, just need to have some "sane" config and needs :D 2019-06-13 21:01:32 As i understand you are security addicted more than me. No browser, no GUI, no JS and i thought maybe you are not using Android. You know Android does have many backgruound service which is sending analytics or something like that 2019-06-13 21:01:46 Also you shouldn't use IOS 2019-06-13 21:02:13 MY_R: when i switched to OpenBSD i will ping you :) 2019-06-13 21:02:22 and remember that every notification going by google servers! :) 2019-06-13 21:02:49 MY_R: absolutely! :) 2019-06-13 21:03:22 revolutionary, give a try to Alpine :) 2019-06-13 21:05:14 MY_R: i will. Now i deleted my LineageOS completely and installed postmarketOS. I am very happy with that although i can't use SMS/Call feature :) I will find a way to test those. I am preparing pmOS to daily usage. I am waiting some packages from Alpine team. As i learned they have some issues on armv7 :/ But i will try Alpine too. Not now because first i have to make ready to use my pmOS :) 2019-06-13 21:09:37 I wish you best luck and lot of entertainment with all that :) 2019-06-13 21:09:50 i hope so too :) 2019-06-13 21:10:15 but please keep in mind i will ping you when i switched to OpenBSD :) 2019-06-13 21:11:54 heh, I won't be so helpful probably since using Alpine :P 2019-06-13 21:12:39 Then i will ping you about Alpine. Ultimately i will ping you :P 2019-06-13 21:12:46 :D 2019-06-13 21:13:44 but ye, remember that OpenBSD put a loooot of effort in documentation 2019-06-13 21:14:09 My to-do-list => Ubuntu->Alpine->Debian->OpenBSD->FreeBSD->gNewSense (i heard from someone Richard Stallman had used this for a while) 2019-06-13 21:15:40 hmm, yes they have good resources 2019-06-13 21:16:54 gnu.org and BSD systems are just dream for me for now 2019-06-13 21:16:59 revolutionary, and what is your favorite window manager? 2019-06-13 21:19:03 I tried all of them and i liked KDE 2019-06-13 21:21:15 This is WAY off-topic 2019-06-13 21:21:35 absolutely, sorry for the noise 2019-06-13 21:22:24 terror, agree 2019-06-13 21:22:33 #alpine-offtopic 2019-06-13 21:35:27 hey community. i am using alpine 3.9.4 in a docker service. however as soon as i install imagemagick the container freezes with zombie 'sync' process . https://pastebin.com/raw/3B3pR2Lw any ideas how to fix this ? 2019-06-13 21:52:07 dreadkopp: it is your environment, not Alpine. 2019-06-14 07:38:49 should the fsck service be in the sysinit level or boot? 2019-06-14 07:48:08 xcko: Dynamic Runlevel: needed/wanted. (it's called by localmount) 2019-06-14 07:48:55 hm. i've noticed my swap isn't being mounted on boot, thought it might be because fsck wasn't showing up with rc-update 2019-06-14 07:49:29 how about the swap service? 2019-06-14 07:49:34 try rcstatus -a 2019-06-14 07:49:45 swap should be boot 2019-06-14 07:51:03 I installed from a rootfs, not from a regular install so I think I missed some services I needed. looks like swap is not a part of boot - I'll add it then 2019-06-14 07:52:40 thanks zcrayfish 2019-06-14 07:52:49 np 2019-06-14 08:22:44 fsck I'd add to boot as well 2019-06-14 08:50:32 Hi folks, i have a fresh installation of alpine 3.9.4 x86_64 that hangs starting busybox crond (after chronyd). takes more than 10 minutes to boot after finally starting crond 2019-06-14 08:52:11 setup options were all default 2019-06-14 08:52:25 eth0 is dhcp 2019-06-14 09:36:36 gonna be frank 2019-06-14 09:36:45 alpine installs way more easily on a laptop than adelie. 2019-06-14 09:49:43 using adelie is an option 2019-06-14 09:52:46 hi how does Alpine generally resolve upstream issues that cause code not to work well on alpine specifically? 2019-06-14 09:52:51 Example: https://github.com/docker-library/openjdk/issues/286 2019-06-14 09:53:07 jmap on alpine doesn't work well, and there's a fix, but, the fix can't be maintained in alpine 2019-06-14 09:53:25 and I assume upstream won't have it either - and anyway won't be applicable to older jvm versions. If upstream takes it, it will be in one of the newer jvms 2019-06-14 09:53:33 we usually open an issue upstream if patching it is non-trivial 2019-06-14 09:54:02 The problem arises for me because I want to interpret a jvm core dump that was taken on openjdk8-alpine 2019-06-14 09:54:33 Ideally I'd integrate the fixed jmap functionality in official alpine openjdk packages 2019-06-14 09:55:05 one second, I'm looking up the Dockerfile for openjdk:8-jdk-alpine 2019-06-14 09:56:43 this is the APKBUILD (synonymous with for instance Arch PKGBUILD): https://git.alpinelinux.org/aports/tree/community/openjdk8/APKBUILD 2019-06-14 09:56:56 the patchset we use can be found @ https://git.alpinelinux.org/aports/tree/community/openjdk8?h=master 2019-06-14 09:56:59 I guess I need to convince someone to take the patch ;/ What I am really asking is "please someone update this patch every time we update openjdk in alpine" 2019-06-14 09:57:08 so I need to beg openjdk maintainer I guess 2019-06-14 09:57:24 correction: https://git.alpinelinux.org/aports/tree/community/openjdk8?h=3.9-stable 2019-06-14 09:58:19 patching the openjdk8 package in alpine is an option as long as it isn't detrimental to other packages that depend on it in the alpine ecosystem 2019-06-14 09:58:54 Right. but it means maintainer has to do extra work every time openjdk has to update (due to security patch for example) 2019-06-14 09:59:16 danieli: If I wanted to use alpine packages on adelie, what kind of terrible trouble am I looking forward to? 2019-06-14 09:59:35 I'm not certain what adelie would do different if it's also using apk, openrc, etc 2019-06-14 09:59:42 priontology: really depends, for some packages it should work fine out of the box, for other packages it could break terribly 2019-06-14 10:00:19 adélie does a fair amount of things differently to how they are in alpine, you could ask for details in their IRC channel 2019-06-14 10:00:30 truth; you were just around for now ;D 2019-06-14 10:00:43 was sad to see that iwd isn't included in their repo 2019-06-14 10:00:48 truth be told i haven't kept up to speed with recent developments in adelie 2019-06-14 10:01:12 codehotter: good point, any suggestions or ideas? 2019-06-14 10:01:33 I'm not sure how to contact the maintainer to plead with them 2019-06-14 10:40:17 Hey guys I have a weird problem 2019-06-14 10:40:41 I have deleted from inittab all ttys besides 0 and 1 and enabled cgroup hierarchy 2019-06-14 10:40:45 now I don't have apt package manager 2019-06-14 10:40:46 what do 2019-06-14 10:40:52 how it happend no idea 2019-06-14 10:40:55 <_ikke_> you mean apk? 2019-06-14 10:41:01 oh 2019-06-14 10:41:04 I have 0 iq 2019-06-14 10:41:14 I always think of apk as .apk on android 2019-06-14 10:41:16 and type apt 2019-06-14 10:41:19 <_ikke_> heh 2019-06-14 10:41:27 <_ikke_> I believe Alpine was first with apk 2019-06-14 10:41:46 I haven't heard about alpine until 2019 2019-06-14 10:42:30 <_ikke_> it was released on 2005 2019-06-14 10:43:12 <_ikke_> s/on/in 2019-06-14 10:44:19 i used to associate 'apk' with android as well, but not anymore 2019-06-14 11:00:18 Is it possible to run the last command with ash like on bash with "!!" ? 2019-06-14 11:01:39 Yeah now I am realizing how bad windows terminal is for remote work, at first I thought dropbear fucks some stuff up, but yeah openssh also changes colors of my terminal completely 2019-06-14 11:05:35 <_ikke_> kubast2: as far as I know, no 2019-06-14 11:05:40 ash does not support !! 2019-06-14 11:06:16 and uh.. the new windows terminal isn't ready for public consumption or to be a daily driver yet as far as i know 2019-06-14 11:07:58 yeah I meant the old one... 2019-06-14 11:08:04 conhost/cmd 2019-06-14 11:59:25 So it turns out that adelie hadn't patched grub so that I could boot with uefi 2019-06-14 11:59:27 ;-; 2019-06-14 11:59:55 Now, with that in mind, is there an easy way that I can build alpine myself with packages I care about built in? 2019-06-14 12:00:37 <_ikke_> with a little bit of setup, it should be fairly trivial (from the alpine side) 2019-06-14 12:00:53 <_ikke_> getting projects to compile properly is a another thign 2019-06-14 12:01:12 <_ikke_> https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package#Setup_your_system_and_account 2019-06-14 12:03:32 Ah, I must have phrased it badly 2019-06-14 12:03:47 I mean to pre-include packages that I care about that are already in the edge repos 2019-06-14 12:03:51 and preconfigure them before installing, etc 2019-06-14 12:04:42 <_ikke_> If you use run-from-ram, you can create an apkovl that contains all the things you need 2019-06-14 12:06:46 Oh, interesting. 2019-06-14 12:07:14 I'll have to mess with that later today. 2019-06-14 12:12:18 Hi guys! :) Can somebody tell me how to persist env variables beyond sessions? 2019-06-14 12:12:31 I´ve tried with /etc/profile, but it requires a login 2019-06-14 12:12:46 and with /etc/environment, like CentOS / RHEL it doesn´t persist 2019-06-14 12:12:59 WhitePantherA: it depends on the shell you are using 2019-06-14 12:13:03 ash 2019-06-14 12:13:13 alpine + docker + kubernetes 2019-06-14 12:13:25 <_ikke_> For what do you need the env variable to persist? 2019-06-14 12:13:26 tried `~/.profile'? 2019-06-14 12:13:53 Yes, danieli, but i need to pass the "--login" and the app is a java that takes the info from the "env" output 2019-06-14 12:14:00 _ikke_: an app needs this... 2019-06-14 12:14:12 <_ikke_> Are you manually running this app? 2019-06-14 12:14:18 <_ikke_> (ie, not a service)? 2019-06-14 12:14:53 with a dockerfile (pipeline with Jenkins, etc.) 2019-06-14 12:15:01 the line that executes is this 2019-06-14 12:15:05 WhitePantherA: you need to set ENV, variable 2019-06-14 12:15:18 ENTRYPOINT ["java", "-cp", "app:app/lib/*", "com.name.of.app"] 2019-06-14 12:15:36 Yes but the ENV variables must be manually set with a script, because it takes info from a file 2019-06-14 12:16:04 modify the java application to take that configuration options from elsewhere than the environment? 2019-06-14 12:16:12 s/options/option/ 2019-06-14 12:16:14 I´ve a container, with a file, need like an "export" worlwide 2019-06-14 12:16:19 Mmmm.... I´m not the software architect 2019-06-14 12:16:20 nor lead 2019-06-14 12:16:23 just devops 2019-06-14 12:16:27 fair enough 2019-06-14 12:16:30 so, impossible to get a change in that level 2019-06-14 12:16:39 how would you solve it with any OS other than Alpine? 2019-06-14 12:16:59 In ubuntu lets say: writing some /etc/environment 2019-06-14 12:17:03 like Variable=Value 2019-06-14 12:17:38 <_ikke_> I would create a simple wrapped script that sets these variables and launches the application 2019-06-14 12:17:45 <_ikke_> this is quite common for java applications 2019-06-14 12:17:48 sounds like the best option, yes 2019-06-14 12:17:55 Yes, but the app runs at the entrypoint 2019-06-14 12:18:07 can you not change the entrypoint? 2019-06-14 12:18:13 If I run the script at the entrypoint i dont have the file containing the info to take the data from 2019-06-14 12:18:26 all that info comes from a container run in helm, using jenkins 2019-06-14 12:18:36 danieli: entrypoint must be like that 2019-06-14 12:18:55 i should add a script once the last container starts and spread the variables across the container 2019-06-14 12:19:26 wonder if busybox ash looks at the ENV environment variable 2019-06-14 12:19:26 can't you change the helm charts (if thats the right word) to set the desired environment for that container? 2019-06-14 12:20:08 kahiru: yes i do, in fact i did, but when execute the script variables are not accessible "worldwide" the node 2019-06-14 12:20:13 danieli: no, it doesn´t 2019-06-14 12:20:30 and when i run it from /etc/profile.d I must set --login 2019-06-14 12:20:41 it means, variables accessible from the session 2019-06-14 12:20:45 not the "system environment" 2019-06-14 12:21:03 https://git.busybox.net/busybox/tree/shell/ash.c?h=1_30_1#n14232 2019-06-14 12:21:19 WhitePantherA: ENV contains a file that should be read for non-login shells 2019-06-14 12:21:19 it doesn't have the `if (login_sh)` check 2019-06-14 12:21:23 yes 2019-06-14 12:21:44 KatolaZ: where? :( 2019-06-14 12:22:07 danieli: could you please point me the line? 2019-06-14 12:22:08 ENV ENV /path/to/some/file 2019-06-14 12:22:09 WhitePantherA: you should set ENV equal to the file to be read at startup by non-login shells... 2019-06-14 12:22:16 that link should point you to the line, look at the URL 2019-06-14 12:22:18 just let it load 2019-06-14 12:22:26 ok 2019-06-14 12:22:27 ENV=~/.myprofile 2019-06-14 12:22:28 syntax highlighting and all the fancy stuff causes it to load slowly 2019-06-14 12:22:30 export ENV 2019-06-14 12:22:41 into /etc/profile 2019-06-14 12:22:42 line 14232 anyway 2019-06-14 12:23:24 awesome, it means if there´s no env set, so it´s going to get the profile 2019-06-14 12:23:25 lets say 2019-06-14 12:23:27 ENV="" 2019-06-14 12:23:34 no WhitePantherA 2019-06-14 12:23:47 it means that is does not load any other file for a non-login shell 2019-06-14 12:24:43 Ok, it should look like this 2019-06-14 12:25:06 FROM openjdk:8-jdk-alpine 2019-06-14 12:25:07 VOLUME /tmp 2019-06-14 12:25:07 ARG DEPENDENCY=target/dependency 2019-06-14 12:25:08 COPY ${DEPENDENCY}/BOOT-INF/lib /app/lib 2019-06-14 12:25:10 COPY ${DEPENDENCY}/META-INF /app/META-INF 2019-06-14 12:25:12 COPY ${DEPENDENCY}/BOOT-INF/classes /app 2019-06-14 12:25:14 COPY ./secrets-to-environment.sh /etc 2019-06-14 12:25:16 ENV ENV /scripts/my-variables 2019-06-14 12:25:21 please use a pastebin service to paste larger snippets of code 2019-06-14 12:25:25 it floods the IRC channel 2019-06-14 12:25:26 ok sorry danieli 2019-06-14 12:25:53 https://pastebin.com/TaPWqguX 2019-06-14 12:26:06 the ENV variable is read by busybox ash, so if your application is not executed by ash, I'm not sure it'll be of much help 2019-06-14 12:26:28 :( 2019-06-14 12:27:21 /etc/environment is read by a PAM module, and Alpine doesn't use PAM by default 2019-06-14 12:27:35 Yeap... been reading 2019-06-14 12:27:51 It wont work with the "ENV path-to-file" option 2019-06-14 12:28:18 if the entrypoint is java directly, then yeah, I don't think it will 2019-06-14 12:29:05 Maybe removing the entrypoint and run at starting in the helm chart both commands? 2019-06-14 12:29:11 1. Script && java ... 2019-06-14 12:29:24 ash --login java something 2019-06-14 12:29:37 Mmm.... 2019-06-14 12:29:39 awesome idea 2019-06-14 12:30:27 i´ll try 2019-06-14 12:30:28 thank you 2019-06-14 12:33:09 good luck, i'm sure you'll figure it out one way or another 2019-06-14 12:39:23 WhitePantherA: ENV=[path_to_file 2019-06-14 12:39:26 export ENV 2019-06-14 12:39:33 put that in /etc/profile 2019-06-14 12:39:51 Ok, all previous values from the configmaps or helm charts are ignored, right? 2019-06-14 12:40:02 o_O 2019-06-14 12:40:06 but /etc/profile isn't loaded in non-login shells 2019-06-14 12:40:12 that's part of the issue KatolaZ 2019-06-14 12:40:26 Yeap, i was thinking what danieli said 2019-06-14 12:40:51 danieli: it is loaded by the login shell 2019-06-14 12:40:59 then the login shell has ENV set 2019-06-14 12:41:09 but they're not using a login shell 2019-06-14 12:41:17 when a non-login shell is spawned, if it finds ENV set, it reads the correpsonding file 2019-06-14 12:41:26 oh OK 2019-06-14 12:41:40 is that a chroot? 2019-06-14 12:41:49 almost, a container 2019-06-14 12:42:20 I see 2019-06-14 12:42:39 depending on the kind of container, it should be possible to set a variable on attach 2019-06-14 12:47:19 hi, when constructing an alpine installation in chroot, the static apk tool displays the error "can't create /dev/null: Permission denied". this is already triggered by baselayout which is the third package to install. I wonder if these cause some critical errors later on 2019-06-14 12:47:57 or is it safe to disregard them for the basic packages. the errors disappear after binding /dev for the chroot 2019-06-14 12:49:15 Either rbind /dev/ or create the nodes as described in the Wiki for the chroot install 2019-06-14 12:49:37 ok, so bind before calling apk 2019-06-14 12:51:12 Cogitri: I used this build as an example and it seems to bind only after calling apk https://github.com/alpinelinux/alpine-chroot-install/blob/master/alpine-chroot-install (lines 358-) 2019-06-14 13:06:38 Huh, that should be reported a b.a.o then, I suppose. Am on phone right now though, so I can't take a closer look at it :) 2019-06-14 13:07:06 Ah, algitbot doesn't function here? bugs.alpinelinux.org it is then 2019-06-14 13:13:35 ok, i'll do that soon 2019-06-14 14:14:00 hey community. recent alpine comes with php7.2 ... how do i get 7.3 ? 2019-06-14 14:14:27 7.3 should be available in edge within the next few hours 2019-06-14 14:14:55 and it will also be in 3.10, which is about to be released within the next few days 2019-06-14 14:14:57 <_ikke_> for most arches, it should already be there even 2019-06-14 14:17:02 bratkartoffel thanks. guess i just wait then XD 2019-06-14 19:58:20 Do I have to be on Edge for wireguard? 2019-06-14 20:02:35 wozencroft: yes 2019-06-14 20:02:43 it'll be in the upcoming 3.10 release though 2019-06-14 20:03:52 danieli: I just saw there are modules for vanilla since Wednesday 2019-06-14 20:04:28 So it does not work yet with vanilla, only from the next release on? 2019-06-14 20:05:29 I meant packages above 2019-06-14 20:06:04 we don't do much except bug/security fixes for packages in previous releases 2019-06-14 20:06:52 'vanilla' usually refers to a non-standard flavor of the linux kernel btw 2019-06-14 20:07:59 danieli: Ok, I see, so I need the edge kernel for a test atm. Thank you for the hints 2019-06-14 20:08:13 no, you need to run edge, it's not about the kernel 2019-06-14 20:08:39 wireguard does not exist on any repositories except for the (unreleased and upcoming) 3.10 ones and the edge ones 2019-06-14 20:09:41 Yes, edge repository with the kernel from that edge/testing repo - not the stable one. Have to build it first... 2019-06-15 07:57:45 what is the alpine equivalent of /dev/disk/by-id 2019-06-15 08:03:22 I don't know if there is... If you are just trying to UUID mount in fstab, you can use blkid to find the UUID 2019-06-15 08:06:45 zcrayfish: i have a virtual partition attached to the server 2019-06-15 08:07:36 its ID shows in the control panel so i need to see which is which inside alpine 2019-06-15 09:14:58 How can I get go installed in 3.9 x86_64 ? 2019-06-15 09:15:28 eject_ck: `apk add go' 2019-06-15 09:15:34 no packages, 2019-06-15 09:15:39 you're going to need the community repository enabled 2019-06-15 09:15:43 disread - community enabled 2019-06-15 09:15:44 sorry 2019-06-15 09:15:46 thanks! 2019-06-15 09:15:48 https://pkgs.alpinelinux.org/packages?name=go&branch=edge&arch=x86_64 2019-06-15 09:15:54 you're welcome 2019-06-15 09:22:42 I have issue running java ... 2019-06-15 09:22:54 ./java 2019-06-15 09:22:54 -ash: ./java: not found 2019-06-15 09:23:05 in fact binary is there 2019-06-15 09:23:25 should be something Alpine Linux specific ) 2019-06-15 09:23:26 run `java' and not `./java' 2019-06-15 09:23:41 java 2019-06-15 09:23:41 -ash: java: not found 2019-06-15 09:23:53 then you need to install the java version of your choice 2019-06-15 09:24:33 From aports.. 2019-06-15 09:26:53 how can I add aports repos? 2019-06-15 09:27:06 modify /etc/apk/repositories 2019-06-15 09:27:28 note that only main and community are available in stable releases, testing is only available in edge (rolling release/development branch) 2019-06-15 09:29:07 what's URL I mean, sorry 2019-06-15 09:29:28 https://wiki.alpinelinux.org/wiki/Aports_tree 2019-06-15 09:29:35 no clue from this page ) 2019-06-15 09:29:56 here's an example for the community repo https://wiki.alpinelinux.org/wiki/Enable_Community_Repository 2019-06-15 09:30:32 eject_ck: https://beta.docs.alpinelinux.org/user-handbook/0.1a/Working/apk.html#_repositories_and_mirrors 2019-06-15 09:30:41 yes, but where is there link to aports? 2019-06-15 09:31:01 there should already be enabled repositories in /etc/apk/repositories 2019-06-15 09:31:06 ouch 2019-06-15 09:31:08 gotcha 2019-06-15 09:31:10 thanks] 2019-06-15 09:31:12 if not, you could add a URL manually (see the guides), or run setup-apkrepos 2019-06-15 09:31:17 I have it 2019-06-15 11:49:00 Is there something wrong with xfce on edge? 2019-06-15 12:01:48 i dont like the deluge version bump at all. cant seem to use it correctly as a thin client, it isnt whitelisted on private trackers 2019-06-15 12:03:07 its in testing/deluge so i guess i kinda asked for it, but still 2019-06-15 12:03:35 needs a fair amount of testing :p 2019-06-16 19:55:57 maybe not Alpine related, but anyone know how to get blink cursor in st (suckless) term? setting or changing source, however 2019-06-16 22:39:07 Hi team, haproxy has released v2.0 how do we go about updating the existing package? 2019-06-16 22:40:26 terror: did you check if it build on alpine/musl 2019-06-16 22:42:19 hmm, 'Target 'linux2628' was removed from HAProxy 2.0 due to being irrelevant and' 2019-06-16 22:46:15 it build as target=generic but with a lot of warnings 2019-06-16 22:55:03 with 'TARGET=linux-musl' it builds on x86_64, armv7 and aarch64 2019-06-16 22:55:33 would be nice to have it in v3.10, haproxy 2.0.0 is LTS 2019-06-16 22:58:31 I'll see if I can get i touch with the haproxy team to add a musl target 2019-06-16 23:00:41 It is already there, I built it with TARGET=linux-musl 2019-06-16 23:02:11 hmm, not in makefile, how it worked then :\ 2019-06-16 23:02:33 s/:\/:|/ 2019-06-16 23:13:38 mps: make TARGET=generic USE_PCRE=1 USE_OPENSSL=1 USE_ZLIB=1 USE_LUA=1 USE_NS=1 LUA_LIB=/usr/lib/lua5.3/ LUA_INC=/usr/include/lua5.3/ 2019-06-16 23:13:46 This works fine for me on edge, no warnings 2019-06-16 23:15:41 maybe 'TARGET=custom'. Tried with 'generic' and the result is same as with 'linux-musl' 2019-06-16 23:17:19 your flags are nearly same as in APKBUILD 2019-06-16 23:19:28 https://pastebin.com/gSJWp8af 2019-06-16 23:21:08 looks good. 2019-06-16 23:22:25 I can check it again, but tomorrow, with hope it is not too late to push it for v3.10. It is late here and I'm going to bed 2019-06-16 23:25:23 no rush mps 2019-06-16 23:25:26 Thanks! 2019-06-16 23:25:43 We should probably contribute a musl target upstream 2019-06-16 23:32:22 ok. see you tomorrow. good nite 2019-06-17 00:54:18 Actually the 'linux-glibc' target builds fine. Just change TARGET=linux2628 to TARGET=linux-glibc. Also USE_PCRE_JIT=1 works in case it's useful (USE_PCRE_JIT: enable JIT for faster regex on libpcre >= 8.32) 2019-06-17 07:07:48 terror: good morning, thanks for info. USE_PCRE_JIT=1 could be good to add (if it works on all arch's) 2019-06-17 07:09:43 I would rather go with TARGET=linux-musl (or custom or generic) instead of TARGET=linux-glibc for 'aesthetic' reasons :) 2019-06-17 07:10:02 I understand, then you need to add all the USE_*. 2019-06-17 07:10:38 APKBUILD have them already 2019-06-17 07:11:39 No no, those USE_ are not enough. 2019-06-17 07:11:51 the linux-glibc defines more of them, check the Makefile 2019-06-17 07:12:08 I meant the linux-glibc TARGET 2019-06-17 07:12:22 TARGET=linux-musl USE_PCRE=1 USE_OPENSSL=1 USE_ZLIB=1 USE_LUA=1 USE_NS=1 LUA_LIB=/usr/lib/lua${_luaver} LUA_INC=/usr/include/lua${_luaver} CFLAGS="$CFLAGS -D__LINUX_NETFILTER_H 2019-06-17 07:13:48 ifeq ($(TARGET),linux-glibc) 2019-06-17 07:13:49 set_target_defaults = $(call default_opts, \ 2019-06-17 07:13:49 USE_POLL USE_TPROXY USE_LIBCRYPT USE_DL USE_RT USE_CRYPT_H USE_NETFILTER \ 2019-06-17 07:13:49 USE_CPU_AFFINITY USE_THREAD USE_EPOLL USE_FUTEX USE_LINUX_TPROXY \ 2019-06-17 07:13:49 USE_ACCEPT4 USE_LINUX_SPLICE USE_PRCTL USE_THREAD_DUMP USE_NS USE_TFO \ 2019-06-17 07:13:49 USE_GETADDRINFO) 2019-06-17 07:13:50 endif 2019-06-17 07:13:59 All those USE will need to be defined 2019-06-17 07:14:44 There's nothing stopping you from patching the Makefile but using linux-glibc is the easiest way. Or talk to upstream and ask for it to be renamed to just linux? 2019-06-17 07:15:38 don't have to redefine them, could just restructure the `ifeq' a bit 2019-06-17 07:17:29 danieli up to the maintainer, I wouldn't deviate from upstream for aestethic reasons 2019-06-17 07:17:30 but if it would be functionally equivalent, it seems a bit over the top to patch it just for it to look better 2019-06-17 07:20:45 all these build flags we can add in 'make ....' no need for patching 2019-06-17 07:22:26 that would be superfluous and would need fixes whenever the flags change 2019-06-17 07:23:54 danieli: true, but we can always clearly see what build flags are used. but I'm not against patch 2019-06-17 07:26:13 At the end of the day, it's up to the maintainer, I think it is ncopa 2019-06-17 07:27:19 I vote for use linux-glibc for now and talk to upstream if it offends you somehow 2019-06-17 07:27:21 :P 2019-06-17 07:27:48 right, -glibc is most complete and looks safe to use 2019-06-17 07:29:08 terror: we just discussed upgrade on #linux-devel, you can join there if you want 2019-06-17 07:30:04 I will build it in about then minutes and give it spin in real working server to see will it work 2019-06-17 07:30:08 #alpine-devel right? 2019-06-17 07:30:15 right 2019-06-17 07:30:41 sorry, just drinking first morning coffee ;) 2019-06-17 07:31:13 Cooking dinner over here 2019-06-17 07:32:24 heh, it's nice to have wife because of these issues ;) 2019-06-17 07:32:37 I do have one of those, gotta help though 2019-06-17 07:44:44 Don't tell anyone, but I think I'm switching back to alpine tonight or tomorrow. 2019-06-17 09:06:24 Hmm. Shouldn't most distributions run from ram? Why wouldn't you? 2019-06-17 09:07:14 because I have a lot of persistent data and software installed 2019-06-17 09:07:23 Because you want to save your work (documents or system upgrades) onto the disk for permanent storage 2019-06-17 09:11:40 to be fair, you can still do that if you're booted from a ramdisk 2019-06-17 09:34:04 Wait--when you say runs from ram in that context, you mean nothing gets flushed to disk? Huh. 2019-06-17 09:34:36 What packages are included in extended versus standard? 2019-06-17 09:34:42 You have to explicitly commit it to the disk 2019-06-17 09:34:53 yeah, it's just a hassle 2019-06-17 09:35:22 and to sensibly run my system from a ramdisk, i would need vastly more memory than my current 32G 2019-06-17 09:36:15 And it doesn't really provide too many advantages these days 2019-06-17 09:36:52 You get diminishing returns from faster storage, so if you compare loading times on an SSD vs on RAM you (usually) can't really notice a difference 2019-06-17 09:38:07 Huh. Interesting. When I install alpine to disk from the extended download, does that automatically configure it to run it from RAM? Is it just a setting? 2019-06-17 09:38:35 <_ikke_> priontology: depends on what you choose when you run alpine-setup 2019-06-17 09:39:01 extended or standard makes no practical difference 2019-06-17 09:39:12 and yes, it depends 2019-06-17 09:43:46 Neat. 2019-06-17 10:53:02 hello, i have a quick noob question: what options do setup-lbu and setup-apkcache need if i want to have everything on the sd card also used for setup-disk? 2019-06-17 10:54:15 the idea is that i can use the card most everywhere and cant install anything permanently without explicit action 2019-06-17 11:36:45 Had a problem this morning where openvpn didn't come up and it blocked sshd from starting. How do I make it optional for networking? 2019-06-17 11:37:00 (make openvpn optional) 2019-06-17 11:41:59 Oh, seems like I have a couple different options; make sshd depend explicitly on net.eth0, set rc_depend_strict="NO", or make openvpn "hotpluggable." 2019-06-17 19:00:36 Hello all. Trying to install Alpine Linux on my SuperMicro X7 motherboard via IPMI. When the system boots, I get a black screen on IPMI (but a normal monitor works OK). 2019-06-17 19:01:35 It seems the console video gets adjusted to 1024x768 - an odd mode that is not supported on my IPMI setup. After installing via normal monitor, I tried various boot arguments like "vga=normal" and "nomodeset", yet the system boots into a graphical console window. 2019-06-17 19:48:56 Got Plasma Desktop on X11 running on Alpine Linux 🎉 2019-06-17 19:49:20 Now I'll make Wayland work and then I'll PR it 2019-06-17 19:50:11 Nice 2019-06-17 20:47:13 Did someone try Ungoogled Chromium on Alpine? Or Iridium? 2019-06-17 21:04:54 TL;DR don't 2019-06-17 21:04:56 Use FF if you want a Browser without Google 2019-06-17 21:05:41 Ungoogled Chrome os at least one major version behind upstream chrome, so you trade being ungoogled for having some funny CVEs from time to time 2019-06-17 21:05:53 And Iridium usually lags behind even more 2019-06-17 23:51:49 my love for you is like a truck berserker 2019-06-18 00:02:21 lmfao 2019-06-18 00:02:33 priontology: just watched clerks again this weekend 2019-06-18 00:02:43 +1 2019-06-18 00:02:44 dude 2019-06-18 00:02:46 mepholic: lol 2019-06-18 00:02:50 also are you still working on adelie? 2019-06-18 00:02:55 yeah 2019-06-18 00:03:07 we're getting closer to 1.0 2019-06-18 07:55:54 I can't HTTP post to the Mandrill REST API from a Go app under Alpine 3.9, whereas it works on OS X. I get the following error: x509: certificate signed by unknown authority 2019-06-18 07:56:04 Does anyone know why this happens with Alpine? 2019-06-18 07:56:46 <_ikke_> Some CA missing in the CA store probably 2019-06-18 07:57:16 <_ikke_> You could try `apk add ca-certificates` 2019-06-18 07:57:48 _ikke_: Thanks! Trying 2019-06-18 08:02:44 It worked! 2019-06-18 08:02:46 Thanks again 2019-06-18 08:04:47 <_ikke_> np 2019-06-18 08:56:21 hi 2019-06-18 08:56:33 has php5 been dropped? 2019-06-18 08:57:43 I'm in edge and php5-cgi is gone 2019-06-18 09:02:32 ScrumpyJack: yes, removed, Date: Thu Nov 8 00:40:26 2018 +0200 'community/php5: remove aport' 2019-06-18 09:08:54 ScrumpyJack: i can recommend migrating from php 5 as soon as possible, it's EOL and running it is extremely risky 2019-06-18 09:09:10 short answer: yes. 2019-06-18 09:13:00 Couple of aaaaaa-moments last night and today. 2019-06-18 09:13:14 pciutils is being a dickhead. usr/share/hwdata/pci.ids is owned by hwids-pci 2019-06-18 09:13:28 so i've got a permaerror in apk because pciutils errors during install 2019-06-18 09:13:52 second aaaaa is that some packages seem to be failing to properly download from 42's mirror, and I'm unsure how i redo the repo selection process 2019-06-18 09:15:04 just randomly, it'll tell me temporary error, try again later, and i never had this before, so i assume it's a repo-being-a-dick error 2019-06-18 09:15:43 watch your profanity 2019-06-18 09:18:08 danieli: lmao 2019-06-18 09:18:23 Also, have you guys noticed that someone asking please on IRC seems a little demanding? 2019-06-18 09:18:29 Or is this just me? 2019-06-18 09:25:50 I can reproduce the pciutils problem. I had to uninstall it entirely to get rid of the error 2019-06-18 09:28:20 PureTryOut[m]: when do i need it? i forgot what installed it. 2019-06-18 09:28:28 i think it might be part of build-base? maybe i'm wrong. 2019-06-18 09:28:32 did you reinstall it afterward? 2019-06-18 09:29:51 `setup-xorg` installed it for me 2019-06-18 09:30:06 Reinstalling just brings back the error 2019-06-18 09:30:30 Problem is that pciutils depends on hwids-pci, so it'll always have the file conflict atm 2019-06-18 09:46:32 PureTryOut[m]: weird. Didn't have it in my last install. 2019-06-18 09:46:41 Why does it need to own that file? 2019-06-18 09:48:52 Idk about 3.9, but the issue is there on edge and probably on 3.10 too 2019-06-18 09:49:01 I don't know why it needs to own that file 2019-06-18 09:49:43 Ah, I'm on edge fully, so probably why 2019-06-18 09:49:52 What client are you using for matrix? 2019-06-18 09:50:12 Riot 2019-06-18 09:51:03 <_ikke_> priontology: so somehow both packages seem to provide that file 2019-06-18 09:51:12 <_ikke_> https://pkgs.alpinelinux.org/contents?file=pci.ids&path=&name=hwids-pci&branch=edge&repo=main&arch=x86_64 2019-06-18 09:51:20 <_ikke_> https://pkgs.alpinelinux.org/contents?file=pci.ids&path=&name=pciutils&branch=edge&repo=main&arch=x86_64 2019-06-18 09:51:40 uh oh 2019-06-18 09:51:53 so why can't they share? 2019-06-18 09:52:12 <_ikke_> Only one package can own a specific path 2019-06-18 09:52:17 <_ikke_> (file) 2019-06-18 09:52:33 <_ikke_> hwids-pci *only* provides that file 2019-06-18 09:52:48 lmao 2019-06-18 09:52:59 <_ikke_> It might be that pciutils started bundling that file 2019-06-18 09:53:11 <_ikke_> As it is now, the dependency on hwids-pci is not required 2019-06-18 09:54:13 <_ikke_> So 2 possible solutions: remove that file from the pciutils package or remove the dependency on hwids-pci 2019-06-18 09:55:11 I'd opt for removing the file from pciutils so we don't have to conflict with hwids-pci 2019-06-18 09:55:28 why not make the package creators themselves duke it out 2019-06-18 09:55:31 possibly over shrimp buffet 2019-06-18 09:55:53 <_ikke_> The package creators or upstream? 2019-06-18 09:56:07 also, so if i show my routing table, even if wlan0 is not connected, it shows that 192.168.1.0/24 is on wlan0 and eth0 2019-06-18 09:56:16 because i previously HAD wlan0 connected, disconnected, then connected eth0 2019-06-18 09:56:21 will that cause problems? 2019-06-18 09:56:43 shouldn't matter if the interface is down 2019-06-18 09:56:45 _ikke_: the individuals that created hwids-pci and pciutils themselves, who bundle the file 2019-06-18 09:57:02 danieli: Ah. And I double-checked it was down with ifdown, which said it was unconfigured 2019-06-18 09:57:13 but I wasn't able to ping 192.168.1.1 with eth0 plugged in and an obvious dhcp address 2019-06-18 09:57:14 so huh 2019-06-18 09:57:26 i've preferred using `ip' for a good while now 2019-06-18 09:57:30 it's decent 2019-06-18 09:57:31 ah what the fuck maybe iwd was still doing it 2019-06-18 09:57:43 what do you do for ip? ip wlan0 down? 2019-06-18 09:57:47 ip link 2019-06-18 09:57:57 or `ip l' 2019-06-18 09:58:17 well.. verbatim: `ip link set down' 2019-06-18 09:58:46 weird; you'd think that me telling iwd to disconnect would have made that easier 2019-06-18 09:58:54 and that it would automatically tear down the iface for me 2019-06-18 09:59:11 <_ikke_> priontology: yes, that's the idea, I was just mentioning how they could solve it 2019-06-18 09:59:14 <_ikke_> rnalrd: ping 2019-06-18 09:59:29 _ikke_: ah ah ah 2019-06-18 10:10:19 Another question: Right now, I'm using "echo 500 > /sys/class/backlight/intel_backlight/brightness" to set my display brightness to a low level, 3500 for bright. However, I have to su instead of sudo for this. Why? 2019-06-18 10:11:16 Because sudo will only invoke `echo` with root permissions, not the redirection to that file 2019-06-18 10:12:45 yes 2019-06-18 10:12:47 The standard trick for that seems to be echo | sudo tee /sys/... >/dev/null 2019-06-18 10:12:51 echo 500 | sudo tee /some/file 2019-06-18 10:13:06 yup 2019-06-18 10:23:49 ah, interesting. 2019-06-18 10:24:00 of course, I _could_ just properly config awesome to do it. 2019-06-18 10:24:12 what are you all using for window managers? 2019-06-18 10:24:20 i think i already asked and heard i3, etc 2019-06-18 10:24:42 <_ikke_> I use awesomewm 2019-06-18 10:24:57 <_ikke_> Let me see what I use to change the brightness 2019-06-18 10:25:38 <_ikke_> apparently xbacklight 2019-06-18 10:25:47 priontology: you can install light apk to control (well) backliht from WM or CLI 2019-06-18 10:26:29 _ikke_: with a config in rc.lua to capture xf86brightness, or whatever? 2019-06-18 10:26:39 mps: do you maintain light, too? ;D 2019-06-18 10:26:52 <_ikke_> priontology: exactly 2019-06-18 10:27:07 i'm surprised that this isn't just automatically done by the display driver 2019-06-18 10:27:12 via the keys 2019-06-18 10:27:40 priontology: no, although I made pkg and eu (nick) also made one which was better than my so I pushed his in repo 2019-06-18 10:27:51 <_ikke_> priontology: http://tpaste.us/vb8O 2019-06-18 10:27:53 heh 2019-06-18 10:29:18 I have lua script for awesome which uses light as backlight control backend 2019-06-18 10:30:07 _ikke_ _o/ 2019-06-18 10:30:44 <_ikke_> rnalrd: apparently pciutils and hwids-pci conflict, while pciutils depends on the latter 2019-06-18 10:30:54 I use light because it worked on ARM's, all other light control programs was 'problematic' ime 2019-06-18 10:32:25 gotta run, i'll check later 2019-06-18 10:32:34 <_ikke_> sure, np 2019-06-18 10:33:37 _ikke_: Uuu, neat 2019-06-18 10:33:46 mps: Oh, that's interesting 2019-06-18 10:33:52 mps: What are you running on ARM? 2019-06-18 10:34:47 don't understand well. what do you mean by 'What are you running on ARM?' 2019-06-18 10:35:48 What ARM device of yours runs alpine and awesome 2019-06-18 10:36:25 ah, for working boxes I use chromebooks 2019-06-18 10:37:18 ARM32 (old one working box) is samsung exynos 5800 Peach Pi 2019-06-18 10:37:58 ah, interesting. 2019-06-18 10:39:11 ARM64 current 'workstation' is rk3399 Samsung One Plus (or is it Plus One, lazy to read exact name) 2019-06-18 10:41:57 eh, not too much lazy :), here is description of my current working box https://www.samsung.com/us/computing/chromebooks/12-14/xe513c24-k01us-xe513c24-k01us/ 2019-06-18 10:42:41 images and marketing babbling more then description, to be fair 2019-06-18 10:46:30 _ikke_: which theme are you using in awesome, by the way? 2019-06-18 10:46:42 <_ikke_> default 2019-06-18 10:46:50 <_ikke_> Not really sophisticated 2019-06-18 10:46:50 is the touch supported, mps? 2019-06-18 10:47:01 _ikke_: interesting. I'd like to solarize the colors 2019-06-18 10:47:09 i'd like some sort of unified solarized machine 2019-06-18 10:47:17 <_ikke_> I like minimalism 2019-06-18 10:47:38 sidenote, but checking in wireshark here, this samsung device 2019-06-18 10:47:45 is fucking spamming the network with arp requests for 192.168.1.1 2019-06-18 10:48:03 like fuck, you'd think it would get it 2019-06-18 10:56:03 _ikke_: how are you monitoring bat life and link state, etc? 2019-06-18 10:57:50 priontology: yes, touch works by fingers and pen device 2019-06-18 10:59:05 neato 2019-06-18 11:00:12 only have issue with alsa, cannot control it well 2019-06-18 11:01:12 and as _ikke_ told, simple awesome 'theme', only text no icons 2019-06-18 11:08:14 that's interesting. 2019-06-18 11:08:32 hmm. back later! eyes getting tired 2019-06-18 11:11:12 <_ikke_> priontology: I've added some widgets 2019-06-18 11:13:36 Hi! I see that 3.9 line contains maven-3.6.0, while edge is at 3.6.1. Will 3.9 be updated or how does versioning / releasing work in alpine? 2019-06-18 11:16:19 <_ikke_> stable branches mainly receive bug/security fixes 2019-06-18 12:45:12 Hi. I'm trying to use qemu-openrc initscripts. However, I'm getting a qemu-system-x86_64: Parameter 'type' expects a netdev backend type 2019-06-18 12:46:37 Here's my conf.d: http://sprunge.us/mVBYUl 2019-06-18 12:46:46 Has anyone experienced this sometime? 2019-06-18 12:47:45 $extra_args seems like it is not an issue, because the same error happens without it. 2019-06-18 18:43:13 Hello! I'm wondering what would be the best approach to creating an ISO image of Alpine which includes my applications/scripts. 2019-06-18 18:44:12 <_ikke_> there is a script in the aports repo for generating ISOs 2019-06-18 18:46:08 I saw "mkimage.sh" and the associated profiles, but I don't see how to include my own applications/scripts. 2019-06-18 18:46:34 <_ikke_> m-a-p: Maybe this helps a bit: https://wiki.alpinelinux.org/wiki/How_to_make_a_custom_ISO_image_with_mkimage 2019-06-18 18:46:41 <_ikke_> I have no experience with it, so I cannot really help 2019-06-18 18:49:16 Thanks, _ikke_, I'll look into it. 2019-06-18 22:52:20 What's everyone using for unicode coverage in terms of fonts? 2019-06-18 22:53:58 Noto 2019-06-18 23:00:17 PureTryOut[m]: Does it cover full width 2019-06-18 23:00:56 Idk tbh 2019-06-18 23:01:12 Wait, are those unicode characters? 2019-06-18 23:01:15 If so then yes 2019-06-18 23:01:28 I believe Noto is the only font that supports every unicode character ever 2019-06-18 23:13:59 that looked fine in Hack Regular Nerd Font Complete Mono 2019-06-19 00:03:48 Hey everyone! I'm new to Alpine, don't know much about it. I was searching for a Linux desktop for my non-tech savvy grandma with a 4 years old laptop. Do you think that Alpine would be a good option ? I want to do as little support as possible 2019-06-19 02:09:17 troll 2019-06-19 02:09:23 ? 2019-06-19 04:19:25 mps: hi come back 2019-06-19 07:05:08 I noticed when using surf (a webkit2gtk browser) on html5test.com, that there are no streaming codecs (vp9, aac, webm, etc) detected. Is that a missing dependency on my end? Is there a seperate package for codecs besides gstreamer and the gst-plugins-{good,bad,ugly}? 2019-06-19 07:33:07 xcko: you might still be missing gst-libav 2019-06-19 07:34:01 mixi: got it already. :/ 2019-06-19 08:29:37 yunfan: Hello! I hope you are good and everything is well. 2019-06-19 08:36:15 yunfan: 3.10 rc3 and up has support for your board. u-boot, serial console and ethernet should work but I'm not sure because I don't have board and can't test 2019-06-19 08:38:57 How come docker-compose isn't available in 3.9? 2019-06-19 08:39:08 It doesn't seem that edgy 2019-06-19 08:39:21 <_ikke_> Because it's still in testing 2019-06-19 08:39:39 docker-compose is? 2019-06-19 08:39:44 <_ikke_> yes 2019-06-19 08:41:19 Really, that's surprising to me 2019-06-19 08:41:40 <_ikke_> There were some dependency issues that affected it 2019-06-19 08:42:09 Oh, so docker-compose don't consider themselves to be in a testing phase? 2019-06-19 08:42:34 <_ikke_> No, that's just regarding Alpine Linux 2019-06-19 08:42:39 I see 2019-06-19 08:43:05 <_ikke_> It would help if you can test it on edge. 2019-06-19 08:43:19 <_ikke_> IF it works fine, it most like can be moved to community 2019-06-19 08:50:47 mps: well , i had police visited 2019-06-19 08:51:00 mps: and yes, safe right now 2019-06-19 08:51:22 mps: weill test that tonight 2019-06-19 08:58:14 If I do install docker-compose from edge, should I also install everything else from edge? Is the alternative prohibited like pacman -Sy is prohibited on arch? 2019-06-19 09:00:40 <_ikke_> yes, mixing edge and stable is not recommended due to dependency conflicts 2019-06-19 09:11:09 I see that edge repository is enabled in alpine:edge in accordance with https://wiki.alpinelinux.org/wiki/Edge, yet `docker run --rm alpine:edge apk add --update-cache docker-compose` fails. I feel a bit silly now asking, but should it be installable like that? 2019-06-19 09:11:23 <_ikke_> You need to include the testing repo 2019-06-19 09:11:41 Ah, of course 2019-06-19 09:11:48 <_ikke_> echo 'http://dl-cdn.alpinelinux.org/alpine/edge/testing' >>/etc/apk/repositories 2019-06-19 09:11:54 Thanks a lot, _ikke_ 2019-06-19 09:31:21 hello, I'm considering alphine as a desktop, not sure if that's a good idea to begin with but I see that in the wiki, a chroot with another distro is presented as a possible solution to running binaries linked against glibc 2019-06-19 09:31:29 how well does this actually work in practice? 2019-06-19 09:31:42 are there any potential common issues? 2019-06-19 09:32:44 I don't use it as of right now because Flatpak fills that hole for me, but a chroot usually works well enough too 2019-06-19 09:33:04 I'm not exactly sure how one would use it with ALSA, but I've always been using pa 2019-06-19 09:36:13 I'm using a chroot right now to run 32-bit Steam on a 64-bit only system, works well. I'm planning to move my laptop to Alpine Linux soon so that will also use that method, should give no problems 2019-06-19 09:36:24 nice nice 2019-06-19 09:36:49 I use Flatpak for that because that's a bit more comfortable 2019-06-19 09:43:05 I'm not particularly against flatpak but I'd personally be picky about what I run through it, given some things I've read about old vulnerabilities staying unpatched since it's the dev's labor to update library versions 2019-06-19 09:44:18 yunfan: usb doesn't work because usb drivers for sunxi are not built in current 3.10 rc's and probably will not work when it will be released (maybe even today) 2019-06-19 09:45:45 now, I can create uboot tarballs, and will continue to work on support for arm boards 2019-06-19 09:47:16 maybe even make separate apk of the kernel with multi platform support for armv7 boards 2019-06-19 11:07:22 Hello, i'm facing a bug with a binary, how can I prevent the package maintainer ? 2019-06-19 11:12:01 I've maybe found: https://bugs.busybox.net/ ? 2019-06-19 11:12:59 <_ikke_> guycaprio: prevent the package maintainer? You mean contact them? 2019-06-19 11:15:52 Yes, I did'nt find any specific email for udhcpc6 2019-06-19 11:18:33 <_ikke_> Depends if it's an issue with busybox itself (then you should contact the busybox mailing list), or the way alpine packages it, in which case it's best to open an issue at bugs.alpinelinux.org 2019-06-19 11:19:40 I'll contact busybox itself then, thank you 2019-06-19 11:21:46 !paste 2019-06-19 11:22:00 Is there a preferred pastebin? 2019-06-19 11:22:32 Anything that doesn't require Javascript just to show some simple text would be nice, like https://paste.sr.ht or https://bpaste.net 2019-06-19 11:23:01 <_ikke_> We have are own: http://tpaste.us 2019-06-19 11:23:20 you can do `apk add tpaste` too 2019-06-19 11:24:13 Oh that's pretty nice 2019-06-19 11:24:31 Okay, handy. I expect to get upgrades for available packages when I run apk upgrade, but it seems like some versions are being held back. http://tpaste.us/YMPP 2019-06-19 11:24:53 How do I figure out why? 2019-06-19 12:02:37 Maybe a little more useful at describing my issue. http://tpaste.us/vb8Q Note that simulating manually update boost-program_options to latest results in no action. 2019-06-19 12:05:13 try to disable the 3.9 repos 2019-06-19 12:06:18 others are tagged repos. @edge, @edgecommunity, @testing. 2019-06-19 12:06:28 I'm probably shooting myself in the foot by doing so. 2019-06-19 12:06:44 Don't try to mix stable and edge together 2019-06-19 12:12:01 I'm getting that idea. Historically, the only package I've been pulling from edge is pdns and friends. Currently, I don't think I have anything installed from edge (/etc/apk/world has no @s) 2019-06-19 12:13:03 Doesn't matter. Those packages might be linked to some other packages which have a higher version in edge than in stable 2019-06-19 12:13:10 It just causes all kinds of issues and shouldn't be done 2019-06-19 12:15:05 Well, short of redeploying this install, what are my options to unbreak it? I don't want to go to edge, I've removed all of the tagged instances of edge from my repositories file. All of the versions I want to upgrade are from v3.9 and not edge. 2019-06-19 12:58:13 made a crappy awk script that verifies all my versions match those in v3.9, except the 4 packages that won't autoupgrade. http://tpaste.us/oaWr 2019-06-19 12:58:35 I could have probably redeployed it by now. I should just do that. 2019-06-19 13:13:33 wozencroft: doesn't look like you have any routes over the wireguard interface? 2019-06-19 13:15:00 wozencroft: please next time use some text paste service, like tpaste.us 2019-06-19 13:16:02 mps: hastebin is a paste service, is it not? 2019-06-19 13:16:12 i use it frequently (assuming it's stable then) 2019-06-19 13:17:13 danieli: There seems to be a routing, but somehow the interface is not in use 2019-06-19 13:17:23 danieli: yes, but for such things text/plain is better, imo 2019-06-19 13:17:44 right 2019-06-19 13:17:59 oh, that's HATEbin, not HASTEbin 2019-06-19 13:18:21 mps: Ok, did not know tpaste.us but looks good 2019-06-19 13:18:30 wozencroft: are you sure that ports on wg0 are correct 2019-06-19 13:19:05 mps: How could I check that? 2019-06-19 13:19:25 and I don't see 'allowed ips:' on wg0 2019-06-19 13:21:47 Endpoint in /etc/wireguard/wg0.conf 2019-06-19 13:21:54 on both peers 2019-06-19 13:22:34 sorry, on [Peer] 2019-06-19 13:22:44 mps: I have Endpoint in .conf only for [Peer] 2019-06-19 13:22:58 ListenPort on [Interface] 2019-06-19 13:23:17 that's ok, i.e Endpoint 2019-06-19 13:23:57 I mean, ListenPort on local and Endpoint for peer 2019-06-19 13:24:14 mps: Using the wg show command, it shows the listening port though 2019-06-19 13:24:24 I see that they are different in you case 2019-06-19 13:24:26 for interface 2019-06-19 13:25:12 Indeed, 38228 and 51820 2019-06-19 13:25:45 uhm, no 2019-06-19 13:26:20 (what a bad paste service, I cannot copy text with mouse) 2019-06-19 13:26:38 mps: You're right, its 45780 in the output 2019-06-19 13:26:48 right 2019-06-19 13:27:11 and, endpoint is 60653 2019-06-19 13:28:03 and, look to your firewal. are these ports (UDP) allowed in 2019-06-19 13:28:34 it is easier to have it to be same port number, imo 2019-06-19 13:28:49 mps: Ok, thank you for the hints, that might be the issue with ports 2019-06-19 13:29:24 I will try to change those ports to be the same and see how it works 2019-06-19 13:30:23 yes, and don't forget firewal 2019-06-19 13:35:27 Ok, thank you 2019-06-19 18:27:30 so 3.10.0 is out \o/ 2019-06-19 18:28:13 <_ikke_> \o/\o/\o/ 2019-06-19 18:55:13 ncopa: a note about the sshguard package 2019-06-19 18:55:21 it does not ship the sample config file 2019-06-19 18:55:32 not a big deal, but it could be annoying 2019-06-19 18:55:40 (not having it at hand) 2019-06-19 18:59:31 any chance it's with sshguard-doc ? 2019-06-19 19:01:36 <_ikke_> that contains only man pages 2019-06-19 19:05:51 not much pkg's in Alpine have example files 2019-06-19 19:06:43 especially system and small pkg's 2019-06-19 19:07:51 yeah, needs some improvement on that front 2019-06-19 19:15:18 mps: mine was not a criticism :) 2019-06-19 19:15:53 I might actually send a patch to the package maintainer 2019-06-19 19:16:41 <_ikke_> KatolaZ: +1 2019-06-19 19:17:20 I do have to admit that the lack of sample config file in busybox inetd got me good the first time I ran it... there was some very subtle difference in syntax from the last inetd I used which caught me by surprise. :) 2019-06-19 19:17:28 KatolaZ: np, I even think we do not need examples 2019-06-19 19:18:55 for some programs examples are outdated for some they are not well written, etc 2019-06-19 19:44:53 mps: I mostly agree 2019-06-19 19:45:23 the only things that's odd is that the sample config file is referred to in the manpage ;) 2019-06-19 19:49:20 true, that is something which shouldn't be in well written man 2019-06-19 21:04:55 Congrats on getting 3.10 out the door! 2019-06-19 21:38:14 Is 3.10 out?? 2019-06-19 21:38:31 latest-stable repo still point to 3.9.3 2019-06-19 21:38:49 3.9.4 I mean 2019-06-19 21:40:12 it is released, but iso images and tarballs are not yet updated 2019-06-19 21:43:07 That's great, congrats team 2019-06-19 21:46:23 terror: thanks, also you helped 2019-06-19 22:04:31 mps: honestly it was nothing. 2019-06-19 22:04:47 I'll do it more often now I know it's useful 2019-06-19 22:06:39 of course, we all start with small things usually 2019-06-19 22:12:33 I've contributed a couple of packages in the past too 2019-06-19 22:13:20 hope mawk made it to 3.10 2019-06-19 22:15:25 mawk is still in testing 2019-06-19 22:26:13 How much does it surprise you guys that most "linux administrators" still don't know shit about basic unix utilities most of the time? 2019-06-19 22:30:41 none. 2019-06-19 23:13:43 anybody know when the downloads page will be updated for 3.10? 2019-06-19 23:18:43 alyx: properly tomorrow (UTC+1) during the day. The files are already available - http://dl-cdn.alpinelinux.org/alpine/v3.10/releases/ 2019-06-19 23:19:07 oh sweet, thank you 2019-06-19 23:19:17 np :) 2019-06-19 23:28:27 priontology: 0 2019-06-19 23:29:12 I showed someone some basic find | xargs one liner and blew their mind 2019-06-20 02:41:03 terror: lmao 2019-06-20 05:10:13 mps: ok i had take the board to work 2019-06-20 06:06:15 mps: are you there? 2019-06-20 06:09:16 yunfan: yes, and good morning or day, whatever time of day is in you TZ 2019-06-20 06:12:33 mps: my time is GMT+0800 2019-06-20 06:13:00 mps: and now i got my board in my hand and the alpine3.10.0 just downloaded 2019-06-20 06:13:10 mps: is that i just dd it to the sd card? 2019-06-20 06:15:52 no, create partition, make FS on first one, mount it, untar tar.gz, 'dd' u-boot, change/edit extlinux, attach serial console or hdmi and then boot 2019-06-20 06:24:18 mps: i had made terrible things.. 2019-06-20 06:24:49 ? 2019-06-20 06:26:14 mps: i had dd the uboot to my host's disk 2019-06-20 06:28:13 mistakes happens :( hope you have backup 2019-06-20 06:28:54 nope, but this host could run 24x7 2019-06-20 06:29:06 so i will find time to restore that 2019-06-20 06:30:57 the default kernel isn't compiled with kexec support, right? 2019-06-20 06:32:26 it is for s390x only, I think 2019-06-20 06:33:25 I see 2019-06-20 06:34:42 mps: ok, i had done untar and dd ubot, how to change extlinux for enabling tty console, i knew you had wrote a article on the wiki, but cant found that, i am reading this https://wiki.alpinelinux.org/wiki/DIY_Fully_working_Alpine_Linux_for_Allwinner_and_Other_ARM_SOCs#Creating_modloop_File 2019-06-20 06:36:00 yunfan: I didn't wrote any wiki article about install alpine on arm32 2019-06-20 06:36:53 although I contemplated to write short guide, and maybe even script to install it 2019-06-20 06:38:30 mps: well , i had run int 2019-06-20 06:38:37 and got the shell 2019-06-20 06:38:41 but not the normal one 2019-06-20 06:38:43 but for sunxi you have to edit /extlinux/extlinux.conf 2019-06-20 06:39:30 you need to edit "APPEND" line 2019-06-20 06:40:02 sorry, that might be the system on the sata disk 2019-06-20 06:40:05 APPEND modules=loop,squashfs,sd-mod,usb-storage,ext4,sunxi-mmc console=${console} 2019-06-20 06:40:11 will plugout and test again 2019-06-20 06:40:55 this line is what worked in my tests 2019-06-20 06:41:42 ok, trying 2019-06-20 06:47:41 same on ppc64le. don't know why kexec is disabled on certain arch, it's very helpful feature. 2019-06-20 06:49:21 so, only on IBM ;) 2019-06-20 06:51:13 mps: still got the rescure shell, here is the output log https://clbin.com/zVXAj 2019-06-20 06:53:40 yunfan: looks like the filesystem is damaged on your card 2019-06-20 06:54:15 mps: ok, you mean the vfat partition for uboot? 2019-06-20 06:54:51 vfat? you should use ext4, it is better 2019-06-20 06:55:32 ok, that's my first choice 2019-06-20 06:55:55 if you use vfat you have to add 'vfat' to extlinux.conf list of modules 2019-06-20 06:56:07 mps: and should i change the partition flags if i use ext4 2019-06-20 06:57:13 I think, no. At least I didn't. but you can 2019-06-20 06:58:55 now it works 2019-06-20 06:59:19 i got the normal login console and finally login 2019-06-20 06:59:54 congratulations :) 2019-06-20 07:00:04 mps: thanks for your hard work 2019-06-20 07:00:52 well, my hard works are on something else ;) 2019-06-20 07:01:44 i am very like alpine's size 2019-06-20 07:03:00 but when i try to use it as our company's base distrio for docker, i found the musl dependencies is till a problem for node and python enviromemnt 2019-06-20 07:03:12 yes, we are trying to keep Small, Simple and Secure 2019-06-20 07:03:53 btw, can i install alpine on my usb disk for dev machine? 2019-06-20 07:03:57 I have no much experience with Docker and python 2019-06-20 07:04:48 you can install it wherever you want, I have one usb as rescue system 2019-06-20 07:04:50 the last time i tried it were break on choosing usb storage 2019-06-20 07:05:27 mps: i use i3 wm and i am a backend developer, what i need is firefox webbrownser and openssh 2019-06-20 07:05:40 but i am a chinese, i still need the xim things works 2019-06-20 07:05:41 there is wiki page about installing non standard on preformated partition 2019-06-20 07:07:14 ok 2019-06-20 07:07:15 I'm not expert in these (xim and similar) 2019-06-20 07:07:42 back to the board, so the ethernet could work ? 2019-06-20 07:08:41 ethernet worked on my three boards, cubox one, and two leemaker boards bananapi and lamobo 2019-06-20 07:10:11 btw, nowadays I prefer Chinese makers of arm boards. they are mostly open and good quality 2019-06-20 07:10:58 mps: well it depends on which one, for all winner, i think its considerable, but i happened had 3 cubieboard 7 2019-06-20 07:11:19 the chip were not that open from my abillities 2019-06-20 07:11:52 well, business as usual :) everyone tries to make more money at first :) 2019-06-20 07:12:43 i dont know why they choose this policy, because by choosing that , they need to maintainling their own distribution 2019-06-20 07:12:47 rockchip (the company) is helpful, imo. 2019-06-20 07:13:04 i am very angry with their debian :[ 2019-06-20 07:13:51 I'm quite satisfied with my rk3399 based chromebook 2019-06-20 07:17:07 I'm worked with Debian for more than 20 years, but switched to alpine first because debian is slow on ARM 2019-06-20 07:17:36 (besides other 'quirks') 2019-06-20 07:18:20 mps: well i had a rk based chromebook 2019-06-20 07:18:45 really, which one 2019-06-20 07:18:55 mps: i am very satisfied with that, except if they could make the flash bigger 2019-06-20 07:19:40 my emmc (flash) is 32GB, quite enough 2019-06-20 07:20:24 of course I have another one in micro SD slot, 128GB 2019-06-20 07:20:57 mps: asus Chromebook Flip C100PA 2019-06-20 07:21:14 yunfan: do you run linux on it 2019-06-20 07:21:17 mps: its quite cheap when i buy it 2019-06-20 07:21:31 mps: nope, i run termux for android on it:D 2019-06-20 07:21:44 the chromeOS's web browser is better 2019-06-20 07:21:51 and the touchpad experiences 2019-06-20 07:22:17 and i had bought a samsung chromebook plus recently 2019-06-20 07:22:20 also arm chip 2019-06-20 07:23:04 hmm, I have alpine on two chromebooks, samsung exynos (32bit) and samsung one plus (64bit). firefox runs quite fine on both 2019-06-20 07:26:04 actually they are my workstations for work 2019-06-20 07:28:17 mps: how about the touchpad? 2019-06-20 07:30:34 ok, i found the ethernet works too 2019-06-20 07:31:59 touchpad works quite fine, and touchscreen also 2019-06-20 07:33:01 btw, can i test if docker works now on these arm boards 2019-06-20 07:34:30 I don't know anything about this 2019-06-20 07:49:35 hmm 2019-06-20 07:59:04 mps: it looks like the sata disk attached to the board wont be mounted to /media automaticlly 2019-06-20 08:00:06 but i could easilly mount it by manual 2019-06-20 08:01:47 yunfan: Alpine is not 'user friendly' for ordinary users, it is user friendly for 'real users' 2019-06-20 08:02:43 personally I don't like that the OS touch any device until I tell to it to do something 2019-06-20 08:03:03 mps: so if i want to use my sata disk, i should store the configure files on mmc and make my own rc script to mount the sata disk ? 2019-06-20 08:03:30 you can add it in /etc/fstab 2019-06-20 08:03:41 got it, i thought there was a bug 2019-06-20 08:04:31 or, install 'udevil' apk and then from cli you can mount/unmount usb flash, disks or other media devices 2019-06-20 08:05:31 well i could mount them by manual currently 2019-06-20 08:05:54 i just want to choose the sata disk as the destination when using setup-alpine 2019-06-20 08:06:18 currently it only allow me to choose from floopy usb mmc and none 2019-06-20 08:06:56 hmm, it should see 'sda' also 2019-06-20 08:07:05 then its a bug 2019-06-20 08:07:32 maybe you need to load ahci driver 2019-06-20 08:07:51 but i could just mount it by manual , 2019-06-20 08:08:02 and the partition on is is also ext4 2019-06-20 08:08:29 aha, ok. didn't tested this 2019-06-20 08:18:44 also reboot commands seems has problem 2019-06-20 08:18:50 it will hang there 2019-06-20 08:19:27 after it says Requesting system reboot 2019-06-20 08:19:29 [ 1226.077869] reboot: Restarting system 2019-06-20 08:20:26 last time I tried reboot worked. will test it again later 2019-06-20 08:22:08 i guess it might caused by sunxi-mmc 2019-06-20 08:22:39 because i then plugout the power and plugin, now i got the same filesystem error 2019-06-20 08:23:08 sounds like the driver wont reponse to the kernel's sig broadcasting 2019-06-20 09:33:07 mps: i found another bug, when i want to use the disk as sys mode, it told me canot find syslinux, so i change repositories of apk to enable community and edge's , and there were still not syslinux in the repo, only grub 2019-06-20 09:35:25 yunfan: syslinux doesn't work on ARM. setup- error message are harmless in this case 2019-06-20 09:36:18 mps: then that setup-disk script need to be change for this i think 2019-06-20 09:58:19 yunfan: probably, all these setup scripts need some work to clean and extend them 2019-06-20 15:22:11 Can alpine be used totally in live mode, to boot from external media and work on ram? 2019-06-20 15:22:24 <_ikke_> yes 2019-06-20 15:22:26 yes 2019-06-20 15:22:57 Oh, the download one gets is both installer and livecd? 2019-06-20 15:24:48 <_ikke_> The livecd is the installer as well 2019-06-20 15:25:12 <_ikke_> (the installer being a script that you run) 2019-06-20 15:26:12 yes, it's kinda like arch's 2019-06-20 15:26:24 you boot into a CLI 'live mode' system 2019-06-20 15:26:43 you can use it in a throwaway fashion, or you can install alpine from there 2019-06-20 15:28:10 Eh, so I cannot use full GUI and desktop in live mode? 2019-06-20 15:28:58 <_ikke_> You can, but you need to install it, Alpine Linux does not come with a graphical environment by default 2019-06-20 15:30:52 I see. Thanks. 2019-06-20 15:32:07 A bit of a letdown, but I guess I'l try it anyways... 2019-06-20 15:32:20 I have been looking for a replacement for this Slitaz distro. 2019-06-20 15:33:25 i used slitaz years ago 2019-06-20 15:33:59 and i don't see why it's a letdown, i can't possibly imagine people expect a super-minimal OS to come preinstalled with GUI 2019-06-20 15:34:05 I had used that for a long time. I liked it because of its real lightness (~50 Mb base), and I could use it virtually for everything I needed: from live backup tasks to even some daily work (just installing missing packages if needed). 2019-06-20 15:34:12 alpine and slitaz are somewhat similar, but they have different goals 2019-06-20 15:34:17 Even some few games. 2019-06-20 15:34:27 you could make your own alpine medium with some preinstalled / preconfigured packages 2019-06-20 15:34:34 should make for a decent slitaz replacement 2019-06-20 15:34:37 Oh, I'm reading 2019-06-20 15:34:45 i honestly don't miss tazpkg much after meeting apk 2019-06-20 15:35:16 <_ikke_> vrd1: like danieli said, that 50MB base image does not have room for any graphical environment 2019-06-20 15:35:30 To finish the slitaz rant, yes, some few games provided they don't need full GPU drivers installed. 2019-06-20 15:35:42 *but* 2019-06-20 15:35:47 needs drivers enough to run X11, but that's just about it 2019-06-20 15:36:01 Everyone: I'm reading indeed. 2019-06-20 15:36:05 your comments 2019-06-20 15:36:07 thanks 2019-06-20 15:36:16 Ok, just to finish: 2019-06-20 15:37:21 in recent years the slitaz guys don't seem interested anymore. Rarely gets truly updated, has problems with many ATI gpus, no x64 nor UEFI, and even their website is down right now. 2019-06-20 15:37:39 vrd1: you might have a look at TinyCore, if thatt's not considered a blasphemy in here ;) 2019-06-20 15:37:56 And, Slitaz *does* boot live by default in GUI with a 50 Mb image! 2019-06-20 15:38:02 I recently tested that. 2019-06-20 15:38:04 yeah, slitaz felt like it fell apart a bit over the years 2019-06-20 15:38:20 With a desktop seemingly based on XFCE 2019-06-20 15:38:57 (Ok, I think the truth is, the slitaz iso is 50 Mb, but it gets expanded while booting live) 2019-06-20 15:44:54 Have you heard about this YUMI tool for USB booting? 2019-06-20 15:45:39 <_ikke_> Doesn't ring a bell 2019-06-20 16:05:34 vri have, but it's been yeeears 2019-06-20 16:05:49 oh, they left, that's why tab completion faild 2019-06-20 16:05:52 failed* 2019-06-20 16:17:57 Hi ! wondering why mongodb was removed from 3.10 release? any hint? 2019-06-20 16:18:36 Esteban__: https://git.alpinelinux.org/aports/commit/testing?id=8a901de31fa055ed591d487e12f8bb9ffcc0df21 2019-06-20 16:18:39 community/mongodb: expunge nonfree software 2019-06-20 16:18:39 Upstream has switched to a nonfree license. 2019-06-20 16:19:20 Mongodb switched to a nonfree license 2019-06-20 16:19:23 yes, SSPL 2019-06-20 16:19:31 It was dropped from debian, void, fedora and others as well 2019-06-20 16:20:09 naturally, yes 2019-06-20 16:23:26 thanks guys ! 2019-06-20 17:22:12 a newbie question: looking at the corresponding aports, it seems that most config files are simply rewritten on package upgrades, regardless of whether they had been changed locally 2019-06-20 17:22:18 is this normally the case? 2019-06-20 17:23:07 (have looked only in openssh/APKBUILD and nginx/APKBUILD 2019-06-20 17:24:06 what are you trying to achieve? 2019-06-20 17:24:22 danieli: upgrade to 3.10 without loosing my locally modified config files 2019-06-20 17:24:30 I know I can back them up 2019-06-20 17:24:39 right, i'm *reasonably* sure it doesn't overwrite the files 2019-06-20 17:24:43 _ikke_: able to confirm? 2019-06-20 17:24:56 it does not look like that from the APKBUILD files 2019-06-20 17:25:01 but maybe apk has some maginc 2019-06-20 17:25:09 ~magic 2019-06-20 17:25:11 ;) 2019-06-20 17:34:13 I think we have .apknew-$(date) for new config files? 2019-06-20 17:48:13 ok, it looks like the answer is in apk-tools/database.c:apk_db_migrate_files 2019-06-20 17:48:22 sorry for the noise, I will read the code :) 2019-06-20 17:56:11 KatolaZ: if user changes config files in /etc upgrade doesn't overwrite them 2019-06-20 17:57:05 new files are installed with added .apk-new 2019-06-20 17:57:16 yep 2019-06-20 17:57:23 I read that, thanks 2019-06-20 17:57:41 I guess that's because /etc has APK_PROTECT_ALL set 2019-06-20 17:57:53 there is update-conf command which does this if user wants to update conf files 2019-06-20 17:57:56 I don't know where that happens, yet, but I will find out 2019-06-20 17:58:04 ok 2019-06-20 17:58:06 thanks mps :) 2019-06-20 17:58:15 very useful 2019-06-20 17:58:57 update-conf -l will show updated conf files 2019-06-20 17:59:43 but, backup them in any case before upgrade if you want to be safe 2019-06-20 18:00:31 mmhhh 2019-06-20 18:00:40 either you trust the code or you don't, right? :P 2019-06-20 18:04:18 "In god we trust, everything we check with PGP", old proverb :) 2019-06-20 18:04:46 s/everything/everything else/ 2019-06-20 18:08:03 I am more on the "everything ww check with PGP. Period." side :D 2019-06-20 18:08:08 s/ww/we/ 2019-06-20 18:10:59 Currently reading the wiki 2019-06-20 18:11:09 I like metaphors 2019-06-20 18:12:44 Erm, sorry, I kept wondering, if there's no way to have a more or less complete desktop environment within 50-60 Mb space, how would you think Slitaz people did? Their ISO still is ~50 Mb! 2019-06-20 18:15:26 What DE do they use? 2019-06-20 18:16:58 Openbox window manager 2019-06-20 18:17:19 and use Busybox 2019-06-20 18:17:26 alpine uses busybox by default 2019-06-20 18:17:43 either way, i have no clue how they build their ISOs 2019-06-20 18:18:11 Well, you can probably still do that with Alpine then. Openbox is available 2019-06-20 18:18:24 There is a portable version of CWM too. 2019-06-20 18:19:10 fvwm and twm are there too 2019-06-20 18:19:30 (cwm being the one OpenBSD uses) 2019-06-20 18:19:34 It really depends on what you mean by "full DE" though. 2019-06-20 18:19:35 and awesomewm, herbstluftwm, i3wm, bspwm 2019-06-20 18:19:39 WHAT, no xmonad!? 2019-06-20 18:19:59 i guess GHC being a pain in the ass on musl/alpine explains it 2019-06-20 18:24:35 I personally wouldn't call any of those DE's. Stuff like LXQt, XFCE, GNOME, Plasma, etc I however do 2019-06-20 18:26:51 You could say that Emacs is a full desktop environment and wouldn't be too far off. But most people would expect things like a web browser and just checking last Firefox binary gives me 54.87 MB compressed 157.08 MB actual. 2019-06-20 18:27:44 yeah, they're all WMs 2019-06-20 18:27:49 and a WM is just a component of a DE 2019-06-20 18:28:02 One of the simplest parts arguably. 2019-06-20 18:28:09 i wouldn't consider browsers a part of the DE for that sake 2019-06-20 18:28:43 i think i consider the golden standard something along the lines of a fresh XFCE/LXDE or gnome-shell install 2019-06-20 18:28:43 But where you draw the line? Is PDF viewer part of one? 2019-06-20 18:28:50 i wouldn't say so 2019-06-20 18:29:05 gnome has evince, it's part of the gnome project, but not really the core gnome DE 2019-06-20 18:30:03 display manager, window manager, 'bars'/application launchers/status bars, etc 2019-06-20 18:30:03 What is then? DE = WM + toolbars + filemanager? 2019-06-20 18:30:03 contemplating running alpine on my laptop....every where i read it states it is not recommended for laptops but servers & pi devices...this still hold up today? 2019-06-20 18:30:34 don't get me wrong, my ramblings about this are ~100% subjective 2019-06-20 18:30:43 i don't have an authoritative answer 2019-06-20 18:31:05 i could throw you a curveball and ask what software an OS consists of 2019-06-20 18:31:13 does it include userland? if so, how much? 2019-06-20 18:31:18 ralpheeee: Well, I use it for my laptop just fine 2019-06-20 18:31:28 Be mindful of musl compatibility though 2019-06-20 18:31:36 ralpheeee: you can run it on your daily driver machine just fine, but running certain software will be annoying 2019-06-20 18:31:53 and some of the technical tools i need for work are not yet (if ever) available on alpine, but that's just me 2019-06-20 18:32:08 but as Cogitri says, beware of prebuilt applications and applications that will not build cleanly on the musl libc 2019-06-20 18:32:21 danieli Cogitri : ahhh need to check packages then.... 2019-06-20 18:32:32 pkgs.alpinelinux.org 2019-06-20 18:32:33 I personally run Alpine on my laptop but mostly recommend Void to others for that purpose. 2019-06-20 18:33:04 i for the most part use use arch, fedora, debian, occasionally rhel, sles 2019-06-20 18:33:25 I guess Void glibC would be an easier option, yes 2019-06-20 18:33:27 CcxWrk: several small issues i have with void.... 2019-06-20 18:33:41 I'm going to run Alpine on my laptop once I have more of KDE packaged 2019-06-20 18:34:41 void can be proceed at glacial pace sometimes with annoying lingering pkg issues 2019-06-20 18:34:43 ralpheeee: I'm pretty sure there are a few. I wouldn't be *slowly* working on own distro/package manager if there was perfect one already. :P 2019-06-20 18:34:47 ralpheeee: i'm curious now 2019-06-20 18:35:26 a while ago i started work on a package manager written in ada, with ideas taken from all around the place 2019-06-20 18:35:38 such as guix and nix 2019-06-20 18:35:56 ostree 2019-06-20 18:36:05 (ostree is not a pkg manager but you get the gist) 2019-06-20 18:36:37 just annoying to read on the web alpine is more "trouble" to use on laptop/desktop ...but for educational purpose def worth while going down the rabbit hole 2019-06-20 18:36:45 danieli: Throw in slashpkg & Gobo/Sabotage in the mix, with configurability >= of Gentoo and you get the idea ;) 2019-06-20 18:37:40 I'm not sure why Alpine would be more trouble on laptop/desktop than other distros, besides a rather old kernel so missing new hardware support 2019-06-20 18:37:48 i never liked gobo much 2019-06-20 18:38:11 PureTryOut[m]: it's more or less about the difference between headless servers and user-facing systems with GUI 2019-06-20 18:38:25 it's just annoying to constantly have to work around software that fails horrendously on musl 2019-06-20 18:38:48 PureTryOut[m]: alpine kernel is not too old, look at some famous distros like debian 2019-06-20 18:38:59 It just lets bad software self-select itself out of your workflow. :P 2019-06-20 18:39:18 mps: I find the Debian kernel to always be too old lol 2019-06-20 18:39:22 danieli: Luckily that's not *that* much anymore though 2019-06-20 18:39:28 And we have patched lots of stuff now 2019-06-20 18:39:30 Anyways, that is manageable 2019-06-20 18:39:47 yeah, it's way better than it used to 2019-06-20 18:39:59 PureTryOut: I find everything in Debian too old lol 2019-06-20 18:39:59 Agreed 😂 2019-06-20 18:45:29 danieli: My main idea is to split the system configuration/build options/version selection part out (implemented in Prolog, with statically linked implementation and proper constraint solver, because this part of package management is nothing but constraint satisfaction problem) which generates package recipes for Sabotage core (which then uses just posix utils aside of one tiny parallel job runner in 2019-06-20 18:45:31 C) 2019-06-20 18:45:50 sounds like it will be a massive suite of software 2019-06-20 18:46:12 anyway, i like prolog, +1 for creativity 2019-06-20 18:46:36 I've grown tired of PMs reimplementing it very poorly. :D 2019-06-20 18:49:07 it's nice to be versed in languages like prolog, erlang, haskell 2019-06-20 18:51:51 Minimalism of base system is core consideration. Or rather result of easy bootstrapability and reproducibility. But sure, if you want to have full configurability for all the package versions and build options that complexity needs to be implemented somewhere. So I wanted to pluck that out into something fitting where it can be actually written as something that can be reasoned about. 2019-06-20 18:53:06 With Ada I've found the biggest pain point to be the lack of reasonable stdlib/OS interface. 2019-06-20 18:53:53 yeah, and there's a certain scarcity of libraries 2019-06-20 18:54:19 If I understand correctly, a DE does use a window manager on its own. If so, what window managers do KDE and GNOME use? Or their window managers don't really have a name on their own? 2019-06-20 18:54:32 scala had the same issues before https://github.com/lihaoyi/os-lib , you had to go through java 2019-06-20 18:54:57 vrd1: kwin and mutter respectively 2019-06-20 18:55:16 gnome used to use metacity but now it uses mutter 2019-06-20 18:55:25 i thought gnome-shell was the wm? 2019-06-20 18:55:29 Is that named mutter? 2019-06-20 18:55:41 Thanks 2019-06-20 18:55:42 that's the "graphical shell" 2019-06-20 18:55:57 it used to be fairly trivial to replace gnome's wm, but it's much harder now. 2019-06-20 18:56:04 oh yeah, they're more tightly knit now 2019-06-20 18:56:07 I can't find the right incantation to use stumpwm as gnome's wm anymore. 2019-06-20 18:56:15 vrd1: ISTR some lightweight DEs just used Openbox (by default). Since what WM should and should not do is fairly well specified there's little to make it a hard requirement to use specific implementation. 2019-06-20 18:56:18 the whole mentality is sort of reflected in devuan's struggles 2019-06-20 18:56:19 That's what I ran for a long time. 2019-06-20 18:56:29 well, the causes of devuan's struggles 2019-06-20 18:57:21 OT? 2019-06-20 18:59:17 to circle back, i'd say kdesktop is analogous to gnome-shell 2019-06-20 18:59:31 .. was.. analogous 2019-06-20 19:04:12 I was starting to quite like Enlightenment, until I was told the reason it doesn't get too much updates is mainly because Samsung stopped sponsoring them. 2019-06-20 19:05:30 After doing some projects on top of EFL professionally I kinda stopped liking it. :] 2019-06-20 19:07:01 Because of the fact I just mentioned? 2019-06-20 19:13:57 Nah, I think they were just freshly acquired by then. And decided that instead of caring about proper modular build they'd rather ship kitchen sink (because that's what the developers want!). Also I've seen the code up close. :D 2019-06-20 19:14:18 Don't talk to me about Genlist. :] 2019-06-20 19:51:28 Does Alpine Linux ISO support both UEFI and legacy BIOS? 2019-06-20 19:52:05 yes 2019-06-20 19:52:18 grub with efi, syslinux with bios 2019-06-20 20:00:03 Does it support a shim file for M$ signature, or one needs to disable secure boot? 2019-06-20 20:08:07 danieli: does it support a shim file for M$ signature, or one needs to disable secure boot? 2019-06-20 20:08:14 i replied because i don't know 2019-06-20 20:08:17 i mean 2019-06-20 20:08:24 i intentionally didn't reply because i don't know 2019-06-20 20:08:32 sorry, had a brainfart 2019-06-20 20:54:06 danieli: if using as run-from-ram system booting from a USB device, can I tell lbu to use the USB itself to save modifications? 2019-06-20 20:54:24 i recommend you ask open questions instead of directing them at specific people, i am a bit preoccupied at the moment 2019-06-20 20:55:06 Sorry 2019-06-20 20:55:11 If using as run-from-ram system booting from a USB device, can I tell lbu to use the USB itself to save modifications? 2019-06-20 23:06:09 Hey everyone. Since latest release, our builds on version 3.9 began to fail due to missing dependency "proj4-dev". I cannot even find it in the package repository anymore. How can I resolve this? 2019-06-20 23:06:27 proj4 was renamed to proj 2019-06-20 23:07:33 Ugh... I'll surely give it a try. But why has been this backwards-incompatible change made in older versions, too? 2019-06-20 23:08:36 no clue 2019-06-20 23:10:44 Latest switching from 3.9.4 to 3.10 might be related 2019-06-20 23:13:02 Yeah, seems like it's fixed by renaming the package. Thanks maxice8. Good night. 2019-06-21 03:27:14 when it is instructed to log, where would iptables log to? 2019-06-21 03:27:38 it's supposed to be /var/log/messages, but nothing shows up and I can see that the rule is seeing traffic 2019-06-21 03:27:47 syslog i would assume 2019-06-21 03:28:09 the syslog conf might need to be such to ensure that the relevant log level is saved 2019-06-21 03:36:18 i wonder what the default is 2019-06-21 03:46:22 that's not good 2019-06-21 03:46:31 I added "-l 5" and now nothing is logging at all 2019-06-21 03:48:35 boost that up to 7, now I'm getting... some things 2019-06-21 03:48:53 mps: yes i guess so, but the most important things prevent my using the board is the shutdown issue i mentioned yesterday 2019-06-21 03:50:56 so default is clearly 8, which means that iptables (logging at level 4) should be showing 2019-06-21 04:10:23 got this working with this: https://it-offshore.co.uk/linux/alpine-linux/55-alpine-linux-lxc-guest-iptables-logging 2019-06-21 04:21:12 <_ikke_> ah, you are it-ofshore on github? 2019-06-21 05:31:57 hello 2019-06-21 05:32:10 I'm trying to install alpine without nuking my home partition 2019-06-21 05:33:05 I'm still not quite sure what's the easiest way but I was gonna mount everything (maybe minus /home because it's not clear to me if it's gonna get formatted) in /mnt 2019-06-21 05:33:13 and then run setup-disk /mnt 2019-06-21 05:33:25 but I can't even mkfs.ext4 2019-06-21 05:33:48 however setup-disk apparently uses it, so it must be somewhere right? 2019-06-21 05:38:29 so I just tried sourcing /lib/libalpine.sh and now I have mkfs.ext4???????? 2019-06-21 05:39:31 right, apparently it was in /sbin but that's not in PATH by default 2019-06-21 05:46:39 now the partitions I make under fdisk don't even show under /dev 2019-06-21 05:47:03 that's just in the virtual machine though, I'm not actually going to make partitions when I actually install it 2019-06-21 05:47:20 but I can't test anything like this 2019-06-21 05:48:19 maybe fdisk didn't tell the kernel to re-read the partition table? 2019-06-21 05:52:56 how do I around that? If that were the case 2019-06-21 05:52:57 Hello.. check out http://0x0.st/zeRa.c . Program halts after compilation via gcc src.c && ./a.out 2019-06-21 05:53:04 Works fine under glibc 2019-06-21 05:53:30 also I keep getting this warning from fdisk, it could be related I guess https://0x0.st/zeRS.png 2019-06-21 05:54:55 well rebooting worked so I guess it's fine 2019-06-21 05:57:39 also I just realized that it is running setup-disk that adds mkfs.* to /sbin 2019-06-21 05:57:44 I really don't get the purpose of that 2019-06-21 05:58:58 does your installer image have hdparm? 2019-06-21 05:59:28 yeah 2019-06-21 06:00:10 "hdparm -z" was what I had in mind 2019-06-21 06:00:17 it tells the kernel to re-read the partition table 2019-06-21 06:00:25 I don't know if it will help, but it certainly won't hurt. 2019-06-21 06:00:50 like I said, I rebooted and it had the same effect, so yeah, the partition table wasn't getting updated 2019-06-21 06:21:31 Anyone checked http://0x0.st/zeRa.c? 2019-06-21 06:21:38 Fails on runtime 2019-06-21 06:28:07 errn0: well for one, you're not null terminating your string. 2019-06-21 06:28:59 errn0: well for one, you are allocating an array with one char, and expecting up to 6 2019-06-21 06:29:02 ... 2019-06-21 06:35:18 I did http://0x0.st/zeRV.c.. still stalls at runtime 2019-06-21 06:35:35 It's funny how glibc ignored that xD 2019-06-21 06:37:52 it's not stalling. change line 10 to printf("Enter: \n"); the line isn't refreshed if you don't put a new line. This can be worked around if you really want. 2019-06-21 06:39:34 if you don't want to put a new line, use fflush() on stdout. 2019-06-21 06:40:54 xcko: I did http://0x0.st/zeRV.c.. still stalls at runtime 2019-06-21 06:41:53 errn0: xcko | it's not stalling. change line 10 to printf("Enter: \n"); 2019-06-21 06:44:14 xcko: Whoa.. that worked.. 2019-06-21 06:44:23 Why glibc ignore that? 2019-06-21 06:44:37 This is interesting 2019-06-21 06:44:44 errn0: ignore what? 2019-06-21 06:44:48 your code was wrong 2019-06-21 06:44:50 not glibc 2019-06-21 06:45:06 yeah musl is behaving correctly 2019-06-21 06:45:12 o_O 2019-06-21 06:45:14 gcc don't care if I don't allocate memory properly in malloc stage 2019-06-21 06:45:24 errn0: this does not depend on gcc 2019-06-21 06:45:33 no compiler will complain for that 2019-06-21 06:45:44 in C *you* must know how you manage your memory 2019-06-21 06:45:46 o_O 2019-06-21 06:45:46 malloc(sizeof(char)) and then pass 6 bytes via fgets, still works xD 2019-06-21 06:45:57 errn0: oyu just don't understand 2019-06-21 06:46:01 I ment glibc.. not gcc xD 2019-06-21 06:46:11 meant* 2019-06-21 06:46:18 it does not work 2019-06-21 06:46:37 it is segfaulting at random, depending on where your program is put in memory 2019-06-21 06:46:48 it will segfault also with musl 2019-06-21 06:46:50 It does... I tried in a Ubuntu VM. xD 2019-06-21 06:46:58 But segfaults in musl xD 2019-06-21 06:46:59 ok, it's a waste of time :) 2019-06-21 06:47:09 errn0: it segfaults AT RANDOM 2019-06-21 06:47:23 it just depends on who own the chunk of memory you get to write to 2019-06-21 06:47:26 sometimes is you 2019-06-21 06:47:34 and the OS is fine with that 2019-06-21 06:47:42 I found ten dollars on the street on a rainy day. can't believe rainy days give you money. 2019-06-21 06:47:46 sometimes is not your process space, and the OS segfaults 2019-06-21 06:47:52 errn0: join ##c if you want more c help 2019-06-21 06:47:55 I see! 2019-06-21 06:48:11 errn0: maybe read "The C Programming Language" first? :) 2019-06-21 06:48:30 +1 for knr 2019-06-21 06:48:52 Gottox: I never found ten dollars on the street. I can't believe streets won't give you money easily... 2019-06-21 06:49:20 :D 2019-06-21 06:51:31 xcko: KatolaZ: Check this: http://0x0.st/zeRw.png 2019-06-21 06:53:00 And this: http://0x0.st/zeRw.png 2019-06-21 06:53:01 and just for the record, errn0 your "stall" was due to the way stdout / terms behave - nothing specific with c. 2019-06-21 06:53:35 errn0: this has nothing to do with the compiler, or the library 2019-06-21 06:53:38 nothing at all 2019-06-21 06:54:17 Oh alright 2019-06-21 06:55:39 mps: FTR, everything went fine on upgrade (re: existing conf files) 2019-06-21 06:55:54 mps: thanks 2019-06-21 07:03:52 KatolaZ: nice to hear 2019-06-21 07:04:48 mps: actually didn't have any doubt after readin the source of apk ;) 2019-06-21 07:05:00 but it's nice to report also what works, not just what does not work 2019-06-21 07:05:29 yes, you are absolutelly right 2019-06-21 07:05:35 the upgrade was quite seamless, to be honest 2019-06-21 07:05:39 I am positively impressed 2019-06-21 07:07:38 I'm right now upgrading one ARM box over RS485 cable/device and it is faster than upgrading debian over ethernet 2019-06-21 07:10:03 yunfan: I'm not in place where I can check if the current stable have problem with reboot on sunxi boards 2019-06-21 07:10:28 will check later and inform you about my tests 2019-06-21 07:14:04 mps: thanks very much, and i need to find a way to restore my mbr on the host :L 2019-06-21 07:16:51 if you saved it before (backup) then restoring shouldn't be problematic 2019-06-21 07:17:27 is it Alpine box or something else 2019-06-21 08:34:21 yunfan: you are right, reboot doesn't work 'out of the box' on current kernel for sunxi boards which I tested 2019-06-21 08:38:37 mps: shutdown either 2019-06-21 08:39:05 true, shutdown also doesn't work 2019-06-21 08:39:19 poweroff 2019-06-21 08:40:03 probably need to enable kernel option for that 2019-06-21 08:41:04 you can rebuild kernel with needed options for you, or use kernel and modules from arch linux 2019-06-21 08:45:57 mps: i dont know how :D 2019-06-21 08:46:08 or you could gave me a base config file 2019-06-21 08:59:09 yunfan: this wiki page could be good starting point https://wiki.alpinelinux.org/wiki/DIY_Fully_working_Alpine_Linux_for_Allwinner_and_Other_ARM_SOCs 2019-06-21 08:59:53 and try different options, as all of us who works with ARM SBC's 2019-06-21 09:00:19 this is good for learning, btw 2019-06-21 09:06:14 mps: can i compileing on alpine :D 2019-06-21 09:07:18 you mean cross compile? not easy 2019-06-21 09:07:57 native compile yes, and this is best option 2019-06-21 13:05:52 "<@_ikke_> ah, you are it-ofshore on github?" nope, just found a solution on his page 2019-06-21 13:16:08 <_ikke_> E-werd: ah ok 2019-06-21 14:38:06 When using Alpine as run-from-ram system, booting from USB device, can lbu be told to save changes within the USB itself? 2019-06-21 14:50:27 <_ikke_> ctcx: That's default behaviour (does require the usb to be writable) 2019-06-21 14:51:29 _ikke_: ouch, so that won't work if I made the Alpine USB via writing ISO with dd? 2019-06-21 14:51:45 <_ikke_> correct 2019-06-21 14:51:58 OUCH 2019-06-21 14:52:53 _ikke_: I'd like to make a multiboot USB, with Alpine and other distros. Would you think this https://linuxconfig.org/how-to-create-multiboot-usb-with-linux could work? 2019-06-21 14:54:52 <_ikke_> Probably, getting the bootloader entry correct is the trickiest part (though you could probably look at an existing setup 2019-06-21 14:55:51 ctcx: I spun up an alpine vm from an iso, passed the usb device to that vm and used setup-bootable in there to create a rw bootable disk 2019-06-21 15:10:07 _ikke_: where can one typically look for the correct bootloader parameters? 2019-06-21 15:12:50 ctcx: is there anything specific you're looking for? 2019-06-21 15:13:46 Making a successful multiboot USB with several live and installation distros without recurring to YUMI 2019-06-21 15:14:08 ah 2019-06-21 15:14:14 that's tricky I guess 2019-06-21 15:25:08 _ikke_: if dd is no good for using lbu to save changes in USB itself, what tool should I use to write Alpine to USB? 2019-06-21 15:26:18 <_ikke_> ctcx: like kahiru said, if you have an alpine system somewhere already, use setup-bootable 2019-06-21 15:26:36 <_ikke_> ctcx: if not, you can manually format the usb, copy the iso files on it, use isolinux and write mbr 2019-06-21 16:02:25 _ikke_: ah, the second one similar to the multiboot link I posted... 2019-06-21 16:04:26 In another thing, can several GPU drivers be installed in Alpine (and preserved with lbu), so I could use this live installation across several different rigs with Intel, ATI, Nvidia? 2019-06-21 16:04:51 <_ikke_> I think it should work yes 2019-06-21 16:05:02 <_ikke_> you can just install multiple xf86-video-* packages 2019-06-21 16:05:48 ^ Hooray! 2019-06-21 16:06:27 Oh, whenever installing a xf86-video, system must reboot, right? 2019-06-21 16:08:31 I don't see why assuming it's built to work with the kernel you are currently running (i.e. you are running a recent kernel) 2019-06-21 16:09:46 My bad. Sorry 2019-06-21 16:10:51 Can I use setup-bootable to save a customized bootable image of Alpine, but in a CD or DVD? 2019-06-21 16:11:36 danieli: hope everything is better for you now (yesterday you were having hard day) 2019-06-21 16:12:00 ah, I'm all right, I was just a bit stressed and tired 2019-06-21 16:13:18 and hm, the process of writing to usb is different to that of writing to CD/DVD 2019-06-21 16:14:06 hello! 2019-06-21 16:14:57 Ok, what would be the process to write customized Alpine image to CD? 2019-06-21 16:15:05 (if any) 2019-06-21 16:15:15 I wanted to install seafile on my alpine-linux, but it turns out that this article( https://wiki.alpinelinux.org/wiki/Seafile:_setting_up_your_own_private_cloud ) is out of date and seafile doesn't seem to be in the repos anymore 2019-06-21 16:20:54 commit 6e5499bca18c723547152560886ab3bb043c6e8f 2019-06-21 16:21:46 Ok, what would be the process to write customized Alpine image to CD? 2019-06-21 16:25:39 @danieli, thanks. At least now I know what's going on 2019-06-21 16:26:02 you might just be able to fix the aport if you put some effort into testing and building it :) 2019-06-21 20:30:20 Hi. I have installed Alpine on Virtualbox and trying to use the auto-resizing feature. It's not working and I think that's because the vboxvideo module is not loaded. What should I do? 2019-06-21 20:31:23 What architecture do I need for ARM_Cortex-A53? Its ARM v8. I see armhf, aarch64, armv7 2019-06-21 20:31:39 <_ikke_> aztron: apk add virtualbox-guest-modules-vanilla virtualbox-guest-additions 2019-06-21 20:32:04 <_ikke_> aztron: not sure if that will fix it 2019-06-21 20:33:05 <_ikke_> IGnorAND: looks like aarch64 2019-06-21 20:33:06 I have both of those pacakges installed already 2019-06-21 20:33:30 _ikke_: thanks, based on what do you say that? 2019-06-21 20:33:56 <_ikke_> https://en.wikipedia.org/wiki/ARM_architecture#AArch64 and https://en.wikipedia.org/wiki/ARM_architecture#AArch64 2019-06-21 20:34:02 apk list -I 2019-06-21 20:34:02 https://en.wikipedia.org/wiki/ARM_architecture#ARMv8-A 2019-06-21 20:34:05 ^^ 2019-06-21 20:34:15 <_ikke_> https://en.wikipedia.org/wiki/ARM_Cortex-A53 2019-06-21 20:36:03 <_ikke_> aztron: does lsmod return vboxvideo? 2019-06-21 20:36:46 No it does not. Only vboxsf and vboxguest is there 2019-06-21 20:36:55 <_ikke_> try modprobe vboxvideo 2019-06-21 20:39:09 modprobe says not found in /lib/modules/4.19.53-0-virt but it's part of vanilla package 2019-06-21 20:40:05 so arm7 and armhf are 32 bit it seems 2019-06-21 20:40:14 _ikke_: shizonic: thanks 2019-06-21 20:40:33 <_ikke_> aztron: what does apk version linux-virt return? 2019-06-21 20:40:55 looks like they use armhf though: https://certification.ubuntu.com/hardware/201810-26543/ 2019-06-21 20:41:09 4.19.53-r0 2019-06-21 20:42:02 np 2019-06-21 20:42:40 <_ikke_> aztron: Ah, you need to install virtualbox-guest-modules-virt, not virtualbox-guest-modules-vanilla (in case you didn't have that) 2019-06-21 20:43:05 <_ikke_> but that one does not have vboxvideo.. 2019-06-21 20:43:41 I have installed both. But virt does not have the module at all. 2019-06-21 20:44:05 <_ikke_> Yeah, just noticed 2019-06-21 20:44:18 <_ikke_> having -vanilla does not add anything 2019-06-21 20:44:54 I tried to manually insmod with full path. vanilla modules just give "Invalid module format" 2019-06-21 20:44:56 <_ikke_> aztron: # vboxvideo.ko won't load with virt kernel as it doesn't have DRM 2019-06-21 20:45:10 <_ikke_> https://git.alpinelinux.org/aports/tree/community/virtualbox-guest-modules-vanilla/APKBUILD#n90 2019-06-21 20:47:50 I chose the virtual download. Should I download any other edition? 2019-06-21 20:49:27 <_ikke_> The standard image is bigger (also for installation), but it would probably work better in this case 2019-06-21 20:49:39 <_ikke_> at least have the module available 2019-06-21 20:51:44 I will just try to hardcode the resolution to see if it works. https://superuser.com/questions/443445/how-can-i-get-virtualbox-to-run-at-1366x768 2019-06-21 22:02:42 Hi, I have set up a custom VESA resolution. I basically tried this: https://www.virtualbox.org/manual/ch09.html#customvesa It shows up in grub but not in xrandr or anywhere within Linux. Any ideas? 2019-06-21 23:36:24 abuild's buildroot has an -a $ARCH option, does it support cross compilation or does that need to be setup manually eg with qemu? 2019-06-22 00:43:04 I'm getting untrusted sig warnings running 'buildroot -C -p ~/buildroot -m dl-3.alpinelinux.org -v edge -a aarch64' on v3.10 when apk is getting the alpine base for the buildroot 2019-06-22 00:57:18 ERROR: http://dl-3.alpinelinux.org/alpine/edge/main: UNTRUSTED signature 2019-06-22 01:31:11 xcko: please report it on https://bugs.alpinelinux.org/ 2019-06-22 02:19:10 https://wiki.alpinelinux.org/wiki/Raspberry_Pi#Traditional_disk-based_.28sys.29_installation If I just want one folder/subfolder to be persistent, I can just mount the 2nd partition as /stage and be done with it right? no need to setup etc? 2019-06-22 02:19:32 sounds 'more' supported as its basically still diskless 2019-06-22 08:19:10 Any sick programs I should know about to make my life better? 2019-06-22 08:19:32 Yes, I'm asking for vague advice in a support channel on an IRC network full of people who can't comprehend vaguery. 2019-06-22 08:21:19 I recommend the GNU coreutils version of "rm"... Good for destroying stuff. 2019-06-22 08:21:22 :) 2019-06-22 08:45:06 priontology: vaguery is the worst enemy of good advice 2019-06-22 08:45:20 KatolaZ: Indeed, but I'm not concerned with good. 2019-06-22 08:45:38 zcrayfish: Are there bugs in gnu's rm? 2019-06-22 08:46:23 https://lists.gnu.org/archive/html/bug-coreutils/2019-06/threads.html <-- Absolutely =) 2019-06-22 08:46:41 priontology: then you will get lots and lots of irrelevant hints 2019-06-22 08:46:52 KatolaZ: Worthwhile for a conversation, though. 2019-06-22 10:06:41 Hey when creating a custom ISO image with mkimage how do you set a default root password / add a user? 2019-06-22 10:20:48 Anyone know how I can write something on /root persistenantly on the alpine armhf architecture? 2019-06-22 10:32:41 IGnorAND: I think you could add the directory to preserve with "lbu add" and then use "lbu commit" to save changes 2019-06-22 10:32:58 aggaz: yeah, thanks just figured out lbu inc 2019-06-22 10:33:28 not sure if it allows dirs yet though or if I should add the files one by one 2019-06-22 10:34:20 cool, it works with dirs :) 2019-06-22 10:34:33 <_ikke_> IGnorAND: note that you need to explicitly commit changes bore they are persisted 2019-06-22 10:34:45 _ikke_: lbu_commit -d? 2019-06-22 10:35:52 <_ikke_> lbu commit should be enough 2019-06-22 10:38:08 _ikke_: it's a raspberry pi, so don't think I need the previous layers 2019-06-22 10:38:44 <_ikke_> it's not docker, afaik it doesn't deal with layers 2019-06-22 10:48:57 -d Remove old apk overlay files. <- aren't overlay docker kinda layers? 2019-06-22 10:49:41 <_ikke_> afaik the overlay file is just a single archive 2019-06-22 10:49:45 <_ikke_> nothing fancy 2019-06-22 10:50:56 Hey when creating a custom ISO image with mkimage how do you set a default root password / add a user? 2019-06-22 10:53:45 <_ikke_> AppXprt: I don't now, sorry 2019-06-22 10:56:55 going to try this in apkovl: sed -i -e "s/^root:[^:]\+:/root:$pass:/" etc/shadow 2019-06-22 10:58:48 Well... can you help me under the $pkgname.pre-install here: https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package 2019-06-22 11:15:24 how are the passwords in /etc/shadow hashed? if Inject a user into /etc/passwd and /etc/shadow would it be the same as creating a user? 2019-06-22 11:16:13 or would it work? are the salts to the hashes still stored in the /etc/shadow user line itself? 2019-06-22 11:18:15 that seems to work from one system to another so I guess salts are still in the shadow 2019-06-22 11:21:55 hi all :-) 2019-06-22 11:24:53 come on I know someone knows this 2019-06-22 11:33:10 What's your question ? you want to add users by editing passwd and shadow files ? 2019-06-22 11:33:24 that's possible 2019-06-22 11:33:40 no I just want to add a root password to custom iso I'm making with mkimage aports tools 2019-06-22 11:33:56 and also another group and user + pass 2019-06-22 11:34:04 but just changing root password would be a great start 2019-06-22 11:34:08 when you add users using tools like adduser/useradd it will copy skel files too 2019-06-22 11:34:13 I might have it 2019-06-22 11:34:31 I mean I'm building alpine custom iso's with aports 2019-06-22 11:34:45 and the default root user has blank pass 2019-06-22 11:35:12 I'm trying to replace /etc/passwd and /etc/shadow in apk overlay (apkovl) 2019-06-22 11:35:20 then you can modify shadow file 2019-06-22 11:35:42 either manually (sed) or using passwd command 2019-06-22 11:36:11 well I need it to be set in iso 2019-06-22 11:36:32 I guess when it is running the mkimage part 2019-06-22 11:37:14 I've never made iso, but it is just a kind of packaging containing files 2019-06-22 11:38:23 yea the overlay might work 2019-06-22 11:38:46 what else is there to users and groups than /etc/passwd /etc/shadow? ( I know I should already know....) 2019-06-22 11:42:48 that worked for root 2019-06-22 11:43:35 well it set a password but won't let me login 2019-06-22 11:44:40 so the hash is wrong 2019-06-22 11:51:53 AH! 2019-06-22 11:51:56 AHA! 2019-06-22 12:07:54 shit don't work 2019-06-22 12:45:12 come on I just need to change root pass in iso 2019-06-22 12:58:49 AppXprt: run the blank password setup, copy /etc to /etc.saved, change root password, diff -ru /etc.saved /etc 2019-06-22 12:59:09 If nothing else it should be informative 2019-06-22 13:00:13 where is that? I need to do this for iso creation 2019-06-22 13:10:15 Start with the iso you have, make changes, observe the results, merge those into new iso, repeat until you're happy 2019-06-22 13:11:40 Using the docker image is likely to be easier. docker diff can show changes across the system 2019-06-22 14:30:01 sorry, I'm not a docker user 2019-06-22 14:31:21 way too difficult to add a password to root for iso creation 2019-06-22 14:46:34 surely someone knows how to add a user to custom iso build with mkimage 2019-06-22 14:59:45 Please someone help me 2019-06-22 15:32:37 <_ikke_> AppXprt: In the weekend it's usually more quite here 2019-06-22 15:32:49 <_ikke_> AppXprt: You'll probably have more success during the week 2019-06-22 17:03:12 Hi 2019-06-22 17:04:11 I succesfully booted and installed alpine extended, no errors 2019-06-22 17:04:34 What do i do now to see the desktop? 2019-06-22 17:04:39 Its diskless 2019-06-22 17:05:17 Alpine does not come with a GUI (desktop) installed by default 2019-06-22 17:05:50 Ok, so i install xorg? 2019-06-22 17:06:14 that's just the display server, you will have to install and configure a window manager (such as i3, awesome, etc.) or a desktop environment 2019-06-22 17:06:49 Thx will look into that 2019-06-22 17:06:50 <_ikke_> danieli: there is a setup-xorg script 2019-06-22 17:07:08 yes, it will do the basics of installing the X server 2019-06-22 17:07:14 iirc drivers too 2019-06-22 17:07:24 <_ikke_> You need to install the proper drivers yourself 2019-06-22 17:07:44 makes sense 2019-06-22 17:08:21 Ok i get it now, im just starting 2019-06-22 17:08:47 I guess i dont gui for running python scripts 2019-06-22 17:09:45 <_ikke_> Nope 2019-06-22 17:12:12 If you want to run KDE Plasma, I recommend reading https://wiki.alpinelinux.org/wiki/KDE 2019-06-22 17:33:18 Hi 2019-06-22 17:33:54 Can i offtop about some other distro? 2019-06-22 17:35:32 No, there is #alpine-offtopic for that 2019-06-22 17:35:42 Thx 2019-06-22 17:36:17 Join 2019-06-22 19:58:23 Hi 2019-06-22 19:59:08 hello 2019-06-22 19:59:14 Is there any manual about installing to ARM? 2019-06-22 19:59:34 I have ADB connection 2019-06-22 20:00:02 ADB connection..? 2019-06-22 20:00:03 And fastboot on both sides, but no connection yet 2019-06-22 20:00:14 I think you might be looking for postmarketOS, not Alpine 2019-06-22 20:00:21 I mean my pc sees the device 2019-06-22 20:00:35 Thx will check that 2019-06-22 20:02:38 I have aarch64 compatible processor in the device 2019-06-22 20:02:46 Cortexa53 2019-06-22 20:03:25 alpine isn't built to run on phones, and it most likely will not without heavy modifications 2019-06-22 20:03:51 however, postmarketOS (an alpine derivative, correct me if I'm wrong) is 2019-06-22 20:06:06 I downloaded alpine version aarch64, wirhmt uboot etc from official site 2019-06-22 20:06:23 with uboot* 2019-06-22 22:10:06 how can I set a root password in alpine-baselayout 2019-06-22 22:13:07 i see passwd, groups, etc in that directory as well as SHA verification hashes for those files in the APKBUILD directory 2019-06-22 22:14:01 also the $pkgname.pre-install never work 2019-06-22 22:17:03 even with exit 0 on end 2019-06-22 22:29:27 think I got it figured out, the pre-install goes in aports-build 2019-06-22 22:29:33 main/aports-build 2019-06-22 22:30:32 that is not likely 2019-06-22 22:30:50 well nothing else works literally 2019-06-22 22:31:07 do you even know what i'm trying to do? 2019-06-22 22:31:27 unless it's building packages on git push, that is the wrong package 2019-06-22 22:31:44 this is all included in aports build tools 2019-06-22 22:32:24 I'm trying to add a user to custom iso creation... I have a custom working profile and apk overlay, but nothing I've tried seems to work, which is a lot... 2019-06-22 22:33:15 I've tried in the profile, overlay, pre-install(s), init, injecting custom passwd, shadow, passwd-, shadow- and group/group- 2019-06-22 22:34:31 the closest I've gotten is with injecting, but breaks everything which I see now why... all these are generating on build with sha256 checksums and uid/gid's are automatically generated too likely 2019-06-22 22:37:17 if you know, please help me out, I've been fighting with this for way too long and it should be super easy 2019-06-22 22:37:47 I'm using mkimage with a custom mkimg.$profile to create custom working ISO's 2019-06-22 22:37:55 the only thing missing is a root passwd 2019-06-22 22:39:16 I don't know a think about mkimage, but maybe you can chroot into the thing and do `paswd`? 2019-06-22 22:39:59 yea but I have to know exactly where it do that because I think it creates users before it even does anything with kernel 2019-06-22 22:40:10 where to* 2019-06-22 22:42:32 it does a variety of things in scripts/genrootfs.sh, not sure if it's relevant to whatever you're trying to do 2019-06-22 22:42:58 AppXprt: http://tpaste.us/65xn 2019-06-22 22:43:26 You can add a script in /etc/local.d to change root password and do other things then delete itself 2019-06-22 22:43:44 that's possible but feels a bit hackish 2019-06-22 22:44:02 Well, if it works :P 2019-06-22 22:44:07 well I think genrootfs isn't called for ISO 2019-06-22 22:44:08 it's pretty much what firstboot does 2019-06-22 22:44:11 pretty sure 2019-06-22 22:44:38 I will try that as I've tried similar with init and cron but not in /etc/local/d 2019-06-22 22:44:41 local.d 2019-06-22 22:45:18 genrootfs seems to be called for other images though like rootfs 2019-06-22 22:46:27 i don't remember how those scripts work to detail 2019-06-22 22:46:39 adding that local.d file in overlay should actually work if it's something like: echo -e "password\npassword" | passwd root > /dev/null 2>&1 2019-06-22 22:47:36 but I tried with an init script that deleted itself and adding it with rc appxprt default 2019-06-22 22:47:37 I'd suggest chpasswd as it's intended for that sort of thing 2019-06-22 22:47:43 but it would never load without service appxprt status 2019-06-22 22:47:50 or service appxprt start 2019-06-22 22:48:01 but other services would start with rc_add service default 2019-06-22 22:48:07 like open-vm-tools 2019-06-22 22:48:40 I tried with /etc/init.d/appxprt.sh and /etc/init.d/appxprt 2019-06-22 22:49:17 and at end had: 2019-06-22 22:49:19 rm /etc/init.d/appxprt.sh 2019-06-22 22:49:20 rc-update del appxprt 2019-06-22 23:02:51 Try something like this: http://tpaste.us/VaLd 2019-06-22 23:04:03 damn that's nice 2019-06-22 23:05:22 way better than my script lolhttps://pastebin.com/GPXPaUyS 2019-06-22 23:05:24 https://pastebin.com/GPXPaUyS 2019-06-22 23:07:03 I can think of two improvements right away. 1) using one call to chpasswd 2) passing already encrypted passwords 2019-06-22 23:10:22 how can I pass already encrypted pass? 2019-06-22 23:10:52 won't that mess up that shadow? 2019-06-22 23:11:27 chpasswd has the --encrypted flag for that 2019-06-22 23:12:43 Add something like this to APPEND http://tpaste.us/mZL7 2019-06-22 23:12:58 then modify the script to use chpasswd --encrypted 2019-06-22 23:14:30 like so: http://tpaste.us/RgR9 2019-06-22 23:15:06 damn nice 2019-06-22 23:17:00 hmm does the /etc/local.d script have to end in .sh? 2019-06-22 23:17:11 I have /etc/local.d/appxprt that didn't execute 2019-06-22 23:21:07 it write and makes it to /etc/local.d on iso, trying with .sh ext 2019-06-22 23:21:55 it has to be executable and end in .start according to the readme file 2019-06-22 23:22:14 https://tpaste.us/65xn 2019-06-22 23:27:07 hmm but your script accepts password from command line? I need it automatic 2019-06-22 23:35:53 it uses kernel args 2019-06-22 23:37:30 feel free to modify it to use wget or whatever you like to get passwords 2019-06-22 23:39:32 well password is set but my script sure didn't work 2019-06-22 23:40:29 wrong password with my script, how can I get your script to set them in script? 2019-06-22 23:42:33 just clear text to start with so I can make sure it's actually working 2019-06-22 23:43:02 I'm thinking I would change /proc/cmdline part? 2019-06-22 23:43:17 cat /proc/cmdline 2019-06-22 23:45:20 that script is seriously epic, but hard for me to fully understand 2019-06-22 23:47:19 can you help me just be able to set a clear text root pass with your script so I can understand that process and then I can figure out how to pipe that to chpasswd --encrypted ? 2019-06-22 23:48:00 I'll make a blog for others facing the same issue, I know this is resolved for docker, but not pure iso creations 2019-06-22 23:50:07 please man 2019-06-22 23:50:18 I really need this to work 2019-06-22 23:51:28 with http://tpaste.us/9g5n you would add something like this to your kernel arguments: passwd_root=SomethingSecure 2019-06-22 23:55:00 with http://tpaste.us/zBv9 you would add a kernel argument like this: http://tpaste.us/nMLx 2019-06-22 23:58:40 guessing that would be kernel_cmdline="nomodeset" 2019-06-22 23:58:48 in the mkimg profile 2019-06-22 23:59:11 sure, just tack on a space and your passwd_root=something argument 2019-06-22 23:59:40 cool trying that 2019-06-23 00:00:17 dude if this works, I owe you... I've been fighting with this for hours 2019-06-23 00:07:54 hmm that doesn't seem to work, when I look at the .start file in has other kernel options but not that one 2019-06-23 00:09:35 it included all the initfs stuff 2019-06-23 00:10:41 like for opt in BOOT_IMAGE=vmlinux-virt root= modules= etc... 2019-06-23 00:10:44 case "" in 2019-06-23 00:14:37 trying in initfs option 2019-06-23 00:14:53 Here's what I get when I test it: # /bin/bash -x /etc/local.d/set_passwords.start # => outputs http://tpaste.us/Lzqa 2019-06-23 00:15:18 I copied /proc/cmdline to /tmp/cmdline so I could test without rebooting 2019-06-23 00:23:04 hmm whats in your /tmp/cmdline 2019-06-23 00:23:23 I see how it works 2019-06-23 00:23:29 let me see if I can get it to work on iso 2019-06-23 00:23:47 http://tpaste.us/ZgLa 2019-06-23 00:25:12 yea it works for me on an installed alpine distro like you said test it 2019-06-23 00:25:19 with those options in /tmp/cmdline 2019-06-23 00:25:42 I understand a lot more now thank you very much, I think I can get it working so I'm going to try it out 2019-06-23 00:38:35 AppXprt: Good luck with it. Here's my latest version: http://tpaste.us/dk0y 2019-06-23 00:38:58 damn, options huh lol 2019-06-23 00:39:10 trying in iso 2019-06-23 00:39:14 I changed it to try to use encrypted passwords or not based on what is passed in the kernel args 2019-06-23 00:39:20 yea 2019-06-23 00:39:25 you're badass 2019-06-23 00:41:10 ahhh shit I had the wrong ARCH case the entire time 2019-06-23 01:06:07 hmm when I add to kernel_cmdline it doesn't work and when I add to initramfs it breaks grub boot 2019-06-23 01:07:50 can you see it in /proc/cmdline when you boot into the system from the iso? 2019-06-23 01:10:28 one sec, rebuild iso 2019-06-23 01:34:53 does anyone know if there is still work being done to port openjdk to musl? 2019-06-23 01:35:10 11* 2019-06-23 01:35:32 IDK but I'm using openjdk 8 with musl for android sdk 2019-06-23 01:36:34 I do know 8 works with alpine, but 11 is lagging behind 2019-06-23 01:40:19 cljm: https://pkgs.alpinelinux.org/packages?name=openjdk11&branch=v3.10 2019-06-23 01:41:28 tcely: fantastic! thanks! 2019-06-23 01:57:32 adding kernel_cmdline="nomodeset passwd_root=PASSWORD" breaks grub from booting 2019-06-23 01:58:12 using the encrypted method that I tested as working in /tmp/cmdline 2019-06-23 01:59:32 initfs_cmdline ? 2019-06-23 01:59:50 modules=loop,squashfs,dasd_mod,qeth quiet passwd_root=PASSWORDHASH ? 2019-06-23 02:15:31 do I need quotes around the password or something? it keeps failing to boot 2019-06-23 02:17:08 or \ escape the $'s 2019-06-23 02:26:54 if your password has any special characters they will need to be escaped 2019-06-23 02:28:23 using kernel_cmdline='nomodeset passwd_root=PASSWORD' is going to be easier than escaping all the $ found in the encrypted form 2019-06-23 02:31:26 yea I tried that and it boots 2019-06-23 02:31:47 but in the .start it doesn't load it even though it is in /etc/cmdline 2019-06-23 02:31:56 one sec I'll paste what is in that 2019-06-23 02:38:17 /proc/cmdline: BOOT_IMAGE=/boot/vmlinuz-vanilla modules=loop,squashfs,dasd_mod,qeth quiet nomodeset passwd_root=password 2019-06-23 02:39:00 .start file for opt: 2019-06-23 02:39:02 for opt in BOOT_IMAGE=vmlinuz-virt root=/dev/vg0/lv_root modules=sd-mod,usb-storage,ext4 pax_nouderef quiet rootfstype=ext4 initrd=initramfs-virt; do 2019-06-23 02:40:22 please don't ghost me, you're the only one able to help me with this apparently :-D 2019-06-23 02:42:14 why does the list after "for opt in" not match /proc/cmdline? 2019-06-23 02:42:31 no idea, this is a fresh iso creation with mkimage 2019-06-23 02:42:51 how are you creating that .start file? 2019-06-23 02:43:21 It looks like it might be running cat /proc/cmdline before the iso is created 2019-06-23 02:45:39 https://pastebin.com/uyMAUjtm 2019-06-23 02:46:04 that is part of the apkovl (apk overlay) 2019-06-23 02:46:47 shouldn't be, it should be creating the file from < instead of executing the cat 2019-06-23 02:47:01 right? 2019-06-23 02:47:09 You need to quote "EOF" and remove lines 3-5 2019-06-23 02:47:15 the #!/bin/sh should be the first line 2019-06-23 02:48:12 using < ahhh 2019-06-23 02:50:57 gotcha 2019-06-23 02:51:17 trying 2019-06-23 02:51:24 Here is a quick example: http://tpaste.us/4yn4 2019-06-23 03:01:31 OMG that worked thank you so much 2019-06-23 03:03:35 AppXprt: congrats 2019-06-23 03:21:47 tcely, you rock... 2019-06-23 03:32:04 Hmmm when setting a password with special chars like $ I should just have to escape them with a \ right? 2019-06-23 03:32:05 \$ 2019-06-23 03:32:25 do I need like passwd_root='\$' 2019-06-23 03:33:05 or should "nomodeset passwd_root=\$" be right... vs "nomodeset passwd_root='\$'" 2019-06-23 03:36:30 kernel_cmdline='nomodeset passwd_root=$example' 2019-06-23 03:37:08 You only need to escape because of using double quotes 2019-06-23 03:37:38 so the $ with show up as $example instead of using it as a var? 2019-06-23 03:37:48 will* 2019-06-23 03:37:59 with single quotes 2019-06-23 03:38:11 nice 2019-06-23 03:38:20 kernel_cmdline="nomodeset passwd_root=\$example" 2019-06-23 03:38:22 I never knew that, you've taught me so much tonight 2019-06-23 03:38:37 hmm yea that's what I thought, but didn't work with multiple \$ 2019-06-23 03:38:43 trying with single ' 2019-06-23 03:45:18 damn, nope... 2019-06-23 03:45:37 it works with standard password though 2019-06-23 03:45:45 say way further ahead 2019-06-23 03:46:08 it always has to be complicated with me 2019-06-23 03:46:21 what ended up in /proc/cmdline? 2019-06-23 03:47:02 lol can't login 2019-06-23 03:48:03 you can login before local runs right? 2019-06-23 03:48:42 guess if I boot to single user mode? 2019-06-23 03:49:13 check the boot configuration or change the user it's changing the password for 2019-06-23 03:52:37 wow it cuts off the $ and the follow char for the first $ and two follow chars for the second $ 2019-06-23 03:52:51 following* 2019-06-23 03:53:21 seen from single user mode 2019-06-23 03:54:03 it may be a coincidence, but the char's being cut off following the $ are always integers (numbers...) 2019-06-23 03:54:18 I have some interesting password combinations 2019-06-23 03:54:40 that sounds like variables are being expanded for some reason 2019-06-23 03:54:53 $6 is going to be an empty variable for example 2019-06-23 03:55:44 you can try single quotes and escaping, but I don't see why that would be needed 2019-06-23 03:56:57 yeap that's exactly what I was trying next 2019-06-23 04:05:24 hmmm yeap that worked 2019-06-23 04:06:07 dude you super rock 2019-06-23 04:06:14 super hacker 2019-06-23 04:50:57 no I think it is the ' ' without the escaping and I must have not been testing right or something 2019-06-23 05:00:01 If grub is used it could be trying to parse variables 2019-06-23 05:00:33 yea you were right, it was ' ' without escaping \$ 2019-06-23 05:00:42 I just didn't upload and test right or something 2019-06-23 05:01:09 you are legit a linux ninja 2019-06-23 05:02:04 Thank you so much man I literally would have never gotten this without you 2019-06-23 11:10:28 Does anybody know if qbittorrent-nox has been recently removed from the edge/testing repo? I swear I saw it last week, but for the life of me I can't find it now =\ 2019-06-23 11:13:07 https://pkgs.alpinelinux.org/packages?name=qbittorrent%2Dnox 2019-06-23 11:13:11 Seems there is no such package 2019-06-23 11:13:53 I searched and couldn't find it just now, yeah -- but it was there in the testing repo last week, so I am just trying to figure out if it's been removed for some reason :o 2019-06-23 11:14:12 Check the Git logs I guess 2019-06-23 11:15:14 I think it was stuck on the builders, or was that another torrent? 2019-06-23 11:15:31 I'm unsure, I only use qbit ^^ 2019-06-23 11:15:52 Just setting up a new VM and went to install and couldn't find the package =P 2019-06-23 11:16:32 xMPLR, it was disabled: https://github.com/alpinelinux/aports/commit/d34df3a1ef83567b13476cda82240c7e8b08e6e9 2019-06-23 11:17:17 Yup, that one of the deps was stuck on the builder 2019-06-23 11:18:12 hrm, any options? Or just wait until it appears in the repo again? 2019-06-23 11:20:01 very descriptive commit message lol 2019-06-23 11:20:09 oh right, there's a comment in the diff itself 2019-06-23 11:20:24 Create a patch to fix it I guess 2019-06-23 11:20:44 PureTryOut[m] I don't think I am that advanced ;) 2019-06-23 11:21:12 In that case you can do nothing but wait sadly 😛 2019-06-23 11:21:31 hehe, I thought that might be the case ^^ *twiddles his fingers* 2019-06-23 11:22:28 I can ask maxice8 about it once he wakes up 2019-06-23 11:22:46 Is he from the US btw? Seems his timezones are a bit different than ours 2019-06-23 11:25:58 He's in BRST 2019-06-23 11:27:32 cool 2019-06-23 11:30:23 BRST? 2019-06-23 11:30:43 Brazil? 2019-06-23 11:31:00 Exactly half way around the globe from me, hehe 2019-06-23 11:40:42 Does anyone actually use Grub on x86_64? It can not install itself because it only has files in `/usr/lib/grub/i386` rather than `/usr/lib/grub/amd64` 2019-06-23 11:42:06 Oh wait, maybe it doesn't matter 2019-06-23 11:42:14 No, it doesn't 2019-06-23 11:42:38 The i386 describes the BIOS and not the OS 2019-06-23 11:42:43 Yeah the error message was just confusing. I forgot I wasn't running as root 😒 2019-06-23 11:42:49 Heh 2019-06-23 12:59:33 Does anyone have virgl working in an Alpine Linux vm? I installed `mesa-dri-virtio` and `virglrendere`, but Xorg keeps complaining that `/dev/dri/card0` isn't available (which it indeed isn't) 2019-06-23 12:59:41 *`virglrenderer` 2019-06-23 13:00:03 Your udev is functional, right? 2019-06-23 13:02:26 Think so? It's just a barebones install so far 2019-06-23 13:02:44 udev is running anyway 2019-06-23 13:06:56 `rc-update | grep udev` shows that udev, udev-settle and udev-trigger are started? 2019-06-23 13:08:07 udev-settle isn't, the others are, although I do have udev-postmount 2019-06-23 13:09:06 Added `udev-settle` to sysinit, rebooted, no luck 2019-06-23 13:11:00 Hm, yeah, I usually only hit that error when the module isn't loaded correctly due to udev not being setup correctly 2019-06-23 13:11:12 Not quite sure what module Virgl needs 2019-06-23 13:11:57 Strange thing is that `dmesg | grep -w VGA` returns nothing, seems the card isn't loaded at all for whatever reason 2019-06-23 13:11:57 Maybe do `lspci -nnk` (not sure if busybox can do that) and check if the virGL card has a module available 2019-06-23 13:12:10 ACTION sent a long message: < https://matrix.org/_matrix/media/v1/download/matrix.org/UeHQiwRpVtStsPrzfVrBOCbe > 2019-06-23 13:13:06 Ah, no joy with busybox lspci 2019-06-23 13:13:07 ACTION sent a long message: < https://matrix.org/_matrix/media/v1/download/matrix.org/YebfCvhQJxXSXbuCmrwjjNGS > 2019-06-23 13:13:10 You'll need pciutils 2019-06-23 13:13:37 Huh, so the module is loaded 2019-06-23 13:13:57 Yeah 2019-06-23 13:51:28 Hi, llvm8 (dependency of the dependency of firefox on arm64) is ready. Thanks! 2019-06-23 13:51:35 now, which is the state of rust (the dependency of firefox on arm64) ? 2019-06-23 13:53:43 <_ikke_> rust is currently on version 1.34, still only on x86_64 2019-06-23 13:54:09 <_ikke_> I think the goal is to get rust 1.35 in aports and get it bootstrapped on more arches 2019-06-23 13:55:11 ok, so it is like 1.35 is a requirement before porting to other archs 2019-06-23 13:55:40 <_ikke_> I don't know, but upgrading is easier than adding new arches 2019-06-23 13:56:27 maybe we can use rust from adelie linux 2019-06-23 13:58:21 bootstrapping a compiler can be tedious, but if one can use an already working one (e.g.: from another distribution) it can be almost as easy as a normal build 2019-06-23 13:59:52 vicencb: I made apk's of rust 1.31 on aacrh64 and armv7 but lost correct APKBUILD's and patches 2019-06-23 14:00:38 I think I will try again if I found some free time 2019-06-23 14:02:27 i am willing to test alpine on my arm64-based desktop, but not having firefox is a blocker issue 2019-06-23 14:03:35 I have firefox-esr on both aarch64 and armv7 but on previous stable (v3.9) 2019-06-23 14:05:45 mps: it is not on upstream alpine, do you have your own APKBUILD? 2019-06-23 14:06:37 for rust yes, I have a lot messed APKBUILDs 2019-06-23 14:07:38 but I have rust apk which works on 3.9, and using them I built standard Alpine firefox-esr 2019-06-23 14:08:25 ok, but, is it possible to build rust and firefox from source on a machine with 4GB of RAM? 2019-06-23 14:09:40 I managed on 4GB with the help from zram swap, it is really slow, take about 12 or more hours, on mmc card 2019-06-23 14:09:59 I'll open a PR for Rust on armhf/armv7/x86 later today 2019-06-23 14:10:14 Will have to do aarch64 in a. It 2019-06-23 14:13:05 But I guess first I'll have to (finally) set up Qemu images for those arches 2019-06-23 14:14:24 posted this on wrong channel: it was firefox-esr version old about 5 month 2019-06-23 14:14:53 Cogitri: when you make PR ping me 2019-06-23 14:15:27 mps: Cogitri: thank you all 2019-06-23 14:17:48 mps: Will do, I guess I'll ping you before that even if I fail to setup the Qemu thinf 2019-06-23 14:18:32 np 2019-06-23 18:13:54 cogitri: Actually i'm on BRT which is UTC -3 2019-06-23 23:36:44 any known issues with fail2ban no longer logging or banning after a log rotation? 2019-06-23 23:58:31 anyone upgraded to 3.10 yet 2019-06-24 00:08:29 ACTION doesnt even know on which version he is 2019-06-24 00:08:45 tsk tsk tsk 2019-06-24 00:08:49 cim209: my raspberry pi is on 3.10 2019-06-24 00:09:23 guess my production is still on 3.9 2019-06-24 00:11:32 I see 3.10 is 4 days old 2019-06-24 00:11:50 but just bug fixes apparently 2019-06-24 00:12:40 gonna test it out on my docker images 2019-06-24 00:20:25 cim209: how do I check my alpine version on my postgres-version-alpine? 2019-06-24 00:20:36 cat /etc/os-release 2019-06-24 00:21:01 cim209: guess my postgres is still 3.9.2 2019-06-24 00:21:29 hmm nginx 17.0 packages aren't in 3.10 2019-06-24 00:22:24 ruby is on 3.9.2 still too 2019-06-24 00:23:17 isn't nginx on version nginx-1.xx? 2019-06-24 00:23:28 nginx is now version 17 2019-06-24 00:23:28 or do you mean 1.17? 2019-06-24 00:23:47 yeah my bad 2019-06-24 00:23:48 1.17 2019-06-24 00:25:18 well, my web docker are based on nginx-latest-alpine and that's still on 1.16 2019-06-24 00:26:49 i've been on nginx 1.17 for a while now 2019-06-24 00:27:27 lemme check why im on 1.16 2019-06-24 00:27:50 I'm on stable: https://hub.docker.com/_/nginx 2019-06-24 00:28:14 idk why they installed nginx from package manager in version 17 2019-06-24 00:28:14 apparently you're on mainline 2019-06-24 00:28:23 they used to build nginx from source 2019-06-24 00:28:23 good lord, just got my first desktop installation of alpine 2019-06-24 00:28:27 i'm in love, it's so fast and pure 2019-06-24 00:28:31 IGnorAND: no, i'm on alpine 2019-06-24 00:28:49 E-werd: what window manager do you use? 2019-06-24 00:28:53 MATE 2019-06-24 00:28:54 mainline is debian 2019-06-24 00:28:59 ACTION hasnt used linux as a desktop in 10+ years 2019-06-24 00:29:08 i'm on XFCE 2019-06-24 00:29:21 I've been using alpine as a server OS steadily for the last few months, fell in love with the simplicity there 2019-06-24 00:29:35 yeah i use alpine as the host OS for my docker setup 2019-06-24 00:29:38 xfce is a name I remember. mate is new 2019-06-24 00:29:45 only thing installed in alpine are bash and docker 2019-06-24 00:29:52 MATE is pretty old lol 2019-06-24 00:29:53 cim209: same here :) 2019-06-24 00:29:54 MATE is just gnome2 renamed 2019-06-24 00:30:04 so if you used gnome before gnome-shell then you already know it 2019-06-24 00:30:05 oow, gnome I know :) 2019-06-24 00:30:50 gnome was supposed to be a simplistic window manger. Supposedly very little that needs to be tweaked, but if it needs to be tweaked its a pain :P 2019-06-24 00:31:05 kde was the othe rway around, EVERYTHING had a gazilion options 2019-06-24 00:31:22 at least that's what I recall 2019-06-24 00:31:51 cim209: running alpine on my pi looks nice too, gonna try and move my dockers on there and see how well they perform 2019-06-24 00:32:18 looks like bash is 5.0 on alpine 3.10 2019-06-24 00:34:30 cim209: i'm on edge right now, installed 3.10 and no upgrades after switch to edge. bash 5.0.0(1)-release 2019-06-24 00:34:58 nice 2019-06-24 00:35:06 hope you are on edge on a desktop lol 2019-06-24 00:35:14 well yeah 2019-06-24 00:35:31 ...and also on my primary server 2019-06-24 00:35:40 it's nothing mission critical, but I wanted hiawatha-webserver 2019-06-24 00:36:14 couldn't you just add the edge to your repo list and remove it after you're done? 2019-06-24 00:39:10 https://kopy.io/wZgLm :( 2019-06-24 00:39:37 well, probably could 2019-06-24 00:41:20 cim209: I dont undrestand why it would break . Maybe fetch the apk from fetch http://dl-cdn.alpinelinux.org/alpine/v3.9/main/x86_64/APKINDEX.tar.gz 2019-06-24 00:41:46 yeah i have to add that to the repo list 2019-06-24 00:41:50 ugh lol 2019-06-24 00:41:56 why are you insisting on specific versions? 2019-06-24 00:42:05 because rolling release 2019-06-24 00:42:09 mmm 2019-06-24 00:42:19 my docker images don't have versions 2019-06-24 00:42:20 cim209: nvm: http://nginx.org/packages/mainline/alpine/ thats your problem 2019-06-24 00:42:35 yeah they don't have a 3.10 branch yet 2019-06-24 00:42:37 those last 2 lines are curious 2019-06-24 00:49:16 yes, it's installing 2019-06-24 00:51:23 i'll upgrade my dev server to 3.10 2019-06-24 00:54:31 so for 3.10 desktops, it comes with lightdm 2019-06-24 01:00:32 cim209: how do you determine if you can go to the new version for prod? 2019-06-24 01:01:16 what do you mean? 2019-06-24 01:14:37 cim209: I upgraded 1 dev machine and my prod miniserver 2019-06-24 01:15:21 testing 3.10 on my docker images first 2019-06-24 01:16:37 The only thing that's broken so far is docker-compose 2019-06-24 01:17:27 And that's upstream fault, it brings a lot of old dependencies and is sort of "abandonware" 2019-06-24 01:17:45 and Swarm Mode is pretty much dead too I recon, all hail Kubernetes 2019-06-24 01:17:48 ah damn 2019-06-24 01:17:54 one of my images has docker compose in it 2019-06-24 01:20:22 To get it going just install python3 and python3-dev and build-base and pip3 install it. It'll compile a whole load of crap but it'll work 2019-06-24 01:20:53 Ideally we should just package it as it is a big PITA 2019-06-24 01:21:05 i have a script that loops and builds through all my dockerfiles so i gotta wait till it finishes 2019-06-24 01:21:23 it should just be a damn binary like the docker client 2019-06-24 01:21:51 Indeed, but for historical reasons it's not. 2019-06-24 01:22:52 And it is the best simple solution so far. docker stack/Swarm Mode thing is no good (no support for capabilities, tmpfs and other things) 2019-06-24 01:23:26 you should just stick with 3.9 for now 2019-06-24 01:25:12 aaaaaaand know I'm getting BAD signature for packages. FFS, I don't have time for this today :) 2019-06-24 01:25:51 lol 2019-06-24 01:26:07 latest-stable repo is borked 2019-06-24 01:26:13 use v3.10 and all's well 2019-06-24 01:34:04 i'm glad php is built from source 2019-06-24 01:47:23 How would I run skype on alpine? 2019-06-24 01:51:01 hmm, how could I get networkmanager to manage ethernet. wireless is working fine 2019-06-24 01:51:18 "device not managed" but I created it 2019-06-24 02:01:51 E-werd: it's like 2019-06-24 02:01:57 nmcli set eth0 managed or some DUMB shit 2019-06-24 02:04:24 nmcli, that's useful 2019-06-24 02:16:46 yeah, that got me pointed in the right direction--thanks priontology 2019-06-24 02:17:08 E-werd: 20 pesos please 2019-06-24 02:17:16 argentine pesos? sure. 2019-06-24 02:19:08 Hell yeah. 2019-06-24 02:19:15 What's that in USD? Two cents? 2019-06-24 02:20:10 Oh, interesting. A peso is about 5 cents now in mexico, 2.x cents in argentina. 2019-06-24 02:22:01 i don't think it's quite that much 2019-06-24 02:22:16 hmm... turns out on alpine, riot-web is just the web app 2019-06-24 02:22:24 noooot what I'm looking for 2019-06-24 03:47:56 hmmm, think I found an abuild issue with 3.10 when creating the cross compiler chain using aports bootstrap.sh 2019-06-24 04:12:09 odd, reboot resolved it 2019-06-24 06:16:21 clandmeter hi, do you have already some Rpi4 images/support available that I could test ? (I am the guy who tried to boot alpine over network at rpi :) ) 2019-06-24 06:28:08 I would have thought it backward compatible 2019-06-24 06:54:02 CvH: morning 2019-06-24 06:54:17 PhoenixMageyou need different blobs etc 2019-06-24 06:55:01 i didnt even know rpi4 was a thing 2019-06-24 06:55:10 ive been too busy with other things 2019-06-24 06:55:19 :D ok likely good that i didn't asked you a month ago :D 2019-06-24 06:56:08 when did it get released? 2019-06-24 06:56:34 1h ago 2019-06-24 06:56:41 ah 2019-06-24 06:56:49 thats a good reason for me to not know about it 2019-06-24 06:56:49 didn't know if you got a sample or not so can't ask ;D 2019-06-24 06:57:09 nah we do not have ties to rpi 2019-06-24 06:57:54 heh its not even on popular news sites yet 2019-06-24 06:57:54 oki :) no problem, I can wait 2019-06-24 06:59:36 bt5.0? heh 2019-06-24 07:07:27 CvH: wasn't that *just* released? 2019-06-24 07:07:40 oh yeah 2019-06-24 07:17:17 yes it did 2019-06-24 07:17:20 and its some upgrade 2019-06-24 07:17:55 yea its much faster, the 4gb ram helps also a lot 2019-06-24 07:18:16 but ofc its currently not as stable as RPi0-3, but that is no surprise 2019-06-24 07:18:35 but you can watch 4k at it :D thats great ^^ 2019-06-24 07:19:33 and bootloader in spi :) thats nice for dev purposes 2019-06-24 07:20:39 i won't have any use for 4K personally, but some more resources will be nice 2019-06-24 07:23:52 it is finally a big upgrade and not just a minor update like rpi0-3 2019-06-24 07:25:01 this is a bit off topic, but one last question: but is it fully open source yet? 2019-06-24 07:26:56 CvH: do you already have one? 2019-06-24 07:29:29 holy hell 2019-06-24 07:30:08 2 micro-hdmi ports, they can apparently drive one 4K display each (2x) 2019-06-24 07:30:38 yes at 30hz 2019-06-24 07:30:41 or single on 60 2019-06-24 07:30:48 and it has pcie lanes 2019-06-24 07:30:53 if i read correctly 2019-06-24 07:31:10 fair enough, it's still impressive for such a small board 2019-06-24 07:31:20 and judging by their twitter, the first person with a rpi4 was someone in their physical store about 15 minutes ago 2019-06-24 07:31:39 i think projects will have received them earlier 2019-06-24 07:32:14 I wish it had sata, at least usb 3 is an improvement 2019-06-24 07:32:23 For my money I'd still get an espressobin 2019-06-24 07:32:49 The 4Gb Rpi4 is $100AUD damn 2019-06-24 07:34:12 that's a bit pricy, but not unexpected 2019-06-24 07:34:42 55 dollar isn't bad 2019-06-24 07:36:09 "The base model still costs $35, while the 2GB RAM model costs $45 and the 4GB RAM model costs $55." 2019-06-24 07:36:29 55 USD is ~80 AUD 2019-06-24 07:36:50 PhoenixMage: where'd you get that figure from? 2019-06-24 07:38:25 Well thats 4 aport bugs logged today, now I know why I dont bootstrap the arm compilers very often 2019-06-24 07:39:22 I'm trying to use NetworkManager on my laptop, but it keeps disconnecting from the wifi. Like, it'll connect, then disconnect within 2 seconds, then reconnect and disconnect again within 2 seconds, and it does this repeatedly till it stops trying after a while 2019-06-24 07:39:38 While it is connected for the few seconds it does manage to download or upload stuff though 2019-06-24 07:39:43 Since I have been a pain and logged the bugs, how do I attach a patch to the case for review? 2019-06-24 07:42:22 CvH: where did you get the update bootloader info from? 2019-06-24 08:03:12 clandmeteryea I have one since ~2-3 months, bootloader can be updated from sd card and is saved in spi (sry no bootloader expert ^^) 2019-06-24 08:03:58 so you can use it without sd card and still have the latest bootloader 2019-06-24 08:04:14 nice for netboot 2019-06-24 08:04:23 thats quite nice, basically like bios 2019-06-24 08:22:24 On a 4Gb with alpine that could be nice to netboot it 2019-06-24 08:22:42 Be interested to see the throughput on the usb3 ports 2019-06-24 08:32:56 that won't happen before we're able to test alpine on an rpi 4 2019-06-24 08:47:54 how will speed of usb3 port depend on Alpine 2019-06-24 08:50:12 mps: I don't think they were aiming at alpine specifically 2019-06-24 08:52:00 my son have acer R13, arm64 with usb3, works nice 2019-06-24 08:53:00 Also, I tested Alpine on it, but didn't benchmarked anything 2019-06-24 08:54:10 but I have seen clearly big speed boost in regard to debian 2019-06-24 08:54:37 where 'big' is subjective 2019-06-24 08:56:52 this is about the new raspberry pi specifically though, not ARM 2019-06-24 08:57:09 I am talking more about usb performance in general on the RPi4, it was pretty bad on the 3 2019-06-24 08:57:40 3 didn't have usb 3 though? 2019-06-24 08:58:18 No but its was bad even for 2. Shared resources from memory 2019-06-24 09:00:34 it's not exactly a supercomputer though 2019-06-24 09:00:36 can't have too high expectations 2019-06-24 09:01:46 I think it was the ethernet and the usb that were shared so you couldnt usb both full speed 2019-06-24 09:02:01 Just went an alternative board that had the disk access I was after 2019-06-24 09:05:11 I guess thats the beauty of having a choice of so many SBUs 2019-06-24 09:05:16 SBCs 2019-06-24 09:05:50 usb/network is solid at RPi4, not to compare with that crap at RPi 0-3 2019-06-24 09:06:48 Well thats a plus 2019-06-24 09:06:59 should probably move this to #alpine-offtopic now that i think about it 2019-06-24 10:20:28 ACTION uploaded an image: Screenshot_20190624_104853.png (89KB) < https://matrix.org/_matrix/media/v1/download/fam-ribbers.com/edoJEOKEQWUDTqXUejNAkNDP > 2019-06-24 10:20:44 Ignore the still ugly font in the window border 2019-06-24 10:24:29 Neat 2019-06-24 10:26:24 ACTION uploaded an image: IMG_20190624_122521_442.jpg (30KB) < https://matrix.org/_matrix/media/v1/download/matrix.exqa.de/jrVVJWbrcAWFfYUcYjIXkWva > 2019-06-24 10:26:40 ACTION goes to updating the wiki article 2019-06-24 10:28:35 Have a look at the KDE article for an example https://wiki.alpinelinux.org/wiki/KDE 2019-06-24 10:28:49 Also, your font needs anti-aliasing 2019-06-24 10:29:37 hmmm odd, ssh-agent isnt working for my yubico on alpine 2019-06-24 10:29:58 i haven't tested my yubikey setup on alpine 2019-06-24 10:30:46 My ubuntu instance works fine to my alpines but alpine to alpine is prompting me for a password 2019-06-24 10:32:54 PureTryOut: that's compression 2019-06-24 10:33:17 Damn, that is some bad compression then 2019-06-24 10:34:15 It's the "small" option of Riot, seems like my homeserver doesn't like "original" for some reason 2019-06-24 10:35:09 Oh interesting 2019-06-24 11:03:16 And edited the GNOME wiki page, thanks for the reminder, PureTryOut 2019-06-24 11:03:19 Np 2019-06-24 11:04:04 About the section "Setting up Xorg", note that that stuff is already done by running `setup-xorg-base` 2019-06-24 11:05:22 Oh, thought it'd only setup the "base" and not the drivers 2019-06-24 11:05:43 Cogitri: is there some Gnome app to notify about new emails? 2019-06-24 11:06:12 Nope, also drivers. It checks the output of `lspci | grep -w VGA` and installs drivers depending on the output. Check `/sbin/setup-xorg-base` 2019-06-24 11:06:57 clandmeter: There's Evolution and Geary, but those are full fledged email clients, I guess you're asking about a systray thingie or something like that? 2019-06-24 11:07:05 PureTryOut: Oh, cool! 2019-06-24 11:07:11 Cogitri: yes 2019-06-24 11:07:21 i dont need an email client 2019-06-24 11:10:02 Hmm..I don't think GNOME offers something like that, especially now that they've removed the systray 2019-06-24 13:40:10 Anyone has documentation on how to get from booting to a Hard drive to boot the same OS from USB? 2019-06-24 13:46:24 alpine latest version 32bits have issue with musl gcc 2019-06-24 13:49:42 do elaborate, please? 2019-06-24 13:50:16 letme show u 2019-06-24 13:56:06 https://termbin.com/mibt 2019-06-24 13:56:57 same all cmds was works fine on alpine 3.9.9 2019-06-24 13:56:59 isn't -luv libuv? 2019-06-24 13:57:17 not working in Welcome to Alpine Linux 3.10 2019-06-24 13:57:20 yeah its libuv 2019-06-24 13:57:24 and you have it installed? 2019-06-24 13:57:41 docker run -it alpine /bin/sh <<< its getting the latest one 2019-06-24 13:57:50 btw 64 bits work fine 2019-06-24 13:57:53 issue only in 32 bits 2019-06-24 13:57:59 docker run -it i386/alpine /bin/sh 2019-06-24 13:58:18 same all commands am using on 64 bits its compiled the binary 2019-06-24 13:58:24 but in 32 bit i got those error 2019-06-24 13:58:25 in the end 2019-06-24 13:58:33 can you verify that you do have libuv installed? 2019-06-24 13:58:43 ls /usr/lib/libuv.so* 2019-06-24 13:58:52 # apk add libuv-dev 2019-06-24 13:58:52 OK: 246 MiB in 81 packages 2019-06-24 13:59:03 # ls /usr/lib/libuv.so 2019-06-24 13:59:20 its musl gcc 8 issue in 32 bits 2019-06-24 13:59:25 in alpine 10 2019-06-24 13:59:38 same thing i did on 3.9.9 2019-06-24 13:59:46 works pretty good 2019-06-24 14:00:23 not sure if you're supposed to use libuv-dev or just libuv 2019-06-24 14:00:41 libuv provides libuv.so.1, libuv.so.1.0.0, libuv-dev provides libuv.so 2019-06-24 14:01:15 i am using those steps since alpine 3.3 to 3.9 all version works fine in 64 bits and 32 bits of older alpine 2019-06-24 14:01:25 but issue only in alpine 3.10 32 bits 2019-06-24 14:01:28 64 bits work fine 2019-06-24 14:04:21 here is the another proof of libuv is exist while cmake lemme show u 2019-06-24 14:04:41 -- Found UV: /usr/lib/libuv.so 2019-06-24 14:04:41 -- Looking for syslog.h 2019-06-24 14:04:41 -- Configuring done 2019-06-24 14:04:41 -- Looking for syslog.h - found 2019-06-24 14:04:41 -- Generating done 2019-06-24 14:04:53 please use a pastebin service when pasting multiple lines, such as tpaste.us 2019-06-24 14:05:02 sorry by mistake 2019-06-24 14:05:12 and you're sure it's not a broken symlink, but an actual file? 2019-06-24 14:05:44 What does 'file /usr/lib/libuv.so' give you? 2019-06-24 14:05:57 # file /usr/lib/libuv.so 2019-06-24 14:06:00 /usr/lib/libuv.so: symbolic link to libuv.so.1.0.0 2019-06-24 14:06:11 install both livub and libuv-dev 2019-06-24 14:06:17 and then try 2019-06-24 14:06:18 letme try 2019-06-24 14:06:27 # apk add libuv 2019-06-24 14:06:27 OK: 246 MiB in 81 packages 2019-06-24 14:06:36 so you have both libuv and libuv-dev installed? 2019-06-24 14:06:40 yeah 2019-06-24 14:07:08 What does 'file /usr/lib/libuv.so.1.0.0' give you? 2019-06-24 14:07:19 /usr/lib/libuv.so.1.0.0: ELF 32-bit LSB shared object, Intel 80386, version 1 (SYSV), dynamically linked, stripped 2019-06-24 14:08:03 https://termbin.com/9ghb 2019-06-24 14:08:25 cmake .. -DUV_LIBRARY=/usr/lib/libuv.so 2019-06-24 14:09:41 same commands i used in 64 bits of Alpine 3.10 2019-06-24 14:09:44 its worked 2019-06-24 14:09:53 issue only in 32 bits 2019-06-24 14:10:05 after that i installed alpine 3.9 again and works there too 2019-06-24 14:10:12 ldd /usr/lib/libuv.so.1.0.0 2019-06-24 14:10:26 PhoenixMage ok let it finish first > [ 68%] Building CXX object CMakeFiles/xmrig-notls.dir/src/workers/CpuThread.cpp.o 2019-06-24 14:10:43 I hate cmake 2019-06-24 14:11:14 same issue again > https://termbin.com/rhfh 2019-06-24 14:11:22 # ldd /usr/lib/libuv.so.1.0.0 2019-06-24 14:11:22 libc.musl-x86.so.1 => /lib/ld-musl-i386.so.1 (0xf7ebb000) 2019-06-24 14:11:28 /lib/ld-musl-i386.so.1 (0xf7ebb000) 2019-06-24 14:13:03 so i should go with 3.9.9 ? 2019-06-24 14:13:25 or this issue will be solve 2019-06-24 14:22:54 So anybody knows how to "clone" my hard drive to my USB stick? 2019-06-24 14:23:47 `dd`? 2019-06-24 14:25:01 Or copy the thing over with `rsync` 2019-06-24 14:25:27 or pv 2019-06-24 14:25:43 pv ? 2019-06-24 14:25:52 Is it gonna be bootable? 2019-06-24 14:26:07 if its a block device it should not matter 2019-06-24 14:26:38 Well I want it to boot from my USB and use the whole SSD for my ZFS raid cache 2019-06-24 14:27:27 So it has to be bootable right ? 2019-06-24 14:29:06 what boot loader are you using? 2019-06-24 14:29:18 How can I tell? 2019-06-24 14:30:04 Is it just a normal x86 computer? 2019-06-24 14:30:24 X64 2019-06-24 14:30:40 same thing 2019-06-24 14:30:45 so probably grub 2019-06-24 14:30:56 Yeah, i'm pretty sure it is 2019-06-24 14:31:46 If it was me I would copy all my files to the USB and then reinstall the grub bootloader on the USB though you may have to make changes if it is using UUID for identification 2019-06-24 14:32:24 Okay, only part i'm not sure is reinstalling grub 2019-06-24 14:32:41 I dont think dd would be the best way because you dont know where on the disk the blocks are and I would assume the usb is smaller then teh ssd 2019-06-24 14:32:56 grub-install /dev/sdX 2019-06-24 14:33:04 Yes, but it's pretty empty 2019-06-24 14:33:16 And the partition I want are smaller on the SSD than on the USB 2019-06-24 14:33:21 When you do a block copy with dd it doesnt care about data 2019-06-24 14:33:34 Still doesnt matter 2019-06-24 14:33:40 Ok 2019-06-24 14:33:53 when you use grub-install note its the drive not the partition 2019-06-24 14:34:13 So like sda and not sda1 ? 2019-06-24 14:34:39 indeed 2019-06-24 14:34:52 obviously it being your usb not sdd 2019-06-24 14:34:58 Yeah ^^ 2019-06-24 14:35:30 You should be able to test to make sure it boots before blowing away the SSD too 2019-06-24 14:36:01 Yes, else i'll be in trouble if it dosen't 2019-06-24 14:36:09 I'm still unsure where to start from though 2019-06-24 14:36:26 If the case is open I would pop the cable to test and boot from USB 2019-06-24 14:36:29 And how do I make the USB copy to RAM on boot? 2019-06-24 14:41:04 What do you mean? 2019-06-24 14:41:52 I want the content of my USB to be copied to my ram when I boot the system 2019-06-24 14:42:47 Why? 2019-06-24 14:43:52 is there any obvious reason that bash in alpine won't load ~/.bashrc on login? 2019-06-24 14:44:10 E-werd: .bashrc is not for login shells 2019-06-24 14:44:14 .profile is 2019-06-24 14:44:25 hm 2019-06-24 14:44:31 so unless you include .bashrc in your ~/.profile, it won't get loaded 2019-06-24 14:44:48 i'll change that step then, thanks for the tip 2019-06-24 14:44:53 np 2019-06-24 14:45:01 man bash would clarify that 2019-06-24 14:45:02 ;) 2019-06-24 14:45:03 working with pyenv, it wants some lines and it was driving me nuts 2019-06-24 15:00:10 PhoenixMage: Cause a USB stick is going to be slower? 2019-06-24 15:00:17 I've got 128 gb of ram 2019-06-24 15:01:43 And alpine is so small, I don't see why I would not 2019-06-24 16:17:15 anyone found that the after install 3.10 manually (setup-disk, etc) and boot newly installed system it's root FS mounted 'ro' instead of 'rw' 2019-06-24 16:18:18 'mount -o remount,rw /' when log in works fine 2019-06-24 16:20:22 FS is ext4 and mount flag is 'rw' in /etc/fstab 2019-06-24 16:21:44 relevant part of '/proc/cmdline' is '/root=/dev/sda1 rw rootwait' 2019-06-24 16:22:32 sorry, 'root=/dev/sda1 rw rootwait' (without slash before 'root') 2019-06-24 17:14:16 hmm, needs fsck after install, so fsck tools for FS should/must be installed by setup-disk 2019-06-24 20:39:24 hello all, I'm trying to create a .img file for a raspberry pi 3 (aarch64), however, i'm struggling to do so. could anyone point me in the right direction as to what is wrong with my script? (located at https://gist.github.com/hoshsadiq/888482befed42a649edf5e868a25cf5b) 2019-06-24 20:40:58 the aim is to use packer with this and configure the alpine image before hand so i can build a pipeline around it and easily create images so i can simply edit scripts and wait a bit and burn the images onto a pi 2019-06-24 21:34:47 I'm looking at the docker init.d file and I'm trying to figure out how to get it to stop filling up the disk with stdout logs 2019-06-24 21:35:14 this template seems to always have a hard coded --stdout argument, is there a good way to not do that? 2019-06-24 21:36:42 maldridge: That's how you get to see WTH your containers are doing. If you don't want that, you can set the logging driver to non 2019-06-24 21:36:46 See https://docs.docker.com/config/containers/logging/configure/#configure-the-default-logging-driver 2019-06-24 21:37:00 That question should go to the docker channel, not Alpine. 2019-06-24 21:38:43 I log my containers to fluentd 2019-06-24 21:39:02 this is logging docker's own actions 2019-06-24 21:39:29 given that it's alpine's init file that lacks options, I still consider this an alpine bug 2019-06-24 21:39:41 though if you'd prefer, I can take the issue to gentoo, which is who the file is copyrighted to 2019-06-24 21:42:26 -l, --log-level string Set the logging level ("debug"|"info"|"warn"|"error"|"fatal") (default "info") 2019-06-24 21:42:34 dockerd --help is your friend too. 2019-06-24 21:43:05 again, this doesn't solve the fact that the openrc file is hard coding certain options 2019-06-24 21:43:32 if my best bet is to not use alpine's openrc files, then say that 2019-06-24 21:44:35 Your best bet is if something doesn't work exactly as you'd like, fix it. The openrc file is fine for 99% of the users and that's why it has the options it has. 2019-06-24 21:45:28 man, people think void is abbrasive 2019-06-24 22:28:30 Hello all, just wondering if anyone else might be able to shed some light on this. I'm trying to build an alpine image with erlang 22.0.4 and I keep running into this: dlsym: Resource temporarily unavailable 2019-06-24 22:28:51 this is only occurring in alpine 3.10. If I build in 3.9, it builds fine 2019-06-24 22:31:49 danieli: ^ I guess? 2019-06-24 22:32:20 beardedeagle: could you elaborate on the specifics of the environment you are using? 2019-06-24 22:32:50 is it docker? erlang/otp images? the ones based on 3.10 or 3.9? etc. 2019-06-24 22:35:49 docker 2019-06-24 22:35:50 https://github.com/beardedeagle/alpine-erlang-builder/blob/otp_22_0_4/Dockerfile 2019-06-24 22:36:08 building from source 2019-06-24 22:39:02 error I'm seeing 2019-06-24 22:39:02 https://gist.github.com/beardedeagle/b4d6b0c6105a8f5e8501df8b33c191b3 2019-06-24 22:39:20 I'll do some testing 2019-06-24 22:39:39 thing is, they're building erlang themselves in that image 2019-06-24 22:39:57 er, the one that image is derived from 2019-06-24 22:41:11 yeah I build images for the community specifically for multistage builds, typically I keep my images up to date but something between alpine 3.9 and 3.10 is causing builds to fail 2019-06-24 22:44:37 I see this issue, but it appears to be fixed by now? https://bugs.alpinelinux.org/issues/9983 2019-06-24 22:49:27 as far as i understand from the previous erlang maintainer, erlang makes use of certain musl internals, and needs to be rebuilt whenever musl is upgraded, but i don't think that's related to this 2019-06-24 22:51:30 ah, i've seen that exact error before, let me just take a moment to remember what it was 2019-06-24 22:51:43 i remember hitting a wall with that at first 2019-06-24 22:51:48 thanks danieli, appreciate the help 2019-06-24 22:52:59 beardedeagle: try patching with this file https://git.alpinelinux.org/aports/tree/community/erlang/safe-signal-handling.patch 2019-06-24 22:56:20 I'll give that a go, thanks 2019-06-24 22:56:49 hm - 21.2.6? that doesn't look quite right 2019-06-24 22:57:06 yeah I just noticed that too 2019-06-24 22:57:39 it does apply though 2019-06-24 22:57:51 iirc i even changed it a bit during the upgrade to 22.0.2 2019-06-24 23:08:54 beardedeagle: did you figure it out? 2019-06-24 23:09:14 this is the 'recipe' / 'template' that defines how the erlang package on alpine is built by the way - https://git.alpinelinux.org/aports/tree/community/erlang/APKBUILD 2019-06-24 23:09:23 analogous with, say, arch's PKGBUILD files 2019-06-24 23:11:44 yeah I'm trying to figure out at which point I would run the patch based off of that. From that file I assume it would be after I autoncif 2019-06-24 23:11:49 autoconf* 2019-06-24 23:13:40 before build and configuratio 2019-06-24 23:13:43 configuration* 2019-06-24 23:13:49 pretty much just after you extract the tarball 2019-06-24 23:14:03 ah, just hop into the folder and patch apply 2019-06-24 23:21:08 bingo 2019-06-24 23:21:11 abuild does it automatically though 2019-06-24 23:23:55 k, it's building. Crossing fingers that it works 2019-06-24 23:34:18 well it made it past that step, not it's failing because of something else. I'll dig on my end now that I've progressed beyond. Thanks for your help. 2019-06-24 23:39:41 actually, you guys/gals might be able to shed some light on this faster than I can too. https://gist.github.com/beardedeagle/f485c46f38170e4d6ed1bde08732b90e 2019-06-24 23:39:55 looks like it's a musl constraint issue? 2019-06-24 23:45:29 hmm, i'd recommend not running `apk add' with so: or cmd: prefixes 2019-06-24 23:45:36 in my experience it's been prone to breaking dependencies 2019-06-25 01:14:27 alpine 3.10 is now available on builds.sr.ht 2019-06-25 01:14:37 sorry for the wait, it wasn't that much work but I may have forgotten to do it <_< 2019-06-25 01:41:15 bash 5.0 is here by default 2019-06-25 07:25:42 Hie, I have been encountering build failures for "GNU Smalltalk". 2019-06-25 07:25:46 http://0x0.st/z2f9.log 2019-06-25 07:28:03 And here is my APKBUILD: http://0x0.st/z2ff.tx 2019-06-25 07:28:07 And here is my APKBUILD: http://0x0.st/z2ff.txt 2019-06-25 07:35:40 ... 2019-06-25 07:36:19 I wonder what ```checking build system type... config.sub: missing argument``` portion fails... 2019-06-25 07:37:03 I have ```makedepends="gdbm-dev readline-dev sqlite-dev zip pkgconf autoconf automake gawk libtool"``` specified. Am I missing anything? 2019-06-25 07:41:57 Looks like config.sub is a part of autotools 2019-06-25 07:52:51 Ok.. I fixed it. There was a small typo in that backslash 2019-06-25 09:29:48 the ``` doesn't render on IRC by the way 2019-06-25 09:54:24 danieli: I think that's dependant on the client 2019-06-25 09:54:47 indeed, but *most* don't, and IRC doesn't differentiate at all 2019-06-25 09:58:00 we are on IRC and 'when you are in Rome speak as Roman' 2019-06-25 10:01:42 It won't render on IRC but it'll send a link to IRC when there is a multiline message 2019-06-25 10:02:46 Also, even though `this won't be rendered as code on IRC`, it'll still look good on Matrix. No reason not to send it really 2019-06-25 10:02:50 ACTION ignores all these ``` 2019-06-25 10:04:06 although I'm considering matrix but missing server for Alpine on ARM 2019-06-25 10:05:05 Missing server for Alpine on ARM? You mean you're missing server side software? Synapse is Python and should run fine 2019-06-25 10:05:12 The Synapse docker images are based on Alpine even 2019-06-25 10:07:18 I have seen 'ruma' written in rust, and waiting for rust on arm to try ruma as server 2019-06-25 10:08:14 Considered synapse, but it looks to complicated to 'my eyes' 2019-06-25 10:08:30 Ruma would be nice, but it's a far away from being usable 2019-06-25 10:08:42 More chance with either Dendrite (written in Go) or Construct (written in C/C++) 2019-06-25 10:08:52 Not sure how Synapse is complicated though 2019-06-25 10:09:02 Construct? didn't know. will look 2019-06-25 10:09:32 when I see server written in python I feel stomach cramps 2019-06-25 10:10:35 that is 'my personal' opinion, I don't like to be judge at the end 2019-06-25 10:59:31 english is hard 2019-06-25 11:02:33 mps: tbf Synapse isn't really great. It has performance and scaling issues. If you can afford to wait for alternative homeservers, I'd do it yes 2019-06-25 11:50:14 I connect to kvm on Alpine using virt-manager on another machine with QEMU+SSH, having set up the polkit rule as per the Alpine wiki (i.e. in /etc/polkit-1/localauthority/50-local.d/50-libvirt-ssh-remote-access-policy.pkla). This worked fine until I upgraded to 3.10. Now I get an 'Unable to connect' error in virt-manager saying 'CheckAuthorization: 2019-06-25 11:50:14 Process PolicyKit1 received signal 9. I can still connect via SSH and use virsh, but the remote virt-manager connection always fails. Any ideas? 2019-06-25 11:51:21 Relevant Wiki page is https://wiki.alpinelinux.org/wiki/KVM 2019-06-25 12:10:33 interesting 2019-06-25 12:10:57 I never had to do anything with polkit to have virt-manager connected to alpine vm 2019-06-25 12:11:36 I thought virt-manager uses spice to have a gui view of vm 2019-06-25 12:12:00 also virt-manager via ssh 2019-06-25 12:13:58 my laptop's ~/.ssh/config has this : 2019-06-25 12:14:09 Host the-host-that-runs-virt-manager-and-alpine-vm 2019-06-25 12:14:10 ProxyCommand ssh gateway-user@gateway -W %h:%p 2019-06-25 12:14:13 HostName %h 2019-06-25 12:14:16 Compression yes 2019-06-25 12:14:19 ForwardX11 yes 2019-06-25 12:15:11 so the gatway stays between my laptop and the virt-manager host. On my laptop I run : $ ssh the-host-that-runs-virt-manager-and-alpine-vm then $ virt-manager. 2019-06-25 12:15:55 fiddlinmacx : ^ 2019-06-25 12:29:25 I have successfully bootstrap a cross compile env, is it possible to use the available packages for my target host within that env? 2019-06-25 12:33:06 If so how? Do I put a repository file in sysroot-armhf and then use apk add some how? 2019-06-25 12:53:53 tmhoang I've had it set up ala the Wiki instructions for quite awhile and it's worked beautifully. The Archlinux wiki says something quite similar. I don't have a gui on Alpine, I'm just running virt-manager on my remote Linux box. I'm using virt-manager to manage libvirt, including creating VMs and snapshots, etc. 2019-06-25 12:57:05 i have the strange behavior that im pretty sure that some days ago the darkhttp package had a init file in it 2019-06-25 12:57:13 and my tests ran fine 2019-06-25 12:57:16 and now its missing 2019-06-25 12:57:23 i found it in the subpackage darkhttp-openrc 2019-06-25 12:57:36 but still... i wonder what happend... or im just blind 2019-06-25 12:57:44 because the package build date is on december 2018 2019-06-25 13:01:55 say I got an alpine installation on a USB, how do I upgrade to a newer release? 2019-06-25 13:02:20 is it enough to download the new ISO and then setup-bootable on to /media/usb ? 2019-06-25 13:08:13 qrvpzvb: yes that should do it 2019-06-25 13:08:49 xsteadfastx: openrc subpkg should automatically be installed when you have openrc installed 2019-06-25 13:12:07 Wouldn't just `sed s/3.9/3.10/g -i /etc/apk/repositories` do the trick? 2019-06-25 13:12:14 Or do we always recommend a reinstall? 2019-06-25 13:16:11 tmhoang considerd using ProxyJump? 2019-06-25 13:16:21 considered* 2019-06-25 13:16:23 aka `ssh -J' 2019-06-25 13:16:38 Cogitri: upgrade should work fine 2019-06-25 13:16:52 dont forget to add --available 2019-06-25 13:16:55 danieli : what is the diff ? And do those related to recent Jenkins CI hack ? 2019-06-25 13:17:29 Cogitri: something like apk -U upgrade -a 2019-06-25 13:17:31 tmhoang: it's not much more than a shorthand for using ProxyCommand like you did 2019-06-25 13:17:36 it's just way more convenient 2019-06-25 13:19:53 fiddlinmacx: yes virt-manager is on remote Linux box. And I don't why one needs to have "libvirtd to remotely control KVM over ssh PolicyKit". What one needs virt-manager/libvirt to control KVM guest ? 2019-06-25 13:20:05 just ssh into that guest and execute the command 2019-06-25 13:20:42 danieli: need to look into ProxyJump. For now I don't have to pass anything to $ ssh command. Easy of mind. 2019-06-25 13:22:25 $ ssh host-behind-gateway. that's all. It doesn't feel like I have a proxy at all :D 2019-06-25 13:23:01 The reason I want to "reinstall" is that I'm on a diskles install and as far as I can tell, the only way to upgrade kernels in by reinstalling 2019-06-25 13:23:41 I have successfully bootstrap a cross compile env, is it possible to use the available packages for my target host within that env? -> yes 2019-06-25 13:23:45 qrvpzvb: correct 2019-06-25 13:23:49 not reinstalling 2019-06-25 13:23:56 you need an upgrade 2019-06-25 13:24:14 setup-bootable has an upgrade flag 2019-06-25 13:24:18 PhoenixMage> If so how? Do I put a repository file in sysroot-armhf and then use apk add some how? --> /sysroot-armhf/etc/apk/repositories should work if I'm not wrong 2019-06-25 13:25:18 clandmeter: which basically just leaves some config files alone, says the wiki 2019-06-25 13:25:53 only one file 2019-06-25 13:26:00 ovl file 2019-06-25 13:26:17 make sure you have a backup of it 2019-06-25 13:26:41 tmhoang: I'm not completely sure what you mean. I don't bother controlling the guest with virt-manager (except for occasionally initiating a Spice connection). I use virt-manager as a remote front-end to libvirt since my Alpine box is headless. I create VMs and manage snapshots with virt-manager, although the snapshots are easy enough to handle wit 2019-06-25 13:26:42 h virsh. I have mostly Windows guests anyway and use RDP/PowerShell Remoting to connect to them. 2019-06-25 13:26:49 danieli: ah ProxyJump seems handy 2019-06-25 13:27:04 it's a *fairly* new option, but yeah, it's decent 2019-06-25 13:32:01 clandmeter: Ah, Alright 2019-06-25 13:34:18 fiddlinmacx: Alpine 3.10 comes with qemu 4.0. Feels like they switched the polkit file format to javascript ? 2019-06-25 13:35:32 tmhoang: Oh? That's interesting. I'll look into that. That's the first clue I've gotten about what changed. 2019-06-25 13:37:57 fiddlinmacx: seems true. https://wiki.libvirt.org/page/SSHPolicyKitSetup says `As of polkit 0.106 the .pkla format is no more`. Alpine 3.9 polkit is 0.105-r9 while 3.10 is 0.116-r0. 2019-06-25 13:39:37 Cogitri: fyi, apk normally only compares version, higher version wins from index. if a package has been rebuild but keeps the same version it will ignore it. adding -a will also compare csum and pull in the new pkg if it does not match. 2019-06-25 13:44:05 tmhoang: Aw man! ;-) OK. Good find. Thanks. 2019-06-25 13:44:55 I have qemu 4.0 installed and no polkit, at least I can't find it on my machine 2019-06-25 13:45:28 fiddlinmacx: if you could eventually fix this with 3.10, please help editting the wiki page. Much appreciated :D 2019-06-25 13:45:58 tmhoang: Yeah. I'll do that once I have it figured out. No problem. 2019-06-25 13:46:33 man pages, here I come. 2019-06-25 16:18:18 Is there any problem with the apparmor package ? 2019-06-25 16:18:50 * Caching service dependencies ... [ ok ] 2019-06-25 16:18:50 * Starting AppArmor ... 2019-06-25 16:18:50 * Caching service dependencies ... [ ok ] 2019-06-25 16:18:50 * arg 2019-06-25 16:18:50 * ERROR: apparmor failed to start 2019-06-25 16:18:53 * Starting AppArmor ... 2019-06-25 16:18:55 * arg 2019-06-25 16:18:57 * ERROR: apparmor failed to start 2019-06-25 16:19:04 Sorry for that, didin,t mean to copy it 3 times :( 2019-06-25 16:58:57 That error really doesn't tell us anything, go see your system's logs 2019-06-25 17:00:55 cogitri: haha good to see ya once more ;) 2019-06-25 17:03:19 Hehe, you're on Alpine too? :) 2019-06-25 17:04:00 I'm on alpine for years 2019-06-25 17:04:55 my master thesis' project based on alpine :) 2019-06-25 17:06:13 *is based 2019-06-25 17:10:34 Oh :o 2019-06-25 17:14:35 wow - that's rad! 2019-06-25 17:15:54 indeed :) 2019-06-25 17:18:58 That's nice 2019-06-25 17:23:36 thx maxice8 2019-06-25 17:23:53 nice to read ya too 2019-06-25 17:24:17 world is small :) 2019-06-25 17:29:10 they've been putting in one hell of an impressive effort 2019-06-25 17:43:30 absolutely confirmed ! 2019-06-25 17:45:45 are able to run alpine on my mobile device too, using postmarket os ... that's unique :) 2019-06-25 17:46:47 Hi! How does one enable the wpa_supplicant to do eap-pwd? 2019-06-25 17:46:59 https://postmarketos.org/ 2019-06-25 17:56:25 Yup, that's pretty cool, but I guess I'll opt for the Librem 5 because I want my phone to "just work" (sort of) 2019-06-25 17:58:41 yeah, that's why it's only on my nightly phone (nexus 5) ... to play around :) 2019-06-25 18:00:11 but it works a lot more "stable" than apple devices IMOH xD 2019-06-25 18:00:50 Cogitri: Would you please point me where those logs are ? 2019-06-25 18:01:17 <_ikke_> most likely in /var/log/ 2019-06-25 18:01:56 And then? 2019-06-25 18:02:31 I have no rc nor no apparmor logs in this folder 2019-06-25 18:05:21 Well, install some syslog daemon (e.g. metalog) and then see /var/log/everything/current 2019-06-25 18:07:12 shizonic: Oh my, maybe I should revive an old Nexus 5 off eBay or smth then 2019-06-25 18:10:21 Haha yes. We've been loving having Alpine as a base so far 2019-06-25 18:10:36 Although we definitely need to move to Alpine stable rather than edge 2019-06-25 18:12:14 cogitri: the Nexus 5 actually fully works on mainline with just a few patches. Only audio is left really 2019-06-25 18:12:16 If you can get your hands on one, I definitely recommend it 2019-06-25 18:12:21 cogitri: I suggest to :) 2019-06-25 18:12:33 Also, pmOS will work fine on the Librem 5 2019-06-25 18:13:33 PureTryOut: Running pmOS on the Librem 5 seems a bit nonsensical though, since I'm paying premium for the software support from Librem. I'd rather get a Pine Phone then, I guess 2019-06-25 18:14:18 You're mainly paying for the development of Phosh really, which will also work on pmOS 2019-06-25 18:14:54 PinePhone is definitely nice though. I have a devkit, I can recommend them :D 2019-06-25 18:15:20 This is largely because of your work on GNOME btw 2019-06-25 18:15:38 Now if only gnome-session were enabled for aarch64 ;) 2019-06-25 18:16:55 Hehe, I can get to that after Rust on more arches :) 2019-06-25 18:19:32 cogitri: or even rust on mobile ... :D 2019-06-25 18:19:52 Well, that's what that'd do :P 2019-06-25 18:20:03 Fractal would certainly be a nice addition to pmOS 2019-06-25 18:20:10 nice ! 2019-06-25 18:21:17 If you you're looking for testers - I'm here with pleasure ... 2019-06-25 18:27:55 clandmeter: i tested it with the alpine:3.9 docker image... and no... it doesnt install darkhttpd-openrc when i do a apk add darkhttpd 2019-06-25 18:32:14 xsteadfastx: openrc on docker? 2019-06-25 18:32:55 shizonic: its not about that openrc works in docker... its about the dependency that darkhttpd installs darkhttpd-openrc 2019-06-25 18:32:59 for the init file 2019-06-25 18:33:17 i deploy a darkhttpd server through ansible to a alpine box and run tests 2019-06-25 18:33:24 in docker through ansible-molecule 2019-06-25 18:33:32 ok, I see 2019-06-25 18:33:39 and since last night my test had an error 2019-06-25 18:33:55 that the init script is missing (where i do some changes on.. document_root) 2019-06-25 18:34:14 without any commit on the ansible role itself 2019-06-25 18:34:26 shizonic: Will do :) 2019-06-25 18:34:42 so i wonder if something changed on the package itself... but it looks like it didnt... and i wonder why this test is failing since last night 2019-06-25 18:35:21 no big problem... because i add darkhttpd-openrc to the dependencies... and test works again... i just wondered :) 2019-06-25 18:36:57 <_ikke_> xsteadfastx: For the record, is openrc installed or not in the container 2019-06-25 18:36:59 <_ikke_> ? 2019-06-25 18:37:25 xsteadfastx: guess it's a subpackage 2019-06-25 18:37:30 <_ikke_> Unless openrc is installed (which is unlikely in a docker container), darkhttpd-openrc is not installed 2019-06-25 18:37:34 <_ikke_> by default 2019-06-25 18:38:09 try pkg add darkhttpd-openrc 2019-06-25 18:38:28 <_ikke_> shizonic: they already mentioned it works if they manually install it 2019-06-25 18:38:34 _ikke_: nope it isnt... and it makes sense and everything but i just wondered why the test succeseded in the first place :) so i thought maybe something changed 2019-06-25 18:39:08 cogitri: :) 2019-06-25 18:39:11 <_ikke_> xsteadfastx: Did you upgrade the image in the mean time? 2019-06-25 18:39:20 <_ikke_> xsteadfastx: the -openrc splic is 'fairly' recent 2019-06-25 18:39:21 https://github.com/alpinelinux/aports/blob/0042419fccace3247609804c5d502defdd594da7/main/darkhttpd/APKBUILD 2019-06-25 18:39:44 _ikke_: yes 2019-06-25 18:40:07 <_ikke_> Hmm, not as recent as I thought 2019-06-25 18:40:08 _ikke_: ok excuse 2019-06-25 18:40:25 _ikke_: thats what i thought... its from 2018 2019-06-25 18:40:33 still strange.... :) 2019-06-25 18:40:36 <_ikke_> 3.7 already had the -openrc subpackage 2019-06-25 18:42:39 _ikke_: so I was right anyway ? 2019-06-25 18:43:26 <_ikke_> shizonic: The premise that it was not necessary before to explicitly install darkhttpd-openrc 2019-06-25 18:45:21 _ikke_: ok 2019-06-25 18:45:50 strange indeed 2019-06-25 18:46:02 <_ikke_> for v3.6 that would make sense 2019-06-25 18:46:19 <_ikke_> starting from v3.7, there is a subpackage, which is not a dependency of darkhttpd 2019-06-25 18:47:10 _ikke_: yes, I saw that... 2019-06-25 18:47:49 <_ikke_> The default_openrc function has install_if=".." defined 2019-06-25 18:48:02 <_ikke_> which makes it automatically install that subpackage if openrc is installed 2019-06-25 18:48:40 so guess it's a bug ... 2019-06-25 18:49:09 <_ikke_> It's not a bug that it's not installed on a docker container 2019-06-25 18:49:13 but not on docker 2019-06-25 18:49:24 <_ikke_> as the default alpine images on docker don't have openrc installed 2019-06-25 18:49:44 that's what I said ;) 2019-06-25 18:50:10 <_ikke_> yes 2019-06-25 18:50:29 sry, my english is bad 2019-06-25 18:50:34 <_ikke_> no problem 2019-06-25 18:50:56 guess there were just some missunderstandings 2019-06-25 18:51:26 we talked about the same 2019-06-25 23:01:10 what's the trick to enable xf86-video-nv? 2019-06-25 23:01:28 I had nouveau on, but it didn't allow a 2nd monitor 2019-06-25 23:01:43 removed that, it appears to be on vesa... poor 2d performance, choppy 2019-06-25 23:01:55 documentation is incomplete for nv 2019-06-25 23:10:25 oh, looks like "linux-firmware-nvidia" is a clue 2019-06-26 01:42:55 Anyone tried compiling VTK for alpine? 2019-06-26 01:45:57 I'm getting Found OpenGL: /usr/lib/libGL.so found components: OpenGL missing components: GLX 2019-06-26 01:46:29 But libglx.so is installed by xorg in /usr/lib/xorg/modules/extensions/libglx.so 2019-06-26 01:47:00 That's not the same thing 2019-06-26 02:21:06 for whatever reason, Engrampa only wants to read ZIP files 2019-06-26 02:23:34 or rather, it doesn't want to read .tar.xz 2019-06-26 02:23:45 or anything else I've thrown at it so far, including .tar.gz 2019-06-26 02:27:40 fwiw, xz is indeed installed 2019-06-26 05:24:35 _ikke_: Doesn't ikke mean "cumming" in Japanese? 2019-06-26 05:25:11 oh dear 2019-06-26 05:41:04 So how would I add packages to my sysroot-armhf? I thoguht when I do CHOST=armhf CTARGET=armhf abuild it would install the makedepend packages into the sysroot-armhf but that doesnt seem to be the case... 2019-06-26 05:41:26 I have put the repository file in the correct place in the sysroot 2019-06-26 09:02:45 ok, I think I have figured out why the APKBUILD deps arent being installed in the sysroot, however now the APKINDEX signature is failing when trying to install packages despite the fact I have the same keys in /etc/apk/keys in both base system and sysroot 2019-06-26 09:49:42 So how would I add packages to my sysroot-armhf? I thoguht when I do CHOST=armhf CTARGET=armhf abuild it would install the makedepend packages into the sysroot-armhf but that doesnt seem to be the case... 2019-06-26 09:50:02 if you are not inside the chroot, you may need to specify --sysroot option somewhere in abuild or apk 2019-06-26 09:50:24 why don't # chroot /mnt/sysroot sh and then apk add packages or abuild in there 2019-06-26 09:50:53 also what is your host ? x86 or armhf ? glibc or musl ? 2019-06-26 10:33:15 Awesome, got Alpine Linux up and running on a LVM on LUKS setup with an additional LVM on LUKS drive and EFI 2019-06-26 10:33:24 That was quite a lot of work lol... 2019-06-26 10:45:28 Luckily ZFS w/ encryption is like 5 commands :P 2019-06-26 10:48:39 PureTryOut: 2019-06-26 10:48:40 I used the alpine installer... and syslvm option. My system is not encrypted yet though. 2019-06-26 10:49:00 Well someone wrote a quite good wiki page for it https://wiki.alpinelinux.org/wiki/LVM_on_LUKS 2019-06-26 10:49:18 I used GRUB rather than syslinux though so I had to change a few things 2019-06-26 10:49:33 tmhoang: I am on x86 building to armhf 2019-06-26 10:50:04 It appears an APKBUILD needs makedepends_host to actually install the packages into the sysroot 2019-06-26 10:50:04 I might change the page and split the bootloader stuff into it's own section 2019-06-26 10:50:13 cogitri: Is that ZFS on single disk would make more sense than using LVM on same setup? 2019-06-26 10:50:47 Sure ZFS design is meta but.. 2019-06-26 10:50:53 tmhoang: I couldnt find that really documented anywhere I ended up reading the abuild script 2019-06-26 10:51:29 Using musl 2019-06-26 10:51:47 chroot isnt an option due to the architecture change 2019-06-26 10:52:10 errn0: It makes more sense imho because of zfs features 2019-06-26 10:52:22 Checksumming and cow is a must have to me 2019-06-26 10:52:52 Snapshots are nice too, but that can be done with LVM too iirc 2019-06-26 10:53:05 cogitri: Does not LVM do COW? 2019-06-26 10:53:43 Don't think so, that's usually the FS' job 2019-06-26 10:55:49 cogitri: "LVM allows you to take a snapshot of your system in a much more efficient way than a traditional backup. It does this efficiently by using a COW (copy-on-write) policy. " 2019-06-26 10:56:00 From Arch Linux wiki 2019-06-26 10:57:26 I think ZFS does COW even on a raw system (without snapshot) 2019-06-26 11:08:59 PureTryOut[m], did you add luks keys to your initrd? 2019-06-26 11:09:55 or just got unencrypted /boot? 2019-06-26 11:09:59 Awesome, got Alpine Linux up and running on a LVM on LUKS setup with an additional LVM on LUKS drive and EFI --> a wiki page ? 2019-06-26 11:10:19 https://wiki.alpinelinux.org/wiki/LVM_on_LUKS 2019-06-26 11:10:23 sorry 2019-06-26 11:11:26 : I think what you are doing is pretty tough. Give qemu armhf a try, your life would be easier. 2019-06-26 11:12:03 last time I was doing s390x, as soon as I got a sysroot, I move it to native glibc-based s390x machine to run inside the chroot over there 2019-06-26 11:12:22 so if I were you I'd use qemu emulation 2019-06-26 11:17:18 MY_R: unencrypted /boot 2019-06-26 11:17:42 ah ok 2019-06-26 11:21:48 I'm not entirely sure what the benefit of an encrypted /boot is 2019-06-26 11:30:26 'evil maid' attack 2019-06-26 11:41:26 errn0: It does that for snapshots, but I'm not quite positive if it does that for normal FS operations too 2019-06-26 11:42:38 ZFS also does it for normal operations, which can avoid data corruption during power outages/hard resets 2019-06-26 11:50:21 I'm looking for someone who can do a bit of contracting for me from time to time. Where's a good place to look? 2019-06-26 11:57:52 instead of an encrypted /boot you can also just put your /boot on a usbstick and keep physical contact with it all the time. 2019-06-26 11:59:44 or sdcard instead of usbdisk, those are smaller and can be worn all the time. 2019-06-26 12:00:49 and if you also do detached luks headers, then your disk can be truly fully encrypted, not only in name 2019-06-26 12:01:35 with an encrypted /boot you still have the problem, that you need something that is not encrypted, to decrypt the boot, so you really only shift the target of the evil maid from /boot to the boot-decryptor. 2019-06-26 12:02:44 p4Wv1qn095FW: evil maid can flash your 'bios' 2019-06-26 12:02:51 yeah 2019-06-26 12:03:11 there's mitigations for that also 2019-06-26 12:03:36 unless the evilmaid resolders flashchips 2019-06-26 12:03:37 yes, build you own computer from components 2019-06-26 12:04:02 nah, thinkpads are pretty well understood against such attacks. 2019-06-26 12:04:19 coreboot/heads + some hw mods go a long way 2019-06-26 12:04:32 start with old MCS51, Z8 or 6502 ;) 2019-06-26 12:19:52 tmhoang: I now have it largely sorted, I hacked around the keys issue, no idea what the root cause is 2019-06-26 14:24:21 After upgrading to 3.10 I'm suddenly getting a dmesg error when connecting to libvirt from a remote computer using virt-manager. 2019-06-26 14:25:27 polkitd[26672]: segfault at 0 ip 00006fed6f13c984 sp 00007ca63b766960 error 6 in libmozjs-60.so.0[6fed6ef98000+4b2000] 2019-06-26 14:25:56 and then polkitd crashes with a segfault 2019-06-26 14:29:37 I just realized that running 'js60' as root produces a segfault. I don't thing this is polkit at all. 2019-06-26 14:44:00 fiddlinmacx: Saw your report on b.a.o, will look into it in a bit 2019-06-26 14:44:12 But js60 is mozjs60, yes 2019-06-26 14:44:18 Without -dbg packages I can't really do anything about it though.. 2019-06-26 14:45:08 Oh. OK. I'm not sure how to proceed. I got as far as I could in the debugging process, but It seems to be js60. Is there anything else I can try? 2019-06-26 14:49:24 Is there a way to automatically install any -doc package together with the main package? 2019-06-26 14:50:01 You can install `doc` (or is it `docs`?), which will install all -doc packages for your installed main pkgs 2019-06-26 14:50:04 Same thing for lang 2019-06-26 14:53:15 For installed main packages? So not for non-main packages, and won't do it automatically when a new package is installed? 2019-06-26 15:20:39 cogitri: I have used ZFS in FreeBSD, it was fantastic. 2019-06-26 15:21:19 But sometimes when I copy data to flash drive, I get corrupted copies. 2019-06-26 15:21:51 I didn't know how to audit what was happening 2019-06-26 15:22:25 Hm, works for me 2019-06-26 15:22:40 It will automatically install them when you install new packages 2019-06-26 15:24:00 It's like this: the default doc()/lang() add a `install_if="$pkgname-$pkgver-r$pkgrel doc/lang` to the package 2019-06-26 15:24:53 cogitri: That what pretty much happens to me. But sometimes, things go out of my expectation. I needed to learn about debugging COW based FS 2019-06-26 15:26:47 ACTION sent a long message: < https://matrix.org/_matrix/media/v1/download/matrix.org/sjXWBDjvKXieCpHMkOSfYHqc > 2019-06-26 15:27:10 I just use mdocml instead of man-db 2019-06-26 15:28:26 Yeah, but a bit strange that it gets installed in the first place 2019-06-26 15:28:54 Maybe you installed man-db manually? 2019-06-26 15:29:10 I did 2019-06-26 15:29:13 I was talking about mdocml 2019-06-26 15:29:42 cogitri: I have used dsbmc-cli.. Maybe I shouldn't. 2019-06-26 15:29:54 Ah, I guess it's just used because we don't really have providers 2019-06-26 15:32:32 Does anyone know what the linux-virt kernel package is meant for? 2019-06-26 15:32:54 fiddlinmacx: virtual machine guests, it's slimmed down 2019-06-26 15:32:59 (virt)ual 2019-06-26 15:33:37 Ah. OK. I'm running Alpine as a Host not a guest. Thanks. 2019-06-26 15:34:44 For some reason my Alpine system is using linux-hardened. Is this a leftover from a previous release? I've upgraded a few times since I installed. 2019-06-26 15:38:11 fiddlinmacx: My Alpine Host runs on Linux-vanilla kernel 2019-06-26 15:38:30 Alpine Edge 2019-06-26 15:44:28 yes, we no longer use linux-hardened 2019-06-26 15:44:35 you can remove hardened and add vanilla 2019-06-26 15:50:07 danieli: Why so? 2019-06-26 15:50:17 M0x32_t[m]1: why we abandoned it? 2019-06-26 15:50:33 I meant why you dropped harden kernel? Maintainence issue ? 2019-06-26 15:50:45 danieli: Yes... 2019-06-26 15:50:58 <_ikke_> M0x32_t[m]1: THey closed it and required you to pay for it 2019-06-26 15:51:43 _ikke_: Grsecurity team? Closed it means they closed access to source code? 2019-06-26 15:51:51 yes 2019-06-26 15:51:52 <_ikke_> M0x32_t[m]1: yes 2019-06-26 15:51:56 linux-hardened was an unofficial fork 2019-06-26 15:52:04 but it essentially became unmaintaind and fell behind 2019-06-26 15:52:27 danieli: Ok. So vanilla is mainline? 2019-06-26 15:52:31 <_ikke_> yes 2019-06-26 15:52:33 yes 2019-06-26 15:52:42 I see 2019-06-26 21:14:51 #canal una consulta alguién sabe como configurar el dockerfile con la imagen python:3.6-alpine y agregar unixodbc-dev para conectarse uno remotamente al mssql ¿? 2019-06-26 21:16:18 english please 2019-06-26 21:16:45 #channel a query someone knows how to configure the dockerfile with the python image: 3.6-alpine and add unixodbc-dev to connect one remotely to the mssql? 2019-06-26 21:17:38 this is my worst code I can not connect to sql 2019-06-26 21:17:44 FROM python:3.6-alpine 2019-06-26 21:17:44 RUN apk update && apk upgrade \ 2019-06-26 21:17:44 && apk add zsh gcc libxml2-dev libxslt-dev \ 2019-06-26 21:17:44 && apk add unixodbc-dev \ 2019-06-26 21:17:44 && apk add --no-cache bash git openssh \ 2019-06-26 21:17:44 && rm -rf /var/cache/apk/* 2019-06-26 21:17:46 RUN echo "[FreeTDS]\n\ 2019-06-26 21:17:50 Description = FreeTDS unixODBC Driver\n\ 2019-06-26 21:17:52 Driver = /usr/lib/odbc/libtdsodbc.so\n\ 2019-06-26 21:17:54 Setup = /usr/lib/odbc/libtdsS.so" >> /etc/odbcinst.ini 2019-06-26 21:17:56 ONBUILD ADD app /app/ 2019-06-26 21:17:58 WORKDIR /app 2019-06-26 21:18:32 please use a pastebin service 2019-06-26 21:21:23 gcastrejon: as clandmeter said, please use a pastebin service - and it soudns like you're looking for #python 2019-06-26 21:22:25 okas tks 2019-06-26 22:14:43 abuild -r installs package deps with apk add 2019-06-26 22:14:47 abuild -R finds them and compiles them 2019-06-26 22:14:55 is there a way to install deps which are present in the repos and compile the rest? 2019-06-26 22:16:10 like -r but -R for missing packages? 2019-06-26 22:16:14 not that I know of, but that could be useful 2019-06-26 22:16:15 yeah 2019-06-26 22:16:49 I think I have some convenience scripts that can do something along those lines, I just need to find them lol 2019-06-26 22:17:28 my use-case is that I generally just want to compile the one package, so -r is preferred 2019-06-26 22:17:32 speeds things up 2019-06-26 22:17:52 but when I need to rebuild a bunch of packages for a new alpine release or whatever, it's easier to use -R than to figure out deps some other way 2019-06-26 22:17:59 since they have to be built in order when the deps don't exist in the repo 2019-06-26 22:18:08 and I'm probably rebuilding most packages anyway at that point 2019-06-27 01:43:09 does anybody have any tips on how to enable to nvidia driver via xf86-video-nv? 2019-06-27 01:43:23 I'm unable to find anything on that package, other than its existence 2019-06-27 01:44:37 nv != nvidia != nouveau 2019-06-27 01:45:00 what is it, then? 2019-06-27 01:45:05 nv is dead 2019-06-27 01:45:36 you *probably* want noveau, if your chip is supported by it 2019-06-27 01:45:42 so what's to be done with nvidia? because nouveau is not really a good solution 2019-06-27 01:45:51 I need dual monitors and it *does not* provide that 2019-06-27 01:45:54 nvidia's proprietary driver will not run on musl libc 2019-06-27 01:46:07 and nv was deprecated almost a decade ago 2019-06-27 01:46:28 ah, I didn't have nvidia cards a decade ago so I didn't realize it was an ancient driver 2019-06-27 01:48:06 well, that's quite a pickle then. I can have dual screen with jittering via vesa, or I can have one smooth monitor and (almost) nothing on the other with nouveau 2019-06-27 01:51:23 if you want to use alpine, it is indeed quite a conundrum 2019-06-27 02:01:15 and that's the problem: i do 2019-06-27 02:24:20 it doesn't even make sense. the screen is on but blank, my mouse pointer moves across it clearly and smoothly. once my cursor is on it, the top 5th of the screen shows a garbled tiled version of my motherboard post image 2019-06-27 02:25:08 there are good reasons linus torvalds flipped nvidia the bird :P 2019-06-27 02:25:54 well I've got two of the damn things 2019-06-27 02:27:10 fglrx was the driver name I was trying to think of earlier, that's the last ATI card I had 2019-06-27 02:27:27 me too, I'm running the proprietary nvidia driver as the only decent option with my current cards 2019-06-27 02:29:13 the unspoken downsides to having good hardware 2019-06-27 02:33:22 it's a shame the firm- and software sucks so bad 2019-06-27 04:03:42 danieli: remember back in the day when, if you wanted decent graphics drivers on Linux, nvidia was your best bet? 2019-06-27 04:03:46 weird how the tables turn 2019-06-27 04:06:41 e-werd, that's literally never been the case 2019-06-27 04:07:37 nvidia was always the hell that new ex-windows converts were stuck with, dealing with crashing X, visual glitches, etc. all the time 2019-06-27 04:08:34 (recalling from my work on mplayer around 2001-2006 or so) 2019-06-27 07:40:54 Aaa.. I want to use lua. The current alpine version is old. Can someone upgrade the lua to latest version 5.3? 2019-06-27 07:41:16 are you on edge or stable? 2019-06-27 07:41:26 danieli: edge 2019-06-27 07:41:48 you could probably make a package update issue or something similar, I don't remember what it's called 2019-06-27 07:41:54 or ask politely in #alpine-devel 2019-06-27 07:44:15 danieli: Thank you. Can I push pull request by editing APKBUILD? 2019-06-27 07:45:44 yes, you can open a PR on it, as long as you test it and make the commit message something along the lines of "main/lua5.3: bump to version x.y.z" 2019-06-27 07:45:57 at least prefix it with "/: " 2019-06-27 07:47:11 danieli: I got you. 2019-06-27 07:49:10 lua5.3 is already in Alpine, even in v3.9 2019-06-27 07:50:33 I'm guessing they're aiming at a patch version or something 2019-06-27 07:52:53 does anybody use bitchx on alpine linux? I installed it via apk -- I do /detach and scr-bx all it does is give a black screen 2019-06-27 07:54:05 I guess I could test it, any specific configuration that triggers it? 2019-06-27 07:54:58 nope 2019-06-27 07:55:00 just default 2019-06-27 08:00:46 Ow.. I use awall, so lua 5.2 is already installed :) 2019-06-27 08:03:15 connection refused for ..... 2019-06-27 08:03:17 is what I get 2019-06-27 08:22:06 danieli: any idea? 2019-06-27 08:22:23 I'm a bit busy with work at the moment, I'll have to look at it a little later 2019-06-27 08:22:30 k 2019-06-27 09:11:08 I've found a way to reliably crash a tab in Firefox on Alpine. Could someone verify for me it happens for you as well? Open gitlab.com, press "sign in" in the top right, then click the "Password" field. It then crashes instantly for me 2019-06-27 09:11:28 ACTION sent a long message: < https://matrix.org/_matrix/media/v1/download/matrix.org/jhmHdKDHOEbguhNJSXhptJXd > 2019-06-27 09:47:00 PureTryOut[m]: Nice to hear that, I felt alone with that bug 2019-06-27 09:47:39 Ah you can reproduce that? I'll make an issue then 2019-06-27 09:48:10 It is also reproducible using https://www.geocaching.com/account/signin and clicking on the password field 2019-06-27 09:49:32 It seems to be the same issue as https://bugs.alpinelinux.org/issues/7454 2019-06-27 09:50:28 I'm starting my netbook to try your url, to see if it is the same issue on your url 2019-06-27 09:51:03 Well the workaround listed there doesn't work for me 2019-06-27 09:51:07 So I don't think it's entirely the same 2019-06-27 09:51:45 Also, that bug was already closed and fixed upstream 2019-06-27 09:54:05 The workaround does not work, but the error message is the same 2019-06-27 09:54:23 I confirm that firefox tab crash using your url 2019-06-27 09:55:20 Seems to be this issue https://bugs.alpinelinux.org/issues/10451 2019-06-27 09:56:22 Yeah right 2019-06-27 10:09:27 Wish me congratulations.. I just did `apk del -f alpine-base` by mistake :) 2019-06-27 10:10:19 M0x32_t[m]11: I did the equivalent of `apk info | sudo apk del' the other day :) 2019-06-27 10:10:25 and I'm a fairly seasoned sysadmin 2019-06-27 10:23:16 danieli: 2019-06-27 10:23:17 haha.. well, I performed a rollback. 2019-06-27 10:23:18 What about you? 2019-06-27 10:23:30 someone else sorted it, it was a container 2019-06-27 10:23:51 danieli: Ah.. I see 2019-06-27 10:36:06 <_ikke_> :-) 2019-06-27 10:36:26 <_ikke_> I installed apk.static and busybox.static 2019-06-27 10:49:43 _ikke_: Oh why? 2019-06-27 10:49:46 I always have busybox static on any system, not just on alpine 2019-06-27 10:52:05 not a bad idea 2019-06-27 10:52:19 M0x32_t[m]11: so he could use it to install the base packages? 2019-06-27 10:52:25 KatolaZ: 2019-06-27 10:52:26 Once when I was using Ubuntu back then, uninstalled NetworkManager somehow, which uninstalled dhcpcd too... 2019-06-27 10:52:27 Busybox saved me from a reboot 2019-06-27 10:52:59 M0x32_t[m]11: dunno about network manager 2019-06-27 10:53:06 danieli: alpine-base and apk-tools are separate I guess 2019-06-27 10:53:14 I know busybox-static has saved my ass a couple of times 2019-06-27 10:53:33 e.g., after accidental removal of libs 2019-06-27 10:53:44 KatolaZ: Busybox has a dhcp client 2019-06-27 10:53:48 M0x32_t[m]11: I had removed *all* packages on the system, both of those included 2019-06-27 10:53:49 Infact, that is what I use anyway 2019-06-27 10:53:55 M0x32_t[m]11: oh, really? :D 2019-06-27 10:54:06 yes, udhcpc 2019-06-27 10:54:13 oh, really? :D 2019-06-27 10:54:15 ;) 2019-06-27 10:54:20 KatolaZ: Yeah... checkout udhcpc -i eth1 2019-06-27 10:54:56 M0x32_t[m]11: I have known busybox for a while, I guess... ;) 2019-06-27 10:55:16 KatolaZ: Most people don't actually need GNU utilities xD 2019-06-27 10:55:36 I am used to FreeBSD base... :) 2019-06-27 10:57:01 M0x32_t[m]11: I started using busybox in early 2000, IIRC... 2019-06-27 10:57:47 KatolaZ: 2019-06-27 10:57:47 I have used it in Android, before that in Ubuntu 2019-06-27 10:58:08 M0x32_t[m]11: those were one-floppy distros 2019-06-27 10:58:40 one-floppy as in one-1.72MB-floppy-with-a-full-system :) 2019-06-27 10:59:04 KatolaZ: Almost :) 2019-06-27 10:59:05 Slackware was fun 2019-06-27 10:59:57 KatolaZ: Imagine GNOME or KDE there xD 2019-06-27 11:00:07 I had busybox even in initramfs on Debian 2019-06-27 11:00:34 mps: busybox actually started as the install env for debian 2019-06-27 11:00:41 in 1995 or 1996 2019-06-27 11:00:44 mps: busybox is part of debian base, so Ubuntu too 2019-06-27 11:02:02 Most distros can actually ship busybox only, most wont even notice if you install bash and busybox unless they break their nonposix scripts 2019-06-27 12:24:53 good day 2019-06-27 12:25:42 I am trying to install deluge from testing, however if I try apk add deluge@testing, it comes back with unsatisfieable deps from community; using apk add -X community deluge@testing seems not to be the right syntax 2019-06-27 12:25:49 Can somebody point me in the right direction? 2019-06-27 12:26:01 mixing stable and edge can break 2019-06-27 12:26:14 it's probably depending on a more recent version of a package in community (newer version in edge than stable) 2019-06-27 12:26:31 or just about any other flavor of dependency hell 2019-06-27 12:29:35 danieli: do you have a recommendation on how to solve it? 2019-06-27 12:29:51 my best suggestion is to upgrade to edge 2019-06-27 13:24:31 hi! i wanna build a bcache-module for mkinitfs. inserting the module is easy, but how do i run a script in the initrd? 2019-06-27 13:25:06 i could edit initramfs-init. but is there a better way? 2019-06-27 13:26:13 romanis: hi, I'd take a look at the zfs module for reference, I'm pretty sure you'll find many similarities there 2019-06-27 13:26:58 anyone know what this is about? https://paste.cointel.pro/?be7f8d4a973689ec#1PcHdOxR9ugpyrBUtKlMaRqH22ugf2o6dcxUVAE7fyM= 2019-06-27 13:27:13 I upgraded from 3.9 to 3.10... apk gives me some errors that im not sure how to fix 2019-06-27 13:28:55 thanks! will look into it 2019-06-27 13:31:17 yeah.... editing initramfs-init is the way to go. 2019-06-27 13:32:19 romanis: are you planinng on packaging it too? 2019-06-27 13:34:54 not yet. but maybe. it shouldn't be too hard. but it would include providing new kernel-parameters for identifying the required bcache-components 2019-06-27 13:36:15 romanis: if you need a hand, let me know. I'm pretty curious about bcachefs but so far I was too lazy to do the whole thing myself 2019-06-27 13:39:13 well ... it depends how much faster my extremely slow HDD (WD-Green something) will be in conjunction with a 40GB SSD cache ;-) 2019-06-27 13:40:07 and ... should i encrypt the bcache device? or add encrypted disks to bcache? hmmm ... what would be faster? 2019-06-27 13:49:57 intuitively dm-crypt over bcache should be faster, but who knows if it holds 2019-06-27 13:54:10 as in, the data needs to be encrypted only once 2019-06-27 14:54:09 ageis: more details ? 2019-06-27 15:13:50 hello, I'm trying to install alpine without wiping out my files, when I run setup-disks it says that I "might have to fix the MBR" but what does that mean? 2019-06-27 15:17:23 I'm trying to use syslinux by the way 2019-06-27 15:17:34 Ir2: That means your mbr partition is corrupted. I advice you to first copy your data to a safe place before doing a fresh install. 2019-06-27 15:18:29 I'm doing everything in a virtual machine 2019-06-27 15:18:37 but how can it be corrupted, that makes no sense 2019-06-27 15:19:26 Ir2: Then probably booting with mbr but trying to do an efi install 2019-06-27 15:20:03 I'm not doing anything efi-specific 2019-06-27 15:21:36 literally all I'm doing is mounting the root partition to /mnt and then running setup-disks /mnt 2019-06-27 15:22:09 Ir2: Then it is a possible bug. 2019-06-27 15:23:29 Are you doing manual install? or via setup-alpine? 2019-06-27 15:24:14 I'm running setup-alpine but only to the point where it runs setup-disks 2019-06-27 15:24:28 I'm running setup-disks separately 2019-06-27 15:24:40 on /mnt, after I mount the partitions there 2019-06-27 15:25:12 well, _the_ partition, I've only tried it with one partition so far 2019-06-27 15:25:34 Ir2: No idea. I did setup-alpine and install was successful with syslvm. Also it installed grub, not syslinux 2019-06-27 15:26:03 I'm pretty sure I could get this to boot easily if I installed grub manually, but I want to try syslinux 2019-06-27 15:33:49 also I don't think it's a bug, there's just no documentation (and what is in the wiki is outdated) for this kind of install 2019-06-27 15:45:46 Ir2, if I remember good that MBR thing isn't critical error and just have to manualy write it to partition later 2019-06-27 15:46:34 Ir2, for example here look at the "note" https://wiki.alpinelinux.org/wiki/LVM_on_LUKS#Installing_Alpine_Linux 2019-06-27 15:46:58 and after got how to make it right 2019-06-27 15:47:57 s/partition/disk sector 2019-06-27 15:48:10 what "note"? 2019-06-27 15:48:40 "Note: The automatic writing of the master boot record (MBR) fails in this step. You will write the MBR later manually to the disk." 2019-06-27 15:49:50 where is that? 2019-06-27 15:50:03 oh wait 2019-06-27 15:50:11 I thought I had clicked on your link but it was something else 2019-06-27 15:51:18 I mean yeah I figured that much, that is very much what the warning says when you're installing 2019-06-27 15:51:28 what I want to know is _how_ to fix it 2019-06-27 15:53:07 Ir2, read under: Installing a bootloader --> Syslinux 2019-06-27 15:53:25 I have already tried that 2019-06-27 15:53:34 didn't work 2019-06-27 15:53:43 you have single partition so can skip creating "boot" stuff 2019-06-27 15:54:44 so set up extlinux stuff and after "dd bs=440 count=1 conv=notrunc if=/mnt/usr/share/syslinux/mbr.bin of=/dev/vda" 2019-06-27 15:55:37 like I said I have already tried it 2019-06-27 15:55:45 but I just get a different error message when booting 2019-06-27 15:56:28 for know I'll try to install grub and then try to setup up syslinux once I have a booting system 2019-06-27 15:56:48 since I get some error when running extlinux --install /boot 2019-06-27 15:57:34 though I have tried the update-extlinux script 2019-06-27 15:57:49 dunno what are your steps but I never had problem with MBR and syslinux or grub 2019-06-27 15:57:55 s/have/haven't/ 2019-06-27 15:58:43 well, that is important step and you shouldn't skip it 2019-06-27 15:59:16 so probably with your MBR is all fine but not with boot loader 2019-06-27 16:00:29 with grub is more fun because have to be sure to mount stuff like /proc /dev /sys 2019-06-27 16:10:59 I can't even do that because busybox's mount doesn't have "--types proc" 2019-06-27 16:11:11 anyone know how to get mouse working in alpine in virtualbox vm? whats the kernel module? 2019-06-27 16:11:50 also grub-install is not working, nor is ext4linux, it's giving some arror like "no path for /boot" or something, I'll give the exact message in a minute 2019-06-27 16:13:18 extlinux: cannot find device for path /boot 2019-06-27 16:13:29 Are you guys facing this issue? Installing clang alone doesn't compile code, requires gcc installed alongside in Alpine linux. 2019-06-27 16:13:31 http://0x0.st/z2k6.txt 2019-06-27 16:13:31 extlinux: cannot open device (null) 2019-06-27 16:14:18 which is the exact same error I got when I tried doing it without the update-extlinux script 2019-06-27 16:44:56 Hello, anyone any idea why upgrading from Alpine 3.8 to alpine 3.9 fails with the following error: 2019-06-27 16:45:02 ERROR: unsatisfiable constraints: 2019-06-27 16:45:08 Huh? Error reporter did not find the broken constraints. 2019-06-27 16:45:16 The only active apk repo's are 3.9 main & community 2019-06-27 16:45:38 Guest88410, how you doing the upgrade, full command? 2019-06-27 16:45:48 apk upgrade -Ua 2019-06-27 16:49:29 <_ikke_> Hmm, i've seen earlier reports for that 2019-06-27 16:49:34 <_ikke_> I don't recall any solution though 2019-06-27 16:52:06 I've found some old github topics talked about: "Looks like apk can't handle the case when two packages provide the same so: file." 2019-06-27 16:54:26 Guest88410, and which packages got under that ERROR: unsatisfiable constraints: ? 2019-06-27 16:54:45 @MY_R none :( 2019-06-27 16:54:54 That's the complete error 2019-06-27 16:54:59 :\ 2019-06-27 16:55:32 My apk/world contains the following packages: acct 2019-06-27 16:55:32 alpine-base 2019-06-27 16:55:32 apk-tools 2019-06-27 16:55:32 ca-certificates 2019-06-27 16:55:32 htop 2019-06-27 16:55:33 linux-virt 2019-06-27 16:55:35 nginx 2019-06-27 16:55:37 nginx-mod-stream 2019-06-27 16:55:39 syslinux 2019-06-27 16:55:41 tinyssh 2019-06-27 16:55:43 ucspi-tcp@testing 2019-06-27 16:55:45 vim 2019-06-27 16:55:45 daaaamn not like that! :D 2019-06-27 16:55:47 zabbix-agent 2019-06-27 16:56:05 Sorry, copy paste :/ 2019-06-27 16:56:22 use a paste service 2019-06-27 16:57:01 @maxice8 will do 2019-06-27 16:57:11 ucspi-tcp@testing 2019-06-27 16:58:41 maybe above @testing causes the issue 2019-06-27 16:58:44 <_ikke_> Guest88410: How does your /etc/apk/repositories look like? 2019-06-27 17:00:05 _ikke_, 06.27 [18:45] The only active apk repo's are 3.9 main & community 2019-06-27 17:00:21 <_ikke_> and a tagged testing repo apparently :) 2019-06-27 17:00:25 yep 2019-06-27 17:00:37 <_ikke_> So yes, that can cause issue 2019-06-27 17:00:39 <_ikke_> issues 2019-06-27 17:00:56 <_ikke_> mixing different versions (ie, stable and unstable) is not supported 2019-06-27 17:01:39 And testing indeed, I'll remove testing 2019-06-27 17:02:09 <_ikke_> but apk will complain when you remove it but still have the packages in /etc/apk/world 2019-06-27 17:04:51 That fixed the problem :) 2019-06-27 17:05:50 @_ikke_ yes but the error was the same, it used to show the package 2019-06-27 17:06:18 <_ikke_> Guest88410: yes, there is probably some edge condition that causes this 2019-06-27 17:08:13 Indeed @_ikke_ & MY_R thnx! 2019-06-27 17:08:26 :) 2019-06-27 17:10:41 I have that with pciutils. It says Error, usr/share/hwdata/pci.ids owned by hwids-pci-20190316-r1 2019-06-27 17:10:54 So fix does not work 2019-06-27 17:12:58 <_ikke_> Yes, I suspect pciutils started to provide that same file 2019-06-27 17:13:04 <_ikke_> and it started to conflict 2019-06-27 17:13:51 @_ikke: So should I remove it then first, and then start to reinstall the newer package? 2019-06-27 17:15:34 <_ikke_> pciutils depends on hwids-pci, so it's going to conflict by definition 2019-06-27 17:18:29 @_ikke_: Indeed, removed hwids-pci pciutils-libs and pciutils and it was all right. But trying to reinstall it gives the same error message 2019-06-27 17:25:45 wozencroft: that is bug which must be fixed soon, I think hwids should be moved to main and hwdata pkg removed 2019-06-27 17:26:49 ncopa: ^ 2019-06-27 17:26:55 what you think 2019-06-27 17:26:57 mps: Ok, good, thank you. Atm just ignore that message then. 2019-06-27 17:29:20 we discussed this problem last night on #alpine-devel, you can look at IRC log 2019-06-27 17:29:44 Ah, ok, missed that discussion 2019-06-27 17:30:32 there is a log on https://dev.alpinelinux.org somewhere in some subdir 2019-06-27 17:31:17 https://dev.alpinelinux.org/irclogs/ 2019-06-28 03:07:45 mps: will that problem be solved in alpine 3.10's release version? 2019-06-28 04:27:38 <_ikke_> yunfan: regarding pciutils? 2019-06-28 04:58:15 _ikke_: nope, the alwinner board's kernel wont reponsed correctly for SIGNAL like reboot poweroff 2019-06-28 05:11:11 <_ikke_> ah, ok 2019-06-28 05:54:50 _ikke_: so it has been solved already? 2019-06-28 06:29:54 yunfan: maybe it will be fixed in some of the next minor alpine releases if someone found what need to be fixed, or maybe I will try but after I return from vacation/s 2019-06-28 06:30:24 mps: aha, sorry for that, enjoy your vocation 2019-06-28 06:31:35 np, but I'm still at work. would be nice if I'm already in some nicer place 2019-06-28 06:32:22 but I don't have time to test mentioned issue because I'm busy with some other things 2019-06-28 06:36:04 ok when you have time 2019-06-28 06:59:08 yunfan: yes, I hope I will find some time for fixes first. ARM needs a lot of work for distributors of ready made systems because there are so much different SBC's and SOC's. All these needs to be tested in different configurations and use case which takes time 2019-06-28 07:00:48 and we are mostly volunteers and not full time paid developers/testers 2019-06-28 07:00:53 mps: thanks for the hardwork, i wish i could contribute some of my time for the job 2019-06-28 07:01:51 you already helped (and you can more :) I think) by testing and reporting issue 2019-06-28 07:04:44 in next few months I will see if it is worth to make separate kernel multiplatform pkg specially for ARM which would 'cover' more features and drivers 2019-06-28 07:06:21 and with the help of the good people from pmOS we already working on improving some parts of ARM, i.e. gpu drivers 2019-06-28 07:06:32 hope we will do a lot more 2019-06-28 07:12:16 mps: by pmOS you mean postmarketOS? 2019-06-28 07:12:50 i had check that, very excited, except that those devices were not accesiable for me in china 2019-06-28 07:13:17 so i just bought a nokia 8810 4G to join the KaiOS hackers 2019-06-28 07:21:18 also I don't have any usable device on which I can install pmOS (yes, I mean postmarketOS) expect one somewhat broken HTC Bravo phone 2019-06-28 07:21:35 but, it can be tested under qemu 2019-06-28 07:24:29 i used to have a htc g7 2019-06-28 07:24:44 but i cant find it any more, it has been past almost a decade 2019-06-28 09:24:51 i don't have any devices supporting pmOS either, and i'm not planning to acquire any 2019-06-28 09:27:16 <_ikke_> I stil have a nexus 5 lying around, might try it on that one 2019-06-28 11:21:10 I can't figure out why this is happening, got 2 machine running alpine on edge branch and when I apk update && apk upgrade, one machine always have 2 more distinct package available... ? Why does it does that, knowing that they hit the same mirror on the same version 2019-06-28 11:24:25 KH405: are they the same architecture_ 2019-06-28 11:24:25 ? 2019-06-28 11:24:33 some packages are only built for certain architectures 2019-06-28 11:25:38 Yes 2019-06-28 11:25:52 Both x64 2019-06-28 11:30:13 hi... i'm trying to get 3.10.0 raspberry pi running on the new 4b model by copying files from the official noobs SD onto the released Alpine 3.10.0 raspberry pi image. is this possible or am i wasting my time? 2019-06-28 11:40:51 rpi4: i don't know 2019-06-28 11:41:04 none of the developers have raspberry pi 4s as far as i know, and i haven't heard of any testing being done on them 2019-06-28 11:42:52 okay - happy to test things if it helps, although i'm sure the devs have other things to do 2019-06-28 11:43:36 i'll try debugging it more with a serial cable and see if i can see how far it is getting into the boot 2019-06-28 11:46:34 there are probably a lot of changes necessary to make it boot and run on the rpi4 2019-06-28 11:50:14 i was hoping it was a case of replacing the bootcode and dropping in the latest firmware / modding some config files, but my prodding in this direction is being so fruitful. i'm sure you are right and there is more to it than that. thanks for the heads-up 2019-06-28 11:50:48 Alpine <3 Alpine 2019-06-28 11:50:58 cheers 2019-06-28 11:51:00 the previous rpis required a modified kernel 2019-06-28 11:51:05 so i assume something similar is still the case 2019-06-28 11:51:23 what we did was diff the rpi kernel against the mainline kernel and apply that, if i remember correctly 2019-06-28 11:57:22 ah sounds like a job for someone more familiar with the process 2019-06-28 11:57:40 happy to sponsor a pi4 though if it helps 2019-06-28 11:58:07 i'm half tempted to just get one and fix it 2019-06-28 11:58:41 it is a nice board i have to say 2019-06-28 12:00:32 I do a lot arm development and use it in different fields, but dislike RPi's 2019-06-28 12:02:36 they're nice for certain hobbyist projects, that's pretty much it 2019-06-28 12:04:12 ah i hear you. this is my first rpi since it's the first one i've considered a fairly reasonable value proposition and it has a large community. 2019-06-28 12:04:26 yes, I just told my personal feeling about them, not want to say are they good or bad 2019-06-28 12:04:38 nod 2019-06-28 12:05:36 afaik, their bootloader is closed/proprietary 2019-06-28 12:07:16 i got a jetson nano recently - dunno if you've played with one but much better value for money even though it is running A57. it surprisingly kicks the Pi 4's ass on performance. probably because NVIDIA didn't skimp on the L2 cache 2019-06-28 12:07:23 Anybody has any idea with my weird problem ? 2019-06-28 12:07:47 oh is it - i thought they were all about open stuff. hmm - doesn't surprise me though 2019-06-28 12:07:59 KH405: is it actually a problem, or is it a curiosity? 2019-06-28 12:08:10 rpi4: yeah, no, that's not really true 2019-06-28 12:08:16 it seems like it but it is not the case 2019-06-28 12:09:13 Curiosity ^^' 2019-06-28 12:09:25 Well could that point out to something wrong ? 2019-06-28 12:10:09 How many packages am I suppose to have available? 2019-06-28 12:10:39 well, perhaps you have two more packages installed on one system than the other? 2019-06-28 12:11:30 It's the distinct packages part 2019-06-28 12:11:44 It's the number of total package available no? 2019-06-28 12:11:58 they aren't available for install if you already have them installed though? 2019-06-28 12:12:35 This wouldn't make sense cause I got 500 package on a system and like 300 on the other 2019-06-28 12:13:05 hmm, i just ran a docker container and did apk update, it says 10342 available. upgrade, then 10333 available 2019-06-28 12:13:33 perhaps it counts packages that were already installed if they aren't the latest version, or something along those lines? 2019-06-28 12:13:55 I got 15010 and 15012 both edge on x64 2019-06-28 12:15:22 Maybe, but I didn't hold back any package though 2019-06-28 12:15:22 i'm not super familiar with the apk code, and it's all undocumented - it's apparently `db->available.packages.num_items' 2019-06-28 12:43:02 KH405, edge with main, community, testing repo and after full upgrade it showing: OK: 15016 distinct packages available / 487 packages installed 2019-06-28 12:44:15 and every not updated package will increase "distinct packages available" 2019-06-28 12:45:55 How can I tell if I got packages that do not update ? 2019-06-28 12:46:56 I mean "outdated" one so just do update and upgrade to be sure system got last available packages 2019-06-28 13:00:15 KH405: the first thing could be comparing $ apk info output of the 2 machines to see which are those 2 packages 2019-06-28 13:00:46 KH405: 2 packages or 200 packages in the difference ? might be linux-firmware-none ? 2019-06-28 13:26:01 Hi there, I've just seen the new release 3.10.0. Why removing MongoDB? And also, how to install it now ? 2019-06-28 13:30:50 tmhoang1: 2 package difference, but i'm just wondering which are they and why it's like that 2019-06-28 13:30:59 AlexisF: you will have to compile it yourself 2019-06-28 13:31:09 we removed it because its new license is considered proprietary 2019-06-28 13:31:25 apk info just prints what is installed, I don't know how this is going to help me 2019-06-28 13:31:25 you might have heard about the SSPL - most of the other larger distros removed mongodb for the same reason 2019-06-28 13:39:32 Anybody has a guide on how to clone my hard drive to my USB and make my USB a ramdisk? 2019-06-28 13:45:34 KH405: apk list -a | sort > machine_a 2019-06-28 13:45:38 same on machine_b 2019-06-28 13:45:40 then diff them 2019-06-28 13:47:36 AlexisF: check out https://git.alpinelinux.org/aports/tree/non-free/mongodb for basic build stuff and patches that might be needed 2019-06-28 13:48:46 that APKBUILD should work and create apk for you that you can then install with apk add 2019-06-28 14:30:24 any idea why mpv does not resolve https urls? and vlc has similar problems, vlc says this: > main tls client error: TLS client plugin not available and looking at all the packages for vlc i don't really know what i should install to make it work. although i would prefer to use mpv if that is possible 2019-06-28 14:32:07 you do have openssl installed, right? 2019-06-28 14:32:15 perhaps it was compiled without support for it, check the APKBUILD for details 2019-06-28 14:41:33 openssl is available yes 2019-06-28 14:42:02 previously mpv supported https urls 2019-06-28 14:42:14 i guess via youtube-dl as a fallback 2019-06-28 14:44:42 `git log --patch APKBUILD` doesnt give any useful hing 2019-06-28 14:45:11 s/hing/hint 2019-06-28 14:53:11 p4Wv1qn095FW, got same error with that link? https://www2.cs.uic.edu/~i101/SoundFiles/PinkPanther60.wav 2019-06-28 15:15:53 huh. no. so it seems to be a problem with youtube-dl? 2019-06-28 15:17:32 nb this is a newly setup system, on the old system it worked well. 2019-06-28 15:27:21 p4Wv1qn095FW, dunno, try install package "ffmpeg", I think youtube-dl using it 2019-06-28 15:27:45 it was installed 2019-06-28 15:31:37 my mpv working fine on edge 2019-06-28 15:42:37 can it directly play youtube links? 2019-06-28 15:43:30 i get Playing: https://youtu.be/\nFailed to recognize file format. 2019-06-28 15:43:40 xiting... (Errors when loading file) 2019-06-28 15:44:44 working fine not only with youtube 2019-06-28 15:46:35 p4Wv1qn095FW, do "ldd /usr/bin/mpv" maybe somehow you missing something 2019-06-28 15:47:15 nah. nothing missing 2019-06-28 15:47:26 alpine 3.10? 2019-06-28 15:47:56 edge 2019-06-28 15:48:07 x86_64? 2019-06-28 16:10:29 How do I make iwd give up an interface's control? I suppose I can just stop iwd. 2019-06-28 16:12:08 MY_R: yes. x86_64 2019-06-28 16:13:19 p4Wv1qn095FW, so you done something wrong because it should just work :\ 2019-06-28 16:24:36 i suspect i forgot to do something. not doing something wrong. since i'm in the process of setting this up. 2019-06-28 17:08:44 Is there a way to live patch alpine's kernel? 2019-06-28 17:09:46 i'm not aware of any way to completely replace it with a new version. of course you can load modules to patch functionality in/out or exit the kernel in memory by hand... 2019-06-28 17:11:25 yes, to my knowledge we don't support any live kernel patching mechanism like kgraft/ksplice/kexec etc 2019-06-28 17:12:36 Is it planned for the future? It's kind of a bummer for a server to not be able to livepatch 2019-06-28 17:14:37 <_ikke_> You'd need something like ksplice, not sure how feasible it would be to bring that to Alpine 2019-06-28 17:15:02 <_ikke_> or kexec 2019-06-28 17:15:50 Or getting a second host and going down another rabbit hole... lol 2019-06-28 17:20:42 kexec is not patching 2019-06-28 17:20:48 it's just fast reboot afaik 2019-06-28 17:21:06 <_ikke_> yes, I just read that indeed 2019-06-28 17:21:12 (or chained boot is probably the more meaningful usage case -- use linux as a bootloader to load the real linux) 2019-06-28 17:22:47 So you guys reboot after each kernel update? 2019-06-28 17:22:52 There's no way around it ? 2019-06-28 17:22:58 <_ikke_> KH405: yes 2019-06-28 17:24:32 <_ikke_> KH405: The problem with high uptime servers is that you are less and less certain that things will work again after an unexpected reboot 2019-06-28 17:24:40 Skipping bios / post delays seems like a good enough reason for kexec to me. Some server boards take a long time on all that stuff 2019-06-28 17:25:14 <_ikke_> Problem is booting the kernel where the hardware is un an undefined state 2019-06-28 17:25:33 <_ikke_> "The new kernel will usually expect all hardware devices to be in a well defined state, in which they are after a system reboot because the system firmware resets them to a "sane" state. Bypassing a real reboot may leave devices in an unknown state, and the new kernel will have to recover from that." 2019-06-28 17:25:46 That problem happens on most boots ;-) 2019-06-28 17:27:25 :-) 2019-06-28 17:27:51 problem is easily solved if all your hardware is on a bus that's physically and logically hot-pluggable 2019-06-28 17:27:57 you just logical-unplug everything 2019-06-28 17:28:14 e.g. reset the usb host controller 2019-06-28 17:28:29 <_ikke_> Does such a system exist? 2019-06-28 17:28:46 Parts of such a system do 2019-06-28 17:28:53 it's easy to have a system where all the hardware that you actually want to use is of that form 2019-06-28 17:28:58 just put everything on usb 2019-06-28 17:30:22 So how will I go if I want my server to be more "Launch and forget" ? 2019-06-28 17:30:41 that generally isn't a good idea 2019-06-28 17:30:53 KH405: managed services? 2019-06-28 17:30:54 unless you have some orchestration system that manages it, anyway 2019-06-28 17:31:46 Well, since we don't install modules into a seperate dir you kind of need to reboot after an update anyway because you won't be able to load modules 2019-06-28 17:31:47 Which makes sense from a "we mustn't take up too much space" perspective, but is sort of painful for everything else 2019-06-28 17:32:38 Cogitri: you couldn't prepopulate before kexec? 2019-06-28 17:32:42 "launch and forget" can be viable if you have no attack surface running 2019-06-28 17:33:13 How can you have no attack surface running ? 2019-06-28 17:33:31 "launch and forget" typically ends with a botnet / Mirai type incident 2019-06-28 17:33:46 it's very rare that there are remotely exploitable kernel vulns (altho one did happen recently, which would have been mitigated if you manually turned off all the wacky and useless tcp extensions) 2019-06-28 17:34:11 tcely, not because of kernel. because of crappy services listening 2019-06-28 17:34:15 tcely: I guess I _could_ make sure that I have all modules loaded, but it's a bit annoying when I upgrade the kernel of my laptop and then my usb soundcard won't work afterwards 2019-06-28 17:34:57 You don't need to load them, just put in place iiuc 2019-06-28 17:38:42 cogitri, i still want a way to force alpine not to remove modules when updating kernel 2019-06-28 17:39:03 it should orphan them and schedule them for deletion after reboot or something 2019-06-28 17:39:12 +1 2019-06-28 17:39:37 or if that's too hard just treat them like config files that require manual deletion or purge 2019-06-28 17:39:58 Sounds good 2019-06-28 17:42:37 I'm wondering If I should change distro, knowing that I only have 1 hosts and don't want to reboot like once, twice a week ... 2019-06-28 17:43:13 <_ikke_> KH405: the kernel is not updated that often 2019-06-28 17:43:26 <_ikke_> especially if you follow a stable version 2019-06-28 17:43:51 I'm on edge.. 2019-06-28 17:44:38 <_ikke_> even on edge it's not updated so often 2019-06-28 17:44:53 <_ikke_> alpine follows the LTS versions of the kernel 2019-06-28 17:45:08 generally you don't need to update kernel unless you need new functionality or there's a cve 2019-06-28 17:45:20 even with edge you could pin kernel to stable if you want 2019-06-28 17:45:34 to avoid automatic updates of it except when there's a security fix for stable 2019-06-28 18:18:50 I will definitely keep an eye out for future development, I really liked alpine but this is a bummer if you don't want to do a second host. I don't wanna put up with Cannonical's Shennanigans, but manually rebooting at random times seems like a PITA. Anyway thanks for the help everyone! 2019-06-28 18:21:09 I don't have any production hosts without redundancy, and when I need to reboot stuff, I do it in maintenance windows 2019-06-28 18:21:22 on my own workstations i don't mind rebooting once in a while if i need a new kernel 2019-06-28 19:11:27 Hey, has anyone got Minecraft working on alpine? 2019-06-28 19:22:36 Should be do-able considering it's Java, I guess 2019-06-28 19:24:41 https://hub.docker.com/r/itzg/minecraft-server/dockerfile 2019-06-28 19:24:53 Someone has 2019-06-28 19:28:13 06.28 [21:19]* Yorick1 has quit (Quit: WeeChat 2.5) 2019-06-28 19:28:20 :P 2019-06-28 19:30:44 i could test it 2019-06-28 19:30:47 can't imagine it doesn't work, it is java 2019-06-28 19:31:05 tcely: i'm *guessing* they meant the client / game, not the server 2019-06-28 19:32:13 danieli: could be you're correct 2019-06-28 19:37:16 Someone did that too. https://hub.docker.com/r/stanleyrya/alpine-minecraft-client/dockerfile 2019-06-28 22:18:33 hey I am using alpine mkimg profiles and genapkovl to generate custom ISO's... I need a package included that isn't in the repo's, but is on GitHub... How would I go about installing this into my custom ISO? https://github.com/sgerrand/alpine-pkg-glibc 2019-06-28 22:20:48 you could precompile it into an apk and include the builder's signing key in /etc/apk/keys/ 2019-06-28 22:23:34 ok cool I'll try that out 2019-06-28 22:27:03 so I can package the already compiled apk's?: apk --no-cache -X https://apkproxy.herokuapp.com/sgerrand/alpine-pkg-glibc add glibc glibc-bin 2019-06-28 22:35:01 yes, you can include it in the installation ISO 2019-06-28 22:36:08 i'm not so sure how applications from our repos will behave then though, if you are installing glibc 2019-06-28 22:38:30 So do you guys think by compiling things like https://github.com/dynup/kpatch it could work? Or there are other issue, like if Livepatch in enabled in the kernel that could make it not work? 2019-06-28 22:42:15 KH405: try it 2019-06-28 22:42:25 that's the only way of knowing for certain 2019-06-28 22:43:09 Will do, but i'm not the most knowledgeable in deep linux stuff 2019-06-28 22:45:22 thing is, unless someone who is needs it or gets paid to fix it, they probably won't 2019-06-28 23:04:07 I have the key being stored in the ISO at /etc/apk/keys/sgerrand.rsa.pub, just not sure where to include the APK though... The Aports build tools have a lot of APK's in source form, but not sure where include the actual apk or when to call apk add 2019-06-28 23:11:27 either in the genapkovl or in the bootstrap 2019-06-28 23:12:49 ahhh I can include the repo (https://apkproxy.herokuapp.com/sgerrand/alpine-pkg-glibc) in the mkimage command line with --extra-repository 2019-06-28 23:18:44 yes, but it will not let you install packages on the target device without internet access 2019-06-28 23:22:45 I got it installed to the ISO directly so it's available on boot 2019-06-29 01:51:09 I am a bit confused about the nonfree repo, mongo-c-driver seems to be there but it is Apache2 licensed 2019-06-29 01:51:36 I have also created a pkg that builds the cxx-driver, would that also be classed as non-free? 2019-06-29 01:54:24 russkel: you are right, it is indeed Apache-2.0 licensed 2019-06-29 01:54:36 I assume it was removed in error when mongodb was removed 2019-06-29 01:54:51 danieli: I can raise a PR that updates it to latest version and back into testing 2019-06-29 01:54:59 russkel: sounds good to me :) 2019-06-29 01:55:13 all ma sources are at work so it will have to wait till monday 2019-06-29 01:55:29 make commit message "non-free/mongodb: upgrade blah, move to testing repo" or something along those lines 2019-06-29 01:55:31 at least that prefix 2019-06-29 01:55:34 er 2019-06-29 01:55:42 non-free/mongo-c-driver 2019-06-29 01:56:00 freudian slip 2019-06-29 01:56:05 will do! 2019-06-29 01:56:15 thank you for noticing and taking the time to ask 2019-06-29 01:56:49 oh my pleasure, I am using it for a project so it benefits all if it's mainlined 2019-06-29 01:57:07 upstreamed* 2019-06-29 01:58:53 there was also another issue I had with compiling the cxx-driver, abuild packaging didn't like a soversion that had text in it 2019-06-29 01:59:26 https://github.com/mongodb/mongo-cxx-driver/pull/645 2019-06-29 02:01:39 i'd suggest asking that in #alpine-devel 2019-06-29 02:01:55 keep in mind that it's super early saturday morning here in central europe 2019-06-29 02:02:51 will do, I will get a proper error message too 2019-06-29 04:35:55 Just got back home, I saw the feature was requested a few years ago, dunno if there's an answer tough! https://bugs.alpinelinux.org/issues/5279 2019-06-29 04:36:34 KH405: no reply 2019-06-29 04:37:06 you asked at 12 am CEST and it is now 6 am CEST 2019-06-29 04:37:16 + it's a saturday 2019-06-29 04:38:18 Yeah, sorry I didin't mean to sound this way, I just wanted to share my findings also this is somehwere EU, but it's a way more reasonable hour for 'Mercia :P 2019-06-29 04:38:47 it's 12:37 AM here 2019-06-29 04:38:50 no worries, just thought i'd point it out as a potential reason why you had no reply yet 2019-06-29 04:38:51 12:38 2019-06-29 04:38:58 AppXprt: most of the alpine devs are in europe 2019-06-29 04:39:10 ahhh gotcha 2019-06-29 04:39:21 Didn't knew that :P 2019-06-29 04:40:14 Altough i'm really surprised that this has sparse documentation as alpine is often refered as a 'firewall/Server OS' 2019-06-29 04:42:21 i wouldn't say all of the buzzwords in the slogan thingy are spot-on accurate 2019-06-29 04:43:22 considering it's a smaller project with few core devs, i guess you could consider it an os for superusers who aren't afraid to get their hands dirty 2019-06-29 04:48:35 me and natanael in NO, carlo and kevin in NL, francesco in IT, kaarle in FI, sören probably in DE 2019-06-29 05:21:31 I find alpine pretty handy, especially with the useful aports customization scripts and such 2019-06-29 05:21:53 I'm using as a server os but not one that runs all the time although I think it would be alright 2019-06-29 07:29:52 russkel: danieli: if you move package it is better to use destination repo in commit msg, i.e. 'testing/mongo-c-driver: move from non-free' 2019-06-29 13:25:53 AppXprt: IMO, the basic goal of a server is to have the most uptime while staying secure ... 2019-06-29 13:28:08 <_ikke_> single server uptime becomes less and less relevant nowadays 2019-06-29 13:30:38 True, in entreprise settings... I don't know alot of people that run a blade server or a cluster in their home :P 2019-06-29 13:31:19 <_ikke_> But in their home system, a small maintenance window to apply updates should not be a problem either 2019-06-29 13:31:33 <_ikke_> you don't need x 9's of uptime in a home setting 2019-06-29 13:31:47 x9 ? 2019-06-29 13:31:55 <_ikke_> x amount of nines 2019-06-29 13:32:35 I mean, I don't NEEED it, but it can be reduced... 2019-06-29 13:33:13 <_ikke_> My personal vps has generally a good uptime, but I just rebooted it to apply a new kernel 2019-06-29 13:33:34 <_ikke_> Less then 2 minutes downtime 2019-06-29 13:33:38 I mean, Of course having Internet redundancy in a home setting might be overkill for what you're going to pay 2019-06-29 13:34:06 <_ikke_> I would not bother with live kernel patching in a home setting 2019-06-29 13:34:21 But having live patching isn't costly af. And it's doable on other Distro 2019-06-29 13:34:42 That's where we disagree :P 2019-06-29 13:34:44 Same, I just reboot every week or so 2019-06-29 13:38:33 I still think this is uneeded ... 2019-06-29 13:39:21 Well, if someone steps up to implement it it's doable on Alpine too :P 2019-06-29 13:39:49 Yeah, that was exactly my point ^^' 2019-06-29 13:40:07 <_ikke_> focus on "if someone steps up" 2019-06-29 13:40:08 Like I won't pay 100$ US a month to have internet fallback 2019-06-29 13:41:15 Again, I totally agree, but shouldn't the focus be high in general as, as I said above, Alpine is refered as a Firewall/Server distro? 2019-06-29 13:41:30 <_ikke_> I don't think those patches come for free (ie, involves work to provide them) 2019-06-29 13:42:21 <_ikke_> KH405: Even actual routers require a reload to get new firmware 2019-06-29 13:42:41 KH405: I'm using Alpine Linux as a desktop distro actually. It's general purpose, it fits most use-cases 2019-06-29 13:43:09 : Same and I love it! 2019-06-29 13:43:59 <_ikke_>: Not really the point I'm trying to make. I just think that downtime should be minimised if they aren't NEEDED 2019-06-29 13:44:15 Which rebooting could not be needed 2019-06-29 13:44:34 KH405: I don't think anyone wants to maximize downtime :k 2019-06-29 13:45:00 It's just that it appears that no one thought live patching was so necessary that that person did something about it 2019-06-29 13:45:02 lol, i'll try to compile kpatch later this week and i'll report 2019-06-29 13:45:12 👍 2019-06-29 13:45:45 30 seconds downtime 2-3 times per month, I would worry more about stability of internet and electricity :D 2019-06-29 13:46:04 Rebooting my server is at least 5 minutes :P 2019-06-29 13:46:24 <_ikke_> Personally I think that delaying reboots might even lead to more issues later 2019-06-29 13:46:27 some TOR node or what? ... 2019-06-29 13:47:12 Nope, but tbf I never understood why it took that long, so that might be more a ''me'' problem :P 2019-06-29 13:47:14 ye I knew people who always avoid reboots to have some high uptime... after 6 months they always were afraid to do it because "but if wont boot!?" :D 2019-06-29 13:47:30 But RAM initialization takes easily 4-5 minutes by itseld 2019-06-29 13:50:36 KH405, so now everyone understand you why you ask about kernel livepatch, you should start from that if didn't before or I missed something :) 2019-06-29 13:51:41 I mean, this is tottaly off-topic, but I have no Idea how to speed up RAM init ^^' 2019-06-29 13:52:48 My laptop reboots in max 45 seconds, which I mean, i'm totally fine, it's not running any services and downtime on a laptop is not a problem ... But a server is all the contrary 2019-06-29 13:53:11 Get another mainboard that doesn't take as long 2019-06-29 13:53:22 But most server boards take ages to POST 2019-06-29 13:53:31 or disable some weird checks, ECC ram or what? :P 2019-06-29 13:53:35 <_ikke_> Or install a hypervisor 2019-06-29 13:53:42 ^ that! 2019-06-29 13:53:43 <_ikke_> so that you just have to reboot the guest 2019-06-29 13:53:44 hypervisor? 2019-06-29 13:53:53 What is it gonna change ? 2019-06-29 13:53:58 A VM 2019-06-29 13:54:02 no more ram init? 2019-06-29 13:54:16 That the VM can reboot within seconds without you having to reboot the actual machine 2019-06-29 13:54:32 I'm using Alpine as my Hypervisor lol 2019-06-29 13:54:44 Well, QEMU and KVM on top 2019-06-29 13:55:52 : I'll definetly check this out on my next reboot :P 2019-06-29 13:56:20 But my point is: Hypervisor is going to be running on top of something ... It just changes the issue of place, the issue is still the same ... 2019-06-29 13:56:40 in your case you could use something what support livepatch to avoid those 5 minutes :D 2019-06-29 13:57:34 or just XEN, dunno, offff topic :) 2019-06-29 13:58:41 Alpine supports Xen and no livepatch :( 2019-06-29 13:59:01 :< 2019-06-29 13:59:22 Again, I don't think the hypervisor is the problem ... 2019-06-29 14:03:08 <_ikke_> "WARNING: Use with caution! Kernel crashes, spontaneous reboots, and data loss may occur!" 2019-06-29 14:03:12 <_ikke_> (kpatch) 2019-06-29 14:05:37 <_ikke_> "Unfortunately, live patching isn't always as easy as the previous example, and can have some major pitfalls if you're not careful." 2019-06-29 14:05:55 <_ikke_> KH405: live kernel patching is not as trivial as you make it out to be 2019-06-29 14:06:57 Yeah, I don't doubt that 2019-06-29 14:07:30 or maybe need find some bios setting to not perform ram checks on warm boot or maybe alpine doing "cold reboot" 2019-06-29 15:03:42 You are not alone :) 2019-06-29 15:38:12 So you shut down 20 VM everytime there's a kernel update? 2019-06-29 15:39:30 <_ikke_> We reboot a lot more servers to update them 2019-06-29 15:39:51 i don't ever have production servers without redundancy if i have the option 2019-06-29 15:51:33 KH405, but nobody telling you to update the kernel EVERYTIME :D can do it only if there is some really bad bug or security risk 2019-06-29 15:53:36 yup, you wouldn't upgrade just for the sake of upgrading, you'd do it if you were at risk of having some security flaw exploited or you really need new features 2019-06-29 15:55:07 Yeah you're right, but that means I have to constantly check CVE's and shit for what is happening 2019-06-29 15:56:40 lol? 2019-06-29 15:58:21 "I care but I don't care" 2019-06-29 15:59:19 you'll have to decide for yourself which path you want to take 2019-06-29 15:59:27 we can't tell you what to do 2019-06-29 15:59:32 06.29 [17:55]* KH405 has quit (Read error: Connection reset by peer) 2019-06-29 15:59:35 <_ikke_> they timed out 2019-06-29 15:59:41 or just pay somebody... 2019-06-29 15:59:44 rip 2019-06-29 15:59:46 i just saw it, thanks 2019-06-29 16:07:48 Sorry I crashed, gotta love windows :( 2019-06-29 16:09:58 i crashed too, but it's not due to my OS, it's some bad hardware i haven't fixed 2019-06-29 17:18:36 Same, bad hardware acting with windows, gotta upgrade this PC soon 2019-06-29 17:21:46 Hi guys, which linux kernel version is in alpine 3.10? 2019-06-29 17:22:43 boeg: 4.19.53 2019-06-29 17:22:44 https://pkgs.alpinelinux.org/package/v3.10/main/x86_64/linux-vanilla 2019-06-29 17:22:47 thanks! 2019-06-29 17:25:10 you're welcome :) 2019-06-29 17:27:22 Is there a good guide somewhere for setting alpine up for a desktop environment for someone who has never used alpine except for via docker? 2019-06-29 17:29:45 boeg: alpine wiki have some (maybe outdated) guides, but don't forget setup-xorg-base 2019-06-29 17:30:11 mps: thanks! 2019-06-29 17:30:26 rest is on your preferences and hardware, i.e. video card driver and maybe firmware needed 2019-06-29 17:31:55 most users have good experience with xfce DE, but there are some other like mate or simple WM's 2019-06-29 17:36:22 Right now I just want to run it in a virtual machine, and was thinking of using i3 2019-06-29 17:36:53 however after logging in as root, the wiki says to run `alpine-install` but it says "not found" 2019-06-29 17:38:25 ah, `setup-alpine` it's called now 2019-06-29 17:38:29 contrary to the wiki :D 2019-06-29 17:43:08 boeg: All scripts are inside /sbin 2019-06-29 17:43:52 errn0[m]: thanks 2019-06-29 17:53:06 hm, my i3wm (on an aarch64 device) launches xinit, i3, and Xorg, but it does not display any text in the window titlebars or i3bar, and no text in the initial config window 2019-06-29 17:53:40 <_ikke_> missing fonts? 2019-06-29 17:53:46 i can launch rofi and a terminal (that works fine) if i rsync over an i3 config file, but even then it does not display any text in the menubar or titlebars 2019-06-29 17:54:01 i wouldn't think so, i have all the 'usual' fonts installed, multiple monospace fonts too 2019-06-29 17:54:21 it works fine in the terminal, which uses the same font (albeit xft rather than pango) 2019-06-29 18:04:08 i always have to fiddle a lot to make stuff like this work flawlessly on alpine :( 2019-06-29 18:07:43 danieli: Install ttf-dejavu 2019-06-29 18:08:17 errn0[m]: that was one of the first things i did 2019-06-29 18:08:24 droid sans mono, liberation, noto too 2019-06-29 18:08:27 danieli: I use dwm.. it worked straight after following this wiki... 2019-06-29 18:08:28 https://wiki.alpinelinux.org/wiki/Dwm 2019-06-29 18:08:46 i'm just using "exec i3" directly in ~/.xinitrc 2019-06-29 18:09:31 danieli, enable log file and check there what is missing 2019-06-29 18:09:41 exec i3 > $HOME/.log-i3-$(date +'%F-%H%M%S') 2>&1 2019-06-29 18:09:42 danieli: Check the config... maybe that needs some modifications. 2019-06-29 18:09:56 i did "exec i3 -V -d all > ~/i3.log 2>&1" 2019-06-29 18:10:03 didn't see anything except useless crap 2019-06-29 18:10:51 danieli: Sounds weird to me. I do not use i3 to verify your claim.. let me check... 2019-06-29 18:11:04 I am installing i3 :) 2019-06-29 18:11:17 errn0[m]: don't get me wrong, i've had several alpine installs where it worked flawlessly after a tiny bit of configuration 2019-06-29 18:11:20 but for some reason it does not in this case 2019-06-29 18:17:41 danieli: Its working here 2019-06-29 18:17:57 what are the exact steps you did, in short terms? 2019-06-29 18:18:09 http://0x0.st/z_PU.png 2019-06-29 18:19:20 danieli: apk add i3wm i3status dmenu and a terminal emulator 2019-06-29 18:19:38 setup-xorg-base, add the right drivers, add some fonts? 2019-06-29 18:19:51 and "exec i3" in ~/.xinitrc? 2019-06-29 18:20:07 danieli: Yes.. exec i3 at xinitrc 2019-06-29 18:20:30 driver was installed for my existing working dwm 2019-06-29 18:20:45 i tried on two aarch64 devices, a raspberry pi and another device 2019-06-29 18:20:53 guess they're f***ed then 2019-06-29 18:21:02 i can try recompiling and see if that sorts it 2019-06-29 18:21:19 danieli: Cool down ;) 2019-06-29 18:21:20 Shit happens all the time 2019-06-29 18:21:34 don't worry, i'm not angry or anything 2019-06-29 18:22:13 umm...i3gaps is prepackaged xD 2019-06-29 18:22:18 just wanted to check whether it was something with my device or configuration, the build for that architecture, or alpine's i3 in general 2019-06-29 18:22:33 cool, i did not realize that, i knew i3wm was in alpine repos but not i3gaps 2019-06-29 18:23:26 Its named as i3wm-gaps 2019-06-29 18:23:46 appreciate the help by the way 2019-06-29 18:24:06 \m/ 2019-06-29 18:34:57 danieli, maybe check if got "/etc/fonts/conf.d/70-no-bitmaps.conf" i3 using some bitmap font as a fallback by default 2019-06-29 18:37:12 MY_R: I don't have that file, and I have not modified any of the config files there 2019-06-29 18:38:29 danieli, and fc-list showing all fonts which you installed? 2019-06-29 18:38:35 yup 2019-06-29 18:38:41 ran fc-cache -fv and they showed up 2019-06-29 18:39:18 so ye weird :\ I would say to just "reboot" but you probbaly done it :D 2019-06-29 18:39:29 multiple times, even :) 2019-06-29 18:40:26 arch specific issue or something missing in repository, who knows, no weird WW and EE in xorg.log? 2019-06-29 18:40:48 had issues with xorg at first but i had forgotten a dependency, X seems to work fine 2019-06-29 18:40:59 it launches and all, i3 starts, but it just doesn't show any text 2019-06-29 18:43:06 danieli, and how many packages got installed with xcb-util* ? 2019-06-29 18:43:39 I have got 7 which are probably i3 deps 2019-06-29 18:43:42 xcb-util, xcb-util-{image,renderutil,cursor,wm,keysyms,xrm} 2019-06-29 18:43:51 7 here too, probably i3 deps 2019-06-29 18:44:20 the only error in xorg's log is that it failed to find data files for quirks in /usr/share/libinput 2019-06-29 18:44:26 so arm specific stuff and problem with render 2019-06-29 18:44:56 installed it now, let's see 2019-06-29 18:45:03 i doubt it'll make any difference at all 2019-06-29 18:45:07 ye that I had to install manualy too xinput/libinput 2019-06-29 18:48:18 i set DISPLAY and XSESSION and took a screenshot with maim 2019-06-29 18:48:22 https://i.imgur.com/zUKlqSB.png 2019-06-29 18:49:27 it looks more like gpu driver issue 2019-06-29 18:49:43 it's possible, i'll try screwing around with it 2019-06-29 18:51:12 but ye, no single font... only separators in status bar :D 2019-06-29 18:51:37 my terminal works fine if i rsync over an i3 config file though 2019-06-29 18:51:41 same font and everything 2019-06-29 18:51:57 but the difference in rendering engine might make a difference 2019-06-29 18:52:49 true 2019-06-29 18:53:01 you were right 2019-06-29 18:53:05 it works perfectly now 2019-06-29 18:54:08 i tried enabling vc4 on it but it did not like that 2019-06-29 18:56:47 danieli, heh, nice, good job :) 2019-06-30 01:48:35 I'm having an error while starting virt-manager: Using the ''memory'' Gsetting backend. Your settings will not be save 2019-06-30 01:48:58 I literally have no clue what is this, could anybody help ? 2019-06-30 01:49:21 I've read that installing dconf could help, but is this really my issue ? 2019-06-30 01:52:21 install dconf 2019-06-30 01:54:55 It didn't work 2019-06-30 01:55:11 can you copypaste the error message to a pastebin service and paste the link here? 2019-06-30 01:55:12 Do I need to start it, or just apk add and i'm done ? 2019-06-30 01:56:32 I'm getting another error now: failed to commit chances to dconf: Failed to execute child process 'dbus-launch' (no such file or directory) 2019-06-30 01:57:08 well, did you install dbus-x11? 2019-06-30 01:57:11 Do I need dbus-x11 ? 2019-06-30 01:57:15 https://pkgs.alpinelinux.org/contents?file=dbus-launch&path=&name=&branch=edge 2019-06-30 01:57:20 Yeah, I didn't :P 2019-06-30 01:57:26 if it's trying to launch that, then yeah, you most likely need it? 2019-06-30 01:57:42 I though dbus-launch was part of dbus 2019-06-30 01:57:49 And not dbus-x11 2019-06-30 01:58:27 i wasn't sure, so i looked it up 2019-06-30 01:59:47 It's working, thanks :D, sorry it's litteraly the first time i'm installing a GUI on a linux machine... There's way more little quircks than I though i would run into ^^' 2019-06-30 02:00:11 alpine tends to do things a bit differently, so it's a good idea to know where to look stuff up, and to do that 2019-06-30 02:00:48 Yeah, when starting virt-manager it gives an error cause it can't find systemd lol 2019-06-30 02:01:06 Of course you can't find it!!! :P 2019-06-30 02:01:18 is it an error or a warning? 2019-06-30 02:02:38 Warning 2019-06-30 02:03:07 good - warnings are not errors 2019-06-30 02:04:53 True, I find it just weird that it checks for systemd 2019-06-30 02:05:06 look into devuan and their struggles 2019-06-30 02:05:20 you'd be surprised at how much software has pretty hard dependencies on systemd 2019-06-30 02:08:58 Why would they do that ? 2019-06-30 02:09:39 I though going away form a systemd distro would free me :( 2019-06-30 02:17:48 Anyway, thanks alot for your help :) Sorry as I said litteraly the first time I install a GUI on a linux machine 2019-06-30 02:18:32 And i'm trying something kinda ballsy with this laptop, I want to replicate Qubes OS on alpine with KVM, dunno in which rabbit hole this is going to lead me tough :P 2019-06-30 03:14:58 I've actually once mapped out the amount of work required for that, a good exercise in itself, and came to the conclusion I'm better off with just understanding the components and their relations than actually giving up my free time to implement all of it 2019-06-30 03:16:32 I mean, I like the challenge and I'm fascinated by how Qubes has solved it, but I'm too fond of my free time outside IT :) 2019-06-30 03:29:43 speaking of KVM, I came across a relatively old tool for running KVM virtuals called kvmtool, it seems pretty good but I'm not sure if it's even developed anymore... looks much cleaner than anything qemu based 2019-06-30 03:48:09 Will definetly check it out! 2019-06-30 03:48:54 and yeah, the Qubes replacement looks kinda daunting, but also looks like I will learn alot doing it. My only fear is f*cking it up and loosing all my progress for some reason ... 2019-06-30 03:56:33 git everything you can and don't be afraid of committing too often :) 2019-06-30 03:58:53 yeah, there's a -lot- that goes into a full distro. I've twice created a distro derivative, it was a lot of work, and only for a derivative 2019-06-30 04:33:50 you'd be surprised at how much software has pretty hard dependencies on systemd <--- ?? 2019-06-30 04:34:04 i've not encountered any. rather just distro configurations that do 2019-06-30 04:36:34 I remember a couple of cases where so far the systemd dependency has been optional, but only a few that have gotten totally dependent on it 2019-06-30 04:36:53 and which have eventually become part of systemd 2019-06-30 04:37:31 one that springs to mind would be gummiboot which is now systemd-boot 2019-06-30 04:38:17 amazingly systemd-boot isn't tottally dependent on systemd in the sense it needs systemd running 2019-06-30 04:39:24 kind of understandable considering it runs before systemd :) 2019-06-30 05:41:56 I've got a dumb question: How is the repo working? Can anyone "upload" a package there without getting checked? I'm asking cause I've noticed every update is always happening during the day. 2019-06-30 05:46:51 And also, i've added to my Fstab, any idea why it isn't working? /10.0.0.50/share /mnt/storage cifs uid=0,gid=0,username=khaos,password=XXXXXXl,file_mode=0777,dir_mode=0777 0 0 2019-06-30 05:49:21 kh405, repo is all built from the aports git repo 2019-06-30 05:49:46 Ok, gotcha 2019-06-30 05:50:09 Sorry for the noob questions, learning slowly :P 2019-06-30 05:55:35 no problem 2019-06-30 06:02:32 TBB: check out firecracker, cloud-hypervisor, and nemu for decent qemu alternatives 2019-06-30 06:05:49 What is wrong with QEMU exactly? 2019-06-30 06:18:34 attack surface 2019-06-30 06:18:44 (for one) 2019-06-30 06:19:19 it's a great project, but it's not always what you need 2019-06-30 10:05:47 anyone got opengl passthru using nvidia-runtime on docker working? 2019-06-30 10:06:30 the standard nvidia image is ubuntu based and I have compiled libglnvd for alpine but doesn't seem to work 2019-06-30 17:28:32 maxice8 (@maxice8:matrix.exqa.de): 2019-06-30 17:28:33 systemd-inertWhatYouUseD is always been a black magic to me. 2019-06-30 18:08:53 how do I do LDAP in alpine? 2019-06-30 19:23:49 qrvpzvb: LDAP for what? 2019-06-30 19:26:19 authentication, most likely 2019-06-30 20:15:43 does anyone know this error when install ssh2 for php? https://kopy.io/TmmNe 2019-06-30 20:48:36 ah, it's compiling