2021-04-01 00:04:55 something's weird mps https://paste.sr.ht/~bl4ckb0ne/67ad01a8f507de312a64d4c64f9b193c4cd4d20d 2021-04-01 00:05:14 OH i know 2021-04-01 00:06:24 fucked up fstab and mmcblk1p1 is not mounted to boot 2021-04-01 00:06:38 so it installed the kernel in a created boot folder on mmcblk1p2 2021-04-01 00:07:56 got eth0 :) 2021-04-01 00:08:00 no sata drives 2021-04-01 00:08:32 5.11.10 r3 was prolly right and not installed to the good place 2021-04-01 00:23:49 usb rtl port shows but doesnt work but that's ok 2021-04-01 05:54:31 ikke: what kind of assistance could it be? maybe i can advise of a prolonged downtime and get some help... 2021-04-01 05:55:16 bl4ckb0ne: I didn't understood you. does eth0 works with linux-edge 5.11.11-r1? 2021-04-01 06:36:35 Marquel: someone who understands the nlplug interface and can figure out why the block devices are not reported at boot 2021-04-01 06:37:31 One suggestion I now recall is looking at the buffer size used in nlplug-findfs 2021-04-01 06:55:58 as i said, i may be able to test, if not help with debug, not to speak of actually solving it ;) 2021-04-01 07:01:49 Would be nice if we found how to reproduce this in a vm, but this might be hw specific 2021-04-01 07:07:11 it is. 2021-04-01 07:07:29 i have two vms running 3.13 with LVM and they do not suffer from that issue. 2021-04-01 07:07:49 and no, i do not have access to similar hardware to test, even if i'd like to. 2021-04-01 07:08:15 though, maybe, i just need to ask the right people the right questions... 2021-04-01 07:08:18 i'll think. 2021-04-01 07:10:54 I have, just would take our x86_64 CI down 2021-04-01 07:11:14 your call ;) 2021-04-01 07:11:40 Not an issue, but would need to have something concrete to test 2021-04-01 07:11:54 well. 2021-04-01 07:12:51 my setup is a single host with a hardware-RAID1 over two SATA-SSDs. the RAID is fully provisioned, with a small GrUB boot area and LVM over the rest of the RAID. 2021-04-01 07:13:19 inside the LVM are numerous volumes, one is /, another is swap, the rest is for different VMs. 2021-04-01 07:13:56 the setup boots fine with 3.12 lts-kernels, while it fails to find its root with 3.13 lts. 2021-04-01 07:14:18 however, the root-volume inside LVM is specified with a UUID on the kernel commandline. 2021-04-01 07:14:28 that's as precise about my setup as i can get. 2021-04-01 07:15:08 UEFI or legacy bios 2021-04-01 07:15:12 ? 2021-04-01 07:15:41 let me see if i can find. 2021-04-01 07:16:54 legacy 2021-04-01 07:19:46 Same for our CI host 2021-04-01 09:55:43 hi, if I run apk upgrade and e.g. openssh is updated, do I need to manually restart the openssh service, so that the update takes effect? 2021-04-01 09:57:04 Yes, you to do it manually 2021-04-01 09:57:43 apk does not manage services 2021-04-01 09:57:43 You could automate it 2021-04-01 09:57:59 Write a script that restarts is after an upgrade 2021-04-01 09:58:05 is = it 2021-04-01 09:59:32 ok, thanks! 2021-04-01 09:59:47 Anytime 2021-04-01 10:21:44 Anyone had openrc start a service on boot when it's not enabled on the runlevel and it shows up in the manual runlevel? 2021-04-01 10:22:15 It's something i have happening on one host and i tried to look everywhere with no success 2021-04-01 10:25:25 It's not enabled on any runlevel, the service isn't started by any 'need' dependecy chains and the service that's symlinked to it has no correlation at all to the behaviour 2021-04-01 10:25:55 caskd: dependency for another service 2021-04-01 10:30:19 mps: why wouldn't it show in hotplugged/needed dynrunlevel then? 2021-04-01 10:31:26 before/after 2021-04-01 10:32:05 oh i see 2021-04-01 10:32:09 lemme look into that 2021-04-01 10:37:39 oh that apparently did it 2021-04-01 10:37:46 switched from need to use 2021-04-01 14:05:01 Hi! 2021-04-01 14:05:05 Are docs down? https://docs.alpinelinux.org/ 2021-04-01 14:10:41 works here 2021-04-01 14:11:17 Yes, now they do, at the time I asked here was nginx 404 2021-04-01 14:22:51 `dd if=/bla/bla of=/bla/bla status=progress` gives me: `dd: invalid argument 'progress' to 'status' 2021-04-01 14:23:07 Do I have to install anything else so statu will work with dd? 2021-04-01 14:24:20 Aha, coreutils pkg was missing 2021-04-01 14:25:10 also consider pv 2021-04-01 14:25:10 frojnd theres an alternative package 2021-04-01 14:25:26 mckaygerhard: which one? 2021-04-01 14:25:31 pv? 2021-04-01 14:25:54 https://wiki.alpinelinux.org/wiki/Alpine_newbie_apk_packages#coreutils_libc_and_utmps_in_alpine read the note 2021-04-01 14:26:20 is a shame tha the alpine wiki are not so well documented like my newbie old pages.. 2021-04-01 14:27:05 wtf is this 2021-04-01 14:27:47 Err maybe someone more knowladgable will know this... 2021-04-01 14:29:29 I'm playing in qemu and am tryuing to install full disck encryption. I've created a temp encrypted container of the sda. But when trying to wipe the container with zeros I get No space left on device: dd if=/dev/zero of=/dev/mapper/to_be_wiped status=progress 2021-04-01 14:29:34 Is this qemu specific? 2021-04-01 14:30:04 umm the disk and the size are wll defined? 2021-04-01 14:31:01 mckaygerhard: no I didn't add any partition table yet 2021-04-01 14:31:36 This is lsblk output: https://termbin.com/p1phc 2021-04-01 14:31:53 if the size was not well efined SIZE OF ISK not partitions or table disk 2021-04-01 14:33:35 also frojnd seem there's no mappe defined so no definition to wipped to! XD 2021-04-01 14:33:54 mckaygerhard: I figured... I canceled during disk setup 2021-04-01 14:34:37 that s common error no problem i made a asemu alpine wiki page too but i canceled cos there's no good work today's 2021-04-01 14:48:24 mps: yes i have eth0 with an ip 2021-04-01 14:48:44 had some fstab fuck up and i ended up installing the kernel not in the mounted boot dir 2021-04-01 14:49:04 i think next sd card clean im gonna do just one whole ext4 partition 2021-04-01 15:14:47 bl4ckb0ne: ok, I think we can't do more with alpine kernels 2021-04-01 15:15:07 bl4ckb0ne: thank you testing all this 2021-04-01 15:15:22 thanks you for your time 2021-04-01 15:15:46 bl4ckb0ne: I don't like to give advises but having separate boot partition is good idea, ime 2021-04-01 15:15:50 gonna continue to poke on my side, i'll tell you if i find anything 2021-04-01 15:17:01 bl4ckb0ne: also I will try to 'keep one eye' on sata driver for helios64 in hope that it will appear soon in mainline 2021-04-01 15:17:38 and, if get any answer from kobol please tell me 2021-04-01 15:18:01 yep 2021-04-01 15:18:14 im mailing the guy that worked on armbian-build 2021-04-01 15:19:26 also I will try to include changes to one of next linux-edge 'normal' releases 2021-04-01 15:19:47 sounds good 2021-04-01 15:28:53 mps havin separate partition today is waste of time.. almost all home install uses same partition.. disk is so moderm today! that is good for server.. when i setup alpine always is a pain on home made install.. (cos i just want to setup dual boot etc) 2021-04-01 15:32:58 mckaygerhard: everyone is free to keep his/her religion 2021-04-01 15:33:21 ACTION clings to his copy of The C Programming Language 2021-04-01 15:34:49 for me having separate boot, root, home and some other partitions is helpfull for backup, restore, move, fix FS problems etc 2021-04-01 15:37:28 I used to do that before I got a high capacity NAS 2021-04-01 15:37:36 now I treat my OS installations as disposable 2021-04-01 15:38:15 of course, there is always a legitimate reason to do it differently in different contexts 2021-04-01 19:12:23 how do i build with gcc6? 2021-04-01 19:13:18 you'd need to build gcc6 2021-04-01 19:13:54 Hello 2021-04-01 19:14:11 ekhas: alpine 3.8 is the last release with gcc 6 2021-04-01 19:14:24 what about gcc-6 pkg? 2021-04-01 19:14:37 i cant use that? 2021-04-01 19:14:48 Doas anyone know how to stop udhcp from automatically setting nameservers from dhcp? 2021-04-01 19:15:09 I have been trying to stop it for about a month but no success so far 2021-04-01 19:15:34 nolem: there is an option, hold on 2021-04-01 19:16:06 RESOLV_CONF=no but i forget where it goes 2021-04-01 19:16:32 RESOLV_CONF=no in /etc/udhcpc/udhcpc.conf 2021-04-01 19:16:42 thanks 2021-04-01 19:17:08 doesn't look like it fits the format but I'll try it and let you know the results 2021-04-01 19:17:29 what gcc-6 pkg for? 2021-04-01 19:17:30 it's not an existing file/directory 2021-04-01 19:17:43 ekhas: oh yes, we still have gcc6 2021-04-01 19:18:31 how do i tell abuild to use gcc6? 2021-04-01 19:19:58 probably by setting the appropriate flags in /etc/abuild.conf 2021-04-01 19:20:00 like CC 2021-04-01 19:20:41 i tried export CC=gcc-6 in terminal 2021-04-01 19:20:53 it didn't work 2021-04-01 19:23:26 i did apk add gcc6 but abuild installs gcc as a dependancy anyway 2021-04-01 19:24:06 yes, apparently gcc6 depends on libgcc 2021-04-01 19:24:13 which is provided by 'gcc' 2021-04-01 19:25:05 no, i can have gcc6 added w/out gcc 2021-04-01 19:26:55 i just did apk add gcc6, yes libgcc gets installed 2021-04-01 19:27:16 w/out gcc 2021-04-01 19:27:27 hmm, yeah, libgcc just depends on musl 2021-04-01 19:28:07 but when i run abuild -r that's whan gcc gets installed and i dunno y 2021-04-01 19:28:57 apkbuild file doesn't have gcc as dep 2021-04-01 19:29:15 recursive dependency? 2021-04-01 19:30:52 hmm RESOLV_CONF=no didn't seem to have worked 2021-04-01 19:31:11 in what exact file did you put it? 2021-04-01 19:32:03 on /etc/udhcpd.conf was there by default 2021-04-01 19:32:05 ok imma try add listed deps manually 2021-04-01 19:32:37 nolem: it should be /etc/udhcpc/udhcpc.conf 2021-04-01 19:32:40 nolem: not that file 2021-04-01 19:32:55 udhcpcd is the daemon config 2021-04-01 19:33:14 ohh fair enough thanks 2021-04-01 19:40:51 that did work thanks ikke 2021-04-01 19:41:02 np 2021-04-01 19:54:58 add all deps from apkbuild by hand 2021-04-01 19:55:12 abuild adds gcc and few others anyway 2021-04-01 21:50:41 hi guys 2021-04-01 21:51:11 im interested in Pl.Deniability 2021-04-01 21:51:19 i did it many times 2021-04-01 21:51:21 https://wiki.alpinelinux.org/wiki/Setting_up_a_laptop#Improved_method_with_plausible_deniability 2021-04-01 21:51:29 but it doesnt work 2021-04-01 21:51:35 somethinh must be wrong 2021-04-01 21:51:43 could some one plz re-check it 2021-04-01 21:51:56 would be great 2021-04-01 21:52:34 i mean re-write and re-test settings and configs 2021-04-01 21:53:53 i dont want to go back to arch linux heh 2021-04-01 21:54:12 arch plain dm-crypt scheme works just fine 2021-04-01 21:54:19 but i want it on alpine 2021-04-01 21:54:34 and this tutorial... 2021-04-01 21:54:50 https://wiki.alpinelinux.org/wiki/Setting_up_a_laptop#Improved_method_with_plausible_deniability 2021-04-01 21:55:02 recheck it please 2021-04-01 21:56:16 should i write it on #alpine-devel ?? 2021-04-01 21:56:20 or is it ok? 2021-04-01 21:56:57 i dont know where to write... 2021-04-01 21:57:03 cuz https://meta.alpinelinux.org/register 2021-04-01 21:57:12 Registration is currently closed. 2021-04-01 22:01:17 this tutorial... 2021-04-01 22:01:28 i mean https://wiki.alpinelinux.org/wiki/Setting_up_a_laptop#Improved_method_with_plausible_deniability 2021-04-01 22:01:33 (Last edited by Nangel on 26 Apr 2020.) 2021-04-01 22:01:37 hell ) 2021-04-01 22:01:54 where are you brother, Nangel? 2021-04-01 22:02:03 hehe 2021-04-01 22:03:28 what doesnt work? 2021-04-01 22:03:51 https://wiki.alpinelinux.org/wiki/Setting_up_a_laptop#Improved_method_with_plausible_deniability 2021-04-01 22:04:00 this tutorial 2021-04-01 22:04:06 looks like outdated 2021-04-01 22:04:56 you are very redundant and not very specific. hard to help 2021-04-01 22:05:11 specific? 2021-04-01 22:05:22 well... 2021-04-01 22:05:54 Luks scheme works just fine 2021-04-01 22:06:44 on the othe hand 2021-04-01 22:06:53 plain dm-crypt device doesnt boot 2021-04-01 22:07:10 thats all 2021-04-01 22:07:43 heh i dont know what to say )) 2021-04-02 03:21:11 does anyone else notice that scrolling in graphical terminal emulators in programs like less and vim feels really sluggish compared to in the virtual console 2021-04-02 03:24:04 (not alpine specific) 2021-04-02 03:31:18 it's not slow 2021-04-02 03:31:20 it's choppy 2021-04-02 03:31:23 that's different 2021-04-02 03:31:25 depends on the terminal you use 2021-04-02 03:31:59 yeah, alacritty and st feel the best to me but still a little unsatisfying 2021-04-02 03:32:03 there's a whole spectrum of responsiveness in graphical terminal emulators 2021-04-02 03:33:23 simulated cursor input will always be crappy 2021-04-02 03:58:54 My /dev/input/mice is missing. Is there some daemon I need to restart? 2021-04-02 03:59:56 Plugging in a USB mouse doesn't add it. 2021-04-02 04:00:52 though it does add /dev/input/eventN 2021-04-02 04:09:49 Never mind, got it on reboot 2021-04-02 10:44:07 Good morning 2021-04-02 10:44:22 Does javaws (for running java stuff like idrac and co.) not exist anymore? I tried ` apk search cmd:javaws` and it does not return any results 2021-04-02 10:45:42 seems to be part of icedtea-web, which is not in edge 2021-04-02 10:46:34 It has been moved to unmaintained 2021-04-02 10:46:41 2edfb76871e79a847245ccfb4b63cbdb729726af 2021-04-02 10:46:52 No reason why 2021-04-02 10:47:06 ikke: thanks a lot. How did you find it that quick? 2021-04-02 10:47:34 https://pkgs.alpinelinux.org/contents?file=javaws&path=&name=&branch=v3.13 2021-04-02 10:48:40 telmich: idrac web console is HTML5 capable for iDRAC8 and younger since at least a year and javaws is not required anymore. 2021-04-02 10:48:41 when I knew the pacakge, I did git log --grep icedtea-web 2021-04-02 10:49:37 i stand corrected: iDRAC7 and younger is HTML5 capable, if updated. 2021-04-02 10:49:44 Marquel: I'm unfortunately dealing with some idrac 6 (!) stuff :-/ 2021-04-02 10:51:34 telmich: my condolences. those things have been out of service for years now. so unless you are running them private i suggest you urgently request a hardware refreshment. 2021-04-02 11:18:41 jjuran: /dev/input/mice is obsolete anyways 2021-04-02 11:20:12 Hello71: My Mac emulator uses it 2021-04-02 11:20:36 what exactly is "mac emulator" 2021-04-02 12:23:24 Hello71: a VM host on a Mac? 2021-04-02 12:33:56 Hello71, wyoung: https://www.v68k.org/advanced-mac-substitute/ 2021-04-02 12:34:19 Well, the fbdev front uses it 2021-04-02 12:34:25 *front end 2021-04-02 12:38:43 Right, I use VBox 2021-04-02 12:40:34 Mac as in 1984 2021-04-02 12:54:15 so it's not "on a mac" 2021-04-02 12:54:41 regardless, mousedev almost certainly won't be removed from the kernel 2021-04-02 12:55:01 so as long as you don't want any new features like smooth scrolling or hotplug or multiple device support you should be fine 2021-04-02 12:58:06 The OS I'm emulating doesn't understand scroll wheels. 2021-04-02 13:14:15 it was happy to understand one button 2021-04-02 14:36:37 yes 2021-04-02 14:37:16 nowadays mac has left and right and middle and also force, it's ridiculous 2021-04-02 14:48:38 gestures =) 2021-04-02 14:48:44 two finger swipe and so on 2021-04-02 14:49:02 Cogitri: do you by any chance do video calls in Google Meet on Firefox? 2021-04-02 14:50:13 Nop, I don't use Google Meet, but I use Jitsi from time to time 2021-04-02 14:54:47 ey guys what do you think about FreeTube, in the middle of full browser <> youtube-dl 2021-04-02 14:57:12 anyone has built it on alpine? 2021-04-02 15:04:47 why some service started but not listed in rc-status ? 2021-04-02 15:07:46 even ls /etc/runlevels/default can found the service, but rc-status --all do not list the service 2021-04-02 15:13:50 what service is wener[m] ? 2021-04-02 15:19:18 nebula.ops-net 2021-04-02 15:21:59 reboot found a error line, nebula.ops-net is the real and virtual service 2021-04-02 15:43:24 fixed by remove provide 2021-04-02 16:25:53 ok great 2021-04-02 17:06:53 artok: yes but that's not a click ;) 2021-04-02 17:07:27 Cogitri: hmm. Well where Jitsi Meet works perfectly, Google Meet likes to crash on me. Annoyingly often 2021-04-02 18:12:26 Is it correct that it's not possible to use Signal desktop on Alpine? Or is there some other client? 2021-04-02 18:24:42 you can use flatpak wozencroft 2021-04-02 18:25:14 but maybe you can compile it from source 2021-04-02 18:34:16 how do i disable modesetting? 'nomodeset' doesn't seem to work 2021-04-02 18:35:44 mikoxyz: where did you put nomodeset 2021-04-02 18:38:31 ikke: in /etc/update-extlinux.conf (did run update-extlinux afterwards) 2021-04-02 18:38:51 mikoxyz: cat /proc/cmdline 2021-04-02 18:41:23 nomodeset shows up there, too :/ 2021-04-02 18:41:52 ok, why do you think it's not set? 2021-04-02 18:43:44 because it's running at 1024x768? 2021-04-02 18:45:19 does your graphics card not support kernel modesetting? 2021-04-02 18:46:59 it does support it (i shouldn't even get video out at this point if it didn't, no?), but 640x480 simply looks better on my crt 2021-04-02 18:47:58 Isn't that a matter of setting the resolution then? 2021-04-02 18:49:11 well, i did try to force kms to set the resolution to 800x600 (which seems to be as low as kms wants to go), but that didn't work either 2021-04-02 19:13:04 donoban I see, flatpak would be ok, thank you for the hint. 2021-04-02 19:29:28 I'm having some issues getting alpine linux to mount a LVM root volume.. I've added lvm to the initramfs, but when I try to boot, the initial ram fs fails to mount my root with "mount: mounting /dev/mapper/vg0-root on /sysroot failed: No such file or directory". I'm at a loss for what it's complaining about - does anyone have any ideas? I know this an issue with my setup and not alpine - 2021-04-02 19:29:31 when I installed from setup-alpine and used lvm, everything worked smoothly 2021-04-02 19:33:53 OK! I figured it out.. The ext4 module wasn't being loaded - I think I have to add this to my mkinitfs configuration 2021-04-02 19:40:52 Hmm; actually maybe I didn't completely figure it out. I can boot just fine if I do "modprobe ext4; mount /dev/vg0/root /sysroot; exit". But even with ext4 in my mkinitfs.conf, I still have to manually load the ext4 driver on boot.. Very strange; not sure why the initramfs wouldn't just load the ext4 module 2021-04-02 19:45:06 Aha! Looks like I needed to provide "rootfstype=ext4" on the kernel command line. All is well now \o/ 2021-04-02 19:45:32 mikoxyz: x86 vbios goes up to something like 1600x1200 2021-04-02 19:46:29 There used to be a hardened-kernel alpine version with grsec patches. Now there is e.g. hardened kernel gentoo or arch. 2021-04-02 19:46:30 i think if you aren't using kms it should use whatever the bootloader sets 2021-04-02 19:46:54 Will there be again a hardened kernel version in the future? Or not to expect? 2021-04-02 19:48:16 there is much less value provided by current hardened kernel. upstream kernel says it's because they've integrated everything useful from pax, pax people claim that it's because grsec is a special snowflake and nobody can replicate it 2021-04-02 19:48:29 the reality is probably closer to the former 2021-04-02 19:49:22 Ok, I see, thank you for the hints 2021-04-02 19:50:21 i don't think alpine is specifically opposed to adding linux-hardened based on anthraxx current work but someone needs to maintain it and test it for every new release etc 2021-04-02 19:52:26 Hello71: the bootloader does seem to run at 640x480, though (no cmdline parameters are set that change the resolution, either) 2021-04-02 19:52:50 cat /proc/fb 2021-04-02 19:53:05 efi or bios? 2021-04-02 19:53:14 bios 2021-04-02 19:53:46 /proc/fb says: "0 radeondrmfb" 2021-04-02 21:10:55 with nomodeset? 2021-04-02 21:11:47 Yeah, Deepin is great, I found it the best for touch as well. 2021-04-02 21:14:55 Hello71: yes 2021-04-02 21:15:12 that doesn't sound right 2021-04-02 21:15:46 indeed, it doesn't 2021-04-02 22:36:04 I had to admit this, but years of Irssi use are coming to an end. 2021-04-02 22:36:11 I really like The Lounge. 2021-04-02 22:36:22 s/had/hate/ ** 2021-04-02 22:36:44 Plopped it on my Pi4 lil personal box. 2021-04-02 23:03:31 fys: what u like ther? to run it in a browser? :/ 2021-04-02 23:04:33 Yeah, convenient for me atm. 2021-04-02 23:16:24 .. spell check is a bonus for idiots like me. 2021-04-02 23:21:24 it's okay, it's just temporary ... u will come back to irssi :/ 2021-04-02 23:25:22 Haha, I'm not uninstalling it. ;) 2021-04-03 00:55:07 what is the process for determining when a programming language's libraries may be included in aports? 2021-04-03 00:57:18 the alpine wiki page "APKBUILD examples:Haskell" suggests that "It is not necessary to create a separate package for a Haskell library" 2021-04-03 00:59:16 which is fine, but it means that packages that depend on haskell libs need to download+install those libs themselves, which means 1) they need access to the network and 2) such libraries will have to get reinstalled every time they are used for a build 2021-04-03 00:59:38 outerpassage: that's why nothing will ever replace C 2021-04-03 00:59:49 (which is a little to my lament) 2021-04-03 01:00:20 though, I'll never stop loving C 2021-04-03 01:00:29 I love C as well :) 2021-04-03 01:00:34 but for example we do package libraries for python 2021-04-03 01:02:04 and it looks like aports even has some ocaml packages included, which is much closer to Haskell 2021-04-03 01:02:30 this came up here https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/19978#note_150060 2021-04-03 01:03:00 haskell not having a stable ABI probably has something to do with it 2021-04-03 01:03:35 this is part of the reason why as soon as a language gets a package manager, I tend to think it's a lost cause 2021-04-03 01:03:40 it's just embracing dep-hell 2021-04-03 01:04:50 c7s: interesting, thanks 2021-04-03 01:07:35 halosghost: amen 2021-04-03 01:08:04 every time i see some install tutorial and it's filled with language specific install commands i sigh out loud 2021-04-03 01:09:48 fys: it's why I'm a touch bummed that zig is gonna get a package manager 2021-04-03 01:09:57 it has such good C-interop, they should just stick with that 2021-04-03 01:10:08 smh 2021-04-03 01:10:12 zig is getting a package manager? that is surprising to me 2021-04-03 01:10:52 And straight to obscurity you go. 2021-04-03 01:11:04 I would have thought they would avoid that kind of thing 2021-04-03 01:14:59 outerpassage: yeh 2021-04-03 01:15:06 🤷 2021-04-03 01:15:14 apparently a lot of people like them… 2021-04-03 01:15:15 🤷 2021-04-03 01:23:28 a lot of people also watched The Big Bang Theory 2021-04-03 01:23:30 just saying 2021-04-03 01:26:58 fys: I agree 2021-04-03 01:43:17 I love C as well :) 2021-04-03 01:43:25 it is statements like this that give me faith in humanity 2021-04-03 01:43:45 I've spent my entire day programming in C, can't get enough of it 2021-04-03 01:45:11 lol 2021-04-03 01:45:32 don't get me wrong, I see so many flaws in it, and it's such a product of its time (and its creators) 2021-04-03 01:45:42 but, I'll switch my personal projects to something else when I see something else better 2021-04-03 01:46:21 (or, admittedly, when they'll benefit from something else; I'm not really a zealot) 2021-04-03 01:46:22 I'll do C++ if I inherit the codebase, I'll convert it to C if that's reasonable... but C++ is "okay" 2021-04-03 01:47:37 with C it's reasonable for a person to have the entire standard library memorized, that is a real benefit for me 2021-04-03 01:50:00 tehcloud: there are several parts of C++ I really enjoy 2021-04-03 01:50:18 (this one may seem really silly, but I love that C++'s ?: can return an lvalue) 2021-04-03 01:50:30 I like C++ strings and templates 2021-04-03 01:50:33 but… it isn't a small language ☺ 2021-04-03 01:50:47 templates are especially useful for handling audio in different formats 2021-04-03 01:51:04 ...and many other things :p 2021-04-03 01:52:04 hehe 2021-04-03 01:52:22 I actually find C's _Generic() enough for me in most circumstances, but there are plenty of times where templates would be welcome 2021-04-03 01:52:29 (because I'm strange) 2021-04-03 04:37:13 https://i.imgur.com/6xXDeZc.png 2021-04-03 04:37:19 When I run lsblk on my server. 2021-04-03 05:15:17 is there something wrong with it fys? 2021-04-03 05:15:39 No, just happy with myself. 2021-04-03 05:15:42 :) 2021-04-03 05:15:45 lol 2021-04-03 05:15:52 actually, where's your sdj :D 2021-04-03 05:16:10 Actually 2021-04-03 05:16:12 good question. 2021-04-03 05:16:21 All drives are accounted for in the arrays. 2021-04-03 05:16:25 ACTION shrug 2021-04-03 05:16:30 lol 2021-04-03 05:16:59 Thank god for UUIDs 2021-04-03 06:21:44 fys: out of curiosity, why choose RAID5 over 10? 2021-04-03 06:24:13 especially with such large drives, completing a rebuild on raid5 with drives over 3TB is really unlikely to complete without any random errors 2021-04-03 06:25:29 based on regular 10^14 BER, 25% chance of single bitflip 2021-04-03 06:25:54 (naive) 2021-04-03 06:26:16 raid6 is a little better here but still falls apart on larger arrays 2021-04-03 06:26:59 EF: I'd rather build a smaller RAID5 than a giant RAID10. 2021-04-03 06:27:05 re-build rather. 2021-04-03 06:27:37 The data is split so it minimizes downtime. 2021-04-03 06:27:51 None of the data is irrecoverable. 2021-04-03 06:30:24 Really the RAID level is there for convenience and to give a fighting chance to not have to re-do that array should I lose a drive. 2021-04-03 06:31:20 if that's the goal then you should have gone with raid6 2021-04-03 06:31:44 in a raid5 of this size, if you lose a drive the rebuild process is rather likely to introduce silent corruption 2021-04-03 06:31:46 The difference between RAID6 and RAID5 is more up to subjective opinion. 2021-04-03 06:31:54 "This size"? 2021-04-03 06:31:58 11TB is small. 2021-04-03 06:31:59 lol 2021-04-03 06:32:16 fys: thanks for your insight. Mine is a large 10 with a small 1 for /boot as could not boot 10 2021-04-03 06:32:16 This was the cheapest possible solution I could have. 2021-04-03 06:32:16 drives of this size 2021-04-03 06:32:40 unless these are enterprise-grade drives you've got at least 25% chance of random error during a rebuild 2021-04-03 06:32:42 4TB drives are considered quite smaller in the media server world. 2021-04-03 06:33:39 I've never in my life heard a number as high as 25% 2021-04-03 06:33:50 And I've rebuilt a lot of RAID5s in my day. 2021-04-03 06:34:10 Both mdadm and hardware. 2021-04-03 06:34:33 I value your insight but everything I've read suggests 5 vs 6 is up to opinion. 2021-04-03 06:34:44 what type of drives are these 2021-04-03 06:35:17 Hold on. 2021-04-03 06:35:31 If what you say is true it's inherent to the RAID level. 2021-04-03 06:35:48 if they've got a BER of 10^15 they're probably fine, but drives rated like that are pretty pricy 2021-04-03 06:36:09 What is this based on? These are numbers I've never heard. 2021-04-03 06:36:22 I'm genuinely interested. 2021-04-03 06:37:16 I've always used RAID5 out of habit from work and then when I went this route I did some re-fresher research and it came down to opinion from everything I read with no obvious winner between the two. 2021-04-03 06:39:07 if you lose a drive and have to rebuild, you're going to need to read the entire dataset on all the other drives. a 10^14 BER is 11.3 TiB, so even a 4 disk 4TB RAID5 is likely to hit an error 2021-04-03 06:39:18 more disks -> more likely to have UCE 2021-04-03 06:39:29 let me look for some article to describe it more clearly 2021-04-03 06:39:48 I understand what you're getting at, it doesn't make RAID6 a clear winner in my eyes. 2021-04-03 06:40:27 Reading over some stuff I remember my mindset now. I am willing to let one of the arrays die for better write performance throughout. 2021-04-03 06:40:51 raid6 extra parity reduces the chance considerably, but still isn't suitable for 100TB+ arrays 2021-04-03 06:41:00 Not my biggest concern. 2021-04-03 06:41:36 As I said, the RAID level is more a matter of convenience than counting on the redundancy factor. 2021-04-03 06:42:01 yeah if the data isn't irrecoverable it's fine :p but bitrot is still annoying 2021-04-03 06:42:35 I could recover from a single array failure from scratch fairly quickly. 2021-04-03 06:42:48 I've got spare drives ready and various ways to restore. 2021-04-03 06:43:24 Write performance is important enough to choose 5 over 6 but not important enough to go 10. 2021-04-03 06:43:43 So, that's where I'm at. 2021-04-03 06:45:40 ah i found this article that summarizes the issue https://www.digistor.com.au/the-latest/Whether-RAID-5-is-still-safe-in-2019/ 2021-04-03 06:46:02 Reading. 2021-04-03 06:46:37 personally i ran into this issue when building a 2.4PB storage cluster in 2018, while doing some initial testing after building all the equipment we repeatedly had random corruption after simulating failure+rebuild 2021-04-03 06:47:39 That's odd.. I've rebuilt lots of RAID-5's much larger than my own. 2021-04-03 06:47:55 I'm trying to remember how large the massive IBM array I had to deal with at an old job was. 2021-04-03 06:48:14 Also, IMO that article is *extremely* one-sided. 2021-04-03 06:48:32 I mean - if that's your mindset - I guess it's completely applicable to you. 2021-04-03 06:48:40 well that's the thing, the rebuilds completed fine, but verifying the data after the fact yielded random corruption multiple times 2021-04-03 06:48:48 But me reading it, my mind immediately goes to "this isn't my use case." 2021-04-03 06:49:12 if your data isn't that sensitive, probably not :p 2021-04-03 06:49:14 I can't say I've ever run into that. 2021-04-03 06:49:35 Honestly this is the first I've ever heard of someone being scared of RAID5. 2021-04-03 06:50:36 i'd imagine enterprise-grade arrays would be fine, since BER is 10^15 or 16 there 2021-04-03 06:50:46 but we were using really cheap consumer grade drives 2021-04-03 06:52:07 Are you familiar with shucking? 2021-04-03 06:52:23 yep 2021-04-03 06:53:17 generally fine, unless they're seagate, who have a tendency to use refurbished RMA drives in their really cheap models 2021-04-03 06:53:21 wouldn't trust those one tiny bit lol 2021-04-03 06:55:11 LOL - I don't trust Seagate drives as far as I can throw them. 2021-04-03 06:55:44 their enterprise grade drives are fine, consumer stuff on the other hand ... no bueno 2021-04-03 08:49:24 Is there any way to know which packages will be updated after the "apk upgrade"? 2021-04-03 08:50:45 apk upgrade -s 2021-04-03 08:50:50 --simulate 2021-04-03 08:51:02 also apk version -l '<' 2021-04-03 08:51:59 Oh wait... 2021-04-03 08:52:34 why i dont see any info about "simulate" in apk -h? :) 2021-04-03 08:52:52 ikke tx! 2021-04-03 08:53:06 apk upgrade -h 2021-04-03 08:53:30 Ah, thats how it works. My bad. Thank you! 2021-04-03 11:46:25 Hi there, I am getting a bit confused. I have done a `docker pull alpine:3.13.4` but inside the container, I am only seeing that it has Kernel 4.19.121. I would have expected something more modern? What am I missing, is there a newer version of upgrade that I have to do? 2021-04-03 11:48:28 adamretter: docker is not a virtual machine 2021-04-03 11:48:37 a container is an isolated process that runs on your host 2021-04-03 11:48:41 so you see your hosts kernel 2021-04-03 11:53:41 @ikke hmm that's interesting, so what happens when my host is a Mac? 2021-04-03 11:54:03 adamretter: on mac, it first starts a virtual machine, and runs the docker container inside that 2021-04-03 11:54:25 So you see the kernel of the virtual machine 2021-04-03 11:56:59 @ikke I see. I just checked the Docker for Mac release notes: "Downgraded the kernel to 4.19.121 to reduce the CPU usage of hyperkit. Fixes docker/for-mac#5044" 2021-04-03 12:05:45 hi guys, I'm wondering of there is some normalized way for running services on restricted cgroup/namespaces with openRC or s6, maybe would it be something considered on new alpine init system? 2021-04-03 12:06:39 donoban: https://ariadne.space/2021/03/25/lets-build-a-new-service-manager-for-alpine/ 2021-04-03 12:06:42 @ikke thanks that all makes sense now 2021-04-03 12:06:51 have a nice day everyone 2021-04-03 12:08:35 uhM I've already read it ikke, there is nothing about cgroups/namespaces 2021-04-03 12:09:51 Right, it's not explicitly mentioned 2021-04-03 12:10:07 But that would be one of the feautres 2021-04-03 12:10:11 features* 2021-04-03 12:10:23 oh great! :D 2021-04-03 12:10:55 I'm suprised that there is no option for doing that on OpenRC 2021-04-03 12:11:10 I'm not 2021-04-03 12:11:12 is it so difficult to do? 2021-04-03 12:11:37 There is not a lot of development happening on openrc 2021-04-03 12:12:25 uhM 2021-04-03 12:12:33 https://wiki.gentoo.org/wiki/OpenRC/CGroups 2021-04-03 12:14:42 I feel it it's pretty hacky but I thinking on try to execute some services with firejail 2021-04-03 12:19:14 why not switch alpine to systemd is proper question 2021-04-03 12:20:04 and glibc, btw 2021-04-03 12:20:13 *snicker* 2021-04-03 12:20:58 a lot of problems on alpine will be solved by these two 2021-04-03 12:22:40 ACTION thinking about using openwrt as workstation 2021-04-03 12:24:11 uhM.. I don't see the point 2021-04-03 12:25:33 do you see cgroups/namespaces as workstation oriented? 2021-04-03 12:26:38 donoban: no, I have in mind 'small, simple, (and maybe) secure' 2021-04-03 12:26:46 firejail is the security sandbox that provides at least one local root per year, right? 2021-04-03 12:27:00 jvoisin: :) 2021-04-03 12:27:26 hehe 2021-04-03 12:28:34 when will people learn that security is process and not product 2021-04-03 12:28:43 no security solution is perfect 2021-04-03 12:29:13 and (absolute security) is illusion, btw 2021-04-03 12:29:43 but I think that the privilege restriction that firejails provides is a good thing 2021-04-03 12:30:29 donoban: I agree, but people have to understand that it is just one 'brick in the wall' 2021-04-03 12:31:23 I was wondering if I could run a service, e.g. tor, cpgroup/namespace isolated using only openRC or alpine std configuration 2021-04-03 12:31:36 I'm not sure, but bubblewrap sound little better than firejail 2021-04-03 12:31:40 another 'brick in the wall' but simpler 2021-04-03 12:32:21 I don't know it very much, I will take a look 2021-04-03 12:35:26 I'm doubtful about cgroups/ns security benefits 2021-04-03 12:35:46 isolation yes 2021-04-03 12:35:48 ACTION has to go 2021-04-03 12:35:54 see you later 2021-04-03 12:36:11 donoban: cul 2021-04-03 13:13:37 Does default kernel for stable release includes zfs module? 2021-04-03 13:14:03 https://pkgs.alpinelinux.org/packages?name=zfs-lts&branch=edge 2021-04-03 13:15:02 hi. is neofetch no longer in the repo, or am i missing some other step? (alpine newbie). i follow https://github.com/dylanaraps/neofetch/wiki/Installation#alpine-linux but it said: ERROR: unable to select packages: neofetch (no such package): required by: world[neofetch] 2021-04-03 13:15:37 it's in the community repo 2021-04-03 13:15:44 https://pkgs.alpinelinux.org/packages?name=neofetch&branch=edge 2021-04-03 13:17:46 I have a Dell Inspiron B120; it has an i686 CPU. Has anyone installed Alpine on this platform? 2021-04-03 13:18:51 probably not that one specifically but it should work at least in theory 2021-04-03 13:20:05 Is there an i686 version somewhere? 2021-04-03 13:20:39 it's not literally an i686 is it 2021-04-03 13:20:48 It is 2021-04-03 13:20:55 It's a very old laptop 2021-04-03 13:21:25 I don't think they have laptops that old 2021-04-03 13:21:29 Intel Pentium M Processor 740 2021-04-03 13:21:55 Well that's odd 2021-04-03 13:22:02 hm 2021-04-03 13:22:09 I'm on stable and did apk add zfs-lts 2021-04-03 13:22:16 Since I found zfs-lts there.. 2021-04-03 13:22:27 friggin Intel -.- 2021-04-03 13:22:31 I keepk getting a ton of ERROR linux formware failed to rename.. 2021-04-03 13:22:56 Did I ran something I shouldnt :P 2021-04-03 13:23:04 I am running this in qemu... 2021-04-03 13:23:05 Hello71: indeed 2021-04-03 13:23:09 frojnd: are you using diskless install 2021-04-03 13:23:19 Hello71: no I'm using standard 2021-04-03 13:23:50 frojnd: that's a kind of boot media, not an installation method 2021-04-03 13:24:14 Welp, guess I'll have to take a weekend... *yay* 2021-04-03 13:24:59 Hello71: Well in which step do I choose diskless or system install? I didn't ran setup-alpine 2021-04-03 13:25:36 Because I am rtrying to partition my root as zfs 2021-04-03 13:25:46 arkho: which distros don't support i686? 2021-04-03 13:27:14 frojnd: you can't install kernel modules on alpine boot media. you need to download extended media instead 2021-04-03 13:28:21 Hello71: wait what? I need to download https://dl-cdn.alpinelinux.org/alpine/v3.13/releases/x86_64/alpine-extended-3.13.4-x86_64.iso <- this? 2021-04-03 13:28:36 Instead of https://dl-cdn.alpinelinux.org/alpine/v3.13/releases/x86_64/alpine-standard-3.13.4-x86_64.iso ? 2021-04-03 13:28:49 arkho: x86 alpine flavor 2021-04-03 13:29:26 Hi,how can I find out the version of kernel installed by setup-disk? 2021-04-03 13:30:19 frojnd: yes 2021-04-03 13:30:57 I need to rebuild the initial ram disk but /mnt/lib/modules doesn't exist 2021-04-03 13:31:11 cvm: you mean uname -a ? 2021-04-03 13:32:03 [07:17:44] I have a Dell Inspiron B120; it has an i686 CPU. Has anyone installed Alpine on this platform? 2021-04-03 13:32:19 arkho: i have alpine in service on i586 class machines still 2021-04-03 13:32:31 Apachez:ah that's it.thanks. 2021-04-03 13:35:08 Apachez: this shows booted kernel, though setup-disk usually install same one 2021-04-03 13:35:37 Hello71: ok let me try that then! 2021-04-03 13:35:52 mps: I chrooted and then ran that command 2021-04-03 13:36:33 cvm: then it shows you host kernel version 2021-04-03 13:37:52 I have run setup-disk /mnt but there isn't any modules directory inside /mnt/lib/ , is that an issue? 2021-04-03 13:38:14 probably 2021-04-03 13:38:55 Oh so probably re running setup-disk should fix it 2021-04-03 14:21:25 Hello71: Apparently this one lmao 2021-04-03 14:21:47 and what led you to this conclusion 2021-04-03 14:22:40 I can't find an i686 boot image. 2021-04-03 14:25:58 x86 as mps said 2021-04-03 14:26:37 ... D'oh, of course. I misunderstood. Thank you all. 2021-04-03 14:27:09 maybe we should set -march=i686 considering linux doesn't support 386 anymore 2021-04-03 14:40:12 I'm getting: 2021-04-03 14:40:13 $ apk add openconnect # openconnect (missing): 2021-04-03 14:40:15 Is the procedure to uncomment http://mirrors.gigenet.com/alpinelinux/edge/community from /etc/apk/repositories 2021-04-03 14:40:17 ? 2021-04-03 14:46:44 Hello, is anyone here using Alpine Linux as on a PC with plasma desktop? 2021-04-03 15:13:21 Hello71: u sure I can install zfs-lts under extended? 2021-04-03 15:14:02 I get the same error messages: https://imgur.com/a/cnjlv9T 2021-04-03 15:14:23 trwww: you can do that 2021-04-03 15:14:42 I would like to install alpine on my primary computer with zfs and whole system encryption... 2021-04-03 15:14:50 alexeymin: no, absolutely nobody, the maintainers just like to add 100s of packages for fun 2021-04-03 15:15:07 frojnd: as i recall either it comes with zfs or you need to do apk add zfs 2021-04-03 15:16:55 I 2021-04-03 15:17:10 Can't I just do mkfs.zfs /dev/sda3 ? 2021-04-03 15:17:15 After installing zfs :D 2021-04-03 15:23:18 Assuming extended does come with zfs support I should follow wiki: https://wiki.archlinux.org/index.php/Install_Arch_Linux_on_ZFS#Installation 2021-04-03 17:34:08 Hello there. Does anyone know how to bridge a wlan interface to lxc? I've been struggling for a couple days 2021-04-03 17:37:00 So far I've seen a lot of contradictory information. Some people say it can't be done, some people say it can, some say it's better to use NAT, but none of the things I've tested have worked for me so far 2021-04-03 17:37:18 We use NAT 2021-04-03 17:37:44 Do you have public IP addresses to assign to your container? 2021-04-03 17:37:44 drfailure: same as other net devices 2021-04-03 17:38:02 create bridge and add wlan to it 2021-04-03 17:38:13 Well I was planning to use the same one as I use on my main machine 2021-04-03 17:38:25 I'm a bit of a noob when it comes to interfaces and linux networking 2021-04-03 17:38:32 so yeah... I don't know very well what I'm doing 2021-04-03 17:38:37 From what I understood, wlan uses different headers 2021-04-03 17:38:54 extra mac-address headers 2021-04-03 17:38:57 ikke: it works fine for me 2021-04-03 17:39:09 nothing special 2021-04-03 17:39:33 802.11? 2021-04-03 17:39:53 you need to enable 4addr and not all endpoints support it properly 2021-04-03 17:40:43 Oh yeah I saw a solution that used that too it's the one that got me the furthest but then dhcp wouldn't work 2021-04-03 18:29:23 Does alpine somehow deals with 'flavors' of the same package in repositories? 2021-04-03 18:34:09 limited 2021-04-03 18:49:24 well, we have for example emacs-nox, emacs-gtk2, emacs-motif, emacs-qt or so 2021-04-03 18:57:48 not alpine question but, anyone know for simple hugo (site generator) guide which could help uninitiated to make site 2021-04-03 19:10:16 Haven't used hugo myself yet 2021-04-03 19:12:25 I use zola for https://arvanta.net/alpine/ and it is fine for this but I want to make something with 'bells and whistles' and as I read and see hugo is good for this 2021-04-03 19:16:40 I tried hugo several times, it's fast and easy, but not as fun as do everything in editor :/ 2021-04-03 19:18:01 maintaing site in vim is ok for sites which don't have changes often 2021-04-03 19:24:05 depends what u want, there are some very simple generators without any extra functions, practically they add just head footer crap, some kind of scripts 2021-04-03 19:24:28 but if want something so called 'featured' hugo is good at least it was easy and fast without too much nonsense 2021-04-03 19:25:04 ACTION handrolls HTML 2021-04-03 19:25:06 🤷 2021-04-03 19:25:09 at least it does not need nodeshit and other 2021-04-03 19:25:26 I also don't change it super-frequently, and I have exactly one blog post (which is still under development since it's article-length at this point) 2021-04-03 19:25:28 I have this already (made by me), but I want/need something with 'bells and whistles' 2021-04-03 19:25:56 install hugo, I think it's without dependencies 2021-04-03 19:26:06 some kind of 50mb bullshit 2021-04-03 19:26:51 the joy of go 2021-04-03 19:26:55 as all go pkgs 2021-04-03 19:29:18 big as wordpress :/ 2021-04-03 19:31:13 g8U4Z1xGzjlm: so practically different package ... just thought about it as most packages are somehow 'full-featured' with a lot of nonsense 2021-04-03 19:53:21 i'm not able to copy large file over samba, there is no error in log too 2021-04-03 20:46:31 define large file? 2021-04-03 20:46:39 2GB, 4GB, or more? 2021-04-03 20:50:37 8g 2021-04-03 20:54:45 https://forums.unraid.net/topic/79454-file-transfer-results-in-error-0x8007003b/ 2021-04-03 20:54:53 looks like windows error 2021-04-04 02:26:04 where can i get the man 3 pages for musl libc? 2021-04-04 02:26:32 doesn't seem to be any corresponding doc package for musl-dev 2021-04-04 02:30:30 there is a package for that iirc… struggling to remember what it is (not on my alpine machine right now) 2021-04-04 03:04:38 I assume it does not exist, there is only man on musl page :/ 2021-04-04 03:07:11 found it, it's in `man-pages` 2021-04-04 03:07:13 :facepalm: 2021-04-04 03:08:21 just learned that https://pkgs.alpinelinux.org/contents lets you search packages by contents. would be nice to have that functionality in apk itself 2021-04-04 03:09:28 man-pages exists in alpine? :/ 2021-04-04 03:10:13 outerpassage: u can search like that using apk too 2021-04-04 03:10:57 how? 2021-04-04 03:12:06 apk-file but it's not part of apk 2021-04-04 03:13:33 oh I meant already installed package with info 2021-04-04 03:14:09 Hello71: ah, thank you! 2021-04-04 03:15:44 where did u find that man-pages ? 2021-04-04 03:16:24 it's in main 2021-04-04 03:16:43 https://pkgs.alpinelinux.org/packages?name=man-pages&branch=edge 2021-04-04 03:18:17 oh u meant man-pages for kernel manual 2021-04-04 03:20:43 yeah. it includes sections other than just the syscall section, which I didn't realize until now 2021-04-04 03:20:51 including libc which is what I was after 2021-04-04 10:29:14 Hi guys, after the git.php.net hack, do you think that enforcing gpg signatures for commits is a good idea? 2021-04-04 10:30:19 dunno if that would help since the accounts used came from lastpass which the reaul user was using 2021-04-04 10:30:44 but it would limit the options to inject unwanted code but then the question might be what is opensource? 2021-04-04 10:31:29 I think if you enforcing using devboxes without internet is a better bang for the buck 2021-04-04 10:47:09 Apachez: if your usually sign your commits with a privatekey in your computer, compromising the public repository they can't fake commits in your name 2021-04-04 10:47:57 unless already pwned your computer you commit stuff from 2021-04-04 10:48:03 which seems to what happend in the php case 2021-04-04 10:48:09 where two of the devs got pwned 2021-04-04 10:48:23 their credentials were copied and then malicious code was commited in their name 2021-04-04 10:48:39 really? I thought that only the public repository was compromised 2021-04-04 10:49:02 but their commits aren't signed, so if you control the repository access you can fake commits easily 2021-04-04 10:49:34 again signing wont help if the signing key is compromised 2021-04-04 10:49:47 and signing is a good thing but dont trust it too much 2021-04-04 10:50:14 a better effect is to do the development from a computer which isnt connected to the internet 2021-04-04 10:50:39 but if you need to share the code in a public place (connected to the internet) 2021-04-04 10:50:47 and when you commit you can use your signing key (of this box) and burn the commit on a cd or such and then take that package to a internetconnected computer and push it to the repo 2021-04-04 10:50:57 you need some way for trust that the public version is the same that what you did offline 2021-04-04 10:51:17 yeah but using key is not differnt from using user/pass 2021-04-04 10:51:23 its a set of characters 2021-04-04 10:51:44 keys often have a fixed size of characters like 2048 bits (256 bytes) 2021-04-04 10:51:51 it's slighly different 2021-04-04 10:52:10 with user/pass you have to trust the centralized server/repository, which handles the authorization process 2021-04-04 10:52:29 not when it comes to actually commiting stuff 2021-04-04 10:52:34 with key signatures the trust is shared between all actors 2021-04-04 10:52:45 again if the dev box is pwned then signing wont help because the evil user will using the real signing key 2021-04-04 10:52:54 so signing in that context can be worser due to false sense of security 2021-04-04 10:53:02 "its signed so its fine" while in fact it isnt 2021-04-04 10:53:12 but again dont get me wrong 2021-04-04 10:53:19 yeah, I believe that the problem with php was the public server being compromised, not the developers computers 2021-04-04 10:53:19 signing by default is a good thing 2021-04-04 10:53:35 but if you want to resolve the current attackvector stop using internet connected devices to do your coding at 2021-04-04 10:54:05 if I recall it correctly some years ago this happend to kernel.org 2021-04-04 10:54:15 the real signing key was used to commit malicious code 2021-04-04 10:54:40 that is the real signing key by one of the devs 2021-04-04 10:54:43 uhM, I remember that they noticed a hack but I think that they didn't find code manipulation 2021-04-04 10:55:15 but its like putting a lock on your front doot 2021-04-04 10:55:30 by todays standards its stupid to not have that if you live in a suburb area 2021-04-04 10:55:33 https://www.securityweek.com/linux-source-code-repository-kernelorg-gets-hacked 2021-04-04 10:55:41 but its not foolproof that not any unwanted will be able to bypass that 2021-04-04 10:55:51 doot=door 2021-04-04 10:57:23 probably altough good is not so important, at least not so much than having eyes looking at any changes, that's the main protection 2021-04-04 10:58:02 "a cryptographically secure SHA-1 hash is calculated to uniquely define the exact contents of that file", I wonder if anyone who still sign up for that quote? ;) 2021-04-04 10:58:16 2011 is that pre-stuxnet? 2021-04-04 10:58:26 no 2021-04-04 10:58:41 2010 2021-04-04 10:58:44 developed since 2004 2021-04-04 10:58:47 err 2005 2021-04-04 10:58:57 utilizing "cryptographically secured hashes" 2021-04-04 10:59:45 and the fork named Flame https://en.wikipedia.org/wiki/Flame_(malware) 2021-04-04 11:00:23 "Flame was signed with a fraudulent certificate purportedly from the Microsoft Enforced Licensing Intermediate PCA certificate authority." 2021-04-04 11:00:37 but its like with https nowadays 2021-04-04 11:00:49 https on its own wont stop malicious stuff from happening 2021-04-04 11:00:54 the internet is fundamentally broken and needs to be reinvented 2021-04-04 11:00:55 but it will narrow down the attackwindow 2021-04-04 11:01:04 I say we just dump it, and while we're at it we go back to horse and buggy 2021-04-04 11:01:17 just scratch IPv6 while we are at it :P 2021-04-04 11:01:26 SLAAC with dns entries from day 1 :) 2021-04-04 11:01:31 well, I am somewhat serious about it being totally broken 2021-04-04 11:01:45 many protocols we rely on were invented in an era before considerations of security 2021-04-04 11:01:46 yeah its a gigantic what we call in swedish "lapptäcke" 2021-04-04 11:01:58 patchwork quilt 2021-04-04 11:02:07 I won't name any names (except SMTP!!!!! UGH!!!!) 2021-04-04 11:02:25 well, any of the older protocols is totally broken 2021-04-04 11:02:27 and its grown to be this large with legacy all over the place so you cant just "start over" 2021-04-04 11:02:35 wasnt there some internet2 back in the days? 2021-04-04 11:02:40 that too seems to have failed :P 2021-04-04 11:02:48 and another problem is that new protocols will be built with surveillance/tyranny built in 2021-04-04 11:03:09 so this is where we get back to the horse and buggy conversation 2021-04-04 11:03:11 the same goes for PSTN with its broken SS7 signaling 2021-04-04 11:03:38 not to mention when security organisations do their best to make things broken by design https://en.wikipedia.org/wiki/Dual_EC_DRBG 2021-04-04 11:03:56 well, that's what I mean... nobody actually wants to build secure protocols 2021-04-04 11:04:27 more time is spent to expand TR069 than to actually fix BGP with signing certs 2021-04-04 11:04:28 not saying I've read the Unabomber's manifesto or anything, but I heard he had something to say about this ;) 2021-04-04 11:04:48 and on top of this all the licensing stuff... so you end up with variants of the same solution 2021-04-04 11:04:49 maybe take this to #alpine-offtopic? 2021-04-04 11:04:58 sure thing :) 2021-04-04 11:05:07 hmmmm... is this actually offtopic? 2021-04-04 11:05:10 wasn't sure 2021-04-04 11:05:35 is the alpine repos signed for commits? 2021-04-04 11:05:35 yes, keep this channel under 10 lines per day like usually! 2021-04-04 11:06:00 darn... so now we have to be silent to at least 15th july to fix the stats? ;) 2021-04-04 11:06:25 ;) only, u typed a lot 2021-04-04 11:06:34 well, the conversation of whether all best practices regarding security are happening is definitely worth having... over and over and over 2021-04-04 11:10:20 well if soft is stable, simple and secure, then those corps can't make much money from it so ... 2021-04-04 11:12:57 back to the original point, signing commits is I think a good idea 2021-04-04 11:13:12 nothing is perfect, but with layers of good practices you are less likely to get hit hard 2021-04-04 11:14:54 yeah and is not very difficult to achieve 2021-04-04 11:21:16 but dont call it a day just because you start to sign your commits 2021-04-04 11:26:25 Hello. Anyone knows how to get g_ether working? I tried to modprobe it on alpine extended edge and it says it's not there 2021-04-04 12:13:41 long time did not use shitty xterm somehow have issue with color 2021-04-04 12:14:34 oh nvm, I am an idiot 2021-04-04 12:15:43 had broken colors bc. used xterm-256-color instead xterm-256color :/ 2021-04-04 12:16:59 ... now world looks much better :/ 2021-04-04 12:28:45 Is there a way to swap the udhcp busybox client with ISC dhcp client? 2021-04-04 12:29:14 I did not find any article to this in the official wiki 2021-04-04 12:34:43 needle: how do you mean? 2021-04-04 12:37:31 I think udhpc is run just with 'dhcp' in your /etc/network/interfaces 2021-04-04 12:37:47 needle: 3.13 or later of before? 2021-04-04 12:38:03 Oh Apachez you are here. I mean I would like to use another dhcp client instead of udhcp. Like f.e. the ISC one. 2021-04-04 12:39:49 helby: that is correct. 2021-04-04 12:40:06 ikke: I am using edge 2021-04-04 12:40:59 well using something else should be just to run that service and tells it what to do in its config no? 2021-04-04 12:41:02 I just thought there would be a oneliner where one can tell alpine linux, use another dhcp client binary and do not use udhcp client. 2021-04-04 12:41:25 yes, adding it as service 2021-04-04 12:41:34 I assume :/ 2021-04-04 12:41:55 it's not a service 2021-04-04 12:42:07 but not sure now, bc. all that ifdownup fuckery 2021-04-04 12:42:19 I know alpine is not gentoo, But in gentoo i can swap the dhcp clients. 2021-04-04 12:42:37 I thought since it uses openrc it might be somewhere in /etc/conf.d/ 2021-04-04 12:42:42 https://github.com/ifupdown-ng/ifupdown-ng/blob/master/executor-scripts/linux/dhcp 2021-04-04 12:43:16 if dhcpcd is installed, it should use that 2021-04-04 12:43:22 yo that looks like something that would be actually useful. 2021-04-04 12:44:07 On edge, that should already be in place 2021-04-04 13:12:10 Yeah as I recall it the symlink in /usr/bin would point to the isc one if you got the isc dhcp client installed 2021-04-04 13:12:53 the builtin dhcp client is just a syntax towards the busybox bin 2021-04-04 14:20:15 i don't think that makes sense. udhcpc is not at all compatible with dhclient 2021-04-04 14:20:32 the dhcp executor supports different clients 2021-04-04 14:20:46 yes 2021-04-04 15:27:12 trying to run palemoon with gcompat installed. Says libxul.so cannot find libmozsqlite3.so 2021-04-04 15:27:23 both files are in the palemoon dir 2021-04-04 15:37:13 ekhas: maybe seeting LD_LIBRARY_PATH? 2021-04-04 15:37:41 ok lemme add it 2021-04-04 15:58:20 @ikke ok it made it past that thx 2021-04-04 15:58:56 seg fault after that 2021-04-04 16:04:38 It would be wonderful in Palemoon was musl and glibc compliant 2021-04-04 16:05:58 says assertion failure invalid stack base unable to setup stack range for JS. segfault 2021-04-04 16:07:38 i tried compiling it with gcc10 it errors out with not found in namespace std 2021-04-04 16:08:38 chromium is such a bloatware 2021-04-04 16:09:04 please let me know when you find a browser that isn't 2021-04-04 16:09:29 tehcloud: w3m? 2021-04-04 16:09:54 i thought palemoon was less so 2021-04-04 16:09:55 hmmm... maybe I should try giving that a shot for a few days as my mainstay 2021-04-04 16:10:31 Palemoon has some... other... issues :p 2021-04-04 16:10:38 palemon is built on sabotage linux, musl based 2021-04-04 16:11:27 i was able to build palemoon on alpine 3.8 with gcc6 2021-04-04 16:11:51 worked great 2021-04-04 16:12:42 afraid gonna have to bring that binary over 2021-04-04 16:13:58 falkon is google free chromium, but u need qt 2021-04-04 16:16:02 tehcloud what pm issues u refer to? 2021-04-04 16:16:41 the upstream developers have not always been enthusiastic about distros packaging it ;) 2021-04-04 16:18:27 lotsa ppl think pm ceo is an ass i remember the openbsd said 4q to palemoon 2021-04-04 16:19:14 cause he is anal about official branding and whatnot 2021-04-04 16:21:32 but i like how the fork grew with its own extensions and all 2021-04-04 16:22:43 the MPL does not grant the ability to use trademarks 2021-04-04 16:23:21 I use the MPL-2.0 for some of my projects because it was the best "weak copyleft" license I could find, and I don't care about trademarks... but I might alter my license if I somehow stumble upon a similar "weak copyleft" license 2021-04-04 16:24:31 the trademark thing is likely why you don't see palemoon available widely in most distros 2021-04-04 16:24:50 ur right on point 2021-04-04 16:25:43 but to a user like me it's all semantics 2021-04-04 16:26:29 i don't care if it 2021-04-04 16:27:03 if it's called newmoon or mypal for winxp lol i know it's palemoon 2021-04-04 16:27:45 is that you volunteering to build the "newmoon" Alpine package ;)? 2021-04-04 16:28:17 no, where? 2021-04-04 16:28:33 i wish i could if i had the knowhow 2021-04-04 16:28:50 it's just a browser ;) 2021-04-04 16:29:14 probably shitty one :/ 2021-04-04 16:29:58 to each his own, im like it 2021-04-04 16:31:25 I had to switch back to X from wayland mostly bc. shitty firefox did weird things 2021-04-04 16:31:44 or fucking gtk crap 2021-04-04 16:32:30 firefox shitty that i agree upon 2021-04-04 16:32:45 well shitty but the best from the crap :/ 2021-04-04 16:32:52 no thx 2021-04-04 16:33:19 the win version takes up almost 1gig 2021-04-04 16:33:27 webkit based are unusable on my machine and chromwhatever will never touch 2021-04-04 16:33:57 spyware not a browser 2021-04-04 16:34:32 I like shit like lynx, but ... these are usable like browsing web with curl :/ 2021-04-04 16:34:35 old hardware? 2021-04-04 16:34:47 thinkpad t420 2021-04-04 16:35:04 webkig is unusable somehow 2021-04-04 16:35:07 webkit 2021-04-04 16:35:37 i5 not bad 2021-04-04 16:35:46 what's that? 2021-04-04 16:36:05 thinkpad t420 intel i5? 2021-04-04 16:36:27 i use sandy bridge chromium and falkon work in alpine 2021-04-04 16:36:41 oh yes, 2520 2021-04-04 16:37:09 don't know, but webkit are like 5 times slower, evern utube shit freezing 2021-04-04 16:37:44 with firefox I can open 30 porn tabs without an issue, it pop ups as fast as systemd features ;) 2021-04-04 16:38:04 compared to my machine urs is a gaming rig 2021-04-04 16:38:32 g2g catch u ltr helby 2021-04-04 16:38:55 what? 2021-04-04 16:39:01 oh 2021-04-04 16:39:17 cya gl with them porn tabs 2021-04-04 16:39:23 yes, see u next week when u make some progress with building browser :D 2021-04-04 16:39:54 gonna write from scratch 2021-04-04 16:44:48 cool, then see u next life 2021-04-04 17:39:28 thanks ikke, very helpful. *gets edge, blows rust off memories of having been here a while ago.* 2021-04-04 17:50:21 any way to downgrade from edge to 3.13? 2021-04-04 17:56:33 changing repositories? 2021-04-04 17:56:42 and reinstall packages maybe 2021-04-04 18:07:56 yes, after changing repositories: apk upgrade -U --available 2021-04-04 18:52:20 love that simplicity in alpine, not much bullshit around :/ 2021-04-05 00:40:06 Really glad that I finally got around to checking out alpinelinux and realizing how absolutely based it is 2021-04-05 00:48:43 indeed, it's based on musl 2021-04-05 00:48:48 and a few other things 2021-04-05 02:52:29 What kde package is needed to get wallpaper configuration in the settings? 2021-04-05 02:53:01 (I've installed kde-applications) 2021-04-05 10:26:26 Hi, does apline linux come with a graphical interface with the install on terminal only? 2021-04-05 10:26:31 *alpine 2021-04-05 10:27:52 Newb101: text mode only 2021-04-05 10:28:40 ok thanks 2021-04-05 11:16:49 hey all :) I need to down grade my php7 to php5. Dont ask, too painful why. Anyhow, pHP5 wors fine i get a nice phpinfo but I cant figure out how to add PDO driver. the core PDO module is there... but I still get Fatal error: Undefined class constant 'MYSQL_ATTR_INIT_COMMAND'. What am I missing? 2021-04-05 11:18:19 https://stackoverflow.com/questions/2424343/undefined-class-constant-mysql-attr-init-command-with-pdo 2021-04-05 11:21:12 OK figured it out. It's because I cant see sh*t anymore after my stroke :( i typed php5-pdo-mysql instead of php5-pdo_mysql. Sigh. 2021-04-05 11:27:00 have a fab day! 2021-04-05 11:34:05 Good morning all. I have a question. Do the GPIO pins work on the rpi3 under alpinelinux? 2021-04-05 11:34:40 I need a serial port to my router and I was thinking of using my rpi3 as serial console 2021-04-05 12:52:21 yes, it is standard linux 2021-04-05 12:52:44 although you will probably have much better results using hardware uart, not bitbanging 2021-04-05 12:52:51 also beware of voltage 2021-04-05 13:49:49 Hello71, thanks for letting me know 2021-04-05 21:28:18 How do i rebase a MR? I do 'git rebase origin/master', fix the conflict and push it. Everything looks ok locally (commit is on top of master, different SHA) but on Gitlab nothing has changed. The MR remains outdated and the commit still has the same SHA. 2021-04-05 21:32:33 did you use the --force ? 2021-04-05 21:34:16 yes, but its not even necessary. It says 'Everything up to date' now (after pushed once). 2021-04-05 21:34:32 but on remote nothing changed. 2021-04-05 21:35:57 When i start cleanly from origin/master and overwrite the files manually and do a force push it works. 2021-04-05 21:36:14 But that can't be the way to do. 2021-04-05 21:37:40 It looks like a bug or as the MR would be 'protected' against force pushes (deletions). 2021-04-05 21:44:45 what push command are you running exactly? 2021-04-05 21:48:52 git push -u bluemax bluemax/php7-initd-umask-fix -f 2021-04-05 21:52:36 I'm no expert, try 'git push -u bluemax php7-initd-umask-fix -f'? 2021-04-05 21:55:38 i do 'git push --force mps' 2021-04-05 22:05:42 Indeed, 'git push -u bluemax php7-initd-umask-fix -f' works. 2021-04-05 22:06:43 Maybe its because my local branch names are identical to the remote ones. Might lead to clashes or confusion. 2021-04-05 22:07:05 thx 2021-04-05 22:07:42 🖒 2021-04-05 23:03:50 Hello, I'm having troubles with apk and would appreciate if someone would help me: APKBUILD https://paste.vpsfree.cz/ckTQgFrT/raw/ , however when I install the package, I do not see /test-chain1-foo file and apk info -L gives me an empty listing. There is not a single warning during the build nor install process, so I'm not sure I understand what is going on. Any pointers? 2021-04-05 23:36:19 /quit/quit 2021-04-06 04:49:22 so, i have asked this before 2021-04-06 04:49:30 and seem to have forgotten 2021-04-06 04:49:37 but is there an input group 2021-04-06 04:49:43 i don't see it in the wiki 2021-04-06 04:50:03 when i start my desktop, i can't move the mouse or issue keyboard commands 2021-04-06 04:52:51 oh, i see there are scripts 2021-04-06 04:53:12 there is 2021-04-06 04:53:22 getent groups 2021-04-06 04:53:24 oh, udev wasn't running 2021-04-06 04:53:40 ok, i really like the scripts. nice touch 2021-04-06 04:53:53 ikke: thanks. i'll return if i take a wrong turn 2021-04-06 04:53:58 o/ 2021-04-06 05:03:58 unless the setup script expects me to run X server as root, it actually made things worse 2021-04-06 05:04:37 setup-xorg-base causes permissions issues for startx 2021-04-06 05:14:43 it does almost nothing 2021-04-06 05:15:26 just install xorg-server xf86-input-libinput eudev and run setup-udev 2021-04-06 05:16:18 regarding mouse and keyboard u probably just missed that input shit 2021-04-06 05:16:42 i added myself to the group 2021-04-06 05:17:05 not sure about groups, I usually always add myself to audio video input 2021-04-06 05:17:18 you've helped though, it's not giving "permission denied anymore" 2021-04-06 05:17:27 " XD 2021-04-06 05:18:36 yeah, audio, video and input are added, along with wheel 2021-04-06 05:19:10 so all run okay? 2021-04-06 05:21:05 segfault at 0x0 2021-04-06 05:21:24 I might have to run through the setup again 2021-04-06 05:22:50 also some wm and .xinitrc can help ;) 2021-04-06 05:23:02 yeah, i have that 2021-04-06 05:23:08 but i'm upgrading what i have and trying again 2021-04-06 05:27:52 it's really mad 2021-04-06 05:30:04 what u expect from xorg? 2021-04-06 05:30:12 it just works for me 2021-04-06 05:34:20 u probably did not fuck up your system before like freq 2021-04-06 06:02:26 ikke: you switched to alpine on desktop? 2021-04-06 06:21:23 aha, liar! 2021-04-06 06:23:14 helby: I wouldn't say so 2021-04-06 06:24:57 helby: can you understand that here are people with different cultural background 2021-04-06 06:26:09 jaja and genders and pronouns ;) 2021-04-06 06:28:17 not these, some cultures appreciate kindness 2021-04-06 06:31:20 No, but I do have vms with alpine 2021-04-06 06:31:48 that does not count ;) 2021-04-06 06:32:15 Why not? 2021-04-06 09:18:32 IDK why mostly only US citizens swear and use vulgar words in IRC. 2021-04-06 09:19:02 Kind of like they did not appreciate the language at all. 2021-04-06 09:19:46 they are intentionally 'educated' to destroy culture 2021-04-06 09:20:06 no all of them, ofc 2021-04-06 09:54:21 https://www.youtube.com/watch?v=PvY6YQH9kE4 2021-04-06 10:00:25 are there transcript of their talk somewhere? I can't understand what they say 2021-04-06 10:07:16 "I must apologise for Wimp Lo, he is an idiot, we have purpously trained him wrong, as a joke." 2021-04-06 10:08:57 I must have thought of that phrase as you mentioned "...they are intentionally 'educated' to destroy culture..." 2021-04-06 10:12:14 aaah, thanks 2021-04-06 10:14:02 happily Wimp Lo didn't had 'big gun', which is not true for ... 2021-04-06 10:39:54 hi, i am in the process of building a sat>ip server with tvheadend and an dvb-s receiver. In order to get this working i need to install the firmware-file dvb-usb-tt-s2400-01.fw which can be found online (for example from OpenELEC's github). But as i installed in diskless mode i now stubled on the problem on adding the firmware-file to /lib/firmware. What would be the best option to add the file? a custom apt-Package with only one file? 2021-04-06 10:49:08 jkoan: I think that you should build your own iso/moodloop 2021-04-06 10:50:01 is there a howto for that? 2021-04-06 10:50:16 it seems that I changed sice last time I did 2021-04-06 10:50:24 https://github.com/alpinelinux/alpine-make-vm-image 2021-04-06 10:53:29 uhM 2021-04-06 10:57:12 "For recent Alpine versions please use the mkimage.sh script in the scripts dir in root of the aports tree." 2021-04-06 10:59:12 https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/scripts/mkimage.sh 2021-04-06 10:59:59 jkoan: I think that that is the right way now, you should clone arpots repository, have a working abuild setup 2021-04-06 11:00:14 and use some custom mkimage.sh script 2021-04-06 11:01:07 hi all 2021-04-06 11:01:24 I say "now" because last time I build a custom iso I used https://github.com/alpinelinux/alpine-iso 2021-04-06 11:01:40 is there anything i could do to find out if somebody is working on upgrading https://pkgs.alpinelinux.org/package/edge/community/x86_64/influxdb to 2.0? 2021-04-06 11:03:36 uhM, last merge is 4 months old 2021-04-06 11:05:07 that was 1 month after 2.0 was released 2021-04-06 11:05:27 i'm not trying to judge here mate 2021-04-06 11:05:54 https://pkgs.alpinelinux.org/flagged?origin=influxdb 2021-04-06 11:05:59 just trying to figure out if this is something that happens when it happens, nothing to be done about, or if it's something that could happen sooner e.g. if one were to contribute something 2021-04-06 11:06:17 its flagged since 14/03, but I says there is no Maintainer 2021-04-06 11:06:32 okay, so there's some kind of flagging system 2021-04-06 11:06:50 hmm and someone also provided a link on how to upgrade 2021-04-06 11:06:58 o there is a commit from 1 month ago 2021-04-06 11:07:04 https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/community/influxdb 2021-04-06 11:07:18 aha and there's a gitlab ... getting closer now 2021-04-06 11:08:11 ACTION signs up 2021-04-06 11:08:53 if it is very urgent for you could try to upgrade the APKBUILD and request a merge 2021-04-06 11:09:02 no, not very urgent 2021-04-06 11:09:16 just trying to figure out how things work 2021-04-06 11:10:06 hehe I'm not any reference just trying to help 2021-04-06 11:10:10 upgrade APKBUILD ... so you're saying if i download https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/community/influxdb then i can try to make it work myself and if i happen to succeed i could also try to get my changes merged? 2021-04-06 11:10:51 I'm not sure but I suppose that nobody worries a good merge request and since the package has no official maintainer... 2021-04-06 11:11:17 ok 2021-04-06 11:11:29 well you should clone whole aports repository 2021-04-06 11:11:34 ACTION goes to read and try to understand APKBUILD file 2021-04-06 11:11:46 and setup an abuild enviroemnt so you can build packages using APKBUILD files 2021-04-06 11:11:50 aha 2021-04-06 11:12:28 if you are docker familiarized I would use dabuild which avoids installing libs and build dependencies on the host 2021-04-06 11:12:35 3bc01ae3a15cafd84e51081e19ea7ae5a3b2e177 2021-04-06 11:13:30 oh, thanks ikke 2021-04-06 11:14:10 So apparently influxdb v2 is more challenging 2021-04-06 11:15:26 ouch 2021-04-06 11:15:32 if it needs docker i'm out 2021-04-06 11:15:53 docker is not required, just a way to build packages in an alpine environment 2021-04-06 11:16:01 You could also setup a VM or something else 2021-04-06 11:16:29 hmm trying out this abuild thing and so far it seems the build is going well 2021-04-06 11:16:45 abuild is a shellscript that interacts with APKBUILDs to build packages 2021-04-06 11:17:31 VM would only be required if i'm unwilling to installer compiler and so on on the bare metal? 2021-04-06 11:17:49 If you run alpine locally, then A VM is not necessary 2021-04-06 11:17:56 k 2021-04-06 11:17:58 the only challenge is to have a clean build environment 2021-04-06 11:18:05 so that you don't miss any dependencies 2021-04-06 11:18:08 hmm 2021-04-06 11:18:25 there is abuil rootbld, which builds aports in a clean chroot 2021-04-06 11:18:26 let's try with this dirty one first, just to get a feel for what he meant by "more challenging" 2021-04-06 11:18:34 nod 2021-04-06 11:20:33 are there some kind of criteria for font packages being ttf-* or font-*? 2021-04-06 11:23:17 c7s: no, but would be nice to use font-* 2021-04-06 11:24:03 maybe font-$name-$type 2021-04-06 11:32:14 right, thanks 2021-04-06 11:45:14 ikke: is there a way to tell abuild to keep things like go installed and now install them again each time it is run? 2021-04-06 11:46:10 abuild -k 2021-04-06 11:47:00 uhM or -K 2021-04-06 11:49:28 thx! 2021-04-06 12:20:49 wow they added a LOT of things to their build system between 1.8 to 2.0 2021-04-06 12:21:17 and "challenging" so far means "not all deps are documented, figure them out yourself" 2021-04-06 12:22:08 and it's like "cool, ppl on twitter have been talking a lot about hipster tool x, let's use that, too" and "woah i always wanted to learn rust, let's put some rust in there" 2021-04-06 12:39:54 funky. we're ****compiling**** something here ... like, it's NOT a binary, it's NOT yet specific to some architecture or libc. and they INSIST on giving an explicit target. the thing compiles just fine if you leave the target out. *sigh* 2021-04-06 12:42:51 okay, got past that, too 2021-04-06 12:43:00 i think i've reached the point where the old maintainer gave up 2021-04-06 12:43:02 /usr/lib/gcc/x86_64-alpine-linux-musl/10.2.1/../../../../x86_64-alpine-linux-musl/bin/ld: /home/jrs/.cache/go-build/pkgconfig/github.com/influxdata/flux@v0.104.0/libflux/lib/linux_amd64/libflux-c558c2c8240a9302f16ae7c3df7b7582944bce914b268fd4b9cce608df811690.a(std-b1047a2788193be4.std.b9wtcb8n-cgu.0.rcgu.o): in function `std::sys::unix::net::on_resolver_failure': 2021-04-06 12:43:04 /rustc/2fd73fabe469357a12c2c974c140f67e7cdd76d0//library/std/src/sys/unix/net.rs:453: undefined reference to `__res_init' 2021-04-06 12:46:07 so in your experience is this something solveable, perhaps with a patch to rust's stdlib? 2021-04-06 12:48:27 seems https://doc.rust-lang.org/beta/src/std/sys/unix/net.rs.html#447 is the culprit fwiw 2021-04-06 12:49:18 Seems like you're trying to build for glibc? 2021-04-06 12:49:36 not sure yet what the influxdb build system is doing 2021-04-06 12:53:56 (going into the cached build folder for that lib and manually doing "cargo build --target x86_64-unknown-linux-musl" works fine) 2021-04-06 13:47:10 What did I miss? After installing zsh and ran it I see: \h:\w\$ 2021-04-06 13:58:26 I missed alpine@alpine> 2021-04-06 13:58:33 grml-zsh-config 2021-04-06 14:16:51 rapha: i wanted to take over maintainership of influx but i highly disagree with the changes they did for 2.0 so idk what to do 2021-04-06 14:17:11 i wonder if there's a way to keep the existing 1.x and 2.x upstream 2021-04-06 14:17:24 without splitting it into influxdb and influxdb1x or something 2021-04-06 14:17:27 like python 2021-04-06 14:20:05 plus the new build system for influx is a mess 2021-04-06 14:20:10 just like telegram-desktop 2021-04-06 14:24:41 "a mess" about covers it, yes 2021-04-06 14:25:06 what's wrong with "influxdb1.8" and "influxdb", though? 2021-04-06 14:30:35 im partial to keeping legacy as "influxdb" and making the new version "influxdb2" 2021-04-06 14:30:49 i'd be happy with either 2021-04-06 14:31:20 My logic is its the path of least resistance 2021-04-06 14:31:26 lol 2021-04-06 14:31:52 No potential for existing things to break due to path changes 2021-04-06 14:32:18 ACTION just wants the new web UI :P 2021-04-06 14:32:24 lol 2021-04-06 14:38:17 (what i'd like even more though is if there was something like influxdb and something like it's web ui, but they were two separate, light-weight projects. alas, can't have everything in life) 2021-04-06 17:32:51 The error encountered was: 2021-04-06 17:32:53 No module named 'jinja2' 2021-04-06 17:32:56 Please search for the python3 version of jinja2 in your distributions packages, or see https://github.com/qutebrowser/qutebrowser/blob/master/doc/install.asciidoc 2021-04-06 17:32:59 If you installed a qutebrowser package for your distribution, please report this as a bug. 2021-04-06 17:33:02 jinja2 is installed 2021-04-06 17:36:15 /b 6 2021-04-06 17:48:36 Is there a flag for apk or a configuration that would say "install -doc automatically, if available"? 2021-04-06 17:49:12 telmich: apk add docs 2021-04-06 17:51:07 mps: you must be joking, that is way to easy even, for any kind of Linux distro. Really? That easy? 2021-04-06 17:51:41 :) 2021-04-06 17:51:55 I'm still be amazed by the simplicity of alpine, ever and every again 2021-04-06 17:52:15 yes, it is meta package which means install doc for installed pkgs 2021-04-06 17:55:22 Hmm...how does it actually work? I'm currently looking at the APKBUILD and it looks like it only depends on mandoc 2021-04-06 17:56:58 mandoc is needed for -doc subpkgs 2021-04-06 17:57:21 telmich: it uses an install_if in the -doc subpackge 2021-04-06 17:57:40 ahhh, nice 2021-04-06 17:57:41 that is proper explanation 2021-04-06 17:57:45 telmich: https://gitlab.alpinelinux.org/alpine/abuild/-/blob/master/abuild.in#L1778 2021-04-06 17:59:44 I'm trying to upgrade community/telegram-desktop which has no maitainer, could I try to be maitainer of it? 2021-04-06 17:59:51 donoban: sure 2021-04-06 18:00:12 great 2021-04-06 18:09:31 how do i use the newest version of a package if there are two options 2021-04-06 18:10:09 What do you mean with 2 options? 2021-04-06 18:10:23 older and newer, based on repos selected 2021-04-06 18:10:39 apk should always prefer the newer, unless there is a version constraint 2021-04-06 18:11:08 it's quite weird that qutebrowser can't find jinja2, but it's instaled 2021-04-06 18:11:25 freq_: might be something different going on 2021-04-06 18:14:48 is it possible you have py3.9 and jinja is installed for py3 2021-04-06 18:14:55 is it possible you have py3.9 and jinja is installed for py3.8 2021-04-06 18:16:17 could be 2021-04-06 18:16:28 freq_: apk info -L py3-jinja2 2021-04-06 18:16:34 apk version python3 2021-04-06 18:16:53 Installed: Available: 2021-04-06 18:16:56 python3-3.8.8-r0 < 3.9.4-r0 2021-04-06 18:17:06 right, you need to upgrade python 2021-04-06 18:17:11 apk add -u python3 2021-04-06 18:17:19 thanks lol 2021-04-06 18:17:33 python3.9 has just been pushed yesterday 2021-04-06 18:18:21 mean while i'm trying to install librewolf.. 2021-04-06 18:18:33 i can stop that and re-test 2021-04-06 18:19:23 what about to update whole system before fuckering ? 2021-04-06 18:19:32 also don't fight with browsers! you will lose ;) 2021-04-06 18:22:32 it's updated and upgraded 2021-04-06 18:22:59 freq_: running apk upgrade --available is recommended as well to get the latest version of everything 2021-04-06 18:23:33 well update and upgrade will upgrade everything anyway 2021-04-06 18:23:51 without nonsense like apk ad -u python3 :/ 2021-04-06 18:24:58 btw upgrade --available will 'upgrade' your builded packages 2021-04-06 18:49:46 is qutebrowser broken for anyone else? 2021-04-06 18:54:53 freq: yes 2021-04-06 18:55:11 ok ty 2021-04-06 18:58:43 u don't listen! stop fighting with browsers :/ 2021-04-06 19:13:35 Has anyone ran into this while pxe booting? https://usercontent.irccloud-cdn.com/file/lsacEf7P/image.png 2021-04-06 19:14:17 Seems the standard iso works 2021-04-06 19:15:06 aaa801: afaik, the netboot does not have any filesystem modules included 2021-04-06 19:15:33 This setup has worked on all of our other hardware, its just these new nucs giving us this issue oddly 2021-04-06 19:15:35 aaa801: I would expect it to run in run-from-ram mode 2021-04-06 19:16:56 Our pxe is nothing special https://pastebin.ubuntu.com/p/7tzbTqGXZ8/ 2021-04-06 19:17:16 Guessing it could be possibly due to the kernel versions? 2021-04-06 19:17:42 Not sure 2021-04-06 19:17:45 Aha im an idiot 2021-04-06 19:17:51 i forgot how semver works.. 2021-04-06 19:18:19 Time to bump this to 3.13 :') 2021-04-06 21:12:18 python3 -m pip3 show pip 2021-04-06 21:12:28 No module named pip 2021-04-06 21:13:35 apk add py3-pip 2021-04-06 21:13:42 Just run: `apk add --no-cache python3-dev py3-pip` and pip/pip3 are in the path 2021-04-06 21:13:59 using last alpine:edge docker image 2021-04-06 21:14:36 ah, edge 2021-04-06 21:14:47 python is in the process of being upgraded to 3.9 2021-04-06 21:15:37 ok 2021-04-06 21:15:43 thanks 2021-04-06 21:16:10 It's in community, so the one that is rebuilt against 3.9 hasn't been uploaded yet 2021-04-06 21:16:34 https://pkgs.alpinelinux.org/contents?branch=edge&name=py3-pip&arch=x86_64&repo=community 2021-04-06 21:16:46 yes 2021-04-06 21:16:51 that's the 3.8 version 2021-04-06 21:17:13 so just wait and see ;) 2021-04-06 21:17:46 Many thanks for the quick support ! 2021-04-06 21:17:48 The issue is that we need to fix quite some packages 2021-04-06 21:17:55 so that is holding it up a bit 2021-04-06 21:19:09 if you don't need edge, :3.13 should be working 2021-04-06 22:19:40 :) 2021-04-06 22:52:07 Hm so no luck on 3.13 pxe either, it seems that modloop is never loaded, only the kernel and initramfs, 2021-04-06 22:53:00 wondering if the kernel config doesn't have the module for this system included possibly, as im basically out of ideas 2021-04-06 22:55:22 can someone explain what actually happend to python on edge? I don't get it 2021-04-06 22:58:46 really makes no sense.. booting 3.13.4 iso works fine and picks up the network devices, just all goes to hell when attempting to use the pxe... 2021-04-06 23:12:13 saying that.. it looks like its just e1000e 2021-04-06 23:33:17 i tired installing filerun on alpine 2021-04-06 23:33:24 installed all php stuff :) 2021-04-06 23:33:27 too much even :D 2021-04-06 23:33:50 php7 unzip.php said conflicts about gmagick vs imagick :) 2021-04-06 23:33:55 removed gmagick 2021-04-06 23:34:11 then cacti-php7 cause php-fpm7 wouldnt run :) 2021-04-06 23:34:23 then filerun index.php complained about ioncube :) 2021-04-06 23:34:43 at that point i shut down the vm and the proxmox (also vm) above it, and poured myself a beer :D 2021-04-06 23:34:52 now i realized there is docker filerun :) 2021-04-06 23:34:56 gonna try that next 2021-04-07 00:03:54 palemoon puts chroimium to some serious shame on 3.13 2021-04-07 00:04:55 was able to bring the pm binary i compiled in 3.8 over to 3.13 haha 2021-04-07 00:06:36 wonder if the latest pm version would compile 2021-04-07 00:07:46 but the 2 year old pm release works so fast on alpine and requires fewer deps 2021-04-07 00:33:13 b 12 2021-04-07 00:33:19 wops 2021-04-07 08:59:31 If I put Alpine Extended on a USB, can I then install Alpine onto that same USB while booted from it? 2021-04-07 08:59:33 Since it runs from RAM 2021-04-07 09:06:29 Newbyte: It is posible, but I don't know exact steps. something must be unmounted. but what? 2021-04-07 09:07:27 got it, thans 2021-04-07 09:07:28 thanks* 2021-04-07 09:31:38 Hi, it seems that telegram-desktop "depends" on Microsoft GSL (which has MIT license), should I try to package it?Or try to use some alternative implementation? I'm not sure if there is some already in Alpine 2021-04-07 09:40:16 there is a GSL package from GNU but it seems something total different 2021-04-07 10:03:39 donoban: I would love to have signal for alpine though 2021-04-07 10:05:08 hehe I have to try it since I'm using more signal than telegram 2021-04-07 10:24:44 hi! how do i upgrade to python3.9? 2021-04-07 10:26:25 Like any normal upgrade 2021-04-07 10:27:05 you mean. 'apk upgrade' with edge enabled? 2021-04-07 10:27:27 that doesn't work for some reason. so i thought there is something else i have to do 2021-04-07 10:40:20 bgs: apk upgrade --available 2021-04-07 10:40:43 But you might have things installed that explicitly depend on python3.8 2021-04-07 10:49:34 ikke: i tried --available but it didn't work 2021-04-07 10:50:22 apk add 'python3>3.9.0' 2021-04-07 10:52:16 i don't actually mind staying on python3.8, but a significant portion of python based software i run broke yesterday, and the suggested fix in this channel was to upgrade python 2021-04-07 10:52:28 In my case the Python upgrade is being held back by a lot of packages not being rebuild yet 2021-04-07 10:52:45 apk add 'python3>3.9.0' displays a bunch of errors 2021-04-07 10:52:57 ACTION < https://matrix.org/_matrix/media/r0/download/matrix.org/rSdWnlLFDQUfpionDCcfhLoD/message.txt > 2021-04-07 10:53:01 bgs: can you show the errors 2021-04-07 10:53:06 ☝️ huge list of packages holding back 3.9 2021-04-07 10:53:11 PureTryOut[m]2: what arch? 2021-04-07 10:53:19 x86_64 2021-04-07 10:53:21 x86_64 community should be uploaded 2021-04-07 10:53:32 But what about testing 😉 2021-04-07 10:53:43 I guess it's mainly the few packages depending on libpython3.8.so.1 2021-04-07 10:54:05 PureTryOut[m]2: testing is for testing purposes only :P 2021-04-07 10:54:13 Obviously lol 2021-04-07 10:54:25 I'm not complaining, just saying that it's a reason why 3.9 would've been hold back 2021-04-07 10:54:35 yes 2021-04-07 10:54:47 ikke: and you say that, but there are tons of packages in testing that have been there for years, upgraded and all, and that I doubt are ever being moved to community 2021-04-07 10:55:01 yes, hence the ':P' 2021-04-07 10:55:19 I'm aware of how testing is used in practice 2021-04-07 10:55:36 A sad state of afairs imo 😢 2021-04-07 10:55:38 yes 2021-04-07 10:55:50 We should make effort improving this 2021-04-07 10:55:59 Talking about that, fcolista could you move obs-studio to community at some point? 2021-04-07 10:56:07 ikke: yes please! 2021-04-07 11:16:03 ikke: my errors from apk add python3>3.9.0: http://okturing.com/src/10818/body 2021-04-07 11:20:12 it works fine here 2021-04-07 12:01:02 PureTryOut[m]2, sue thing 2021-04-07 12:01:06 *sure 2021-04-07 12:01:30 Yay thanks! 2021-04-07 12:01:31 have you had a chance to test it? 2021-04-07 12:01:44 Yeah I use it quite regurarly 2021-04-07 12:02:00 Works as well as on any other distro afaics 2021-04-07 12:03:24 If I need to add ms-gsl for telegram-desktop, should both go to testing? 2021-04-07 12:04:39 also I would like if there is some problem iwth ms-gsl itself or it's a more linux likely alternative that I should try to use 2021-04-07 12:05:41 "The GSL is a header only library, hence, you can use the functions and types of the library quite easily. The best-known implementation is the one from Microsoft, hosted at Github: Microsoft/GSL. ", uhM, that means that I could just download it for the build process ? 2021-04-07 12:07:42 s/I would like/I would like to know/ 2021-04-07 12:08:42 how we are easily sold self to big corporations :c 2021-04-07 12:10:31 ng doesn't have ideals and principles 2021-04-07 12:13:29 ng? 2021-04-07 12:14:50 I didn't expect to find that dependecy when tried to build telegram :\ 2021-04-07 12:15:23 Not Good, though some thinks it means New Generation :) 2021-04-07 12:16:00 https://aur.archlinux.org/packages/microsoft-gsl-git/ , 12 arch packages depending on it 2021-04-07 12:16:12 apk search telegram-desktop 2021-04-07 12:16:46 uhM, yes, it is outdated and unmaintained 2021-04-07 12:17:19 I don't see any mentions of gsl deps 2021-04-07 12:17:42 I don't know when they added that dependency 2021-04-07 12:17:52 -DDESKTOP_APP_USE_PACKAGED_GSL 2021-04-07 12:18:41 maybe it contain gsl in tarball 2021-04-07 12:18:45 uhM 2021-04-07 12:19:49 I built it yesterday, no problem, then switch to pkgver=2.7.1 and it complained about missing ms-gsl 2021-04-07 12:20:41 medion:~/alpine/aports/community/telegram-desktop/src$ find | grep gsl 2021-04-07 12:20:42 ./tdesktop-2.7.1-full/Telegram/ThirdParty/GSL/include/gsl 2021-04-07 12:21:07 maybe it is a problem with cmake finding it :) 2021-04-07 12:21:27 could be 2021-04-07 12:21:36 I will try again this evening 2021-04-07 12:31:12 Are you trying to upgrade Telegram? 2021-04-07 12:40:44 yes 2021-04-07 12:42:37 have you tried it? I see that you updated it last time 2021-04-07 12:54:08 somehow my "/usr/lib/jvm" directory and therefore "/usr/lib/jvm/default-jvm" symlink is missing. any idea which package provides these files? 2021-04-07 12:55:11 "/usr/bin/java" is now a dead symlink to "../lib/jvm/default-jvm/jre/bin/java". is suspect this broke when updating from 3.12 to 3.13. 2021-04-07 13:06:02 hi, on alpine 3.13.4, libreoffice writer saves fine, but draw fails with error "General Error, General input/output error". I'm running a ramdisk setup and my system is plain 2021-04-07 13:06:20 can someone test so I see is this my system or libreoffice itself ? 2021-04-07 13:06:41 simply run: libreoffice --draw 2021-04-07 13:06:47 and save the document 2021-04-07 13:09:41 I found the solution, it seems if you do "doas apk add libreoffice-draw" it will not pull all needed dependencies 2021-04-07 13:09:49 the program will run and function fine, but wont save! 2021-04-07 13:10:14 doing "apk add libreoffice", will pull all of it and saving will work 2021-04-07 13:42:53 > have you tried it? I see that you updated it last time 2021-04-07 13:42:53 caskd was working on it, so I left it to him. it seems he hasn't got around to actually finalising the update though 2021-04-07 13:43:16 how do those replies come through on IRC btw? 2021-04-07 13:44:02 Newbyte: uh, yeah the build system was hell of a mess 2021-04-07 13:44:20 i tried rewriting the entire build system but then i've reached some weird dependency loops 2021-04-07 13:44:25 so i gave up 2021-04-07 13:44:29 rewriting it..? 2021-04-07 13:44:34 yeah 2021-04-07 13:44:39 to have it more uniform 2021-04-07 13:44:44 and not the mess that is now 2021-04-07 13:44:47 ^^' 2021-04-07 13:45:09 oh, I hope it was an easy task, I've never fixed an APKBUILD :D 2021-04-07 13:45:16 if nothing else testifies that it is a mess it's that upstream replaced their build instructions with "use Docker" 2021-04-07 13:45:54 caskd: could you share your APKBUILD rewrite? 2021-04-07 13:46:12 I think he means more than APKBUILD rewrite 2021-04-07 13:46:20 it wasn't rewriting the apkbuild, it was the entire cmake files 2021-04-07 13:46:52 i can share what i had so far but i didn't get far enough as telegram api depended on mtproto and back 2021-04-07 13:46:58 should i? 2021-04-07 13:47:07 mtproto depends on tdesktop? 2021-04-07 13:47:15 on some files, yeah 2021-04-07 13:47:23 seems like it at least from the source files 2021-04-07 13:48:33 well I have the intention to ugprade it but it seems that it will be probably too difficult for me 2021-04-07 13:48:58 you're gonna have to fix some things and the dependency spaghetti is getting worse every new version 2021-04-07 13:49:19 i kinda decided i won't really do a lot anymore until they fix that first 2021-04-07 13:49:34 have you tried to package signal? 2021-04-07 13:50:45 i don't use signal so no 2021-04-07 13:52:21 https://git.redxen.eu/caskd/tdesktop-new-cmake 2021-04-07 13:52:26 here's what i've done so far 2021-04-07 13:53:59 buy windows 2021-04-07 13:54:25 ok I will take a look thanks 2021-04-07 14:02:23 Hi, I have a fresh install and added networkmanager, disabled and also removed references to eth0 in /etc/interfaces, stopped networking daemon, and still I get a message that the device is "strictly unmanaged" in nmtui. "# nmcli d" reports unmanaged state. 2021-04-07 14:02:47 8> I can manually bring the interface up, configure IP, Netmask Route and DNS to get a working connection. 2021-04-07 14:03:18 "# nmcli n" reports enabled 2021-04-07 14:16:18 mm28: I've run into that before usually adding an [ifupdown] managed=true to /etc/NetworkManager/NetworkManager.conf helps 2021-04-07 14:28:16 wsinatra, I added that to my vm, and additionally it was already in my raspberry pi which is my target system, would the daemon /etc/init.d/networking conflict, or is it required? 2021-04-07 14:36:19 It does not make a difference either way, but I see on my debian based vm that networking.service shows active. 2021-04-07 14:46:04 mm28: networking shouldn't conflict. You probably need it to have it come up regardless. Uusually as a sysinit service 2021-04-07 14:46:11 ah he quit 2021-04-07 14:52:23 and now he's back! 2021-04-07 15:31:48 Alright. Hello 2021-04-07 15:32:24 I was wondering what happened to the xf86-input-keyboard and xf86-input-mouse packages 2021-04-07 15:32:47 jana0000: what I understood is that they are deprecated and not maintained 2021-04-07 15:32:52 and not necessary 2021-04-07 15:33:17 90fc54683ff5d8890f2dba76343d976a2196466b 2021-04-07 15:33:25 #12362 2021-04-07 15:33:39 thats a nice bot 2021-04-07 15:36:23 It's very useful :) 2021-04-07 15:37:23 ACTION pets algitbot 2021-04-07 15:37:26 good bot 2021-04-07 15:40:37 expanded storage i see 2021-04-07 15:40:54 https://wiki.alpinelinux.org/wiki/Draft_Release_Notes_for_Alpine_3.14.0 2021-04-07 15:41:27 oh nice 2021-04-07 18:18:10 hooray qutebrowser is updated 2021-04-07 19:17:29 that's not bad really 2021-04-07 19:17:33 one day dip 2021-04-07 19:17:37 i'll stick with alpine 2021-04-07 19:26:37 #61 2021-04-07 19:27:03 :\, I suppose that algitbot only works for aports 2021-04-07 19:32:26 I was yesterday trying to do this https://gitlab.alpinelinux.org/alpine/docker-abuild/-/issues/61 , I think that the best "naming" could be using reuse tag for current image, when run dabuild with DABUILD_REUSE=true looks for alpinelinux/docker-abuild:reuse , if exists it uses it as tag, after running container it commits it to the new reuse tag 2021-04-07 19:39:21 hi. How to do apk update && upgrade with --no-cache or something? 2021-04-07 19:39:48 FreeBDSM: what do you want to achieve? 2021-04-07 19:40:05 ikke: less files on file system 2021-04-07 19:40:23 apk upgrade -U --no-cache 2021-04-07 19:40:44 ikke: do I need to to `apk update` prior to it and after it? 2021-04-07 19:40:51 to do* 2021-04-07 19:41:08 -U is short for --update 2021-04-07 19:41:36 in any case, apk upgrade will automatically do an update if there are no cached repos 2021-04-07 19:48:29 ikke: how to tell it to ignore certs? 2021-04-07 19:48:39 --alow-untrust 2021-04-07 19:48:46 --allow-untrusted 2021-04-07 19:49:25 ikke: doesn't work for upgrade: `apk -U --no-cache --allow-untrusted upgrade` 2021-04-07 19:54:37 I'm kind of stuck: I need ca-certificates package to be able to install certificates to the system so that I could use apk with our corp proxy of upstream repo to be able to use apk without --allow-untrusted 2021-04-07 19:55:26 and you can only use https? 2021-04-07 19:55:39 yes 2021-04-07 19:56:48 I wish --allow-untrusted would work 2021-04-07 19:56:58 yeah, that's more for untrusted packages 2021-04-07 19:57:04 not for https 2021-04-07 19:58:02 was there an announcement somewhere that python is getting upgraded and that edge will be broken in the meantime? 2021-04-07 19:58:14 no 2021-04-07 19:58:25 we don't announce these kinds of things 2021-04-07 19:59:26 bgs: edge is constantly updated and can break for all sorts of reasons 2021-04-07 19:59:49 ikke: is there a way to emulate apk fetch via wget? 2021-04-07 20:00:01 I could then wget ca-certificates 2021-04-07 20:00:06 just download the package 2021-04-07 20:00:20 where from? I want to do this from apk repo 2021-04-07 20:00:22 wget https://mirror/pkg.apk 2021-04-07 20:01:00 https://dl-cdn.alpinelinux.org/alpine/v3.13/main/x86_64/ca-certificates-20191127-r5.apk 2021-04-07 20:01:24 https://dl-cdn.alpinelinux.org/alpine/v3.13/main/x86_64/ 2021-04-07 20:01:42 The packages are listed there 2021-04-07 20:01:46 ikke: that's nice! 2021-04-07 20:01:49 ikke: i am aware of that. but i didn't expect this kind of breakage to be within expectations for edge 2021-04-07 20:02:04 let's see if I could grep it to get link to the correct .apk 2021-04-07 20:04:43 weird, I wget'ed similar page (but of our mirror of your repo) and it saved index.html with text 'This apk proxy repostirory is not directly browseable at this URL.' 2021-04-07 20:04:44 :( 2021-04-07 20:05:04 FreeBDSM: it depends how the mirror is setup 2021-04-07 20:05:15 apparently they disabled index 2021-04-07 20:05:30 they probably did :( 2021-04-07 20:05:30 (or did not explicitly enable it) 2021-04-07 20:05:40 But if you know the package name, you should be able to wget it 2021-04-07 20:05:41 it's configured on nexus 2021-04-07 20:05:51 ikke: what if it changes over time? 2021-04-07 20:06:08 I wish there were aliases like ${pkg_name}-latest.apk or something 2021-04-07 20:06:15 FreeBDSM: You'd have to download the APKINDEX file and extract the package name from there 2021-04-07 20:06:21 there is not 2021-04-07 20:06:29 is there some standard way for creating a root fs snapshot before doing 'apk upgrade'? 2021-04-07 20:06:55 donoban: not something specific for alpine 2021-04-07 20:07:06 things like zfs / btrfs / lvm snapshots could work 2021-04-07 20:07:46 my plea: please do not ruin the image of alpine by getting politically correct 2021-04-07 20:07:55 well, APKINDEX is an archive with APKINDEX file inside and that file is of weird format 2021-04-07 20:08:06 and some way for executing some command after upgrading? I'm using btrfs, I think that I could be useful running edge 2021-04-07 20:08:10 I grep'ed ca-certificates against it 2021-04-07 20:08:24 if something gets totally broken you can just reboot and change subvol id 2021-04-07 20:08:27 and it returned some lines like P:ca-certificates-cacert 2021-04-07 20:08:32 yes 2021-04-07 20:08:49 The latter denotes a specific field 2021-04-07 20:09:07 each package is separated by an empty line 2021-04-07 20:09:30 oh wow 2021-04-07 20:09:35 yeah, just realized that 2021-04-07 20:09:45 so I see 'V:20191127-r2' 2021-04-07 20:09:53 apk-tools 3.0 will have a different format 2021-04-07 20:09:53 A:x86_64 2021-04-07 20:10:05 oh, so it's gonna break soon, awesome :D 2021-04-07 20:10:14 Not sure how soon it's ready 2021-04-07 20:10:23 I suppose that just a dayly automated snapshost should be enough 2021-04-07 20:11:00 FreeBDSM: but it's going to be a binary format, so less easy to manually parse 2021-04-07 20:11:45 I wish apk then could be used like `apk genlinkstoapk ca-certificates --allow-untrusted` that's yield like 'https://dl-cdn.alpinelinux.org/alpine/v3.13/main/x86_64/ca-certificates-20191127-r5.apk' 2021-04-07 20:11:53 (but with my mirror, ofc) 2021-04-07 20:12:33 but apk would need to be able to download the APKINDEX first 2021-04-07 20:12:34 but, actually, it is just weird that there's no '--no-check-certificate' for apk 2021-04-07 20:12:56 ikke: yeah, I suggested a stupid thing. 2021-04-07 20:13:10 I think it just needs --no-check-certificate, just like wget has 2021-04-07 20:13:35 wondering, why do you need ca-certificates? 2021-04-07 20:13:58 ikke: to install certificates into the system to be able to use apk properly 2021-04-07 20:14:16 https works out of the box without ca-certificates as well 2021-04-07 20:14:48 we don't have access to your repos, we only have access to our local mirror of your repo 2021-04-07 20:15:37 yes, I understood that 2021-04-07 20:15:50 then what do you mean by 'https works'? 2021-04-07 20:15:59 I need to install packages from alpine mirror 2021-04-07 20:16:37 FreeBDSM: apk-tools depends on ca-certificates-bundle 2021-04-07 20:17:04 so that should already be installed 2021-04-07 20:17:32 https://pkgs.alpinelinux.org/contents?file=&path=&name=ca-certificates-bundle&branch=edge&repo=main&arch=x86_64 2021-04-07 20:17:53 so that's why I wonder why you need ca-certificates 2021-04-07 20:17:58 ikke: then maybe I just don't know how to use it properly. I have 2 .cer files, how to import them into system? 2021-04-07 20:18:21 aha, you want to add custom root certs 2021-04-07 20:18:37 usually I put them to /usr/local/share/ca-certificates/ and then ran `update-ca-certificates` 2021-04-07 20:18:39 technically, you could just append them to the existing list 2021-04-07 20:18:52 ikke: where is that list? 2021-04-07 20:19:27 /etc/ssl/certs/ca-certificates.crt ? 2021-04-07 20:19:33 /etc/ssl/cert.pem ? 2021-04-07 20:19:37 I don't get it 2021-04-07 20:19:54 thost files are just text files with all root certs concatenated 2021-04-07 20:20:01 /etc/ssl/cert.pem is a sym-link 2021-04-07 20:20:10 /etc/ssl/certs/ca-certificates.crt looks like it 2021-04-07 20:20:13 yes 2021-04-07 20:20:25 so just concatenate your PEM encoded certificates to that file 2021-04-07 20:20:28 and it should work 2021-04-07 20:23:14 looks like wget doesn't respect no_proxy env 2021-04-07 20:23:26 no_proxy/NO_PROXY 2021-04-07 20:25:35 https://gitlab.alpinelinux.org/alpine/apk-tools/-/issues/10681 2021-04-07 20:26:49 ikke: weird 2021-04-07 21:15:26 ikke: I can confirm on latest stable that wget doesn't respect no_proxy/NO_PROXY 2021-04-07 21:15:57 if, obviously, http_proxy+https_proxy+HTTP_PROXY+HTTPS_PROXY are set 2021-04-07 21:16:44 1https://github.com/gliderlabs/docker-alpine/issues/259 2021-04-07 21:18:07 so, the bug is actually upstream (busybox) 2021-04-07 21:18:12 okay, thanks, got it 2021-04-07 21:20:04 what's the difference between openjdk8-jre and openjdk8-jre-base? 2021-04-07 21:21:32 ah, the former brings just 3 bins 2021-04-07 21:21:48 appletviewer, policytool and one more policytool 2021-04-07 21:27:52 why can't I install package from locally downloaded .apk file if I only have 'main' repo, but the file was downloaded from 'community' repo? 2021-04-07 21:29:42 ah, nvm, the error was different 2021-04-07 21:30:42 so after adding certs to /etc/ssl/certs/ca-certificates.crt and installing curl (which in turn installs ca-certificates) apk forgets about those concatenated certs and I need to re-import them 2021-04-07 22:11:54 how to create user in alpine non-interactively? 2021-04-07 22:12:01 adduser asks for password 2021-04-07 22:12:20 ah, -D 2021-04-08 01:16:51 FreeBDSM▸ instead of updating that generated ca-certificates.crt file manually, install the ca-certificates package and add your certificate to /usr/local/share/ca-certificates with a .crt extension and run update-ca-certificates 2021-04-08 01:17:37 programmerq: to install it one needs certificates 2021-04-08 01:17:40 I had to do a bit of digging because I couldn't remember the specific directory to add it to. 2021-04-08 01:17:54 https://gitlab.alpinelinux.org/alpine/ca-certificates/-/blob/master/update-ca-certificates.8 - here's the man page source 2021-04-08 01:18:02 the man page is in the ca-certificates-doc package 2021-04-08 01:18:14 I made it work, the problem is solved 2021-04-08 01:18:25 it's just that solution is 2-step instead of 1 step 2021-04-08 01:18:29 FreeBDSM▸ it's fine to bootstrap it manually, but you'll want to add it to that directory so the next time update-ca-certificates is run you don't lose that cert again. 2021-04-08 01:18:53 that's step 2 2021-04-08 01:18:55 and yes, it will be run as you do regular package upgrades and maintenance 2021-04-08 03:43:50 hello! i am trying to install an hp multifunction printer. i could install the printer part, but not the scanner part. with other distros i used the hp-setup script for both. i installed and set up sane, and tried simple-scan, but it only recognizes my webcam. "scanimage -L" mirrors the result from simple-scan. i checked the arch wiki and it points me back to the hp-setup tool. is it possible to add the scanner at all? thank 2021-04-08 04:26:09 I simply don't know, but someone who might know would probably need the model number and/or USB ID. 2021-04-08 04:30:36 lol 2021-04-08 06:21:16 rodper45: there used to be a line in a config file that needed uncommenting. Let me check 2021-04-08 06:25:15 rodper45: https://wiki.voidlinux.org/SANE 2021-04-08 06:35:01 ignore ^^^. May be true but thought I was on Void's channel! 2021-04-08 09:15:21 hello to everyone 2021-04-08 09:16:24 i would like to install alpine on an old laptop and saw, that there is no wiki entry for installing from chroot 2021-04-08 09:17:38 i can handle the installation i think but i would like to document the process in the wiki if that's ok? 2021-04-08 09:18:09 is there a way to contribute in this way? 2021-04-08 09:20:25 dk_: something like this https://arvanta.net/alpine/armhf-setup/ 2021-04-08 09:21:02 dk_: https://wiki.alpinelinux.org/wiki/Alpine_Linux:Contribute 2021-04-08 09:21:07 or this https://arvanta.net/alpine/install-alpine-on-acer-r13-chromebook/ 2021-04-08 09:25:28 yes exactly.... 2021-04-08 09:26:39 i the installation on rpi is pretty much what i had in mind 2021-04-08 09:27:59 instead of installing directly from the repos i would have loop mounted the iso and first installed that 2021-04-08 09:31:39 ok, seems i didn't take enough time to read through the wiki and how to contribute :) 2021-04-08 09:32:07 will work from the link you gave me, thanks 2021-04-08 09:35:57 bye for now and thanks for the answers 2021-04-08 10:09:17 hello, short note on my way through the wiki, so you can understand why i asked 2021-04-08 10:09:28 i have i686 architecture 2021-04-08 10:10:34 i went from tutorials and howtos to https://wiki.alpinelinux.org/wiki/Alpine_newbie_install_manual#x86_64_x86_32_x86 2021-04-08 10:13:33 and from there, there is a link for a missing wiki page https://wiki.alpinelinux.org/wiki/Alpine_Install:_from_alpine_mirror_to_an_external_disc_by_chroot 2021-04-08 10:18:10 dk_: there is a shell script for chroot install 2021-04-08 10:19:44 thank you donoban 2021-04-08 10:20:02 I used it but I don't remeber where is it :\ 2021-04-08 10:21:48 https://github.com/alpinelinux/alpine-chroot-install 2021-04-08 10:22:55 default does a x86_64 install but you can change with -a ARCH option 2021-04-08 10:24:02 i just thought it might be a good idea to have the information to read through directly in the wiki, because there is a link 2021-04-08 10:24:37 i will further look into the installation process at the weekend 2021-04-08 10:25:43 I tried to do it following the wiki and finally ended using that sript, and then removing it and use docker-abuild 2021-04-08 10:25:44 thanks for the info 2021-04-08 10:26:08 but I only wanted some isolated enviroment for building packages 2021-04-08 10:26:55 have to go for now, bye 2021-04-08 10:27:04 ok, see u 2021-04-08 10:57:57 Hi there. 2021-04-08 12:23:58 is archivemount supposed to work on alpine? I see the package in 'community' repo, but after trying it - I got an error: `fusermount: fuse device not found, try 'modprobe fuse' first` and `modprobe fuse` returns another error: `modprobe: can't change directory to '/lib/modules': No such file or directory` and there is actually no such dir, only /lib/modules-load.d/ 2021-04-08 12:26:36 however, installing the package installed fuse and fuse-common as deps, but I see that there is also fuse3 available in the repo 2021-04-08 12:27:51 found this closed ticket https://gitlab.alpinelinux.org/alpine/aports/-/issues/10861 2021-04-08 12:29:57 well, I thought I'd just mount .tar.gz with kafka scripts to a dir to save some space from unpacking it... 2021-04-08 12:32:37 however small the difference (just 5MB: 53.2MB vs 58.2MB) - it is probably not worth the hassle, but still would be nice to have some info somewhere about the prerequisites for this to work 2021-04-08 12:39:11 FreeBDSM: Are you running Alpine in a container or as actual OS? 2021-04-08 12:39:30 Cogitri: on OpenShift Container Platform 2021-04-08 12:55:01 Cogitri: actually, it doesn't work in a container on an actual OS as well 2021-04-08 12:55:55 CentOS 7 with an old kernel (3.10.0-1127.19.1.el7.x86_64) 2021-04-08 13:29:15 hi folks, i have a realy simple question. where can i find the "potential deliverable source code? (tar.gz)". i have find multiple sources, but i want to now where the dev, are deploy the last state of alpine. THX 2021-04-08 13:37:43 Hi make i anything wrong? :| 2021-04-08 14:06:17 FreeBDSM: you can't load modules in a linux container 2021-04-08 17:44:41 If I want to boot off a USB with an overlay, do I need to create a custom ISO or is there some easier way? 2021-04-08 17:47:17 you don't need a custom iso for that, you can use any that works for you and setup-lbu https://wiki.alpinelinux.org/wiki/Alpine_local_backup 2021-04-08 17:55:36 bl4ckb0ne: I will leave this channel, you can contact me on #alpine-devel or privately. sorry 2021-04-08 17:55:53 or by mail, ofc 2021-04-08 18:02:27 What up with the .apk-new anyway to remove them all and is it safe? 2021-04-08 18:04:32 donoban: I just tried it and it said "Read-only file system" 2021-04-08 18:05:33 ouch yes if you want to save the overlay in the same usb you boot you have to create it with an specific method, not just with dd 2021-04-08 18:07:44 ikke, do you know anything about the .apk-new files? 2021-04-08 18:08:00 lucidone: Check this https://wiki.alpinelinux.org/wiki/Create_a_Bootable_USB 2021-04-08 18:08:27 "Copy ISO content to USB stick as individual files" 2021-04-08 18:09:56 Jenkler: just that apk creates them when files have been modified 2021-04-08 18:10:24 ikke, must be safe to remove them after boot right? 2021-04-08 18:10:41 ikke, any tool to clear them. cant find any dock about it 2021-04-08 18:11:49 update-conf 2021-04-08 18:17:59 ikke, thanks ;) 2021-04-08 19:23:26 is there something like /etc/skel in alpine? 2021-04-08 19:24:08 what is /etc/profile used for? 2021-04-08 19:25:31 /etc/profile is sourced by login shells 2021-04-08 19:26:09 doesn't seem so :( 2021-04-08 19:26:33 Do you have a login shell? 2021-04-08 19:26:43 looks like OCP's permissions circumcision kung-fu beats alpine :( 2021-04-08 19:27:28 I run alpine right now in a container in OCP (Red Hat OpenShift Container Platform, basically a custom k8s) 2021-04-08 19:27:32 /etc/skel should work as well 2021-04-08 19:27:37 it randomizes uid of my user 2021-04-08 19:27:52 I then can only write just to /tmp and nowhere else 2021-04-08 19:28:12 maybe on purpose? 2021-04-08 19:28:38 yes. and when I open interactive terminal - it somehow runs sh, not the default user's shell of choice 2021-04-08 19:29:22 I thought `printf '\n/bin/bash' >> /etc/profile` before creating user would help 2021-04-08 19:29:29 but it still runs sh 2021-04-08 19:30:03 the only working solution I found so far is using ln -s to make /bin/sh point to /bin/bash 2021-04-08 19:30:22 but then (actually dunno if due to this reason or not) - I can't run tmux 2021-04-08 19:30:25 /etc/profile is only sourced for login shells 2021-04-08 19:30:32 if ot 2021-04-08 19:30:43 if it's just executing /bin/sh, you don't get a login shell 2021-04-08 19:30:50 oh 2021-04-08 19:31:24 then is there a way (other than ln -s) to make it run bash on statup? 2021-04-08 19:31:32 I don't know 2021-04-08 19:32:49 without `ln -s` at least tmux works fine (only shows the same error when I quit a window) 2021-04-08 19:54:26 EF Ok, i tried adding my user to the lp and scanner groups, as the Void Wiki cites. no luck though. ill keep on digging 2021-04-08 19:56:47 rodper45: try the checking hpaio in dll.conf part 2021-04-08 19:57:08 EF will do! 2021-04-08 20:26:52 donoban: Thanks! The bootable USB link worked for what I wanted. 2021-04-08 20:26:53 oh great :) 2021-04-09 00:56:21 Is `setup-disk` well tested with GPT? 2021-04-09 02:24:33 hey, in the past my intel graphics have worked fine on installation, this time with the latest release the load seems really high watching youtube and stuff. has anything changed in the process of setting up xorg? 2021-04-09 03:25:10 it does seem to be using the intel graphics driver in the Xorg log 2021-04-09 07:34:06 Has someone made a build for tor browser on alpine? currently I'm using a custom firefox profile or a chroot 2021-04-09 09:20:22 blubshep: It seems hard to build, there was a request for it that was finally closed. But you can use flatpak com.github.micahflee.torbrowser-launcher 2021-04-09 09:55:38 Hmm my laptop doesn't receive the 3.9 Python upgrade. `apk add python3=3.9.4` tells me that it breaks `world[python3=3.9.4]` but satisfies 2021-04-09 09:56:24 Maybe because some package hasn't been rebuilt yet 2021-04-09 09:56:35 Afaik all packages have been rebuilt already? 2021-04-09 09:56:42 https://pkgs.alpinelinux.org/contents?file=&path=%2Fusr%2Flib%2Fpython3.8%2F*&name=&branch=edge&arch=x86_64 2021-04-09 09:56:45 On x86_64 at least 2021-04-09 09:57:05 Oh hmm 2021-04-09 09:57:16 That is quite a list 2021-04-09 09:57:17 The builders are idle on x86_64 though 2021-04-09 09:57:40 Seems like ncopa just missed a few packages 2021-04-09 09:57:48 I guess they haven't got bumped 2021-04-09 09:59:28 Yup 2021-04-09 10:19:47 It would be nice if it could list which packages though 2021-04-09 10:22:30 Doesn't pkgs.a.o do that with that query? 2021-04-09 10:27:56 I mean, when you use apk to install it 2021-04-09 10:28:04 It's not clear which packages are actually holding it back 2021-04-09 10:45:48 hum, i wonder why those got missed 2021-04-09 10:50:46 ikke: ah yes, that'd very nice 2021-04-09 10:51:35 I hope apk-3 improves on that 2021-04-09 11:01:22 Yeah that's I guess my biggest complaint with apk, it's error messages are often vague 2021-04-09 12:36:01 https://github.com/rapid7/metasploit-framework/pull/14833 update your alpine :) 2021-04-09 12:46:06 If you use ACF :-) 2021-04-09 12:46:37 if you use ACF :) 2021-04-09 12:49:17 what is acf? 2021-04-09 12:49:39 https://wiki.alpinelinux.org/wiki/Alpine_Configuration_Framework_Design 2021-04-09 12:49:53 a lightweight alpine-specific webmin 2021-04-09 12:50:35 wow. there is an inhouse webmin? til. 2021-04-09 14:34:48 are older packages accessible somewhere? 2021-04-09 14:35:49 No, sadly not 2021-04-09 14:36:19 packages from http://dl-cdn.alpinelinux.org/alpine/edge/testing/aarch64/ are replaced by the new ones right? 2021-04-09 14:36:29 Yes 2021-04-09 14:36:51 ooooh lucky me i got an apk locally 2021-04-09 14:41:09 how long does it typically take for packages to move from edge & testing into community and a release? 2021-04-09 14:41:40 does it perhaps entirely depend on the maintainer(s) of it? 2021-04-09 14:41:49 The maintainer decides it, yes 2021-04-09 14:42:08 After it's in community, it will be part of the next release 2021-04-09 14:44:47 I see, thanks! 2021-04-09 14:51:43 Giving feedback to the maintainer they you tested the package can certainly help 2021-04-09 15:22:40 did anyone see my question about intel graphics? cpu load seems very high on 3.13.4 2021-04-09 15:39:05 i think youtube increased default resolution 2021-04-09 15:46:46 echoline: could you paste your qestion again? 2021-04-09 15:47:16 Hello 2021-04-09 15:47:50 Is it normal that after updating and rebooting alpine edge it's not mounting my crypt root anymore 2021-04-09 15:49:28 Shit I guess next install I'm not doing edge 2021-04-09 16:08:39 Sigh 2021-04-09 16:15:59 MY-R: I have intel graphics in two machines, just reinstalled 3.13.4 on one and cpu load seems really high watching YouTube or stuff like that 2021-04-09 16:16:40 I'm also trying to set up an nvidia k80, I doubt that's relevant but could be 2021-04-09 16:18:36 Can you list installed packages filtered by repository? 2021-04-09 16:21:52 how do I do that? 2021-04-09 16:25:06 echoline: youtube in firefox is terrible for me too on alpine recently. Maybe try running the same video in a video player and see if the problem persists 2021-04-09 16:25:39 hmm 2021-04-09 16:27:09 xordspar0: not that I know of 2021-04-09 16:27:26 Ok, thanks. 2021-04-09 16:39:28 does apk even track that? 2021-04-09 16:40:07 Hi I got disconnected. Is it only happening to me? 2021-04-09 16:49:03 No issues here 2021-04-09 16:49:19 Are there unit/integration tests for `setup-disk`? 2021-04-09 16:49:42 Hmm that sucks 2021-04-09 16:50:04 Hello71: You can see the repo an individual package was installed from with `apk policy`. 2021-04-09 16:51:00 i don't think that tells you where it was installed from, it tells you where it can be installed from 2021-04-09 16:51:33 since alpine has (mostly) disjoint repositories it should turn out pretty close 2021-04-09 16:51:59 hm, or does it? 2021-04-09 16:52:11 Ah, I think you're right. 2021-04-09 16:53:03 Is luks2 on grub already a thing by default? 2021-04-09 16:53:17 Does it support it now I mean 2021-04-09 16:53:46 no, but encrypted /boot is of dubious value anyways 2021-04-09 16:55:12 Why 2021-04-09 16:56:10 what is the threat model 2021-04-09 16:56:32 "it's more secure 2021-04-09 16:56:40 lucidone: I don't think so 2021-04-09 16:57:32 it would probably be more useful to build some qemu tests for alpine-conf in general 2021-04-09 16:57:45 yes 2021-04-09 16:58:14 boot from usb, setup-alpine, shut down, unplug usb, reboot, log in 2021-04-09 16:58:47 i think other distros have this, but i'd wager a guess they are very heavy (not "alpine spirit") 2021-04-09 16:58:48 sudo cmp alpine-standard-3.13.4-x86_64.iso /dev/sdb 2021-04-09 16:58:51 cmp: EOF on alpine-standard-3.13.4-x86_64.iso after byte 139460608, in line 613423 2021-04-09 16:58:56 Does that mean write was successful? 2021-04-09 16:59:31 most likely 2021-04-09 16:59:51 although for more certainty you need to sync; echo 3 > /proc/sys/vm/drop_caches first 2021-04-09 17:01:30 I plugged into laptop and I see "Security Boot Fail" 2021-04-09 17:01:41 Hm maybe I have smth in my bios that is holding usb back 2021-04-09 17:02:24 more likely "secure boot failed" but sure 2021-04-09 17:02:25 Erase all secure settings did the trick 2021-04-09 17:05:43 I can ran alpine out of ram. I just noticed I don't have any hard drives in laptop 2021-04-09 17:11:21 i was thinking of setting up an android vm for use with my mobile banking app (only way to deposit checks via the internet, it seems) 2021-04-09 17:11:50 has anyone done that in alpine with like qemu? or a chroot? 2021-04-09 17:59:05 there are a number of commercial packages 2021-04-09 17:59:21 i think the most popular option currently is android-x86 2021-04-09 18:00:13 which i think is not strictly commercial but is apparently mostly sponsored by online gambling websites 2021-04-09 18:36:11 thats dope 2021-04-09 18:36:23 looks like there is an anbox package, maybe i will try that 2021-04-09 19:50:37 bank applications probably won't run on this stuff though 2021-04-09 19:50:49 because they think "open source = insecure" 2021-04-09 20:31:51 yeah if banks want insecure software, they want to PAY for it 2021-04-09 20:40:34 banks don't actually care about buying secure software, they care about having someone to blame 2021-04-09 20:41:00 and open source projects don't meet that need 2021-04-09 20:48:37 That's why I was forced to use RHEL for so many years. 2021-04-09 20:48:38 lol 2021-04-09 20:51:45 i work at a bank, can confirm 2021-04-09 20:58:16 i didn't actually work for a bank but i worked for a major financial institute in the marketing dept's it dept. 2021-04-09 20:58:24 ... yes some depts got their *own* IT dept 2021-04-09 20:58:46 "caring about someone to blame" over doing things properly is 100% accurate 2021-04-09 20:59:01 people who make directing desicions have little to no technical knowledge 2021-04-09 21:02:16 yep 2021-04-10 00:04:33 seems to be a question of which features the banking app uses, and whether those features are faithfully implemented by anbox/android-x86 2021-04-10 00:05:32 im sure banking apps work fine on lineageos which would seems to violate any requirement that there is "somebody to blame" 2021-04-10 00:05:54 well, if those are some DRM/Trustzone/Google Play services/similar things, then I'd say "probably not" 2021-04-10 00:06:16 (for the "supported by anbox") 2021-04-10 01:21:28 crappy bank apps tend *not* to work on lineage 2021-04-10 01:21:42 especially with addonsu installed 2021-04-10 01:22:11 usually you can patch over these with magisk though, but i'm not sure if that works in emulators 2021-04-10 03:09:31 qman__: Not only that, but the PCI requirements are a bit prescriptive in some ways, as well. 2021-04-10 13:35:16 Hmm, seems setuptools is broken? I can't install a local python package with `pip install .`, it reports that "egg_info" is an invalid setup.py command 2021-04-10 13:36:25 PureTryOut[m]2: yeah, we also seem to have issues with building pakcages 2021-04-10 13:36:28 packages 2021-04-10 13:36:52 see #alpine-devel; 2021-04-10 13:38:29 Ah good, will follow there 2021-04-10 14:41:03 Hi I'm having a problem after an upgrade today where it installed pipewire -- namely that pactl does not detect pipewire-pulse 2021-04-10 14:41:08 https://tpaste.us/49Jw 2021-04-10 14:41:15 looks like it's not registering itself with dbus properly? 2021-04-10 14:44:19 all of: pipewire, /usr/bin/pipewire -c pipewire-pulse.conf, and pipewire-media-session are running 2021-04-10 14:46:16 hmm that's weird, last time i started pipewire it started pipewire-pulse, but that seems to not have happened -- unless running with the pipewire-pulse.conf is equivalent 2021-04-10 14:47:38 is pipewire-pulse installed? 2021-04-10 14:47:48 yes 2021-04-10 14:49:18 /usr/bin/pipewire-pulse is present and when running it it tells me the service is already running, I'm guessing because for whatever reason pipewire started another instance with `-c pipewire-pulse.conf`, i would kill all of them and try again, but they don't actually respond to SIGTERM, and SIGKILL might leave dbus state (I don't know if it would interfer with dbus registering -- although that 2021-04-10 14:49:24 doesn't seem to be working at the moment anyway, so) 2021-04-10 14:50:10 I'll add my user to the rtkit group, restart, and try again 2021-04-10 14:51:23 hm, doesn't exist, so I can assume that my user has the right group permissions I guess 2021-04-10 14:55:54 I'd run the pulseaudio binary, since pipewire is supposed to be able to work in tandem, except it doesn't exist despite the package being installed. 2021-04-10 15:08:28 ok so, it's appeared to register as org.pulseaudio.Server, however running `pactl -s org.pulseaudio.Server [info|list]` both fail with connection refused, which I guess is 'progress' of sorts 2021-04-10 15:12:20 Check with `pactl info` what Pulse server software is running 2021-04-10 15:12:24 If any 2021-04-10 15:15:13 I already have, if you read above it fails with 'connection refused' 2021-04-10 15:15:37 Interesting 2021-04-10 15:16:03 Oh wait, you're manually starting pipewire? 2021-04-10 15:16:13 Afaik it needs to be enabled early in your session, and manually is probably too late 2021-04-10 15:16:18 There is no entry under /etc/init.d/ 2021-04-10 15:16:33 Yeah cause it has to run under your user, not as a system service 2021-04-10 15:16:37 Right 2021-04-10 15:16:41 It should be started with `/etc/xdg/autostart/pipewire.desktop` 2021-04-10 15:16:52 why should a .desktop be the main... 2021-04-10 15:17:12 Hmm? That's just how the XDG standard works and how programs are autostarted in DE's 2021-04-10 15:17:21 Uh, I am not running a DE 2021-04-10 15:17:23 But I'm assuming you're not running a DE that implements that XDG standard 2021-04-10 15:17:35 I'm running XMonad, and pulseaudio has worked fine before, DBus is running 2021-04-10 15:17:40 Like I've been reading forums threads for over 3 hours now, reading different ways to diagnose the problem, including active issues. I am not new to diagnosing and solving problems with linux distributions otherwise i wouldn't be running alpine linux as my main distribution haha 2021-04-10 15:18:03 Then I guess XMonad does something special for PulseAudio that it doesn't do for Pipewire? 🤷‍♂️ 2021-04-10 15:18:42 pulseaudio was previously started when needed, so when firefox or pavucontrol was loaded, XMonad does not care about the programs that run under it, it simply renders them 2021-04-10 15:19:06 I'm not sure why starting it after.. dunst.. or redshift, would matter 2021-04-10 15:19:59 Like I can add it to my autostart or .xinitrc and restart Xorg, but I'm not really sure _why_ that would change anything specifc. There shouldn't be anything started after or during XMonad that cares in that way 2021-04-10 15:20:45 Also, it was started before any programs that would make use of DBus were opened, fwiw 2021-04-10 15:23:36 a lot of the things I'm turning up with the connection refused issue point to the configuration format, but this is a clean install from latest -- that being the problem would mean the packaged configs are out of date 2021-04-10 15:24:42 The packaged configs are not specific to Alpine, they come straight from the build system 2021-04-10 15:24:57 You can check if there are files with a `.apknew` prefix in the Pipewire config directory 2021-04-10 15:26:03 Those do not exist because this is a fresh upgrade and before I just ran pulseaudio. I haven't upgraded in three months and the switch happened after the second-previous upgrade haha 2021-04-10 15:26:20 however... running strace on pactl seems to show that the pulseaudio libraries are not installed for pactl to work 2021-04-10 15:27:08 Did you switch on purpose btw or "on accident" as in it just got installed without you telling apk too? 2021-04-10 15:28:06 On accident. I installed vlc-qt and it purged pulseaudio, so I decided to go with a full upgrade and a reboot, and pavucontrol (and my bluetooth headphones) stopped working 2021-04-10 15:28:13 I've read that pipewire-pulse still requires the pulseaudio libs to be installed to run properly, so I guess that points to a solution. 2021-04-10 15:28:18 Ah yeah, probably best to just reinstall PulseAudio 2021-04-10 15:28:22 I've tried 2021-04-10 15:28:32 https://postmarketos.org/edge/2021/04/02/pipewire-pulse/ 2021-04-10 15:28:37 the pulseaudio package is installed but no longer provides /usr/bin/pulseaudio 2021-04-10 15:28:48 `apk fix pulseaudio`? 2021-04-10 15:29:22 I've done apk fix and the only problem is an exfat man page being owned by two packages 2021-04-10 15:29:34 Well do what the blog post I linked suggests 2021-04-10 15:29:37 It worked for others 2021-04-10 15:29:46 Yeah, I could pin pulseaudio, but since this is the Future(tm) I'd rather find an actual solution, not a workaround 2021-04-10 15:30:02 You don't have to pin pulseaudio, that is not what that blog posts suggests 2021-04-10 15:30:24 It just forces pipewire-pulse to be removed in favor of pulseaudio by creating a conflict on it 2021-04-10 15:31:37 well, pulseaudio _is_ installed, just the binary is missing, and most of the library files that pactl requires to work, apparently 2021-04-10 15:32:09 But is pipewire-pulse uninstalled? Please make sure. And do `apk fix pulseaudio`, not just `apk fix` 2021-04-10 15:32:43 I'm not interested in uninstalling pipewire, because I would like to get this working as-is, not supplant it with pulseaudio 2021-04-10 15:33:27 I did not say uninstall pipewire, I said uninstall pipewire-pulse. Although I guess you can, I really wouldn't have both pulseaudio and pipewire-pulse installed at the same time 2021-04-10 15:33:51 oh! I see what you mean, hm 2021-04-10 15:38:26 hm, how is pipewire-pulse provided through apk? 2021-04-10 15:38:50 aports and the online packages interface only shows `pipewire` 2021-04-10 15:38:54 https://git.alpinelinux.org/aports/tree/community/pipewire/APKBUILD#n49 2021-04-10 15:39:06 -pulse is a subpackage of pipewire 2021-04-10 15:39:47 i see 2021-04-10 15:40:22 So in this case, pipewire-pulse is not providing any of the libraries that pactl requires to work, which is why this error exists 2021-04-10 15:40:32 https://tpaste.us/MQg8 2021-04-10 15:41:07 No that is incorrect 2021-04-10 15:41:19 pactl works fine for me on my pipewire system with pipewire-pulse 2021-04-10 15:41:19 you can see this via this grepped strace (`strace -f -o /tmp/pipe.txt pactl info && grep ENOENT /tmp/pipe.txt`) 2021-04-10 15:41:56 Well, pipewire-pulse provides pulseaudio, and pipewire-pulse does not provide any of the given libraries, so your system might have upgraded without purging? maybe? 2021-04-10 15:42:13 No it is purged, I made sure of it 2021-04-10 15:42:20 all of the given libraries are provided, however, by pulseaudio 2021-04-10 15:44:32 pactl looks for libpulse.so*, which is not provided by pipewire: https://pkgs.alpinelinux.org/contents?file=libpulse.so*&path=&name=&branch=edge 2021-04-10 15:44:48 can you run the strace and grep to a tpaste for me? 2021-04-10 15:44:58 that will at least show the difference in how pactl is behaving 2021-04-10 15:45:09 Make sure libpulse is installed (not pulseaudio itself) 2021-04-10 15:45:53 apk didn't install it when I asked, so it's installed 2021-04-10 15:46:10 That is not making sure 😛 Do `apk list -I libpulse` 2021-04-10 15:46:24 libpulse-14.2-r5 x86_64 {pulseaudio} (LGPL-2.1-or-later) [installed] 2021-04-10 15:46:34 Hmm ok, that is the only thing pactl needs 2021-04-10 15:46:42 Could you run `ldd /usr/bin/pactl`? 2021-04-10 15:46:45 (and post the output) 2021-04-10 15:47:08 these seem to be libraries loaded at runtime 2021-04-10 15:47:13 libpulse.so.* comes from libpulse which clearly is installed 2021-04-10 15:47:19 https://tpaste.us/E1qV 2021-04-10 15:48:03 I even did `apk fix libpulse`, just in case, before running that ldd command 2021-04-10 15:48:09 Yeah ok everything found and no missing symbols at least. So as far as we can see, pactl should "just work" ™️ 2021-04-10 15:48:12 re-running pactl again gives me this: https://tpaste.us/0Ee8 2021-04-10 15:48:20 I've again only grepped for ENOENT 2021-04-10 15:48:34 (specifically: cat /tmp/pipe.txt | grep ENOENT | tpaste) 2021-04-10 15:48:59 can you post a pipe.txt with a grep for 'open' ? 2021-04-10 15:49:58 because at the moment you're looking for symptoms, whereas doing that will show how pactl is _acting_, which is the true difference here that we need to compare, IMHO 2021-04-10 15:51:56 https://tpaste.us/qg8B 2021-04-10 15:52:03 thanks! 2021-04-10 15:52:15 Well one weird thing in your strace output is that it's not looking in /usr/lib for things like libFLAC.so.8 2021-04-10 15:52:35 well the first difference that shows for me is that mine is looking under /usr/lib/pulseaudio 2021-04-10 15:53:11 whereas yours is simply looking under /usr/lib, oh except for libpulse 2021-04-10 15:53:59 your system might be inconsistent -- or mine, but your pactl is finding /usr/lib/pulseaudio/libpulse.so.0 -- which alpine linux packages (shown earlier) says does not exist at that path 2021-04-10 15:54:19 oh, that' an ENOENT, nvm lol 2021-04-10 15:55:26 I doubt my system is inconsistent lol 2021-04-10 15:56:11 It definitely finds it at `/usr/lib/libpulse.so.0` which is the right place 2021-04-10 15:56:36 But yeah your system seems to not be looking in `/usr/lib` for whatever reason 2021-04-10 15:57:59 ok I grepped for 'open' and my native is on the right, your foreign pactl output is on the left : https://tpaste.us/yN6k 2021-04-10 15:58:19 full output at https://tpaste.us/N81v 2021-04-10 15:58:40 *full output of my pactl strace log | grep open 2021-04-10 16:01:39 uhhhh, oops ignore that, paste messed up when I posted in the url to my terminal 2021-04-10 16:04:23 I might have a solution 2021-04-10 16:06:10 yep 2021-04-10 16:06:22 pipewire isn't respecting XDG_CONFIG_DIR 2021-04-10 16:06:29 or at least, my pipewire isn't 2021-04-10 16:08:42 Interesting, I haven't changed that so I wouldn't know 2021-04-10 16:09:37 hm, ok so _something_ to do with where pactl expects things to be, I'm sure of it, but my solution wasn't exactly functional 2021-04-10 16:10:09 basically pipewire seems to be creating the following files/dirs when run: ~/pulse/, ~/pulseverbose.log 2021-04-10 16:13:19 so what I'm thinking is that pactl expects the files in ~/pulse to be at a different location, or something of the old pulseaudio configuration was left and is causing it to fall over 2021-04-10 16:14:38 what's the output of your XDR_RUNTIME_DIR? 2021-04-10 16:14:41 mine is unset 2021-04-10 16:15:13 `/run/user/1000` 2021-04-10 16:17:12 ok, that's useful -- also this https://github.com/flatpak/flatpak/issues/4058 seems to be relevant 2021-04-10 16:24:15 Hi, is there any reason why the automatic raid done by setup-alpine may fail to boot? 2021-04-10 16:24:48 I'm just trying to figure out if there is anything special I need to do after I go through setup-alpine using 2 disks in sys configuration 2021-04-10 16:25:02 markpash_: generally it should just work 2021-04-10 16:25:06 what is failing? 2021-04-10 16:25:26 the initramfs seems to fail to mount the sysroot 2021-04-10 16:26:11 I think the UUID that was placed in the extlinux config was incorrect 2021-04-10 16:26:31 Maybe this? https://gitlab.alpinelinux.org/alpine/aports/-/issues/12325 2021-04-10 16:26:53 PureTryOut[m]2: figured it out. My pactl is looking for pulse/ in /var/run (which is a symlink to /run) -- changing permissions on /run (as a test, this isn't a permanent solution) to root:alx g+w, and then setting XDG_RUNTIME_DIR to /var/run/ enables pactl to run properly. 2021-04-10 16:28:04 @ikke: not running lvm :( but I just rebooted another time and it worked fine...... 2021-04-10 16:28:18 this is not good, going to reboot a few more times and see what's wrong 2021-04-10 16:28:21 markpash_: oh ok, strange 2021-04-10 16:28:36 So essentially the only problem is that it's unable to find the runtime data that it needs, maybe there's a configuration file I can edit so I can place it somewhere it doesn't require permissions. And this I assume would be a more general solution too -- so I'm wondering where I can write this where other people can find it(?), possibly the mailing list? 2021-04-10 16:28:44 markpash_: Yeah, but I'm not sure that issue is specifically limited to lvm, so was wondering if that was the same issue 2021-04-10 16:29:30 pavucontrol works fully now as well, there are some random errors from firefox resetting a connection but everything works 2021-04-10 16:30:41 @ikke: thanks, I'll try and get to the bottom of this and see if it's bug or some other issue. This damn machine takes 5 mins to boot though :( 2021-04-10 16:31:39 markpash_: Heh, the machine I had issues with similar 2021-04-10 16:31:53 Took me a while to get to the bottom of that, so I feel you :) 2021-04-10 16:32:30 PureTryOut[m]2: thanks for that bit of info, and for your assistance btw 2021-04-10 16:32:38 No problem! 2021-04-10 16:33:00 It might be worth asking about this on Pipewire upstream btw 2021-04-10 16:34:06 @ikke: yeah, I hope it's an issue on my end, because I really like the alpine installer and how easy you can have a RAID1 boot disk setup in under 2 mins 2021-04-10 16:34:42 yeah, I imagine it's a general solution to the problem, too. I'll dig into the pactl code first to see if i can put this in a more coherent place that doesn't require setting permissions, however 2021-04-10 16:42:17 So, I have an interesting one: The alpine installer won't boot on my system unless I remove the drives. Without the plugged in, it boots fine: when it's booted, I can plug them in. But if they are plugged in during boot it will NOT go past realizing they exist. 2021-04-10 16:45:37 boot order (EFI before CSM?) issue? Are the drives USB or on some kind of add-on card? (or raid card?) 2021-04-10 16:46:08 bios boot, just booting to USB: It loads the kernel, and gets to seeing sda / sdb / sdc: then sits there 2021-04-10 16:46:26 it is going through an LSI sas card: mpt3sas driver 2021-04-10 16:52:42 why did I think it was mpt3sas... it's mpt2sas 2021-04-10 16:56:18 actually, this isn't even flashed to IT mode 2021-04-10 17:22:40 @ikke: still the same problem, I am in the emergency recovery shell, and doing ls /dev shows the /dev/sd* devices but not the md devices. Is there a way to troubleshoot from this emergency shell? 2021-04-10 17:24:13 markpash_: can you try: nlplug-findfs -p /sbin/mdev 2021-04-10 17:24:46 what goes in place of ? 2021-04-10 17:25:45 cat /proc/cmdline 2021-04-10 17:25:53 root= 2021-04-10 17:26:29 cool, let me go try that 2021-04-10 17:32:43 @ikke: I tried that and ls /dev/md0...md2 appeared. I didn't know what to do next though. 2021-04-10 17:34:49 markpash_: That's the same issue as in that ticket then 2021-04-10 17:35:25 Workaround was to use an older kernel 2021-04-10 17:35:35 Still trying to find the actual solution 2021-04-10 17:37:18 @ikke: great, ok so after finding the md devices, how do I get it so I can continue from the emergency shell? when I tried just exit I got a kernel panic? Or should I just go install a 3.12 iso? 2021-04-10 17:37:53 markpash_: boot from the livecd again, mount the disk, chroot in it, and install the kernel from edge 2021-04-10 17:38:19 great, I'll give that a go, thanks! 2021-04-10 17:38:28 sorry, kernel from 3.12 2021-04-10 17:39:10 oh, I think installing the IT firmware actually fixed it lol 2021-04-10 17:39:30 eeey it did, awesome 2021-04-10 17:44:19 @ikke: so is this actually a kernel bug or the alpine packaging of the kernel being problematic? missing configs? 2021-04-10 17:48:14 Or changes in the kernel 2021-04-10 17:48:23 Some interaction 2021-04-10 17:48:38 markpash_: could you please add information about your setup to that issue? 2021-04-10 17:49:30 Yup will do, thanks again 2021-04-10 18:04:38 @ikke: I retried with a 3.12 iso, same problem :/ this is really annoying now. 2021-04-10 18:05:03 not frustrated with you or anyone, just the fact it takes sooooo long to then get an error 2021-04-10 18:06:59 you reinstalled everything? 2021-04-10 18:08:53 yup 2021-04-10 18:10:03 for me, it was working as soon as I switched to a 3.12 kernel 2021-04-10 18:12:26 yeah I downloaded and went through the setup with alpine-standard-3.12.6-x86_64.iso 2021-04-10 18:12:43 I hit the same wall. 2021-04-10 18:25:12 still works after nlplug-findfs? 2021-04-10 18:34:33 hey guys. when i try to install thunderbird i get https://tpaste.us/YoPM. im on edge branch and i enabled the testing repos. updated and upgraded my system, i dont know what to do else 2021-04-10 18:37:43 g0r3: looks like thunderbird would need to be rebuilt 2021-04-10 18:41:12 ikke: so this is not something i can do? 2021-04-10 18:41:18 No 2021-04-10 18:41:24 okay 2021-04-10 18:41:27 tahnks 2021-04-10 18:44:49 i think this is actually a longstanding bug? 2021-04-10 20:49:00 g0r3: there is an open issue about thunderbird (https://gitlab.alpinelinux.org/alpine/aports/-/issues/12017) 2021-04-10 20:50:22 I did try to have more logs, but without success so far 2021-04-11 17:06:54 https://github.com/ifupdown-ng/ifupdown-ng/issues/142 been having the same problem using dhclient with alpine and ifupdown-ng and dual-stack. 2021-04-11 17:07:38 it ended up running dhclient -6 -v manually to generate a dhcp ipv6 request for the box. 2021-04-11 17:09:05 Until I installed dhcpcd again. and works out ouf the box, for both ip address families. 2021-04-11 17:36:24 but swapping dhclient with dhcpcd worked using ifupdown-ng, no manual configuration needed. 2021-04-11 18:29:49 is there 'redo' in busybox vi? 2021-04-11 18:48:59 hi, I installed alpine on my main desktop, I can't get wayland working, I get a black screen 2021-04-11 18:52:02 "DBus user socket was not found", maybe is related 2021-04-11 18:55:59 is it some kind of DE? 2021-04-11 18:56:06 or sway? 2021-04-11 18:56:19 KDE 2021-04-11 18:56:33 but that error is firefox via firejail 2021-04-11 18:56:52 god help you then :/ 2021-04-11 18:57:08 I'm not sure what groups should I add myself 2021-04-11 18:57:23 I used sway, added to video, audio, input 2021-04-11 18:57:26 well in this laptop it's working fine 2021-04-11 18:57:34 let's try 2021-04-11 18:57:41 for firefox to work in session started with dbus-run-session 2021-04-11 18:57:51 donoban: is elogind installed? 2021-04-11 18:57:59 uhM 2021-04-11 18:58:09 but it was just sway, not kde, there is ton of shits, I have no idea about tat 2021-04-11 18:58:24 I have elogind-inhibit on my path 2021-04-11 18:58:29 is that? 2021-04-11 18:58:54 is a sesison manager like sddm? 2021-04-11 18:59:07 no, it's not a session manager 2021-04-11 18:59:33 elogind-inhibit is part of elogind, yes 2021-04-11 18:59:49 I've never used wayland, so cannot be a lot of use 2021-04-11 19:00:26 well I think that I will try upgrading to edge 2021-04-11 19:00:35 or don't use shitty kde 2021-04-11 19:00:58 thats an option too yes 2021-04-11 19:01:01 u can't do anything with that monster, if hundreads of pieces are glued together 2021-04-11 19:01:52 also I can't enable compression with btrfs, I add 'compress=zstd' to fstab but it seems ignored 2021-04-11 19:03:03 oh it entered but segfaulted 2021-04-11 19:07:18 I have like 600MB on /lib/firmware, can't I remove things from there? 2021-04-11 19:08:59 u well u have firmware what u installed 2021-04-11 19:09:21 also probably really just god can help you if u hacking whole machine :/ 2021-04-11 19:11:16 well I just did: setup-disk -m sys /mnt/ 2021-04-11 19:11:18 donoban: if you know what firmware modules you need, you could install the individual packages 2021-04-11 19:11:28 and tehre was already the firmware files 2021-04-11 19:12:14 what package installed all them? 2021-04-11 19:12:21 yes, linux depends on linux-firmware-any, which in turn installs all firmware modules 2021-04-11 19:12:24 yes by default linux instlal all firmware 2021-04-11 19:12:35 linux-tls? 2021-04-11 19:12:36 yes 2021-04-11 19:12:58 but if u instll any or some of them, then it will not push whole firmware crap 2021-04-11 19:13:34 https://pkgs.alpinelinux.org/packages?name=linux-firmware-*&branch=edge&arch=x86_64 2021-04-11 19:14:47 but I need to remove linux-tls, then install some custom firmware, and finally reinstall linux-tls again? 2021-04-11 19:14:52 no 2021-04-11 19:14:58 it's not a hard dependency 2021-04-11 19:15:09 ah I can't just remove 2021-04-11 19:15:12 I can* 2021-04-11 19:16:00 first u have to install at least something, that it does not pull all crap 2021-04-11 19:16:40 ok 2021-04-11 19:16:54 apk add linux-firmware-... 2021-04-11 19:17:10 You can also install linux-firmware-none, then everything should be removed 2021-04-11 19:17:55 iwlwifi 0000:05:00.0: loaded firmware version 46.4d093a30.0 9260-th-b0-jf-b0-46.ucode op_mode iwlmvm , Im trying to find this one 2021-04-11 19:19:29 https://pkgs.alpinelinux.org/contents?repo=main&page=2&file=%2Aiwl%2A&arch=x86_64&branch=edge&name=linux-firmware-%2A 2021-04-11 19:19:35 I guess linux-firmware-other 2021-04-11 19:20:27 hehe ok 2021-04-11 19:20:29 so let's try reboot 2021-04-11 19:22:45 I'm gonna try to enable compression via kernel arg 2021-04-11 19:24:08 can I debug the commands that the initfs does? should it be aware of 'compress=zstd' on /etc/fstab? 2021-04-11 19:25:41 append debug_init to cmdline 2021-04-11 19:26:38 uhM nice thanks 2021-04-11 19:31:38 uhM, I don't see any additional log 2021-04-11 19:31:50 ahh, I didn't regerenated grub.cfg :( 2021-04-11 19:41:00 uhM 2021-04-11 19:42:07 I'm not sure "who" is the responsable of applying the 'compress' option 2021-04-11 19:42:27 Me neither, but it would not be initramfs-init 2021-04-11 19:42:39 So either an option in fstable or something the kernel supports 2021-04-11 19:43:35 I would like to enable it before restoring my backup :S 2021-04-11 19:43:51 I'm gonna try with rootflags 2021-04-11 19:44:00 yeah, would make sense 2021-04-11 19:44:33 if there was some missing module in the initrams, it would log some error on dmesg? 2021-04-11 19:44:57 I guess in dmesg 2021-04-11 19:50:39 it works! :P 2021-04-11 19:51:27 /dev/dm-0 on /home type btrfs (rw,relatime,compress=zstd:3,ssd,space_cache,subvolid=257,subvol=/@HOME) 2021-04-11 19:51:43 but this is pretty uggly, it should work with /etc/fstab 2021-04-11 19:52:47 uhM... 2021-04-11 19:52:59 [ 16.707371] BTRFS info (device dm-0): setting incompat feature flag for COMPRESS_ZSTD (0x10) 2021-04-11 19:53:01 bad news 2021-04-11 20:00:58 this should be enough for the initramfs loading the zstd module? 2021-04-11 20:01:01 localhost:/etc/mkinitfs/features.d$ grep zstd * 2021-04-11 20:01:02 btrfs.modules:kernel/crypto/zstd 2021-04-11 20:55:30 any idea what handle xorg logs in ~/.local/share/xorg ? would like to move it somewhere else 2021-04-11 20:57:49 xorg itself I assume 2021-04-11 21:00:04 probably not confiruable? did not find mentinoned 'log' anywhere in xorg files 2021-04-11 21:00:32 in xorg read about 'logfile' but it is mentioned it works only for root xorg 2021-04-11 21:02:22 probably startx --logfile or something like that during start would work, will try after smoke break :/ 2021-04-11 21:03:23 was looking into it, but could not find a suitable option for it 2021-04-12 07:15:11 Hi all :-) 2021-04-12 07:16:06 Do you have a list of the security enforcements added to alpinelinux ? 2021-04-12 07:18:29 I install as little as possible, that's my aproach :/ 2021-04-12 07:21:02 btw. I m surprised that vi/nvi is not in repositories, as vi in busybox is really crippled 2021-04-12 07:23:08 even 'undo' acts somehow weird 2021-04-12 07:24:35 I was thinking to default security measure like PIE compilation (https://git.alpinelinux.org/aports/tree/main/gcc/APKBUILD#n325) or fortify-headers (https://pkgs.alpinelinux.org/package/edge/main/x86_64/fortify-headers) 2021-04-12 08:26:02 Hello. I want to use alpine for creating a ssh/SOCKS5 proxy. I want to secure my ssh so my users will not be able to do any harm. Users will connect from windows machines, via plink. I was NOT able to setup a non-interactive shell for them, not I was able to properly restrict them to their home folders. Also, I am unable to find a restricted shell in alpine, like rbash. Can someone point me to some information ? 2021-04-12 09:31:30 sorinello: Alpine includes bash 2021-04-12 09:31:42 rbash is bash -r *OR* a symlink named "rbash" pointing towards bash. 2021-04-12 09:33:47 More on that here: https://www.gnu.org/software/bash/manual/html_node/The-Restricted-Shell.html 2021-04-12 09:39:40 Consider using the "Match User" and "ChrootDirectory" directives in sshd.conf also. 2021-04-12 09:48:59 zcrayfish, I am using Match User 2021-04-12 09:49:02 with PermitTTY no 2021-04-12 09:49:16 unfortunately, ChrootDirectory does not allow me to login, I get a broken pipe error 2021-04-12 09:49:53 My guess there would be missing /dev entries inside the chroot. 2021-04-12 09:49:56 but PermitTTY no seems to be enough, since I cannot issue commands from CLI if I connect 2021-04-12 09:50:06 zcrayfish, yeah, you are right 2021-04-12 12:53:27 kmmndr: i don't think we have a centralized list 2021-04-12 12:54:26 sorinello: "restricted" shells are inherently insecure 2021-04-12 12:55:03 it would be better to use a proper socks server, or better, a udp vpn 2021-04-12 13:09:12 Hello71, can you tell me a supported socks5 server by alpine ? 2021-04-12 13:25:05 dante, shadowsocks, probably some others 2021-04-12 13:25:32 wireguard is much faster though 2021-04-12 16:47:36 Is it mailing list abandoned like wiki? 2021-04-12 16:49:03 alpine-devel is still actively used. I suppose you refer to alpine-users? 2021-04-12 16:52:49 the most funny is probably infra :/ 2021-04-12 16:53:17 Yes, that one is not really used anymore 2021-04-12 16:53:23 users is just dead 2021-04-12 16:55:20 it is used by spammers ... if it's not used, why it is abandoned the way that make an impression of abandoned project 2021-04-12 16:56:13 I read some about mailing list, forum, gitlab ... did not get output from it 2021-04-12 16:56:51 helby: It is surprising that nvi isn't packaged. It would be super easy to do though. 2021-04-12 16:57:30 xordspar0: probably yes, that's why it's surprising to me, that nobody was interested during so long time 2021-04-12 16:58:38 as nvi is practically capable for fraction of size compared to bloated monsters like neo/vim 2021-04-12 17:44:42 hi, i'm trying to configure alpine to store overlay files on the same media that it boots from, but i'm unable to mount this media writable - i should be able to store overlays on the same media alpine boots from right? 2021-04-12 17:45:22 funnyone: yes, but the iso is isofs, which cannot be writable 2021-04-12 17:45:54 You need to create a bootable media with a writable filesystem 2021-04-12 17:46:09 setup-bootable can do that, or you can create one manually 2021-04-12 17:50:22 ikke: thankyou, i will have a go 2021-04-12 18:39:12 the more I update my Linux machine the more I understand why it will be used as desktop anytime soon 2021-04-12 18:42:25 feel free to share your breakthroughs :/ 2021-04-12 18:42:49 audio was already messy with pulseaudio a decade ago 2021-04-12 18:43:02 now with pipewire I no longer have sound on edge :} 2021-04-12 18:43:23 relevant xkcd https://xkcd.com/927/ 2021-04-12 18:44:14 +not 2021-04-12 18:45:44 funny, I never used pulse or pipe or other nonsense, always went with defualt like alsa or sndio/bsd and my audio always works 2021-04-12 18:46:32 sndio <3 2021-04-12 18:46:49 well on openbsd it worked well for me 2021-04-12 18:46:55 if only Linux devs could do one thing and doing it well instead of proliferating alternatives 2021-04-12 18:48:02 the most funny things is, that most alternatives are usually based on other already broken alternatives :/ 2021-04-12 18:48:08 heh 2021-04-12 18:48:36 today I thought about editors, why the fuck somebody created vim :/ 2021-04-12 18:48:47 to add 20MB of useless shit on top of vi? 2021-04-12 18:48:57 haha 2021-04-12 18:49:08 colours, buffers/split and completions? 2021-04-12 18:49:20 I can't live without vim's macro to be honest 2021-04-12 18:49:34 okay, what about next 19MB? 2021-04-12 18:49:34 my colleagues always O_O when they look at me refactoring some code at amazing speed 2021-04-12 18:50:12 neovim might be a step in the right direction 2021-04-12 18:50:16 there are editors which do all and more with like 10% of size of vim 2021-04-12 18:50:39 yes, neovim has only 17Mb with all crap 2021-04-12 18:50:49 helby: you might like https://github.com/martanne/vis 2021-04-12 18:51:01 yes vis is okay, but it reminds me vim 2021-04-12 18:51:20 much lightweight but practically nothing special 2021-04-12 18:51:34 kakoune is bigger but at least it has visual shit, which helps a lot 2021-04-12 18:52:49 it's so funny with these editors, to install 20MB vim, to just install additional 20 plugins on top of it 2021-04-12 18:53:15 then work in line like with sed or awk 2021-04-12 18:53:23 with colors 2021-04-12 18:54:10 on other side ... 'ed' ... that's something :/ 2021-04-12 18:55:21 just two days ago, fucked up fstab, so have to edit it, but both 'vi' and 'ed' require /tmp so not one launched :D 2021-04-12 18:55:43 actually vi will complain about backup files IIRC but you can still use it 2021-04-12 18:55:56 I was not able, it was on openbsd 2021-04-12 18:56:11 not sure how it is here, but as busybox, this shit will probably run even without alectricity 2021-04-12 18:56:17 ele 2021-04-12 18:57:01 but vi is cool, relatively simple, small and enough for a lot of jobs 2021-04-12 18:57:18 and part of POSIX 2021-04-12 18:57:36 well I m hearing that often, but I m not sure if that is feature or bug 2021-04-12 18:58:03 yes it's quite surprising 2021-04-12 18:58:15 bc. lot of software is kept ancient with tons of useless historical shits just bc. of posic, unix, this and that 2021-04-12 19:00:25 still interested in how u get to the situation, that you installed pulseaudio 2021-04-12 19:00:40 why would adult do it? :) 2021-04-12 19:02:12 desktops like GNOME/Plasma require it, also unless you run a raspberry pi with a unique output you're pretty much required to use pulseaudio 2021-04-12 19:03:03 well, why would adult install gnome/plasma? :) 2021-04-12 19:03:05 with pulseaudio you can mix volume per app and obviously change output sinks on the fly while applications emitting sound are running 2021-04-12 19:03:17 mix volume? 2021-04-12 19:03:22 which is definitely required for bluetooth headsets, HDMI audio, dockable stations, and so on 2021-04-12 19:03:29 like listen 3 radio station? and every on different level? 2021-04-12 19:03:45 no like listening to a podcast and playing a game to quote one example 2021-04-12 19:04:20 alsa can't do it with mixerctl? 2021-04-12 19:04:40 or whatever is there, I never needed that 'feature' 2021-04-12 19:04:42 probably, but nobody use sndiod on Linux except some few enthusiasts 2021-04-12 19:05:18 I have still never seen a good solution for hotplugging audio devices on linux 2021-04-12 19:05:19 oh yes, that was from sndio, used openbsd until like tomorrow, so still mixed 2021-04-12 19:05:33 how not? 2021-04-12 19:05:43 nshp, maybe you never had a dockable laptop then 2021-04-12 19:05:45 I plug whatever and it plays 2021-04-12 19:05:46 it's pretty useful 2021-04-12 19:06:11 helby: that's true of your analog audio jack, sure 2021-04-12 19:06:16 aka you have a headet plugged into a dock station at work, you leave your job with your laptop and boum, the sound is now on the speakers 2021-04-12 19:06:24 not at all true with bluetooth outputs or other digital outputs 2021-04-12 19:07:10 yes audio jack, even it's name telling me, it's for an audio :/ 2021-04-12 19:07:13 no idea how you'd do that without hardcoding specific audio outputs in your asoundrc 2021-04-12 19:07:39 bluetooth is crap, always turned it off in bios 2021-04-12 19:07:43 and even then, it's a pain to switch between them 2021-04-12 19:07:59 it will be always pain, it's shitty bluetooth 2021-04-12 19:08:11 it's like shitty pulseaudio or gnome 2021-04-12 19:08:11 helby: right, you are exactly the audience for which everything in linux works 2021-04-12 19:08:13 :) 2021-04-12 19:08:16 the audience which refuses to use things :P 2021-04-12 19:08:54 nshp: please type slower, my one line calculator can't read so fast lines :d 2021-04-12 19:09:01 nshp, oh excuse me I didn't see your first message correctly :P 2021-04-12 19:09:08 bluetooth works just fine with every other thing I own, and it works almost fine with pulseaudio. I just have to ignore the fact that it eats 90% of my CPU time while doing so 2021-04-12 19:09:33 that's not working if that nonsense it 90%cpu 2021-04-12 19:09:53 agreed. as said, pulseaudio is crap. I wish there were another solution. I just haven't seen one that works yet. 2021-04-12 19:10:02 networkmanager was harder with the CPU 2021-04-12 19:10:03 u can mine crypto instead and buy 10$ mini radio :d 2021-04-12 19:10:18 yes, networkmanager is great companion with pulseaudio 2021-04-12 19:10:29 I never need that shit too :/ 2021-04-12 19:10:51 NetworkManager + PulseAudio + Avahi and get your heater for free 2021-04-12 19:10:53 networkmanager is a different story. it is entirely pointless so far as I can tell, as dhcpcd + wpa_supplicant can completely replace it 2021-04-12 19:11:08 with no real loss of functionality 2021-04-12 19:11:11 networkmanager is the OpenJDK of networking 2021-04-12 19:11:27 well ip link, ifconfig supplicant and other 10 scripts... then u never know how to set up things 2021-04-12 19:12:37 alpine has now if down up updown conf and obviouslly ip link 2021-04-12 19:12:53 like 100 different shits for one thing 2021-04-12 19:13:20 dhcpcd and wpa_supplicant are usually enough 2021-04-12 19:13:22 to just run fucking dhcpcd :) 2021-04-12 19:13:33 helby: different people have different needs 2021-04-12 19:13:35 it's how my laptop is configured 2021-04-12 19:13:46 helby: the alpine base is small and thus limited in functionality 2021-04-12 19:13:51 there is no one size fits all 2021-04-12 19:14:09 ikke: well bootloader is not so small ;/ 2021-04-12 19:14:22 helby: and ifupdown-ng and iproute2 serve different purposes 2021-04-12 19:14:55 ikke: probably nobody know just those 7 people who were around it 2021-04-12 19:14:56 I boot my laptop with efibootmgr 2021-04-12 19:15:03 pretty awesome 2021-04-12 19:15:16 yes efistub obviously 2021-04-12 19:16:49 wanted to check networking, but when I saw these ifupdown up down ifconfig ip link supplicantos I lost interested for now as using always ethernet 2021-04-12 19:17:20 can't even imagine to manage one connection with one tool and other with something else 2021-04-12 19:17:45 helby: why haven't you yet built that tool that does everything 2021-04-12 19:18:02 bug free, ofcourse 2021-04-12 19:18:07 ikke: I still believe in you 2021-04-12 19:18:09 unless the kernel has mechanism to handle wireless connections you're pretty much required to launch a daemon to do it 2021-04-12 19:18:55 helby: I have no desire to build such a tool, in fact, I think it's better to have dedicated tools for the job 2021-04-12 19:19:12 yes dedicated obviously 2021-04-12 19:19:27 but not that, if speed of connecton is 10mb use this tool, if speed is 20 use another one 2021-04-12 19:19:35 if connection is coming from left, use something else :/ 2021-04-12 19:19:42 I can already guess the response from Linus if someone send a patch so that the kernel does wireless auto-connect and search by itself 2021-04-12 19:19:50 "why the kernel should f*cking care?" 2021-04-12 19:20:03 if google will ask, it will be there next day :/ 2021-04-12 19:30:18 markand: all these issues can be addressed in new distro :) then there will be 19786 distros :/ 2021-04-12 19:43:24 even ffmpeg complain about run dir bc. wanna put pulse file there even pulse nowhere close 2021-04-12 19:52:11 linux kernel already has dhcp for many years 2021-04-12 19:53:12 for netboot, I assume? 2021-04-12 19:53:30 though, that's ipxe 2021-04-12 19:54:55 it's somewhat orthogonal: ip=dhcp is for mounting root fs 2021-04-12 19:55:07 it dates from before initrd/initramfs 2021-04-12 19:55:20 nowadays the logical thing would be to remove it, but "muh backwards compat" 2021-04-12 19:56:20 the ancient way of nfs root is ip=dhcp nfs=10.0.0.1:/nfsroot root=/dev/nfs or something like that, no initrd or initramfs required 2021-04-12 20:04:40 will see how it will looks after 30 years of adding features and removing almost nothing 2021-04-12 20:06:53 if 'ed' will survive, everything will be okay and even fixable with rm utility 2021-04-12 20:07:09 ['''''''''''''''''' 2021-04-12 20:08:09 reading little bit wiki :/ that's fun too 2021-04-12 20:08:19 started from the beginning 2021-04-12 20:08:43 now on FAQ: How to I remove CDROM? 2021-04-12 20:09:25 it probably dates to time Hello71 mentnioned :/ 2021-04-12 20:11:36 ... most external links are dead already obviously, did not survive millenium ;) 2021-04-12 20:13:17 Ah, my cat typed that, haha. 2021-04-12 20:15:08 tried exit ed I bet 2021-04-12 20:15:47 :D 2021-04-12 20:16:29 Hi there 2021-04-12 20:16:34 helby: You're reading the whole wiki starting from the beginning? If you're bored and need something to do, I have list for you! 2021-04-12 20:16:43 Can someone help me figure out why wifi can't find anything? 2021-04-12 20:16:58 iwlist wlan0 scanning gives me no scan results 2021-04-12 20:17:16 I tried with ntui, nmcli... 2021-04-12 20:17:23 mtui* 2021-04-12 20:17:26 xordspar0: no whole wiki, who would read that nonsense? just checking little bit, bc. never list it if maybe something really ancient is there ... some gem :d 2021-04-12 20:17:42 My wireless card is: 02:00.0 Network controller: Intel Corporation Wireless 7260 (rev 73) 2021-04-12 20:17:42 here we go ... wifi 2021-04-12 20:18:08 Yeah wifi 2021-04-12 20:18:30 Should work out of the box afaik 2021-04-12 20:19:15 Or maybe iwlwifi isn't loaded 2021-04-12 20:19:36 you will figure out sooner or later :/ 2021-04-12 20:20:16 Figure what out? 2021-04-12 20:20:37 holy shit, now I remember while reading wifi shit ... 2021-04-12 20:21:29 just recollected the last reason I went from alpine to openbsd for some time ... it was fuckering with wifi :d 2021-04-12 20:21:47 I just had to saw that wiki page 2021-04-12 20:22:54 Hm iwlwifi is loaded 2021-04-12 20:23:04 6 different commands to make it work + add it sa service 2021-04-12 20:23:20 wireless-tools is ancient 2021-04-12 20:23:24 we should remove it from setup-interfaces 2021-04-12 20:23:47 frojnd: running as root? 2021-04-12 20:23:50 why? I think nobody use it anyway ;) 2021-04-12 20:24:16 iwlist [iface] scanning just shows the last scan results rather than starting a scan if you're unprivileged iirc 2021-04-12 20:24:26 nshp: ah ok 2021-04-12 20:24:33 Let me use nmcli then 2021-04-12 20:24:44 iw dev wlan0 scan 2021-04-12 20:24:59 Ah iw is even beter 2021-04-12 20:25:22 don't scan, just add name and password 2021-04-12 20:25:31 more steps u left, better 2021-04-12 20:25:43 oom wroom 2021-04-12 20:26:03 Gotta see if everyting is working properly first :P 2021-04-12 20:26:51 It works proper not sure why nmtui didn't find stuff before 2021-04-12 20:27:10 I will ditch networkamanger 2021-04-12 20:29:18 nmcli won't find bssid 2021-04-12 20:29:19 Just iw 2021-04-12 20:29:24 That's interesting 2021-04-12 20:36:50 Well not even wpa_cli in interactive mode can't find bssids 2021-04-12 20:38:15 it's not bc. networkmanager took over? 2021-04-12 20:38:54 I deleted it 2021-04-12 20:39:01 Service is not running 2021-04-12 20:39:25 I figured I don't need networkmanager for just one wifi 2021-04-12 20:39:34 Home wifi 2021-04-12 20:39:38 Forever home wifi 2021-04-12 20:40:31 not sure if it does not require restart, but I really did not fuckering with that long time 2021-04-12 20:41:00 Restart of the steel? 2021-04-12 20:41:16 Hm let me try 2021-04-12 20:41:55 no idea, I restarting like crazy when playing with that kind of shit 2021-04-12 20:44:13 Why it won't find bssids? 2021-04-12 20:44:23 iwlwifi is loaded which is reqired for this wifi card 2021-04-12 20:44:26 iw finds bssid 2021-04-12 20:44:34 But I can't authenticate with iw for wpa2 2021-04-12 20:52:38 Maybe it has something to do with LOWER_UP 2021-04-12 20:52:48 My primary laptop's wifi: BROADCAST,MULTICAST,UP,LOWER_UP 2021-04-12 20:52:57 Vs new one with Alpine: BROADCAST,MULTICAST,UP 2021-04-12 20:54:45 did u try to delete installed crap/disable running network services and just run 'setup-interfaces'? 2021-04-12 20:57:58 Ha no I didn't good suggestion 2021-04-12 20:59:42 Hm it did ask me for the pass but ip a s didn't show interface to have any IP 2021-04-12 20:59:52 It also found BSSids 2021-04-12 21:01:29 So not sure what's going on 2021-04-12 21:13:11 iw doesn't support wpa 2021-04-12 21:13:14 only wep 2021-04-12 21:13:21 you need to use wpa_supplicant or iwd 2021-04-12 21:13:48 setup-interfaces should do all this 2021-04-12 23:32:50 how do i delete the cache files not currently installed? 2021-04-12 23:42:51 how do i delete the cache apks that r not installed? 2021-04-13 00:09:21 apk cache clean ? 2021-04-13 00:11:32 didn't work 2021-04-13 00:12:58 that's for older versions, not uninstalled 2021-04-13 00:15:47 idk if this is officially supported, but it didn't blow up my system: I simply rm'd all of the apk files from the cache directory. I then rebuilt the cache with `apk cache download' 2021-04-13 00:17:50 that's an option but needs extra bandwidth 2021-04-13 10:12:19 Gooood morning linuxian Alps ! 2021-04-13 10:31:37 2021-04-13 10:31:38 |" 2021-04-13 10:40:32 I have troubles setting up limits on alpine ; I set up a `realtime` group with rules in `/etc/security/limits.conf`, but they don't show up in `ulimit`. I already did `apk add shadow`. I'm running postmarketos' alpine. Duckducking around I found people asking the same question but no solution that apply to me 2021-04-13 10:45:38 Today I just reinstalled and only configured wlan0 and it's working 2021-04-13 10:45:52 So default installation for standart image wlan works 2021-04-13 10:47:36 obviously ;) 2021-04-13 10:47:59 Can i set static ip so it will boot faster? 2021-04-13 10:48:07 I chose dhcp but it's taking extra 2 seconds 2021-04-13 11:15:59 rerun setup-interfaces 2021-04-13 11:16:25 but if dhcp takes 2 seconds maybe something wrong with network 2021-04-13 11:17:03 la_fleur[m]: alpine does not support /etc/security/limits.conf 2021-04-13 11:17:11 DHCP server is probably attempting a ping to ensure the address is free first. 2021-04-13 11:17:15 Hello71: with pam (shadow) it should work 2021-04-13 11:18:48 hm 2021-04-13 11:20:43 shadow provides /bin/login 2021-04-13 11:31:25 uhoh, someone broke the pelican package :< 2021-04-13 11:33:15 I'll just reserve IP on router and assing static IP on laptop 2021-04-13 11:35:38 Btw since alpine doesn't have systemd how do I measure boot time? systemd-analyze is tool I'm using on other distros. What is recommended for non systemd? 2021-04-13 11:36:02 Time untill system gets to the tty1 2021-04-13 11:42:05 frojnd: done by bootchartd I believe, part of mkinitfs 2021-04-13 11:46:50 well few seconds to get lease is normal I assume 2021-04-13 11:47:06 at least it's standard time for me 2021-04-13 11:47:12 on ethernet 2021-04-13 11:48:34 To get all tha man pages I have to install mandoc-doc ? 2021-04-13 11:49:16 now I am thinking about it, udhcpc is not a deamon? that I don't need to wait to get lease during boot? 2021-04-13 11:49:32 as it's probably the slowest part during boot 2021-04-13 11:50:32 frojnd: for bootchartd I assume you need to add "bootchart" to the features list in /etc/mkinitfs/mkinitfs.conf and then rebuild the initramfs 2021-04-13 11:51:43 minimal: add 'boothcart' or 'bootchartd' 2021-04-13 11:52:32 "bootchart" as that's the name of the feature file (in /etc/mkinitfs/features.d/) 2021-04-13 11:52:44 the software is called bootchartd 2021-04-13 11:56:41 minimal: thank you. I've installed apk add mandoc man-pages mandoc-apropos less less-doc and mkinitfs-doc to get me started 2021-04-13 11:58:05 After mkinitfs I probably have to reboot? 2021-04-13 11:58:14 So changes are applied 2021-04-13 12:01:28 yes, you can't measure the boot time if you don't reboot ;-) 2021-04-13 12:01:46 But how do I actually see bootchart stuff 2021-04-13 12:01:50 There is no man for it 2021-04-13 12:01:54 No doc 2021-04-13 12:02:18 I just did sudo mkinitfs 2021-04-13 12:02:30 ==> initramfs: creating /boot/initramfs-lts 2021-04-13 12:02:35 And that was it 2021-04-13 12:02:43 it writes to /var/log/bootchart.tgz 2021-04-13 12:02:52 use to source Luke! https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/bootchartd.in 2021-04-13 12:02:53 Aaa! 2021-04-13 12:02:59 s/ to/ the/ 2021-04-13 12:04:19 haven't used it myself so not sure exactly what is contained in the created tgz file 2021-04-13 12:05:28 No bootchart.gz in /var/log/ hm 2021-04-13 12:06:43 my /etc/mkinitfs/mkinitfs.conf features="ata base ide scsi usb virtio ext4 bootchart" 2021-04-13 12:07:07 was your initrd also before that called /boot/initramfs-lts ? 2021-04-13 12:07:38 frojnd: are you using Alpine is "sys" disk mode or running from ram? 2021-04-13 12:07:44 minimal: sys 2021-04-13 12:08:49 wdtTP2O82Kft: what do u mean before? From first boot I see lts install alaso minimal installation iso includes lts I believe 2021-04-13 12:09:20 right, I only use sys myself. That should have worked. As wdtTP2082Kft pointed out did it create an initramfs file of the same file as was previously used? Also you should check the contents of the created initramfs file to verify the bootchart stuff was activated 2021-04-13 12:10:17 your Syslinux or Grub config will define the name of the initramfs file loaded during boot... 2021-04-13 12:15:27 Which file are you 2 talking about? 2021-04-13 12:15:54 that depends on which of Syslinux and Grub you are using 2021-04-13 12:16:27 5.10.29-0-lt 2021-04-13 12:16:31 5.10.29-0-lts 2021-04-13 12:16:43 And bootloader is grub 2021-04-13 12:17:19 I just ran sudo mkinitfs 2021-04-13 12:17:54 ok, so I'd assume that /etc/grub/grub.cfg refers to "/boot/initramfs-lts" 2021-04-13 12:18:30 minimal: On that I just have /etc/grub.d/ dir 2021-04-13 12:19:02 typo, meants /boot/grub/grub.cfg 2021-04-13 12:20:23 /boot/grub/grub.cf have initrd /boot/initramfs-lts 2021-04-13 12:22:19 I also just rebooted and pressed 'e' key to see initrd /boot/initramfs-lts inside commands 2021-04-13 12:22:36 So this should be good right? 2021-04-13 12:23:11 so if your kernel is 5.10.29.0-lts then first you can check what would be put into a new initramfs file (without actually updating it) by running "mkinitfs -l 5.10.29-0-lts" 2021-04-13 12:23:54 and actually update it by running "mkinitfs -q 5.10.29-0-lts" 2021-04-13 12:24:29 you don't actually need the "-q" (that's quiet mode) 2021-04-13 12:25:22 I see ./sbin/boothartd https://termbin.com/mtum 2021-04-13 12:26:46 seems to have put the required files (defined in /etc/mkinitfs/features/bootchart.files) into the initramfs 2021-04-13 12:27:53 Yeah, they seem to be included 2021-04-13 12:28:22 minimal: -q modes is quiet, but even just mkinitfs is quiet... 2021-04-13 12:28:31 I didn't see anything while running that command 2021-04-13 12:29:17 frojnd: I'm looking at my Ansible playbook where I've automated this so guess I was just making sure it was quiet :-) 2021-04-13 12:29:34 also seems the kernel version parameter is optional 2021-04-13 12:30:32 minimal: when running mkinitfs -l ? 2021-04-13 12:31:18 in all situations, look at output of "mkinitfs -h" 2021-04-13 12:31:32 I just low how easily I can build a specific directory env with another architecture thanks to apk and then cross-compile using it as sysroot with clang <3 2021-04-13 12:32:13 markand: that would be something of interest to me. Any guide on that? 2021-04-13 12:32:15 so assuming you've run "mkinitfs" without the "-l" it will have actually updated the initramfs file (check its timestamp has changed) and in theory a reboot should run bootchartd during boot to create the file in /var/log/ 2021-04-13 12:32:36 minimal: yeah I've just ran sudo mkinitfs without -l 2021-04-13 12:32:44 frojnd, yes I keep you up to date once I get some experiments working :) 2021-04-13 12:32:45 minimal: timestamp ? 2021-04-13 12:33:08 "ls -l" for the file and look at the date/time 2021-04-13 12:33:27 Oh for ./sbin/boothartd you mean 2021-04-13 12:33:47 Or all the files defined in /etc/mkinitfs/features.d/bootchart.files 2021-04-13 12:34:13 nope, I mean "ls -l /boot/initramfs-lts" to see that its time has changed to provide mkinitfs changed it 2021-04-13 12:34:20 s/provide/prove/ 2021-04-13 12:36:03 minimal: https://termbin.com/miirf 2021-04-13 12:36:17 This time it took at least 10 seconds to do it 2021-04-13 12:36:29 The first time I was running it it was done in 1 or 2 seconds 2021-04-13 12:37:09 But still no bootchart.tgz inside /var/log directory 2021-04-13 12:37:40 yeah, I just mde the same changes on a VM here and rebooted - no file either 2021-04-13 12:37:54 Maybe it's a bug 2021-04-13 12:37:57 so in theory that should have been it but obviously something's not right 2021-04-13 12:38:05 looking at the code now.... 2021-04-13 12:38:24 I did say I'd not actually tried this before ;-) 2021-04-13 12:38:24 OOOOH 2021-04-13 12:38:42 Maybe I don't have tar 2021-04-13 12:38:59 tar -zcf "$BOOTLOG_DEST" header $pacct *.log 2021-04-13 12:39:19 I have ... I have.. it's inside core packages 2021-04-13 12:42:55 frojnd, https://pastebin.com/ppvc3xaS 2021-04-13 12:44:00 Thank you markand 2021-04-13 12:51:57 frojnd: although its now in the initramfs I don't think its actually being run during boot 2021-04-13 12:53:15 frojnd: ah, think I see the issue 2021-04-13 12:54:36 minimal: it says here: http://www.bootchart.org/docs.html that boot logger (/sbin/bootchartd) is run by the kernel instead of /sbin/init. This can be achieved by modifying the GRUB or LILO kernel command line 2021-04-13 12:54:52 kernel /vmlinuz-2.6.10 ro root=/dev/hda1 init=/sbin/bootchartd 2021-04-13 12:55:56 nope, just added " chart" to the boot command line. However with that its running but having problems - stuck with errors looping during boot :-( 2021-04-13 12:56:07 so best you don't try that yet 2021-04-13 13:08:09 frojnd: looks like /dev/null doesn't exist when bootchartd is being run early in boot and so the script is giving errors. Investigating further 2021-04-13 13:15:59 frojnd: its going to take a while to work out what the issue is - I'll let you know what I find 2021-04-13 13:19:53 minimal: only if ur also gonna have some fun with it, but thanx for fixing this issue 2021-04-13 13:39:43 hello, please help me debug our problem a little bit. We have k8s with pods based on different OSes and running on different worker-nodes. The problem is that there seems to be an issue with DNS resolution on some worker nodes, but it has different results when we try simple curl requests from those pods to diagnose the issue: testing from alpine-based containers results into `curl: (6) could not resolve host: ....${namespace}.svc. 2021-04-13 13:39:43 cluster.local`, while the same curl request from a CentOS7-based pod (running on the same worker node) results into successful response, but with HUGE delay of ~4 seconds 2021-04-13 13:40:33 so it looks like we have some dns resolution issue on some of our worker nodes in our k8s cluster, but I don't get why curl requests behave so much differently between alpine-based pods and centos-based pods 2021-04-13 13:40:46 could someone, please, shed some light on what's happening here? 2021-04-13 13:43:55 since it all happens in k8s - it is k8s dns and k8s controls /etc/resolv.conf in pods. in both kinds of pods (centos-based and alpine-based) contents of that file are like `search ${pods_namespace}.svc.cluster.local svc.cluster.local cluster.local ocp.ourcompany.internaldomain local` + `nameserver ${ip}` + `options ndots:5`. 2021-04-13 14:07:44 i don't think musl supports search domains 2021-04-13 14:08:30 also k8s is abusing the .local tld reserved for mdns but that's a separate issue 2021-04-13 14:10:14 wait no it does 2021-04-13 14:10:29 maybe your udp is broken 2021-04-13 14:10:36 check with dig/drill 2021-04-13 14:13:24 Hello71: could you expand on what exactly (and how) to check? 2021-04-13 14:19:07 nslookup returns response in 1ms 2021-04-13 14:21:46 Is that ${namespace} literal? 2021-04-13 14:26:37 ikke: no 2021-04-13 14:26:54 ikke: there's a string like ournamespace-prod 2021-04-13 14:27:13 Ok 2021-04-13 14:29:11 https://i.imgur.com/mkXdrR4.png :/ 2021-04-13 14:30:16 https://i.imgur.com/qpHGgmz.png different domain 2021-04-13 14:33:54 don't use nslookup or host, they're inconsistent between platforms and give misleading results 2021-04-13 14:34:19 dig/drill for raw dns, getent hosts on linux for libc reuslts 2021-04-13 14:39:51 Hello71: https://pkgs.alpinelinux.org/packages?name=dig&branch=v3.13&arch=x86_64 0 resoolts 2021-04-13 14:41:21 Part of bind-tools 2021-04-13 14:41:53 Sorry 2021-04-13 14:42:01 It's drill 2021-04-13 14:42:06 2 ll 2021-04-13 14:45:49 no, I mean that there's no dig for alpine 2021-04-13 14:45:56 so I re-built image with drill 2021-04-13 14:46:21 Well, then bind-tools, like I mentioned 2021-04-13 14:52:04 https://pkgs.alpinelinux.org/contents?file=dig 2021-04-13 14:53:33 drill gave weird result 2021-04-13 14:54:01 it shows that only full fqdn line ${service}.${namespace}.svc.cluster.local gets resolved 2021-04-13 14:54:33 neither ${service}.${namespace}.svc.cluster nor ${service}.${namespace}.svc nor ${service}.${namespace} get resolved 2021-04-13 14:56:56 dig confirmed that behavior 2021-04-13 14:57:55 dig ignores search domains 2021-04-13 15:01:13 I also tried `dig` with `+noignore` and got similar results (only fqdn with full suffix gets resolved) 2021-04-13 15:02:08 same with `+notcp +noignore` 2021-04-13 15:02:50 I don't get it 2021-04-13 15:03:06 so why does curl fail with 'could not resolve host' then? 2021-04-13 15:13:29 can you strace curl? 2021-04-13 15:23:26 Hey, I see this pkg: https://pkgs.alpinelinux.org/packages?name=kannel&branch=edge . But in my docker config, when I specify "FROM alpine:edge" and try to apk add kannel, it says it's not found. What am I missing? 2021-04-13 15:24:37 its in the testing repo 2021-04-13 15:28:44 Is there a way I can specify that? 2021-04-13 15:30:05 have echo 'https://dl-cdn.alpinelinux.org/alpine/edge/testing' >> /etc/apk/repositories 2021-04-13 15:31:20 that's it thanks. I'm gonna just keep it there for the duration of this pkg install, then remove it 2021-04-13 15:32:05 aha I see this is -X option to apk 2021-04-13 17:50:53 Hi. I'm trying to use liburing on alpine but I get an error saying a required header () is not available. Is there a quick fix for this? 2021-04-13 17:54:38 https://pkgs.alpinelinux.org/contents?file=swab.h&path=%2Fusr%2Finclude%2Flinux&name=&branch=edge 2021-04-13 17:54:42 uringating: looks like linux/swab.h is part of the linux headers, so you will want to `sudo apk add linux-headers` 2021-04-13 17:54:50 thank you 2021-04-13 17:55:34 what do you guys think of uring 2021-04-13 17:55:54 I've been hearing it's a lot better than epoll and similar to kqueue 2021-04-13 17:56:13 No experience with it, but the idea sounds good 2021-04-13 17:56:41 epoll always felt like a pain to me so it looks promising 2021-04-13 18:02:53 ... tried to clean my keyboard on thinkpad, now have to buy new keyboard ;) 2021-04-13 18:03:18 what did you clean it with 2021-04-13 18:04:00 well I took out every key :/ some of them broke obviously ;) 2021-04-13 18:05:16 why didn't you just use compressed air 2021-04-13 18:07:35 not so dirty 2021-04-13 18:07:59 it was sticky, bc. shitty glue on some keys, as there were glued labels 2021-04-13 18:08:05 so it was hard to clean up 2021-04-13 18:08:18 I spent probably 3 hours with it :d 2021-04-13 18:08:34 now it's like new :) just 3 keys are broken 2021-04-13 18:08:56 honestly I'm not even gonna ask helby 2021-04-13 18:09:33 it's T420, it should not break :/ 2021-04-13 18:11:46 anyway wanted to change it, bc. previously broke trackpad middle button :/ 2021-04-13 18:16:27 well nice chat I have to go uringate (have a piss) 2021-04-13 18:31:03 helby, I heard of someone who didn't bother who replace a broken 'q' letter, so they always kept q in paste buffer 2021-04-13 18:31:12 s/who/to/ 2021-04-13 18:33:00 lol 2021-04-13 18:33:20 My laptop keyboard is broken as well, but I use an external keyboard now 2021-04-13 18:35:35 https://impressive.net/people/gerald/images/dilbert/condescending-unix-users.gif 2021-04-13 18:35:59 pracitally fixed it now with other buttons, but G can't fix. bc. on thinkpad there is fuckin trackpad, so the key has different corner :/ 2021-04-13 18:37:16 however I can try to cut some part from other key :/ 2021-04-13 18:39:34 the issu is, I am continuing to break other keys ... 2021-04-13 18:41:21 you got issues 2021-04-13 18:41:26 problems 2021-04-13 18:41:44 you should buy keyboard on garden sales 2021-04-13 18:41:54 ok, thanks for your help earlier today folks 2021-04-13 18:56:35 holy ... this is now really custom keyboard 2021-04-13 18:56:43 custom keyboards are the most expensive :/ 2021-04-13 18:57:54 had to cut corner from one key, the issue was, during that process I broke another 3 keys :/ 2021-04-13 18:58:42 at least it's practically usable until I get new one 2021-04-13 19:05:03 Hello. Is is normal for the who command not displaying users connected via ssh ? 2021-04-13 20:05:51 you need to use 3.13+ and install and enable utmps 2021-04-13 20:33:58 hello, out of curiosity how does alpine decide what apps to include in the repos ? 2021-04-13 20:37:11 tiotags: volunteers providing and maintaining an apkbuild 2021-04-13 20:38:23 Software should have an appropriate license 2021-04-13 20:39:56 by appropriate I assume you mean any open source one, right ? 2021-04-13 20:41:38 so I assume only really popular programs have a chance ? 2021-04-13 20:47:59 no. you can volunteer to provide and maintain an apkbuild for some really obscure free software 2021-04-13 20:53:14 I hardly know how to do basic things with linux, maintaining software is probably beyond my ability (never tried though) 2021-04-13 20:53:15 We have very obscure packages in the repo as well 2021-04-13 21:05:16 Hello71: I just tryed to install "who" (coreutils) and run into the same problem like sorinello 2021-04-13 21:05:42 did you install and enable utmps 2021-04-13 21:05:46 Hello71: coreutils depends on utmps 2021-04-13 21:05:52 did you enable it 2021-04-13 21:05:53 Yes 2021-04-13 21:06:14 You mean start utmpd? 2021-04-13 21:06:18 yes 2021-04-13 21:06:19 yes 2021-04-13 21:06:20 did you log back in 2021-04-13 21:07:24 Yes, just tryed loged out and back in 2021-04-13 21:07:41 rc-service utmpd status 2021-04-13 21:07:41 * status: started 2021-04-13 21:09:06 The leading cause of massive uncontrollable erections is gigantic assess according to my penis. 2021-04-13 21:09:44 I tryed becourse same problem here, month ago ;) 2021-04-13 21:15:49 Hello71: Or do I have to start the service (utmpd) "enable to default" and reboot ... yust start it is not enough? 2021-04-13 21:24:00 Hi Nice I need somehelp with the usb boot 2021-04-13 21:24:18 Hi Nice I need to meet everybody. I need  some help with the usb boot 2021-04-13 21:31:48 Ok, I yust tryed and added "utmpd" to default and rebooted the VM ... no changes 2021-04-13 21:32:36 Btw. I'm on 3.13.4 and Kernel 5.10.29-0-virt 2021-04-13 21:37:38 Hello Allo 2021-04-13 21:37:45 Hello All 2021-04-13 21:44:28 Hello All, 2021-04-13 21:44:29 I need a little help with the ISO Boot. I am having issue that say Mounting boot media failed. 2021-04-13 21:44:29 Could somebody help 2021-04-13 21:46:32 do you have all the message or something? 2021-04-13 21:46:48 Ok let me write it here. 2021-04-13 21:47:52 Mounting boot media failed. 2021-04-13 21:47:52 initramfs emergency recovery shell launched. Type `exit` to continue boot 2021-04-13 21:47:53 sh: can't access tty: job control turned off 2021-04-13 21:49:18 I am buring the iso using MAC through unetbootin 2021-04-13 21:52:37 is there anything before? 2021-04-13 21:52:56 No nothing before that 2021-04-13 21:53:10 `blkid`? 2021-04-13 21:53:57 (that means, « what's the output of the command "blkid"? ») 2021-04-13 21:55:14 Give me moment I burning the new fresh one before trying it again. 2021-04-13 21:57:20 Ok Got the result of blkid. Let's me write it down 2021-04-13 21:59:04 ` 2021-04-13 22:00:22 "/dev/sdb1: LABEL="ALPINE" UUID="xxxx-xxxx" TYPE="vfat" 2021-04-13 22:01:06 "/dev/sda:  UUID="xxxx-xxxx" TYPE="vfat" 2021-04-13 22:01:15 Above are the result 2021-04-13 22:02:22 Can anyone explain the behavior of ifupdown-ng and requires? I would expect it to bring up the required interface before the one that requires it, but that does not seem to be the behavior. 2021-04-13 22:03:26 The docs seems to say it should work the way I am thinking, but that is not what we see when networking starts. 2021-04-13 22:05:03 hm... i don't actually remember how our utmps is supposed to work 2021-04-13 22:05:20 Lampang: unetbootin is not recommended 2021-04-13 22:06:11 it has too many modes that mangle the image 2021-04-13 22:06:41 Hello71You recommending I should go with dd instead 2021-04-13 22:06:59 in fact i think it exclusively has mangling modes 2021-04-13 22:07:06 best to use dd (very carefully) 2021-04-13 22:08:24 Yes. I am trying to avoid that. I don't want to mistakenly type the drive path 2021-04-13 22:08:32 Ok let's me try 2021-04-13 22:11:11 Hello71: Thanks a lot! If you (ore someone else) have an Idea yust let me know in here ... 2021-04-13 22:12:01 Have to go to sleep soon ;) 2021-04-13 22:15:19 Hello71 Just rerun initiate using `dd` should be done soon 2021-04-13 22:18:03 Hello71 2021-04-13 22:18:03 Just retried using dd. 2021-04-13 22:18:04 I am still getting the same issue 2021-04-13 22:18:16 what dd command did you run 2021-04-13 22:18:26 and how much ram do you have 2021-04-13 22:19:03 sudo dd if=/[FULL_PATH]/alpine-standard-3.13.4-x86_64.iso of=/dev/disk2s1 2021-04-13 22:19:08 I have ram like 16GB 2021-04-13 22:23:13 Try sync after dd is complete? 2021-04-13 22:24:22 I have already pull out the disk. 2021-04-13 22:24:22 Let me retry it. 2021-04-13 22:24:23 Once the dd is done. 2021-04-13 22:24:23 I just type sync and enter or you want me to append to the command 2021-04-13 22:29:45 Just sync. 2021-04-13 22:29:50 done 2021-04-13 22:30:15 It flushes the write buffer and does not complete until it is finished. 2021-04-13 22:30:53 I am getting the same error 2021-04-13 22:31:12 Hm, weird. Not sure then. 2021-04-13 22:31:39 did u check ancient wiki? not sure if it's still applicable, but there is one shit about changing device name 2021-04-13 22:31:44 https://wiki.alpinelinux.org/wiki/Create_a_Bootable_USB 2021-04-13 22:32:10 The error you cited is output from mount? 2021-04-13 22:33:00 Or error from Alpine while booting? 2021-04-13 22:34:00 The error is from the alpine usb disk. 2021-04-13 22:34:01 Once I booted it, it show that error. 2021-04-13 22:34:29 It's like the boot menu come up then I press enter then it come here 2021-04-13 22:35:11 the option in the menu is `Linux lts` 2021-04-13 22:37:33 also if u have another usb, u can maybe try classic installer and then instead setup-alpine do setup-bootable and point it to another usb 2021-04-13 22:37:43 I assume it should work 2021-04-13 22:38:34 Maybe try mounting the stick to see if you get a similar error? 2021-04-13 22:38:35 helby I did check there and try something. 2021-04-13 22:38:35 I might use it in the wrong way. I will try find another usb and try 2021-04-13 22:39:21 shrizza Now with using dd I could not mount it on mac 2021-04-13 22:40:04 Oh yeah, Mac... Not sure if it can mount ext. 2021-04-13 22:41:31 I think with dd you should sdX instead of sdx1crap 2021-04-13 22:41:44 helby concerning the ancient wiki. 2021-04-13 22:41:44 Can you help adapt the line if I have current file from the iso. 2021-04-13 22:41:45 From the wiki I am not sure that I did make the right edit 2021-04-13 22:42:15 wiki is not sure too, don't worry 2021-04-13 22:42:51 hm, yes, it should be /dev/disk2 2021-04-13 22:43:02 i thought there was another p for partition 2021-04-13 22:44:12 Ah yea, you gotta dd to the entire disk and not patition. Good catch. 2021-04-13 22:44:22 Hello71You mean to use of=/dev/disk2 instead of of=/dev/disk2s1 2021-04-13 22:44:31 Ok let's me try again. 2021-04-13 22:48:39 done. Let's see if it boot properly 2021-04-13 22:48:50 Still the same error 2021-04-13 22:49:23 did you sync 2021-04-13 22:49:32 also it will run faster if you put bs=1M 2021-04-13 22:49:45 or bs=1084576 2021-04-13 22:50:47 Yes I did use sync 2021-04-13 22:50:56 This is is a bit change. when I ran blkid 2021-04-13 22:51:30 Now I am also getting /dev/sdb2 with vfat and /dev/sdb1 type "iso9660" for alpine 2021-04-13 22:54:27 I am retrying for another time 2021-04-13 22:56:33 Yes after another time still the same issue 2021-04-13 22:57:05 Probably will try again with the fresh usb later 2021-04-13 22:57:26 ah, well at least blkid output is reasonable 2021-04-13 22:57:39 what happens if you try to mount them manually? 2021-04-13 22:57:45 how 2021-04-13 22:58:25 I would like to try. afontain_ can you help recommend way 2021-04-13 22:58:45 maybe: mount /dev/sdb1 -t iso9660 -o loop 2021-04-13 22:59:06 ... shouldn't you add a mount point 2021-04-13 22:59:26 then maybe: mount /dev/sdb1 -t iso9660 -o loop /mnt 2021-04-13 22:59:56 crazy gnu extensions 2021-04-13 23:00:33 afontain_You mean in the grub right. 2021-04-13 23:00:33  I am in now 2021-04-13 23:00:43 not in grub 2021-04-13 23:01:13 at the "initramfs emergency recovery shell launched" stage 2021-04-13 23:01:24 hopefully you can type commands? 2021-04-13 23:01:28 (not sure) 2021-04-13 23:05:43 with that command I am getting cant find /dev/sdb2 in /etc/fstab 2021-04-13 23:06:35 with "mount /dev/sdb1 -t iso9660 -o loop /mnt"? 2021-04-13 23:08:59 ok i could mount it 2021-04-13 23:09:20 in there I am seeing three folders apks, boot and efi 2021-04-13 23:10:58 hmm 2021-04-13 23:11:12 well, I'm not sure what's the problem with your iso 2021-04-13 23:11:25 @fo 2021-04-13 23:11:32 @afon 2021-04-13 23:11:54 is the initial error still "Mounting boot media failed." 2021-04-13 23:11:57 ? 2021-04-13 23:12:21 Yes I am curious as well. 2021-04-13 23:15:03 Guys I have got to go. 2021-04-13 23:15:03 Very appreciate for your help today. 2021-04-13 23:15:04 I will try with the fresh usb. If this resolve later today. 2021-04-13 23:24:53 ... after cleaning, this keyboard is better than new one :/ 2021-04-13 23:25:14 it cost me 4 keys, so it's custom now :/ 2021-04-13 23:26:06 well, exactly 4 and half, as I need tu cut some part from one ;) 2021-04-14 00:23:55 Can anyone else access the docs? 2021-04-14 00:24:02 https://docs.alpinelinux.org/ 2021-04-14 00:24:07 I get a 404 error 2021-04-14 00:24:33 CAPTCHA_REQUIRED: working fine 2021-04-14 00:24:52 Really? 2021-04-14 00:24:57 Who maintains it? 2021-04-14 00:24:59 I just get 404 2021-04-14 00:28:44 CAPTCHA_REQUIRED: maybe try do "ctrl + shift + r" in your web browser 2021-04-14 00:30:00 I did refresh it a few times 2021-04-14 00:30:20 Oh hey 2021-04-14 00:30:23 It's back now 2021-04-14 00:30:25 That was weird 2021-04-14 00:30:30 Probably a transient error 2021-04-14 00:44:35 that was short, but interesting story btw ;0 2021-04-14 05:23:26 Dammit. 2021-04-14 05:23:28 https://github.com/jlesage/docker-makemkv/blob/master/makemkv-builder/README.md 2021-04-14 05:23:36 This is what I was was assuming. 2021-04-14 05:28:16 "The CLI, distributed as a binary built from private source code." 2021-04-14 05:29:34 so just a proprietary application 2021-04-14 05:56:37 ikke: Yes, but I still was reserving hope. 2021-04-14 12:29:54 i like ikke 2021-04-14 15:10:57 hi, when installing to a raspberry pi, following https://wiki.alpinelinux.org/wiki/Raspberry_Pi, traditional disk based (sys) installation, it's said to setup-disk -o /media/mmcblk0p1/MYHOSTNAME.apkovl.tar.gz /stage 2021-04-14 15:11:26 but there's an error: ext4 is not supported. Only supported are: vfat 2021-04-14 15:11:44 Did I miss or messed something? 2021-04-14 15:27:17 the /boot needs to be on fat due to limitation in rpi bootloader 2021-04-14 15:31:34 maybe setup-disk should print a message 2021-04-14 15:31:47 we've had this question like half a dozen times already 2021-04-14 15:59:25 sorry, but didn't understand what to do. /boot is a directory of a vfat filesystem of /dev/mmcblk0p1 mounted on /media/mmcblk0p1 2021-04-14 16:35:11 Hi all, I've been trying to figure our a good way to pull all CVE's for Alpine. Do you guys have a tracker of some sort just for security? 2021-04-14 16:35:38 xeon65: wait until friday ;-) 2021-04-14 16:35:47 (if all goes well) 2021-04-14 16:36:25 ikke: Not sure how to interpret that, are you saying you're releasing a tool just for that? 2021-04-14 16:36:51 xeon65: for now: https://gitlab.alpinelinux.org/alpine/aports/-/issues?scope=all&utf8=%E2%9C%93&state=opened&label_name[]=tag%3Asecurity 2021-04-14 16:36:55 xeon65: yes, exactly 2021-04-14 16:37:25 ikke, ah perfect, thanks for your quick response 2021-04-14 17:04:53 xeon65: also: https://secdb.alpinelinux.org/ 2021-04-14 17:27:54 I get a nginx 405 not allowed when logging in to the forum. I haven't used in a while. Is it not working only for me? 2021-04-14 17:30:51 Alpine Linux has a forum? 2021-04-14 17:30:55 had 2021-04-14 17:33:50 https://dev.alpinelinux.org/~clandmeter/other/forum.alpinelinux.org/forum.html#collapse-12 2021-04-14 17:35:39 funny :/ 2021-04-14 17:39:33 is is there some actual info about how to get proper chroot? in cases as broken bootloader or something like that 2021-04-14 17:39:56 last night I needed something like that :/ 2021-04-14 17:42:47 after like hour fuckering with lynx in wiki I found something and even it look like it works, it did not 2021-04-14 17:43:06 I mounted /, /boot/efi, sys, proc, dev 2021-04-14 17:43:13 not sure what need more 2021-04-14 17:45:05 also there is one annoying little thing during every boot kernel.unprivileged_bpf_disabled is unknown key 2021-04-14 17:46:00 I commented out that shit in /lib/sysctl.d/alpine.conf don't know if it's like no more used in 'edge' or where is the issue with it 2021-04-14 17:48:56 that forum btw. looks clean 2021-04-14 18:03:27 helby: did you have issues creating a chroot? 2021-04-14 18:03:45 mounting boot, proc, proc/sys and dev works for me 2021-04-14 18:04:19 hey, i was wondering, why does the /etc/init.d/localmount script have "keyword -lxc" set? that way it does not get executed in lxc containers, which is not very helpful 2021-04-14 18:04:59 ikke: yes, it 'worked' somehow for me at least grub or insalling anything did not show errors, however whatever I did I was still not able to boot, had to reinstall :/ 2021-04-14 18:05:32 ikke: the issue was also I was and still I am not what everything and how to mount ... it's like 5 lines, but that fucking info is nowhere for alpine 2021-04-14 18:11:30 The downloads page at https://alpinelinux.org/downloads/ has a "Mini Root Filesystem" version for "use in containers and minimal chroots". But after extracting that and extracting to a directory (such as ./alpine), and then installing openrc with `systemd-nspawn --directory=./alpine apk add openrc` and then "booting" the container with `systemd-nspawn --boot --directory=./alpine` gives errors such as "Service `hwdrivers' needs non 2021-04-14 18:11:30 Is there documentation for how the mini root filesystem is INTENDED to be used somewhere? 2021-04-14 18:11:30 existent service `dev'" and "can't open /dev/tty1: No such file or directory". 2021-04-14 18:13:27 curiousgeorge: afaik, systemd-nspawn requires systemd in the container, which does not work on alpine due to the use of musl 2021-04-14 18:14:21 curiousgeorge: an apparently /dev is not mounted in the chroot 2021-04-14 18:14:31 ikke: no it doesn't - it just has some convienence methods if the container also happens to have systemd (such as a service that will set up networking automatically). 2021-04-14 18:15:14 You need to make sure at least expected filesystems are mounted 2021-04-14 18:15:57 ikke: `man systemd-nspawn` states "Note that systemd-nspawn will mount file systems private to the container to /dev, /run and similar. These will not be visible 2021-04-14 18:15:57 outside of the container, and their contents will be lost when the container exits." 2021-04-14 18:16:18 did you verify this is the case? 2021-04-14 18:16:27 ikke: that is the benefit of systemd-nspawn vs chroot 2021-04-14 18:18:08 And the hwdrivers servers should not be required in containers 2021-04-14 18:18:11 service* 2021-04-14 18:18:58 ikke: `man systemd-nspawn` states "Use a tool like dnf(8), debootstrap(8), or pacman(8) to set up an OS directory tree suitable as file system hierarchy for systemd-nspawn containers". In other words, every other distro just works. 2021-04-14 18:19:08 https://tpaste.us/5VNE 2021-04-14 18:19:32 curiousgeorge: that does not include virtual filesystems like devfs procfs sysfs, etc 2021-04-14 18:20:00 They need to be mounted one way or another 2021-04-14 18:20:06 ikke: do you know how to mount them manually? 2021-04-14 18:22:58 curiousgeorge: and just to be clear, the minirootfs is mostly geared towards things like docker, where you don't 'boot' the system with init, so that indeed might require a bit more work 2021-04-14 18:24:50 ikke: systemd-nspawn does indeed mount /proc, /dev, and /sys. 2021-04-14 18:26:50 can you check the contents of /dev? 2021-04-14 18:27:10 it for example complains about /dev/tty1 2021-04-14 18:30:57 In order to get alpine linux to boot with `systemd-nspawn -b -D ./alpine` first `echo 'pts/0' >> /etc/securetty`, and then `echo 'console::respawn:/sbin/getty 38400 console' >> /etc/inittab` (both in the container filesystem) and comment out all the lines starting with 'tty/' in /etc/inittab. 2021-04-14 18:31:48 This gets it to boot, and you can see that the /dev, /proc, and /sys directories are populated, but it is still not correct, for instance `machinectl login alpine` or `machinectl shell alpine` don't connect correctly. 2021-04-14 18:32:59 https://github.com/lxc/lxc-templates/blob/master/templates/lxc-alpine.in#L277 2021-04-14 18:33:12 maybe you can find some answers there 2021-04-14 18:33:21 contents of /dev are: char fd net pts stderr tty 2021-04-14 18:33:21 console full null random stdin urandom 2021-04-14 18:33:21 core mqueue ptmx shm stdout zero 2021-04-14 18:34:14 /dev/tty1 is missing 2021-04-14 18:34:46 ikke: do you know how to create it? When booting other os with systemd-nspawn it seems to be there. 2021-04-14 18:35:30 hmm, not sure what is ought to create it in this case 2021-04-14 18:35:44 You see in that script I linked to how they are created for lxc containers 2021-04-14 18:35:57 but the man page mentioned you are not allowed to created device nodes 2021-04-14 18:37:38 does lxc run containers, or virtual machines (ie. with or without kernel)? 2021-04-14 18:37:52 containers 2021-04-14 18:38:35 In a similar fashion as what you try to do know with systemd-nspawn 2021-04-14 18:38:37 It is possible to just install alpine in a container the normal way, but I guess I was thinking of somebody took the time to extract the essence of Alpine Linux into a miniroot why not use that (especially since the .tar.gz is only ~2 MB). 2021-04-14 18:39:15 But I can't find any documentation about what they had in mind, or how they intended the miniroot to be used. 2021-04-14 18:39:27 curiousgeorge: base image for docker :) 2021-04-14 18:40:12 I truly don't see a point to Docker anymore - it is nothing but additional complexity (in my opinion). 2021-04-14 18:40:27 https://github.com/alpinelinux/docker-alpine/tree/v3.13/x86_64 2021-04-14 18:41:09 curiousgeorge: for us it's an easy way to deploy applications in a documented way 2021-04-14 18:41:44 I was big into it when it first came out, but then spent too much time fighting Docker and ultimately realized it is just an abstraction over kernal capabilities that systemd-nspawn exposes much more directly. 2021-04-14 18:42:19 Basically Dockerfiles + docker-compose is what's the benefit for us 2021-04-14 18:43:13 I understand about Docker being easy to deploy, but my experience running postgresql (for example) with the official docker image vs. just extracting the alpine miniroot, running `apk add postgrsesql postgresql-contrib` and then coping over my postgresql.conf file is just unnecessary headache. 2021-04-14 18:44:21 But in that example I am running postgres directly in the container - now I want to be able to run the complete openrc init system. That think about container images not coordinating (ie. one per service) was a neat trick by the Docker people to convince the world that a limitation of their system is a feature 2021-04-14 18:45:05 it is not - there is no reason not to run an init system in a container. I do it all the time with fedora in the container (systemd-nspawn) and it works great. Now I want to get more lightweight. 2021-04-14 18:45:26 sure, it's just one 'deployment' model 2021-04-14 18:45:56 ie, instead of running a binary on the host, run it 'isolated' in a container 2021-04-14 18:46:36 I really want to get to the point of booting the alpine miniroot with systemd-nspawn and using openrc as the init system. 2021-04-14 18:47:13 I understand 2021-04-14 18:47:40 But there is something about Alpine where it doesn't boot properly out of the box. Several other OS's such as Fedora, Ubuntu, debian do boot out of the box with systemd-nspawn. 2021-04-14 18:48:16 But also with the other OS I am using their bootstrapping systems (pacman, dnf, etc.) instead of a prepackaged miniroot. 2021-04-14 18:49:22 You can bootstrap with apk as well 2021-04-14 18:51:04 ikke: yeah - I somehow got it into my mind that the miniroot was purpose designed for containers so I should use that - but maybe I need to rethink that 2021-04-14 18:51:16 thanks for the help 2021-04-14 18:52:03 those other systems use systemd and are glibc based 2021-04-14 18:52:39 At the end of the day, I was thinking about the miniroot filesystem the way you think about the docker image (small, self contained, and fixed for each version). 2021-04-14 18:52:44 https://wildwolf.name/a-simple-script-to-create-systemd-nspawn-alpine-container/ 2021-04-14 18:55:52 ikke: yeah, I saw that. That is where I got those two modifications I wrote earlier. They are bootstrapping instead of using the miniroot, but I guess my basic question remains - why doesn't alpine boot out of the box without these modifications - and who/how to talk to the person that can get it updated to work smoothly? 2021-04-14 20:11:00 hi, is there some way to get apk to use metatadata/index from a specific date 2021-04-14 20:11:11 so I can have consistent results when I do apk install 2021-04-14 20:11:41 CustosLimen: sadly, we do not have the space to keep older packages 2021-04-14 20:11:53 I see, fair enough, just checking :)( 2021-04-14 21:12:21 hello, i've installed the armv7 version of alpine linux 2021-04-14 21:12:32 was having trouble installing packages 2021-04-14 21:12:41 i followed this page: https://wiki.alpinelinux.org/wiki/Raspberry_Pi 2021-04-14 21:13:43 however when i try to type something like "apk add vim" 2021-04-14 21:13:59 i get an error like "ERROR: unable to select packages: vim(no such package): required by: world[vim] 2021-04-14 21:14:24 i've read the wiki page on package management and wasn't able to find any solutions to this 2021-04-14 21:15:01 rob128: what does apk update return? 2021-04-14 21:15:35 3.31.5 [/media/mmcblk0p1/apks] OK:65 distinct packages available 2021-04-14 21:15:54 cat /etc/apk/repositories 2021-04-14 21:16:22 i had an issue with mirrors during the install 2021-04-14 21:16:28 that might be why 2021-04-14 21:16:28 oh 2021-04-14 21:16:33 let me get the output 2021-04-14 21:17:08 "/media/mmcblk0p1/apks" 2021-04-14 21:17:30 right, you need to have actual repositories in there 2021-04-14 21:17:33 yeah 2021-04-14 21:17:35 that makes sense 2021-04-14 21:17:43 so i should just pick a mirror link and add it in 2021-04-14 21:17:47 https://dl-cdn.alpinelinux.org/alpine/v3.13/main/ 2021-04-14 21:18:01 or yes, from any mirror you desire 2021-04-14 21:19:20 so if i wanted edge i'd just do https://dl-cdn.alpinelinux.org/alpine/edge/main/ 2021-04-14 21:19:53 without /main 2021-04-14 21:19:58 oh, sorry 2021-04-14 21:20:01 yes 2021-04-14 21:20:11 but, you should not mix edge and stable 2021-04-14 21:20:25 like debian 2021-04-14 21:20:26 right 2021-04-14 21:21:41 if u want even to find some packages, then also /community and /testing 2021-04-14 21:24:05 https://dl-cdn.alpinelinux.org/alpine/community/ and https://dl-cdn.alpinelinux.org/alpine/testing/ then? 2021-04-14 21:24:06 cool thing is to test all mirrors with setup-apkrepos then f to find one not on other side of eart 2021-04-14 21:24:10 h 2021-04-14 21:58:16 i ran setup-apkrepos and ran f for mirrors and half the repos on /etc/apk/repositories are /edge 2021-04-14 21:58:21 should i change to edge 2021-04-14 21:59:11 also should i use lbucommit bc i'm in diskless mode using persistent storage 2021-04-14 22:00:30 nvm the edge repos are commented out 2021-04-14 22:13:57 edge is edge - rolling, if you are in 3.13 or whatever number you are like in stable or whatever it is calling 2021-04-14 22:15:07 Hi! Just noticed over pavucontrol, that Alpine is not using my headset mic and is defaulting to use the laptop internal one. I've noticed that it detects that it's outputting the audio to 'Headphones' 2021-04-14 22:15:21 so u keep .../3.13/... or /edge/ whatever u want 2021-04-14 22:15:22 I wonder if there's anything related to some group or something on Alpine 2021-04-14 22:15:44 porcellis: "audio" perhaps 2021-04-14 22:15:47 i don't use pulse 2021-04-14 22:16:36 Yeah, I'm on plugdev and audio 2021-04-14 22:16:44 c705: alsa over jack directly? 2021-04-14 22:16:51 nope, just alsa 2021-04-14 22:17:08 I've tried just alsa on my arch install and couldn't make it work :( 2021-04-14 22:17:21 what did not work? 2021-04-14 22:17:24 i learned a very long time ago to leave all the fucking around with audio to external equipment 2021-04-14 22:17:31 alsa just works 2021-04-14 22:18:02 alsa didn't work with firefox for a while, but then they built support back in. besides that, never had any issues, but then again, i don't do much 2021-04-14 22:18:12 only issues people have are always wiht pulseshit 2021-04-14 22:18:19 indeed 2021-04-14 22:18:28 pulseaudio is the worst thing that ever happened to desktop Linux 2021-04-14 22:18:31 avoid 2021-04-14 22:18:39 Yeah, I'm having fuckton of issues with mic on a desktop, I sort of give it up 2021-04-14 22:18:42 well I know some other 'worst' thing ;) 2021-04-14 22:18:57 helby, I am willing to put pulse above all of those, and I know what you are referring to 2021-04-14 22:19:15 same author too 2021-04-14 22:19:36 yes, we're all putting systemd on that list... but there are other things 2021-04-14 22:19:57 i dunno, systemd kind of takes the cake 2021-04-14 22:20:02 of course, there is also grub :/ 2021-04-14 22:20:10 anyway, avoid pulse, and if you are on a desktop computer, consider getting a discrete soundcard with hardware mixing 2021-04-14 22:20:41 haha 2021-04-14 22:21:11 +100 for hardware mixing 2021-04-14 22:21:42 porcellis: did u even try just with default alsa and check things with alsamixer or amixer? 2021-04-14 22:22:15 I've removed pulse, will try on default alsa 2021-04-14 22:22:53 On my desktop, I bought a usb soundcard and audio keep working, but no fucking way to make mic work 2021-04-14 22:23:03 well for most poeple using speakers and headset ... usind discrete soundcard is ... pointless 2021-04-14 22:23:57 Not really, internal soundcards tend to be really meh 2021-04-14 22:24:04 what for? 2021-04-14 22:24:08 is it good practice to install all packages as root? 2021-04-14 22:24:16 for all these youtube what people listening? 2021-04-14 22:24:29 For listening to decent audio 2021-04-14 22:24:38 rob128: regular user can't install package I assume 2021-04-14 22:24:43 yeah 2021-04-14 22:24:51 rob128: You don't really have a choice, you have to install things as root 2021-04-14 22:24:52 i've heard sudo is bloated and insecure 2021-04-14 22:25:02 rob128: oh, check 'doas' 2021-04-14 22:25:04 considering installing doas 2021-04-14 22:25:05 yeah 2021-04-14 22:25:08 I use doas 2021-04-14 22:25:35 it's the first thing I install everywhere, then just permit nopass ... and I am okay 2021-04-14 22:26:01 not much fuckering with bullshit 2021-04-14 22:26:47 I keep my physical workstation secure by having blank keycaps 2021-04-14 22:27:27 yes, it helps agains grandma :) 2021-04-14 22:29:12 for regular people, it's enough to use tilling window manager without a bar or anything 2021-04-14 22:29:29 they will not figure out how to do anything on machine 2021-04-14 22:29:38 oh yeah... blank keycaps, tiling window manager... I'm set 2021-04-14 22:29:44 ;) 2021-04-14 22:30:02 blank keycaps are nice btw 2021-04-14 22:30:05 helby, what tiling window manager do you use? 2021-04-14 22:30:47 well, technically it's not tilling, but like hybrid :/ cwm 2021-04-14 22:31:07 100KiB little beast :/ 2021-04-14 22:31:26 damn, mine is 144 2021-04-14 22:31:30 I'm so bloated 2021-04-14 22:31:36 if there is no cwm, I would be probably on sway/wayland 2021-04-14 22:31:42 too big :/ 2021-04-14 22:31:45 what's that? 2021-04-14 22:31:48 spectrwm 2021-04-14 22:32:41 it will be probably simple like this one 2021-04-14 22:32:50 just that u can have windows in X :/ 2021-04-14 22:32:51 it's the right balance between dwm and i3 for me 2021-04-14 22:33:18 it does all the stuff I'd probably patch dwm to do if I wanted to spend my time doing that 2021-04-14 22:34:11 that's good 2021-04-14 22:35:39 sometimes I don't prefer only tilling, as practically every window is mostly fullscreen 2021-04-14 22:36:11 but sometimes I want little task, like update, check package or check mail, so I open a terminal, but don't wanna split it with anything let's browser 2021-04-14 22:36:23 u know just pop terminal, do shit, close it and continue 2021-04-14 22:36:45 yep... I do use floating windows for some stuff 2021-04-14 22:36:55 like the program I am spending 99% of my time building... 2021-04-14 22:36:56 obviously on bigger screen I would probably use some 100% tilling, but on 14' laptop 2021-04-14 22:37:27 what are u building? 2021-04-14 22:37:41 jejeje... the world will find out in a month or so :) 2021-04-14 22:37:44 Using tiling windows on laptop is a must have 2021-04-14 22:38:05 why? splitting that small screen in half? 2021-04-14 22:38:25 mostly, yeah 2021-04-14 22:38:29 most important is probably that's keyboard based 2021-04-14 22:38:56 That as well 2021-04-14 22:39:00 Most touchpads are shi 2021-04-14 22:39:04 *shit 2021-04-14 22:39:07 of course can't imagine now using classic shit with bar, panels and mouse menu shit 2021-04-14 22:42:38 it's addictive like vi styl key navigation 2021-04-14 23:25:40 Well, I can hear sound perfectly 2021-04-14 23:25:43 But no mic yet 2021-04-14 23:33:29 did u check alsamixer? 2021-04-14 23:35:18 u can check some shit about alsa on archwiki 2021-04-15 01:00:26 I've managed to make it work, although Firefox still says I'm not using anything 2021-04-15 01:00:35 I can use arecord to record audio and works perfectly fine 2021-04-15 01:00:38 ¯\_(ツ)_/¯ 2021-04-15 01:01:08 with alsa? 2021-04-15 05:14:58 trying to configure doas, the only line in /etc/doas.conf is permit persist :wheel 2021-04-15 05:15:17 however whenever i try using it as the user account i get the response "permission denied" 2021-04-15 05:15:47 bc. u no member of wheel group 2021-04-15 05:16:15 but u can put just line 'permit persist rob as root' 2021-04-15 05:16:19 aren't all users added members of the wheel group by default? 2021-04-15 05:16:28 I don't think so 2021-04-15 05:16:46 but if u only user, u don't need to fuckering with group 2021-04-15 05:17:08 doas is pretty simple, u can check man page for it 2021-04-15 05:17:50 like instead of persist if u put 'nopass' u don't have to write password like idiot ;) 2021-04-15 05:20:58 nah i prefer using persist 2021-04-15 05:21:58 u can limit it also to just certain commands, not everything 2021-04-15 05:45:48 im trying to setup an alpine xorg,xrdp,lightdm,xfce4 setup in a VM using virt-manager so to try to RDP in using xfreerdp 2021-04-15 05:46:40 i have everything installed and am able to start xrdp on the vm then connect using xfreerdp but I cannot get past the login screen (which is xord and not lightdm) .. am I going about this all wrong and expecting weird behavior? 2021-04-15 05:47:16 %s/xord/xorg/g 2021-04-15 05:48:14 i followed the alpine wiki with the setup-xorg-base script 2021-04-15 05:52:06 the xdm login screen? 2021-04-15 05:54:48 I can image what a nonsense is xrdp, however if u don't see lightdm or whatever display manager u decided to use, then you did not start it probably in services 2021-04-15 05:55:28 as normally you should not see console login if using display manager 2021-04-15 05:56:14 but virtual manager and xrdp crap with whole DE ... it can't work comfortably for a political reason ;) 2021-04-15 05:57:05 entropygain: are you just having trouble seeing what is being typed into xdm? 2021-04-15 06:17:30 i can see what is being typed but once i press ok nothing happens and the login screen just returns 2021-04-15 06:18:35 helby: i did try to start lightdm before trying to rdp and same thing 2021-04-15 06:18:37 xdm 2021-04-15 06:19:01 xdm should not start if u have lightdm I assume 2021-04-15 06:19:02 i haven't added it to services though so that it automatically starts 2021-04-15 06:20:15 the weirdest thing is when i start lightdm within the virt-manager graphical console the mouse works for about a second and then stops along with keyboard 2021-04-15 06:20:25 i assume that is a whole separate thing though and through xrdp it would work fine 2021-04-15 06:20:35 (input seems to work fine with xdm) 2021-04-15 06:20:47 i just can't seem to login though 2021-04-15 06:21:33 xfreerdp -> xdm -> type user/pass -> press ok -> screen freezes -> back to login screen/unfreezes (textboxes still have input) 2021-04-15 06:21:36 try to disable xdm and lightdm at all and use xinitrc 2021-04-15 06:22:17 also u can check ~/.local/share/xorg/... or /var/log/xorg something for x logs 2021-04-15 06:22:21 was going to say, that sounds like the behavior of XDM when .xinitrc has an issue. 2021-04-15 06:23:01 how do you disable xdm? it doesn't seem to be with rc 2021-04-15 06:23:03 also usually I add myself to audio, video, input, install some font, not sure if all is still ncessary 2021-04-15 06:23:15 it has to, if it started itself 2021-04-15 06:23:18 those being groups? 2021-04-15 06:23:31 not even sure how u install it, if u want to use lightdm with xfce 2021-04-15 06:23:34 well i have to be hoenst i don't know if it is using xdm .. i just assume that is the default xrdp dm 2021-04-15 06:23:48 oh 2021-04-15 06:23:48 i didn't excplicitly install xdm 2021-04-15 06:23:55 it's some xrdb, no idea what's that 2021-04-15 06:26:36 hmm Xorg.log has a bunch of "qxl_surface_create: bad bpp: 1" 2021-04-15 06:27:10 weird cause I have EnableSurfaces set to "False" 2021-04-15 06:29:57 not sure how that xrdp log, it can help to tell you why u can't login probably as it's most probably related to that xrdp 2021-04-15 06:30:23 pretty weird that something like that exist :/ 2021-04-15 06:32:26 error connecting to sesman 2021-04-15 06:32:49 that is what the xrdp.log is giving me 2021-04-15 06:33:15 that's something inside xrdp, so your issue will be probably with that, no xorg setting, but ... 2021-04-15 06:33:30 with trillions layers it's hard to say 2021-04-15 06:34:24 how would you tackle what I am trying to accomplish? 2021-04-15 06:34:44 no idea what you trying to accomplish 2021-04-15 06:34:50 can't you do it without xrdp? 2021-04-15 06:35:34 i just want to be able to run xfce4 on alpine and remote desktop into that 2021-04-15 06:36:00 i chose xrdp arbitrarily tbh is there something better i can make use of? 2021-04-15 06:39:04 oh rc-service xrdp-sesman start I think i am on to something 2021-04-15 06:47:17 i get a sesman connect ok now and login successful for display 10 2021-04-15 06:47:32 but it's just a teal screen 2021-04-15 06:47:50 it feels like i need to set up xrdp to use lightdm some how? 2021-04-15 06:48:23 not sure if that as to be even used if u logged in somehow 2021-04-15 06:48:32 probably xrdp take care of that ... 2021-04-15 06:49:00 maybe set up xrdp to login to xfce, really have no idea how xrdp works, but probably it uses own login nonsense 2021-04-15 06:52:45 ohai 2021-04-15 06:53:04 ohla 2021-04-15 06:53:09 someone has a idea why VBoxClient --vmsvga seems ineffective in an Alpine Linux guest under virtualbox ? 2021-04-15 06:53:20 ohla 2021-04-15 06:53:24 the service is also correctly running 2021-04-15 06:53:31 oops sorry wrong terminal up+enter 2021-04-15 06:59:30 oh looks like you need xf86-video-vbox which is only available through testing 2021-04-15 07:12:25 why do you think this is what i need? 2021-04-15 07:13:37 in the sesman logs what i get is "X server for display 12 startup timeout" and "another Xserver might already be active on display 12 - see log" 2021-04-15 08:12:17 entropygain: so how it is going? do u see at least rad on that teal screen? ;) 2021-04-15 08:12:45 rat 2021-04-15 16:14:52 Is it so ? I found many alpine wiki pages that rely on it. Could you tell more ? Can anyone confirm or infirm ? My conclusion so far is that alpine does support PAM auth and thus `limits.conf` with the `shadow` package, but postmarketos is using PAM in an unusual manner with its session starter: [tinydm](https://gitlab.com/postmarketOS/tinydm). 2021-04-15 16:34:31 install it and try it, if it's in repository, probably it has some purpose 2021-04-15 16:40:00 I guess you got me wrong - I _am_ using it inside my postmarketos install, where it comes as default. And I suppose `limits.conf` _does_ work in vanilla alpine. I just wanted to mention it so the conversation doesn't finish on "limits.conf not supported" 2021-04-15 16:45:43 what machine u use with postmarket? 2021-04-15 16:46:16 Samsung A3 2021-04-15 16:47:56 u can use it as phone? or just like display for some linux? 2021-04-15 17:02:50 I need it as a linux box ATM, so I haven't tested the calls and SMS things. For the rest it's pretty usable with phosh. Plasma works fine too, although not as lean. Even gnome works fine, though it doesn't deal with the screen's size so good. 2021-04-15 17:11:42 how can two most bloated DEs works well on that machine? 2021-04-15 17:12:36 3/4G would work with it? using regular sim 2021-04-15 17:13:28 thought about that, even I don't need a smartphone, but some extra tiny linux machine would be nice to have 2021-04-15 17:19:33 Yes, I was quite surprised too ; all of them ran quite smoothly. Snapdragon 410 _is_ a nice architecture, cheap and efficient. 3G and 4G do work with it. But I think it still would still "forget" to ring for calls and SMS if into crust (sleep) mode. 2021-04-15 17:23:10 well, but for mobile data, probably okay 2021-04-15 17:27:21 sure 2021-04-15 17:27:23 I would assume that using limits.conf and actually having it work would require that whatever login method the user is using be configured in PAM. 2021-04-15 17:27:57 last i checked, sudo isn't built with pam support 2021-04-15 17:28:02 not in alpine anyways 2021-04-15 17:32:40 seems correct 2021-04-15 17:33:00 su and login from the shadow package are though 2021-04-15 17:33:20 Yes, I installed `shadow` allright 2021-04-15 19:12:57 tar: ./bcm2711-rpi-4-b.dtb: Cannot change ownership to uid 1000, gid 1000: Operation not permitted 2021-04-15 19:12:57 tar: Exiting with failure status due to previous errors 2021-04-15 19:13:15 tar xvf ~/Downloads/alpine-rpi-3.12.0-aarch64.tar << with sudo 2021-04-15 19:13:25 am fallowing this guid https://wiki.alpinelinux.org/wiki/Raspberry_Pi_4_-_Persistent_system_acting_as_a_NAS_and_Time_Machine 2021-04-15 19:14:03 from what i see it copys stuff to sdcard just it fails at setting premmision/ownership right? 2021-04-15 19:14:24 isent that normal cuz its fat32 partrition? 2021-04-15 19:38:22 gnjurac: yes 2021-04-15 19:42:41 ok just wanted to confirm 2021-04-15 19:55:13 hi, i've got alpine installed on a machine without internet connectivity, however there's a few packages i would like to install from the repos, is there a way i can go about installing them? 2021-04-15 19:56:21 apk add `package` 2021-04-15 19:56:39 assuming that you have the apks avalible on your installation medium 2021-04-15 19:57:02 i don't have them available on the install media 2021-04-15 19:57:40 and you don't have internet access 2021-04-15 19:58:06 u have put them anywere u wish then install them with the same command 2021-04-15 19:58:33 funnyone: I guess apk fetch --recursive 2021-04-15 20:00:59 okay, thank you all 2021-04-15 20:34:19 hey funnyone grab the packages you want, put them on the alpine system anywhere and do `apk install --allow-untrusted /path/to/your/package.apk`. Beware that you will have to resolve dependencies manually 2021-04-15 20:35:34 --allow-untrusted is not necessary is they come from the alpine repos 2021-04-15 20:57:26 hello 2021-04-15 20:58:02 i have a question that i can't seem to find a definitive answer to via documentation alone 2021-04-15 20:58:21 would be very grateful for any amount of help 2021-04-15 20:59:23 i will just.. describe my issue. feel free to ignore 2021-04-15 21:01:56 i'm attempting to scrape the secdb json in order to identify the "fixing" versions of Alpine, but i'm not sure what that means in the context of vulnerabilities in upstream products 2021-04-15 21:02:45 no: What do you mean? 2021-04-15 21:03:36 this is harder to articulate than i anticipated 2021-04-15 21:04:11 i'm working to build a sort of system that communicates the version of Alpine that is affected by a vuln 2021-04-15 21:05:13 fyi, tomorrow we plan to launch a security tracker of our own 2021-04-15 21:05:30 but i don't think that is compatible with the entire package maintenance process that alpine has 2021-04-15 21:05:38 (wow, very good to know) 2021-04-15 21:06:14 i have an "Affected Products" table as well as an "Affected Packages" table that i'd like to populate using secdb 2021-04-15 21:06:33 (the product in this case being Alpine, with a corresponding version) 2021-04-15 21:06:54 right, but not sure if you can really call a specific version of alpine vurnerable 2021-04-15 21:07:36 the package piece of that is very straightforward. if a package's version's "secfixes" node contains a CVE ID, it's clear that the version of the package cited is the patched package 2021-04-15 21:07:43 but yes right 2021-04-15 21:08:00 once we updated the package, that version is no longer vulnerable 2021-04-15 21:08:15 fixes are backported to other versions of alpine all the time 2021-04-15 21:08:24 correct 2021-04-15 21:09:00 iso's might contain vulnerable versions, but one system upgrade later (which is part of the installation) and you should be fine 2021-04-15 21:09:49 for certain packages, we might make new releases, but for most packages, it does not make sense to create a new release 2021-04-15 21:10:27 that might be important. so, if i'm looking at https://secdb.alpinelinux.org/v3.10/main.json, for example, version 3.11 of Alpine (the major release) was likely vulnerable, hence the patch 2021-04-15 21:11:03 hm 2021-04-15 21:11:21 the major/minor version are just the twice per year 'major' releases 2021-04-15 21:11:44 3.10 was released 1.5 years ago, 3.11 1 year ago 2021-04-15 21:11:58 every so often, a patch releases is done 2021-04-15 21:12:01 3.10.x 2021-04-15 21:12:13 this triggers new iso's, images and docker images to be created 2021-04-15 21:12:28 is that what prompts posts like this? https://alpinelinux.org/posts/Alpine-3.11.6-released.html 2021-04-15 21:12:37 yes 2021-04-15 21:12:55 but that's certainly not required for people to get fixes 2021-04-15 21:13:06 you don't explicitly upgrade to 3.11.6 2021-04-15 21:13:08 right! there's always the option to manually upgrade 2021-04-15 21:13:17 the package itself, i mean (right?) 2021-04-15 21:13:31 yes 2021-04-15 21:13:51 just regular system upgrades gets you these updates 2021-04-15 21:14:00 (apk upgrade) 2021-04-15 21:14:06 got it 2021-04-15 21:14:40 given that information, i THINK i have a compromise in mind. do you mind if i run it by you as a sanity check? 2021-04-15 21:14:43 so the patch releases are not required for users to get security patches 2021-04-15 21:15:54 Feel free to discuss it here 2021-04-15 21:16:08 yes, that makes sense to me, but i'm struggling to communicate it effectively to a colleague 2021-04-15 21:16:58 .. 2021-04-15 21:17:02 sorry 2021-04-15 21:17:04 keyboard 2021-04-15 21:17:32 a distro is much more complex than just a single piece of software 2021-04-15 21:17:39 if a fix for a vulnerability appears in the secdb feed of a given Alpine version, it's safe to assume that natively, at least, the package repositories contained the vulnerable software. otherwise there would be nothing to patch 2021-04-15 21:17:52 yes 2021-04-15 21:17:59 that's correct 2021-04-15 21:19:23 so if i used the "Affected Product" table to store the version of Alpine that is/was natively affected, then use the "Affected Packages" table to store the fixed versions and packages, those fields would provide the information necessary for remediation 2021-04-15 21:20:16 if my major version (ignoring minor releases) is marked as affected, i should consult the affected packages table and upgrade to be safe 2021-04-15 21:23:51 yeah, if a certain version is marked as fixed, the package version before that can be considered vulnerable 2021-04-15 21:24:09 but we do not keep track of what version introduced the vulnerability 2021-04-15 21:24:37 that's totally understandable, and i'm not sure i really need that. as long as i know the fixed version 2021-04-15 21:24:43 i really appreciate your help 2021-04-15 21:24:51 so it could potentially be that version 1.5.2 introduced the vulnerability, and 1.5.3 fixed it 2021-04-15 21:25:01 but that means 1.5.1 is not vulnerable either 2021-04-15 21:25:24 right. assuming is not a good idea here 2021-04-16 00:06:41 hello, how can I have a command / commands run at boot (not a service)? how about at reboot/power down ? 2021-04-16 00:21:26 If it's not a service, you could use an @reboot entry in crontab to start something only on startup. 2021-04-16 00:22:17 /etc/local.d scripts are another option (see the readme file there) 2021-04-16 00:23:28 The latter option would also allow you to run it on shutdown 2021-04-16 00:24:47 zcrayfish: oh perfect, thanks. I went the /etc/local.d route and got what i wanted done :) 2021-04-16 00:25:15 awesome, no problem 2021-04-16 15:25:39 any clue for get nice libreoffice decoration? 2021-04-16 17:35:25 This is a long shot.. Has anyone issues with bittlbe_typing_notice.py plugin for WeeChat in Alpine? 2021-04-16 17:36:38 I think that plugin is broken 2021-04-16 17:37:02 Oh. :( 2021-04-16 17:42:05 I just tried the exact same setup on Gentoo and it works there. :/ 2021-04-16 17:47:52 I also noticed that I have an old version of WeeChat even though a newer one is available? 2021-04-16 17:48:00 According to the website. 2021-04-16 17:48:25 (https://pkgs.alpinelinux.org/packages) 2021-04-16 17:48:34 Hund: Are you on edge? 2021-04-16 17:48:41 No. 2021-04-16 17:48:58 Then you need to selct the version that you have 2021-04-16 17:49:33 Select version? 2021-04-16 17:49:42 Where it says "edge" 2021-04-16 17:50:21 Oh 2021-04-16 17:50:23 lol 2021-04-16 17:50:36 I assume it's the newest version? 2021-04-16 17:50:55 cat /etc/alpine-release 2021-04-16 17:51:08 3.12.7 2021-04-16 17:51:17 then you need to select 3.12 2021-04-16 17:51:21 Why don't I have 3.12? 2021-04-16 17:51:28 ? 2021-04-16 17:51:29 3.13* 2021-04-16 17:51:52 Because you did not upgrade to 3.13? 2021-04-16 17:52:12 I have installed every update that's available. 2021-04-16 17:52:18 You need to update the repositories in /etc/apk/repositories 2021-04-16 17:52:28 And how should I now that? 2021-04-16 17:52:31 ^ 2021-04-16 17:52:35 it's a text file 2021-04-16 17:53:03 You see v3.12 in the urls there 2021-04-16 17:53:09 You need to change it to v3.13 2021-04-16 17:53:24 after that, run apk upgrade --update --available 2021-04-16 17:53:27 I didn't ask how. 2021-04-16 17:54:05 I asked how I should be able to know when to change the repos if it doesn't let me know about it? 2021-04-16 17:54:09 https://wiki.alpinelinux.org/wiki/Upgrading_Alpine 2021-04-16 17:55:11 It doesn't say anything about that there though. 2021-04-16 17:55:30 apk does not tell you that you should upgrade to a new major version 2021-04-16 17:55:35 brb, I'm going to reboot the computer after the update. 2021-04-16 17:55:38 section 1.2 2021-04-16 18:00:30 Hund: new releases are anounced at several places 2021-04-16 18:00:32 https://alpinelinux.org/ 2021-04-16 18:00:51 https://lists.alpinelinux.org/~alpine/announce 2021-04-16 18:00:57 here 2021-04-16 18:00:59 twitter 2021-04-16 18:01:01 Thanks 2021-04-16 18:01:12 I should subscribe to the mailing list then. 2021-04-16 18:01:19 https://alpinelinux.org/releases/ 2021-04-16 18:02:33 Note that for the patch releases, you do not need to do anything 2021-04-16 18:09:13 or use edge and u get rid of that fuckering 2021-04-16 18:55:54 hi 2021-04-16 18:59:35 so i am trying to install alpine on rpi4 in persistant mode fallowing that wiki guide but when i hit setup-disk -m sys /mnt it complains it wont cuz its not vfat, how will i have premisions and stuff if its not ext4 and in guide it says to do mkfs.ext4 2021-04-16 19:00:36 At least the bootloadr needs to be vfat 2021-04-16 19:00:42 . 2021-04-16 19:00:52 bootloader is left vfat 2021-04-16 19:01:09 there is 2 partritions on sdcard 2021-04-16 19:02:13 . 2021-04-16 19:03:05 https://wiki.alpinelinux.org/wiki/Raspberry_Pi_4_-_Persistent_system_acting_as_a_NAS_and_Time_Machine 2021-04-16 19:04:05 first is vfat and second ext4, boot should go to vfat but want my root to go to ext4 2021-04-16 19:04:53 i dont really get whats the point of diskless intall 2021-04-16 19:05:17 It's mostly meant for server installations 2021-04-16 19:06:06 hmm but how can i host ngnix when it wont save 2021-04-16 19:06:47 1) there is lbu that persists changes 2021-04-16 19:07:10 2) you can still mount volumes that persist data\ 2021-04-16 19:08:41 so if i want to host this can work mount an external usb other partrition and put my html there 2021-04-16 19:08:51 sure 2021-04-16 19:09:22 did you actually mount /mnt/boot 2021-04-16 19:09:41 if you skip partitioning then setup-disk won't magically know where to mount stuff 2021-04-16 19:10:07 Hi there. 2021-04-16 19:10:34 I installed sway and after I kill it, I see this line: backend/session/logind.c:665 User has no sessions 2021-04-16 19:10:38 hmm i mount my second partitio. to /mnt guide says you should run on that 2021-04-16 19:10:41 diskless is nonsense, don't do it :/ 2021-04-16 19:10:44 And Coundln't find an active session or a greeter session 2021-04-16 19:10:51 How do I solve these two? 2021-04-16 19:11:05 If anyone is using X with alpine 2021-04-16 19:11:41 mount /dev/mmcblk0p2 /mnt thats what guide says 2021-04-16 19:11:49 frojnd: dbus and then run sway with dbus-run-session sway 2021-04-16 19:12:11 line before you do mkfs.ext4 2021-04-16 19:12:24 line after crys :( 2021-04-16 19:12:26 that also allow to open another win/tab in firefox from for example terminal 2021-04-16 19:13:17 so am i doomed 2021-04-16 19:13:37 formating other partrition to vfat sound bad 2021-04-16 19:14:30 also as wlroots it's probably still not build in alpine with seatd, if u don't want sway run as root, you have to use elogind crap 2021-04-16 19:14:49 helby: ok, I've started it like u said but now I get: XDG_RUNTIME_DIR is not set in the environment. Aborting. Before that I was running it: XDG_RUNTIME_DIR=/temp sway 2021-04-16 19:15:42 https://wiki.alpinelinux.org/wiki/Wayland 2021-04-16 19:16:12 u have to put that shit or temporary run just "XDG_RUNTIME_DIR=/tmp sway" 2021-04-16 19:16:16 i go try find some other guide 2021-04-16 19:17:12 Gnjurac: wiki is ancient, not everyting has to be correct :/ 2021-04-16 19:17:39 Gnjurac: you didn't answer if you mounted /mnt/boot 2021-04-16 19:18:32 i did not i only mounted second partritio. to /mnt 2021-04-16 19:18:55 should i mount boot into it? 2021-04-16 19:20:16 usually it's like ext4 to /mnt and fat to /mnt/boot or /mnt/boot/efi 2021-04-16 19:22:00 helby: not sure how to add that shit 2021-04-16 19:22:12 I've installed elogind but not sure how to configure it 2021-04-16 19:23:19 https://wiki.alpinelinux.org/wiki/Wayland maybe this will help 2021-04-16 19:24:18 Hello, For some reason grub-mkconfig insists on adding 'nomodeset' to the kernel command line, afaik this only happens to me on alpine 2021-04-16 19:24:40 I haven't really looked into the full working of the script but any leads where I should check? 2021-04-16 19:25:04 frojnd: I assume, u just have to enable it as service with 'rc-update add elogind add' 2021-04-16 19:25:39 if u want use just that script from wiki, u just copy it and create 'xdg_runtime_dir.sh' in /etc/profile.d/ 2021-04-16 19:26:26 how to mount boot when i dont see it mount /boot /mnt/boot , i manuly created boot dir on mounted p2 2021-04-16 19:27:10 yes 2021-04-16 19:27:11 hmm or do i first mount someere p1 then mount boot part to 2021-04-16 19:27:19 Thank you helby 2021-04-16 19:27:32 I really think alpine would need better wiki :/ 2021-04-16 19:27:34 mount firt ext4 to /mnt , create /mnt/boot then mount fat to /mnt/boot created 2021-04-16 19:28:41 Gnjurac: btw. why u just don't run 'alpine-setup' to skip all this nonsense? 2021-04-16 19:29:04 i did run that but thats for diskless setup 2021-04-16 19:29:27 oh, any reason for dkisless ? 2021-04-16 19:29:36 whay 2021-04-16 19:29:44 i dont want diskless 2021-04-16 19:29:52 want normal install 2021-04-16 19:29:59 alpine-setup is for normal installation 2021-04-16 19:30:19 even for diskless 2021-04-16 19:30:45 https://wiki.alpinelinux.org/wiki/Raspberry_Pi 2021-04-16 19:30:54 it will ask during installation what disk to use, if whole disk then if regular installation u choose 'sys' 2021-04-16 19:31:20 ok will reboot an try agai. wiki says its diskless only 2021-04-16 19:32:30 don't know if installing on raspberry is different then on x86... I doubt 2021-04-16 19:32:39 try normal run setup-alpine 2021-04-16 19:34:13 it asks where to store configs 2021-04-16 19:34:29 what is that step? 2021-04-16 19:34:31 i think rpi4 is diskless wiki says that 2021-04-16 19:34:47 probably ... 2021-04-16 19:34:48 its the step after ssh 2021-04-16 19:35:05 so no question about disk? 2021-04-16 19:35:38 nah just after that asks where to store apk cache 2021-04-16 19:36:15 yes, that's diskless 2021-04-16 19:36:21 sorry if mising letter but is hard to type on this phone 2021-04-16 19:37:11 anyone runnig rpi4 persistant? 2021-04-16 19:37:50 or i guess i can run it diskless but kinda dont get whats the point of that 2021-04-16 19:38:28 no point obviously :/ 2021-04-16 19:38:38 check wiki, there is something about that 2021-04-16 19:38:52 also probably ancient, but some boby talks about persisent 2021-04-16 19:39:21 but why is default to diskless ... pretty weird 2021-04-16 19:39:53 ye i started with that wiki quide guess can tru find fome newer 2021-04-16 19:40:08 or try void 2021-04-16 19:45:05 void has at least recent docs, relatively well written 2021-04-16 19:51:43 just still no offical rpi4 tar, there some w8ting on merge req 2021-04-16 19:52:55 so just continue where I stopped you with 'regular installation' :d 2021-04-16 19:54:58 I'm trying to install the openvpn package, https://pkgs.alpinelinux.org/packages?name=openvpn&branch=edge&repo=main 2021-04-16 19:55:26 according to this the latest version is "2.5.1-r0" however when I run "apk update && apk upgrade && apk install openvpn" I'm being served version 2.5.0-r1... 2021-04-16 19:55:30 does anyone know why? 2021-04-16 19:55:57 johnny56: cat /etc/alpine-release 2021-04-16 19:56:16 You are probably running alpine 3.13 2021-04-16 19:57:44 ikke: yes I'm running 3.13.4 2021-04-16 19:57:47 y? 2021-04-16 19:57:53 https://pkgs.alpinelinux.org/packages?name=openvpn&branch=v3.13&repo=main 2021-04-16 19:58:00 You looked at edge 2021-04-16 19:58:20 oh... 2021-04-16 19:58:48 alright I'll just switch to edge then tyty 2021-04-16 19:59:01 Is there a specific reason you need 2.5.1? 2021-04-16 20:00:02 yeah there are quite a few fixes in 2.5.1 including the memory leak 2021-04-16 20:00:06 and a few crashes 2021-04-16 20:00:24 ok, we can backport it to 3.13 2021-04-16 20:00:54 yeah if it's not too much hassle that'll be great just rebuilding my docker image with edge atm instead 2021-04-16 20:01:36 No, should be available soon 2021-04-16 20:02:04 cool thx! 2021-04-16 20:05:20 johnny56: https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/20549 2021-04-16 20:05:36 should normally vi from busybox use /tmp? like some kind of vi-recover or something like that what I remember ... I think on X it worked, but it looks no in wayland/sway 2021-04-16 20:06:10 it acts weird without it ... but can be session issue ...? 2021-04-16 20:07:05 it even does not 're-wrap' lines if window is resizes, not sure if that happened in X 2021-04-16 20:08:27 all I know is that its very limited 2021-04-16 20:08:43 even kakoune does not rewrap lines now, not sure why, vis does it well 2021-04-16 20:08:54 ikke: yes busybox vi is limited a lot 2021-04-16 20:09:13 but don't remember well, I think in X it worked well 2021-04-16 20:09:45 👍👍 2021-04-16 20:10:03 tried to 'make 2021-04-16 20:10:23 ' nvi but I am so 'pro' at building that I gave up soon ;) 2021-04-16 20:11:44 I think I miss some package or something, bc. this rewrapping was not issue before even in sway 2021-04-16 20:12:52 oh fuck I think I know ;) I builded wlroots without x11backend 2021-04-16 20:13:06 probably it use some shit about window size from there ... 2021-04-16 20:13:44 still some editor can deal with it 2021-04-16 20:39:17 hi what package provides the script(1) utility? 2021-04-16 20:42:06 oh ok util-linux 2021-04-16 20:42:22 it sucks that on https://pkgs.alpinelinux.org/contents the path cannot have an ending / 2021-04-16 20:42:25 or it does not find anything 2021-04-16 23:25:38 don't close my irc window pls ;) 2021-04-17 11:49:27 Hi all :-) 2021-04-17 11:49:54 What are the recommandations for packaging games on alpinelinux ? 2021-04-17 11:50:19 Do we prefer /usr/share or /usr/share/games folder ? 2021-04-17 11:51:01 (I've packaged both, and I want to normalize if possible) 2021-04-17 14:52:21 The typing notice issue with Bitlbee is solved. Now I have another issue with bitlbee thouhg. 2021-04-17 14:54:03 It can't save the settings because it runs as "nobody" which seems to have no permissions at all to write to the bitlbee directory. 2021-04-17 15:11:09 kmmndr: seems that `/usr/share/games` is used by a lot of games in Alpine Linux. I don't believe there are any hard rules on it though 2021-04-17 15:51:55 Hello all! i just installed alpine yesterday and sincerely, i love it, is just what i have been looking for for long time. I'm setting and installing what i need, and i like to ask you two questions, i installed and i have been using links browser, but i would like to use links2, i couldn't find it :-/ The other question, i installed tor, no tor browser, i preffer to use tor with links2 and the other programs. I do: #serv 2021-04-17 15:51:55 VirtualAddrNetwork 10.192.0.0/10 2021-04-17 15:51:55 AutomapHostsOnResolve 1 2021-04-17 15:51:55 Transport 9040 2021-04-17 15:51:55 art, it says is ok, and then when i check with #service tor status, it says status: crashed :-( I edited /etc/tor/torrc adding 2021-04-17 15:51:55 DNSport 53 Any idea what's happening? thank you. 2021-04-17 15:52:49 alpinista: your first message was cut-off at "the other programs. I do: #serv" 2021-04-17 15:54:34 oh! sorry for the mess, can't you read the rest of the message? 2021-04-17 15:55:30 alpinista: the next line starts with "art, it says is opk" 2021-04-17 15:57:13 service tor start, it says is ok, and then when i check with #service tor status, it says status: crashed :-( I edited /etc/tor/torrc adding VirtualAddrNetwork 10.192.0.0/10 AutomapHostsOnResolve 1 Transport 9040 DNSport 53 Any idea 2021-04-17 15:57:14 ppening? thank you. 2021-04-17 16:12:43 Hello all! i just installed alpine yesterday and sincerely, i love it, is just what i have been looking for for long time. I'm setting and installing what i need, and i like to ask you two questions, i installed and i have been using links browser, but i would like to use links2, i couldn't find it :-/ The other question, i installed tor, no tor browser, i preffer to use tor with links2 and the other programs. I do: servi 2021-04-17 16:12:43 VirtualAddrNetwork 10.192.0.0/10 AutomapHostsOnResolve 1 Transport 9040 DNSpot 53 Any idea what's happening? Thank you. 2021-04-17 16:12:43 rt, it says is ok, and then when i check with service tor status, it says status: crashed :-( I edited /etc/tor/torrc adding: 2021-04-17 16:27:38 uhM, if rc-status says crashed could be realted with /var/run problem 2021-04-17 16:27:57 are you running 3.13 ? 2021-04-17 16:28:07 alpinista: you don't need to paste your question (incorrectly formatted again) every 15 minutes 2021-04-17 16:32:49 sorry 2021-04-17 16:33:14 yes, 3.13 2021-04-17 16:35:17 thank you donoban, i'm going to check, thank you 2021-04-17 17:34:50 any way to have wine32 on 64 alpine? 2021-04-17 17:35:19 chroot 2021-04-17 17:35:29 there is no multilib support atm 2021-04-17 17:35:50 ok 2021-04-17 17:38:46 Hello71 were u able to get it going in chroot or u just say try it? 2021-04-17 17:40:45 well chroot obviously works 2021-04-17 17:41:46 ok 2021-04-17 17:43:45 you just have the limitations of chroot 2021-04-17 17:44:47 although actually with wine because you have drive mapping you can probably do like mount --rbind / chroot/mnt then bind X: -> /mnt 2021-04-17 17:46:26 i'll try it 2021-04-17 17:46:32 thx 2021-04-17 17:48:20 Is there some reason what discourage maintainers of packages to update them to the latest version, or it's just not their interest? 2021-04-17 17:48:39 some people are busy 2021-04-17 17:48:44 or how that shit works on alpine? 2021-04-17 17:48:47 some people are not aware that there are updates 2021-04-17 17:48:50 some people moved on 2021-04-17 17:49:01 some packages might have issues with newer versions 2021-04-17 17:49:03 all sorts of reasons 2021-04-17 17:49:31 I guess you are still complaining about firefox 2021-04-17 17:49:35 but these reasons are also on other distros, but somehow did not experience on rolling to be so long behind 2021-04-17 17:49:48 ikke: no firefox, there are many packages that are behind 2021-04-17 17:50:10 you could consider updating them yourself... 2021-04-17 17:50:11 mostly anything I check is behind 2021-04-17 17:50:28 helby: maybe you should start maintaining some packages 2021-04-17 17:50:31 I did, but I m not much fun of building myself as it polute my system with a lot of crap 2021-04-17 17:50:32 and by doing so you could get a feel how easy fun this is... 2021-04-17 17:50:36 then you can make sure they are up-to-date 2021-04-17 17:50:45 you can build everything in clean chroots. 2021-04-17 17:52:20 still it's not eficient that many people will build them on their machine if these can be build once for everyone 2021-04-17 17:53:15 well, that's why you should contribute back your updates! 2021-04-17 17:53:23 are you fishing for excuses now? 2021-04-17 17:53:29 helby: sometimes a maintain may build a new version and tests it locally for a while to make sure it works properly before doing an MR 2021-04-17 17:53:31 nope 2021-04-17 17:54:02 how that works? if I send bump version back, then maintaner will have time? :/ 2021-04-17 17:54:32 as many times it's just change version number 2021-04-17 17:54:46 if you test it out that your bump version indeed compiles and works, the maintainer will be happy and probably will have time to merge yes. 2021-04-17 17:54:46 There are volunteers watching the list of merge requests 2021-04-17 17:55:05 so even if the maintainer is absent, trivial upgrades still get merged 2021-04-17 17:55:11 oh okay 2021-04-17 17:55:12 but bumping a version is often not enough. 2021-04-17 17:55:17 no? 2021-04-17 17:55:22 depends 2021-04-17 17:55:32 well, take for example libreoffice 2021-04-17 17:55:44 well, that's a very complex suite of software 2021-04-17 17:55:47 if you manage to get it to the latest version, you'll understand my words. 2021-04-17 17:55:49 I am not talking about browser or libreoffice :/ that I will not even try 2021-04-17 17:55:52 one of the packages I maintain did 3 releases in 2 days recently......hard to keep up with that 2021-04-17 17:56:00 minimal: I know right :) 2021-04-17 17:56:48 ACTION maintains a package with no release in over a decade 2021-04-17 17:57:13 huh i did two new releases today for the same sw. the first release was by accident, before pushing the changes... /o\ 2021-04-17 17:57:26 I generally like to give a new release of something a couple of days before creating a MR in case there's any issues/breakages spotted upon release 2021-04-17 17:57:48 c7s: thanks for keeping up so long up to date ;) 2021-04-17 17:57:58 it 2021-04-17 17:58:04 helby: are you using edge? 2021-04-17 17:58:21 yes edge 2021-04-17 17:58:34 rolling does not mean bleeding edge 2021-04-17 17:58:49 yes, but edge means it :/ 2021-04-17 17:59:16 I just want to understand how it works in alpine 2021-04-17 17:59:35 helby: are you sure about that? 2021-04-17 17:59:36 as I experience not much edge :/ 2021-04-17 18:00:00 […]edge" is the name given to the current development tree of Alpine Linux. It consists of a APK repository called "edge" and contains the latest build of all available Alpine Linux packages 2021-04-17 18:00:02 helby: You probably ignore the numerous packages that are updated very quikly 2021-04-17 18:00:11 if i run host alpine diskless, is it ez to have chroot on a disk? 2021-04-17 18:00:24 ekhas: sure 2021-04-17 18:00:36 ok 2021-04-17 18:00:52 ikke: I am not complaining, more like asking 2021-04-17 18:01:32 helby: the goal is to have packages up-to-date, but some things are not immediately updated for reasons 2021-04-17 18:01:36 software is complex 2021-04-17 18:01:41 things are interdependent 2021-04-17 18:02:02 we build for 8 arches 2021-04-17 18:02:04 you want _edge_? go git everything and have fun 2021-04-17 18:03:04 this machine would die if I start to build firefox :/ 2021-04-17 18:03:56 helby: there are some packages where upstream has a stable release (e.g. 1.6.x) and a development release (e.g. 1.7.x) and the Alpine maintainer only packages stable release versions and people see the package is far behind what they perceive to be the latest version (and sometimes they submit MRs to update to what is a dev version) 2021-04-17 18:04:00 ikke: is it worth to bump versions of some packages then sent merge or it's more likely just annoying for maintainer if he is not doing it himself? 2021-04-17 18:04:23 minimal: yes, that I understand 2021-04-17 18:04:37 helby: at least make sure it builds, then it should not matter 2021-04-17 18:05:05 ikke: then I sent just merge back? 2021-04-17 18:05:20 You open a merge request in gitlab 2021-04-17 18:06:56 in gitlab? like from web? :/ 2021-04-17 18:07:25 You typically first clone aports, update the package there, build it 2021-04-17 18:07:33 yes 2021-04-17 18:07:43 then you push your branch to your fork and open a merge request 2021-04-17 18:07:50 which could be from the webinterface 2021-04-17 18:08:16 normally with git it will works ye? 2021-04-17 18:08:24 without web 2021-04-17 18:08:37 for making merge request, you cannot just use plain git 2021-04-17 18:08:44 but there are command line tools that can make merge requests 2021-04-17 18:08:47 aha 2021-04-17 18:08:50 there's also a mailing list 2021-04-17 18:08:58 yes, that dead list 2021-04-17 18:09:06 i thought there is a cli tool to create mrs/ 2021-04-17 18:09:11 wdtTP2O82Kft: yes 2021-04-17 18:09:14 glab for example 2021-04-17 18:09:51 I think checking wiki in regards of this is useless ... 2021-04-17 18:09:53 the aports mailing list is not dead, you can send patches there 2021-04-17 18:09:56 'glab mr create --remove-source-branch --allow-collaboration --fill --head kdaudt/aports --yes' 2021-04-17 18:09:56 or is there some recent info? 2021-04-17 18:10:13 https://wiki.alpinelinux.org/wiki/Creating_patches#Creating_a_merge_request 2021-04-17 18:11:52 PureTryOut[m]2: thanks for your answer :-) 2021-04-17 18:13:57 ghub was bought by ms. who owns glab? 2021-04-17 18:14:08 Still independent 2021-04-17 18:14:10 also ms 2021-04-17 18:14:16 ms owns the entire internet if you didn't hear 2021-04-17 18:14:27 this chat is now copyright 2021 microsoft 2021-04-17 18:14:31 well google owns also something 2021-04-17 18:14:35 with amazon 2021-04-17 18:15:09 mention all three and it almost looks like a conspiracy :D 2021-04-17 18:15:17 :eyes: 2021-04-17 18:18:23 at least alpine is not primary on github 2021-04-17 18:22:02 We used to be :P 2021-04-17 18:29:26 ye? long time? 2021-04-17 18:29:41 well everyone was there mostly some time 2021-04-17 18:30:31 ekhas: you could also make init script to create chroot from apk cache 2021-04-17 18:32:25 Hello71 yea im reading the alpine chroot wiki as im a total noob. but im saving ur suggestions to a txt file 2021-04-17 18:32:36 anyway today I have to make 'beep' working in sway 2021-04-17 18:38:05 i noticed the default mount -t ntfs in alpine is read only? is it cuz its busybox? 2021-04-17 18:38:32 adding ntfs-3g works but im curious 2021-04-17 20:07:07 we should probably remove CONFIG_FS_NTFS too 2021-04-17 20:07:17 arch did a couple months ago 2021-04-17 21:56:46 What do you use to get glibc stuff working (when you need to) chroots are very messy imo 2021-04-17 22:18:25 gcompat sometimes works 2021-04-17 22:22:24 Hello71: it only worked once for me :( I tried to get alpine-pkg-glibc working but that really didn't at all 2021-04-17 22:23:00 I wonder if it's not to difficult to create something like akin to how bedrock linux works 2021-04-17 22:23:43 or maybe a binary uses namespaces to create bind mounts to replace the paths 2021-04-18 02:09:40 Hello71: what is gcompat 2021-04-18 02:10:30 CAPTCHA_REQUIRED it's a shim for musl-libc that lets programs use glibc-specific features basically 2021-04-18 02:10:51 gcompat is to glibc like wine is to windows 2021-04-18 02:12:11 Oh so you run it like gcompat program 2021-04-18 02:12:24 i think so (i havent personally used it yet) 2021-04-18 02:12:26 Can it run steam games? 2021-04-18 02:12:40 I don't mind a chroot for that and even prefer it 2021-04-18 02:13:09 Question is though, can you give a container access to your X11 session? 2021-04-18 02:13:40 https://code.foxkit.us/adelie/gcompat 2021-04-18 02:13:47 guess im wrong here, musl seems to be linked against gcompat 2021-04-18 02:14:03 yeah you can use chrooted programs with another xorg session 2021-04-18 02:14:23 just bind-mount your /tmp in the chroot, and use xhost accordingly 2021-04-18 02:14:26 brb 2021-04-18 02:14:41 (/tmp is what has the xorg connection socket) 2021-04-18 02:17:08 musl is not linked against gcompat 2021-04-18 02:17:19 the reverse is true (on alpine anyways) 2021-04-18 02:21:22 Is there an easy to use script or something that sets up a steam chroot automaticlly? 2021-04-18 02:40:32 i have a system at home im about to decommission because it's getting old and having issues, so i may experiment with gcompat on it before completely taking it offline 2021-04-18 02:41:28 >If the values specified during build were correct, programs built for glibc 2021-04-18 02:41:29 should run normally on your musl host. 2021-04-18 02:41:33 oh i guess i misread the readme 2021-04-18 02:41:43 so ideally it "just works" 2021-04-18 04:36:00 Hello! 2021-04-18 04:36:42 That sounds like a superhero name 2021-04-18 10:19:39 hey,is there a git repo for Alpine site ? the docs link on the main site (docs.alpinelinux.org) is 404ing. 2021-04-18 10:21:35 lists.alpinelinux.org is also 404ing 2021-04-18 10:30:42 jeff__: they work fine for me 2021-04-18 10:31:03 they are back now for me too 2021-04-18 10:31:08 ah ok :) 2021-04-18 17:33:37 Good evening 2021-04-18 17:34:14 o/ 2021-04-18 17:34:33 I've another interesting case in which ifupdown-ng fails with incorrect parsing, where busybox does the right thing. I've seen a dependency in alpine-base that requires ifupdown-ng (seems to be removed in the current version) - what is the plan for ifupdown-ng at the moment? 2021-04-18 17:35:26 Also I found another scenario in which the DAD detection script creates in infinite loop: when there is no link on an interface 2021-04-18 17:36:08 I think we should really remove that script, as the kernel has internal DAD detection and this script does not give additional value, but breaks booting systems from time to time 2021-04-18 17:36:32 (with the latter I am referring to /etc/network/if-up.d/dad) 2021-04-18 17:38:42 telmich: ifupdown-ng is the default now 2021-04-18 17:38:47 (it's pulled in by openr) 2021-04-18 17:38:50 openrc* 2021-04-18 17:39:09 grr 2021-04-18 17:39:33 If there are problems, they should be fixed 2021-04-18 17:40:19 So this interfaces file fails to boot with ifupdown-ng: https://www.nico.schottelius.org/temp/interfaces_apurouter1p5 2021-04-18 17:40:29 But it works with Debian and with busybox ifupdown 2021-04-18 17:41:29 You should open an issue on https://github.com/ifupdown-ng/ifupdown-ng 2021-04-18 17:41:44 Will do 2021-04-18 17:42:49 Interesting, seems vlan and ifupdown-ng are conflicting 2021-04-18 17:43:49 ... almost get excited that can remove another useless package :/ 2021-04-18 17:44:10 https://wiki.alpinelinux.org/wiki/Release_Notes_for_Alpine_3.13.0#Switching_from_busybox_ifupdown_to_ifupdown-ng 2021-04-18 17:44:25 ifupdown-ng has native vlan support 2021-04-18 17:45:06 ok... shortly breaking my system for recreating the error 2021-04-18 17:46:26 telmich: It seems ifupdown-ng also has native mtu statements 2021-04-18 17:46:37 https://github.com/ifupdown-ng/ifupdown-ng/blob/cc067126118f36193130da3827184129c3dfaa00/tests/fixtures/vlan-complex.interfaces#L8 2021-04-18 17:47:16 Hmm, I might have run into a problem of an older ifupdown-ng that vanished with `apk upgrade -a` 2021-04-18 17:47:17 Nice! 2021-04-18 17:47:51 right, you do want the latest ifudown-ng version indeed 2021-04-18 17:49:11 and, if you want, you can use cidr notation for the netmask 2021-04-18 17:49:24 for ipv4 as well 2021-04-18 17:50:29 And you can combine ipv4/ipv6 2021-04-18 18:00:52 That actually looks nice 2021-04-18 18:05:57 You kinda convinced me that ifupdown-ng is a good idea just by seeing the syntax 2021-04-18 18:06:13 Guest46799 do you use openvpn or protonvpn app? 2021-04-18 18:06:26 Funny, I just see the bridge-vids, and bridge-pvid that I only saw on mikrotik before 2021-04-18 18:07:09 interfaces-bridge(5), that is quite interesting. It is a very "funky" syntax, if you compare it to actual switches. I was not aware that this is an upstream Linux feature 2021-04-18 18:08:15 Because in the switch world, an interface is access port x or can be tagged with vlans a, b, c and the bridge `bridge-vids` syntax basically defines that on a per bridge basis 2021-04-18 18:08:22 As opposed to per interface 2021-04-18 18:09:33 telmich: I understood that the bridge interface on linux is a layer 3 interface connecting to the actual to the layer 2 bridge 2021-04-18 18:09:47 Gosh, there is bridge (as a command) now, which actually looks interesting: https://unix.stackexchange.com/questions/556735/linux-vlan-aware-bridges-and-trunk-ports 2021-04-18 18:10:04 telmich: brctl is old :P 2021-04-18 18:10:40 iproute2 handles it now 2021-04-18 18:12:12 telmich: yeah ifupdown-ng works well with "mtu", "vlan-raw-device" and "vlan_id" directives, been using those for a while now 2021-04-18 18:13:25 Thank you both for the updates - I need to see that the routers get an up-to-date ifupdown-ng everywhere then 2021-04-18 18:15:20 So time moves on. First ifconfig deprecation, then brctl is old, I assume vconfig is also deprecated - recently I learned that bird2 has radvd support, it gets pretty crazy these days 2021-04-18 18:16:12 haha, yes, things change indeed 2021-04-18 18:16:43 don't forget "ss" taking the place of "netstat" 2021-04-18 18:16:59 I'm actually quite happy with nft, which is much more sane compared to iptables 2021-04-18 18:17:29 I haven't looked into nftables yet 2021-04-18 18:17:58 minimal: ss is one of those things that really did not go into my mind. In German, tulips are "Tulpen" and you usually learn `netstat -tulpn` as a starter 2021-04-18 18:18:12 telmich: Same in duch 2021-04-18 18:18:16 but ss should support that as well 2021-04-18 18:18:35 ss -tulpn 2021-04-18 18:21:00 can 'ss' do same what 'netstat -s' doing? 2021-04-18 18:25:16 MY-R: https://tpaste.us/nWLv 2021-04-18 18:27:03 telmich: in recent Debian version iptable tools are actually using nftables behind the scenes. 2021-04-18 18:29:39 ikke: https://tpaste.us/zxDW 2021-04-18 18:33:22 Cannot find an option that gives exactly that info 2021-04-18 18:34:49 ye, that is why 'ss' doesnt replace 'netstat' for me :\ 2021-04-18 18:34:56 minimal: true, I've seen that and the transition was a bit bumpy, when you dist-upgrade and then have nftables by default, but ip(6)table rules still installed 2021-04-18 18:36:38 MY-R: I assume this info is hidden inside /proc or /sys and netstat is just extracting it from there 2021-04-18 18:40:32 and doing it very well 2021-04-19 11:59:27 Are entries in /run/openrc supposed to stay after reboot? I was expecting them to be cleaned up, but they aren't. 2021-04-19 12:09:10 I was thinking all of /run was tmpfs . . . 2021-04-19 12:12:33 Yeah it is.. But I have files that dates from 2020 there hum.. 2021-04-19 12:13:59 any chance your hardware clock is set wrong, giving those files old dates before NTP kicks in? 2021-04-19 12:14:08 that's my wild guess :) 2021-04-19 12:16:19 That would be a good explanation. It's raspberry pi and afaik there's not hwclock 2021-04-19 16:10:32 tomleb: have you the swclock init.d service enabled? 2021-04-19 16:16:36 minimal: I don't, let me try that 2021-04-19 16:17:30 hi all, i am attempting to install alpine on a raspi4, and confused regarding the arch 2021-04-19 16:17:40 i believe raspi4 supports aarch64, but all alpine howtos say use armhf 2021-04-19 16:17:50 is that bc the doc is slightly out of date, or is that bc aarch64 doesnt run yet 2021-04-19 16:17:58 aarch64 should work 2021-04-19 16:18:18 ye, aarch64 working great on rpi4 2021-04-19 16:19:26 how are you working around this? https://github.com/raspberrypi/firmware/issues/1529 2021-04-19 16:19:44 i dont have uart, but my raspi says cant find vmlinuz, and indeed there is none in the tarball 2021-04-19 16:19:54 i don't know if it's the same issue, but i'm a bit lost right now tbh 2021-04-19 16:20:09 first time alpine 2021-04-19 16:20:27 i am attempting this: https://wiki.alpinelinux.org/wiki/Classic_install_or_sys_mode_on_Raspberry_Pi 2021-04-19 16:20:44 tomleb: as the RPI has no hardware RTC then its timekeeping is "problematic" - eve using ntp to sync time the clock will be wrong at boot until NTP sync occurs and so some files will have incorrect timestamps 2021-04-19 16:21:28 minimal: Yeah, right now the files on /run/openrc have a date of August 2020. swclock didn't help with that. 2021-04-19 16:22:23 yeah, swclock is a fudge to kick in before NTP occurs 2021-04-19 16:22:43 tomleb: which files have that date? they've probably being copied into /run from elsewhere during boot 2021-04-19 16:23:30 ah, /run/openrc - yes they are copied into place. OpenRC keeps its services dependancy "database" and some other stuff in /run/openrc (from memory) 2021-04-19 16:23:49 For example, /run/openrc/options/alertmanager/start_time, which also contains the wrong date as the start time. 2021-04-19 16:26:56 There's the savecache service that saves the following files "depconfig deptree rc.log shutdowntime softlevel" to /lib/rc/cache, but I couldn't find anything else which would backup the options/* files for example 2021-04-19 16:28:28 tomleb: I assume you see "clock skew" messages during boot - that's "normal" for a RPI due to lack of RTC 2021-04-19 16:28:50 And in my /media/mmcblk0p1/*.apkovl.tar.gz I don't see anything under either /var/run and /run. 2021-04-19 16:29:22 maybe it using date from initramfs files/dirs 2021-04-19 16:30:01 Where would those be located ? 2021-04-19 16:30:25 (In case it makes any difference, I'm using diskless installation right now) 2021-04-19 16:30:36 tomleb: what is your Alpine version installed and "uname -a" 2021-04-19 16:31:20 Alpine v3.12, Linux soju 5.4.84-0-rpi4 #1-Alpine SMP PREEMPT Wed Dec 16 11:53:58 UTC 2020 aarch64 Linux 2021-04-19 16:31:24 tomleb: it wont change anything if you wont provide correct time before services start 2021-04-19 16:31:36 yeah makes sense 2021-04-19 16:32:04 tomleb: you're using Alpine in diskless mode so the timestamp comes from the files inside the original packages, I'm more familiar with sys mode where I've installed Alpine on a SD card and so timestamp there can change 2021-04-19 16:32:15 tomleb: and looks like you not updating your kernels but probably only system 2021-04-19 16:33:14 tomleb: /var/run and /run are created at system startup, there are not contained in any packages but when openrc starts up it may copy files into there 2021-04-19 16:37:14 minimal: Hum, interesting. I had to install wireguard-rpi4 to be able to use wireguard on this server. iirc it installs a new kernel with support for wireguard.. Maybe that's why I have an old kernel 2021-04-19 16:37:45 tomleb: all my RPIs have RTCs fitted as there's no other way to have the system time correct all the time otherwise - even with NTP enabled the clock will be wrong early in the boot sequence (and so in any logfiles written then) 2021-04-19 16:39:29 minimal: Is there no packages that keeps the last known time and set the time to that on next boot? Otherwise I'll just have to live with wrong dates.. Or maybe there's a way to wait for NTP to be ran before starting a service 2021-04-19 16:39:39 tomleb: you didnt install any new kernel, in diskless mode you have to unpack firmware/dts/kernels etc from rpi4 tarball to /media/mmc*/boot and replace it 2021-04-19 16:40:48 tomleb: swclock service should do that 2021-04-19 16:40:54 MY-R: Yeah, that's done with wireguard-rpi4 (and a little script to mount rw /media/mmc*/boot, symlink /media/mm*/boot to /boot, install wireguard-rpi4, remount ro /media/mmc*) 2021-04-19 16:41:48 tomleb: but also you have stuff in /media/mmc*/ dir which should replace too, apk wont do it 2021-04-19 16:42:52 ikke: swclock service doesn't fix time-related issues completely - openrc itself uses the system time before it starts any init.d services (including swclock) and so the time jumps during booting which causes those "clock skew" messages during boot and shutdown on a RPI 2021-04-19 16:43:18 minimal: was just replying to what tomleb said 2021-04-19 16:43:25 "Is there no packages that keeps the last known time and set the time to that on next boot? " 2021-04-19 16:45:31 ikke: I'd already pointed out swclock to him - it is fine if you reboot but don't work (correctly/accurately) if you shutdown and then boot again later as it (from memory) sets the time on boot using the timestamp of one of openrc's data files - if you reboot immediately the time will be basically correct (several seconds slow) but if you shutdown, and wait (e.g.) 20 minutes before booting then the system time will be 20 minutes slow 2021-04-19 16:46:24 yes, ofcourse, that's inherent to how swclock works 2021-04-19 16:46:43 without an RTC, that's about best you can get after boot 2021-04-19 16:46:55 until ntp can correct it 2021-04-19 16:47:27 minimal: In my case that'd be good enough really. Better than seeing timestamps 1 year old. Let me try rebooting again. 2021-04-19 16:47:47 (Unfortunately, znc runs on that rasp pi, so I might miss some messages, brb) 2021-04-19 16:47:52 The alternative is to let the clock start from epoch until ntp kicks in 2021-04-19 16:47:54 yeah its a fudge. I did create a patch for openrc to "improve" this fudge more but decided it wasn't worth submitting a MR. 2021-04-19 16:50:54 isn't the current alternative setting it to the / mtime? 2021-04-19 16:51:15 instead I got the RTC drivers built into the RPI kernels so that with a 3rd party RTC addon fitted the system time is set once the kernel is loaded 2021-04-19 16:51:45 Hello71: isn't that basically what swtime does? 2021-04-19 16:51:52 description="Sets the local clock to the mtime of a given file." 2021-04-19 16:52:08 i think swclock additionally touches a file on shutdown 2021-04-19 16:52:20 right 2021-04-19 16:52:23 you can do one without the other 2021-04-19 16:54:06 yeah as it touches a file that's why I said if you reboot the time will be roughly correct whereas if you shutdown and later boot it will be wrong (i.e. set to the time of last shutdown) 2021-04-19 16:54:57 Isn't that logical? 2021-04-19 16:55:20 Ok I'm back. I have manually saved the clock with swclock --save and rebooted but it looks like it didn't properly saved the clock. I'm still seeing Aug 2020. Do we know where the clock is saved ? 2021-04-19 16:55:23 without an rtc, there is no way have accurate time on boot 2021-04-19 16:56:11 yes, I was just pointing out that it means early log entries will have incorrect timestamps and the time may jump during boot (e.g. NTP sync) and may cause "clock skew" boot/shutdown messages 2021-04-19 16:56:25 #define RC_SHUTDOWNTIME RC_SVCDIR "/shutdowntime" 2021-04-19 16:57:41 ikke: agreed, hence why I have RTCs fitted. however until I got the RTC driver built into the kernel (rather than a module) there were still issues with a RTC as the time was wrong until the kernel module was loaded and hwclock init.d ran to set system clock from RTC 2021-04-19 16:58:04 yeah, you would need to have something on very early boot 2021-04-19 16:58:33 without swclock, logs would jump from 1-1-1970 2021-04-19 16:58:37 to current time 2021-04-19 16:59:18 no as the RTC driver is compiled into the kernel as as the HWTOSYS (or whatever its called) kernel option is on its the kernel itself that sets the system clock from the RTC very early in boot 2021-04-19 16:59:19 in diskless mode could at least use date/time from .apkovl.tar.gz file :P 2021-04-19 16:59:37 just a quick question. the installer's "sys" mode is the same as the documentation's "diskless", yes? 2021-04-19 16:59:47 opv: no 2021-04-19 16:59:53 no, sys mode is regular disk mode 2021-04-19 17:00:01 and 'data' is diskless 2021-04-19 17:00:02 opv: sys is where you install Alpine onto a "disk" 2021-04-19 17:00:03 diskless is "none" mode if I remember correctly 2021-04-19 17:00:11 opv: data is hybrid 2021-04-19 17:00:18 diskless, + /var mounted 2021-04-19 17:00:40 so shutdowntime is saved to /lib/rc/cache/shutdowntime when /etc/init.d/savecache runs. So with diskless alpine, this file isn't kept by default when running `lbu commit -d` 2021-04-19 17:01:06 Which is why to me it appears that swclock doesn't work 2021-04-19 17:01:22 tomleb: best would be get rtc module for rpi 2021-04-19 17:01:28 tomleb: an RTC for a RPI is only about £5 / $5 2021-04-19 17:01:37 do i understand it right that in "diskless" mode i have some form of remote config store 2021-04-19 17:02:06 Yeah, to be honest, I don't really need the dates to be perfect. swclock working would be good enough. Is there a way to start services AFTER ntp changed the clock? 2021-04-19 17:02:57 need ntp-client? 2021-04-19 17:03:18 tomleb: well the ntp service depends on the networking service so some services (such as networking) need to already running 2021-04-19 17:04:09 ikke: I will try that. Is it possible to add this to a runlevel? Say I wanted a new runlevel `after-ntp`. Sorry, this probably is more an openrc question than an alpine question 2021-04-19 17:06:24 opv: In diskless mode nothing is kept except things that are whitelisted and committed with `lbu`. Whatever you commit with lbu will be written in /media/mmc*/*.apkovl.tar.gz. These apkovl.tar.gz files un unpacked when you reboot. 2021-04-19 17:06:28 tomleb: looking at boot log ntp client start at begining of "default" runlevel right after firewall 2021-04-19 17:06:53 tomleb: yes but more conventional way would be echo rc_need=ntp-client >> /etc/conf.d/whatever 2021-04-19 17:07:02 not sure the exact variable name 2021-04-19 17:07:11 i see. thank you 2021-04-19 17:08:01 Hello71: Right, just seems easier to do it at runlevel level if possible, that way I don't need to specify this in every conf.d I want. I'll play with these things later tonight. Thank everyone! 2021-04-19 17:08:24 note that ntp clients take a while before they set the time. If you can reach an ntp server, something like rdate will set approximate time instanteneously 2021-04-19 17:09:25 tomleb: in "/etc/rc.conf" uncomment line: rc_logger="YES" and you will see what and when something start so maybe you wont need adjust anything 2021-04-19 17:10:06 and that how fast time will update depend on client and configuration, "chronyd" can do it very fast 2021-04-19 17:12:55 Anybody able to help out an Alpine n00b? Is there a way to get Name Service Switch (NSS) installed so I can dynamically generate user passwd data (as from a NIS+ service or something)? It doesn't look like something the Busybox world is intended to do. 2021-04-19 17:13:44 it's not related to busybox 2021-04-19 17:14:01 nss is a glibc feature and is not (directly) supported by musl 2021-04-19 17:14:11 Well, for example, the `getent` command doesn't load NSS modules ever. 2021-04-19 17:14:19 also isn't nis+ super obsolete 2021-04-19 17:14:21 Oh, okay. Yes. the library. 2021-04-19 17:14:39 yes but getent doesn't load nss in the first place, it calls libc which may use nss 2021-04-19 17:14:39 I'm not after NIS+, per se, just something like it -- that works through NSS. 2021-04-19 17:15:14 you can install musl-nscd which is mostly compatible with glibc style nss modules 2021-04-19 17:16:23 nsss is a more fully featured package but requires recompiling all your packages that you want to use nss with 2021-04-19 17:20:13 I see an nss-pam-ldapd package and wonder if that may get me close. 2021-04-19 17:22:40 the documentation says some raspi options like 'gpu_mem' can only be set in config.txt, they dont work when set in usercfg.txt 2021-04-19 17:23:03 that is correct - but config.txt says it is being automatically overwritten "on upgrade" 2021-04-19 17:23:18 what conditions constitute an upgrade, and/or how to workaround? 2021-04-19 17:23:23 opv: just don't upgrade :p 2021-04-19 17:23:50 on diskless, only if you extract a new file 2021-04-19 17:24:08 not sure about sys 2021-04-19 17:36:46 :-/ 2021-04-19 18:01:23 opv: usercfg.txt working fine 2021-04-19 18:02:13 amazing news from new version of Firefox: "PDF forms now support JavaScript embedded in PDF files." ;) 2021-04-19 18:02:41 in Alpine "config.txt" you have line "include usercfg.txt" so you dont have to touch config.txt at all 2021-04-19 18:02:47 opv ^ 2021-04-19 18:14:01 MY-R: yes. *but* the gpu_mem param is not processed when in usercfg, only when in config 2021-04-19 18:14:02 opv: but hmm probably you are right and gpu_mem in other file than config.txt is ignored when other settings working... I set gpu_mem=16 but vcgencmd get_mem gpu return default 76 2021-04-19 18:14:15 ye, kind of weird 2021-04-19 18:15:06 i set up a local.d script with a simple if :) 2021-04-19 18:15:22 i cant seem to have it run on shutdown tho can i? 2021-04-19 18:16:26 opv: /etc/local.d/README 2021-04-19 18:44:29 Hi, how can I use alpine to send an email without opening the composer? 2021-04-19 18:45:49 very carefully 2021-04-19 18:46:15 wrong chennel, sorry 2021-04-19 18:46:53 ##alpine plz. 2021-04-19 18:46:55 :) 2021-04-19 18:47:25 thank you :) 2021-04-19 19:28:50 minimal: are you using "initial_turbo" setting in your rpi's usercfg.txt? for me on diskless setup it cut off system boot time by 15 seconds from 55s to 40s 2021-04-19 19:31:58 40s boot time? 10 years old thinkpad can do that for 5 ;) 2021-04-19 19:33:28 MY-R: haven't tried that - I'm using sys mode rather than diskless for my RPIs and they are running (via PoE) all the time so I'm more interested in getting the power usage as low as possible 2021-04-19 19:34:17 helby: he's talking about booting off SDcard, not a HDD or SSD 2021-04-19 19:34:26 minimal: max what can set is 60 seconds so after all it will drop to your current setting or until your cpu gov will change 2021-04-19 19:35:16 minimal: I know, but using something like that is nonsense 2021-04-19 19:35:25 isnt about sd or hd, it is diskless setup where all apks are installed from early boot so decompresing/installing them + apkovl.tar.gz taking time 2021-04-19 19:36:02 what's the point to run diskless? 2021-04-19 19:36:29 MY-R: true but as I'm powering via PoE and some of the RPIs have USB HDDs attached I need to ensure the power draw at power on (including HDD) doesn't exceed what's available via PoE 2021-04-19 19:36:43 u can find 100GB HDD in trash ;) 2021-04-19 19:36:43 minimal: oh ye true 2021-04-19 21:54:38 Hello71: with your suggestion of rc_need=ntp-client, was it literally ntp-client? Or do I need to substitute with something else (eg: chronyd)? 2021-04-19 21:54:45 not sure 2021-04-19 21:54:54 i think it might be "ntp" 2021-04-19 21:58:53 Heh, maybe wasn't a good idea to reboot the pi when I ask someone a question.. 2021-04-19 21:59:19 tomleb: 23:54 Hello71 │ i think it might be "ntp" 2021-04-19 22:01:25 Thanks. I tried rc_need=chronyd and that didn't work, will try other things (or do a research first) 2021-04-19 22:03:19 tomleb: like I said before, chronyd will start basically at the begining of "default", even if you put it in "boot" then will start after network 2021-04-19 22:03:24 i don't think chronyd waits for sync by default 2021-04-19 22:03:35 then it still will adjust clock with some short delay 2021-04-19 22:03:43 and yes, unless you enable rc_parallel then you just need to check service order 2021-04-19 22:03:55 and whether chrony actuallyw aits 2021-04-19 22:05:31 "provide ntp-client ntp-server" in /etc/init.d/chronyd, so yeah I guess I can also use rc_need=ntp-client 2021-04-19 22:08:16 By default FAST_STARTUP=no and so chronyd runs in the foreground .. So I'd think it does block.. hum.. 2021-04-19 22:10:54 iirc depends on if you use initstepslew 2021-04-19 22:12:01 That's the default value: initstepslew 10 pool.ntp.org 2021-04-19 22:12:14 yeah, you need to use initstepslew 2021-04-19 22:17:54 So it should be fine then.. Hum.. At this point it seems like I'd save a ton of time just buying an rtc module 2021-04-19 22:19:19 I'll be testing a few things and therefore won't be able to see any messages, but thanks again for helping 2021-04-19 23:04:40 tomleb: these are the RTCs that I use: https://www.amazon.co.uk/DS3231-Precision-Module-Compatible-Raspberry/dp/B07GZHTKCS/ref=sr_1_2_sspa?dchild=1&keywords=raspberry+pi+rtc&qid=1618873440&sr=8-2-spons&psc=1&spLa=ZW5jcnlwdGVkUXVhbGlmaWVyPUEzVEQxQVBKUEdTMjhMJmVuY3J5cHRlZElkPUEwODQwNDI0TDk4UU41Q0xPQk1BJmVuY3J5cHRlZEFkSWQ9QTAxNTQzMTAzU0ZLSEYxOUI4VE5JJndpZGdldE5hbWU9c3BfYXRmJmFjdGlvbj1jbGlja1JlZGlyZWN0JmRvTm90TG9nQ2xpY2s9dHJ1ZQ== 2021-04-19 23:33:05 minimal: Im using same one :) 2021-04-19 23:35:35 MY-R: you know since I had the RTC driver put into the Alpine RPI kernel that you no longer need to run the hwclock init.d service? you can run osclock init.d (which actually is a stub to do nothing) *IF* you have NTP setup as the kernel, rather that the OS will copy the system clock back to the RTC every 10mins automatically once NTP is synced 2021-04-19 23:35:57 s/that /than / 2021-04-19 23:38:59 minimal: I think we spoke about it time ago and in my case I couldnt disable hwclock service, it always coming back enabled :P 2021-04-19 23:40:43 you should be able to "rc-update del hwclock boot, rc-update add osclock boot" - though for diskless I guess you'd need to use LBU to write the change 2021-04-19 23:44:41 minimal: I need reboot my pi so will check it one more time but in chrony I have "rtcsync" so is like you said, kernel taking care about it :) 2021-04-19 23:45:54 if you do a "dmesg" and look for mention of "RTC" you should see the system setup its management of the RTC<->System clock 2021-04-19 23:47:01 minimal: yep, that working: [ 0.833388] rtc-ds1307 1-0068: setting system clock to 2021-04-19T19:04:28 UTC (1618859068) 2021-04-19 23:47:07 "CONFIG_RTC_SYSTOHC=y" is the setting you should see in /boot/config-lts that the System Clock to RTC auto copy is enabled, it just needs to see a NTP sync to be activated 2021-04-19 23:47:47 the same applies for x86 and x86_64 - you can use init.d/osclock rather init.d/hwclock 2021-04-19 23:48:13 will see after reboot if can disable hwclock at all :) 2021-04-19 23:49:27 hwclock removed from boot, added osclock to boot, lbu ci and brb 2021-04-19 23:54:50 minimal: it is some diskless quirk which bringing back hwclock service, I think somebody pointed it out in some config 2021-04-19 23:55:10 could be - never really used diskless 2021-04-19 23:58:05 but rtc working fine, with chrony as a ntp server for my local network even if internet is down then still can serve time from rtc, I know it isnt safe but still better than nothing or 1970 ;) 2021-04-20 00:00:23 RTC is fine in that scenario, the DS3231 keeps good time 2021-04-20 00:01:34 you could get a RPI GPS board or a radio clock receiver if you really wanted for when your Internet's down 2021-04-20 00:05:42 I was thinking about it but gps need little time to sync and in place where I could put gps receiver visibility on sky is still low, so I would need at least 2 gps in two different place and one with at least some battery backup 2021-04-20 00:06:21 I have a RPI GPS board and a DCF77 (Germany radio signal) receiver 2021-04-20 00:06:54 however you need 3+ sources to determine accurate time so should get a MSF (UK) radio signal receiver too :-) 2021-04-20 00:08:15 dcf I need check first if will get any signal at all, Im living in middle of city with lot of noise and between buildings, even radio and gsm suck here 2021-04-20 00:09:25 I thought about using mobile network as time source but their accuracy in general is crap 2021-04-20 00:09:41 oh yeah it is :) 2021-04-20 00:11:05 it used to be only down to per-minute accuracy, think it might have improved with 4G/5G (haven't worked in mobile since the 3G days) but I've seen some operators elsewhere in the world with very shoddy timekeeping 2021-04-20 00:13:12 I noticed here it is 1 minute +/- 2021-04-20 00:15:12 MY-R: CDF77 range diagram: https://en.wikipedia.org/wiki/File:Dcf_weite.jpg 2021-04-20 00:19:47 minimal: Im from poland, I know Im in range but none of my window point to West + ether here isnt clear at all :P 2021-04-20 00:31:19 it's future proof to look at East so 2021-04-20 08:13:27 Hi there 2021-04-20 08:21:57 general kenobi 2021-04-20 10:03:25 hey i am facing some problems with xorg setup 2021-04-20 10:03:34 is anyone online? 2021-04-20 10:07:30 lmao 2021-04-20 10:07:32 ok 2021-04-20 16:15:04 hi 2021-04-20 16:15:26 any idea how to install gnu icecat on alpine 2021-04-20 16:15:49 Seems like it has not been packaged 2021-04-20 16:16:01 I found that really weird 2021-04-20 16:16:19 Maybe there are challenges 2021-04-20 16:16:31 what do you mean 2021-04-20 16:16:53 compattibility issues or something, making it difficult to build on alpine 2021-04-20 16:17:01 hmm 2021-04-20 16:17:04 Or maybe no one had interest in packaging it yetr 2021-04-20 16:17:13 also 2021-04-20 16:17:26 is it normal that firefox has 801 file descriptors open 2021-04-20 16:17:30 Cannot find any references in our issue tracker 2021-04-20 16:17:44 it looks a bit excessive 2021-04-20 16:17:45 I don't keep count 2021-04-20 16:18:05 I was trying to run something and I saw I had hit the limit 2021-04-20 16:18:19 lsof | grep firefox | wc -l 2021-04-20 16:18:22 is returning 801 2021-04-20 16:18:40 feels like a lot considering I have only like 5 websites open 2021-04-20 16:18:49 and even most of those fds should have been closed after load 2021-04-20 16:19:47 63134 for me (on my non-alpine system) 2021-04-20 16:20:37 holy fuck what 2021-04-20 16:20:43 63k fds open? 2021-04-20 16:20:46 all by firefox? 2021-04-20 16:20:54 seems like it 2021-04-20 16:21:02 If it's a good representation 2021-04-20 16:21:27 then firefox is probably fucked 2021-04-20 16:21:33 what the fuck 2021-04-20 16:23:06 ikke: 63k is a bit execessive. On Alpine I have 1273 currently 2021-04-20 16:23:18 * ikke: 63k is a bit excessive. On Alpine I have 1273 currently 2021-04-20 16:23:22 that's still a whole lot 2021-04-20 16:23:31 Sure but way better than 63k 2021-04-20 16:23:47 But this amount is probably normal 2021-04-20 16:24:05 I really doubt it 2021-04-20 16:24:32 based on your extensive knowledge of linux, using... lsof|grep? 2021-04-20 16:25:58 based on the fact that it doesn't make sense that a browser with most connections closed has 63 thousand file descriptors open 2021-04-20 16:26:38 With "this amount" I meant my 1.2k, not their 63k 2021-04-20 16:26:46 still pretty strange 2021-04-20 16:26:52 why does it need so many fds open 2021-04-20 16:27:11 several assumptions you are making are not correct. including, but not limited to: browsers don't have keepalive, lsof is a good way to measure fds, threads don't exist... 2021-04-20 16:27:48 I have some websocket connections open and yes some keep alive sure 2021-04-20 16:28:02 but I have 3 tabs opened and 1000 file descriptors 2021-04-20 16:28:05 oh, and most importantly, your central assumption that it's better to repeatedly close and reopen files instead of caching open files 2021-04-20 16:28:33 my central assumption is that it's good to close unused file descriptors 2021-04-20 16:28:45 is that wrong 2021-04-20 16:28:45 it's called caching. 2021-04-20 16:28:56 well shit how do I make it cache less bullshit 2021-04-20 16:29:14 you're measuring useless things 2021-04-20 16:29:31 see also: linuxatemyram 2021-04-20 16:29:47 fuck me 2021-04-20 16:30:10 "how do i turn off the page cache" 2021-04-20 16:30:34 you done playing playa 2021-04-20 16:32:13 done wasting my time trying to get information into your thick skull 2021-04-20 16:32:55 so you done 2021-04-20 17:26:30 Hey, I install alpine to a virtualmachine, but it's stuck on 'starting chronyd' 2021-04-20 17:26:51 installed* 2021-04-20 17:27:30 wpeace: i have not seen that issue before, sorry... i don't know if it will time out and move on or not 2021-04-20 17:29:38 It moved on but it took like 15 minutes. Do you have any suspected causes? Could it because the vm does not have enough resources? 2021-04-20 17:30:55 why people use chrony? 2021-04-20 17:30:56 wpeace: could it be (lack of) entropy related perhaps? 2021-04-20 17:31:11 there is busybox for that crap :/ 2021-04-20 17:31:32 I use it because it's default :| 2021-04-20 17:31:43 that's weird, default should be busybox 2021-04-20 17:31:50 helby: busybox's ntpd is only a minimal ntp implementation so, for example, it does not support some of the security features 2021-04-20 17:31:57 it works and does not require any package to be installed 2021-04-20 17:32:12 minimal: security feature to adjust time? :/ 2021-04-20 17:32:45 if somebody has desire to install tons of package, at least one good quality is opentp for it 2021-04-20 17:32:55 helby: security "hack" to tell a machine the wrong time to defeat some time-related security (i.e. allow the use of an expired account by setting back clock) 2021-04-20 17:32:59 chrony is the default option is why 2021-04-20 17:33:22 wpeace: what's this VM running on? KVM/qemu? Virtualbox? something else? 2021-04-20 17:33:28 virtualbox 2021-04-20 17:33:57 when I looked this problem up, I saw something about entropy 2021-04-20 17:34:00 yes default, choosen busybox, then slowly for everything there will be package 2021-04-20 17:34:07 ok, not sure if virtualbox exposes the host machine's RNG to VMs to provide them with "good" level of entropy 2021-04-20 17:35:58 you can check the level of entropy once its booted by: "cat /proc/sys/kernel/random/entropy_avail". What's its value? 2021-04-20 17:36:52 also openntpd is much smaller than chrony and has some control check with a web server 2021-04-20 17:40:15 3118 2021-04-20 17:40:20 sorry I was slow to reply 2021-04-20 17:40:39 wpeace: that seems fine, so I don't think low entropy was a factor 2021-04-20 17:41:05 wpeace: did you see anything in the logfiles to show why the delay during boot? 2021-04-20 17:41:11 I iwll check 2021-04-20 17:41:25 uninstall that nonsense ;) 2021-04-20 17:41:36 openrc doesnt log anything by default 2021-04-20 17:41:39 is that set in alpine? 2021-04-20 17:41:50 in daemon 2021-04-20 17:42:07 sorry, nope 2021-04-20 17:43:45 wpeace: no files in /var/log such as "messages" and "daemon" and "chrony/chrony.log" ? 2021-04-20 17:45:06 minimal: a possibility is that chrony may be configured to wait until it gets sync with NTP server before proceeding (that's configurable) 2021-04-20 17:46:45 wpeace: in /etc/chrony/chrony.conf change "FAST_STARTUP=no" to "FAST_STARTUP=yes"? 2021-04-20 17:51:14 There is a chrony folder in var/log/ but it's empty 2021-04-20 17:53:43 wpeace: try the config change I mentioned ^^^ 2021-04-20 17:54:23 change your nick to 'complicated' 2021-04-20 17:54:36 u don't deserver 'minimal' ;) 2021-04-20 17:55:28 okay, I added 'FAST_STARTUP = yes" to that file 2021-04-20 17:55:39 helby: minimal software that does what I want, not minimal software full stop 2021-04-20 17:56:16 helby: do you run only "hello world" on your computer as its about as minimal as you can get? 2021-04-20 17:56:30 'hello helby' only 2021-04-20 17:57:03 is openntpd not enough? 2021-04-20 17:57:06 oh I am having an issue where 'echo "secret_new_root_password" | chpasswd' returns 'chpasswd:missing new password' 2021-04-20 17:57:15 better to fight with 'entropy, fast_startup and other nonsense? 2021-04-20 17:57:37 spend one hour with some ntp nonsense? 2021-04-20 17:57:48 helby: in my case I don't believe it support NTS yet, also (from memory) it doesn't support working in an offline mode 2021-04-20 17:59:24 helby: the entropy "nonsense" was nothing to do with chrony - if I machine pauses during boot in general then low-entropy is a common reason for that (especially for embedded devices, ARM boards, servers in general, etc) 2021-04-20 18:00:23 okay, but change your nick ;) 2021-04-20 18:00:57 busybox ntpd is also sntp only, so less accurate 2021-04-20 18:03:14 still for most people is enough 2021-04-20 18:04:19 wpeace: did that config change and a reboot make any improvement? 2021-04-20 18:07:32 oh, I didn't re boot to check 2021-04-20 18:08:04 I will tell you next time I reboot 2021-04-20 19:48:08 Hi all. I'm running Alpine linux (arm64) on QEMU running on a RPi. It works very well, except if I setup QEMU to use a VNC display for the console I get strange characters at the prompt. It's fine when using SPICE or SSH. Other distros are fine, which is why I'm asking here. The characters are a left arrow [123;11R but to make things even stranger, the '123' increases by one each time I hit the enter key. Any ideas folks? 2021-04-20 19:57:40 hi 2021-04-20 19:58:56 sorry for the stupid question, i'm running apk add openssh, then "/etc/init.d/sshd start" fails, probably because openrc-run command which is the shebang of the script, doesn't exist 2021-04-20 19:59:15 is that because i missed something, or is it lacking some sort of dependency? 2021-04-20 20:00:13 what is the context? 2021-04-20 20:00:20 where are you running it 2021-04-20 20:03:12 in a docker container 2021-04-20 20:04:48 right, there is no openrc in a docker container 2021-04-20 20:05:15 ok, i shall run sshd "directly" then? 2021-04-20 20:06:01 running sshd in docker does not make a lot of sense 2021-04-20 20:06:11 but, if you want to, you should run it directly, yes 2021-04-20 20:26:08 so sometimes diskless mode just straight up doesnt load pkgs, however loads the config apkvol.. anyone ever seen this 2021-04-20 20:26:53 i can reboot like 10 times and 8 out of 10 itll fail to load all pkgs and 2 times itll work 2021-04-20 20:27:14 mla: never noticed that when I still used diskless mode 2021-04-20 20:27:20 the system still boots, its just pkg contents are missing; and its a barebones system 2021-04-20 20:29:54 strange thing is, its dependent on the machine too 2021-04-20 20:30:19 like on old x60 thinkpad boots properly 100% of time; on newer machine its 20% of time boot works 2021-04-20 20:30:30 ive tried efi/enabling/disabling too.. 2021-04-20 20:33:12 Do you get an error? 2021-04-20 20:36:23 no error its just sometimes the pkgs wont be there; i enabled debug_init and it looks like it "thinks" its ok loading all the pkgs 2021-04-20 20:37:05 about to start hacking this initram script, or give up i dunno 2021-04-20 20:37:19 i've already hacked mine 2021-04-20 20:37:32 mla: and `apk fix`? 2021-04-20 20:37:44 sounds like you have a bum package in your initramfs 2021-04-20 20:38:12 hm lets see 2021-04-20 20:46:35 apk fix runs clean... when it a actually boots properly 2021-04-20 20:46:54 right, would be interesting to see what happens when it does not 2021-04-20 20:46:56 in dmesg i see utf8 is not a recommened IO charset for FAT filesystems 2021-04-20 20:47:01 wonder if thats related? 2021-04-20 20:47:22 should not 2021-04-20 20:47:54 when it doesnt boot properly apk fix complains that /media/usb/apks/../APKINDEX.tar.gz doesnt exist 2021-04-20 20:48:17 and yeah /media/usb is empty on failure scenario 2021-04-20 20:48:40 but i dont get why my configs.. are there when it fails to boot and load pkgs / /media/usb 2021-04-21 03:05:08 Hey 2021-04-21 03:07:13 Hello. 2021-04-21 03:07:44 I am struggling to connect to ethernet upon first boot, I am not sure the command I need to use to possibly solve the issue by following the instructions on here Configure_Networking#Ethernet_Configuration 2021-04-21 03:07:57 when I installed alpine yseterday, I did not have this issue (I had a different issue) 2021-04-21 03:08:19 I did  a test, I could connect before rebooting 2021-04-21 03:11:12 actually, I am logging in without typing the password 2021-04-21 03:11:21 something could be wrong with how I have virtualbox 2021-04-21 03:11:57 Could be. 2021-04-21 03:12:07 If your hypvervisor has issues, it'll def. trickle down. 2021-04-21 03:12:44 fixed it 2021-04-21 03:12:51 I had to remove the iso 2021-04-21 03:13:33 What I want to test is installing the nix package manager on alpine (it is designed to be distro agnostic) 2021-04-21 03:15:00 I did something like that a long time ago with Slackware. 2021-04-21 03:22:52 Oh, what do you use alpine for? 2021-04-21 03:22:55 if you do 2021-04-21 03:39:48 Oh sorry, I was AFK 2021-04-21 03:40:01 I use alpine as my router and docker host with alpine docker containers 2021-04-21 03:40:02 mostly 2021-04-21 03:40:11 Oh, he left. 2021-04-21 03:40:42 I use The Lounge and don't do as well checking the tab as I did irssi in a term. 2021-04-21 03:40:48 :\ 2021-04-21 04:27:45 hello, recently i've been getting errors with my certificates when i use lynx and irc, i couldn't connect to the ssl version of freenode because of a certificate error 2021-04-21 04:28:11 i tried updating it with the update certificate command but it didn't work 2021-04-21 04:29:46 !help 2021-04-21 04:30:21 :! 2021-04-21 04:32:11 kelvinf: make sure system time is correct 2021-04-21 04:33:03 Was lxde removed? 2021-04-21 04:33:15 it is mentioned on the wiki in multiple places, but doesnt show up on the package manager 2021-04-21 04:33:28 Check aports git history 2021-04-21 04:33:52 i'm using weechat, does it display the time anywhere? 2021-04-21 04:34:01 I don't know that. 2021-04-21 04:34:04 on the bottom left 2021-04-21 04:34:06 for me 2021-04-21 04:34:10 I suppose you have cmdline access, so just run date on it 2021-04-21 04:34:12 above the name? 2021-04-21 04:34:15 yes 2021-04-21 04:34:30 it says 13:29, so i'm assuming it's not correct 2021-04-21 04:34:41 At least 5 minutes off. What TZ are you in? 2021-04-21 04:34:50 CEST 2021-04-21 04:34:53 A couple of hours shouldn't matter though. 2021-04-21 04:35:17 kelvinf: what is the error you are getting? 2021-04-21 04:35:26 weechat shows my system time 2021-04-21 04:35:41 i'm gonna leave the channel and come back, i really should've installed tmux on this thing 2021-04-21 04:35:41 Well, then your system time also isn't correct 2021-04-21 04:45:25 alright the time was incredibly screwed up 2021-04-21 04:45:32 it said that it was january 5th or smth 2021-04-21 04:45:38 so i reset the date and time 2021-04-21 04:45:43 and the error is fixed 2021-04-21 04:46:14 thank you 2021-04-21 04:46:49 wait is the 6697 room different from the 6667 room? 2021-04-21 04:46:52 i'm an idiot 2021-04-21 04:46:56 No 2021-04-21 04:46:59 Same room 2021-04-21 04:47:08 oh ok 2021-04-21 04:50:18 kelvinf: consider enabling an ntp client 2021-04-21 04:50:42 kelvinf: are you running this on an rpi or something? 2021-04-21 04:50:44 i do have one more question: what are some ways that my clock could have ended up three months behind 2021-04-21 04:50:51 yes i am running it on an rpi 2021-04-21 04:51:03 an rpi does not have a real-time clock 2021-04-21 04:51:09 i see 2021-04-21 04:51:21 so when you boot, it starts at january 1st 1970 2021-04-21 04:51:31 So I suspect it was off for more then just a few months 2021-04-21 04:51:35 nah it was jan 2021 2021-04-21 04:51:39 hmm, strange 2021-04-21 04:51:40 Where is the xorg log file located? 2021-04-21 04:51:46 i guess that might be the default date 2021-04-21 04:51:52 i don't think i have xorg 2021-04-21 04:51:55 it's just tty 2021-04-21 04:52:02 ah okay 2021-04-21 04:52:13 which is why i mentioned tmux earlier 2021-04-21 04:52:14 I am texting out alpine in a vm because I am thinking about using it over Debian 2021-04-21 04:52:17 on my laptop 2021-04-21 04:52:26 oh yeah i started with debian 2021-04-21 04:52:28 wpeace: ~/.local/share/xorg/Xorg.log 2021-04-21 04:52:32 thx 2021-04-21 04:52:55 wpeace sorry i didn't know you were asking for help, i thought my xorg file had something to do with the time lol 2021-04-21 04:53:03 ol 2021-04-21 04:53:21 I've been contemplating switching my workstation to Alpine. 2021-04-21 04:53:34 I'm getting annoyed with Debian too. 2021-04-21 04:53:41 oh I don't even have a .local 2021-04-21 04:53:42 i used devuan for a bit 2021-04-21 04:53:47 Well 2021-04-21 04:53:49 When I say Debian 2021-04-21 04:53:50 it's nice 2021-04-21 04:53:52 I do mean Devuan. 2021-04-21 04:53:54 :P 2021-04-21 04:53:55 ah 2021-04-21 04:54:11 The only box I have w/ systemd is my Plex server. 2021-04-21 04:54:27 And at that point it's a single purpose application black hole. 2021-04-21 04:54:34 It can do whatever it wants. 2021-04-21 04:54:51 I use arch on my desktop because it is just easier to like get packages for anything, but for my laptop, I wanted something stable (doesnt update often), so I went with debian testing because buster is really old atm 2021-04-21 04:55:12 i use devuan ceres on desktop 2021-04-21 04:55:12 I use devuan unstable atm 2021-04-21 04:55:13 I have been having some issues related to things in the normal repository not being in the main repository 2021-04-21 04:55:24 I mean testing* 2021-04-21 04:55:49 I have been finding things in the normal repo are not accessible in testing, I am not sure if that is intentional or not 2021-04-21 04:55:51 testing is only built for edge 2021-04-21 04:56:07 Well I will use Bullseye when it comes out 2021-04-21 04:56:08 Not sure what you mean with accassible? 2021-04-21 04:56:16 is testing more up-to-date than edge? 2021-04-21 04:56:28 oh I am talking about Debian btw 2021-04-21 04:56:35 oh 2021-04-21 04:56:35 kelvinf: testing is a repo on the edge branch 2021-04-21 04:56:39 wpeace: aha 2021-04-21 04:56:45 My router has as a much too big 250GB mSATA SSD that I keep a local backup of x86_64 and aarch64 alpine repos for edge] 2021-04-21 04:56:58 @ikke i figured, i wasn't sure though 2021-04-21 04:57:44 atom /var/www/lighttpd/html/dist # du -sh * 2021-04-21 04:57:44 52G alpine 2021-04-21 04:57:49 :D 2021-04-21 04:58:53 I just wrote a little one liner so that when i want to update my router, it pulls a fresh rsync of the repository to keep it 1:1 then updates from the local backup 2021-04-21 04:59:01 then all packages are immediately available to my network 2021-04-21 04:59:26 what os does your router use? 2021-04-21 04:59:29 Alpine 2021-04-21 04:59:31 oh 2021-04-21 04:59:33 lmao 2021-04-21 04:59:33 :P 2021-04-21 04:59:43 It's a PCEngines APU2 2021-04-21 04:59:56 https://www.pcengines.ch/apu2.htm 2021-04-21 04:59:57 idk much about routers lmao 2021-04-21 05:00:19 seems interesting 2021-04-21 05:00:30 i guess i could look into setting one up sometime 2021-04-21 05:00:41 It's pretty straight for-ward. 2021-04-21 05:00:43 foreward 2021-04-21 05:01:18 not to be pedantic but it's spelled "straightforward" usually 2021-04-21 05:01:28 Is it? 2021-04-21 05:01:30 I didn't know that. 2021-04-21 05:01:31 lol 2021-04-21 05:01:36 usually when i've read it yeah 2021-04-21 05:01:47 Google says you're right 2021-04-21 05:01:51 maybe that's a usa thing idk 2021-04-21 05:01:51 So I'll assume you're wrong 2021-04-21 05:01:53 ;P 2021-04-21 05:01:53 oh good 2021-04-21 05:01:53 jk 2021-04-21 05:02:01 :) 2021-04-21 05:02:21 I got really sick of my consumer router, then I got really sick of how bloated pfsense was. 2021-04-21 05:02:53 After configuring the same services and more on the same machine, I went from a consistent 1GB of memory used to about 200mb. 2021-04-21 05:03:16 So using alpine will make the router viable, longer. 2021-04-21 05:03:42 Oh apparently it's only using 145mb atm 2021-04-21 05:03:46 and that's during an update 2021-04-21 05:04:34 neat 2021-04-21 05:05:04 It also doubles as Microsoft KMS server thanks to vlmcsd. 2021-04-21 05:05:05 alright i should go 2021-04-21 05:05:09 later! 2021-04-21 05:05:13 later! 2021-04-21 05:05:26 Nice guy. 2021-04-21 05:09:33 I am trying to debug Xorg rn, the error message is " 2021-04-21 05:09:33 Error loading shared library libEGL.so.1" 2021-04-21 05:09:34 I tried one suggestion I found on the arch forums to create this symlink to no effect, another one said to install libGLU, but that doesnt exist in the repo. I think most of the errors were related to a driver issue with arch and something with fedora, this leads me to believe I might not have the right drivers, however I have the ati and radeon 2021-04-21 05:09:34 driver currnetly, 2021-04-21 05:11:45 mesa-egl provides that library, is it installed? 2021-04-21 05:12:16 I believe it was, but I will doublecheck 2021-04-21 05:21:50 That didn't seem to have any effect 2021-04-21 05:24:48 I am running it in a vm 2021-04-21 05:24:55 maybe I need some sort of package related to that 2021-04-21 07:03:31 Hello, I am having an issue with virt manager on alpine, for some reason I can't enable sound in one of my vms and this is bugging me. 2021-04-21 07:04:11 It works out of the box when I had my vms on Debian but now that I am running vms off of Alpine none of the sound cards work. 2021-04-21 07:05:12 These aren't windows vms, they are regular linux ones. 2021-04-21 08:18:18 hi 2021-04-21 08:19:20 I noticed that rebooting to my @EDGE snapshot from my current system, docker stops working :\ 2021-04-21 08:19:25 some btrfs related problem 2021-04-21 08:22:50 (currently @ROOT and @EDGE are identically :D) 2021-04-21 09:06:14 hey there! ubuntu scared me of for good now. so i am switching to alpine for my personal server. i am a complete novice to alpine. usually using voidlinux. 2021-04-21 09:06:40 Welcome! 2021-04-21 09:07:05 My issue at the moment is, that i want to produce an sdcard to boot a rasPi 4 with running sshd, since its buildin without display output 2021-04-21 09:07:52 s/buildin/built in 2021-04-21 09:12:25 you can setup an overlay with alpine lbu which ssd configured and autostarted 2021-04-21 10:08:13 s/ssd/ssh 2021-04-21 10:17:23 @freenode_donoban:matrix.org: i tried this guide https://wiki.alpinelinux.org/wiki/Raspberry_Pi_-_Headless_Installation 2021-04-21 10:17:40 no network device showing up so far 2021-04-21 11:32:58 uhM, maybe you need some kernel module or firmware, look in dmesg 2021-04-21 12:42:07 no output at all... 2021-04-21 12:42:19 still trying to get something atm 2021-04-21 13:13:35 finally got HDMI out. eth0 was DOWN. 2021-04-21 13:28:15 does firefox run on alpineV 2021-04-21 13:28:19 s/V/?/ 2021-04-21 13:28:57 Yes, https://pkgs.alpinelinux.org/packages?name=firefox 2021-04-21 13:29:02 thanks. 2021-04-21 13:31:29 Well, I finally jumped to Alpine on my workstation this evening. Got X and the usual utilities going.. I'll finish tomorrow. :D 2021-04-21 13:45:48 hey guys, can anyone tell me if there is any info on the role the file ../cache/installed plays in cache/package management? I haven't been able to find any yet 2021-04-21 14:04:01 okay. i am still struggling to get networking up and i did setup for the third time. after reboot the settings are gone. 2021-04-21 14:04:10 hints? 2021-04-21 14:12:21 hey im trying to figure out why libraries aren't getting installed into /opt/musl/x86_64-linux-musl/lib, i'm trying to static compile one tool and it says libmagic is missing, i've installed that lib, it went to /usr/lib/libmagic.so*, ./configure still says it's missing unless i manually copy it to /opt/musl/x86_64-linux-musl/lib, what i'm doing wrong? why these libs wouldn't be installed to the musl 2021-04-21 14:12:27 lib dir ? 2021-04-21 14:19:03 uhM, most libs are in /usr/lib I don't know exactly the purpose of /opt/musl/x86_64-linux-musl/lib 2021-04-21 14:19:40 so why would configure script look in /opt/musl/x86_64-linux-musl/lib and ignore /usr/lib ? 2021-04-21 14:19:41 how did you install libmagic? 2021-04-21 14:19:51 apk add libmagic 2021-04-21 14:20:00 and apk file-dev 2021-04-21 14:20:06 add* 2021-04-21 14:20:19 ah sorry, was searching on pkgs.alpinelinux.org but had a wrong field 2021-04-21 14:20:37 /usr/lib/libmagic.so.1 it seems std 2021-04-21 14:20:51 what script is ignoring /usr/lib ? 2021-04-21 14:21:15 the rpm's one /usr/src/rpm-4.16.1.3/configure 2021-04-21 14:21:32 but it's like that with other soft i try to compile 2021-04-21 14:21:49 for some reason it looks in /opt/musl/x86_64-linux-musl/lib and not in /usr/lib 2021-04-21 14:22:02 there is something inside /opt/musl/x86_64-linux-musl/lib ? 2021-04-21 14:22:09 In my current install it doesnt' exist 2021-04-21 14:22:16 but I use docker images for building packages 2021-04-21 14:22:42 maybe you can try with a symbolic link 2021-04-21 14:23:43 where did you get the rpm source? 2021-04-21 14:24:18 https://rpm.org/wiki/Releases/4.16.1.3.html 2021-04-21 14:24:37 im thinking about reinstalling alpine ;P 2021-04-21 14:25:06 I feel that they assume that musl is something "alternative" and base in /opt/ 2021-04-21 14:25:24 but in alpine I doubt that that path is used 2021-04-21 14:26:55 do you want RPM support in alpine? 2021-04-21 14:27:24 I don't expect too much rpm packages work directly on alpine 2021-04-21 14:28:32 Note that there already is a rpm package for Alpine 2021-04-21 14:28:36 It could use a maintainer though 2021-04-21 14:30:30 And an update 2021-04-21 14:31:03 and there are some rpm base distro which packages work on alpine? 2021-04-21 14:35:32 I don't think so. The package is more about inspecting rpms and stuff I think 2021-04-21 14:43:52 i have alpine running on a raspi4 and would like to control the fan 2021-04-21 14:43:57 i found this: https://www.raspberrypi.org/forums/viewtopic.php?t=256848 2021-04-21 14:44:44 but the mentioned README doesnt exist, and i'm not quite sure how to set the mentioned params in alpine 2021-04-21 14:45:12 certainly cant open the dtbo files 2021-04-21 15:28:27 https://raspmer.blogspot.com/2019/02/control-cooling-fan-with-gpio-fan.html opv, this link seems more useful 2021-04-21 15:30:57 ty fys 2021-04-21 15:31:12 np 2021-04-21 15:35:24 I am on 3.11 release. Is that the latest main release http://ix.io/2WRZ 2021-04-21 15:39:41 2021-04-14 Alpine 3.10.9, 3.11.11 and 3.12.7 released 2021-04-21 15:40:23 I saw that, but, was not sure if 3.11 is just a branch name or such. 2021-04-21 15:41:46 i really need your help... i got alpine running okayish i got tinysshd to start but files keep on disappearing. 2021-04-21 15:42:55 talking about authorized_keys for example 2021-04-21 15:44:47 also some other configs. and i cant figure what i need to do, in order to make files persistent. reading up on editing overlay files manually 2021-04-21 15:45:15 that article copies the article linked in the forum which was already criticized for having a terrible circuit 2021-04-21 15:45:59 joe9: https://alpinelinux.org/releases/ 2021-04-21 15:51:11 To be fair I just googled the config file, I didn't look at the circuit lol 2021-04-21 15:51:18 :D 2021-04-21 15:57:26 using alpine:3 - RUN echo 'http://dl-cdn.alpinelinux.org/alpine/edge/community' >> /etc/apk/repositories - RUN apk add -U py3-pip - I end up with an error: importlib.metadata.PackageNotFoundError: pip 2021-04-21 15:57:31 Any thoughts on what the heck I'm goofing up? 2021-04-21 15:58:08 Yes, you care combining stable branches and edge 2021-04-21 15:58:36 edge has python 3.9, the latest stable version has 3.8 2021-04-21 15:58:40 so you cannot combine them 2021-04-21 15:59:31 (You could use pinned repositories to get only certain packages from edge, but that can steal break) 2021-04-21 15:59:53 ikke: thanks for fixing the pelican package, it's now working again ♥ 2021-04-21 16:00:14 ikke thanks, will dig. 2021-04-21 16:00:19 Can't remember anymore what I did, but you're welcome :) 2021-04-21 16:00:42 with the latest-stable, I cannot find sudo anymore. Has it been replaced? 2021-04-21 16:00:57 I used 3.11 earlier and sudo worked fine in it. 2021-04-21 16:00:59 no 2021-04-21 16:01:12 I can do apk add sudo if that is the solution. 2021-04-21 16:01:13 https://pkgs.alpinelinux.org/packages?name=sudo&branch=v3.13 2021-04-21 16:01:24 ok, thanks. 2021-04-21 16:01:30 You need to install it, yes 2021-04-21 16:02:25 thanks, that helped 2021-04-21 16:05:54 ikke interestingly, I get the same error in 3.12, and 3.13, so I guess I don't understand what I'm mixing up. It works in 3.11. 2021-04-21 16:06:32 Joel: Might be that some package in 3.11 pulled sudo in 2021-04-21 16:06:36 which it no longer does 2021-04-21 16:06:42 (abuild for example no longer depends on sudo) 2021-04-21 16:06:59 and interestingly, I can't find py3-pip on https://pkgs.alpinelinux.org/packages for v3.11 2021-04-21 16:07:21 ah, sorry, mixing up converstations 2021-04-21 16:07:29 joe9 and joel look similar :P 2021-04-21 16:07:37 oh shit, they sure do. haha. 2021-04-21 16:08:26 hi, I have an odd issue, that I get Access Denied when I try to connect to mysql on RDS from Alpine based php image (I tested PHP and mysql-client) 2021-04-21 16:10:27 do we actually have alpine:3 somewhere? 2021-04-21 16:10:42 huh, we do 2021-04-21 16:10:54 yes 2021-04-21 16:11:04 was there alpine:2? 2021-04-21 16:11:16 Joel: I don't believe so 2021-04-21 16:11:25 that was before docker :P 2021-04-21 16:11:31 wat 2021-04-21 16:11:39 Hmm, maybe not 2021-04-21 16:12:26 well, dockers first release was one year before alpine 3.0 2021-04-21 16:16:18 Joel: since alpine 3.12, python3 does not come with the pip module anymore 2021-04-21 16:16:52 ikke even if I install python3, I get the same error 2021-04-21 16:17:07 the issue is that you pull in py3-pip from edge 2021-04-21 16:17:22 ohh the repo line, instead of 3.13 2021-04-21 16:17:30 Yes 2021-04-21 16:17:51 https://pkgs.alpinelinux.org/contents?file=&path=&name=py3-pip&branch=edge&repo=community&arch=x86_64 2021-04-21 16:18:01 if you look there, py3-pip from edge is built for python3.9 2021-04-21 16:18:04 Yeah, sorry, I missread your first reply to me about that 2021-04-21 16:18:09 so if you install that, python3.8 is not going to find it 2021-04-21 16:19:18 ikke yeah that was it *facepalm*, I should never be putting edge when adding a repo, double idiot move. 2021-04-21 16:19:39 You can also do: apk add -X 2021-04-21 16:19:54 oh interesting, that seems way cleaner 2021-04-21 16:20:35 ikke thanks times a million again 2021-04-21 16:21:23 no problem 2021-04-21 16:22:59 but it still won't work if you use edge 2021-04-21 17:54:22 is there any practical reason to use some kind of 'seat management' like seatd with sway? 2021-04-21 17:54:35 it makes not much sense 2021-04-21 17:55:03 like now it runs without root, but before it ran with root, so 2021-04-21 17:55:05 if you have several users locally at the same time, 2021-04-21 17:55:35 that's probably 'session issue' I assume 2021-04-21 17:55:49 not connected by ssh or so, but really several mice, several keyboards, several monitors 2021-04-21 17:56:12 why? 2021-04-21 17:56:20 without it it will be not handled? 2021-04-21 17:56:44 in inferior linux distros, systemd/logind takes care of it 2021-04-21 17:57:02 don't even mention that crap please :/ 2021-04-21 17:57:36 the system needs to recognize which user is associated with which device in such a multiseated situation. 2021-04-21 17:57:38 as systemd takes care of lot of things so 2021-04-21 17:57:52 so for one user it makes no sense 2021-04-21 17:57:53 ? 2021-04-21 17:58:11 It makes sense to have as little code as possible running as root 2021-04-21 17:58:19 yes, but 2021-04-21 17:58:28 sway was able to get root without it 2021-04-21 17:58:33 why it would be not able to get it with it 2021-04-21 17:58:34 Running your UI and everything that it spawns as root seems rather horrific 2021-04-21 17:58:42 no like that 2021-04-21 17:58:53 sway ran normally as user 2021-04-21 17:59:07 just one process ran as root somewhere ... if not using seatd 2021-04-21 17:59:53 Seatd is smaller than elogind and only does things needed for compositors i believe 2021-04-21 17:59:58 on seatd desc. it tells me, that it take ccare of like gpu, input and shit, that app don't need root access to use them 2021-04-21 18:00:19 but anyway, that shit without it works, so it means, ,they can get access to these things without seatd 2021-04-21 18:00:38 exception: yes much smaller ,I m trying it, but don't see a reason :/ 2021-04-21 18:00:55 had to build wlroots, bc. that in repo is outdated like rest of packages 2021-04-21 18:01:32 is sway the only wayland engine table to use seatd? 2021-04-21 18:01:56 well there are not many in alpine repo anyway 2021-04-21 18:12:01 I think all wlroots 0.13 based compositors can 2021-04-21 18:12:09 yes 2021-04-21 19:38:37 let's try again to boot "edge" :P 2021-04-21 20:06:22 arg same problems 2021-04-21 20:06:56 I guess that docker uses his owns btrfs subvolumes for images and they are not cloned when I create my @EDGE snapshot 2021-04-21 20:52:12 beh all docker stuff is nested with my @ROOT subvolume 2021-04-21 21:12:17 what would people do without all that dockers, kubernetes and other nonsense? ;) 2021-04-21 21:15:40 helby: they probably would find something else 2021-04-21 21:17:14 koollman: depends on what some corps would offer them 2021-04-21 21:18:15 I do like some of the ideas. But it's hard to separate the 'neat/smart' parts from the 'overly complex' ones 2021-04-22 07:28:51 ACTION upgrading to edge :D 2021-04-22 07:43:08 And living on it :-) 2021-04-22 07:49:03 hehe, let's reboot before assuming I'm alive :D 2021-04-22 07:52:29 speaking of.... i am constantly locking up ssh access after rebooting. this time i just changen resolv.conf und did lbu commit -d 2021-04-22 07:52:57 i am rewriting the overlay file after every failed try... is that a known iddue? 2021-04-22 07:53:18 couldnt find anything about that. 2021-04-22 07:53:50 theoretically resolv.conf should not interfere with ssh on a local network imo 2021-04-22 07:54:54 maybe it's stopping some service to properly start and sshd service is not started too 2021-04-22 07:55:12 do you can reproduce it? old resolv.conf works always, new resolv.conf fails? 2021-04-22 08:00:46 just found the problem 2021-04-22 08:01:07 i got a configuration user. i added the authorized _keys manually to he overlayfile. 2021-04-22 08:01:20 it is discarded with every commit. 2021-04-22 08:01:57 there isnt even a homedierectory for the user any more after commit > reboot 2021-04-22 08:02:12 you can add new directories to lbu 2021-04-22 08:02:56 https://wiki.alpinelinux.org/wiki/Alpine_local_backup#Include_special_files.2Ffolders_to_the_apkovl 2021-04-22 08:03:13 the e.g. is an authorized_keys for root 2021-04-22 08:19:35 jjuran: rebooted and ended on a system without network (even without ext4/vfat modules) until I realized it was booted on wrong btrfs subvol and there is a missmatch between kernel and modules version :D 2021-04-22 08:19:51 all fine now 2021-04-22 08:20:25 All's well that ends btr :-) 2021-04-22 08:20:56 I felt a small moment of panic hehe 2021-04-22 08:25:32 uhM, utmp and utmps are compatible? 2021-04-22 08:27:33 I get "Warning: cannot find /var/run/utmp" but there is a 0size file o /var/run/utmp 2021-04-22 08:50:10 Migration from Devuan to Alpine is almost complete. Relatively straight-forward, however I've noticed Chromium a bit finnicky compared to Chrome. 2021-04-22 08:50:26 Not sure if it's me or what. 2021-04-22 08:51:00 (Haven't looked hard enough yet, but it's still minor.) 2021-04-22 08:51:34 don't you like firefox fys ? 2021-04-22 08:51:38 Nope. 2021-04-22 08:51:53 thecnicall or emotional? 2021-04-22 08:52:28 Based on the phrasing of the question I'm going to decline to comment further. 2021-04-22 09:00:51 good morning all. I was wondering if kubernetes works fine on alpine-linux aarch64 2021-04-22 09:01:17 I am going to deploy a 5 nodes kubernetes cluster, and ideally I'd like to use alpinelinux 2021-04-22 09:02:01 but I have never heard anyone running a kubernetes cluster on alpine aarch64, hence the question 2021-04-22 09:05:12 sorry fys :\ I was just curiously about why do you prefer one over the another but I feel that there are many decisions that are just some kind of preference without a deep rationale behin (starting with myself) 2021-04-22 09:05:52 I personally use just firefox because I like mozilla a few more than google, I don't have any idea about what is better :( 2021-04-22 09:06:44 Simple answer: just use the browser that does what you want it to in the best way 2021-04-22 09:09:02 not so simple to measure "in the best way" 2021-04-22 09:10:13 Latrina: it seems only supported on edge so maybe is not very stable yet 2021-04-22 09:10:31 donoban, I see. 2021-04-22 09:10:39 I will use debian then 2021-04-22 09:11:01 I need to learn the product before messing around with what works and what doesn't 2021-04-22 09:22:28 anyone here using zfs on rasPi? 2021-04-22 09:41:59 hi all ! 2021-04-22 09:42:21 Is there a way to force the architecture with apk ? 2021-04-22 09:42:32 I didn't find the option in the documentation 2021-04-22 09:43:58 why would you want to farce architecture? 2021-04-22 09:44:49 Forcing it would be enough, but farcing it might be fun ! 2021-04-22 09:45:01 short answer : Mac m1 2021-04-22 09:45:42 long answer, I have a package missing in the aarch64 repository of the AL I'm using 2021-04-22 09:46:25 NORMALLY, rosetta should be able to handle a x86 package, but I got to tell apk to download the right architecture 2021-04-22 09:53:30 Tumulte i can tell you that /etc/apk/arch contains the architecture as a single line string. so you can change it globally. for a specific package i am really not sure. 2021-04-22 09:54:29 mh, that's a nice starting point, thanks 2021-04-22 09:54:35 (it's docker BTW# 2021-04-22 09:54:39 maybe append it in the next line and see if you can find the different arch 2021-04-22 09:54:46 yeah i figured that 2021-04-22 09:54:50 I should have made it clear 2021-04-22 09:54:56 ah nice 2021-04-22 09:56:20 but i am also an alpine noob so maybe someone here has the option you are searching for 2021-04-22 10:01:00 Hi everyone! 2021-04-22 10:01:30 I've recently had an issue moving from 3.12 to 3.13 - the repositories file now appears to be https:// instead of http:// - this is fine for most people I appreciate. I'm behind an evil corporate web proxy. 2021-04-22 10:02:01 I can sed the file and change https to http and it still works, it feels like an inferior workaround. I can do an http update just to get `update-ca-certificates` and then i can update my certs file and use https like normal... still a bit bodgy 2021-04-22 10:02:32 Has anyone worked around it in an easy yet less bodgy way? Or had this issue? 2021-04-22 10:03:35 It's a bit chicken and egg.. I need to use APK to install update-ca-certificates so that I can install things with APK :-D 2021-04-22 10:04:18 can anyone take a look at this log: termbin.com/5d9u 2021-04-22 10:07:12 i am trying to install zfs-rpi and i get a lot of writing errors. 2021-04-22 10:56:05 ph8: did u try go download the apk-tools-static with wget? 2021-04-22 10:56:07 Might work 2021-04-22 10:59:40 Or download the apk file for thst package with wget and extract or something like in the wikipage for making a chroot 2021-04-22 11:06:11 I did read a bit about downloading the package direct, I saw some posts saying that between alpine versions the dependencies it also requires can change a lot so it can be a bit of a nightmare to maintain 2021-04-22 11:06:30 But I guess that's the nature of our restrictive corporate environment! Thanks for thinking around the problem 2021-04-22 11:07:33 Np 2021-04-22 11:58:20 Latrina: you may also consider k3s which we have had for some time 2021-04-22 12:55:01 is there a policy for packagin in alpine linux? 2021-04-22 12:55:43 i must ask cos in gambas we are updating the guidelines for packagin.. i try to fit the most need in all used distros, i used alpine for some task 2021-04-22 12:59:45 ncopa, ikke, clandmeter is there a policy for packagin in alpine linux? naming scheme guidelines? 2021-04-22 14:36:22 https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/COMMITSTYLE.md idk if this is what u need 2021-04-22 14:42:32 Hello71, thanks, however I need to learn the product as is for the time being 2021-04-22 14:42:49 than I can certainly go for slimmed down versions 2021-04-22 14:43:33 a fine tuned and slim down product is always my preference. But I need to learn the whole product first. That's the way I work at least. 2021-04-22 15:07:26 hi all, i'm having trouble PXE booting a raspi4. i am setting cgroups options in the cmdline.txt, but they seem to get ignored 2021-04-22 15:07:32 this doesn't happen on persistent install, where everything is fine and dandy 2021-04-22 15:07:39 what am i doing wrong? 2021-04-22 18:32:18 morning some fuckering with router, then whole day without internet :/ 2021-04-22 18:42:16 holy this channel is as dead as possible :/ 2021-04-22 19:05:10 yeah, a whole ten minutes without someone responding 2021-04-22 19:05:25 opv: i don't think cmdline.txt is read for pxe 2021-04-22 19:05:38 i think the point of pxe is that you can manage it remotely 2021-04-22 19:06:21 Hello71: yes, PXE provides a cmdline.txt to the raspi 2021-04-22 19:06:54 i did this: https://gitlab.alpinelinux.org/alpine/aports/-/issues/12628 2021-04-22 19:09:06 after long time I bought usb to save at least my config files 2021-04-22 19:09:35 with the cheapest router as bonus 2021-04-22 19:10:37 now I can probably get second old thinkpad and use it as kind of server :/ 2021-04-22 19:11:12 this one is okay, but broke so many keys last time I cleaned it 2021-04-22 19:12:55 probably better choice as fuckery with arm? 2021-04-22 19:13:04 even bigger and noisy 2021-04-22 19:13:17 but x86_64 2021-04-22 19:22:26 then with irc bouncer I will maybe see more than 5 lines in this channel during 24 hours :/ 2021-04-22 22:18:51 It looks like there was a libcgroup pkg in testing some years ago, but no longer. I couldn't find any info in the wiki or mailing list archives, and the forums seem down. Anyone know what the story is? 2021-04-22 22:20:22 no one was using it or maintaining it i imagine 2021-04-22 22:21:54 👍 thanks c705 2021-04-22 22:55:43 ncopa: As long as it's not just me, that's enough for my sanity for now. LOL 2021-04-23 08:21:50 hello! can anybody tell me why 'apk add pdns' on alpine:edge (docker hub) reports 'ERROR: unable to select packages: pdns (no such package): required by: world[pdns]' ? 2021-04-23 08:21:56 https://pkgs.alpinelinux.org/packages?name=pdns&branch=edge looks fine to me 2021-04-23 08:22:52 apk update 2021-04-23 08:23:00 i did :) 2021-04-23 08:23:07 apk search pdns 2021-04-23 08:23:18 pdns-recursor-4.4.3-r2 pdns-recursor-doc-4.4.3-r2 pdnsd-doc-1.2.9a-r5 pdnsd-1.2.9a-r5 pdns-recursor-openrc-4.4.3-r2 2021-04-23 08:23:19 (plus newlines) 2021-04-23 08:23:29 AS you can see, there's no pdns package. 2021-04-23 08:24:01 x86_64? 2021-04-23 08:24:03 yes 2021-04-23 08:24:07 oh 2021-04-23 08:24:15 the url i posted does not have x86_64 2021-04-23 08:24:16 No x86_64 package yet. 2021-04-23 08:24:25 https://pkgs.alpinelinux.org/packages?name=pdns*&branch=edge&arch=x86_64# 2021-04-23 08:24:27 ) 2021-04-23 08:24:28 :) 2021-04-23 08:24:30 thanks 2021-04-23 08:24:41 i'd expect to get the previous package then, is that weird? 2021-04-23 08:24:52 Depends on if it exists. 2021-04-23 08:25:28 > arch="all !aarch64 !x86_64" # Failing tests on aarch64 2021-04-23 08:25:36 In pdns APKBUILD 2021-04-23 08:25:57 ah, where can i find that? 2021-04-23 08:26:15 Source for package pdns of any architecture is the same. 2021-04-23 08:26:21 Click the package name, then "git repository" 2021-04-23 08:26:24 Then APKBUILD. 2021-04-23 08:27:25 ah there, yes 2021-04-23 08:27:58 do you know where i can find build logs? 2021-04-23 08:28:32 You can probably find them when CI ran for the commit 2021-04-23 08:28:44 Check commit 694e9951a6713d2e297bb564ec154741524d8d5b 2021-04-23 08:29:13 thanks 2021-04-23 08:29:16 yw 2021-04-23 08:38:38 ah options="!check" # Failing on new boost 2021-04-23 09:01:34 test failure reproduced 2021-04-23 09:03:08 How can I set environment variables for an openrc init script that doesn't source a file? I'm looking in particular at /etc/init.d/watchdog 2021-04-23 10:14:28 ryan-c: you can export them in /etc/conf.d/watchdog 2021-04-23 10:18:42 Hello, what will be the nginx version for alpine 3.14 ?1.18 or 1.19 ? 2021-04-23 10:18:55 https://pkgs.alpinelinux.org/package/v3.13/main/x86_64/nginx 2021-04-23 10:20:17 1.20 is in edge 2021-04-23 10:21:43 ikke hello :) yes it is, What will be the next version of nginx for alpine 3.14 logically? 2021-04-23 10:22:43 1.20 2021-04-23 10:23:25 MY-R every package in edge will be in 3.14 ? 2021-04-23 10:24:03 GreyXor: yeah 2021-04-23 10:25:24 when will edge be 'frozen' for 3.14? 2021-04-23 10:25:25 GreyXor: https://alpinelinux.org/releases/ "There are several release branches for Alpine Linux available at the same time. Each May and November we make a release branch from edge. The main repository is typically supported for 2 years and the community repository is supported until next stable release." 2021-04-23 10:25:36 ah 2021-04-23 10:26:07 👌 2021-04-23 10:26:11 Habbie: already is almost in that state so wont be any huge changes 2021-04-23 10:26:25 ah 2021-04-23 10:27:05 I need an administrator of the Gitlab infrastructure. I have a technical problem with 2021-04-23 10:27:46 GreyXor: you can join #alpine-infra and leave there message 2021-04-23 10:37:28 ikke: thank you! 2021-04-23 11:17:51 is there a way to have an openrc extra command mark a service as stopped?P 2021-04-23 11:18:28 zap 2021-04-23 11:19:47 how do i call that from the /etc/init.d/... script? 2021-04-23 11:20:32 oh, within an openrc init script 2021-04-23 11:20:51 yeah, i'm writing a script for the full version of watchdog, as the busybox version didn't have some features i needed 2021-04-23 11:21:21 so i'm adding a `keepalive` function that stops the watchdog daemon and replaces it with a wd_keepalive daemon so that the system doesn't reboot 2021-04-23 11:21:40 which works fine, except when i try to start the service again, openrc says it's already started 2021-04-23 11:21:56 i do call the stop function from my init script before starting wd_keepalive 2021-04-23 11:22:35 ryan-c: mark_service_stopped (https://www.mankier.com/8/openrc-run) 2021-04-23 11:22:40 https://up1.rya.nc/#6YIt4Jc-Sg_JtGCO is what I have currently 2021-04-23 11:22:43 okay, lemme try that 2021-04-23 11:23:15 Ah, yes! That did it! 2021-04-23 11:23:17 thanks again 2021-04-23 12:47:20 hello guys. 2021-04-23 12:48:03 in edge privoxy comming all set config files with .new extension so privoxy fail to start. 2021-04-23 12:48:42 some body can fix this or is intentionally leave that? 2021-04-23 12:50:46 .apknew? 2021-04-23 12:51:16 ikke: no /etc/privoxy/ (all config files as .new extension( 2021-04-23 12:51:39 so have to rename and privoxy start fine. 2021-04-23 12:52:02 i see this more than a month. 2021-04-23 12:52:19 Right, it's named like the 2021-04-23 12:52:21 i got no much time change the job and im outside all the time. 2021-04-23 12:52:38 let me see in aports. 2021-04-23 12:52:44 that in the package apperently 2021-04-23 12:53:32 yeah. i build to void. and have to rename. 2021-04-23 12:54:23 right now im on edge, because qt5 crash qutebrowser 2021-04-23 12:54:58 when stable update qt5 and qutebrowser2 switch back to stable. 2021-04-23 12:57:15 ikke: got to go have nice day. and keep alpine up! 2021-04-23 12:57:31 thanks 2021-04-23 14:09:03 i guess i found the origin of my problems regarding the kernel package for zfs 2021-04-23 14:18:21 the /lib/modules folder is on a ro filesystem, but mount does not indicate why... 2021-04-23 14:18:53 any thoughts? (alpine 3.13@rpi4) 2021-04-23 14:20:21 Modloop 2021-04-23 14:30:56 Hi. I'm running Proxmox at home. Is there a way to download a cloud-init enabled image and if not, how do i make one myself? 2021-04-23 14:31:26 We're still working on that 2021-04-23 14:32:08 There is a cloud-init package 2021-04-23 14:32:30 is there a way i can make my own template? 2021-04-23 14:33:00 i just want to be able to deploy with cloud-init through terraform so i can automate the whole stack 2021-04-23 14:33:15 I guess so, but I'm not familiar with proxmox or cloudinit 2021-04-23 14:35:13 i think cloud-init is being executed on first boot and it will write a file to determine if it's the first boot 2021-04-23 14:35:48 Yes, the package in alpine should already support that 2021-04-23 14:37:04 There is a setup-cloud-init script 2021-04-23 14:38:55 really? haven't seen. i gotta test that, hold on 2021-04-23 14:39:46 You need to install cloud-init first 2021-04-23 14:41:09 already doing that on a test vm 2021-04-23 14:41:33 There I'd also a Readme file included 2021-04-23 14:41:56 o/ 2021-04-23 14:42:25 someone already written up on how to setup a fully headless raspberry pi install with the generic arch minirootfs? 2021-04-23 14:42:43 we don't want to use the diskless one because it does not mix well with our appliance 2021-04-23 14:43:13 the sys install wiki page works fine but you actually need a screen and keyboard plugged in 2021-04-23 14:43:35 @ikke: https://imgur.com/a/QLqanaI looks good already 2021-04-23 14:44:07 markand: don't have a screen and keyboard? 2021-04-23 14:44:53 Gaming4LifeDE, no, when we build our software we would like to automate the creation of the appliance image, so no user interaction to create it 2021-04-23 14:45:22 for example we are using raspbian at the moment, I mount the disk image and I install our software on it and then its fine 2021-04-23 14:45:31 but the problem is that I want us to use alpine instead ;p 2021-04-23 14:46:34 how do you want the installation process to work in the end? like a pre-configured image you can just dump onto a raspi? are any other supporting services (like pxe) allowed? 2021-04-23 14:48:38 my idea is to generate a new minimalist .tar.xz archive pre-configured each time we release a new version that we extract on a correctly partitioned SD card. the appliance has no internet connectivity so basically our process is "flashing" the SD card like this 2021-04-23 14:49:01 but the diskless one is insuitable for us because we need persistent rw storage 2021-04-23 14:50:32 for example if it was a generic x86 machine I would just download the minirootfs, chroot to install some packages and install our software and that's it 2021-04-23 14:51:52 I guess the generic arm minirootfs could be done as long as I add the required raspberry pi files too, but there are so many I don't know which ones are really required 2021-04-23 14:57:59 maybe minirootfs + linux-rpi2 package is just enough (and a correct config.txt/cmdline.txt) :D 2021-04-23 15:06:12 work perfectly, alpine <3 2021-04-23 15:18:55 Gaming4LifeDE: Hi, I'm the Alpine package maintainer for cloud-init 2021-04-23 15:19:35 markand: how about you make an image you can just dd onto a fresh sd card and use something like lift (https://github.com/bjwschaap/alpine-lift) to do config and to deploy your application? then you send send out fully prepares pis 2021-04-23 15:19:49 minimal: wanted to tag you, but you weren't in this channel :) 2021-04-23 15:20:04 minimal: cool to see the maintainers on here (especially for the package you're trying to figure out) 2021-04-23 15:20:09 ikke: was taking care of important business - having a coffee ;-) 2021-04-23 15:20:12 ha 2021-04-23 15:20:29 so for proxmox: cloud init works on boot but apparently it can't read from ConfigDrive mounted via SATA 2021-04-23 15:21:02 oh btw: please mention me if you want me to respond, i'm trying to build and image for cloud-init 2021-04-23 15:21:09 I read in the readme it looks for a volume called cidata 2021-04-23 15:22:17 ikke: cidata is used by the NoCloud data source 2021-04-23 15:22:28 ah ok 2021-04-23 15:23:00 Gaming4LifeDE: are you trying to build a prepared VM for Alpine? 2021-04-23 15:23:18 minimal correct 2021-04-23 15:24:15 easiest way is to install Alpine inside a chroot, add cloud-init to it, do any other config in it, and then write that out as a disk image so boot the disk image 2021-04-23 15:25:00 i want to practice and build my homelab in a fully automated way. I'm currently using terraform to deploy alpine lxc containers (which i built a custom image for too) and deploy software with pyinfra or ansible 2021-04-23 15:25:14 you'll need to decide which DataSource to use and enable it in the cloud-init config - in /etc/cloud-init/cloud.cfg (from memory) 2021-04-23 15:25:37 minimal proxmox's way of doing things is making a vm, installing everything your want (including cloud-init) and then converting it to a template 2021-04-23 15:25:47 and then you just clone it 2021-04-23 15:26:10 don't really like it this way myself but that's just how it is 2021-04-23 15:27:51 so then you could "apk add cloud-init" in your VM, run setup-cloud-init, tweak the cloud-init config, and write that as your template. However the act of booting the original VM in the 1st place may well change settings that uou don't want in your template (i.e. hostname, network config, etc) 2021-04-23 15:28:39 minimal after completing the config, you shutdown and then convert the vm into a template. templates are not bootable, you can only clone them 2021-04-23 15:29:28 Gaming4LifeDE: I have my own Ansible playbook to create a file to use as a disk image, loop mount it, chroot inside it and install alpine etc, and then unmount it - result is disk image ready to go 2021-04-23 15:29:53 and how would i use that with proxmox? 2021-04-23 15:30:46 Gaming4LifeDE: I meant that when you install Alpine in a VM in the first place (in order to prepare a template) you want to make sure that host-specific config doesn't end up in your template 2021-04-23 15:31:16 ohhh, doesn't the setup-cloud-init script wipe host-specific data? 2021-04-23 15:32:09 I'm not familiar with Proxmox, but its just a Docker and KVM/qemu host OS right? For local VMs here I use "qemu-img" to convert VM RAW disk images into QCOW2 files for qemu use 2021-04-23 15:32:29 lxc and kvm/qemu but yes 2021-04-23 15:33:28 Gaming4LifeDE: "setup-cloud-init" does exactly what it says, it sets up cloud-init :-) If (for example) you start sshd on your installed VM it will create SSH host keys - you don't want them ending up in your template as then all VMs created from template will have same host keys 2021-04-23 15:33:46 s/it will/sshd will/ 2021-04-23 15:34:26 makes sense. is there any way to make an install iso out of a template which would just auto deploy the image to the first harddisk? 2021-04-23 15:35:13 I've been meaning to write a simple shell script (proably derived from Alpine make-vm-image and or alpine-chroot-install scripts) to create cloud-init enabled images for KVM/qemu use. 2021-04-23 15:36:13 that would be amazing actually. i wonder how hard it would be. you'd basically need a minimal os (so probably another lapine setup) which would just deploy the prepared image 2021-04-23 15:37:34 i also thought about if it'd be possible for cloud-init to nuke itself after doing it'S job 2021-04-23 15:37:40 normally when you boot Alpine from an ISO its a "run from RAM" - if you provided that VM with a disk device (from qemu/KVM) then you could use "alpine-setup" and pick "SYS" mode to install on the disk device 2021-04-23 15:38:14 basically yeah. and i think sys mode would be the only feasible mode for cloud or cloud-like setups really 2021-04-23 15:38:31 Gaming4LifeDE: I'm actually using cloud-init for disk images I build for VMs and also physical x86_64 and Raspberry Pi machines 2021-04-23 15:38:40 interesting 2021-04-23 15:39:31 how do you do that? have the cloud-init data on a webserver and poll it on first boot? 2021-04-23 15:39:51 Gaming4LifeDE: cloud-init certainly wasn't designed for "run from RAM" setups but it might work ok, haven't tried it 2021-04-23 15:41:02 i'd think you'd get the generated image, dump that into the filesystem of the iso file and then when you boot the iso it would just dd onto the first disk 2021-04-23 15:41:59 for physical machines I create a very small "cidata" partition with user-data.yaml and network-config.yaml files used at boot time - for Raspberry PI for example I ""dd" the disk image onto multiple SDcards and then just need to edit those 2 files on each SDcard before booting (as I'm using "static" networking) 2021-04-23 15:43:08 why don't you use static dhcp? (instructing the dhcp server to hand out a specific ip to a mac address) 2021-04-23 15:43:16 Gaming4LifeDE: yeah for x86_64 physical I have a stripped down Alpine OS which I boot and "dd" the disk onto the SSD/HDD. I haven't yet setup a PXE Network equivalent version 2021-04-23 15:44:02 Gaming4LifeDE: its a bit hard to dhcp boot your dhcp server and you'd typically want some/all servers to have static IP addresses ;-) 2021-04-23 15:44:21 ehh ok makes sense for the dhcp server 2021-04-23 15:45:22 basically with cloud-init you don't install it on an already running "normal" system, you use it when building disk images to be later booted 2021-04-23 15:45:49 well you need to build the image somehow 2021-04-23 15:46:47 yupe, I guess you could use something like Packer to automate an install in a VM, I preferred using a chroot install 2021-04-23 15:47:45 the team build the official Alpine images for AWS (which currently use a cloud-init alternative) are using Packer to create the images 2021-04-23 15:47:53 s/build/building/ 2021-04-23 15:49:24 oh a collaegue of mine is having an issue with packer and alpine, i don't really know what exactly is going on but oh well 2021-04-23 15:50:15 i also think cloud-init is a lot more elegant than packer (i mean, autotype is the most basic and unreliable form of automation) 2021-04-23 15:52:12 well they're targetting 2 different tasks - Packer is for preparing images with whatever content is in them, cloud-init is a boot-time configuration-related tool 2021-04-23 15:52:29 so i'm really clueless about how to make this work properly... 2021-04-23 15:52:55 minimal they're using it for both... 2021-04-23 15:53:23 so if you add a 2nd disk device to a VM Packer can't do anything as its not used on the running VM whereas cloud-init can (depending on config) partition and setup filesystems on it automatically :-) 2021-04-23 15:57:26 basically 2021-04-23 16:14:20 i g2g. Thanks for you insights, minimal 2021-04-23 16:21:43 Gaming4LifeDE: sorry, on phome 2021-04-23 16:21:53 ping me later for a chat 2021-04-23 16:21:54 minimal: Gaming4LifeDE left 2021-04-23 16:58:40 ikke: thanks. helps a lot :P 2021-04-23 17:07:53 just a single word :P 2021-04-23 17:10:28 I would also be interested in a cloud-init image of alpine for my proxmox setup. It's not a big deal, would just be handy. 2021-04-23 17:20:17 programmerq: I've been meaning to find some time to create a script for "baking" such disk images for KVM use 2021-04-23 17:21:24 if you do, feel free to ping me to be a guinea pig. proxmox is basically just a kvm gui. 2021-04-23 17:22:02 I'll see if can find some time after next week (when the next upstream release of cloud-init is due out) 2021-04-23 17:22:19 I never got the hang out of kvm or whatever virtual machine 2021-04-23 17:44:16 proxmox makes it pretty simple, and I can still drop down to a config file to feed raw options if I need to (rare) 2021-04-23 18:04:17 hi guys, after lot of upgrades on edge with "apk upgrade -a" and reboot, I noticed that konsole starts in / instead my home dir 2021-04-23 18:19:14 anyone knows how to disable GNOME from suspending when my external screen shut off? 2021-04-23 18:19:17 it drives me crazy 2021-04-23 18:19:39 in settings I've disabled every option related to automatic suspend 2021-04-23 18:22:17 oh looks like it's in privacy -> screen lock, rather than in power 2021-04-23 23:38:39 anyone use perl on alpine? I'm trying to figure out how to get cpan. searching the repo isn't helping me much, they all seem to be tools around cpan. 2021-04-23 23:39:03 (it's for a super important reason - someone told me about ascii acquarium) 2021-04-24 03:56:22 How can I restore my configuration on reboot after choosing the store on usb option during setup? 2021-04-24 03:59:43 using a VPS- do i need to create a small partition on the disk and mount it? 2021-04-24 04:05:08 Payupsir: Which mode is this? data/sys/none ? 2021-04-24 04:05:18 data 2021-04-24 04:05:38 What do you mean by configuration ? 2021-04-24 04:05:44 File in /etc? 2021-04-24 04:06:07 when running setup-alpine during first boot 2021-04-24 04:06:23 and configuring hostname, keyboard layout etc. 2021-04-24 04:06:36 My configuration is not persisting after reboot 2021-04-24 04:06:55 Right 2021-04-24 04:07:00 You need to use lbu 2021-04-24 04:07:23 I'm in a little over my head /: should probably start with an easier distribution. 2021-04-24 04:07:35 It's actually pretty simple 2021-04-24 04:07:40 Let me explain to you 2021-04-24 04:08:40 With lbu, you have directories you want to "keep" or "save" for when you reboot. When you do `lbu commit -d`, this takes all these files/directories and puts them in a *.tar.gz file. In your case, it's done on your usb key. 2021-04-24 04:09:03 When alpine reboots, it will unpack this big *.tar.gz file, restoring your configuration. 2021-04-24 04:10:34 Neat! is there a way to run Alpine in RAM on a virtual private server where I can't provide a physical USB key? 2021-04-24 04:13:41 i guess this is an edge case as most are doing things in Docker 2021-04-24 04:14:38 That's the none option (diskless). Essentially, everything runs in RAM. You can still use lbu if you have any other storage (sdcard, cdrom, etc). 2021-04-24 04:15:22 The data option also runs from RAM, except stuff in /var/ are stored on disk/usb/whatever. It's like an hybrid between sys and none. 2021-04-24 04:16:51 I have 32GB of SSD allocated to me, can i store my system configuration (Hostname, Time-Zone, Keyboard Layout) on this storage? 2021-04-24 04:17:19 yeah I don't see why you couldn't 2021-04-24 04:18:02 How would i do that? just a simple lbu commit usb after running setup-alpine then reboot? 2021-04-24 04:20:11 sorry, probably a if you need to ask use a different distro situation lol 2021-04-24 04:20:54 I'm not entirely sure because I've just used `none` so far. But I think all you need to do is format properly the SSD, make sure it mounts at the right place, and then configure lbu to store its config to that drive. I think setup-disk helps you do that for you, but not 100% sure 2021-04-24 04:21:28 Look at this file: /etc/lbu/lbu.conf 2021-04-24 04:24:11 Do i even need to reboot after install? 2021-04-24 04:24:30 Do any of you use Alpine on PowerPC? If so, is that a desktop? 2021-04-24 04:24:53 is that necessary, I've always done it with debian as i assumed it was best practice 2021-04-24 04:26:04 After installing a new package or the entire system? 2021-04-24 04:26:16 the system 2021-04-24 04:26:50 Payupsir: Good question, I'd say follow the instructions (I'm assuming we have some for this..) 2021-04-24 04:27:42 For example, for raspberry pi I use this: https://wiki.alpinelinux.org/wiki/Raspberry_Pi_-_Headless_Installation which does say reboot at the end. 2021-04-24 04:28:40 https://unix.stackexchange.com/questions/382943/how-to-load-alpine-linux-backed-up-configuration-on-boot 2021-04-24 04:29:23 found this which is sorta similar to my question- when the solution says "create a disk" does this mean format and create a small partition? 2021-04-24 04:30:47 Yeah, it's saying to partition the disk and mount it on the system with fstab 2021-04-24 04:31:55  lbu ci 2021-04-24 04:32:03 is this shorthand for lbu commit? 2021-04-24 04:32:32 Yeah, you can see that using `lbu --help` 2021-04-24 04:34:49 Thanks, can i create this partition with lvm? 2021-04-24 04:37:33 afaik it doesn't matter how you create it, as long as you have something mounted+writable at the end 2021-04-24 04:44:19 Do i need to create this partation before the disk is partitioned for the data configuration? 2021-04-24 04:49:48 I have some learning to do apparently 2021-04-24 04:50:54 How many disks do you have ? 2021-04-24 04:56:57 one 2021-04-24 04:57:12 the 32GB ssd 2021-04-24 04:59:29 Data Disk Mode 2021-04-24 04:59:30 This mode is still accelerated by running the system from RAM, however swap storage and the whole /var directory tree gets mounted from a persistent storage device (two newly created partitions). This location holds e.g. all log files, mailspools, databases, etc., as well as lbu backup commits and the package cache. The mode is useful for having 2021-04-24 04:59:30 RAM accelerated servers with amounts of variable user-data that exceed the available RAM size, and to let the entire current system state (not just the boot state) survive a system crash according to the particular filesystem's guarantees. [Fixme: Storing lbu configs to disk is not auto-configured after configuring the data partition, one still has 2021-04-24 04:59:31 to select saving configs to "none" first (the new data partition is not listed), and to manually set e.g. LBU_MEDIA=sda2 in /etc/lbu/lbu.conf and execute echo "/dev/sda2 /media/sda2 vfat rw 0 0" >> /etc/fstab afterwards, to have the partition (here sda2) used.] The boot device may remain to be the one initially used, and can even be immutable 2021-04-24 04:59:31 (read-only). 2021-04-24 04:59:38 sorry for wall of text 2021-04-24 04:59:58 can i not just lbu commit to the partition this creates? 2021-04-24 05:03:41 Yeah, it says this location holds the lbu backup commits 2021-04-24 05:07:08 ok so how would i do a backup? 2021-04-24 05:07:16 sorry I'm 100% noob 2021-04-24 08:17:06 Hi, I'm trying to do an upgrade on edge. It seems some python3 packages have been upgraded to python3.9, but my python3 package still stays on 3.8 2021-04-24 08:17:16 How can I figure out how to upgrade python3 to 3.9? 2021-04-24 08:18:37 parazyd: first: try apk upgrade --available 2021-04-24 08:18:57 Doesn't help 2021-04-24 08:20:13 oh, it seems `apk del python && apk upgrade` helped 2021-04-24 08:20:17 The former deleted python2 2021-04-24 09:09:13 is it /etc/profile.de/crap.sh run after every spin of new terminal if login shell? 2021-04-24 09:09:43 did not expect that spam 2021-04-24 09:15:35 seems like it 2021-04-24 09:45:51 yes, makes sense, I did not think about that 2021-04-24 10:44:35 Hello :) If I would like to see some package moved to community from testing, is there some formal procedure? Or should I just send a patch and wait for someone to merge it? 2021-04-24 10:45:57 gray-_-wolf: are you the maintainer? 2021-04-24 10:48:52 No. So the process is to ask maintainer to do so? 2021-04-24 10:49:40 You could create an MR and ask the maintainer to approve 2021-04-24 10:53:40 Thanks will try :) 2021-04-24 11:54:01 Anyone willing to try my openrc-exporter package? Would want some feedback on the software and the packaging. (It's a Prometheus exporter for OpenRC services) 2021-04-24 12:45:07 sure, that's actually something I've been wanting to monitor 2021-04-24 13:25:02 when using alpine in diskless mode does it just unpack all the apks at boot? 2021-04-24 13:25:36 yes 2021-04-24 13:26:34 is there a simple way to delete apks that aren't currently installed? 2021-04-24 13:38:09 apk cache clean 2021-04-24 13:39:09 that leaves the cache directory with some apks in it that i've definitely uninstalled 2021-04-24 13:45:22 hi there, does `ash` source a ".bashrc-like" file upon startup? I see there's /etc/profile and some files under /etc/profile.d but I seemingly have to source them manually. Is there somewhere I can configure it? 2021-04-24 13:45:40 also, does the bashrc syntax work for it? 2021-04-24 13:50:48 krimmspare: there is ~/.profile /etc/profile netc /etc/profile.d/*.sh 2021-04-24 13:50:56 s/netc/and 2021-04-24 13:51:03 but they are only sourced for login shells 2021-04-24 13:53:24 ikke: thanks, just noticed I don't have ~/.profile under my $HOME 2021-04-24 13:53:53 You need to create it 2021-04-24 13:53:58 I suppose that once I create that one it's sourced automatically when I log in? 2021-04-24 13:54:03 yes 2021-04-24 13:54:03 got it. 2021-04-24 13:54:22 does the bashrc syntax work? 2021-04-24 13:54:33 ash is not bash 2021-04-24 13:54:51 it's mostly similar, but bash has some additions that ash (posix sh) does not have 2021-04-24 13:54:54 ok, does it work with sh syntax? 2021-04-24 13:54:58 ah, right 2021-04-24 13:54:58 yes 2021-04-24 13:55:13 got it, I'll experiment with it then 2021-04-24 13:55:27 so no arrays, no [[ 2021-04-24 13:55:42 cool, no bash-specific stuff 2021-04-24 13:55:47 thanks! 2021-04-24 13:56:01 and btw, let me thank you all for this awesome project 2021-04-24 13:56:26 'been using it for a few days and I'm hooked already! 2021-04-24 13:56:36 You're welcome :) 2021-04-24 14:03:43 Alpine r00lz 2021-04-24 14:03:49 2021-04-24 14:18:44 fys: :D 2021-04-24 14:32:34 ;) 2021-04-24 16:55:36 is ash the same sh in busybox? 2021-04-24 16:56:09 or there is some extra over sh? 2021-04-24 17:02:47 helby: ash is the shell provided by busybox 2021-04-24 17:03:08 on alpine 2021-04-24 17:03:19 If by "extra over sh" you mean differences between ash and a POSIX shell, ash has a few niceties built-in that aren't POSIX 2021-04-24 17:05:34 yes, I mean where can I find differencies between both 2021-04-24 17:05:49 or about ash 2021-04-24 17:06:04 https://en.wikipedia.org/wiki/Almquist_shell 2021-04-24 17:06:08 Cogitri: not sure what sh is in busybox 2021-04-24 17:06:22 ash 2021-04-24 17:06:50 only yes? sh will produce the same? 2021-04-24 17:08:59 so from wiki I got that it's dash of some version :/ 2021-04-24 17:09:13 never thought that manual for shell is in wiki ;) 2021-04-24 17:09:19 sh is symlinked to busybox 2021-04-24 17:09:28 which includes ash for sh 2021-04-24 17:09:44 yes, both are symlinked to busybox obviously 2021-04-24 17:09:54 I was just not sure if both are the same 2021-04-24 17:09:57 yes 2021-04-24 17:10:25 thanks 2021-04-24 17:10:54 if I want some kind of manual for ash or what everything it's included, should I then check that version of dash? :/ 2021-04-24 17:11:10 or just some version from many of sh? 2021-04-24 17:12:34 You should check the busybox implementation, sadly their documentation is terse 2021-04-24 18:42:32 who said /etc/udhcpc/udhcpc.conf was a thing 2021-04-24 18:42:38 < 2021-04-24 18:42:42 theres no mention of it in busybox source or anything else 2021-04-24 18:42:55 It's our dhcp script that sources it 2021-04-24 18:43:07 i grepped in /etc/init.d and /etc/conf.d 2021-04-24 18:43:11 /usr/share/udhcpc/default.script 2021-04-24 18:43:11 is it somewhere else 2021-04-24 18:43:14 oh 2021-04-24 18:43:15 thanks 2021-04-24 18:43:23 chasing around scripts is fun 2021-04-24 18:43:50 yup 2021-04-24 19:04:40 yes it's fun, will see when you will be without internet :/ 2021-04-24 19:05:50 what 2021-04-24 19:39:01 opal: just joking about fuckering with udhcpc 2021-04-24 19:59:44 Any way to view hard drive serial numbers using a default tool or file system (such as `/sys`)? Can't find the typical /sys/block//device/serial` file 2021-04-24 20:00:47 smartctl? 2021-04-24 20:00:48 I guess it depends on the driver whether that's exposed? 2021-04-24 20:02:26 Just a normal hard drive over sata. Wouldn't really need any special drivers, right? 2021-04-24 20:03:34 Doesn't the setup-alpine script display the serial# when asking which to format as well? 2021-04-24 20:04:01 Specifically setup-disk ig 2021-04-24 20:04:56 blkid? 2021-04-24 20:15:19 lsblk but that's from util-linux and i think also requires /sys 2021-04-24 20:15:25 check that /sys is actually mounted 2021-04-24 20:18:41 I checked on a few servers, I don't see it either 2021-04-24 20:18:44 just 'serial' missing 2021-04-24 20:22:02 ikke `blkid` only show the UUID 2021-04-24 20:22:09 yeah 2021-04-24 20:22:16 I'm wanting serial 2021-04-24 20:23:23 For example, this is the model number `/sys/block/$d/device/model` 2021-04-24 20:25:35 ikke do you have another linux install besides alpine to cat out `/sys/block/$d/device/serial` on? Just to see what I'm talking about 2021-04-24 20:26:05 On my arch install I don't have it either 2021-04-24 20:26:15 Hmm 2021-04-24 20:26:17 Odd 2021-04-24 20:29:13 wwid number is not serial? 2021-04-24 20:31:15 Serial is like make/model 2021-04-24 20:32:41 Similar to `/dev/disk/by-id` 2021-04-24 20:33:07 https://unix.stackexchange.com/questions/121757/harddisk-serial-number-from-terminal 2021-04-24 20:33:40 All of those aren't with default commands or fs 2021-04-24 20:33:56 yup 2021-04-24 20:35:06 ikke Since you're on arch, do `ls /dev/disk/by-id` First letters are the vendor. Next is the model. Last bits are the serial# 2021-04-24 20:35:28 For whatever reason, alpine doesn't have by-id 2021-04-24 20:35:37 You need udev for that iirc 2021-04-24 20:35:55 yes 2021-04-24 20:36:04 Yea 2021-04-24 20:36:16 wwid is like serial but not 2021-04-24 20:36:25 Maybe udev is why /sys doesn't show serial either? 2021-04-24 20:36:31 no 2021-04-24 20:36:39 for me wwid is the same like serial 2021-04-24 20:36:52 helby: where do you get the wwid from? 2021-04-24 20:37:21 /sys/block/sda/device/wwid 2021-04-24 20:37:52 right 2021-04-24 20:38:25 wwid isn't the serial at all 2021-04-24 20:38:29 it's the same like that udevadm info .... shit at least for me 2021-04-24 20:40:46 yes, it's that id 2021-04-24 20:40:56 in by-id 2021-04-24 20:41:34 udev has both ata model/serial and wwn 2021-04-24 20:42:07 although some vendors may use similar serial and wwn 2021-04-24 20:42:30 i'm pretty sure serial was never in /sys for sata/scsi 2021-04-24 20:42:35 only ide and nvme 2021-04-24 20:42:48 btw alpine has 'by-id' 2021-04-24 20:42:59 yes, *if you install udev* 2021-04-24 20:43:01 oh 2021-04-24 20:43:04 Yea... 2021-04-24 20:43:04 for the... fourth time? 2021-04-24 20:43:10 Pretty much 2021-04-24 20:43:22 my laptop with nvme does not have serial 2021-04-24 20:43:31 ikke: did you check device/ 2021-04-24 20:43:34 yes 2021-04-24 20:43:57 so you are more lightweight :0 2021-04-24 20:45:16 I can get it from udev if necessary. Probably will always have it installed anyways. Just didn't want to install a tool I wouldn't already be installing for the most part 2021-04-24 20:45:39 /sys/block/nvme0n1/device/subsystem == ../../../../../../class/nvme/? 2021-04-24 20:46:31 https://github.com/torvalds/linux/commit/1e496938b6aec67b7cfd06614164fee0fe9f13e3 since 4.15-rc1 2021-04-24 20:47:31 `../../../../../../../bus/scsi` for sata usually 2021-04-24 20:47:41 yes, i know 2021-04-24 20:47:44 Idk about nvme 2021-04-24 20:47:57 i think there was some fiddling with libata at some point 2021-04-24 20:53:36 Thanks for the help guys tho 2021-04-24 20:55:23 funny thing is i think busybox hdparm handles this 2021-04-24 20:55:45 although for some reason it's broken on model and firmware rev 2021-04-24 20:59:21 cvoges12: you can join mps in counting packages and being insistently opposed to a... 132 kB disk usage 2021-04-25 00:46:34 . [Fixme: Storing lbu configs to disk is not auto-configured after configuring the data partition, one still has to select saving configs to "none" first (the new data partition is not listed), and to manually set e.g. LBU_MEDIA=sda2 in /etc/lbu/lbu.conf and execute echo "/dev/sda2 /media/sda2 vfat rw 0 0" >> /etc/fstab afterwards, to have the 2021-04-25 00:46:34 partition (here sda2) used.] The boot device may remain to be the one initially used, and can even be immutable (read-only). 2021-04-25 00:46:50 is this still applicable when using the data option during initial setup 2021-04-25 01:06:28 ? 2021-04-25 06:17:11 Is /etc/init.d/networking gone from latest openrc in edge? 2021-04-25 06:17:39 What's the replacement? 2021-04-25 06:24:46 parazyd: it's not gone, but for some reason, upgrading removes it 2021-04-25 06:25:01 parazyd: apk fiex openrc should restore it 2021-04-25 06:27:37 Gotcha, thanks 2021-04-25 06:44:53 apk fix* 2021-04-25 08:25:23 hi, i'm new to apine-linux, can you please brief me, do i install alpine image on a bare metal or it needs to be installed into a docker/virtual machine ? 2021-04-25 08:25:43 It can certainly be installed on bare metal 2021-04-25 08:25:57 won't it be limited ? 2021-04-25 08:26:04 limited in what sense? 2021-04-25 08:26:11 We use alpine on bare metal for everything 2021-04-25 08:26:31 oh, had i known :( 2021-04-25 08:26:48 It's a full OS, not just limited to docker / vms 2021-04-25 08:27:16 ok, 2021-04-25 09:06:40 I never understood what dockers do 2021-04-25 09:09:30 docker is infrastructure to build and distribute chroots + infrastructure to run these chroots with support around networking and volumes 2021-04-25 09:10:19 where each chroot typically runs a single executable (which in turn can run other executables), which is called a container 2021-04-25 09:12:22 containers were not something docker invented, but they made it convenient to build and run these containers 2021-04-25 09:33:41 aka orchestration 2021-04-25 09:56:28 Hello everybody 2021-04-25 09:56:42 I compiled oscam-patched and I don't know how to start it automatically 2021-04-25 09:56:43 please help me with an automatic startup script 2021-04-25 09:56:43 thank you 2021-04-25 09:57:14 adria_: Welcome 2021-04-25 09:57:19 do you want to run it as a service? 2021-04-25 09:58:14 yes,or to boot start automatically 2021-04-25 09:58:34 Maybe start with https://github.com/OpenRC/openrc/blob/master/service-script-guide.md 2021-04-25 09:59:24 now I start it manually /usr/local/bin/oscam 2021-04-25 09:59:49 I'm new to linux 2021-04-25 10:01:21 adria_: You could try to start with https://tpaste.us/E1lJ 2021-04-25 10:01:32 in /etc/init.d/oscam 2021-04-25 10:04:46 after reboot it does not start automatically 2021-04-25 10:05:12 Did you create that service? 2021-04-25 10:05:19 yes 2021-04-25 10:05:27 and it's working when you start the service? 2021-04-25 10:05:32 rc-service start oscam 2021-04-25 10:05:47 yes 2021-04-25 10:05:54 work 2021-04-25 10:06:12 rc-update add oscam 2021-04-25 10:06:40 service oscam added to runlevel default 2021-04-25 10:06:49 Now it should start at boot 2021-04-25 10:07:20 I'm testing now 2021-04-25 10:10:53 https://i.postimg.cc/nLYdZrtR/IMG-20210425-130840.jpg 2021-04-25 10:11:46 it doesn't start anymore, I can't connect ssh 2021-04-25 10:28:16 I stopped oscam from the web interface and the server started 2021-04-25 10:29:24 Does oscman have options to fork to the background? 2021-04-25 10:35:01 I tested this script 2021-04-25 10:35:15 https://git.alpinelinux.org/aports/tree/community/oscam/oscam.initd?h=3.3-stable 2021-04-25 10:35:44 same problem 2021-04-25 10:36:05 it doesn't start anymore, I can't connect ssh 2021-04-25 10:37:33 add to /etc/init.d/oscam 2021-04-25 10:37:36 command_background =" yes " 2021-04-25 10:39:10 * oscam: error loading /etc/init.d/oscam 2021-04-25 10:39:10 * ERROR: oscam failed to start 2021-04-25 11:06:11 <[diablo]> Good afternoon #alpine-linux , hope you're all well. Hoping someone can help here. So I'm trying to do an NFS mount , and passing the option no_root_squash, and I get back "Invalid argument" 2021-04-25 11:06:19 @ikke I solved 2021-04-25 11:06:26 adria_: nice 2021-04-25 11:06:27 @ikke 2021-04-25 11:06:29 <[diablo]> any ideas please? 2021-04-25 11:07:11 @ikke 2021-04-25 11:07:13 I solved 2021-04-25 11:07:18 yes, I rad 2021-04-25 11:07:21 read 2021-04-25 11:07:24 https://tpaste.us/0Enx 2021-04-25 11:07:43 <[diablo]> oh fwiw, also "mount.nfs: an incorrect mount option was specified" 2021-04-25 11:07:49 I don't know if it's correct 2021-04-25 11:09:36 Not sure if the start_stop_daemon_args are necessary 2021-04-25 11:14:27 I deleted start_stop_daemon_args 2021-04-25 11:15:07 works 2021-04-25 11:15:15 thank you 2021-04-25 12:34:20 Hello, does anyone has any idea what happened to mkmkr and where could I that tool? It was briefly mentioned on the mailing list, but I did not manage to find any links. 2021-04-25 12:36:05 gray-_-wolf: The maintainer abandoned it, and switched to glab 2021-04-25 12:36:10 an existing tool 2021-04-25 12:39:44 currently trying lvgl on a raspberry pi 4 running alpine to create a embedded app 2021-04-25 12:39:52 using raw linux fb == totally awesome 2021-04-25 12:40:24 once complete I think I can even get rid of openrc to reduce the number of packages entirely 2021-04-25 13:01:22 I am using tinyalsa which include 'tinyplay' crap, using it for 'beep' notification :) however it always hangs if something else is using sound and it stays like that until another program will finish with sound 2021-04-25 13:02:17 I would not like to use 4MB package alsa-utils with depends just to unmute and adjust volume of audio 2021-04-25 13:03:13 alsa can be configured for mixing inputs, iirc, but I don't recall the details. 2021-04-25 13:04:10 it works well for anything, just that part - tinyplay can't get access or whatever 2021-04-25 13:10:10 "i installed alsa library without software mixing, why doesn't it do software mixing" 2021-04-25 13:10:17 -.- 2021-04-25 13:11:33 I have alsa-lib too, as ffmpeg-lib requires it 2021-04-25 13:12:01 but it's probably not enough to 'volume control' 2021-04-25 13:38:12 obviously there is one another annoying issue 2021-04-25 13:38:34 in wayland I can't get ffplay to play videos, while it plays, it does not show that window with video 2021-04-25 13:39:37 installed some mesa-gl shit as it requires otherwise complains about some sdl whatever 2021-04-25 13:41:12 of course ffplay is crap, but if ffmpeg-libs already installed as I want firefox play also porn :/ it's relatively lightweight player as everything it needs is already installed 2021-04-25 13:42:35 so ffplay in wayland somehow acts like with '-nodisp' options, it does not shows me video window :/ 2021-04-25 13:42:46 that's really minimal video play 2021-04-25 13:42:54 does ffplay even support wayland 2021-04-25 13:43:00 if not, find something that does. 2021-04-25 13:43:03 no idea 2021-04-25 13:43:11 did not find anything that it does not support 2021-04-25 14:46:52 ffplay is test program 2021-04-25 14:46:57 not intended for real use 2021-04-25 14:47:11 not sure it even has vsync 2021-04-25 15:11:38 well it's media player 2021-04-25 15:11:53 and normally it play audio video without an issue ... in x 2021-04-25 15:12:51 again, its intended purpose is testing; it is not intended to be a media player. try mpv or mplayer. 2021-04-25 15:39:06 someone is using xtetv? 2021-04-25 16:47:53 someone is using xtetv? 2021-04-25 16:47:55 https://github.com/xteve-project/xTeVe 2021-04-25 16:49:49 please help me with an automatic startup script 2021-04-25 17:45:11 there is no longer mesa-dri-vc4 for armv7? 2021-04-25 17:45:20 it was presentin 3.10 2021-04-25 17:47:07 ah they have been merged in -gallium 2021-04-25 18:18:24 I think everyting will be in gallium now, only old shit for my potato stays in classic :/ 2021-04-25 19:23:49 it's slow as hell 2021-04-25 19:23:57 can't even play a NES game on retroarch 2021-04-25 19:24:03 glxgears says 30FPS though 2021-04-25 19:24:50 firefox-esr crashes on armv7 but firefox is fine o aarch64 2021-04-25 19:25:00 pretty hard to select the perfect arch to use for rpi 2021-04-25 19:25:29 I've been told that aarch64 isn't supported for 3D rendering as the official foundation does not provide support 2021-04-25 19:25:29 i am struggling with alternating the modloop for use on an diskless mode. i am quite fine with overlays now, but still struggling with a good way to install the modules via apk. i am now trying to copy the files from the apk local cache and the one from /lib/modules, then generate a squashfs from that. but there must be a intented way thats better... right? 2021-04-25 19:43:55 what did u expect to play AAA games on some shitty arm crap? 2021-04-25 19:54:07 helby, I run NES games on a pi zero, so there is definitely an issue there 2021-04-25 19:54:27 I don't know what lakka folks are doing to improve performances 2021-04-25 20:27:38 Hi, is there any tutorial of installing alpine with EFIStub as 'bootloader'? 2021-04-25 20:37:04 kernelc: https://www.cogitri.dev/posts/04-secure-boot-with-unified-kernel-image/ 2021-04-25 20:37:16 u can ignore that secure boot crap 2021-04-25 20:38:16 very simple, just one command practically 2021-04-25 20:41:53 helby: Thank you, I will test it 2021-04-25 20:44:12 after adding entry to uefi with efibootmgr once, then after every kernel upgrade I just run install binutils and gummiboot, then run that obj command to update .efi 2021-04-25 20:47:17 helby: Is there any better approach? like kernel update hooks or something… sorry if I ask about something obvious I am switching to alpinelinux 2021-04-25 21:03:31 kernelc: well after update of kernel, mkinitfs is ran, so check manual for mkinitfs if u can add that 'obj' command somewhere 2021-04-25 21:04:11 I did not fuckering with it much, as kernel does not update every day, so can do it manualy 2021-04-25 21:04:46 mostly bc. I don't keep installed binutils and gummiboot bc. of that 2021-04-25 21:38:14 helby: did you traied to do it with luks? 2021-04-25 21:43:28 kernelc: never touched anything related to luks 2021-04-25 23:24:55 Hi. I just installed Alpine for the first time on a virtualbox machine and trying to run sway (have elogind running). Sway gives this output and stuck https://pastebin.com/raw/r7DphETc 2021-04-25 23:24:56 what can I do? 2021-04-25 23:42:32 first, consider not using virtualbox 2021-04-26 00:21:24 Hey 2021-04-26 00:22:27 Anybody using the noto fonts? 2021-04-26 00:23:23 Piping terminal output through less is making it show escape sequences, like "ESC[1;34m". 2021-04-26 00:23:37 Any idea on how to fix this? 2021-04-26 00:28:16 which terminal? 2021-04-26 00:28:21 xterm 2021-04-26 00:28:52 have you tried the -R option in less? 2021-04-26 00:29:35 Oh 2021-04-26 00:29:45 It works now. 2021-04-26 00:29:53 cool 2021-04-26 00:29:56 Thanks 2021-04-26 00:30:02 np 2021-04-26 00:30:08 I am trying to install alpine on my raspberry pi zero w and if I go though the setup and select diskless when I reboot the pi does not boot up. No matter what I do I can't install it 2021-04-26 00:36:05 Hello, looking at service file for docker, it looks like it does not depend on iptable. Should it depend on iptables? I think I'm running into race when docker starts up, then iptables and they clear all chains the docker created. So just curious why is there no dependency. 2021-04-26 01:30:29 Hello71, the same issue with qemu :( 2021-04-26 01:37:55 LOLCause: not sure about virtual/qemu shit, but there is usually problem if u don't have mesa-dri-shit driver, maybe add your user to video group, but no idea how it works in virtual env 2021-04-26 01:41:47 helby, my user is in video group and I have mesa-dri-* installed 2021-04-26 01:42:24 on machine, it worked out of box just following very minimal alpine wiki sway instruciton 2021-04-26 01:44:51 helby, OMG! I didn't install eudev🙊 2021-04-26 01:44:57 hi, i downloaded the latest alpine-linux to test it out, but it fails to format the disk in virtualbox. has anybody else had this issue? 2021-04-26 01:45:09 now sway working, thank you :) 2021-04-26 01:45:14 I thought about it but was not sure how it works in virtual env 2021-04-26 01:45:32 i guess alpine is not intended to be installed to disk, rather it's more like a live environment? 2021-04-26 01:45:45 no 2021-04-26 01:45:47 I think it can be installed to disk 2021-04-26 01:46:09 I use is on my only machine 2021-04-26 01:46:13 it 2021-04-26 01:46:15 maybe there's something i'd have to tweak in vbox then. i'll try qemu sometime when i have a spare moment 2021-04-26 01:46:40 it boots fine and all the questions are fine, but when i enter disk to write to the hard disk it freezes on the creating disk message 2021-04-26 01:47:08 johnjay: u can ask LOLCause , he just did the same, but changed to qemu then 2021-04-26 01:47:19 ah ok 2021-04-26 02:39:17 is the install media located on an microSD card? jason123santa 2021-04-26 02:39:40 the ISO 2021-04-26 02:40:52 yeah I have extracted the .tar.gz to my sd card 2021-04-26 02:40:59 and ran setup-alpine 2021-04-26 02:42:01 but no matter what I do it will not install and work 2021-04-26 02:44:48 bc. whole diskless mode is nonsense 2021-04-26 02:45:48 also if I remember well, u have to install it to other media than installation 2021-04-26 02:45:57 as installation is read only 2021-04-26 02:46:39 So I can't have alpine on my microsd card and install it to my microsd card? 2021-04-26 02:49:31 when u boot it, check 'mount' option if it's mounted r what I assume, as read only 2021-04-26 02:49:44 but I never did diskless crap as it's pointless 2021-04-26 02:50:07 diskless is the only way to install it on the pi 2021-04-26 02:53:54 if u have enough storage it should be possible to install it normally 2021-04-26 02:54:27 it does not work on the raspberry pi 2021-04-26 02:54:36 only diskless work 2021-04-26 02:54:53 so then only what it will write to some storage is lbu 2021-04-26 02:54:59 everything else will run from ram 2021-04-26 02:55:01 but I still can't get diskless to work 2021-04-26 02:57:00 I think u have create one writable partition on that storage 2021-04-26 02:57:14 where u can save your lbu cache or whatever is it called 2021-04-26 02:57:59 otherwise nothing is stored and after reboot u have nothing as ram is not persisent 2021-04-26 03:14:19 I can try doing that 2021-04-26 03:14:24 thanks 2021-04-26 03:15:10 normally when u setup-alpine diskless what u get that it does not work? 2021-04-26 03:15:28 when it finish, u get some error or what? 2021-04-26 03:17:10 just check all that outdated nonsense in wiki, something from there will work :/ 2021-04-26 03:18:33 it's even mentioned there, it's possible to install if persistent after you finish diskless successfuly 2021-04-26 03:42:56 Is it possible to make autologin with /sbin/getty which is symlink to /bin/busybox ? 2021-04-26 03:54:34 oops, didn't know I can run /bin/login directly :) 2021-04-26 04:26:05 hello, anyone running wireguard on alpine using the wg-quick up/down scripts on startup/shutdown? 2021-04-26 04:26:53 wondering what the best way to bring up wireguard after boot 2021-04-26 04:27:18 I put it in the interfaces file 2021-04-26 04:27:33 I don't use the wg-quick 2021-04-26 04:29:13 you can still use wg-quick with the interfaces file, though, just add an up line after your main interface 2021-04-26 04:33:10 qman__: thanks! 2021-04-26 05:01:57 qman__: to clarify, would I just append "post-up wg-quick wg0.conf" to the preferred interface entry? 2021-04-26 05:13:23 ok, that worked, well post-up wg-quick up wg0 worked, thanks for the guidance 2021-04-26 06:26:20 shit, tried everything, cat set cursor theme :/ 2021-04-26 06:26:26 *can't 2021-04-26 07:05:20 Hi folks. I've been using Alpine for couple months for develepment purposes without any issues now. Yesterday I had to use pip in order to install some dependencies, but I had an error about TLS/SSL certificates, but I didn't had this like one or two weeks ago (the last time I needed to install python packages). I first thought that there was something broken on pythonhosted.org's servers, so I waited about a day for it to get "fixed". Seems 2021-04-26 07:05:20 like the issue wasn't related to the servers though (the status page says it is up). I've been getting some errors like this when I try to install anything: https://p.sonu.ch/fd80.txt . I've been using v3.13, so I though maybe switching to edge would "fix" it somehow. I did that, upgraded my packages and all, but it also doesn't work. Looking at the GH issues, I can only see people having issues with some sort of proxy settings, but I don't h 2021-04-26 07:05:20 ave any. So, I checked the IRC logs of last two months, but couldn't find anything related. Any clues? :D 2021-04-26 07:06:19 srfsh: `date` returns the correct date + time? 2021-04-26 07:06:44 I didn't think of that. Let me check. 2021-04-26 07:07:03 No, it doesn't... 2021-04-26 07:07:18 It is says it is Apr 19. :D 2021-04-26 07:08:01 Try fixing that then, having an incorrect date will break HTTPS 2021-04-26 07:08:19 Yeah, doing it now. That's kind embarrassing, though. :D 2021-04-26 07:08:41 hi, anyone online ? 2021-04-26 07:08:58 srfsh: Heh, no worries, easy to miss that :D 2021-04-26 07:10:38 Cogitri: Yes, that was really it. Thanks for the reminder! 2021-04-26 07:10:53 Great that that fixed it :) 2021-04-26 07:11:34 Although I thought I had an NTP daemon. 2021-04-26 07:23:07 Hello, I'm trying to boot Alpine from USB for installation to the disk but it seems that it doesn't matter how I do it, the final USB produced from the x86 ISO is not bootable (not even detected by the BIOS). What are my options here? 2021-04-26 07:24:33 I was considering going the Raspberry Pi approach and manually format and extract the files to the USB drive, but I don't see any tar files for that architecture (only netboot ones). Is there any way I can extract it from the ISO? 2021-04-26 08:10:05 Hi, I am looking for a solution for autologin on tty1, is that feasible in alpine? Google has not helped me 2021-04-26 08:11:43 edit /etc/inittab and replace a getty with login. You'll need to read the manual for more in-depth information. 2021-04-26 08:12:41 already tried, does not work 2021-04-26 08:13:20 zcrayfish, you have a link for me? 2021-04-26 08:24:20 Alpine linux can be installed on the microsd card? 2021-04-26 08:24:44 for the Raspberry Pi 3b + 2021-04-26 08:25:41 zcrayfish, i have trid this, but not work ==> tty1::respawn:/sbin/getty -a foo 38400 tty1 2021-04-26 08:26:02 *tried 2021-04-26 08:26:11 I want to run adguard home, xtetv, oscam, openvpn and tvheadend in 2021-04-26 08:26:11 docker 2021-04-26 08:31:16 is there a plan to upstream llvm/clang patches? the trivial ones like https://gitlab.alpinelinux.org/alpine/aports/-/raw/master/main/clang/10-add-musl-triples.patch 👀 2021-04-26 08:48:49 root 2021-04-26 08:52:25 mate works well on Alpine, that's nice 2021-04-26 08:53:04 but there are no "mate" global package that bring required dependencies to run the desktop entirely 2021-04-26 10:21:50 anyone know a make implementation that offers only POSIX make functionality? 2021-04-26 10:22:23 Do you want to use one? 2021-04-26 10:22:48 you could try building with both gnu make and bsd make, but tbh you should just use remake. 2021-04-26 10:25:46 Thermi: I'd rather not use the GNU extensions. it seems like some mental overhead to keep track of what exactly is POSIX compliant and what's not 2021-04-26 10:25:54 Sheila: why should i just use remake? 2021-04-26 10:28:19 the better question is why you're looking specifically for strict POSIX compliance here 2021-04-26 10:29:47 does the above not make that clear? 2021-04-26 10:30:27 I see you asking if there is a make that is strictly compliant, but I don't see you saying why you need that. 2021-04-26 10:32:26 I want my software to be portable. I want it to be easy to compile on as many systems as possible. So I want to try to make sure I only use POSIX make functionality in my makefile(s). It seems the easiest way to do that is just to have a tool that exposes such functionality, rather than having to test it on a bunch of implementations. 2021-04-26 10:34:16 just use GNU make extensions, literally everyone provides GNU make by now (yes, even the BSDs). the only reason to care here is if 'runs on Solaris 1.0' or so is a desired goal. 2021-04-26 10:37:57 i.e. you would really need to reach to find an environment where it's not available at all. 2021-04-26 10:39:24 ok, thanks for the suggestion 2021-04-26 11:03:36 I am looking for a solution for autologin on tty1, is that feasible in alpine? Google has not helped me 2021-04-26 11:04:45 hsp: You need agetty for that 2021-04-26 11:05:39 once installed, you need to switch that console to agetty in /etc/inittab, and add -a 2021-04-26 11:05:51 agetty is not in repo aarch64 2021-04-26 11:06:10 It's part of util-linux 2021-04-26 11:09:54 thanks, found 2021-04-26 11:16:01 ikke, nut work :( 2021-04-26 11:16:14 tty1::/sbin/agetty --autologin holgi --noclear 38400 linux 2021-04-26 11:16:30 ikke, but work :( 2021-04-26 11:16:42 It did work for me when I tried it 2021-04-26 11:17:24 strange 2021-04-26 11:18:08 that should be okay? ==> tty1::/sbin/agetty --autologin holgi --noclear 38400 linux 2021-04-26 11:19:42 yes 2021-04-26 11:20:42 but it does not work here 2021-04-26 11:20:56 What are you trying? 2021-04-26 11:24:22 ikke, only edit the inittab 2021-04-26 11:24:28 and then? 2021-04-26 11:24:52 reboot 2021-04-26 11:26:11 try: tty1::respawn:/sbin/agetty --autologin holgi --noclear 38400 tty1 2021-04-26 11:26:55 ok 2021-04-26 11:29:14 same, not work 2021-04-26 11:30:04 I do not understand 2021-04-26 11:35:05 ikke, can you show me your entry? 2021-04-26 12:40:01 if you aren't actually using a serial connection you can just do /bin/login -f user 2021-04-26 12:43:47 no, tty1 2021-04-26 12:46:47 i thimk on aarch64 is different 2021-04-26 13:33:02 How can I mount an ext4 formatted disk on startup with a specific user and group on an diskless Alpine installation? 2021-04-26 15:54:29 <[diablo]> hey good afternoon #alpine-linux , guys anyone tell me how I can see any associated scripts of a package, for example where they add a group and user please? 2021-04-26 15:56:20 [diablo]: that things are usually done in a pre-install script 2021-04-26 15:56:38 https://gitlab.alpinelinux.org/alpine/aports/ 2021-04-26 15:57:03 look the package and check if has some pre-install script 2021-04-26 15:59:12 <[diablo]> hey donoban , no way of seeing using the apk command no? 2021-04-26 15:59:53 you could download the apk and unpack it 2021-04-26 16:00:17 I don't believe that apk has some command that exposes directly that information 2021-04-26 16:01:19 <[diablo]> ah ok , thank you donoban appreciate that, cheers 2021-04-26 16:01:32 :) 2021-04-26 17:03:29 what else export $PATH like /etc/profile ? I did not export or set path anywhere, but my user has in path some bins more times 2021-04-26 17:03:41 root does not have that issue 2021-04-26 17:04:09 /etc/profile.d/*.sh 2021-04-26 17:04:13 ~/.profile 2021-04-26 17:04:18 there is nothing 2021-04-26 17:04:26 I mean what export path 2021-04-26 17:04:42 I practically have default everything 2021-04-26 17:05:00 probably nothing wrong for usage, but I don't like that some 'bin' is there few times 2021-04-26 17:05:52 PATH='/bin:/usr/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin' 2021-04-26 17:06:05 that env report me 2021-04-26 17:06:15 what shell? 2021-04-26 17:06:20 ash 2021-04-26 17:06:27 without user profile 2021-04-26 17:06:46 did you try to grep for PATH? 2021-04-26 17:07:10 like whole system? 2021-04-26 17:08:02 /etc and $HOME for starters 2021-04-26 17:08:24 export PATH="$PATH:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin" That's what I have in /etc/profile 2021-04-26 17:08:49 nothing else with path env what I know and I checked probably all system 2021-04-26 17:08:57 Then you are expanding PATH 2021-04-26 17:09:08 so if PATH already had those values, you are now duplicating them 2021-04-26 17:09:28 This is in my /etc/profile: 2021-04-26 17:09:31 export PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin 2021-04-26 17:09:41 i did not touch path part in /etc/profile 2021-04-26 17:09:46 me neither 2021-04-26 17:09:55 Ah, that might be something new 2021-04-26 17:10:00 I thought about it, that extra $PATH there is ... 2021-04-26 17:10:44 https://gitlab.alpinelinux.org/alpine/aports/-/commit/f440e2db3f139f3e1b611e1f964acfcc93890fcf 2021-04-26 17:10:46 but if I remember well, that shit is default there 2021-04-26 17:11:35 so it's for secure these path if I set up new? 2021-04-26 17:13:48 sorry, that sentence does not make sense to me 2021-04-26 17:15:09 hM, I have the same PATH 2021-04-26 17:15:22 ahh no 2021-04-26 17:15:26 I meant, that extra '$PATH' assure that these paths are kept after anything else add another path 2021-04-26 17:15:30 PATH=/usr/local/bin:/usr/bin:/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin 2021-04-26 17:15:45 yes there are duplicates :\ 2021-04-26 17:16:02 my system can't containt duplicate, it has to be clean like virgin 2021-04-26 17:16:31 when I set up in foot germinal login shell, I had even more doubles ;) 2021-04-26 17:16:34 helby: yes, WSL sets paths entries that were overwritten by the original profile 2021-04-26 17:16:56 okay, thanks, so assume it's safe to remove that crap 2021-04-26 17:16:59 everytime profile is sourced in nested sessions 2021-04-26 17:17:06 bc. nobody use windows anymore in this world 2021-04-26 17:22:22 hope alpine will not require microsoft account bc they will change something in windows 2021-04-26 17:23:09 Microsoft account, forced telemetry, nice menu with 'helpful' suggestions, all comming soon(tm) 2021-04-26 17:23:30 Oh, and ofcourse reboot prompts and forced reboots 2021-04-26 17:23:44 "For your benefit" 2021-04-26 17:26:30 cool, hope console will have blue color too 2021-04-26 17:37:04 i am waiting for amd ryzen 5950x in laptop, so i can just install Alpine Linux on bare metal with bspwn (like this guy's setup: https://www.youtube.com/watch?v=UP3vVAxbq4c). that will make me a happy camper for next three years. 8) 2021-04-26 17:41:24 I would be okay with new keyboard for my 10 years ThinkPad ;) 2021-04-26 17:41:55 nobody selling some in my region and I pretty nice broke old one 2021-04-26 17:45:53 that guy should contact Stallman, he has looking like real Saint ... 2021-04-26 17:47:47 my ThinkPad gave up on me in March, 11 days before the three years warranty ended (ssd disaster). i spent 2400 euros on it in 2018. for me, laptops (even macs) last for 3-4 years, and then i simply change them.. 2021-04-26 17:48:31 i could have replaced the SSD, had i not opened it up myself to see what's up and broke a (seemingly important) wire -.- 2021-04-26 17:48:43 with that cpu u can probably compile browser :/ 2021-04-26 17:49:29 I bought this for 180 EUR, T420 not so long ago and it's good 2021-04-26 17:49:51 but tried to clean up keyboard too hard and broke many keys 2021-04-26 17:49:53 petrj: have it fixed 2021-04-26 17:51:32 good that still in warranty so 2021-04-26 17:51:46 but new laptops are shit 2021-04-26 17:54:02 my 3-years warranty ran out 11 days after the ssd crash. i didn't noticed it until after few days that i could have claimed the warranty. there there was a "oh, well" moment. specing: lol ok, i don't know what lenovo support will charge or how many days they will take, but i will call them tomorrow. 2021-04-26 17:56:51 petrj: dont send it to lenovo lol 2021-04-26 17:57:04 send it to a real repair shop 2021-04-26 17:57:19 you know, the ones that actually know how to fix a mobo other than replacing it whole 2021-04-26 18:01:45 lel yea good idea. thinkpad/lenovo support in finland is generally good, but sometimes they take a few days (i sent my laptop once before when it was still under warranty). but now that it is out of warranty, i guess they may overcharge compared to other repair shop. (much easier to buy a new laptop though '=D ) 2021-04-26 18:06:25 well if in warranty it's better if they replace for new one, why not 2021-04-26 18:07:06 than some boby with blowlamp fuckering with your machine :/ 2021-04-26 18:08:52 i will format windows and install alpine bare metal anyway :D i do a lot of docker containers, vagrant and general development stuff cross-platform, so SSDs on this side of the house give up fast. 2021-04-26 18:10:21 can't compare old thinkpads to your new machine, but here, u can replace everything with one screwdriver :/ 2021-04-26 18:11:41 1600x900 display does not shine, but enough for porn :/ 2021-04-26 18:13:31 why does setup-xorg-base install so much 2021-04-26 18:14:08 wouldn't you just need xorg-server? 2021-04-26 18:17:44 u need font and gpu driver, xinit ... 2021-04-26 18:17:49 input 2021-04-26 18:18:32 installed a lot bc. mesa 2021-04-26 18:18:42 helby: some oems like tuxedo allow u to replace 2021-04-26 18:19:05 that's very nice, that they allow :/ 2021-04-26 18:19:40 still if these 'light' are glued together, hard to change anything 2021-04-26 18:20:23 cool project is mmt reform laptop, replacable things too 2021-04-26 18:20:31 Wirh allow i meant u being able to replace 2021-04-26 18:21:37 yes, checking them now, looks nice 2021-04-26 18:22:52 is there a way to undo setup-xorg-base? i just wanted to test out dwm. was thinking about writing a script to uninstall all the packages once i was done with it but wasn't sure if there was something there already 2021-04-26 18:23:26 There is nothing direct to undo it 2021-04-26 18:23:54 alright i guess i'll have to write the script then 2021-04-26 18:24:08 apk del xorg-server xf86-input-libinput eudev mesa 2021-04-26 18:25:13 that was fast script ;/ 2021-04-26 18:25:36 It's not perfect, but does most of it 2021-04-26 18:25:40 alright 2021-04-26 18:25:58 it's not manjaro, it does not polute your system with 100 config files 2021-04-26 18:26:04 lol 2021-04-26 18:26:40 if u will not build xorg yourself, u will install these packages anyway if using X 2021-04-26 18:27:23 hm 2021-04-26 18:29:16 I disabled all crap in sway and got 660KiB package insted of 5000 something 2021-04-26 18:30:19 yup that command gets rid of 55 packages 2021-04-26 18:30:57 https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-xorg-base.in 2021-04-26 18:32:03 so not much happens with that command as u can see 2021-04-26 18:32:40 just ancient xorg has some dependencies :/ 2021-04-26 18:32:44 yeah 2021-04-26 18:32:51 i read the script i had in /sbin 2021-04-26 18:33:05 it also calls setup-udev which isn't in my sbin from some reason 2021-04-26 18:33:48 yes, first it install eudev 2021-04-26 18:34:16 ah is setup-udev in the udev-init-scripts package? 2021-04-26 18:35:49 y 2021-04-26 19:06:29 Folks, how does one prevents GTK's file-chooser from creating a Desktop folder? 2021-04-26 19:08:39 a cronjob running `.purge-desktop.sh` would be one "hammer" solution for that -.- 2021-04-26 19:12:30 hate that shit too 2021-04-26 19:13:39 even don't like if some crap create Download folder 2021-04-26 19:15:26 can't you just set the Desktop folder to your home folder 2021-04-26 19:15:42 on your ~/.config/user-dirs file? 2021-04-26 19:19:19 petrj: Or hammering the GTK devs. 2021-04-26 19:19:48 helby: Yeah 2021-04-26 19:20:07 eletrotupi: I don't have that file. 2021-04-26 19:20:27 create it 2021-04-26 19:20:58 man xdg-user-dirs-update 2021-04-26 19:21:36 man 5 user-dirs.dirs 2021-04-26 19:22:17 I used xdg dir but it has no impact on that 2021-04-26 19:22:35 but I had it in profile 2021-04-26 19:22:49 directed it to /var/empty 2021-04-26 19:27:31 eletrotupi: man: No entry for xdg-user-dirs-update in the manual. 2021-04-26 19:27:58 eletrotupi: man: No entry for user-dirs.dirs in section 5 of the manual. 2021-04-26 19:29:32 Which package contains those man pages? 2021-04-26 19:29:58 you should add the docs meta peckage 2021-04-26 19:30:03 *package 2021-04-26 19:30:33 There are no manpages for those, though 2021-04-26 19:30:58 "--disable-documentation # Broken, tries to load a network resource but fails" 2021-04-26 19:34:22 really? 2021-04-26 19:34:31 well, I didn't tested on my alpine install 2021-04-26 19:34:46 anyway you can easily find that man page on the net 2021-04-26 19:41:12 eletrotupi: Didn't work. 2021-04-26 19:42:17 :| 2021-04-26 19:42:56 Weird, most of the apps I run here always behave well reading that file 2021-04-26 19:49:22 they had to disable documentation, bc. it tries create ~/Desktop directory 2021-04-26 19:50:31 Wouldn't surprise me. 2021-04-26 19:51:15 I remove often also .local 2021-04-26 19:51:18 Lots of stuff writes to ~/ 2021-04-26 19:51:29 some like to create that nonsense 2021-04-26 20:04:25 not sure if 'export LANG=en.utf8' is correct in /etc/profile ... never saw that style 2021-04-26 20:05:07 and even I delete these lang and char and system keep in env for LC_COLLATE and LANG = en_US.UTF-8 2021-04-26 20:05:26 while it is okay, nothing agains Americans, but I am not US type :/ 2021-04-26 20:05:42 no idea from where it comes if nowhere to set up 2021-04-26 20:06:01 maybe from Microsoft! 2021-04-26 20:07:30 I assume it can't come from keymap 2021-04-26 20:07:51 oh, I am idiot, nvm 2021-04-26 20:07:56 lol 2021-04-26 20:08:23 but still default 'en.utf8' looks weird :/ 2021-04-26 20:08:33 probably even does not exist 2021-04-26 20:09:06 change it to C.UTF-8 and run like hell. :) 2021-04-26 20:11:14 i don't speak C ;) 2021-04-26 20:13:02 why doesn't setup-xorg-base install x11? 2021-04-26 20:13:18 i tried compiling dwm and it sead that x11/xlib.h doesn't exist 2021-04-26 20:13:21 *said 2021-04-26 20:13:54 how would i install x11? 2021-04-26 20:14:02 do i have to compile from source? 2021-04-26 20:15:57 https://pkgs.alpinelinux.org/package/edge/community/x86_64/dwm 2021-04-26 20:16:17 https://git.alpinelinux.org/aports/tree/community/dwm/APKBUILD 2021-04-26 20:16:33 i was thinking of compiling dwm from source 2021-04-26 20:16:34 It does install X11, just not development headers 2021-04-26 20:16:37 ah 2021-04-26 20:16:51 since dwm's config is mainly done from its source code 2021-04-26 20:16:54 check `cat /bin/setup-xorg-base` 2021-04-26 20:17:02 Uh sorry, in `/sbin` 2021-04-26 20:17:10 aha $PATH 2021-04-26 20:18:32 i see xorgserver and xf86 input 2021-04-26 20:18:39 kelvinf: once u try cwm u will never touch anything else in X ;) 2021-04-26 20:18:50 i hear cwm is good 2021-04-26 20:35:24 ikke, thanks for the link to the aports script 2021-04-26 20:36:18 np 2021-04-26 22:44:04 helby: What do u like most about CWM? 2021-04-26 22:44:45 it's simple without nonsense, but powerfull enough 2021-04-26 22:46:27 I use mostly window maximized and each in different group display as tilling feature in sway is mostly useless for me on 14' screen 2021-04-26 22:47:00 but sometimes I want open fast terminal for some shit, but tilling in sway u know ... 2021-04-26 22:47:44 only drawbag and it's huge drawbag, that's it's X, so even it is light, X requires trillions of 'X files' around system, what can't smell 2021-04-26 22:47:47 i, too, happen to enjoy simple and non-ceremonial things. give me a terminal and web-browser and i'm all set. 2021-04-26 22:47:57 yep 2021-04-26 22:48:42 I still don't get why so many users want all kind of tray bar or other shit what take just place in screen 2021-04-26 22:48:58 watching nonstop of clock or cpu freq :/ 2021-04-26 22:50:49 if u live with just browser and terminal, then u don't need any menu or whatever, bc. u have 2 keybindings for terminal and browser at all 2021-04-26 22:52:12 check even APKBUILD for cwm for example https://git.alpinelinux.org/aports/tree/testing/cwm/APKBUILD 2021-04-26 22:52:32 it can't be even simplier 2021-04-27 00:36:30 that really is nice 2021-04-27 00:37:01 helby: I've been experimenting with just having conky, no tray. when I want to see the time I toggle show desktop. 2021-04-27 00:37:41 I just 'date' in terminal for time :/ 2021-04-27 00:37:46 if needed 2021-04-27 00:38:13 I prefer the status line in tmux for such nefarious purposes. 2021-04-27 00:38:30 still using dwm at the moment, I have my own program to collect system status info 2021-04-27 00:38:35 I work with a lot of people in different time zones, so with conky i have it setup to display 3 or 4 2021-04-27 00:38:42 one of which is a time/date block that is formatted exactly the way I want 2021-04-27 00:38:43 😛 2021-04-27 00:39:11 cwm is very mouse centric, isn't it? using openbox on my linux partition ATM 2021-04-27 00:39:38 🤷 2021-04-27 00:39:39 never used it 2021-04-27 00:39:56 sorry that was meant for helby 2021-04-27 00:39:59 ahh 2021-04-27 00:40:51 no mouse, it's keyboard first class 2021-04-27 00:41:00 if I use mouse, I would not use that 2021-04-27 00:41:02 I used to use tiling wms, but then I realised the most tiling I ever did was two windows side by side which is easy to fake with a floating wm 2021-04-27 00:41:17 well exactly that cwm is doing 2021-04-27 00:41:17 would be nice if I had a massive screen though, I mostly do work on a laptop 2021-04-27 00:41:33 it's practically stacking, but u can split 2 windows vertical or horizontal 2021-04-27 00:41:34 helby: I'll have to check it out. I'm a big fan of openbsd stuff, like tmux and doas 2021-04-27 00:41:40 what I prefer more than only tilling in sway 2021-04-27 00:41:44 any godo guides? 2021-04-27 00:42:33 helby: I'm slowly working on my own WM, and I know it'll be tiling, but I've not decided how flexible I'm going to allow it to be 2021-04-27 00:42:53 helby: despite using dwm, I barely use most of the tiling functionality; I spend most of my time in monocle 2021-04-27 00:42:58 well depends on what u used to, I can't imagine to use default kebybind, I somewhere found some sane config, so copied keybinding from there and works well for me 2021-04-27 00:43:36 me too, as this is small screen I use one window per screen with groups/workspaces 2021-04-27 00:44:29 btw. in cwm there is already built in something like 'simplified dmenu' 2021-04-27 00:44:39 I use rofi 2021-04-27 00:44:49 it's fine 2021-04-27 00:45:03 I've not found a launcher that's perfect for me yet 2021-04-27 00:45:17 well what I can launch is just firefox or terminal, so for me these are just 2 keybindings in wm, so without a launcher 2021-04-27 00:45:25 hehe 2021-04-27 00:45:56 rofi looks interesting... you guys are giving me a lot of stuff to check out :) 2021-04-27 00:46:08 don't like gui, annoying like every fucking gtk shit create trillion folders everywhere in my home 2021-04-27 00:46:27 helby: ever heard of window swallowing? 2021-04-27 00:46:37 You should start mounting your home as tmpfs. 2021-04-27 00:46:53 naaaa, I have few configs there too :/ 2021-04-27 00:47:01 helby: https://github.com/salman-abedin/devour (I'm considering switching to this entirely instead of a launcher) 2021-04-27 00:47:04 but .cache is already there 2021-04-27 00:47:07 which… might be a touch overkill… 2021-04-27 00:47:10 🤷 ☺ 2021-04-27 00:48:49 bc. firefox like to put there a lot of crap 2021-04-27 00:49:10 I don't like much these 'fancy' crap like launcher 2021-04-27 00:49:33 even if I needed some, I would use something what can launch programm without flashing whatever at me 2021-04-27 00:49:41 like cwm is doing ... just text 2021-04-27 00:49:59 When using data disk mode 2021-04-27 00:50:07 is this still relevant? 2021-04-27 00:50:08 . [Fixme: Storing lbu configs to disk is not auto-configured after configuring the data partition, one still has to select saving configs to "none" first (the new data partition is not listed), and to manually set e.g. LBU_MEDIA=sda2 in /etc/lbu/lbu.conf and execute echo "/dev/sda2 /media/sda2 vfat rw 0 0" >> /etc/fstab afterwards, to have the 2021-04-27 00:50:09 partition (here sda2) used.] The boot device may remain to be the one initially used, and can even be immutable (read-only). 2021-04-27 00:52:32 ? 2021-04-27 00:52:58 nobody know what is relevant, wiki is from 1990 and after that, nobody touched diskless or data mode :/ 2021-04-27 00:55:33 no one uses data or diskless mode? 2021-04-27 00:55:39 find that hard to believe 2021-04-27 00:56:03 no idea, I can't imagine that somebody really use it 2021-04-27 00:57:52 I'd like to use it in diskless mode for a small static webserver 2021-04-27 00:58:03 why? 2021-04-27 00:58:11 no 1gb disk space? 2021-04-27 00:58:17 is diskless or data mode no longer supported/maintained? can i get some clarification from a dev 2021-04-27 00:58:22 well probably .5 2021-04-27 00:58:33 RAM is always going to be faster than a disk 2021-04-27 00:58:47 yes, especially during reboot 2021-04-27 00:58:59 it is supported, I more joking than be serious 2021-04-27 01:08:19 u can check what setup-alpine with other scripts like setup-disk are doing 2021-04-27 01:08:38 https://gitlab.alpinelinux.org/alpine/alpine-conf 2021-04-27 01:09:09 ok thank you 2021-04-27 01:09:13 that's helpful 2021-04-27 01:09:17 how do i use lbu commit 2021-04-27 01:09:25 just lbu commit and it will automatically save /etc 2021-04-27 01:09:26 ? 2021-04-27 01:09:34 or do i need to specify the device 2021-04-27 01:10:25 not sure about your data mode, never used it 2021-04-27 01:10:50 normally on disk I set up lbu in /etc/lbu/lbu.conf 2021-04-27 01:11:06 where is place for disk 2021-04-27 01:12:25 as I use it for some kind of backup of my configs 2021-04-27 01:13:13 check lbu it's simple but cool, u can add any directory or file as include/exclude 2021-04-27 01:18:09 very confused 2021-04-27 01:18:15 do i need to mount the disk in /media 2021-04-27 01:18:20 do use it for lbu? 2021-04-27 01:23:16 no idea, try it :) as mentioned, can also be set to 'floppy' :d 2021-04-27 01:24:13 I assume 'LBU_MEDIA should point to anything u mount in 2021-04-27 01:24:32 usb is probably used there, as it's by default used in fstab by mdev or whatever 2021-04-27 01:25:42 let it commented and lbu commit, it will probably ask for media, then u will figure out what to type there 2021-04-27 02:02:55 Tempora10: as someone who runs a website off of alpine... it never occured to me to use diskless mode 2021-04-27 02:03:34 though the idea is interesting 2021-04-27 04:32:48 anyone utilizing Alpine Linux as a webserver and running it in RAM? 2021-04-27 04:53:14 what's the question 2021-04-27 07:58:40 hi, anyone succesfully set up exim on alpine-docker ? i'm a bit confused here, 2021-04-27 09:48:03 kodi 2021-04-27 10:54:32 Is there any web feed for the news section? 2021-04-27 10:54:53 There is an rss icon 2021-04-27 10:55:00 next to Alpine News 2021-04-27 10:55:35 Not for me. 2021-04-27 10:55:51 I even searched the source code. 2021-04-27 10:56:04 https://alpinelinux.org/atom.xml 2021-04-27 10:56:24 Thanks! 2021-04-27 10:56:37 It's on the home page 2021-04-27 10:56:46 Not on https://alpinelinux.org/posts/ 2021-04-27 10:57:45 I noticed that. :) 2021-04-27 14:29:50 Hi, I'm trying to make a bootable USB from the x86 ISO by dd'ing it to the device, but the final USB is not recognized by the BIOS of the target machine 2021-04-27 14:30:15 the options I'm passing to dd are as described in the Wiki pages 2021-04-27 14:30:49 should I do manual partitioning, Raspberry Pi-style, with the ISO instead? 2021-04-27 14:31:24 I've tried with two different drives but the result is the same 2021-04-27 14:31:34 Normally dd-ing the iso should work 2021-04-27 14:32:31 that's what I usually do, too. The drive gets readable in a file manager, but not when I put it to boot 2021-04-27 14:33:19 I was wondering if there were any flags or formatting I had to set to it 2021-04-27 14:34:26 how old is this machine 2021-04-27 14:35:05 ca. 2013 or 2014 I think? I've USB-booted in it before 2021-04-27 14:36:39 bios or efi? 2021-04-27 14:37:28 EFI 2021-04-27 14:37:52 but looking at hte boot menu, there's a hint I can change the mode to BIOS 2021-04-27 14:38:12 (actually, should I try that?) 2021-04-27 14:39:32 did you turn off secure boot 2021-04-27 14:41:02 deeeeeeeeeeerp nevermind, switching to legacy BIOS booting did the trick 2021-04-27 14:41:22 sawry, didn't think of trying that 2021-04-27 14:41:42 but yeah, I didn't have Secure boot on eihter 2021-04-27 14:41:51 thanks Hello71! 2021-04-27 14:42:10 efi is usually better 2021-04-27 14:42:45 That's what I thought too, but for some weird reason it insisted not to boot from it 2021-04-27 14:43:21 eh, at least it worked, I'll switch it back once I'm finished with the installer 2021-04-27 16:23:48 mps ncopa ikke there's a "Replace" in alpine pacjkagin as on debian when a package are renamed? 2021-04-27 16:23:59 stop highlightin us 2021-04-27 17:11:58 Is usb gadget mode for the pi zero w support on alpine linux? 2021-04-27 17:12:25 it's upstream 2021-04-27 17:13:57 So I can just add the text to the files like on raspberry pi os and it would work? or is there usb gadget mode by default 2021-04-27 17:50:12 what's the problem ikke ? i am working not joking 2021-04-27 17:50:59 PICCORO: I'm not joking either. It's not polite to constantly highlight people to get their attention. You can just ask your question in the channels and if anyone can answer it, they will 2021-04-27 18:04:44 PICCORO: https://wiki.alpinelinux.org/wiki/APKBUILD_Reference#replaces 2021-04-27 20:33:43 I am working with an Alpine image to get something compiled which requires lz4.so, and I installed both lz4 and lz4-dev. This only installs the lz4 binary and not the lz4.so object/library, which I need for compilation to continue. Anyone know how to get that? I assumed the -dev version would include libs. 2021-04-27 20:34:33 yes, it should 2021-04-27 20:35:16 but there is no lz4.so 2021-04-27 20:35:28 ok turns out there is ANOTHER package called... lz4-libs 2021-04-27 20:35:34 right 2021-04-27 20:35:49 I'll try that 2021-04-27 20:36:01 but that provides liblz4.so* 2021-04-27 20:36:26 oh noes 2021-04-27 20:36:27 which is provided by lz4-dev 2021-04-27 20:36:46 https://pkgs.alpinelinux.org/contents?file=liblz4.so&path=&name=&branch=edge 2021-04-27 23:26:23 hello, seems like direnv is gone from edge/testing, can i get pointed at “why?” (apologies if i’m mistaken, i don’t know much about alpine) 2021-04-27 23:34:23 mamercad: Seems to be in community now (community/direnv) See commit c0649817ae984b0c77f70b72fdcdc08737f7ef93. 2021-04-27 23:35:39 Data disk mode is not working with lbu 2021-04-27 23:36:12  [Fixme: Storing lbu configs to disk is not auto-configured after configuring the data partition, one still has to select saving configs to "none" first (the new data partition is not listed), and to manually set e.g. LBU_MEDIA=sda2 in /etc/lbu/lbu.conf and execute echo "/dev/sda2 /media/sda2 vfat rw 0 0" >> /etc/fstab afterwards, to have the 2021-04-27 23:36:12 partition (here sda2) used.] The boot device may remain to be the one initially used, and can even be immutable (read-onl 2021-04-27 23:36:22 using that for guidance 2021-04-27 23:36:56 anyone have experience with this? 2021-04-27 23:42:09 tomleb: i'm running 3.13.5 (docker latest) and i have "community" in /etc/apk/repositories and i can't `apk add direnv` ... any idea what i'm doing wrong? 2021-04-27 23:44:29 tomleb: i figured it out, i needed edge/community, thank you 2021-04-28 00:44:43 i also figured my problem out 2021-04-28 00:45:13 went ahead with diskless (running everything in RAM) instead of data mode (hybrid) 2021-04-28 00:45:35 there were some intricacies running it in a Guest VM 2021-04-28 00:45:48 getting LBU to work specifically 2021-04-28 00:50:28 was the wiki written by someone who English is a second language? 2021-04-28 00:50:36 it's very hard to understand 2021-04-28 01:06:59 Want chroot on a disk but all i got is ntfs partition. Is it possible to do something like a virtual partition from a file sitting on ntfs? 2021-04-28 01:12:22 yes but performance will be total shit 2021-04-28 01:12:52 much better to resize partition 2021-04-28 01:13:06 Payupsir: some of it 2021-04-28 01:13:17 it's a wiki, that's how it works 2021-04-28 01:14:13 slow is ok i'm still learning 2021-04-28 01:15:26 is vhd the way to go for me? 2021-04-28 04:33:57 mamercad: Now that it's in community, it will be part of the next release, so in 3.14, you no longer need to pull in edge 2021-04-28 07:30:06 having so much fun with alpine on our IOT stuff at work 2021-04-28 07:30:26 busybox's fbsplash is nice to hide boot process 2021-04-28 07:36:05 looks like alpine's init script support fbsplash but not heavily documented right now 2021-04-28 10:27:31 donoban: telegram uses openal-soft 2021-04-28 10:27:45 if that isn't compiled with pa, then it uses only alsa 2021-04-28 10:28:09 oh thanks 2021-04-28 10:28:48 It seems that bubblejail doesn't have an option predefined for alsa access 2021-04-28 11:54:45 I wonder what raspberry (raspbian) folks do to make their desktop/variant much faster than a raw aarch64 alpine 2021-04-28 11:55:02 like even moving my mouse on my alpine pi 4 is slow as hell 2021-04-28 11:57:20 If you use the raspberry pi alpine images it has the raspberry pi kernel made for the pi 4 2021-04-28 11:57:35 markand: maybe they have different kernel timer freq? 2021-04-28 11:59:17 first thing I see is that it's using fullhd rather than 4k 2021-04-28 12:02:07 The raspberry pi alpine images are made to work on the pi with the tweaks while the aarch64 alpine image is a generic aarch64 alpine 2021-04-28 12:09:09 Hello, 2021-04-28 12:09:21 Is it recommended to install alpine linux on microsd card? 2021-04-28 12:09:30 I want to install on a Raspberry PI 3 Model B + 2021-04-28 12:12:54 I want to install sys mode 2021-04-28 12:15:27 adria_, you can do it through various ways 2021-04-28 12:15:58 for sys mode I use the chroot method + minirootfs (aarch64) using qemu-user-static so you can chroot from your x86_64 machine 2021-04-28 12:16:09 then I install alpine-base linux-rpi4 and all other stuff I need 2021-04-28 12:16:35 there's a procedure to use the installer but I consider the minirootfs much quicker 2021-04-28 12:17:01 + raspberrypi-bootloader, otherwise there is no start.elf files 2021-04-28 12:18:13 I'm new to linux 2021-04-28 12:24:38 if you're new to Linux I think you should first take your habits using a more general distro like Raspberry Pi OS images first 2021-04-28 12:47:13 I did a normal install by flashing it to a microsd card and running setup-alpine with my pi zero w plugged into a monitor. The only install that would work is the diskless mode. I do know there is a way to install it sys but with setup-alpine it never worked. 2021-04-28 13:16:46 ERROR: unable to select packages: 2021-04-28 13:16:55 i keep getting this error when i apk something 2021-04-28 13:19:14 looks like apk is broken 2021-04-28 14:02:38 metbsd, can you show the full logs and the command you're typing ? 2021-04-28 14:21:34 i remember hearing that the default kernel is nerfed to make hosting an wifi AP more difficult 2021-04-28 14:21:58 anyone know about this issue, and does it apply to alpine? 2021-04-28 14:22:44 i think openwrt might be the only distro that doesn't have that problem 2021-04-28 14:43:52 How to make the system suspend and hibernate? 2021-04-28 14:48:12 from what? 2021-04-28 14:49:14 The lid close event and also from a command, like the one. 2021-04-28 15:08:16 alpine is standard linux, so use standard linux commands 2021-04-28 15:08:20 echo mem > /sys/power/state 2021-04-28 16:11:32 Nice, the command worked. 2021-04-28 16:11:50 I'll tweak around and see if I can adapt it to detect the lid closure. 2021-04-28 16:11:54 Thanks 2021-04-28 16:13:47 acpi 2021-04-28 16:21:47 i've just installed and ran `apk add docs`. I have a man(1) manpage but none for ps(1) ed(1) or any other common commands. is that normal? does the docs package not pull in _every_ manpage? 2021-04-28 16:22:15 arwn: it pulls in the -doc subpackage for commands that provide them 2021-04-28 16:22:28 ps by default is provided by busybox, which does not have man pages 2021-04-28 16:23:08 same with ed 2021-04-28 16:23:09 OOoh so the vi/ps/ed/grep/ls/cat or whatever is in coreutils aren't gonna have manpages 2021-04-28 16:23:39 If you install coreutils, you will get coreutils-doc 2021-04-28 16:24:17 But you you use the busybox variants, you should use the -h / --help output of each command 2021-04-28 16:24:21 and that's the gnu coreutils though? 2021-04-28 16:24:40 yes 2021-04-28 16:24:53 gocha. i dont need that. thanks for the information 2021-04-28 16:24:58 -h is fine 2021-04-28 16:45:33 ikke: Yup, acpi did it ^^ 2021-04-28 20:04:41 anyone has an idea why resolvconf -u keeps generating an empty /etc/resolv.conf file? although dhcpcd acquired lease for both eth0 and wlan0? 2021-04-28 20:04:50 and obviously pinging 8.8.8.8 works fine 2021-04-28 20:06:41 Is dhcpcd not hooked properly in resolvconf for some reason? 2021-04-28 20:08:25 good question, but openresolv is correctly installed 2021-04-28 20:12:05 check if the dhcpcd hooks exist. 2021-04-28 20:13:39 yes it is in usr/lib/dhcpcd/dhcpcd-hooks/20-resolv.conf 2021-04-28 20:13:46 resolvconf -l lists nothing though 2021-04-28 20:15:46 Is /etc/resolv.conf a file or symlink? 2021-04-28 20:15:51 simple file 2021-04-28 20:19:54 what's resolvconf? 2021-04-28 20:20:12 scripts that dhcpcd usually use to generate /etc/resolv.conf 2021-04-28 20:20:49 u use some package for dhcp? 2021-04-28 20:21:12 dhcpcd + wpa_supplicant 2021-04-28 20:22:44 resolvconf can collect dns information from multiple sources and then combine it into resolv.conf 2021-04-28 20:25:06 evidently it can ... but not the case :/ 2021-04-28 20:31:17 well, uninstalling openresolv and dhcpcd generate its own /etc/resolv.conf but still no nameserver in it 2021-04-28 20:32:09 let see if the bug doesn't come from my ISP box 2021-04-28 20:32:24 Do you even receive any DNS server configuration? 2021-04-28 20:43:25 I don't think so: https://paste.centos.org/view/6cc1d7fc 2021-04-28 21:03:40 run tcpdump, dump packets and disect them with wireshark 2021-04-28 21:03:45 then you can see what is exactly transmitted 2021-04-28 21:33:34 My internet worked before I went to lunch. I'm in hyper-v with a bridge adapter. I can ssh into my vm as well as ping 1.1.1.1/8.8.8.8 but I can not ping example.com. I did a restart of the networking service but what else can I try. I didn't change any config files yet other than `etc/abuild.conf`. I am using dhcp. 2021-04-28 21:35:35 you know maybe it's just cause my internet is piss.. 2021-04-28 21:36:08 it works now so whatever :) 2021-04-28 21:36:26 problem solved :/ 2021-04-28 22:54:12 Hello Everyone.. Where can I put these options? grub.cfg doesn't seem configured to accept these GRUB_DEFAULT=0 2021-04-28 22:54:12 GRUB_TIMEOUT_STYLE=hidden 2021-04-28 22:56:36 in other distros it's in the file /etc/default/grub 2021-04-28 23:02:37 alpine use the same /etc/default/grub 2021-04-28 23:42:22 helby it appears that grub wasn't setup, it's extlinux, that's why the file is not there. Thank you 2021-04-28 23:42:51 good for u, the last command with grub I remember was 'apk del grub' ;) 2021-04-28 23:43:04 haha 2021-04-28 23:43:16 how it comes that u have extlinux? 2021-04-28 23:43:40 not uefi? 2021-04-28 23:44:34 it's the default bootloader (syslinux) and the conf file is in /boot/extlinux.conf 2021-04-28 23:44:51 https://wiki.alpinelinux.org/wiki/Bootloaders 2021-04-28 23:45:38 Currently, I can boot with both UEFI or Legacy 2021-04-28 23:47:08 However I needed GRUB because I can't find how to execute kernel command for 1GB hugepage in syslinux only in grub with the option GRUB_CMDLINE_LINUX_DEFAULT="default_hugepagesz=2M hugepagesz=1G hugepages=3" 2021-04-28 23:47:55 probably default if not uefi, I never get anything else than grub 2021-04-28 23:49:51 I haven't setup alpine before as full install. I use Diskless configurations.. not sure if this will differ as bootloader 2021-04-28 23:50:04 but by default I can boot both UEFI and Legacy 2021-04-28 23:53:26 just a note that Rufus used to make the USB bootable. 2021-04-29 00:04:06 Hevan what are you using diskless for? curious cause most people do a traditional install afaik 2021-04-29 00:04:19 Haven't found a use case for it yet myself 2021-04-29 00:06:03 I'm using alpine as Docker Host with very few services in old hardware.. storage isn't needed, like OpenVPN server, devleopment ... etc and I access via SSH 2021-04-29 00:06:13 there is no disk installed, all loading from USB to RAM 2021-04-29 00:09:49 cool beans 2021-04-29 00:09:58 so it's already configured with my settings and it's loaded once.. then USB can be removed after boot up if wanted.. if there are changes you can commit again and save changes. 2021-04-29 00:10:12 Yep! 2021-04-29 00:10:24 although to save changes the usb would need to be present? 2021-04-29 00:11:15 that's cool, how old is the hardware lol? pentium era? 2021-04-29 00:13:44 Yeah.. must be present or you can save the changes somewhere and upload via network.. but at the end put them back in the USB.. all the changes will be saved as compressed tar file 2021-04-29 00:13:48 https://wiki.alpinelinux.org/wiki/Alpine_local_backup 2021-04-29 00:14:06 Actually not that old.. but low powered 2021-04-29 00:14:48 Atom-x5-z8350 4 Cores 2021-04-29 00:15:00 it's less than 25W power consumption I believe 2021-04-29 00:15:07 as whole sys 2021-04-29 00:20:42 is it some kind of low end laptop or something else? 2021-04-29 00:22:52 mini PC 2021-04-29 00:25:19 how is it usable for these tasks? not slugish? thinking about something and cheap of that kind 2021-04-29 00:31:55 as GUI system it's sluggish and slow, just for basic tasks of browsing/connecting to TV, add like 5 seconds for each thing compared of using normal PC and Oh! sorry there is disk installed but because I don't really consider it there... it's that kind of bad soldered eMMC. However it's usable. 2021-04-29 00:32:19 but using Diskless to make the whole system works in RAM. makes it better for my use case 2021-04-29 00:34:09 This CPU has AES.. it's good for OpenVPN Server. 2021-04-29 00:35:51 You can see reviews or compare by benchmark score compared to the low end CPUs you tried before to get better idea of how it performs. 2021-04-29 00:35:57 https://ark.intel.com/content/www/us/en/ark/products/93361/intel-atom-x5-z8350-processor-2m-cache-up-to-1-92-ghz.html 2021-04-29 00:37:06 yes checked it before little bit 2021-04-29 02:22:09 Is the backend of git.alpinelinux.org down? 2021-04-29 02:58:29 looks like 2021-04-29 03:02:29 cmiiw, but alpine's package installation process is "the fastest" compared to any linux distribution or any other operating system? 2021-04-29 03:02:34  i have used at least 30 different operating systems in past 25-26 years, i have not seen any package installation faster than this. 2021-04-29 03:02:53 and i'm not saying coz i'm on #alpine-linux channel :) 2021-04-29 03:04:08 my question/curiosity being, why doesn't it pisses off other package management guys, and why haven't they improved the performance since now it is evident that it can be made better? 2021-04-29 03:07:42 don't tell me about it, used openbsd before, that's probably the slowest one 2021-04-29 03:08:33 the time I upgraded firefox on openbsd I can probably install whole system in alpine :/ 2021-04-29 03:10:11 I would also add, that boot is probably also the fastest one, just waiting for lease from udhcp slows down it little bit 2021-04-29 03:10:11 freebsd, netbsd, windows, openbsd, fedora, debian, ubuntu, gentoo, void, arch, solaris, illumos ... i have tried many, this 'installation time' is my pet peeve, and it NEVER improves over years and years, decade. 2021-04-29 03:11:32 alpine linux has sliced the bullsh!t in half and solve the problem with a big FU for all the rest. 2021-04-29 03:13:17 many OS use a lot of shit from ancient time for whatever reason 2021-04-29 03:13:37 a lot of nonsense 2021-04-29 03:15:55 exactly, i mean even torvalds was pissed in debcon few years back. why haven't anyone bothered to improve anyway. i feel that there is a contrived bullsh!t notion 'if it is fast, it must be insecure' 2021-04-29 03:16:41 well, simple and fast is not good as business model for those corporations ... so 2021-04-29 03:17:00 so they have to fuck up things 2021-04-29 03:17:10 to provide support 2021-04-29 03:17:46 but there are many stupid examples, like freebsd, u install that shit and u get 3 firewalls packages, not one is ON by default 2021-04-29 03:18:03 yup, they do find security vulnerabilities in those systems from time to time as well. keeping it simple is actually also making it less risky. 2021-04-29 03:18:11 but freebsd has many other nonsense there so 2021-04-29 03:19:19 but openbsd inluce a lot of useless ancient nonsense too even they like to strip out code, they install bullshit 2021-04-29 03:20:08 e.g. try calculating timing of 'apt install g++' on ubuntu. 'yum install g++' on centos/fedora, 'pkg install g++' on freebsd, 'pkgin add g++' on netbsd, 'brew install g++' on macOS. then compare with our gold standard 'apk add g++' 2021-04-29 03:21:56 usually u get package already installed/upgrade before other OS just check version :/ 2021-04-29 03:22:42 on other hand alpine wiki is worse than bad 2021-04-29 03:24:27 yup they calculate like million hashes, check the health of 'other' packages (which you are NOT installing) and update indexes for EVERY freakin install command execution. 2021-04-29 03:24:37 haha, yes 2021-04-29 03:24:58 some maybe check atomic time 2021-04-29 03:25:33 i really wish it improves, someone need to make positive criticism to impact the change 'cut the bullsh!t style package manager' 2021-04-29 03:25:48 don't use these systems 2021-04-29 03:26:28 right, but some of us do work on different systems in our daily jobs. 2021-04-29 03:26:55 quit the job, be free ;) 2021-04-29 03:27:26 so we don't have a choice but suffer from these atrocities of 'slow package managers' under the guise of security :/ 2021-04-29 03:27:46 can't imagine to work at some shitty company to use all that kubernetes and whatever other nonsense workflows 2021-04-29 03:30:42 well, if you are hiring to work on all-alpine place, i'll quit. but otherwise i gotta pay bills brov! -.- 2021-04-29 03:32:05 but u can at least rest little bit or go to smoke break during your work 2021-04-29 03:32:21 if boss ask something u just answer, 'I am installing one package' :D 2021-04-29 03:35:46 lol we have a separated docker laying in our CI with pre-installed packages, to improve the timings. 2021-04-29 03:36:03 layer* 2021-04-29 07:28:13 hi, can you please assist in installing exim in alpine ? 2021-04-29 07:32:52 what's the problem? 2021-04-29 07:33:32 i'm trying to install exim4 smarthost into a docker image, but i can't find much articles about it 2021-04-29 07:42:48 what's similar command for dpkg-reconfigure in apine ? 2021-04-29 07:54:38 afidegnum: I'm not sure if apk has some dpkg-reconfigure equivalent 2021-04-29 07:54:50 right, you can probably just do 'apk add exim' and follow generic guides online for configuring 2021-04-29 07:55:10 there is no equivalent to dpkg-reconfigure since apk doesn't do configuration 2021-04-29 07:55:11 there is a command for check new conf files 2021-04-29 07:55:22 it just installs default configs and not much more 2021-04-29 07:55:23 do you know what command is danieli ? 2021-04-29 07:56:12 I remember something along the lines of update-exim4.conf but I can't remember if that's third party or not 2021-04-29 08:01:30 ok 2021-04-29 11:43:14 guys, is k3s on alpine stable for production usage? 2021-04-29 12:27:50 hello everyone. I need some help with alpine audio. I recently installed alpine into my laptop, but it does not recognizes audio. I followed the wiki: wiki.alpinelinux.org/wiki/ALSA  but still not working 2021-04-29 12:28:46 I also installed pulseaudio 2021-04-29 12:29:21 When I type "alsamixer" to show mixer, it says there in the Card AND Chip "PulseAudio" 2021-04-29 12:35:00 I logged out, and login, now shows the card "HDA ATI SB" with chip Conexant CX20588" but I have no idea how to test 2021-04-29 12:40:25 you don't need pulseaudio for audio, pulseaudio is a layer above alsa. so if alsa works, pulse should also work. if alsa doesn't work pulse will not. 2021-04-29 12:40:36 get alsa working and the rest comes automatically 2021-04-29 12:40:47 does your audio work under other linux distros? 2021-04-29 12:42:05 yes, it works 2021-04-29 12:42:33 alpine is the first distro I test that audio doesn't work for some reason :/ 2021-04-29 15:03:44 Anybody running Qutebrowser with a working webview? 2021-04-29 15:03:54 I've got this error: 11:55:54 ERROR webview tabbedbrowser:show_error_page:956 Renderer process crashed 2021-04-29 18:32:58 hi, can I please get some info on how the CVE database vulnerabilities are updated for Alpine ? 2021-04-29 18:34:10 Devs manually add secfixes comments to APKBUILDs, secfixes.alpinelinux.org collects them 2021-04-29 18:39:13 ack. What are the chances of missing any vulnerabilities ? is there a page where I can see any existing vulnerabilities on alpine packages? I understand you build from the latest. 2021-04-29 18:40:06 madhurranjan: We recently launched https://security.alpinelinux.org/ 2021-04-29 18:40:14 That sources from the nvd CVE feed 2021-04-29 18:42:13 We still need to tweak things so that the CVE's are correctly mapped to the packages we have, but it should already cover most 2021-04-29 18:42:44 excellent! 2021-04-29 18:45:40 I think if these are tweaked , the criticism against alpine handling vulnerabilities will reduce and will allow enterprises to more freely adopt this at scale. Looks like debian/Redhat with their vuln setup are some key things that push alpine supporters within these orgs out of some conversations 2021-04-29 18:46:11 thanks for the awesome work. 2021-04-29 18:47:35 another point - redhat/debian has a lot of vulnerabilities open against glibc and there is 0 for musl 2021-04-29 18:47:56 is there any thought on how to resolve these or where I can find the vuln against musl ? 2021-04-29 18:49:56 https://security.alpinelinux.org/srcpkg/musl 2021-04-29 18:50:50 madhurranjan: what do you mean with resolving these? 2021-04-29 18:52:58 ikke Thanks for pointing it out. I was blind! 2021-04-29 18:56:16 I feel safe now 2021-04-29 18:56:41 ... until madhurranjan brings some enterprises from debian/redhat space 2021-04-29 18:58:13 haha I deserved that:) 2021-04-29 19:06:52 Just to close out this, the project https://github.com/aquasecurity/trivy does scanning as well and in their github page says "High accuracy Especially Alpine Linux " 2021-04-29 19:07:27 heh 2021-04-29 19:07:35 please read the recent mailing list threat on alpine devel 2021-04-29 19:08:10 https://lists.alpinelinux.org/~alpine/devel/%3CAM6PR03MB471180AD19195D25E1BC462AB3409%40AM6PR03MB4711.eurprd03.prod.outlook.com%3E 2021-04-29 19:08:21 sigh 2021-04-29 19:10:51 it's always pleasure to read these 10 meters long lines in mailing list 2021-04-29 19:11:48 yup 2021-04-29 19:12:44 at least our bot send links well 2021-04-29 19:16:32 ack 2021-04-29 19:16:33 having wrapping would be nice indeed 2021-04-29 19:16:54 my mail client wraps it 2021-04-29 19:34:51 windows email clients don't count 2021-04-29 19:34:59 What is windows? 2021-04-29 19:39:13 check that little help button in your menu :/ 2021-04-29 22:17:11 tried to add wireguard in interfaces such as: wiki/Configure_a_Wireguard_interface_(wg) says 2021-04-29 22:17:22 but instead of eth0, I used wlan0. Didn't seem to work in 3.13. 2021-04-29 22:17:48 so instead, just used wg-quick in a script. that worked. 2021-04-29 22:18:23 (requires wlan0) I mean 2021-04-29 22:48:05 Anybody know the best way to downgrade from edge to non-edge? 2021-04-29 22:48:51 I guess removing the edge repos and then running `apk upgrade -a` 2021-04-29 22:50:59 This will remove version pins if you have any though 2021-04-29 22:51:14 If you do local development it's good to keep in mind 2021-04-30 00:39:06 helby: i shared this link with aptitude team: https://ariadne.space/2021/04/25/why-apk-tools-is-different-than-other-package-managers/. they do ack the perf issues and have the bottlenecks identified. 2021-04-30 00:39:28 it's a non-trivial change for them, but still in the realm of possibilities to accomplish. lets see how the future looks like for other package managers. 2021-04-30 00:39:48 what does this have to do with alpine? 2021-04-30 00:40:21 we were discussing the general pain yesterday. 2021-04-30 00:40:38 my comment is continuation of that discussion. 2021-04-30 00:41:08 oh ok 2021-04-30 00:41:11 [13:06:51] Just to close out this, the project https://github.com/aquasecurity/trivy does scanning as well and in their github page says "High accuracy Especially Alpine Linux " 2021-04-30 00:41:33 these security companies (which largely are all shit) 2021-04-30 00:41:42 repackage https://secdb.alpinelinux.org 2021-04-30 00:42:02 and charge anywhere from $0 to $4999 monthly for their services per server 2021-04-30 00:42:26 :) 2021-04-30 00:42:33 i am not a fan of those companies and their products 2021-04-30 00:43:20 too much koolaid of "safety" and "security" has been spun in past two decades. very few people questioned the merits ig. :D 2021-04-30 00:43:42 for example trivy is just a ripoff of some other scanner 2021-04-30 00:43:56 how do you write a scanner with all those features in only 480 commits ? 2021-04-30 00:43:58 give me a break 2021-04-30 00:44:55 oh i see 2021-04-30 00:44:59 everything is squashed 2021-04-30 00:45:01 well still 2021-04-30 00:45:13 all these scanners i look at is cut and paste code from other scanners 2021-04-30 00:45:22 petrj: will check that soon 2021-04-30 00:45:34 i plan to write about that particular flavor of FOSS grift soon 2021-04-30 00:45:45 especially aquasec 2021-04-30 00:45:49 they pissed me off already :) 2021-04-30 00:45:56 good :) 2021-04-30 00:46:17 good read, actually someone shared it, i just advocated value of keeping things simple and boring for few hours. :) 2021-04-30 00:46:37 fwiw the aptitude author is a cool dude 2021-04-30 00:48:01 yup, they were very receptive. i was a bit nervous whether it will turn into a matter of pride and they being defensive.. but they were quite open 2021-04-30 00:48:06 How would I get usb gadget mode working on my pi zero w? 2021-04-30 00:48:32 the same steps as for the raspberry pi os did not work 2021-04-30 00:48:33 they weren't the ones i was trying to help out with that article though 2021-04-30 00:48:42 but instead ikey doherty's new package management effort 2021-04-30 00:50:36 ah i didn't realized that it's your article. sorry. but excellent post. 8) 2021-04-30 00:53:40 q66: you have a raptor ppc64 board, right? 2021-04-30 00:55:55 two 2021-04-30 00:56:02 sup? 2021-04-30 00:59:55 so, i have a raptor board at my colo, which i have installed 2x radeon 5700 series cards in for transcoding with OBS 2021-04-30 01:00:04 but OBS keeps crashing :) 2021-04-30 01:00:17 alpine kernels are built with CONFIG_PPC_64K_PAGES=y 2021-04-30 01:00:26 i'm surprised it runs at all 2021-04-30 01:00:37 amdgpu has been problematic with 64k pages 2021-04-30 01:00:43 well i am not running X on the card 2021-04-30 01:00:48 i am running it in Xvnc 2021-04-30 01:00:54 iirc the kernel module didn't even come up 2021-04-30 01:01:09 hmm 2021-04-30 01:01:11 but that was on polaris, so maybe navi was spared of it 2021-04-30 01:01:23 so what i am hearing here is 2021-04-30 01:01:35 i should try to CONFIG_PPC_4K_PAGES=y instead 2021-04-30 01:01:43 sure, feel free to try 2021-04-30 01:01:47 we use 4k pages @ void 2021-04-30 01:01:53 i think everyone should, really 2021-04-30 01:02:27 IBM want 64K page kernel for some reason 2021-04-30 01:02:48 sure, because they think power is for single purpose servers that run their one thing and don't care about anything else 2021-04-30 01:03:14 i'll make a 4k kernel pagesize variant 2021-04-30 01:03:32 we already have linux-virt 2021-04-30 01:03:33 for any other use, 4k pages will perform much better, and even in cases where 64k pages per form better, there is still no benefit to them because you can just hugepages + madvise 2021-04-30 01:03:40 *perform 2021-04-30 01:03:59 so i can have linux-4kpages 2021-04-30 01:04:01 plus 4k pages have tons of other benefits like actually making guard pages possible 2021-04-30 01:04:01 or somethng 2021-04-30 01:04:03 on ppc64le 2021-04-30 01:04:18 lots of software disables guard pages with large pages beacuse they become wasteful 2021-04-30 01:04:37 though maybe not in an IBM server with 256+G RAM :/ 2021-04-30 01:04:52 q66: have you had success with amdgpu on ppc64le with 4k pages? 2021-04-30 01:05:04 well i run it as my primary workstation, so make a guess :) 2021-04-30 01:05:22 i guess i'll make a 4K page variant kernel 2021-04-30 01:05:27 and then try that 2021-04-30 01:05:33 it's worth a try 2021-04-30 01:05:36 seems cheaper than paying remote hands fees 2021-04-30 01:05:38 :P 2021-04-30 01:05:38 (i have the exact same card btw, 5700xt) 2021-04-30 01:06:07 i had to pay some outrageous price for a pair of them 2021-04-30 01:06:12 thanks ethereum miners 2021-04-30 01:06:16 yeah nowadays that's just going to be the case :/ 2021-04-30 01:06:18 i bought mine ages ago 2021-04-30 01:06:28 for like... €350 2021-04-30 01:06:34 i just want to be able to send alpineconf stream at 2160p60 2021-04-30 01:06:39 some folks have received BeagleV board (riscv64gc compatible ISA) yesterday. would it help if aports team get one (or three) of them? :) 2021-04-30 01:06:40 that card could do it 2021-04-30 01:06:45 so i got two of em 2021-04-30 01:06:47 for... 1700$ 2021-04-30 01:06:49 :| 2021-04-30 01:06:51 i am not santa but one i can donate :D 2021-04-30 01:06:51 oof 2021-04-30 01:06:59 that's a lot of money 2021-04-30 01:07:18 (its good to have a spare) 2021-04-30 01:07:33 i wish i had some other card around these days since navi on aarch64 is very broken and i want a functional card for my honeycomb 2021-04-30 01:07:37 if you dont have a spare while producing an event, 2021-04-30 01:07:39 i currently have a 5500xt in there... 2021-04-30 01:07:47 what tends to happen is that the card dies in the middle of the event 2021-04-30 01:07:52 but the gpu prices right now really make me not want to do that 2021-04-30 01:07:54 or at least that's what i have seen :p 2021-04-30 01:08:08 but then, maybe it'll give me an incentive to actually fix the damn thing 2021-04-30 01:08:44 I have an RX 560. (also a 1070 Ti, but I want to keep the AMD and sell the nVidia, frankly) 2021-04-30 01:08:49 yeah, i was originally going to use one of my honeycomb boards (i have 3 colo'd for various things) for this but then you said navi is fucked 2021-04-30 01:08:52 on HC 2021-04-30 01:08:56 i think that was you 2021-04-30 01:08:57 yeah 2021-04-30 01:08:58 it was 2021-04-30 01:08:59 one of the void devs did 2021-04-30 01:09:09 is the rx560 any better on hc? 2021-04-30 01:09:11 i originally contributed the kernel support for aarch64/amdgpu 2021-04-30 01:09:28 the thing is, the way the code is written, it becomes pretty unreliable on anything but x86 and power 2021-04-30 01:09:36 Sheila: that's what i havei n my desktop, its very passable 2021-04-30 01:09:40 so it was later reverted since it occasionally hung and stuff 2021-04-30 01:09:55 and to actually restore the support properly, the whole DC thing will need a refactor 2021-04-30 01:10:11 can somebody just make a gpu that does not suck ? 2021-04-30 01:10:19 i dont even care about the gpu in this case 2021-04-30 01:10:25 i just need the encoder ASIC 2021-04-30 01:10:27 :D 2021-04-30 01:10:28 i'm looking forward to testing the new intel ones as soon as i can get my hands on them 2021-04-30 01:10:42 risc-v people are working on an open GPU-adjacent block, iirc 2021-04-30 01:10:55 Sheila: if you mean libresoc, that got changed from riscv to power 2021-04-30 01:11:02 f 2021-04-30 01:11:02 like, NVENC would be fine 2021-04-30 01:11:09 if there were a free driver for it 2021-04-30 01:11:12 btw. does apk add !crap like disable installing that package? did not find it in man 2021-04-30 01:11:19 helby: yes 2021-04-30 01:11:29 hidden gem 2021-04-30 01:11:51 but defaults are sane; pit of success 2021-04-30 01:12:05 btw, what's the best way to make a compile-time option in an apkbuild? i don't see there being actual support for user-defined options 2021-04-30 01:12:13 q66: does not yet exist 2021-04-30 01:12:16 q66: open to ideas 2021-04-30 01:12:42 ACTION wanted to have some easy way to disable llvm in mesa 2021-04-30 01:15:24 it has to be possible to run OBS headless 2021-04-30 01:15:53 mainly so i could keep the amount of local changes minimal anyway, now that i actually fixed webkit on musl it seems i'll be actually deploying alpine for the customer after all 2021-04-30 01:20:21 in the past people usually installed software, now they are 'deploying' ;) 2021-04-30 01:20:21 btw, do the llvm patches get upstreamed? at least few of them seem trivial. 2021-04-30 01:22:17 (or is it #llvm who usually pick them up?) 2021-04-30 01:25:23 i can upstream few if it is ok? :) 2021-04-30 01:25:25 (needed to try some things with llvm-libunwind on sabotage/void-musl at the same time) 2021-04-30 01:26:32 buh 2021-04-30 01:26:54 i built 4K page size kernel and that helped a little, but OBS still crashes under Xvnc 2021-04-30 01:27:11 maybe i just plug lantronix spider into the card :D 2021-04-30 02:10:28 q66: ... gentoo? 2021-04-30 02:11:59 Ariadne: does ffmpeg not do the needful? 2021-04-30 02:16:31 Hello71: needs to output 2160p60 video stream 2021-04-30 02:17:07 i don't see how that's related 2021-04-30 11:23:40 o/ 2021-04-30 11:24:03 for a raspberry pi 3 model a+, is is linux-rpi or linux-rpi2 that we must install? (assuming armv7) 2021-04-30 11:53:50 looks like its pi2 according to the image 2021-04-30 12:02:06 Hello71: we need the ability to mix multiple video streams into an output stream 2021-04-30 12:02:31 Hello71: so OBS acts as video mixer in that case :) 2021-04-30 12:02:31 ffmpeg can do this, albeit more clunkily 2021-04-30 12:04:08 I wish obs had a simple "export to ffmpeg command" function for static filter chains but unfortunately I think it implements most filters internally. I think ffmpeg has equivalents of most though 2021-04-30 12:04:44 hmm, i can read up on how to do video mixing with ffmpeg i guess :) 2021-04-30 13:45:50 good afternoon folks 2021-04-30 13:46:55 I am in the process of building a alpine aarch64 roots for the rpi 4b, however for some reason the boot hangs right after mounting the root partition 2021-04-30 13:47:04 [ 4.017940] EXT4-fs (mmcblk0p2): mounted filesystem with ordered data mode. Opts: (null) 2021-04-30 13:47:04 [ 4.027091] Mounting root: ok. 2021-04-30 13:47:20 any idea of what the reason might be? I don't get no more output after that 2021-04-30 13:48:18 the rootfs was put together under a ubuntu aarch64 on the same device using 2021-04-30 13:48:22 ./sbin/apk.static -X http://uk.alpinelinux.org/alpine/latest-stable/main --arch aarch64 -U --allow-untrusted --root . --initdb add alpine-base 2021-04-30 13:49:02 I am wondering if the issue might be related to permission corruption 2021-04-30 14:12:39 How does one limit the number of getty ttys to, say, three? 2021-04-30 14:13:26 I've edited rc_tty_number=3 in /etc/rc.conf, but it doesn't seem to have any effect 2021-04-30 14:33:39 kiedtl: look at /etc/inittab and comment the gettys you don't want/need. 2021-04-30 17:56:12 even bot is dead, he should write something, that it's not so dead here 2021-04-30 17:58:56 algitbot: thanks 2021-04-30 18:06:49 what is the pipewire status integration? 2021-04-30 18:31:17 algitbot: date 2021-04-30 18:31:23 algitbot: time 2021-04-30 20:13:07 caskd: Huh, thanks for that tip. 2021-04-30 23:28:37 Latrina: what is "permission corruption"