2024-09-01 08:56:35 a 2024-09-01 10:41:53 a suspense 2024-09-01 14:46:41 I'm using setup-bootable to build a custom alpine live-usb. I've noticed that it places a copy of ldlinux.c32 and ldlinux.sys at the root of the disk, but these don't appear to be used by anything? as in, if I delete them it still boots fine. The official iso only has ldlinux.c32, and it's inside /boot/syslinux rather than the filesystem root. Anyone know why setup-bootable does this differently? 2024-09-01 14:48:33 oh, and it looks like setup-bootable also puts a second copy of ldlinux.c32 inside /boot/syslinux, to match the iso 2024-09-01 22:12:11 anyone try playing around with alpine+flatpak steam on the steam deck? curious if the controller works fine and if the refresh rate slider/gpu freq slider work out of the box or if the correct knobs can be exposed to make them work 2024-09-01 22:12:31 running nixos on mine currently but kinda missing alpine tbh 2024-09-02 01:14:49 How do i properly from inside a script pass a sting with spaces onto a script? i tried this one ./te.sh "$line4" 2024-09-02 01:16:12 i get a problem of filenames being cut at the first space counted from the right 2024-09-02 01:45:50 Let me see if i can paste all 7 lines in the script in case someone got a better idea to solve it all :) 2024-09-02 01:45:56 while IFS= read -r line; do 2024-09-02 01:45:56 ./te.sh "$line4" 2024-09-02 01:45:56 line4=$(echo "$line" | cut -b 45-) 2024-09-02 01:45:56 if [[ "$line" == *".mcr"* ]]; then 2024-09-02 01:45:56 cat "$line4" 2024-09-02 01:45:56 fi 2024-09-02 01:45:58 done < "$1" 2024-09-02 01:46:43 worked :) 2024-09-02 03:50:13 Babba: there's nothing wrong with the code you've provided; more likely whatever te.sh is doesn't handle parameters with spaces correctly 2024-09-02 09:29:11 how to set up loginctl? 2024-09-02 09:29:25 https://wiki.alpinelinux.org/wiki/Elogind 2024-09-02 09:29:34 i read this and loginctl wont detect my session 2024-09-02 09:30:10 summershine:~$ doas rc-service polkit status 2024-09-02 09:30:10 summershine:~$ doas rc-service elogind status 2024-09-02 09:30:10 * status: started 2024-09-02 09:30:10 * status: started 2024-09-02 12:18:51 Hello, 2024-09-02 12:18:51 I'm encountering issues while building Docker images with Alpine Linux. The build fails with the following error: 2024-09-02 12:18:54 "WARNING: updating and opening https://dl-cdn.alpinelinux.org/alpine/v3.19/main: temporary error (try again later)" 2024-09-02 12:18:57 This occurs across multiple Alpine versions. It seems the APK repositories might be temporarily down. Has anyone experienced this or have suggestions for resolving it? 2024-09-02 12:19:04 Thanks in advance! 2024-09-02 12:20:32 if you pop the URL (https://dl-cdn.alpinelinux.org/alpine/v3.19/main/) into your browser, you will probably see that the mirror is up 2024-09-02 12:20:57 that error message is a bit terse, but it probably points to a connectivity issue inside your build environment 2024-09-02 12:23:00 Thank you for the suggestion. I will check the network settings and connectivity within my Docker build environment to address this issue. 2024-09-02 12:26:49 Also, http seems to be the recommended proto, not https 2024-09-02 12:27:22 Could be part of what causes the issue you're having 2024-09-02 12:27:38 i use https myself and haven't had an issue. but if your clock is desync'ed enough, ssl cert verification could fail ya 2024-09-02 12:27:39 s/recommended/target/ 2024-09-02 12:27:56 peerl3ss, you aren't :) 2024-09-02 12:29:08 Just for an initial header exchange to tell your client to connect over http 2024-09-02 12:29:16 At least that's what I can see from here 2024-09-02 12:29:51 definitely uses ssl for me 2024-09-02 12:30:15 well, tls :) 2024-09-02 12:30:15 Can you try: curl -I https://dl-cdn.alpinelinux.org/alpine/v3.19/main 2024-09-02 12:30:23 Yeah ;) 2024-09-02 12:32:53 not sure what that's about, but i get an http 200 when i dl a pkg apk via https and curl 2024-09-02 12:33:21 I get a 301 to http://dl-cdn.alpinelinux.org/alpine/v3.19/main/ 2024-09-02 12:33:35 i get that on the landing page for some reason. odd 2024-09-02 12:33:39 That's… Weird that you don't get it too 2024-09-02 12:33:45 ah :D 2024-09-02 12:33:46 browser doesn't redirect tho 2024-09-02 12:33:46 ok 2024-09-02 12:33:55 strange 2024-09-02 12:33:59 That's an nginx issue if you leave out the final / 2024-09-02 12:34:12 ohh 2024-09-02 12:34:24 ohhh 2024-09-02 12:34:32 Would it be fixeable? 2024-09-02 12:35:12 good question, I haven't found a good solution yet 2024-09-02 12:35:13 I mean is it a configuration issue, or an inherent nginx problem 2024-09-02 12:35:17 ok 2024-09-02 12:35:29 Thanks for clarifying anyway ikke :) 2024-09-02 12:35:46 mkerkeni, sorry for the noise, can go on! 2024-09-02 12:35:55 Problem of nginx being behind a reverse proxy that does the tls termination 2024-09-02 12:37:00 right, it'd have to force the scheme instead of taking automatically from the server type 2024-09-02 12:38:46 They say: http://nginx.org/en/docs/http/converting_rewrite_rules.html 2024-09-02 12:39:19 Hummm, though I don't know what rules are actually there, I'll shut up. :) 2024-09-02 12:39:23 FWIW, wget says (among other): « URL transformed to HTTPS due to an HSTS policy » 2024-09-02 12:42:14 Got it, I'll check the nginx rewrite rules and HSTS policies. Thanks for the info! 2024-09-02 13:04:07 Hey all, I got it fixed! It was a time sync issue, as suggested. I synced the time in the Docker container, and now everything's working fine with the apk commands. Thanks for the help! 2024-09-02 14:55:37 lvscan show me inactive volumes as shown in the pic. Kindly need command line to make them active. Reference to the picture as actual setting going on, please 2024-09-02 14:56:29 ACTION uploaded an image: (89KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/aYwppzFhJBuZcAcdtITvAxLY/2024-09-02-19-49-50-967.jpg > 2024-09-02 14:56:56 the basic syntax iirc is lvchange -ay / 2024-09-02 14:57:52 usually tho i just do vgchange -ay to activate all volumes at once 2024-09-02 14:59:14 Thankyou 2024-09-02 15:00:22 sure, both the man page and --help are pretty cumbersome :P 2024-09-02 15:59:39 My machine keeps periodically rebooting suddenly (not a clean reboot; more like suddenly dies and then boots back up). Trying to figure out how I can try to get to the bottom of this in Alpine. Any ideas on pinning down the problem? 2024-09-02 16:01:41 RAM passes memtest86 (though the RAM stick used to have a partner that failed the memtest so the failed stick is gone) 2024-09-02 16:03:10 Might be worse problem than just a bad memory module 2024-09-02 16:03:19 SSD passes SMART health assessment. 2024-09-02 16:03:39 My current suspicion is bad motherboard. I've already replaced the PSU. 2024-09-02 16:11:24 ok 2024-09-02 16:11:25 yeah :/ 2024-09-02 16:30:03 grossaffe[m]: i experienced this before, your motherboard is probably shorting to the case 2024-09-02 16:30:20 Oh, that's an interesting thought. 2024-09-02 16:30:43 grossaffe[m]: i had the same issue (random reboots) when i ran alpine on a raspberry pi. the problem was indeed bad hw, but i had to narrow it down slowly by elimating other potential causes 2024-09-02 16:31:38 if you ran afford to sit and watch your dmesg logs until the issue recurs, you can run `dmesg -w` to tail them. you may get a kernel panic and a hint at the particular problem before the machine resets 2024-09-02 16:33:31 another thing to look at is file system corruption. you can reinstall all packages with `apk upgrade -a`. to detect corruption, generate a message authentication code for each file before and after this step. depending on the file system you chose, this may be relevant 2024-09-02 16:34:25 This is a new system. And this problem has been persistent across clean installs, so very much doubt package corruption to be the culprit. 2024-09-02 16:35:59 i suspected the SD card on the pi was dying but had to rule this out. but i think it is fair to say that sudden reboots can cause corruption and make the system even more instable 2024-09-02 16:37:01 i assume you're not running any out-of-kernel modules. those can cause issues sometimes 2024-09-02 16:37:55 I'll take a look at re-installing the motherboard and see if that helps. This is a mini-ITX board in a small case. As for corruption, Yes, I could see the sudden reboots potentially causing further issues to the filesystem (I see when it reboots that it is always fixing some fileystem issues, which isn't shocking given it doesn't shut down properly). 2024-09-02 16:37:55 And yes, I am using a pretty vanilla Alpine setup. No custom kernel. Just trying to make myself a little media server. 2024-09-02 16:39:32 yeah, i'd say getting the panic is your best bet. does the system ever just freeze instead of reboot? 2024-09-02 16:40:28 if not, would make sense if power issue 2024-09-02 16:40:46 It used to have freezing problems before I swapped out the PSU, I think. 2024-09-02 16:41:39 PSU was the only part I bought used for this build (the RAM was previously used, too, but by me and was known working; this machine killed one of my sticks) 2024-09-02 16:42:50 hmm, yeah that does make me suspect power and overvoltage protection 2024-09-02 16:43:09 err overcurrent 2024-09-02 16:43:13 But that PSU has since been replaced with a brand new EVGA PSU. 2024-09-02 16:43:38 My fear is that the old PSU may have damaged the motherboard, though. 2024-09-02 16:44:21 mhmm. i'd say all bets are off if a component died already 2024-09-02 16:50:04 i think the chipset (assuming intel here) runs on the same 3.3V rail as the memory modules 2024-09-02 16:51:34 AMD. Went with an AM5 APU (little did I realize at the time that AMD's AV1 encoder was broken for 1080p video) 2024-09-02 16:56:23 oic... another idea is to try inducing a failure with a stress test. you may be able to isolate the component that is defective if kernel panic logs are not fruitful 2024-09-02 16:57:59 i remember getting failures more often when i was compiling. this page has lots of good ideas: https://wiki.archlinux.org/title/Stress_testing 2024-09-02 16:58:20 My machine is h appy to fail at idle. 2024-09-02 16:59:57 haha, mine was too. but more prone with memory operations and cpu load 2024-09-02 17:00:51 Okay, I'll add stress test to my to-do list. 2024-09-02 17:01:22 you can run dmesg -w in tmux in one window and the stress test in another 2024-09-02 17:28:42 ACTION uploaded an image: (106KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/EgGyZjdJKcvXjXRYISEacFqm/2024-09-02-22-28-23-587.jpg > 2024-09-02 17:28:44 Need help with this 2024-09-02 17:47:40 how is alpine compared to kicksecure 2024-09-02 17:48:58 That assumes we know what kicksecure is 2024-09-02 18:02:26 i have not used kicksecure but i have used whonix. assume it's p similar 2024-09-02 18:02:41 oh, he left :P 2024-09-02 18:05:19 It might be https://www.kicksecure.com/ > Kicksecure is a secure by default OS always based on the latest security research. 2024-09-02 18:05:31 Never heard of it. 2024-09-02 18:08:35 ok, and i never tried Alpine OS, do you guys reccomend it as a daily, or only as a portable live os 2024-09-02 18:09:32 We absolutely recommend it as your daily os! 2024-09-02 18:09:58 (Depends on what you want to achive actually) 2024-09-02 18:10:52 my biggest reservation about recommending it for daily use is not being able to install multiple kernel versions natively. at least, not that i know of. i prefer running it virtualized so i can seamlessly roll back to previous kernels with file system snapshots 2024-09-02 18:10:53 well, kinda not have the OS spying on me. Do a bit of video and photo editing and run some social apps 2024-09-02 18:11:32 "2024-09-02-22-28-23-587.jpg" <- ? 2024-09-02 18:11:32 thanks for input 2024-09-02 18:13:20 why would you want to roll back to previous kernels / im still new to linux distros 2024-09-02 18:14:16 literally every linux distro out there doesn't spy on you 2024-09-02 18:15:34 user_: you might run into some issue with the newer kernel. maybe a kernel module failed to compile and you didn't notice before you rebooted. perhaps you upgraded from alpine 3.19 to 3.20 and some hardware isn't working the way you expect. it is nice to roll back in such situations 2024-09-02 18:16:48 i haven't tried kicksecure, but the impression i get is that it's basically a hardened install like whonix. except whonix is distributed in two parts: a gateway and a workstation. this is done to prevent malware on the workstation from deanonymizing your tor connection. the two projects have some relation to one another but i can't remember what 2024-09-02 18:18:33 if you just want a no nonsense distribution with sensible defaults and that doesn't spy on you, i generally recommend debian for people newer to linux 2024-09-02 18:21:26 Hardened distributions like kicksecure and whonix exist to make it harder for malicious (or exploited) programs to spy on you, no distribution itself spies on you (at least not the popular ones), like Ermine said 2024-09-02 18:24:27 I mean i migrated to Linux because win did unreasonably much spying. Then i stumbled upon kicksecure and now Alpine. And i have a bit of truble with kicksecure and apps to run on it. and it looks kind of dull, so i will definitly try Alpine. Thanks for the input guys 2024-09-02 18:26:24 if you don't have the technical know-how to troubleshoot and resolve hardening-related issues, it's not ideal to use them. hardening breaks things and that's why most distributions and kernels are not super hardened by default 2024-09-02 18:27:44 unfortuneatly, hardening is probably the way to go forward, more or less. 2024-09-02 18:28:15 have you looked into qubes at all? if you really want a secure distribution, it's the one i would recommend 2024-09-02 18:31:12 one nice thing about qubes is that you can pretty seamlessly run mainstream distros for the applications you need. it gives you a radical amount of control over sandboxing that is kind of mind-blowing. nice balance of compatibility and security 2024-09-02 18:35:48 thanks peerless. Yes but do you think Qubes is suitable as daily OS? I did run it before with whonix from usb. Now Im looking for a better daily experience. Will have a go with it again. thanks 2024-09-02 18:37:19 Do you know if Mullvad is working on Alpine? 2024-09-02 18:38:24 you can use just about any openvpn or wireguard service on linux 2024-09-02 18:38:59 i ran qubes for about a year as a daily. it worked well. i did end up running back to debian because security and convenience are at odds, and i had no real need for that level of isolation. there are things i learned from qubes that i brought into my daily practice though. it was a great learning OS 2024-09-02 18:43:11 security claims are hard to verify, because stuff changes all the time. but defaults do matter in the comparisons. 2024-09-02 18:45:26 you could just take a mainstream distro and apply some hardening techniques yourself. if things break, revert the changes you made. this is my bible on that subject: https://madaidans-insecurities.github.io/guides/linux-hardening.html 2024-09-02 18:48:06 thanks! 2024-09-02 18:48:08 or follow a STIG for AlmaLinux or SUSE https://www.stigviewer.com/stig/red_hat_enterprise_linux_9/ 2024-09-02 18:51:48 having gone down this rabbit hole years ago myself, i'd recommend avoiding hardened kernels and honestly, just use systemd lol. yes it's more attack surface but it makes service sandboxing trivial 2024-09-02 18:54:19 if you really want to compile your own slimmed down kernel, you can do so on fedora and follow the RHEL guidance for the most part 2024-09-02 18:54:58 gentoo is not for the faint of heart lol 2024-09-02 18:58:53 people are not good about setting constraints for themselves; which is why sane, well documented defaults matter. that's where the conversation starts. it's harder to have a conversation around innumerable security guides or wikis or youtube videos, primarily oriented around systemd, which alpine is not. 2024-09-02 19:01:22 downloaded alpine now, need to find somewere to run the iso from 2024-09-02 19:07:28 can you recomend an app for easy acces to apps 2024-09-02 19:07:45 like libray folder kind of thing 2024-09-02 19:13:10 for me, the main pitch of running alpine as your daily driver is having a personal bias towards simplicity. as a downstream consequence of security-conscious decisions e.g. running openrc instead of systemd, ash instead of bash, and busybox instead of gnu coreutils, the system is far less complex than are most distrubitons 2024-09-02 19:13:43 qubes is a decent amount of work but highly secure and compatible. source distributions are way too much work for me and keeping tabs on upcoming kernel features and flags is just painful. alpine "just works" (at least if you don't need GNU libc in which case there is a compatibility layer that mostly works) 2024-09-02 19:20:44 still, my main system is a hypervisor and i prefer systemd to openrc for that purpose. i also think it's better to learn systemd than openrc for general futureproof linux knowledge, as this is where most distributions have gone. likewise with gnu coreutils vs busybox 2024-09-02 19:23:55 sorry if i have triggered all of the openrc diehards :'D i DO like it quite a bit but not for everything 2024-09-02 19:54:07 can you turn pmos into alpine? but i wanted to use unl0kr 2024-09-02 20:15:16 scorpion2185[m]: seems alpine-to-pmos isn't well-supported: https://gitlab.com/postmarketOS/pmaports/-/issues/106 . the other direction has no documentation that i can find. in theory reversing https://wiki.postmarketos.org/wiki/Existing_Alpine_installation should work but i have no experience trying this and expect there be dragons... to get unl0kr on alpine, you will have to pull https://github.com/droidian/unl0kr and compile 2024-09-02 20:17:38 you might want to first test that you can compile unl0kr in an alpine virtual machine. it looks to me that you can obtain all of the dependencies that are not statically linked from alpine's package list 2024-09-02 20:21:55 in fact, you could try all of this in a virtual machine first to get the procedure down pat before messing with your real device 2024-09-02 20:26:04 unl0kr is in pmos repo it's from pmos actually, problem is https://gitlab.alpinelinux.org/alpine/mkinitfs/-/issues/18 2024-09-02 20:26:21 when i turned alpine into pmos it messed up my tablet 2024-09-02 20:27:02 it's in buffybox "bundle" https://gitlab.com/postmarketOS/buffybox 2024-09-02 20:28:45 i wanted to install alpine on a pinephone using pmos repo masked for kernel etc. 2024-09-02 20:28:55 i suspect you don't want to mix pmos and alpine repos if your goal is to wind up with an alpine system, but alpine does not have unl0kr 2024-09-02 20:29:28 that's why you mask them 2024-09-02 20:30:01 does apk use the repo that's before in the file? 2024-09-02 20:32:23 not sure... the wiki page seems to imply that order matters 2024-09-02 20:51:48 seems like you'll have to hack quite a bit to get it working and maintain your hack to keep it working. given that an init.sh refactor exists in pmos, the biggest remaining blocker on a prospective MR is the creation of a regression test suite. you may be able to ping #alpine-devel about the feasability of this if you are not interested in rolling your own init.sh hack 2024-09-03 15:59:11 peerl3ss: I have followed https://wiki.alpinelinux.org/wiki/LVM_on_LUKS#Creating_and_Mounting_the_File_Systems guide for installing alpine full disk encryption. when at last I `reboot` and pullout USB. it booted to alpine and ask for encrypted password. but whenever I reboot again, I shows no bootable device found 2024-09-03 16:00:07 my laptop is Toshiba satellite pro NB10t 2024-09-03 16:00:25 I do have uefi 2024-09-03 16:19:13 Velo_Frost[m]: to be clear: you followed the uefi/gpt instructions, and ignored the dos/mbr instructions? 2024-09-03 16:22:14 opinion, that page should be split into two pages. one for uefi/gpt and another for dos/mbr. the subject matter gets more confusing when it tries to handle both. 2024-09-03 16:24:52 i followed uefi/gpt 2024-09-03 16:25:33 is dos/mbr good to try when facing the problem as i discussed above? 2024-09-03 16:26:10 moreover any other user guid for full disk encryption 2024-09-03 16:30:05 Toshiba satellite pro NB10t bootloader key 2024-09-03 17:10:04 Velo_Frost[m]: dos/mbr always works 2024-09-03 17:10:31 good to hear 2024-09-03 17:10:45 but why is the problem with uefi 2024-09-03 17:11:35 there shouldn't be but, in either case, everything has to be done right. 2024-09-03 17:11:38 and for dos/mbr my bootloader be uefi or CSM? as I only have these two options in bootloader 2024-09-03 17:12:47 invoked: all right, can I make a list of commands that I have to use with setup. if I am missing anyone and you can guide please 2024-09-03 17:15:06 Velo_Frost[m]: there is also setup-alpine which will do this for you 2024-09-03 17:15:39 that is not encrypting lvm 0n luks 2024-09-03 17:16:22 # setup-keymap... (full message at ) 2024-09-03 17:21:51 https://wiki.alpinelinux.org/wiki/LVM_on_LUKS 2024-09-03 17:22:32 i was just suggesting it as a more turnkey thing. 2024-09-03 17:32:02 in my own case -- usually -- i manually do mbr with grub, encrypted root, cryptkey, vg name == $HOSTNAME 2024-09-03 17:33:27 i thought about doing some MR with that as an option, but i'm not sure it would be in acceptable form. 2024-09-03 17:34:27 there's still things that can go wrong no matter what approach 2024-09-03 17:53:11 # setup-keymap... (full message at ) 2024-09-03 17:58:05 Velo_Frost[m]: > <@surya_x3:matrix.org> # setup-keymap... (full message at ) 2024-09-03 18:00:01 * ```... (full message at ) 2024-09-03 18:00:51 my apology 2024-09-03 18:03:19 Velo_Frost[m]: don't confuse your efi partition with your boot partition 2024-09-03 18:03:47 this is also why mbr is simpler (2tb disk limitation aside) 2024-09-03 18:03:50 invoked: these are two different things? 2024-09-03 18:03:54 yes. 2024-09-03 18:04:14 some people make them one thing, but they know exactly what they are doing 2024-09-03 18:04:26 lvcreate -L 16G vg0 -n boot 2024-09-03 18:04:32 and 2024-09-03 18:04:54 i have to step out. bbl 2024-09-03 18:05:03 mkfs.ext4 /dev/vg0/boot 2024-09-03 18:05:14 ok. all my steps are ok? 2024-09-03 18:17:57 https://paste.mozilla.org/ajwfbQxe 2024-09-03 18:17:57 kindly check the paste in for full disk encryption lvm on luks. Are these commands correct 2024-09-03 18:19:13 if some correction there is edit button on top to correct 2024-09-03 18:19:14 please 2024-09-03 18:31:18 clandmeter: friendly greetings :) 2024-09-03 18:32:07 just created a wiki account and saw there's a contribute artwork .. i flew about swiss alps and did some photos, so if there's an image you like and want to use: 2024-09-03 18:33:13 https://www.flickr.com/photos/aiei/sets/72157650144744776/with/24400951973/ (for some reason doesn't load for me, but maybe for you?) 2024-09-03 18:36:59 works for me. there are some really nice pictures in it 👍️ 2024-09-03 18:37:37 take with a canon 5d2 100mm 2024-09-03 18:37:40 fossdd[m]: can you check the paste in plz 2024-09-03 18:38:29 I Boot to USB media after 1st reboot when alpine full disk encryption is installed 2024-09-03 18:38:57 one time it Boot to sda than grub cannot find media device 2024-09-03 18:40:59 fossdd[m]: thanks. i should probably dig out the 1000 others, which are offline 2024-09-03 18:41:20 one alpine background image for every release? 2024-09-03 18:42:43 i'm mainly interested in small and desktop (multimedia, video, audio) and gnustep, so i will try to contribute. and because i like small, consistent and non-distracting, i'm using this as background: https://github.com/alexmyczko/autoexec.bat/blob/master/gray.png 2024-09-03 18:42:57 (works really well on the eyes, no matter if it's day or night) 2024-09-03 18:45:12 Velo_Frost[m]: your GRUB_CMDLINE_LINUX_DEFAULT is not correct. it should be, e.g. "modules=sd-mod,usb,ext4,nvme,keymap,whatever,whatever [options]" 2024-09-03 18:45:38 gone again 2024-09-03 18:46:53 invoked: all right it should contain (,)? 2024-09-03 19:13:25 and do I have to install syslinux? 2024-09-03 19:14:25 is it a necessary part? to write MBR without system partition. 2024-09-03 19:14:42 as I use uefi 2024-09-03 19:43:24 All right invoked I have successfully booted to encrypted alpine linux. what change I have to made so I don't Boot back to media installation grub window 2024-09-03 19:43:42 any post installation process I have to do 2024-09-03 19:54:46 i don't know what's wrong. but when i type date, i get today 2097. and when i uptime says 24855 days... 2024-09-03 19:55:08 lol 2024-09-03 19:55:12 How is it in the future? 2024-09-03 19:56:47 what's an rdate server that works? 2024-09-03 19:58:00 Never used rdate 2024-09-03 19:58:21 ok chrony says it's running, but it's still 2097 2024-09-03 19:58:38 alex_: what does chrony tracking say? 2024-09-03 19:58:40 chonyc 2024-09-03 19:59:23 it says many letters. which line do you want? i can't copy paste from this UTM 2024-09-03 19:59:33 last offset 2024-09-03 19:59:48 +0.005905933 seconds 2024-09-03 19:59:53 and the ref time it shows right 2024-09-03 19:59:59 huh 2024-09-03 20:00:12 alex_: im going to switzerland saturday. ill make some pics :) 2024-09-03 20:00:31 clandmeter: enjoy!. i'm in poland wed-sun 2024-09-03 20:00:48 how did you shoot them? 2024-09-03 20:01:05 ikke: ok /bin/date is busybox and i must confess randomly run upx on /usr/bin, /bin, /usr/sbin, /sbin binaryes, including busybox* 2024-09-03 20:01:23 clandmeter: sit in a small airplane, aim with camrea and click? 2024-09-03 20:01:35 oh, so upx may have broken things? 2024-09-03 20:01:48 let me upx -d busybox and see 2024-09-03 20:02:25 busybox is back to 935k from 509k and date is still wrong 2024-09-03 20:02:53 i'm not sure what exactly broke things, everything looked fine until i wanted to debug/fix ruptime 2024-09-03 20:03:21 clandmeter: Maybe you can make a picture like https://dev.alpinelinux.org/~kevin/alpine_wallpaper_2.png :P 2024-09-03 20:03:31 i guess just click and it stopped working 2024-09-03 20:06:27 ok i run rdate -s time.nist.gov and it gets even better Sat Oct 11 04:34:15 CEST 2160 2024-09-03 20:07:58 ikke: ill see what I can do 2024-09-03 20:09:44 ok date 100322092024 more or less fixed the time, but uptime is still wrong... 2024-09-03 20:10:56 hah using strace, checking /proc/uptime ... some seconds later date is wrong again.. 2024-09-03 20:12:33 stopping chrony, setting time, running date, wrong :( 2024-09-03 20:13:20 ok the virtual machine arm64, has been running a few days, with sleep (lid down up) hosted on macOS but this should not happen, trying reboot 2024-09-03 21:33:08 "All right invoked I have..." <- I have to install grub-install --target.............. again to boot my installed alpine linux. and I have to mount root,boot,sys,proc,swap and than install grub and finally umount all and reboot. than I am only able to boot to installed alpine linux in grub. what is the issue here? 2024-09-03 21:33:33 any body who has the same issue 2024-09-03 21:39:02 Do I need to install windows along side? as I only have installed alpine linux on whole system 2024-09-03 22:32:05 whydo you want windows? 2024-09-03 22:33:59 Occasionally for FW updates, but honestly... once the Linux system runs it hardly matters. I do have one laptop where I regret wiping Windows, as it has a graphics driver bug in every Linux distro: black screen after every mode switch, only fixable by a suspend/resume cycle. 2024-09-03 22:34:20 I do slightly wonder if a FW upgrade would help, who knows. 2024-09-03 22:34:41 But I also don't want to wipe the OS... 2024-09-03 23:18:21 I'm trying to write Alpine's standard x86_64 ISO to a USB drive and it's giving me hell. Tried downloading the image again, checked the hashes and signatures, nothing. Etcher did give me this error (https://imgur.com/a/ofAHwvA) but I don't know anything else. 2024-09-03 23:18:40 It just loops me back to the boot select menu with no error or anything. 2024-09-03 23:19:02 I even tried making a new GPT table and extracting the ISO contents to a FAT32 partition. 2024-09-03 23:19:19 Also wrote out /dev/zero to my flash drive to clear it, nothing. 2024-09-03 23:27:15 "whydo you want windows?" <- I cannot seet grub after installing alpine. its just show on 1st reboot when I install alpine. login and when I want to reboot and start laptop again, it shows **"Reboot and select proper Boot device or insert Boot Media in selected Boot device and press any key"** 2024-09-03 23:27:29 s/seet/see/ 2024-09-03 23:29:22 grub was installed (chroot) grub-install --target=x86_64-efi --efi-directory=/boot/efi 2024-09-03 23:30:58 efi partition is not encrypted which is sda1 and other sda2 partition is encrypted. 2024-09-03 23:36:54 GRUB is installed but the menu is not shown at boot when reboot. I only see GRUB once and for all and that when I finished all installation ALpine and type reboot. 2024-09-04 05:59:40 Velo_Frost[m]: i've run into some issues with EFI installs that were a bit difficult to troubleshoot 2024-09-04 06:01:10 it's been awhile and my memory is vague, but there was some issue with having seperate EFI and boot partitions in my case 2024-09-04 06:01:35 so what i wound up doing is: combine the /boot and /boot/efi mount points into a single partition, formatted as vfat 2024-09-04 06:03:34 on that unified partition, set the type code in gdisk (assuming you are using GPT disks) to EF00 2024-09-04 06:06:49 it took awhile to find that that arrangement worked and now all my alpine VMs are set up that way. it's worth a shot in your case 2024-09-04 07:59:59 "so what i wound up doing is..." <- I have efi boot partition on sda1 (not encrypted) 2024-09-04 07:59:59 Do you mean to combine them to one? 2024-09-04 07:59:59 And addationally created lvcreate -L 2G vg0 -n boot  in an encrypted volume sda2 2024-09-04 08:06:45 Where as I have mounted as below... (full message at ) 2024-09-04 08:20:15 Velo_Frost[m]: yes, this is what i had to do to get GRUB to boot Alpine on an UEFI system (though I did not use LVM) 2024-09-04 08:21:39 well, hold on. not quite "yes" 2024-09-04 08:22:08 the idea would be to combine boot and EFI on sda1 and not encrypt either 2024-09-04 08:22:52 300MB is more than enough space for both 2024-09-04 08:24:10 another option would to try a bootloader other than GRUB. i prefer GRUB, so i was willing to compromise my ideal partition layout to keep it 2024-09-04 09:16:36 "the idea would be to combine..." <- What would be the command line for both if using sda1 2024-09-04 09:17:59 And also mount command for this 2024-09-04 09:21:05 that depends on how you have your disks set up. if you had just one partition on the sda device, you could run "sgdisk -o -N 1 -c 1:boot -t 1:ef00 /dev/sda" 2024-09-04 09:21:34 i would imagine you have your lvm partition on sda also, so it's probably better to do the partitioning interactively with `gdisk` 2024-09-04 09:21:44 I have 2 partation 2024-09-04 09:22:01 peerl3ss: I did with parted 2024-09-04 09:22:38 ohh, you're doing mbr 2024-09-04 09:24:07 Sad... (full message at ) 2024-09-04 09:24:28 peerl3ss: No its got but MBR is protected 2024-09-04 09:24:45 * No it is GPT but MBR is protected 2024-09-04 09:25:17 s/Sad/Sda/ 2024-09-04 09:26:03 ok, gotcha. i usually use gdisk to set up GPT disks with protected MBR so i don't know the syntax off the top of my head 2024-09-04 09:27:45 it should be the same as this section: https://wiki.alpinelinux.org/wiki/LVM_on_LUKS#UEFI_with_GPT_disklabel 2024-09-04 09:28:52 i would 300M instead of 200M for the combined partition but honestly 200M is probably enough 2024-09-04 09:29:36 then you just skip the `lvcreate` for boot (command #2) of https://wiki.alpinelinux.org/wiki/LVM_on_LUKS#LV_Creation_for_UEFI/GPT 2024-09-04 09:31:05 `mount -t vfat /dev/sda1 /mnt/boot` to mount. in this case, efi is just a subdirectory of /boot not a mount point itself. you may need to mkdir /mnt/boot/efi 2024-09-04 09:44:21 All right. I too follow the lvm on luks guide which you have mentioned. But you said earlier that you didn't use lvm.  2024-09-04 09:45:02 yeah, i don't use encryption either, though i'd recommend GRUB for that 2024-09-04 09:47:47 but... if you are just using LVM for a swap partition and the root, i think you could certainly get by without it 2024-09-04 09:47:53 Thing is when I followed this guide using encryption and lvm. Uefi and grub. I boot to alpine successfully. Everything work fine. But once I reboot, I don't see any grub menu and all I see black screen asking for bootable device, press any key. 2024-09-04 09:49:09 hmm. did you reboot with the USB taken out? 2024-09-04 09:51:32 I than again boot from USB, setup-interfaces  select wlan0 for WiFi and than install cryptsetup and decrypt sba2 and mount all things. Chroot and install grub than reboot. Unplug USB immeditly and alpine boot to grub menu to select alpine or uefi settings. 2024-09-04 09:52:55 i'm not sure what to make of your being able to boot into your system successfully only once. in my case, i wasn't able to get past the bootloader unless i combined /boot and /boot/efi 2024-09-04 09:53:55 sometimes removing a device can change the boot order in the UEFI compared to when you had it in. you might just need to manually create a boot entry when that happens 2024-09-04 09:53:58 peerl3ss: Yes. I took USB out on 1st booting the machine, at that point grub work well. Login to alpine. But once login without USB in, and if I reboot, than the error occured 2024-09-04 09:55:33 peerl3ss: How can I manage manual 2024-09-04 09:56:07 depending on your motherboard, you can probably do this interactively in the settings menu when you switch the PC on 2024-09-04 09:56:08 I think the problems is grub. Which disappeared 2024-09-04 09:57:01 it could be that you're not getting to grub because the boot entry for it went away 2024-09-04 09:58:38 if you 'boot from file' in the uefi menu and select the disk containing your EFI partition, the file to boot from is `boot\bootx64.efi` (assuming you are on x86_64 CPU architecture` 2024-09-04 09:59:25 peerl3ss: Oh yes, that I select ssd than man than usb 2024-09-04 09:59:47 that's supposed to be efi\boot\bootx64.efi 2024-09-04 09:59:57 peerl3ss: Yes this is what happen. But I don't know how to manage 2024-09-04 10:00:02 (which the firmware would find automatically if it existed) 2024-09-04 10:00:18 I don't think grub installs itself there by default though 2024-09-04 10:00:51 on my system, i don't have the extra `efi` dir. but that's probably because i combined EFI and boot into one partition 2024-09-04 10:01:22 i would try boot from file and see if that works. if it does, you can manually add the new entry afterwards 2024-09-04 10:02:05 maybe a result of moving files around during said combining 2024-09-04 10:02:46 peerl3ss: I too have no extra setting in uefi in boot settings. Its just uefi,CBM or CSM thing and secure boot toggel. 2024-09-04 10:02:48 e.g. if you had /boot and /boot/efi, but told grub to install to /boot, and then moved /boot/efi to /boot... 2024-09-04 10:02:49 it has something to do with the --efi-directory=/boot/efi parameter to grub-install 2024-09-04 10:03:22 but it nevertheless has to be [efi_partition]\efi\boot\ from the firmware point of view (so e.g. /boot/efi/efi/boot from the linux perspective) 2024-09-04 10:03:41 ye, i just have /boot/efi/boot 2024-09-04 10:04:15 Velo_Frost[m]: there should be a list of boot menu entries in your settings somewhere. usually the boot from file option is in/near there 2024-09-04 10:04:41 if there's no "boot from file" look for "launch efi shell" 2024-09-04 10:06:12 anyway the point I wanted to make is that if the file were at the correct location [efi_partition]\efi\boot\bootx64.efi the firmware would most likely find and boot it automatically 2024-09-04 10:06:42 so if it somehow happens to be at just \boot\bootx64.efi, better figure out why that is the case first 2024-09-04 10:07:02 because no bootloader install puts it there... 2024-09-04 10:07:14 peerl3ss: Its not there. I search the whole uefi boot settings. I have Toshiba satellite pro mini laptop 2024-09-04 10:07:46 yeah not all firmwares have it 2024-09-04 10:07:54 is there an "efi shell" option though? 2024-09-04 10:08:33 grawity: No 2024-09-04 10:09:02 I will send the pic of my uefi screen in while. I'm on way to home 2024-09-04 10:10:26 what files do you have in the efi partition? 2024-09-04 10:10:39 well, what *.efi files 2024-09-04 10:18:42 ACTION uploaded an image: (52KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/LyWCTYQZlKYBMAilTguLfGGQ/2024-09-04-15-15-44-021.jpg > 2024-09-04 10:18:43 ACTION uploaded an image: (65KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/UdsNfOjwDITvsiFUoLukMAKo/2024-09-04-15-16-03-855.jpg > 2024-09-04 10:18:46 ACTION uploaded an image: (78KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/VRttZyffCdTQrEAPAbkgtQAp/2024-09-04-15-16-24-345.jpg > 2024-09-04 10:18:49 ACTION uploaded an image: (62KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/jdvgosvjPuNKxUkoUgjiPOEB/2024-09-04-15-16-37-048.jpg > 2024-09-04 10:18:52 ACTION uploaded an image: (52KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/tVJhsXydXFlHkGRUgpgEVgDL/2024-09-04-15-16-50-256.jpg > 2024-09-04 10:18:54 ACTION uploaded an image: (65KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/uRKuuzmyYMhSCgpdhuyduXkY/2024-09-04-15-16-58-719.jpg > 2024-09-04 10:18:57 ACTION uploaded an image: (63KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/VixldaXrqzbiHzPIsueSGgYE/2024-09-04-15-17-20-700.jpg > 2024-09-04 10:18:58 .. 2024-09-04 10:19:00 These are overall setting I have. No more no less 2024-09-04 10:19:05 ACTION uploaded an image: (62KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/FDRltQXOWXTYxVQkdVZKuWPb/2024-09-04-15-17-31-403.jpg > 2024-09-04 10:19:10 ACTION uploaded an image: (70KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/XnNwmSxRlmZWpMsxmwoZgemY/2024-09-04-15-17-52-721.jpg > 2024-09-04 10:19:40 ok 2024-09-04 10:19:51 what files do you have in the EFI partition? 2024-09-04 10:20:05 "what files do you have in the..." <- I don't understand what you mean. Sorry 2024-09-04 10:20:17 literally a list of files 2024-09-04 10:20:45 grawity: Let me check them 2024-09-04 10:31:13 "what files do you have in the..." <- Should I have to mount efi partition 1st to see the constant?  2024-09-04 10:32:57 mount -t vfat /dev/sda1 /mnt/boot/efi 2024-09-04 10:33:34 yes. you can list files with `find /boot -name '*.efi'` 2024-09-04 10:40:31 ACTION uploaded an image: (92KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/saKEssFboJEzZUMwGrJGJKuR/2024-09-04-15-40-08-821.jpg > 2024-09-04 10:40:33 ACTION uploaded an image: (127KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/hjTWAVNwAvPdKOGEBdaKCeOu/2024-09-04-15-40-17-318.jpg > 2024-09-04 10:41:21 ok, looks like the boot entry is indeed missing 2024-09-04 10:42:29 "yes. you can list files with `..." <- What is the exact command.  2024-09-04 10:43:14 if you have efi mounted to /mnt/boot/efi, it is `find /mnt/boot -name '*.efi'` 2024-09-04 10:45:25 peerl3ss: Wait I just didn't mount anything. I'm on bootable USB. Let me mount root,proc,sys et  2024-09-04 10:46:35 you can recreate the missing boot entry with efibootmgr: `efibootmgr -c -d /dev/sda -L Alpine` 2024-09-04 10:49:31 what file path does that default to? 2024-09-04 10:51:43 Command not ececuted 2024-09-04 10:52:20 find /mnt/boot-name'*efi' 2024-09-04 10:52:49 Boot-name is sda1 or it is as it is 2024-09-04 10:53:02 first of all you're missing a few spaces in there 2024-09-04 10:53:10 it was `/mnt/boot` separately from `-name` 2024-09-04 10:53:24 the `/mnt/boot` is wherever you mounted sda1 2024-09-04 10:53:34 so like if you did `mount /dev/sda1 /mnt/boot/efi` as you said previously 2024-09-04 10:53:46 then do `find /mnt/boot/efi -name "*.efi"` 2024-09-04 10:55:26 grawity: Work 2024-09-04 11:00:09 * Checking 2024-09-04 11:00:47 ACTION uploaded an image: (108KiB) < https://matrix.org/oftc/media/v1/media/download/AUoNVSB77nz8zsWI8uBFX71g6O2cIQr78JIRk9LZecSbAzzF8YEWeiVRBhU5HDZTwVyJp75SfxQbEpW4khrXnNtCeRvmki7gAG1hdHJpeC5vcmcvTE96WUZUR25RZmNxUklkaWh6SGt6cUN5 > 2024-09-04 11:01:47 interesting. no `bootx64.efi` 2024-09-04 11:02:14 yeah as I said grub-install does not install one by default 2024-09-04 11:02:37 some distros patch it to install both, I know debian does, but from what I remember, alpine doesn't 2024-09-04 11:02:56 otherwise the paths look ok though 2024-09-04 11:03:05 How can I add that 2024-09-04 11:03:20 you can either try adding the full boot entry with efibootmgr, like peerl3ss suggested before 2024-09-04 11:03:39 efibootmgr -c -d /dev/sda -l '\EFI\grub\grubx64.efi' -L Alpine 2024-09-04 11:03:46 `efibootmgr -c -d /dev/sda -l '\EFI\grub\grubx64.efi' -L Alpine` 2024-09-04 11:04:07 or you can temporarily copy the grubx64.efi to EFI/boot/bootx64.efi 2024-09-04 11:04:11 To sda or sda1 2024-09-04 11:04:21 efibootmgr wants the whole disk so just sda 2024-09-04 11:04:30 (it has a separate option for the partition number) 2024-09-04 11:04:35 Ok 2024-09-04 11:06:29 Which option is good to go now? 2024-09-04 11:07:03 ACTION wonders how he got his bootx64.efi on Alpine... 2024-09-04 11:07:21 if you added it, does the option now show up in `efibootmgr`? 2024-09-04 11:07:34 you may need to change your boot order so that the new entry is at the top of the list 2024-09-04 11:08:19 peerl3ss: maybe you installed grub with --removable once upon a time 2024-09-04 11:09:01 Let me install it first 2024-09-04 11:09:22 my guess is that it's an artifact of an old bootloader install 2024-09-04 11:09:37 Anyway I should installing chroot? 2024-09-04 11:10:00 As now I'm from usb 2024-09-04 11:10:25 i don't think you need to install it again 2024-09-04 11:11:48 I'm saying I haven't installed the efibootmgr.... Should I install in chroot or just over here as I'm bootloading from usb 2024-09-04 11:12:22 oh, it's fine to install it outside of chroot. probably a good idea to add it to your chroot as well tho 2024-09-04 11:12:38 Ok 2024-09-04 11:19:25 ACTION uploaded an image: (93KiB) < https://matrix.org/oftc/media/v1/media/download/ATeyTZ9tKBe2jsJ3TOi23zV4qfqBbexwqaw6qEk0GILhRp0m71C7hyIurgvKGn3Sm-DpSs9sn5OPrZhbdQ5QTIJCeRvnozYgAG1hdHJpeC5vcmcvU0psR01VaHNBdHNrQ29yTHNWdFp6Uk1X > 2024-09-04 11:20:24 looks good 2024-09-04 11:21:08 I mean 2024-09-04 11:21:24 the efibootmgr part looks good, but you've got to pay attention to the error messages from "choot" 2024-09-04 11:22:27 Oh that chroot /mnt 2024-09-04 11:28:35 I run it again. All well no errors 2024-09-04 11:29:10 Now its boot0004 2024-09-04 11:30:27 anyone knows some workaround for setting up a br0 interface without any attached network? I've just use it for qemu, works fine but I need to manual 'brctl addbr br0' 2024-09-04 11:32:20 ACTION uploaded an image: (134KiB) < https://matrix.org/oftc/media/v1/media/download/ARt-f24w8RiVVO8DVqkJwqkADVqI22K2CapKS0J0Dh5DL7xbjf1O-SVgWkI0r45dsUGp-wePewzusgueNs-F89FCeRvoYIYQAG1hdHJpeC5vcmcvSFRvZWV3aHZrU3ljaHphZXFCSGxvQ0pJ > 2024-09-04 11:34:50 that will work, but you don't need the duplicate 'alpine' boot entry 2024-09-04 11:35:31 No! It created 2 2024-09-04 11:35:31 Boot0000 2024-09-04 11:35:31 Boot0004 2024-09-04 11:41:05 you can delete the original with `efibootmgr -d /dev/sda -B -b 0` 2024-09-04 11:46:05 donoban: what is your current config for the bridge? I don't remember needing a workaround 2024-09-04 11:46:43 donoban: ip link add dev br0 type bridge 2024-09-04 11:54:07 https://tpaste.us/BJY9 2024-09-04 11:54:12 that it's the current 2024-09-04 11:54:48 ikke: but how do it persist reboot? 2024-09-04 11:56:36 I think that on my old laptop it worked fine without any bridge-port spsecified but probably is related with that laptop having an eth0 interface 2024-09-04 11:56:53 maybe since this don't have it something messes inside ifupdown... 2024-09-04 12:11:24 donoban: that is the configuration i use in proxmox and it persists across reboots. though i also add `bridge-fd 0` (probably doesn't matter) 2024-09-04 12:20:45 donoban: probably difference between ifupdown and ifupdown-ng 2024-09-04 12:21:26 That config looks fine, any issue with it? 2024-09-04 13:15:51 Does anyone use CUPS with KDE? With lpadmin and lp groups, users still have difficulty saving printer settings. Through some of the docs I have noticed a suggestion to use lppasswd to create a CUPS password digest file, but I don't know if this should be necessary with PAM, and I actually don't know what package to find lppasswd in. 2024-09-04 13:15:51 logread doesn't output anything relevant to troubleshooting this, and the only plasma prompt I get is "forbidden" after a few prompts for passwords. 2024-09-04 13:16:24 I can probably manually define these settings through through config file, just curious about what may be causing this 2024-09-04 13:24:46 ACTION uploaded an image: (134KiB) < https://matrix.org/oftc/media/v1/media/download/Af0o1Jnge2ndYZk3bcuJngPQ4sj8P5o9mBB2UD8naOd8lmD2yx36hXeAoopoI6qT3kO_tY9qd-Y7srI3NdhIN4dCeRvuz1WQAG1hdHJpeC5vcmcvbEplSmFKSUt0Y0dUYk5wVGVCQkZCdEVI > 2024-09-04 13:25:10 "you can delete the original with..." <- Seems good now 2024-09-04 13:25:58 Any further steps? So once rebooted I cannot lost this file from here 2024-09-04 13:27:36 you should be good to reboot. i'm not sure what caused your boot entry to disappear in the first place tho. would suggest recording the efibootmgr commands in case it happens again 2024-09-04 13:28:17 Does this need to be executed... (full message at ) 2024-09-04 13:29:10 And as I mounted all the below. Should I amount... (full message at ) 2024-09-04 13:29:51 since you were able to boot into the system before the entry was deleted, you shouldn't need to reinstall grub. yes, umounting before reboot is best practice 2024-09-04 13:30:46 Let me  exit chroot and amount all.  2024-09-04 13:30:49 peerl3ss: How? 2024-09-04 13:31:11 By recording what you mean sorry 2024-09-04 13:32:08 write `efibootmgr -c -d /dev/sda -l '\EFI\grub\grubx64.efi' -L Alpine` in a place where you can access it even if you can no longer boot again 2024-09-04 13:32:10 Since I run in chroot that command line you provided. Should I also run outside chroot. Or there is no need? 2024-09-04 13:34:30 it would be a good idea to run efibootmgr without arguments outside the chroot just to make sure the entry exists there 2024-09-04 13:34:38 peerl3ss: I have written it in my diary. But when I didn't boot I have to do all this process again.mount umound decrypt chroot and than this command line? 2024-09-04 13:35:34 actually, you can just run the command without mounting anything or decrypting next time 2024-09-04 13:37:34 Okayyy 2024-09-04 13:39:10 Any way boot0000* shows Alpine and HD(1, GPT, xxxxxx-xxxxx (UUID) 2024-09-04 13:39:10 What is this uuid and why HD not sda  2024-09-04 13:39:39 peerl3ss: But I have to plugin USB drive right? 2024-09-04 13:39:57 that is a GPT partition UUID 2024-09-04 13:40:29 hopefully you can do it without the USB 2024-09-04 13:46:04 "that is a GPT partition UUID" <- Than this is matching with sda1 and sda2 2024-09-04 13:46:32 ACTION uploaded an image: (44KiB) < https://matrix.org/oftc/media/v1/media/download/Ad0sq13fxz_RWXP8EdkO42T9wVnXkJtIVhLuhBTZqat9bJUNMyXRY3PcG6oXGbyXpNdDIFT0JJWu_OBk4Fh605NCeRvwDjywAG1hdHJpeC5vcmcvbVF3dG9LVFlWam1oQ3VrcVRUTW96Z3Bh > 2024-09-04 13:47:43 Velo_Frost[m]: Is this correct or should add ,  2024-09-04 13:48:01 that is correct without spaces 2024-09-04 13:48:56 "hopefully you can do it without..." <- Cannot do so because it only show after reboot about insert media. Press any key give me same error again and again 2024-09-04 13:49:21 okay, but it works if you plug in the USB drive? 2024-09-04 13:49:51 peerl3ss: You mean mine is correct? 2024-09-04 13:50:05 peerl3ss: Yes, I boot to Alpine installation from start 2024-09-04 13:50:08 yes, no spaces 2024-09-04 13:50:15 s/I/It/ 2024-09-04 13:50:23 err, no commas. only spaces lol 2024-09-04 13:51:26 peerl3ss: No space? Like atabaseide 2024-09-04 13:51:26 Or 2024-09-04 13:51:26 ata base ide ..... 2024-09-04 13:51:36 what you have in the file now is correct 2024-09-04 13:51:45 peerl3ss: Hahahah. I got confused 2024-09-04 13:51:47 Apology 2024-09-04 13:52:05 well, that is a conundrum you're in... seems your UEFI resets the boot entries whenever you add a boot device like the USB drive. but you need the USB drive to edit the boot entries when you cannot boot into Alpine 2024-09-04 13:52:49 and you cannot add a boot entry from the settings menu either =I 2024-09-04 13:54:12 Yes, I have to plug USB than starting alpine installation process, mount and umount stuff and than the command line to login to the actual Alpine Linux installed on my machine 2024-09-04 13:54:59 ACTION uploaded an image: (64KiB) < https://matrix.org/oftc/media/v1/media/download/ATr-c2jhcSVmcamD7OimGJA3G15aVzt6S22mvOv0VV0-M_gi6yGgiW4nJ2XcCkRvnaWwivt4VaHdO-oyDj3v-npCeRvwifhQAG1hdHJpeC5vcmcvdE9jUXBvTEhXSnJQYWlyckNrUEluTlRn > 2024-09-04 13:55:46 Check these values above, specially the grub_cmdline_Linux_default 2024-09-04 13:55:53 Is it correct  2024-09-04 13:56:23 looks ok. what does 'efibootmgr' show when you are in the installer again? 2024-09-04 13:56:56 And should I add GRUB_DISTRIBUTER="Alpine" 2024-09-04 13:58:16 i've never used that setting in /etc/default/grub. i would say no because it appears to be cosmetic 2024-09-04 13:58:24 peerl3ss: It show no HD (....uuid) 2024-09-04 13:58:24 Mean that vanished 2024-09-04 13:58:24 And yet the uuid is not matching to any of my sda1 and sda2 2024-09-04 13:58:47 peerl3ss: Ok 2024-09-04 14:00:17 it's not clear to me why it keeps vanishing. i think if it were me, i'd just give up on UEFI and try a BIOS boot install instead :) 2024-09-04 14:02:53 peerl3ss: That is my final card hahah 2024-09-04 14:03:09 ACTION uploaded an image: (111KiB) < https://matrix.org/oftc/media/v1/media/download/Ac0U4t-na6k1mcl1llRVUaaWpJxtv2AIQmQyb8fSqBts87YeCvFAqJh4edG3U02uSaUEERfMITyqrIbBVsfxx-VCeRvxAakQAG1hdHJpeC5vcmcvSU5NU0pPeGZKYmphdmF3cE1ISGxVYVlV > 2024-09-04 14:04:33 Anyway how can I check uuid of sda 2024-09-04 14:05:28 i think you want the PARTUUID, i.e. `blkid -s UUID -s PARTUUID /dev/sda1 2024-09-04 14:10:53 ACTION uploaded an image: (85KiB) < https://matrix.org/oftc/media/v1/media/download/AXen82Mrnudmnss3sOQ1DAf69sB_EgUd4NU8MN1_y5aTY30srqDnd19gk6wQh8zsSqW1AhifkPUvPluUjCPR2T9CeRvxctOAAG1hdHJpeC5vcmcvWHh0aVBKc3hMWkNiSlJMZmh4YVdvUEhD > 2024-09-04 14:10:54 "i think you want the PARTUUID, i..." <- Let me check 2024-09-04 14:12:18 Velo_Frost[m]: It is as of sda1 vfat in the above pic. Doesn't crosspond with efiboormgr boot0000 uuid 2024-09-04 14:13:01 that is the UUID again not the PARTUUID 2024-09-04 14:13:21 Hi folks ! I'm new to IRC so forgive me if I make netiquette mistakes... 2024-09-04 14:13:22 OK 2024-09-04 14:14:16 Going to unmount and rebooting. Plunging USB out before reboot 2024-09-04 14:15:19 welcome philfr! have fun 2024-09-04 14:16:38 I'm trying to use kernel-hooks to create a linux efi UKI image and copy it to the EFI partition. But the kernel hook runs before kmod and mkinitfs hooks. Before I was using the grub hooks, but I'd like to remove grub as it is useless with UKI. Can I do that ? 2024-09-04 14:19:04 ACTION uploaded an image: (44KiB) < https://matrix.org/oftc/media/v1/media/download/AZZ3cEytf9GRIU2qDJv7aetgaSqQHx1ptUQ0DtD_tkMc7LcUXyb1vO3kjhIFwsUf4BGc6twMc83yEK3jTqlXnYpCeRvx6smAAG1hdHJpeC5vcmcveWNsTG5rZ3lzTmpBaE1KbHpiWVdyeFJu > 2024-09-04 14:20:10 hey, that's an improvement 2024-09-04 14:20:41 ACTION uploaded an image: (82KiB) < https://matrix.org/oftc/media/v1/media/download/AXSkdfNyTA6RyVoiqn745UST8BAvaH_ZGHSSzErfIw6zDFQn4oQxCTvy_y2Q-tKXR4uVNN70OE_YcMsSLz1bkCpCeRvyAoxgAG1hdHJpeC5vcmcvYkpOTE11SW12RGhhdndsR3VVaWpTeWdU > 2024-09-04 14:21:00 First reboot after exiting chroot and unplug USB. It give me grub menu and boot into alpine installed on my machine 2024-09-04 14:21:19 matrix guys could add some kind of tinyURL :) 2024-09-04 14:22:24 ACTION uploaded an image: (107KiB) < https://matrix.org/oftc/media/v1/media/download/AXCW31H23cif4cRcEP_ZFLDX3qcsSTlhlhtgaiQW1541o_4Lll05g91ak5rE9hC8-V47sP8dxZTO6lSc3ULARMBCeRvyG5egAG1hdHJpeC5vcmcvSEFNT01tZmtpR0pCTnZjaXpNSHd6TWZh > 2024-09-04 14:22:39 Rebooting again to see if I get boot again or not 2024-09-04 14:23:05 And BIG NOOOOOOOOOOoooo 2024-09-04 14:23:32 ACTION uploaded an image: (49KiB) < https://matrix.org/oftc/media/v1/media/download/AZ8sUHzqvvGf4M3A4t-j-VvALlNQn72-cNHskZkt9d4kjyMOkd8Yj70C5dHbYQVvV8HVOI12q8drZRCn8Y7rUJJCeRvyLDSAAG1hdHJpeC5vcmcvZ1lhTnNXeFhlTm1mZGpycnpPTXNpRWZC > 2024-09-04 14:23:58 Time to say good buy to Uefi. Going for bios 2024-09-04 14:24:42 peerl3ss: ikke the problem is that it just doesn't create br0 interface but probably it's just a ifupdown-ng bug... I vaguely remember that it skips the bridge creation in some circunstances (reading directly on source) 2024-09-04 14:25:43 @_oftc_peerl3ss:matrix.org  I am so glad for your help and understanding me the commands. And all others who show calmness to see my post in the room. Once again thankyou everyone. 2024-09-04 14:27:13 oic donoban... i just use ifupdown 2024-09-04 14:27:34 "hey, that's an improvement" <- This thing work from day one. Problem is reboot and than boot to alpine 2024-09-04 14:27:41 hmm.. I think that I use the default 2024-09-04 14:28:09 ifupdown-ng 0.12.1 2024-09-04 14:28:14 -ng is the default for alpine yes 2024-09-04 14:28:20 ah ok 2024-09-04 14:30:27 philfr: it's been a while since i've made UKI but that's what EFISTUB is for, right? 2024-09-04 14:30:32 @_oftc_peerl3ss:matrix.org I have DM you, kindly 2024-09-04 14:33:33 invoked: yes. A single .efi file groups the UEFI stub, the kernel, the initramfs and command line options. The BIOS directly boots into linux without going through a bootloader. Works great and boots fast 2024-09-04 14:38:07 you should be able to uninstall grub just fine (no warranty) 2024-09-04 14:39:08 i mean, be prepared to boot off iso/whatever just in case, but if you have UKI working i don't see how it would break. 2024-09-04 14:40:53 I did remove grub, and it works fine, but when the kernel is updated, the hooks are called in a wrong order for me. 2024-09-04 14:42:12 sorry, misunderstood you. not sure i can help you with that, at least not at the moment. 2024-09-04 16:20:42 Hello dear all. I am looking to run althttpd through inetd as running it standalone doesn't support 80 and 443 simultaneously. I discover xinetd is dead, and then I discover inetd doesn't seem to be in busybox no more. apk search doesn't give me anything. Could you please tell me where to go? Thank you. 2024-09-04 16:27:48 PAM for polkit-1 adjustment works for fingerprint, but cant get cups pam file to work :(. so close lol. Anyone ever had this working? 2024-09-04 16:33:26 Also, why is a lot of what is contained in PAM config for SDDM not redundant? 2024-09-04 16:38:10 sunspot: xinetd ought to still work even if 'dead' 2024-09-04 17:23:31 it appears CUPS config needs a reference to root user for systemgroup, not just lpadmin 2024-09-04 17:57:55 Hi, after latest update (edge) my sway/wayland setup is borked 2024-09-04 17:58:02 firefox seem working, but foot (terminal) Thunar and fuzzel (launcher) are not working 2024-09-04 17:58:20 foot shows black window, Thunar fully transparent window 2024-09-04 17:58:48 swaybg doesn't show anything and swaybar works 2024-09-04 17:59:01 I suspect mesa 24.2.1 2024-09-04 17:59:07 my GPU is AMD rx580 2024-09-04 17:59:20 I'm not sure what should I report in issues 2024-09-04 17:59:21 Aelspire: yes there seems to be an issue (which i've confirmed myself on one machine) 2024-09-04 17:59:29 nothing suspicious in dmesg or /var/log/messages 2024-09-04 18:01:52 invoked: So this is known issue? 2024-09-04 18:03:00 mentioned a few times, we don't really know what's the cause of this yet 2024-09-04 18:12:14 hmm, I cannot downgrade to older mesa without breaking whole graphic shell 2024-09-04 18:12:31 yeah uh, older mesa packages probably won't be on the mirrors anymore 2024-09-04 18:12:31 is downgrade possible from cache? 2024-09-04 18:12:42 you can try `apk policy mesa` and see which ones are available 2024-09-04 18:14:21 Aelspire: i can get foot to work under weston with drm backend 2024-09-04 18:14:40 weston seems to be working ok but i just started messing around with it. 2024-09-04 18:14:49 to me this suggests a problem with wlroots and the new mesa 2024-09-04 18:14:59 but i'm not smart enough to have a real opinion. 2024-09-04 18:18:00 other things (besides foot) which didn't work/display properly after the mesa update are working under weston. eg, keepassxc. 2024-09-04 18:19:58 invoked: do you also have an amd gpu? 2024-09-04 18:20:09 i've seen xulfer mention amd as well 2024-09-04 18:20:31 yep 2024-09-04 18:22:19 damn, now i'm scared to upgrade mesa on my main pc 2024-09-04 18:23:34 i have two alpine daily drivers just for this occasion 2024-09-04 18:24:08 the one i'm typing to you on now is intel, though 2024-09-04 18:24:12 I've tried --force-broken-world and --force with installing apk from cache 2024-09-04 18:24:24 both options just purge half of system 2024-09-04 18:24:25 lol 2024-09-04 18:24:38 yeah, just sit tight or switch to 3.20 2024-09-04 18:29:44 weston just exited on its own, randomly... new mesa is interesting for sure. 2024-09-04 18:31:20 looking at https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31019, but i'm not quite sure if it's even related 2024-09-04 18:36:25 drm/mesa stuff is not in my wheelhouse at all, but debug output is not showing me anything interesting. 2024-09-04 18:38:35 anyway, back to the salt mines. ($dayjob) 2024-09-04 18:50:02 okay, can reproduce the same issue locally 2024-09-04 18:50:06 bye then 2024-09-04 18:50:12 had to manually launch a terminal with xwayland 2024-09-04 18:50:17 is it possible to install alpine linux with GPT/ Bios legacy and lvm on LUKS along side? 2024-09-04 18:51:07 has anyone tried it? 2024-09-04 18:53:56 I have that running on a laptop 2024-09-04 19:05:07 sigh 2024-09-04 19:05:17 tried recording the mesa graphical artifacts 2024-09-04 19:05:32 but the bitrate required to see anything is way too high 2024-09-04 20:11:37 just adding a datapoint on the new sway issue here. I can start sway, my alacritty terms run but the 2024-09-04 20:12:03 mouse cursor and top status bar/menu bars are all uninitialized data / garbled. 2024-09-04 20:13:23 on sway startup, I'm getting a (os error 32) from xdg-desktop-portal-WARNING no skeleton to export cast proxy message ... 2024-09-04 20:17:01 Activated service 'org.freedesktop.impl.portal.desktop.wlr' failed: Process org.freedesktop.impl.portal.desktop.wlr exited with status 1 2024-09-04 20:17:28 that is the only error I saw in /var/log/messages, will drop to vtty and see what's there next 2024-09-04 20:22:54 fwiw, currently trying to build latest mesa commit from main 2024-09-04 20:24:06 ok. I'm operational enough and this is on my "edge" workstation. just wanted to report it (i am on 24.2.1-r0) it appears 2024-09-04 21:15:16 i've put one of my older Mesa builds on https://dev.alpinelinux.org/~ptrc/mesa-24.1.2-r1.tar.gz 2024-09-04 21:15:22 i have no idea which flags it was built with, etc. 2024-09-04 21:15:31 but it works, so 2024-09-04 21:19:14 Well the main things that changed (that I saw) was the drop of swrast, and the inclusion of llvmpipe in the gallium drivers in the mesa package. I did try just popping swrast back in but I didn't have time to see if I did it properly. 2024-09-04 21:20:54 i tried that as well 2024-09-04 21:21:24 rebuilding 24.2.1 with adding softpipe back and generally bringing the amd64 options back to how they were 2024-09-04 22:01:42 hm how did you force those mesa packages in? I'm not as quick on the docs in just console. 2024-09-04 22:09:09 I have a stupid question. 2024-09-04 22:09:38 Will the busybox DD on the ISO tell me when it's done (I'm using it to write random garbage to my SSD) 2024-09-04 22:09:56 you need to clarify what you mean by "tell" 2024-09-04 22:10:02 'dd bs=1M if=/dev/random of=/dev/sda' 2024-09-04 22:10:02 It'll tell you there's no more space 2024-09-04 22:10:10 will it drop me back at the prompt 2024-09-04 22:10:13 yes 2024-09-04 22:10:19 okay 2024-09-04 22:10:21 thx 2024-09-04 22:10:35 dd'ing faster than anybody 2024-09-04 22:35:00 xulfer: `apk add mesa*.apk` should install them directly 2024-09-04 22:35:28 or well, if you built it yourself, you can also just do `apk add 'mesa<24.2'` 2024-09-04 22:37:22 Hi ptrc 2024-09-04 22:37:26 hiya 2024-09-04 22:42:08 Whats the best method to use steam on alpine, i dont like the flatpak 2024-09-04 22:42:38 you're not going to like the answer then 2024-09-04 22:42:43 yeah.. 2024-09-04 22:42:50 i mean, flatpak is the most convenient 2024-09-04 22:43:27 i personally have a rube goldberg machine of bwrap scripts running in rootful Xwayland and an Ubuntu chroot 2024-09-04 22:43:34 I used distrobox for a little while but it always ended up breaking on me 2024-09-04 22:43:36 but it breaks once in a while 2024-09-04 22:43:53 Yeah same with distrobox 2024-09-04 22:57:22 ptrc: your files worked for me, btw 2024-09-04 22:58:08 On https://wiki.alpinelinux.org/wiki/NetworkManager where it says to put 2024-09-04 22:58:17 [main] dhcp=internal plugins=ifupdown,keyfile [ifupdown] managed=true [device] wifi.scan-rand-mac-address=yes wifi.backend=wpa_supplicant 2024-09-04 22:58:33 into /etc/NetworkManager/NetworkManager.conf' 2024-09-04 22:58:52 should I replace [device] with [wlan0]? 2024-09-04 23:18:44 Yep downgrading works. Thanks ptrc! 2024-09-04 23:22:08 I got NetworkManager all set up but now I can only configure connections as root. I added myself to plugdev and netdev and still no dice 2024-09-05 08:51:22 After I downgrade to mesa 24.1.5 (build locally at commit 2ace92581be6f4ae2ba47921cf71feeb56845f58), most things work fine. But grim and wf-recorder only capture complete black images and videos, have no error logs. 2024-09-05 08:51:24 And a flatpak qt5 application (com.tencent.WeChat) also displays a black window. 2024-09-05 08:58:21 Not sure if it's just for me. I am using the latest commit of wlroots and sway. 2024-09-05 09:00:27 good morning, what to do whenever suspend isn't working under non-root user? (loginctl suspend) 2024-09-05 09:14:40 tl;dr session starts screen locker, puts devices like wifi card into standby mode, but the PC by itself isn't entering suspend mode. Only under root is possible. 2024-09-05 11:03:05 the issue happens only on wayland. Whenever I use x11 as display server, it can suspend normally. 2024-09-05 11:07:41 do you mean you run wayland as root or do you mean you run loginctl as root? 2024-09-05 11:16:01 I run KDE with wayland as non-root and loginctl hangs at suspend/hibernation. 2024-09-05 11:16:55 is nvidia involved? 2024-09-05 11:18:35 try directly 2024-09-05 11:19:03 My GPU is a builtin one from AMD A8-6500 2024-09-05 11:19:03 echo mem > /sys/power/state 2024-09-05 11:19:31 execute the command as root? 2024-09-05 11:20:03 Not sure if non root can do it 2024-09-05 11:20:53 has to be as root, yes 2024-09-05 11:24:10 executing the command works 2024-09-05 11:26:50 executing loginctl suspend first, then the echo command causes the Resource busy error for that sys path 2024-09-05 11:28:30 do you have anything new in dmesg while loginctl is trying to make the suspend call? 2024-09-05 11:33:26 informations 2024-09-05 11:33:53 an information that wifi disconnected as well as bluetoith devices 2024-09-05 11:34:24 as the end the PM: suspend entry (deep) and shortly after suspend exit 2024-09-05 11:37:05 in messages file an information from elogind that it tried to recreate sddm session but failed 2024-09-05 11:37:44 sddm runs on x11 while KDE on wayland 2024-09-05 13:37:44 Is it recommended to install rust toolchains and tools through apk? 2024-09-05 13:39:11 or are these packages related to kernel dev? 2024-09-05 13:45:19 No, they are not dedicated for kernel dev, you can use them normally 2024-09-05 13:45:29 All rust packages in aports are built with the packaged rust compiler 2024-09-05 13:45:42 there seems to be cranelift, gcc, and llvm related compiler files which is cool. I dk if all that comes in with normal script install 2024-09-05 13:46:27 oh sweet. Will the rust package bring in analyzer, rustup, cargo? Or should I install those too 2024-09-05 13:46:50 also, the apk fonts made my morningggg haha 2024-09-05 13:47:15 usually I have a way bigger pain grabbing nerd fonts haha 2024-09-05 13:47:18 cargo is a separate package 2024-09-05 13:48:52 Thanks. Do you know if rust will bring in most of this:... (full message at ) 2024-09-05 13:49:01 (minus the completions) 2024-09-05 13:49:54 I don't yet understand if "meta packages" are a thing in apl 2024-09-05 13:49:57 *apk 2024-09-05 13:50:24 so far it seems that it is not, but I am still naive / ignorant here 2024-09-05 13:50:26 There are metapackages, but not for everything 2024-09-05 13:50:58 That last multi-line message resulted in a 404 link on irc fyi 2024-09-05 13:51:53 Ah sorry, it was just a copied output from apk search rust which showed clippy, analyzer, bindgen, wasm, etc 2024-09-05 13:52:25 Rust doesn't seem like a meta package, but there is so much in it that it may be 2024-09-05 13:59:23 Basically is this necessary? (and similar for cargo where there are many cargo-* packages too): 2024-09-05 13:59:24 apk add rust rust-analyzer rust-analyzer-doc rust-bindgen rust-clippy rust-dbg rust-doc rust-gdb rust-lldb rust-src rust-wasm rustfmt rustup rustup-bash-completion rustup-fish-completion rustup-zsh-completion 2024-09-05 14:00:33 It depends what you need to do. you can get quite far with just rust and cargo 2024-09-05 14:00:48 The completion packages are installed automatically if you have the shell installed 2024-09-05 14:01:32 Ah, I guess if it is a metapackage it should state "Meta package for build base" in Description? 2024-09-05 14:02:49 That is good to know about completion packages. I work with wasm, embedded, and standard targets. 2024-09-05 14:03:38 Note that I'm not a rust developer, so I don't know what you need for typical development 2024-09-05 14:03:59 but things like rust-dbg is only necessary if you try to debug the compiler itself 2024-09-05 14:04:06 That's fine. Mainly just trying to understand how to delineate a metapackage 2024-09-05 14:04:21 If you want to know for sure, check the contents 2024-09-05 14:04:24 https://pkgs.alpinelinux.org/contents?branch=edge&name=build%2dbase&arch=riscv64&repo=main 2024-09-05 14:04:25 ya, assyming gdb may be similar 2024-09-05 14:04:35 You see that this package has no files 2024-09-05 14:04:59 for a metapackage? 2024-09-05 14:05:01 yes 2024-09-05 14:05:11 but there is afaik no way to search for packages without files 2024-09-05 14:05:12 perfect, that is great to know. Thank you 2024-09-05 14:05:34 build-base is a metapackage, and it had that in description, which is why I got confused 2024-09-05 14:06:14 no contents though, so will figure out how to check contents on cli. Thanks so much 2024-09-05 14:06:42 apk info --contents , but that requires the package to be installed 2024-09-05 14:07:12 apk --help 2024-09-05 14:07:19 oops lol 2024-09-05 14:11:21 apk search -d woulda been nice if packages aligned on metapackage description 2024-09-05 14:12:42 do you think apk search --has-origin rust could indicate a metapackage? 2024-09-05 14:15:03 No, those indicate all the subpackages for rust, but they will not necessarily automatically be installed (the whole idea of subpackages is to be able to not install everything if you don't need it) 2024-09-05 14:15:43 good to know, thank you 2024-09-05 14:16:11 Does PostmarketOS use APK? 2024-09-05 14:16:45 search seems yes, cool 2024-09-05 14:18:37 yes, they do 2024-09-05 14:36:23 is there some backups of alpine linux forum I can read online? I dont consider reddit to be a valid source of information 2024-09-05 14:37:16 what alpine linux forum? 2024-09-05 14:38:59 wasn't there an official forum subdomain for this project and it got shutdown recently? 2024-09-05 14:40:03 recently is more then 5 years ago 2024-09-05 14:40:24 https://dev.alpinelinux.org/~clandmeter/other/forum.alpinelinux.org/ 2024-09-05 14:42:48 yeah I was able to find that link I was curious whether there isn't something else 2024-09-05 14:46:58 did anybody get qsv working with ffmpeg? im getting `Failed to set value 'qsv=qs@va' for option 'init_hw_device': Unknown error occurred` with ffmpeg (jellyfin-ffmpeg specifically) 2024-09-05 14:48:13 joseph_ii: there is nothing else. We don't want to set something up that ends up being a waste land 2024-09-05 15:15:51 ptrc: saw your bug on fdo. thanks for that. for more context i'd also confirm i've not been able to repro the problem on non-amd 2024-09-05 15:17:33 ptrc: but it's important to note i was able to get things working on weston... 2024-09-05 15:17:59 oh, do you wanna add that as a comment? 2024-09-05 15:19:19 sure, maybe. once i understand that i'm not doing something stupid. i used the drm-backend on weston 2024-09-05 15:19:34 there is a wiki 2024-09-05 15:19:49 https://wiki.alpinelinux.org/wiki/Main_Page 2024-09-05 15:20:07 with an account there seem to be no restrictions to edit 2024-09-05 15:20:24 so you can update info as things become more clear 2024-09-05 15:20:41 there is also the gitlab account 2024-09-05 15:21:00 https://gitlab.alpinelinux.org/ 2024-09-05 15:22:11 i dk how much issues are used, but worth a peak 2024-09-05 15:23:41 ptrc: unless i'm using weston wrong, it suggests a regression with mesa but with wlroots specifically. it'd probably be a good idea if someone else with amdgpu tested the broke things with weston so that i know i'm not trippin'. 2024-09-05 15:23:46 The wiki edit syntax isn't quite markdown which sucks, but if you edit a page it is easy to see 2024-09-05 15:24:11 ohh, hm 2024-09-05 15:24:13 I have amdgpu on my other machine, but I haven't tested alpine on that yet 2024-09-05 15:24:32 i'm gonna update my laptop to latest mesa ( also amdgpu ) and check something like gnome or kde 2024-09-05 15:24:38 required pulling in firmware from linux kernel and updating intiramfs on deb 2024-09-05 15:24:45 that works too 2024-09-05 15:24:49 but I didn't experience mesa issues 2024-09-05 15:24:57 what kernel are you on? 2024-09-05 15:25:25 andar1an[m]: edge, lts 2024-09-05 15:25:26 I think alpine uses a newer kernel, which seemed to not require that change for deb 2024-09-05 15:25:44 flatpak may also solve mesa issues for some apps 2024-09-05 15:26:34 I wish I could test this now haha, sorry I can't yet. Making sure laptop is set up right before getting alpine on server and messing with btrfs raid 2024-09-05 15:27:25 I have a 7800xt, and that was problematic on release, but hasn't been bad these days 2024-09-05 15:31:00 andar1an[m]: this is a new bug specific for amd cards triggered by the mesa 24.2.1 update. likely has nothing to do with the kernel 2024-09-05 15:31:27 I have been seeing that bug, all I saw as a solution was to downgrade mesa 2024-09-05 15:31:31 or use flatpak 2024-09-05 15:32:04 I wonder how pinning packages works in alpine 2024-09-05 15:32:51 seems pretty straight forward to set "sticky" versions: https://superuser.com/questions/1055060/how-to-install-a-specific-package-version-in-alpine 2024-09-05 15:33:32 dk how one may propagate that for a tree 2024-09-05 15:34:07 maybe with --has-origin or -r flag you can get a list of what relies on mesa and pin those too 2024-09-05 15:35:20 why I love rust cargo.toml haha 2024-09-05 15:38:47 ptrc: one last (important) detail. weston was able to run foot/etc fine, but weston itself would SIGABRT randomly after 5 or so minutes. hence why i'm not sure about it. 2024-09-05 15:39:17 running gnome or kde would probably be a good test. 2024-09-05 15:42:09 what is weston? 2024-09-05 15:42:27 wayland's reference compositor 2024-09-05 15:42:31 ah wayland comp 2024-09-05 15:42:36 cool 2024-09-05 15:43:21 Steam was having issues with wayland for steam link, would be really curious to see what deal is when I get alpine on server 2024-09-05 15:43:37 but that was for video, mesa is audio right? 2024-09-05 15:43:57 oh nvm, I misremembered 2024-09-05 15:44:45 invoked: I am seeing broken mouse cursor and preview pop-ups completely garbabled since the update to the latest mesa on edge FWIW. my system is intel cpu and amd GPU 2024-09-05 15:45:39 yeah. that's consistent with what i think some have seen with sway. (i run river, not sway, but both have wlroots in common) 2024-09-05 15:46:00 and I use sway 2024-09-05 15:46:15 ack, roger that, thanks 2024-09-05 15:46:25 in river's case i just get black windows. nothing garbled 2024-09-05 15:46:33 for the affected programs 2024-09-05 15:47:06 the bar and launcher aren't visible at all. 2024-09-05 15:47:53 mine are visible, but they contain multi-color (looks like uninitialized data) so lots of multi-colored pixels 2024-09-05 15:47:57 I got that on older mesa versions with remote play - likely unrelated, but that is issue I referred to 2024-09-05 15:49:21 some component that is in common with how the mouse cursor, the little firefox tab hover-preview-popups, and the "bar" are all problematic on my system. 2024-09-05 15:49:57 https://github.com/ValveSoftware/steam-for-linux/issues/6148 2024-09-05 15:50:43 This reflects a lot of what you are metioning, both black screen and artifacts 2024-09-05 15:50:53 i tried as a test to use @ptrc's built images but I am using LLVM18 tool chain so I couldn't install them without mucking further with my system 2024-09-05 15:51:56 yeah. i would have just switched to 3.20 but... river things... 2024-09-05 15:51:58 i have to dust off my aports setup to see about trying some of my own packages eventuall after $dayjob 2024-09-05 16:06:38 anyone knows if is it possible to run pipewire just for audio without dbus and any graphics? 2024-09-05 16:09:18 ikke could help 2024-09-05 16:09:21 :-D 2024-09-05 16:09:34 With what 2024-09-05 16:09:37 donoban: i thought pipewire needs dbus, but https://wiki.alpinelinux.org/wiki/PipeWire says otherwise 2024-09-05 16:09:45 oh, he is probably busy with more important things :) 2024-09-05 16:09:54 it depends. 2024-09-05 16:10:04 I have no experience with pipewire 2024-09-05 16:10:20 long time no see sire, i hope things are good 2024-09-05 16:10:36 donoban: i'm running pipewire without graphics, but with a dbus user bus 2024-09-05 16:10:40 and it works just fine 2024-09-05 16:10:46 uhmm, thanks invoked I'm gonna try editing config 2024-09-05 16:11:05 ptrc: how do you start dbus? 2024-09-05 16:11:47 kinda cursed setup with dinit and pam-rundir 2024-09-05 16:12:12 can I don something like, dbus-run-session ..¿ash? 2024-09-05 16:12:34 ash is a terminal, which exits if there is not a tty attach, so, no 2024-09-05 16:12:42 uhmm 2024-09-05 16:12:46 s/terminal/shell 2024-09-05 16:12:48 look 2024-09-05 16:12:58 You need a graphical terminal at least 2024-09-05 16:13:00 https://tpaste.us/xyPJ 2024-09-05 16:14:02 uhm.. 2024-09-05 16:14:05 this is weird 2024-09-05 16:14:10 /home/donoban # pipewire 2024-09-05 16:14:11 [W][06110.528582] mod.jackdbus-detect | [module-jackdbus-: 216 on_is_started_received()] Failed to receive jackdbus reply: org.freedesktop.DBus.Error.ServiceUnknown: The name org.jackaudio.service was not provided by any .service files 2024-09-05 16:14:13 let's see if works.. 2024-09-05 16:14:41 i'm not ptrc but https://okturing.com/src/20614/body is how i do it. 2024-09-05 16:15:00 wow, it's working 2024-09-05 16:15:20 my problem is runing alsa directly it starts to beeeeeeeeeeeep after some seconds 2024-09-05 16:15:35 probably something kernel related but let's see if pipewire resists 2024-09-05 16:16:01 thanks invoked 2024-09-05 16:16:40 beh 2024-09-05 16:16:43 beeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeep 2024-09-05 16:16:46 :) 2024-09-05 16:21:42 ikke: hope you are doing good, dont know if you remember me 2024-09-05 16:22:05 I do, namaste 2024-09-05 16:23:44 great! namaste namaste 2024-09-05 16:23:47 :-P 2024-09-05 16:39:27 Is it possible to test modifications to setup-*.in scripts easily? 2024-09-05 16:40:49 what may that process look like? 2024-09-05 16:41:34 https://gitlab.alpinelinux.org/alpine/alpine-conf/-/tree/master/tests 2024-09-05 16:41:43 Ya, I have that 2024-09-05 16:41:57 I am just not sure how the process may work 2024-09-05 16:42:20 Do you mean how to run those tests, or? 2024-09-05 16:42:35 No, I mean if I want to try to update setup-disks for example 2024-09-05 16:42:45 how could I test those changes on my own machine 2024-09-05 16:43:33 would it involve replacing specific file in installation media? 2024-09-05 16:44:23 Oh, you want to actually test it in place, not just using the test suite 2024-09-05 16:44:52 Yes, correct. I can look into tests, but would rather iterate if possible 2024-09-05 16:45:47 I think I am in a good place with daily driver setup, the last thing I want to investigate on laptop before moving to server it btrfs subvols during installation 2024-09-05 16:49:49 on an exciting note, literally lost 0 functionality which is really nice, except for my printer driver isn't available, driverless print doesn't seem to work, and gutenprint doesn't support this model yet, but maybe by time it is in stable it will. Printer adds just fine though 2024-09-05 16:55:17 It's a magical feeling when init runs and it isn't a million services haha 2024-09-05 17:51:02 I can reproduce the mesa regression with sway and wl_shm-based wayland clients 2024-09-05 17:51:21 cc ncopa (mesa maintainer) 2024-09-05 17:54:49 ddevault: https://gitlab.freedesktop.org/mesa/mesa/-/issues/11846 if you'd like to add something there 2024-09-05 17:55:15 thanks 2024-09-05 18:56:16 ddevault: what kernel version? I wonder if it is a regression in the kernel or in mesa 2024-09-05 18:57:33 ncopa: i see it on 6.6.49-0-lts 2024-09-05 18:58:35 has anyone seen it on 6.6.48 2024-09-05 18:58:39 or less? 2024-09-05 19:17:05 i'm using the old/ptrc mesa on 6.6.49. by that logic the kernel should be fine, correct? 2024-09-05 19:17:30 that's a good point actually 2024-09-05 19:17:44 invoked: no, 6.6.49 could be the kernel that introduced the regression 2024-09-05 19:18:32 ACTION quits computers 2024-09-05 19:19:48 but again i would just like to mention... i worked around it using weston. 2024-09-05 19:21:49 reproduced it on x86_64 6.6.47-0-lts 2024-09-05 19:21:57 so probably not related to the kernel 2024-09-05 19:23:06 actually on 6.6.47 it's even weirder 2024-09-05 19:23:20 sway? 2024-09-05 19:23:23 yeah 2024-09-05 19:23:26 some things are broken 2024-09-05 19:23:28 but not all things 2024-09-05 19:23:32 and not in the same way 2024-09-05 19:23:53 the cursor is gone, swaybg is gone, but glxgears doesn't break, and foot only sometimes renders a blank window.. 2024-09-05 19:25:35 did you test kde or anything by chance 2024-09-05 19:25:44 anyone affected with integrated amd graphics? 2024-09-05 19:27:38 not me donoban 2024-09-05 19:28:00 Thank you Ikke for recommending the tests dir, I have never seen kyua and atf-sh before 2024-09-05 19:28:02 ACTION tries to determine his GPU model \_/ 2024-09-05 19:28:17 however, glxgears runs smooth and no artifacts for me at 120fps with no artifacts 2024-09-05 19:28:31 glxinfo 2024-09-05 19:29:42 donoban: it's broken for me on a laptop with Ryzen 7 4750U 2024-09-05 19:29:43 so yeah 2024-09-05 19:29:45 Device: AMD Radeon 780M (radeonsi, gfx1103_r1, LLVM 18.1.8, DRM 3.54, 6.6.48-0-lts) (0x15bf) 2024-09-05 19:30:20 one thing I have wondered about is where the id's come from in the setup-disk.in script, or the tests. 2024-09-05 19:30:21 e.g. biosboot) id=21686148-6449-6E6F-744E-656564454649 ;; 2024-09-05 19:30:21 I thought it may be a uuid at first for a partition, but I don't think so based on the script 2024-09-05 19:30:27 oh, I'm on safe kernel 2024-09-05 19:32:30 I guess, trying to understand why the uuid if it is one is static 2024-09-05 19:34:31 invoked: how did you actually launch weston? it doesn't even start for me from drm, with weston-backend-drm imstalled 2024-09-05 19:39:09 weston --backend=drm --debug --log=weston.log 2024-09-05 19:41:03 hm, it iterates over input devices, then doesn't do anything.. 2024-09-05 19:41:14 i have weston,weston-backend-drm,weston-backend-wayland,weston-clients,weston-shell-desktop,weston-terminal,weston-xwayland installed 2024-09-05 19:41:37 you have to give it a minute after launching. it's weird. i'm probably using it wrong. 2024-09-05 19:42:20 you'll know it's up when the little shell icon is in the upper left corner. click on that, it opens weston-terminal 2024-09-05 19:42:56 i'm a chimpanzee with a keyboard 2024-09-05 19:44:11 let me know if it SIGABRTs on you 2024-09-05 19:44:12 to me it sounds like you're the one using it correctly, because for me it just exits back into tty, not even a screen blink or anything 2024-09-05 19:44:31 ah 2024-09-05 19:44:33 it segfaults 2024-09-05 19:44:33 lol 2024-09-05 19:45:15 i didn't read the manual. is there a manual? not gonna read it. 2024-09-05 19:46:04 i should probably retest it to see if it still aborts with the old mesa, though 2024-09-05 19:46:19 ACTION goes to do that 2024-09-05 19:47:49 oh interesting. it starts right up now, and it has a background wallpaper it didn't have before 2024-09-05 19:48:14 so clearly something is borked with mesa. 2024-09-05 19:55:48 this informs me that if i upgrade mesa again, testing kde/gnome would reveal more breakage there also 2024-09-05 19:56:14 it's not some idiosyncratic thing with wlroots. 2024-09-05 20:06:23 busybox `sed -n l` doesnt work? 2024-09-05 20:41:57 Hey there! I wanted to let you know that I have a Telegram channel where I share some amazing Verified sauce and soft cashout... (full message at ) 2024-09-05 21:13:26 what text editor do you use on alpine? codium flatpak is annoying me 2024-09-05 21:13:52 andar1an[m]: code-oss is in alpine, and works quite okay 2024-09-05 21:14:05 thanks, will look 2024-09-05 21:36:12 code-oss is in testing, but there are also host-spawn and flatpak spawn configuration lines one can add to access system shells and hopefully path 2024-09-05 21:37:15 hopefully it lets me find rust-analyzer and direnv lol 2024-09-05 21:41:40 at least can edit the .desktop start command with an env flag 2024-09-05 21:43:30 emacs and vim work too 2024-09-05 21:57:51 I came from mech eng lol. I need some kind of gui haha 2024-09-05 21:57:58 We like pictures 2024-09-05 21:58:37 Helix is also on here though, and so is lapce 2024-09-05 21:59:13 tested those and they work well 2024-09-05 21:59:28 ack, i came from the cdc and wyse terminal era so I don't like to use my mouse unless I have to :-) 2024-09-05 21:59:33 I love when it's available through apk, so much faster and easier 2024-09-05 21:59:41 haha nice 2024-09-05 21:59:52 I wish I could be like that 2024-09-05 21:59:55 OG 2024-09-05 22:00:10 however, i used to sysadmin about 50 SGI computers in the early '90s that all ran ProEngineer 2024-09-05 22:00:24 which you might know a thing or two about 2024-09-05 22:00:39 wait, nopjmp, are you the nopjmp i knew or just a random person with the same username? lol 2024-09-05 22:00:55 I never used pro engineer. NX, Catia, Solidworks, autodesk, but never pro engineer haha 2024-09-05 22:01:03 did you hear good things? 2024-09-05 22:01:25 I lie, it is PTC. I used Creo 2024-09-05 22:01:29 the aerospace and mech-e labs used it for years, I just needed to manage the license server and install the irix boxes 2024-09-05 22:01:56 My fav was NX. But recently I like following FreeCAD and a friend showed me Leap71 2024-09-05 22:02:12 they have a pecogk geometry kernel in the works which is neat 2024-09-05 22:02:24 my life is now complete as in 1990 i supported proengineer and now I have like 5 solidworks users but they are all on windows 2024-09-05 22:03:02 I used to teach and sell swx haha. Nice 2024-09-05 22:03:20 Ya, CAD loves Windows. Starting to see change though 2024-09-05 22:03:46 FEA had abaqus for a while on linux, but I think you can get many FEA suites on linux now 2024-09-05 22:05:43 have you seen the https://zoo.dev ai startup yet? hope they make some headway 2024-09-05 22:05:57 Yes, they will be at rust.conf 2024-09-05 22:06:07 the challenge with AI in CAD space though, it it is often redundant 2024-09-05 22:06:24 you can easily leverage part families and component libraries already 2024-09-05 22:06:53 better for optimization and surfacing work I think. Easy to generate standard components as is 2024-09-05 22:07:02 i've seen people doing cad in blender 2024-09-05 22:07:27 Engineering CAD or animation and toys? 2024-09-05 22:08:07 engineering stuff... assemblies, cnc models, whatever 2024-09-05 22:08:39 wow, I have tried blender, but I never saw that? Do they have any certs? 2024-09-05 22:09:05 no i doubt it, i think people are just using it for fun 2024-09-05 22:09:47 I wonder what that assembly performance was like haha 2024-09-05 22:10:04 0 clue how they did CNC work though haha 2024-09-05 22:10:21 plugin? 2024-09-05 22:10:45 it's not my area (its been decades since i've done anything except view) 2024-09-05 22:10:47 Blender-CAM. That's awesome hahaha 2024-09-05 22:10:57 (it is a plugin) 2024-09-05 22:11:14 but you know, game engines are extremely powerful so maybe this is a sign of how things are changing 2024-09-05 22:11:39 anyway this is kind of #alpine-offtopic 2024-09-05 22:11:45 True, but I am more interested in what I saw from Leap71 2024-09-05 22:11:57 with NX my favorite way to model was composition based on primatives 2024-09-05 22:12:03 Sorry, I will jump there 2024-09-05 22:12:19 no worries, i'm not a cop 2024-09-05 22:12:47 (that's exactly what a cop would say) 2024-09-05 22:12:53 Is there an alpine off topic? 2024-09-05 22:13:10 I can't get from what you shared 2024-09-05 22:13:22 yep, #alpine-offtopic 2024-09-05 22:13:32 but it's probably not bridged to matrix 2024-09-05 22:13:50 Ah. I saw an IRC client in flatpak 2024-09-05 22:13:53 can give that a go 2024-09-05 22:14:02 Never really used IRC 2024-09-05 22:14:08 i don't know how matrix works. i'm boycotting my brain from all of ... that 2024-09-05 22:16:12 I'm working towards baremetal orchestration over ipv6 only networks. I love that stuff haha 2024-09-06 01:32:42 FYI: i built my own mesa apks from aports and downgraded to 24.1.2 and my sway issues are resolved, I didn't try 24.1.5; will stay here until we know more 2024-09-06 03:12:20 we have an issue on ipv6-only networks, which we would like to discuss with alpine folks: https://mailarchive.ietf.org/arch/msg/v6ops/0CGYcHYFCi-qtodX3ApsQjZZVio/ (also brought this up in the musl irc, but the solution we'd push for involves creating an "/etc/pref64.conf" and that wouldn't be managed by the libc but by a system daemon, so it ends up being relevant here too) 2024-09-06 07:35:41 ikke, how to setup loginctl on alpine linux? i read this and still failed: https://wiki.alpinelinux.org/wiki/Elogind 2024-09-06 07:36:11 andar1an[m], wow, "Never really used IRC" 2024-09-06 07:36:44 i remember when i was 15 and trying to understand what is IRC.. what a time 2024-09-06 08:32:19 How can I get connect to my WiFi fi router and have a static IP for my alpine linux 2024-09-06 08:35:54 anyone knows how could I reproduce this on linux? https://github.com/JollyFrogs/Backlighter/blob/eff9910859ed0fc20c47c5e4db5a0419308b4311/Program.cs#L73 I first tried with py3-usb but ended with resource busy errors and yesterday started to look at solaar source code and I'm not sure if I should focus on the HID layer 2024-09-06 08:36:38 my humble purpose is that the keyboard stops backlight automically during day hours and turn on on night :) 2024-09-06 08:57:15 you don't just "use" IRC.. it uses you 2024-09-06 08:59:27 Velo_Frost[m]: there are many options for that, probably the best is just setup your fixed ip on the router DCHP config 2024-09-06 09:44:06 "Velo_Frost: there are many..." <- Any other method to do on my machine rather than router? 2024-09-06 09:46:31 yes, how do you handle the network interface? did you use setup-interfaces? 2024-09-06 09:47:37 check your /etc/network/interfaces and change dhcp with static, then add "address xx.xx.xx.xx" and "netmask 255.255.255.0" 2024-09-06 09:51:20 With ifupdown-ng, you can combine those 2024-09-06 09:51:47 address w.x.y.z/24 2024-09-06 10:07:32 "check your /etc/network/..." <- I have to add gateway as well? 2024-09-06 10:08:20 I mean, unless you don't want internet access 2024-09-06 10:15:09 iface wlan0 inet static... (full message at ) 2024-09-06 10:15:23 Is it correct? 2024-09-06 10:16:08 After editing these values I get inet addr as local host. 127.0.0.1 2024-09-06 10:16:12 Weird 2024-09-06 10:16:36 you're probably looking at the wrong interface 2024-09-06 10:17:17 and you have to restart ifupdown after making changes 2024-09-06 10:21:05 ifup -f $iface 2024-09-06 10:28:03 "you're probably looking at the..." <- Interface is correct. I just add # before auto wlan0 and if wlan0 inet dhcp. Than add the above in down line 2024-09-06 10:28:31 Than #service networking restart 2024-09-06 10:30:54 "ifup -f $iface" <- This is not executing 2024-09-06 10:31:54 pastebin the output of 'ip addr show' 2024-09-06 10:35:02 ACTION uploaded an image: (300KiB) < https://matrix.org/oftc/media/v1/media/download/AR0VKtCbdZVc9iZhBWqtzKfFY2SRoahfUuvRb9nbhGksx3_8yVEeyy1hra5B7ZFBgdC_dpdo8W0J9mx_WSfbgMBCeRyJ5IXwAG1hdHJpeC5vcmcvaVRjYm5HTnhha1ZIcXlYWmFwTGhjV2xP > 2024-09-06 10:35:39 yeah, that's literally looking at 'lo' 2024-09-06 10:35:46 not wlan0 2024-09-06 10:36:12 Than how will I set to look at wlan0 2024-09-06 10:36:51 well, step one is stop using ifconfig; use ip commands instead; like 'ip addr show' 2024-09-06 10:37:05 Ok 2024-09-06 10:37:41 does 'ip addr show' show a wlan0? 2024-09-06 10:37:55 Yes showin 2024-09-06 10:38:18 ACTION uploaded an image: (83KiB) < https://matrix.org/oftc/media/v1/media/download/AUNeuMrNFGlWwplq-ErmCxcoDBNMKcg_fQDaZJDxIo6VjnxydwkcGHNlNRBLA8yXFmqdNcaSmt3Ojyh5TN5Qy8tCeRyKFH9gAG1hdHJpeC5vcmcvRmxCV0xRRWRGS1Nyc3lRTHhzcHdPa3VG > 2024-09-06 10:38:47 that says you're probably not actually connected to your wifi 2024-09-06 10:40:27 dwfreed: Its when I change the settings. If I change to auto wlan0. Than I get the IP, otherwise static IP config make it down 2024-09-06 10:41:47 OK done. I remove the I face wlan0 inet dhcp line completly 2024-09-06 10:41:57 And now getting the static ip 2024-09-06 10:42:29 Previously I just put #iface wlan0 inet dhcp 2024-09-06 10:47:13 on edge, anyone else having trouble with the default graphical file chooser? specifically in telegram? 2024-09-06 10:47:31 i get segment fault as soon as i call the file chooser 2024-09-06 10:47:55 btw, wish one could replace the file chooser.. 2024-09-06 10:48:48 Velo_Frost[m]: how looks your current interfaces? 2024-09-06 10:50:19 IP addr show me the static IP address I have configured 2024-09-06 10:51:05 it _could_ be my setup, where i use old config files... 2024-09-06 10:55:59 I mean the /etc/network/interfaces file 2024-09-06 11:08:37 ACTION uploaded an image: (74KiB) < https://matrix.org/oftc/media/v1/media/download/Afo8ExF9a-lwSLl19fje23KQx_ss62QTKC1rpBkpmG7NYazTCmpfptixvesELAg3QcIcMHtKKHwbGLSwutZrQl9CeRyL0IbAAG1hdHJpeC5vcmcvcnFTVmJueGJlelVzcGZtWFR3R29XVVB1 > 2024-09-06 11:42:31 "I mean the /etc/network/..." <- they are just fine. I followed the script above 2024-09-06 11:42:48 I have now a static IP thankyou 2024-09-06 11:45:00 now I want to install docket-rootless. still confused about installing apk. Alpine linux docker wiki has 3 packages, I mean docker, docker-rootless-extras and docker-cli-compose. I think I need only one out of these? 2024-09-06 11:45:20 any Idea, who have installed rootless-docker?? 2024-09-06 11:56:43 no idea, always used std docker 2024-09-06 12:00:03 all right. how can one reset Alpine to initial settings 2024-09-06 12:00:10 like we do reset in Android? 2024-09-06 12:00:56 reinstall 2024-09-06 12:08:35 damn long process hahaha 2024-09-06 12:09:08 need a good guide to install rootless-docker on alpine Linux. other than its own wiki 2024-09-06 12:10:40 a container, or OS-level virtualisation is faster than reinstall 2024-09-06 12:11:03 (sandbox, jail, namespace, kernel slice etc) 2024-09-06 12:12:54 https://en.wikipedia.org/wiki/OS-level_virtualization?useskin=vector#Implementations 2024-09-06 12:17:43 the other option is mtree(8) comparison with the system specification (directory/file lists sync) for a system cleam state restoration 2024-09-06 12:18:24 you can also use snapshots in file systems if you have that option/capability, but it's low assurance / validity typically (unclean slice in time, non-system atomic) 2024-09-06 12:49:15 if you run raspberry diskless can one umount the sd card after booting from it? 2024-09-06 12:50:52 if you run diskless, is it even mounted? 2024-09-06 12:51:43 been a while since i messed with that stuff, but you have to have something to boot from, usually the sd card on raspberry, then i guess it loads the system into RAM and run from there 2024-09-06 13:07:30 You probably need to stop the modloop service 2024-09-06 13:19:09 does Alpine need both docker, docker-rootless-extral packages to rune rootless. or just docker-rootless-extra is sufficient? 2024-09-06 13:20:14 what I see is that docker sub packages include docker-rootless-extra package and 2024-09-06 13:20:14 docker-rootless-extra sub package also include docker. very confusing 2024-09-06 13:20:29 s/rune/run/ 2024-09-06 13:29:34 Velo_Frost[m]: i haven't used docker-rootless but I've been using podman rootless for a while on alpine 2024-09-06 13:30:12 great. how is podman compared to docker. pro and cons? 2024-09-06 13:30:27 basically - make sure cgroupsv2 is up and running and you have configured /etc/sub[ug]id 2024-09-06 13:30:53 all right 2024-09-06 13:31:27 the user you have created, is it un-previliged and did you add that to wheel? 2024-09-06 13:32:55 I mean you must have installed rootless-podman with sudo or daos? 2024-09-06 13:36:03 no, the podman package will include all the necessary sub-packages to run rootless by default 2024-09-06 13:36:25 but yes, you need root or (sudo/doas) privileges to install packages with apk 2024-09-06 13:37:11 "apk add podman" as root should get it installed for you 2024-09-06 13:38:23 ok 2024-09-06 14:22:17 as for how podman compares to docker, podman is more stable, has more features, and is more secure out of the box 2024-09-06 14:25:58 sometimes that extra security can get in the way if you're trying to do something specific, but generally it's a good thing 2024-09-06 14:30:00 yeah, i've been using rootless podman I guess since the 3.x days to run production gitlab instances :-) Works well for me. (alias docker=podman FTW!) 2024-09-06 14:30:52 What about podman-compose? 2024-09-06 14:32:01 i've not used podman-compose myself but have promised to try it the next time I need a multi-container setup for something 2024-09-06 14:37:14 Does anyone have a preferred system monitoring tool for daily driver that works well with openrc? System Monitor seems to require cgroupsv2 and systemd for applications view 2024-09-06 14:39:10 * system monitoring gui tool for 2024-09-06 15:10:55 ACTION just uses simple cli tools but hears many people use conky 2024-09-06 15:14:30 modprobe tun 2024-09-06 15:14:53 rmmod irc 2024-09-06 15:19:31 hi everybody 2024-09-06 15:19:43 hi 2024-09-06 15:26:38 i want to use alpine linux as a host for virtual machines 2024-09-06 15:28:14 i recent install tiny windows on my laptop but i get poor performance until install intel rapid storage driver, donwloaded from hp support center 2024-09-06 15:29:40 if i install alpine linux on that laptop it will have same poor performance? how i can install intel rapid storage driver on alpine linux? 2024-09-06 16:24:30 hello 2024-09-06 16:25:12 my mkimage error "wpa_supplicant=coreutils: unable to select package (or its dependencies)" 2024-09-06 16:52:30 hi 2024-09-06 16:58:02 hi 2024-09-06 17:06:26 MarcoGT: It seems to be a RAID implementation 2024-09-06 17:07:16 hi ikke thanks for answer 2024-09-06 17:08:39 Do you have multiple disks in that laptop? 2024-09-06 17:08:44 i was reading on forums and it seems as installing the driver on windows enables AHCI mode, linux supports it natively right? 2024-09-06 17:08:52 no, just one disk 2024-09-06 17:09:02 hdd sata disk 2024-09-06 17:12:09 linux should support that natively, yes 2024-09-06 17:12:16 Just make sure your bios is set to ahci 2024-09-06 17:12:53 my concern is about why windows clean installed works but poor performance, when i install this intel rapid storage technology driver performance boost 2024-09-06 17:14:02 so i suspect if i install alpine linux clean install, performance will be poor too, so i need to install some kind of driver to achive high performance? 2024-09-06 17:14:56 Not everything that applies to windows will also apply to linux 2024-09-06 17:15:15 Where in windows you need to manually install all kinds of drivers, linux comes with many drivers built-in 2024-09-06 17:21:08 ok thank you i will try 2024-09-06 18:58:28 sweet lord, finally got codium flatpak to work on alpine with rust analyzer. 2024-09-06 18:58:28 For flatpak they have language sdk's you can add, and then need to ensure a correct env variable exists in sandbox to leverage, as well as spawning terminals through host and potentially giving more open fs access (i just went with full host access, didn't want a sandbox, just to run on musl system lol) 2024-09-06 18:58:28 Can't wait till it is in apk - they have an alpine reh (remote extension host) build, but I didn't feel like teasing that apart with normal build. 2024-09-06 18:58:48 I am assuming code-oss is november which is exciting too 2024-09-06 18:59:13 snaps an flatpak make me want to quit computers lol 2024-09-06 21:33:01 can anyone talk to us about our ipv6 issue? 2024-09-07 00:04:31 Soni: the vast majority of people are going to ignore meta questions... In order to get a response, I'd suggest explaining exactly what your issues is. 2024-09-07 09:21:08 The Alpine Linux is so buggy when I boot it inside QEMU: two graphical windows open, one of them not responding to any input... 2024-09-07 12:33:31 Hi eveyone!!! 2024-09-07 12:34:09 zcrayfish: talked about it with musl devs, something about SIIT-DC/pref64 support in libc 2024-09-07 13:46:22 quick sanity check - with busybox mkdir, it is not possible to create multiple dirs with 1 command? I believe based on searching that you can't, just wan't to confirm if anyone knows 2024-09-07 13:47:27 e.g. with gnu may be mkdir -p /some/dirs/{one,two,three} 2024-09-07 13:47:38 andar1an[m]: you are confusing what the shell does and what mkdir does 2024-09-07 13:48:00 I thought mkdir was part of busybox? 2024-09-07 13:48:02 mkdir can make multiple directories, but {a,b} is non-posix shell syntax that busybox ash does not support, but bash and zsh do 2024-09-07 13:48:18 andar1an[m]: it's not mkdir that parses the {a,b} syntax 2024-09-07 13:48:22 ah ok, thank you 2024-09-07 13:49:03 I didn't know that was non-posix 2024-09-07 13:49:26 so it would be sus or something unrelated? Need to remove from scripts 2024-09-07 13:49:48 If you want to support many environments, you should not use it indeed 2024-09-07 13:50:44 merci - seems like you can just pass multiple dirs anyways without the curlies 2024-09-07 13:50:49 yes 2024-09-07 13:51:06 the shell does that normally 2024-09-07 13:51:11 it expands {a,b} to a b 2024-09-07 13:51:24 and foo/{a,b} to foo/a foo/b 2024-09-07 13:52:47 I found a posix cheatsheet. Gonna cross reference new things from now on haha 2024-09-07 14:00:28 Or directly look at the specs 2024-09-07 14:12:58 so it would be sus or something unrelated? Need to remove from scripts 2024-09-07 14:13:06 sus is just posix + curses 2024-09-07 14:14:02 TIL 2024-09-07 14:14:56 so basically posix is lowest level, and if you are posix compliant you are sus compliant? unless using curses? lol 2024-09-07 14:22:34 I’m not aware of something like “SUS compliance” for applications. Implementations can conform to SUS, be certified by The Open Group, and thenceforth use the trademark “UNIX.” That’s a bunch of nonsense. 2024-09-07 14:24:17 Applications can conform to POSIX (or conform to POSIX with extensions, or conform to POSIX with national stuff, or whatever); that’s all defined in POSIX. 2024-09-07 14:26:16 One of the major things I noticed on my learning journey from mech to software was the shear amount of layers. Why I enjoy embedded learning is it can make things more simple, and why the minimalism of alpine feels refreshing to me. 2024-09-07 17:01:46 i tend to agree with google that the default shell choice should be bash, unless you have a compelling reason to deviate https://google.github.io/styleguide/shellguide.html#s1.1-which-shell-to-use 2024-09-07 17:16:50 Which version of bash? 2024-09-07 17:29:48 how to run a hello-world.sh file in alpine? I have done the chmod +x 2024-09-07 17:30:02 ./hello-world.sh 2024-09-07 17:30:12 That's not different from any other linux distro 2024-09-07 17:32:34 it says -sh: ./hello-world.sh: not found :( 2024-09-07 17:34:01 probably the shebang 2024-09-07 17:34:43 sorry, shutting up 2024-09-07 17:35:21 should I not give the shebang directive? 2024-09-07 17:37:05 How is the shebang? Maybe its bash and you just dont have it installed 2024-09-07 17:37:35 that is the kind of issue i am implying 2024-09-07 17:37:47 ooh yeah that must be the case, the shebag is #!/bin/bash 2024-09-07 17:56:24 quinq: RHEL 8 still uses bash 4, but you can use bash 5 unless you know you can't 2024-09-07 18:08:16 xD 2024-09-07 19:20:47 is this wiki article correct in that some commands are "optional"? https://en.wikipedia.org/wiki/List_of_POSIX_commands 2024-09-07 19:40:29 frag: most likely, yes 2024-09-07 19:40:55 #!/usr/bin/env bash is nice too. But ash is part of busybox no? 2024-09-07 19:41:22 I don't think I have hit a scenario where I needed more than ash on machine setup yet 2024-09-07 19:43:43 or what is available through busybox I should say. And if you want you can just install 2024-09-07 19:46:19 ash is nice imo, for both scripts and interactive 2024-09-07 19:48:15 I have bash, zsh, and fish to play with. Still don't know what I want to call home. The only thing I am not sure about yet is completions in ash (if possible) 2024-09-07 19:49:07 ash doesn't have completions and is not very extensible 2024-09-07 19:49:20 have you messed with alacritty? 2024-09-07 19:49:57 That was on my list too, but on this laptop doesn't make sense 2024-09-07 19:50:04 I'm using it now 2024-09-07 19:50:12 do you enjoy? 2024-09-07 19:50:31 It's a fine terminal 2024-09-07 19:51:05 even forgot why you would need completion, for anything but filenames.. 2024-09-07 19:51:32 recommending commands 2024-09-07 19:51:50 fish does a nice job there 2024-09-07 19:52:22 makes discoverability of a cli easier 2024-09-07 19:53:12 I think a lot of young developers may not be as familiar with man pages or cli help since a lot of what one does is docs through internet 2024-09-07 19:53:30 especially as platforms obfuscate lower level functionality 2024-09-07 19:54:08 and then queue llm's haha 2024-09-07 19:54:13 i find command completions very useful 2024-09-07 19:54:25 +1 2024-09-07 19:54:45 hi all 2024-09-07 19:55:26 howdie 2024-09-07 20:02:08 do the alpine test suites in setup that use kyua need a kyuafile? 2024-09-07 20:02:30 you mean for alpine-conf? 2024-09-07 20:02:40 yes, sorry 2024-09-07 20:03:27 I think I can reference 1 test at a time as a shell 2024-09-07 20:03:35 *shell script 2024-09-07 20:03:55 but wondering if there is supposed to be a top level kyuafile to run all, or if it is somehow generated 2024-09-07 20:04:28 https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/Makefile?ref_type=heads#L108 2024-09-07 20:05:00 ah cool, thank you. I am used to cargo tests lol 2024-09-07 20:05:35 seems easy to just make a simple kyuafile too 2024-09-07 20:06:04 this is pretty great. I never knew how to test bash - the mock disks are slick 2024-09-07 20:06:20 I was living in caveman test in place days lol 2024-09-07 21:02:58 Is there a better way to see relevant error output from tests other than `kyua --loglevel=debug test`? 2024-09-07 21:02:58 I am starting with "I don't have everything I need installed". Currently kyua and atf-sh are installed. Most tests fail but about a quarter pass. 2024-09-07 21:02:58 This info is not clear as to why tests are failing. 2024-09-07 21:04:02 this specific wording makes me think there may be a better way to investigate: failed: atf-check failed; see the output of the test for details [0.011s] 2024-09-07 21:07:15 ah, maybe I need to query the sqlite db generated for details 2024-09-07 21:07:37 There is a command for that 2024-09-07 21:08:08 cool, thanks. definitely see some messages in there 2024-09-07 21:08:19 will be easier with command 2024-09-07 21:09:26 seems to not be finding some scripts called libalpine.sh often 2024-09-07 21:09:35 kyua report --verbose 2024-09-07 21:10:19 merci. Seems what is in repo is libalpine.sh.in, so I may be running the tests wrong 2024-09-07 21:10:59 You need to run `make` to build it\ 2024-09-07 21:11:38 the libalpine.sh also? I just ran make for the tests/kyuafile. Shoulda looked through rest. Thank you 2024-09-07 21:12:05 yes 2024-09-07 21:12:27 .in are build input files 2024-09-07 21:13:24 I have never seen those before. I thought it was some special extension for scripts during install, not input haha 2024-09-07 21:13:32 derp 2024-09-07 21:16:23 Thanks so much, most pass now 2024-09-07 21:16:53 last one is likely due to permissions 2024-09-07 21:17:22 should tests be run as root? 2024-09-07 21:18:08 that seemed to do it. WOO 2024-09-07 21:46:31 https://sendvid.com/aqgpr9ff# 2024-09-07 21:51:13 do not click that. 2024-09-07 21:57:29 roger 2024-09-07 21:58:17 what could go wrong 2024-09-07 21:58:25 ACTION clicks 2024-09-07 21:58:34 ok 2024-09-07 21:58:48 but are you ok? 2024-09-07 21:58:58 no 2024-09-07 21:59:04 i am vindicated 2024-09-07 21:59:11 i was not looking for vindication, to be clear 2024-09-07 21:59:14 but here we are 2024-09-07 21:59:50 how do you know 2024-09-07 21:59:54 maybe you were all along 2024-09-08 00:40:05 i made a merge request to upgrade electrum wallet from 4.5.4 to 4.5.5: https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/71444 but it seems the maintainer has been away for a while (at least it shows no activity on gitlab) 2024-09-08 00:40:56 this version was also released months ago (may 29), so i think he won't maintain it anymore. should i change myself to the maintainer? 2024-09-08 07:01:21 Hi! I'm trying to compile the "st" terminal emulator, and I can see a message saying that the "tic" program, used to set up some terminal info, not found. How can I find the package that includes "tic"? 2024-09-08 07:03:22 uwumeowmeownyaa: https://pkgs.alpinelinux.org/contents?file=tic&path=&name=&branch=edge 2024-09-08 07:07:38 Why is it in ncurses package? That's so silly... 2024-09-08 07:08:00 Thank you very much, now I know how to resolve such situations :-) 2024-09-08 07:08:46 Because it's ncurses that provides it in the first place? 2024-09-08 07:20:09 I have changed to edge repositories. Apk update and apk upgrade. Almost 166 packaged upgraded.  2024-09-08 07:21:17 I have lvm on luks. Need password  2 times. First grub and than booting alpine. After upgrade it only need paraphrase on grub  2024-09-08 07:21:45 Can I revert back downgrade everything to stable release 2024-09-08 07:22:25 Velo_Frost[m]: change the repositories, then run apk upgrade --update --available 2024-09-08 08:04:26 uwumeowmeownyaa, it's also possible to do things like: apk add cmd:tic 2024-09-08 08:19:59 Can someone help me with this dcker setup 2024-09-08 08:20:29 I'm trying since 3 days. 2024-09-08 08:26:25 Further errors reports from docker-rootless-setuptool.sh install are that path variables 2024-09-08 08:26:25 docker-rootless-setuptool.sh install  prompt errors/info that docker-rootless.sh  need to be run manually. When $docker-rootless.sh it hang the terminal counsel. Sometime parent child error and containerd error. 2024-09-08 08:26:55 The manual running of script is because of non systemd 2024-09-08 09:03:33 quinq: This is a very cool command too! 2024-09-08 09:06:40 Also, I have set up an Alpine Linux virtual machine image, did some configurations, installed packaged, etc... Now I would like to save the progress to be able to do the exact same setup in the future. Is there any command in Alpine Linux to save everything I have configured, or should I just write a simple note and store it alongside the OS image somewhere safe? 2024-09-08 09:24:22 You could use tar 2024-09-08 09:30:12 I'll copy the image then, but thanks :-) 2024-09-08 09:44:07 'k :) 2024-09-08 11:33:39 any hope for https://gitlab.alpinelinux.org/alpine/mkinitfs/-/issues/18 ? 2024-09-08 11:34:14 I don't think on a short term 2024-09-08 12:27:09 scorpion2185[m]: it's an impactful change, with the potential of breaking many existing setups 2024-09-08 13:30:12 it could have new version alongside current one i think 2024-09-08 13:42:06 someone writing tests around it probably would speed things up 2024-09-08 14:24:51 asking this again: is it possible to umount the media you booted from after starting up diskless? it says "resource busy", but i dont know how/why 2024-09-08 14:25:08 frag: I mentioned you need to stop the modloop service 2024-09-08 14:28:14 ok, ty ikke! 2024-09-08 14:28:53 That will unmount the modloop which is located on the boot media 2024-09-08 14:31:02 okok, no idea what it is tho hehe 2024-09-08 14:31:39 It contains the kernel modules and firmware, and is used to prevent those files eating up all your ram when put into tmpfs 2024-09-08 14:31:42 looks like I need to make something custom-amde for unl0kr. https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/nlplug-findfs/nlplug-findfs.c#L652 right? 2024-09-08 14:32:40 sounds important ... 2024-09-08 14:33:04 frag: It is only necessary if you need to load more modules 2024-09-08 14:33:51 okok 2024-09-08 16:51:38 what's need to be done to get deterministic network interface names? because eudev-netifnames is already installed? 2024-09-08 16:53:07 Can I use btrfs + luks right? 2024-09-08 16:53:07 uhm i can encrypt only home. 2024-09-08 16:54:32 maybe a kernel parameter for deterministic network interface names `GRUB_CMDLINE_LINUX="net.ifnames=0"` i ngrub 2024-09-08 16:54:38 not sure 2024-09-08 17:02:12 can I forward a disk (/dev/sdX) to alpine VM to install alpine in /dev/sdX? 2024-09-08 17:39:47 i once had a btrfs install corrupt the luks header on an encrypted partition. hope btrfs is much more solid these days 2024-09-08 17:42:07 that's interesting - the linux vfs layer should have prevented that 2024-09-08 17:43:01 i've not had problems with luks/btrfs but i never suggest it to anyone. i always tell everyone to stick to ext4 2024-09-08 17:46:26 xfs is also rock solid :) 2024-09-08 17:46:36 I'm going btrfs media server/NAS. Figure there are benefits to putting filesystem metadata in raid 1. Maybe do compression and raid 1 for non-media storage. 2024-09-08 17:47:21 Really wish btrfs raid 5/6 weren't broken, because that seems like it would be perfect for a home NAS. 2024-09-08 17:48:36 if it were me doing that, i'd go with zfs 2024-09-08 17:49:08 I won't deal with an out-of-tree filesystem. Plus, I like btrfs' ability to mix-and-match drive sizes. 2024-09-08 17:51:41 That capability is part of what makes it perfect for home use. I'm not a corporation ordering a pallet of HDDs at a time. I'll pick up a drive when I have a need and if the value is right, and that means there will be a good deal of mismatched drives, which btrfs raid handles well. 2024-09-08 17:51:51 you can mix drives on zfs 2024-09-08 17:52:25 I believe mixing drives on ZFS means all of your drives are treated as if they were the same size as your smallest drive. 2024-09-08 17:52:28 but anyways i'm not gonna argue the point except to say that it's a data risk thing you have to make, especially when you start getting into btrfs advanced features 2024-09-08 17:52:48 yes, that's a caveat 2024-09-08 17:53:21 zfs on linux seems to have earned a good reputation. was nervous about it for many years. but i still use a hardware RAID card lol 2024-09-08 17:54:01 zfs is much more mature code, even for linux, especially for complicated features like btrfs is attempting now 2024-09-08 17:54:20 i'm not talking shit about btrfs, but fs work is hard to do right 2024-09-08 17:54:26 When do Oracle's patents expire, thus freeing linux to bring ZFS into the kernel? 2024-09-08 17:56:30 no idea. 2024-09-08 17:57:05 it's not really a patent issue, it's a license issue 2024-09-08 17:57:45 License become irrelevent if the patent is expired. 2024-09-08 17:57:49 grossaffe[m], incorrect 2024-09-08 17:59:56 That was the understanding I had come to regarding ZFS, but I don't recall the details of why, so it's not a hill I'll die on. 2024-09-08 18:00:30 the main reason it's not merged is because the license is incompatible 2024-09-08 18:00:51 grossaffe[m], ok good :) 2024-09-08 18:01:33 grossaffe[m], i for one meanwhile wasn't even aware if there perhaps was a patent question -besides- the license problem 2024-09-08 18:01:36 but the license problem is enough 2024-09-08 18:01:47 copyright protection typically lasts 70 years. i wonder how that will pan out for software with undesirable licenses. it's still a moot point for software that is actively updated, but still 2024-09-08 18:02:53 https://en.wikipedia.org/wiki/List_of_copyright_terms_of_countries has a nice table 2024-09-08 18:02:59 i assume there would be some things to adjust if it were to be integrated in mainline, but overall it definitely seems to be the most mature "modern filesystem" by a long shot 2024-09-08 18:03:01 peerl3ss: 70 years after the death of the author. Used to be just 14 years with the option to renew another 14 years. Back before Disney bought the government. 2024-09-08 18:03:36 oof, that's a long wait 2024-09-08 18:04:56 grossaffe[m]: what is your concern with running zfs as a module? are you tracking the LTS kernel? 2024-09-08 18:05:01 ext4 is a good filesystem. it's reliable. people move away from it without good reasons. 2024-09-08 18:05:36 invoked, when i use btrfs, it's for snapshots or dedup - i'd love for you to tell me "well actually ext4 has those now" 2024-09-08 18:06:04 ext4 + lvm will get you to snapshots 2024-09-08 18:06:21 lvm gets you a lot of extra things, but yeah, i mean those are reasons. 2024-09-08 18:06:29 peerl3ss, ok yes, that's true 2024-09-08 18:07:00 zfs send, checksums, datasets with different policies, raid that isn't dependent on one raid card not dying, etc 2024-09-08 18:07:05 peerl3ss, not as convenient, but i bet the tooling exists 2024-09-08 18:07:11 peerl3ss: I just don't want anything out-of-tree on my device. I've escaped nvidia's clutches and I'm not going back. But also I just see more value in the way btrfs approaches multi-device filesystems for a home media server with drives added piece-meal. 2024-09-08 18:09:58 also, depending on the kind of data being stored, btrfs' transparent compression can be convenient 2024-09-08 18:10:26 very much a thing in zfs too ;) 2024-09-08 18:10:45 socksinspace: yes, that was wrt what invoked said 2024-09-08 18:11:07 btw, i learned xfs has dedup now too 2024-09-08 18:11:12 ah, missed the contexts :) 2024-09-08 18:11:43 (but it should be noted that btrfs/xfs dedup is not identical to zfs dedup) 2024-09-08 18:11:59 neat! xfs is very actively developed 2024-09-08 18:12:21 I am looking at trying the compression for my documents. Put 'em in full raid with all the drives and compress. Probably won't make a big difference in my storage, though. Most of the data will probably be video, which is already compressed and would not benefit from filesystem-level compression. 2024-09-08 18:13:33 i get not wanting external modules, just for the hassle, but i trust zfs more than btrfs (not based on personal experience luckily :D ) so it outweighs that "problem" for my use 2024-09-08 18:14:40 i would not trust btrfs to hold my tax returns 2024-09-08 18:14:44 If ZFS had an equivalent to btrfs' raid, it would at least make me consider adding it as an out-of-tree module. 2024-09-08 18:15:16 so btrfs can raid over varying sizes? how does that work? do you decide the replication factor per filesystem? 2024-09-08 18:15:53 is that the one where even the btrfs devs say it might loose your data? 2024-09-08 18:16:08 there's vdo for compression and dedup. never used it though. in theory, you make a vdo device then mkfs.ext4 on the vdo device and voila you have those features. 2024-09-08 18:16:31 "xfs is also rock solid :)" <- Where have you been sir 2024-09-08 18:17:07 Rather than mirroring discs, it mirrors chunks of data. And you choose how many copies of that data should be kept, and it will spread them out amongst the different devices. So btrfs raid1 will keep 2 copies of your data in your array. btrfs raid1c3 will keep 3 copies of your data in your array. 2024-09-08 18:17:30 As for data loss, that's the parity-raid (raid5/6) write-hole issue. 2024-09-08 18:17:44 grossaffe[m], right, and you can pick that per filesystem on the volume? 2024-09-08 18:17:47 It doesn't affect btrfs raid 1. 2024-09-08 18:17:55 ahhh ok 2024-09-08 18:18:08 but i had data loss on a btrfs file system that was unraided (same HW raid card) 2024-09-08 18:18:19 invoked, TIL, thanks 2024-09-08 18:18:36 i am not brave enough for hw raid :D 2024-09-08 18:19:17 You can pick different raid levels by directory if you choose. And you can do different raid levels for different types of data. So you can go without raid for "data" data, and go with raid for filesystem metadata so that even if things get corrupted somewhere, you'll have a functioning filesystem rather than losing everything. 2024-09-08 18:19:24 grossaffe[m], ok, neat 2024-09-08 18:19:26 btrfs has rough edges (ie do regular backups). again, fs work is very tricky stuff 2024-09-08 18:19:30 i figure i can just take my drives to a data recovery place if my card dies or w/e... you can plug in a similar card and the drives will show up ok 2024-09-08 18:19:43 all this said, my /home has been on btrfs for 4 years without trouble now 2024-09-08 18:20:14 in a single threaded system fs work would be easy, i guess. but alas 2024-09-08 18:20:23 also other work :D 2024-09-08 18:20:29 Velo_Frost[m]: i've been working! 2024-09-08 18:21:18 That's what make btrfs so compelling, if only it was more mature. Being able to use raid 5/6 for media that can always be replaced without having to halve my drive space would be nice. 2024-09-08 18:21:27 peerl3ss: All right. Hope everything going good 2024-09-08 18:25:19 Hey, could someone who has write access to the alpine-conf repository check merge request 211? It's a very simple change that fixes what I would consider a big bug. Many thanks. 2024-09-08 18:39:17 @_oftc_peerl3ss:matrix.org have something for you in your DM. 2024-09-08 19:00:45 i use virt-manager how can I add physical disk? sd card it is shown as /dev/sddx 2024-09-08 19:03:52 just set the path as the source path 2024-09-08 19:04:24 ACTION uploaded an image: (10KiB) < https://matrix.org/oftc/media/v1/media/download/AX10KmOC1DVqFsbPpIsk0uttwZ1nN7vA0lJIjpM1JmmbMrzNesXRKznmtGcD4gFUJ2DesIupULPN9imOMlOlsGxCeR1L1ZIwAG1hdHJpeC5ib3R0b21zZXJ2aWNlcy5jbHViL2FSWmlYZk9yRHphdFNlcmFtYXdvZkViSQ > 2024-09-08 19:04:38 ACTION uploaded an image: (13KiB) < https://matrix.org/oftc/media/v1/media/download/Aawp2R1SYG9XWeAr6qSgoFlTJu_dhpCtbJRErMrhbgU7s8jurHWFW1vkbck4Wp3qW18uwCxunJqxAcxtsZxnzQFCeR1L2MuwAG1hdHJpeC5ib3R0b21zZXJ2aWNlcy5jbHViL3dUa2hnYm5rWmRMbERSbW9MeVhhQUZtbw > 2024-09-08 19:05:10 2024-09-08 19:05:21 Add hardware -> VirtIO VSOCK ? 2024-09-08 19:05:48 i've benn searching and this looks good https://serverfault.com/questions/261703/kvm-guest-using-physical-sata-disk/261907#261907 2024-09-08 19:06:16 >Add hardware -> VirtIO VSOCK ? 2024-09-08 19:06:17 you just add a storage device like normal and set the path in /dev as the source, that has nothing to do with it 2024-09-08 19:06:43 i am not expert of virt-manager , how add it as normal? 2024-09-08 19:07:32 i guess it's Storage then 2024-09-08 19:07:44 Add Hardware -> Select or create custom storage -> Manage -> Browse Local -> Locate disk 2024-09-08 19:09:05 no need to add a new storage pool 2024-09-08 19:09:59 i see and do I add `index="N"/>` in XML? to `` 2024-09-08 19:11:37 sorry for confusion, no. that's only there while the VM is running. will be populated automatically 2024-09-08 19:21:43 does someby know a guide for btrfs+luks? 2024-09-08 19:24:48 no sorry no luks cause i cannot use unl0kr 2024-09-08 19:25:14 https://wiki.alpinelinux.org/wiki/Install_Alpine_on_a_btrfs_filesystem_with_refind_as_boot_manager 2024-09-08 19:29:33 how do I run setup-alpine so that it will stop before making the disks? 2024-09-08 19:29:46 so I can use btrfs 2024-09-08 19:34:43 no way 2024-09-08 19:34:49 install in semi-automatic mode 2024-09-08 19:49:46 ROOTFS The filesystem to use on /, defaults to ext4, also allows ext2, ext3, (flat) btrfs and xfs. 2024-09-08 19:49:52 what's flat? 2024-09-08 19:53:41 env variable 2024-09-08 20:06:50 hello 2024-09-08 20:10:32 it made tmpfs /tmp tmpfs nosuid,nodev 0 0 in fstab 2024-09-08 20:21:37 how force removal of dep of a package ? 2024-09-08 20:23:40 can I hack it somehow without rebuild a package? 2024-09-08 20:30:50 scorpion2185[m]: The hacky way is to install the dep as a virtual package 2024-09-08 20:30:55 apk add -t 2024-09-09 02:19:08 does anyone use helix with alpine? Do you have lsp servers working properly? 2024-09-09 02:20:06 wanted to tinker with this and am noticing weird issues with lsp servers where if I install as root, even if bins are in /usr/local/bin and permissions are correct, system user running hx --health will not register the lsp servers 2024-09-09 02:20:31 as root it finds them, and installing the lsp servers as user in home dir also works 2024-09-09 02:21:19 but I seem to be unable to get the binaries installed either as user or root to work for the other 2024-09-09 02:22:05 a workaround could be to install the lsp servers as root, and as user, but that seems silly lol 2024-09-09 02:26:23 this is an apk for helix. have been playing around with it today. very nice. no issues from i could see. 2024-09-09 02:26:31 s/this/there/ 2024-09-09 02:31:29 That is what I am using 2024-09-09 02:31:37 have you tried to use lsp servers? 2024-09-09 02:46:17 i thought so, but after running the --health option, not so sure... looking into it now 2024-09-09 02:48:11 I can get health to work for root and user respectively if the lsp servers were installed in the respective spaces 2024-09-09 02:50:10 If installed in home, root can see, but I have lsp issues errors. 2024-09-09 02:50:35 if installed by root, even with -rwxr-xr-x, no bueno for user 2024-09-09 02:51:12 one weird thing i notice in the health report is that is cautions that /usr/bin/runtime does not exist... which is not a valid dir for such purpose 2024-09-09 02:52:07 mine will cautiion for `$USER/.config/helix/runtime 2024-09-09 02:52:37 so root would be root/.config and my user would be username/.config 2024-09-09 02:52:46 mine did also, i created an empty one 2024-09-09 02:53:23 from what I read that is just a warning 2024-09-09 02:53:32 and it looks in multiple locations 2024-09-09 02:53:36 e.g. line above is Runtime directories: /root/.config/helix/runtime;/usr/share/helix/runtime 2024-09-09 02:54:03 don't know if that is accurate though 2024-09-09 02:56:56 yeah 2024-09-09 02:59:47 well, I am out of juice for the night haha. I will smack my head against this next weekend maybe. Hopefully rested brain will see better haha. Have good one 2024-09-09 03:00:23 you also 2024-09-09 03:01:20 think i'm going to just stick with vis anyway 2024-09-09 04:17:07 andar1an[m]: from what i understood it downloads complete precompiled runtimes that expect glibc 2024-09-09 04:25:26 n/m that may be a different editor that I was thinking of 2024-09-09 04:56:16 andar1an[m]: if they count, only have gopls and zls installed at the moment from aports and hx --health detected them 2024-09-09 04:57:50 for other runtime components would just get them from upstream, the aports packages for those (the tree-sitter grammars etc.) didn't work for me, like the aports issue described 2024-09-09 05:20:58 I was wondering, we don't have gnugo in alpine at the moment, do we? I am referring not to google's go, but the go engine: https://www.gnu.org/software/gnugo/ 2024-09-09 05:23:56 We have kigo as a frontend, but it seems to be unusable without gnugo installed 2024-09-09 10:02:07 "andar1an: from what i understood..." <- I have gcompat installed, can look into more later. I have not used gcompat before. The current issue seems to do with permissions though 2024-09-09 10:03:45 "andar1an: if they count, only..." <- Thanks. Try hx --health as both root, and your user. I installed from npm and cargo. Didn't know tree-sitter grammars were available through apl 2024-09-09 10:03:55 * Thanks. Try hx --health as both root, and your user. I installed from npm and cargo. Didn't know tree-sitter grammars were available through apk 2024-09-09 10:06:09 Rust-analyzer seems fine, but it also seems that rust components are installed in both /root/.cargo, and $HOME/.cargo. 2024-09-09 11:43:35 to boot a pinephone with twoboot can I just install u-boot-pinephone from pmos? 2024-09-09 11:48:40 #postmarketos 2024-09-09 12:30:05 I can't wait to get a device for postmarket haha 2024-09-09 13:51:11 Waiting to see you online @_oftc_peerl3ss:matrix.org proxmos 2024-09-09 14:32:11 Hi there. I installed alpine as my daily-driver distro this past weekend, and I must say, it has been amazing. I have a couple of issues however, that are yet to be resolved. 2024-09-09 14:36:55 Firstly, the issue thats hardest to explain, but I'll do my best - my terminal sometimes gets stuck, I guess, but only visually until I input another character. For example, when I hop down half a page in vim, using , it would sometimes look like nothing happened, but in fact it simply didn't render the next frame. After I input another character, I see the result of my 2024-09-09 14:36:57 last motion. 2024-09-09 14:37:50 What terminal? Or do you mean just console? 2024-09-09 14:38:50 Was just about to say, I'll make sure if it is not specific to foot term, but iirc it happened in tty also. I'll get back to you on that one, ty for reply. 2024-09-09 14:39:17 it happens pretty frequently, but at random times 2024-09-09 14:40:42 what card 2024-09-09 14:42:07 it's intel integrated graphics, my processor is an Intel Celeron N4000 1.10GHz 2024-09-09 14:42:55 does that help? Just checked in my tty, and can't reproduce it there, so might have been mistaken. 2024-09-09 14:44:11 Before I switched, I used foot term on ubuntu without this issue, but I'll test another terminal - maybe the alpine foot package is newer and introduced an issue perhaps 2024-09-09 14:46:09 I should probably mention, I did follow the how-to guides on the alpine wiki to install the proper drivers. Will list the ones I installed. 2024-09-09 14:49:34 cool. it vaguely sounds to me like a hw accel issue... try a different terminal, like alacritty or weston-terminal and see if it recurs there (to rule out foot itself) 2024-09-09 14:49:35 ok, so seems it is foot thats at fault, or should I say something is afoot 2024-09-09 14:50:32 yeah, sry. In hindsight as I'm typing my question I realised what I was supposed to do to debug. Mb for the timewaste guys. 2024-09-09 14:50:47 Oh no it's fine 2024-09-09 14:51:11 check dmesg for i915 ... there is something extra you need to do if it's i915, iirc 2024-09-09 14:51:12 You might want to see if your drivers need sorting out. 2024-09-09 14:52:13 so `doas dmesg | grep i915` gets me the following output 2024-09-09 14:52:47 [ 18.964864] i915 0000:00:02.0: [drm] Unknown revid 0x06 2024-09-09 14:52:49 [ 19.002365] i915 0000:00:02.0: vgaarb: deactivate vga console 2024-09-09 14:52:51 [ 19.002500] i915 0000:00:02.0: [drm] couldn't get memory information 2024-09-09 14:52:53 [ 19.002671] i915 0000:00:02.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=io+mem:owns=io+mem 2024-09-09 14:52:55 [ 19.003422] mei_hdcp 0000:00:0f.0-b638ab7e-94e2-4ea2-a552-d1c54b627f04: bound 0000:00:02.0 (ops i915_hdcp_ops [i915]) 2024-09-09 14:52:57 [ 19.057723] [drm] Initialized i915 1.6.0 20201103 for 0000:00:02.0 on minor 1 2024-09-09 14:52:59 [ 19.060586] snd_hda_intel 0000:00:0e.0: bound 0000:00:02.0 (ops i915_audio_component_bind_ops [i915]) 2024-09-09 14:53:01 [ 19.064012] fbcon: i915drmfb (fb0) is primary device 2024-09-09 14:53:03 [ 19.082194] i915 0000:00:02.0: [drm] Finished loading DMC firmware i915/glk_dmc_ver1_04.bin (v1.4) 2024-09-09 14:53:05 [ 19.091742] i915 0000:00:02.0: [drm] fb0: i915drmfb frame buffer device 2024-09-09 14:53:07 [ 1221.542453] mei_hdcp 0000:00:0f.0-b638ab7e-94e2-4ea2-a552-d1c54b627f04: bound 0000:00:02.0 (ops i915_hdcp_ops [i915]) 2024-09-09 14:53:13 next time use a paste service pls 2024-09-09 14:53:29 ah ok np, will do 2024-09-09 14:54:57 that looks ... ok 2024-09-09 14:55:15 if you have sysfs-utils installed, run 'doas systool -m i915 -av' 2024-09-09 14:56:46 what that does is check to see what options are set for the i915 module you have loaded 2024-09-09 14:57:11 Can you recommend a paste service to use? 2024-09-09 14:58:45 not a recommendation but https://bpa.st/ is fine 2024-09-09 15:00:08 I guess I just share this link then, right? https://bpa.st/56AJE 2024-09-09 15:00:26 why not ours 2024-09-09 15:00:28 https://tpaste.us/ 2024-09-09 15:01:31 sure. i keep falsely remembering it as tpaste.net 2024-09-09 15:03:05 https://tpaste.us/WxyN 2024-09-09 15:03:16 damn, thats sick :) 2024-09-09 15:04:32 i'm just comparing your output to one of my systems running i915 defaults 2024-09-09 15:04:37 please hold 2024-09-09 15:05:48 np, I'm totally lost at this point. Knowledge gap becoming more and more apparent. 2024-09-09 15:07:10 no differences. what does 'lspci | grep VGA' say? 2024-09-09 15:09:12 One sec, looks like I don't have lspci, and it's not named like that on apk. Do you perhaps know the package name? 2024-09-09 15:09:18 lspci is part of the 'pciutils' package. 2024-09-09 15:09:32 ah ok ty 2024-09-09 15:12:13 did you check to see if alacritty/something else reproduces the problem 2024-09-09 15:12:21 https://tpaste.us/ggLP 2024-09-09 15:13:53 so many lakes. i'll have to look to see if that has any special handling 2024-09-09 15:14:04 oh yeah I mentioned that earlier. I couldn't reproduce on tty or alacritty. So yeah maybe I should rather just look at the open foot issues or smth and figure it out from there. 2024-09-09 15:15:40 yeah. if it's foot defaults it may be a bug. 2024-09-09 15:15:40 Is there a term for this kind of issue? I had trouble explaining it earlier, and tried searching for a lot of different things to no avail. 2024-09-09 15:16:22 it's a rendering problem 2024-09-09 15:17:48 but it's hard to say. if it's foot defaults, foot may be exposing a bug in something else. either way it's good to report it from a user pov with the dmesg/systool/lspci you shared here. 2024-09-09 15:19:11 Will definitely do, and thanks for trying to help invoked and xulfer. You're champs. 2024-09-09 15:19:30 i don't think you mentioned what compositor you're using 2024-09-09 15:19:50 Oh yeah I'm running sway. 2024-09-09 15:20:28 yeah, all that should work on defaults. mention that it works fine in alacritty in the bug. 2024-09-09 15:21:18 The other question I wanted to ask is how do some people get sub 100mb memory use out of alpine? My system is pretty lean I feel like, with not many services running, and I'm idling at 400mb. 2024-09-09 15:21:45 Which is a big improvement still from ubuntu, but I'm curious how they manage. 2024-09-09 15:22:25 ulimit would probably be the first thing 2024-09-09 15:24:14 Ah ok, I'm unfamiliar. Will check that out, ty. 2024-09-09 15:27:33 Anyway, thanks again. This was fun :D. I'm out though, my head hurts. Cheers. 2024-09-09 15:28:30 another satisfied customer 2024-09-09 15:56:13 andar1an[m]: hx --health shows gopls and zls as enabled either as root or as user 2024-09-09 15:56:40 I have it working somewhat now 2024-09-09 15:56:45 just messing with zellij now 2024-09-09 15:57:14 glad to hear you got it working 2024-09-09 15:58:35 The tree-sitter packages are a nice add, though I am curious if building from git works well too 2024-09-09 15:58:51 cargo installs for lsp's have been good because can just target musl 2024-09-09 16:00:07 thanks for help all 2024-09-09 17:41:48 if i install fro virt-aarch64 image will it install without necessary things for non virt machine? 2024-09-09 17:52:51 how avoid that? in case 2024-09-09 17:54:54 Why do you even use virt image for bare metal? 2024-09-09 18:47:15 cause i do not have aarch64 desktop 2024-09-09 18:49:02 i should use generic arm i guess 2024-09-09 18:49:41 https://dl-cdn.alpinelinux.org/alpine/v3.20/releases/aarch64/alpine-uboot-3.20.3-aarch64.tar.gz for pinephone 2024-09-09 19:07:42 (i forwared sd to virt machine) 2024-09-09 19:07:53 forwarded 2024-09-09 19:09:14 if install uboot tar.gz and I use one partition do I mark it bootable? 2024-09-09 19:09:30 parted ${DISK_IMAGE} set 1 boot on? 2024-09-09 19:21:21 seems so in https://wiki.alpinelinux.org/wiki/Alpine_on_ARM 2024-09-09 19:22:00 wait but tar.gz is not filesystem 2024-09-09 19:47:00 Which timezone are you in? [UTC] 'UTC' is not a valid timezone on this system damn 2024-09-09 19:47:06 i guess i have to start over 2024-09-09 19:47:39 Sounds like you don't have a working network connecting 2024-09-09 19:47:42 connection 2024-09-09 19:50:18 i was able to ping 2024-09-09 19:51:52 wiki states something about doing it mnually 2024-09-09 19:53:22 https://docs.alpinelinux.org/user-handbook/0.1a/Installing/manual.html#_keyboard_layout 2024-09-09 20:09:33 setup-disk coul have boot loader and kernel option 2024-09-09 20:22:21 how install packages as deps of one package? 2024-09-09 20:24:48 -t 2024-09-10 03:32:05 Argentum is looking for experienced IRC operators to moderate our #politics channel on Snoonet. please /join #help on irc.snoonet.org for further details. 2024-09-10 05:11:45 Anyone else having issues with wifi cards recently? My driver has crashed twice now, and it recovered by resetting the hardware, with a new wlanN interface. 2024-09-10 05:13:19 Haven't noticed anything yet on my laptop 2024-09-10 05:25:00 "with wifi cards recently" is extremely broad 2024-09-10 05:26:31 https://bugzilla.kernel.org/buglist.cgi?bug_status=NEW&bug_status=ASSIGNED&bug_status=NEEDINFO&bug_status=REOPENED&field0-0-0=product&field0-0-1=component&field0-0-2=alias&field0-0-3=short_desc&field0-0-4=status_whiteboard&field0-0-5=content&field1-0-0=alias&field1-0-1=short_desc&field1-0-2=status_whiteboard&field1-0-3=content&field2-0-0=product&field2-0-1=component&field2-0-2=a 2024-09-10 05:26:33 lias&field2-0-3=short_desc&field2-0-4=status_whiteboard&field2-0-5=content&field3-0-0=product&field3-0-1=component&field3-0-2=alias&field3-0-3=short_desc&field3-0-4=status_whiteboard&field3-0-5=content&no_redirect=1&order=changeddate%20DESC%2Cbug_status%2Cpriority%2Cassigned_to%2Cbug_id&query_format=advanced&type0-0-0=substring&type0-0-1=substring&type0-0-2=substring&type0-0-3=s 2024-09-10 05:26:35 ubstring&type0-0-4=substring&type0-0-5=matches&type1-0-0=substring&type1-0-1=substring&type1-0-2=substring&type1-0-3=matches&type2-0-0=substring&type2-0-1=substring&type2-0-2=substring&type2-0-3=substring&type2-0-4=substring&type2-0-5=matches&type3-0-0=substring&type3-0-1=substring&type3-0-2=substring&type3-0-3=substring&type3-0-4=substring&type3-0-5=matches&value0-0-0=Microcode 2024-09-10 05:26:37 &value0-0-1=Microcode&value0-0-2=Microcode&value0-0-3=Microcode&value0-0-4=Microcode&value0-0-5=%22Microcode%22&value1-0-0=SW&value1-0-1=SW&value1-0-2=SW&value1-0-3=%22SW%22&value2-0-0=error&value2-0-1=error&value2-0-2=error&value2-0-3=error&value2-0-4=error&value2-0-5=%22error%22&value3-0-0=detected&value3-0-1=detected&value3-0-2=detected&value3-0-3=detected&value3-0-4=detected 2024-09-10 05:26:39 &value3-0-5=%22detected%22 2024-09-10 05:26:41 Wow that url... o.o 2024-09-10 05:28:31 [974821.327277] iwlwifi 0000:00:14.3: WRT: Collecting data: ini trigger 4 fired (delay=0ms). 2024-09-10 05:28:34 [974821.327294] ieee80211 phy1: Hardware restart was requested 2024-09-10 06:09:47 intel wifi drivers suck 2024-09-10 06:14:02 dwfreed: amd bluetooth pedestrians lick 2024-09-10 11:07:48 how can I install alpine on pinephone? i see that linux-lts has it https://pkgs.alpinelinux.org/contents?file=*pinepho*.dtb&path=&name=&branch=edge&repo=main 2024-09-10 11:07:55 setup-disk install grub 2024-09-10 11:09:24 but i did not configure BOOTLOADER 2024-09-10 11:50:18 scorpion2185[m]: you might want to look at pmOS as a starting point. They carry several patches for audio and modem to work. 2024-09-10 17:00:49 yes, but they have different initram and custom u-boot. for now I am trying to boot 2024-09-10 17:01:17 tried before to install all 2024-09-10 17:28:20 looks like that there is a bug with aarch64 standard image 2024-09-10 17:28:34 when setting timezone I was stuck again. 2024-09-10 19:13:09 hi 2024-09-10 19:13:45 is alpine x86 only? 2024-09-10 19:20:33 see the downloads page 2024-09-10 19:20:36 would you be offended if i told you to just look at the website 2024-09-10 19:20:53 ahh, beat me to it :D 2024-09-10 19:49:51 Howdy friends, I'm doing a baremetal install of Alpine and I'd like to use efibootmgr instead of grub -- is there a preferred way to set that when running setup-alpine? 2024-09-10 19:51:25 l0rd_hex_: I don't think setup-alpine itself supports it, but you can do the disk layout (incudling bootloader) manually, mount the root partition, and then run setup-disk /path/to/mount 2024-09-10 19:52:33 ikke: thanks, it's been a while since I've tried it but I think setup-disk wants to install grub too 2024-09-10 19:53:00 l0rd_hex: it should not do that if you provide a path to a mounted partition 2024-09-10 19:54:37 In anycase if you provide BOOTLOADER=none 2024-09-10 19:56:42 thanks, I'll give that a shot 2024-09-10 20:09:44 any suggestion to how boot pinephone? 2024-09-10 20:10:27 scorpion2185[m]: WhyNotHugo gave you the advice to look at PostmarketOS 2024-09-10 20:10:36 which is dedicated to phones 2024-09-10 20:11:03 they have different initram and custom u-boot 2024-09-10 20:12:07 i should check u-boot package. i think PP is like some pine board 2024-09-10 20:12:20 supporting a bunch of arm boards requires doing stuff like that 2024-09-10 20:12:34 does somebody know how set grub + uboot (arch64)? 2024-09-10 20:12:48 i think pmos does that for PPP 2024-09-10 20:12:51 alpine is probably never going to support booting a bunch of phones 2024-09-10 20:13:07 bbl 2024-09-10 20:33:30 I wanna put alpine on a usb, on a readonly mounted partition and have a separate home, var and other such partitions where I'd only ever update it atomically, i.e. chroot, perform updates, write and then reboot, is data disk mode suitable for this? 2024-09-10 20:34:13 ashie: not sure if it's suits you, but that sounds a lot like a run-from-ram system where changes are persisted using lbu 2024-09-10 20:43:45 ikke: I did not know about lbu, do I understand it correctly that `commit` creates an overlay file and when the system boots it runs through all the overlays and applies them? 2024-09-10 20:44:02 It's just a single overlay, but yes 2024-09-10 20:46:31 so a single overlay file that gets updated then, I see... how does updating the kernel work? I assume that cannot go into an overlay file because it is needed to boot the system 2024-09-10 20:47:29 correct. Traditionally that would mean you have to update the boot media 2024-09-10 20:47:48 There is an update-kernel script to update it on the fly, but that would require the boot media to be writable 2024-09-10 20:50:57 if everything from the boot media gets loaded into ram does would the partition still be available to be mounted and written to? 2024-09-10 20:51:12 s/does // 2024-09-10 20:51:52 yes, if it's a suitable filesystem (not isofs) 2024-09-10 21:03:46 thanks a lot, I think I now know enough to figure stuff out 2024-09-10 21:57:29 We are looking for an experienced IRC operator to moderate our #politics channel on Snoonet. please see argentum in the channel #help on irc.snoonet.org for further details 2024-09-11 08:19:38 I think sof firmware in edge broke (again) in relation to kernel 2024-09-11 08:20:35 At least on the dell xps 9320 does not have sound anymore, neither with -lts nor -edge kernel 2024-09-11 12:39:29 hi 2024-09-11 12:40:51 someone uses mtp with alpine linux? 2024-09-11 12:43:27 can't write with simple-mtpfs to a samsung galaxy, read is ok 2024-09-11 12:51:56 i use simple-mtpfs and it works mostly. sometimes i need to replug to make it work. also the phone needs to be in mtp mode, not charge only... 2024-09-11 12:52:23 i mostly use it to download pix. but i also delete them, so that's writing... 2024-09-11 12:58:50 I was about to install alpine on an old macbook that was on the drawer but I just gave up because of the broadcom wireless LOL 2024-09-11 13:02:26 I have been using various GNU/Linux and Linux distributions for several years, with no Wi-Fi functionality at all 2024-09-11 13:02:44 Because I have a really long cord. 2024-09-11 13:04:32 what is "really old" 2024-09-11 13:04:41 oh, you said just old 2024-09-11 13:04:47 question remains 2024-09-11 13:05:49 if it's sufficiently old you just need firmware :P 2024-09-11 13:06:35 its an air 2015 2024-09-11 13:07:20 bcm4360 seem super annoying by what I have searched, I think it will go back to drawer 2024-09-11 13:08:19 that's not old at all 2024-09-11 13:08:31 and yeah probably annoying 2024-09-11 13:44:13 i got a few macbook airs to work on alpine. (people keep giving me these damn things when they upgrade, and i don't say no) 2024-09-11 13:45:18 offhand i can't recall specifics but it's mostly getting the wifi driver correct 2024-09-11 13:47:00 what i haven't done yet: booted alpine on m1 macs, like what asahi does 2024-09-11 13:47:10 I actually run Alpine on a 2012 Macbook Pro. 2024-09-11 13:47:47 It's like 20x faster to use for web browsing than macOS Catalina (which is what Apple provides for it to reinstall) 2024-09-11 14:11:52 hmm does alpine has a "live cd" for rescue operations? 2024-09-11 14:14:57 all of the install images are live cds i think. 2024-09-11 14:17:34 seem to all be minimal, but actually that might be enough. 2024-09-11 14:18:49 you can install packages if you boot the iso after setup-interfaces 2024-09-11 14:20:26 Or make your own with mkimage 2024-09-11 14:20:26 https://wiki.alpinelinux.org/wiki/How_to_make_a_custom_ISO_image_with_mkimage 2024-09-11 14:22:00 The extended image had more tools included 2024-09-11 14:23:14 ah right 2024-09-11 14:33:06 non of them have cryptsetup? 2024-09-11 14:37:08 You should be able to apk add cryptsetup if you've set up the network first. 2024-09-11 14:37:34 iirc all of them have cryptsetup, but i mostly use extended 2024-09-11 14:40:07 ikke is there an list of what is included? 2024-09-11 14:40:17 oh, just had to install it, worked without internet! 2024-09-11 14:47:10 frag: yes, check /etc/apk/repositories 2024-09-11 14:48:04 fabricionaweb: https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/scripts/mkimg.standard.sh?ref_type=heads#L37 2024-09-11 14:49:54 perfect, thank you 2024-09-11 15:07:31 hmm could lukOpen, but cant mount the mapper file, "invalid argument", e2fsck /dev/mapper/void, suggest superblock is corrupted .. 2024-09-11 15:08:15 managed to move /* to /1 on a void laptop, thats why im trying to do this... 2024-09-11 15:08:37 hard to force it to shutdown by holding down power button 2024-09-11 15:09:46 is alpin good for laptops btw? last time i tried i think it sapped the battery.. with void i can leave it in a shelf for a month 2024-09-11 15:12:26 *had to force it to shut down by holding the down power button 2024-09-11 15:17:06 it can be, it's just linux. void probably has more in its defaults 2024-09-11 15:18:25 it's rarely about whether it does or doesn't ... it mostly comes down to how much magic you expect and what you're willing to configure yourself. 2024-09-11 15:21:53 neither void nor alpine is likely to sap more battery than the other imo 2024-09-11 15:22:35 though what do you mean by "sap the battery" 2024-09-11 15:22:43 should've asked that first 2024-09-11 15:45:01 i thought it sounded like a suspend/hibernate issue to me. which anyways is tricky with luks... 2024-09-11 15:45:08 fwiw all i use is alpine as my desktop in both personal and $job. that's not a recommendation, or warranty. any linux can do it, the "how" varies. 2024-09-11 16:42:49 even after mkfs.ext4 /dev/sda, and running setup-alpine and choosing sda, crypt, it asks me to enter passphrase for _sda3_??? 2024-09-11 16:47:23 yes, probably 2024-09-11 16:47:41 sda1 is /boot, sda2 is swap and sda3 is / 2024-09-11 16:47:53 if you used the alpine-setup method 2024-09-11 16:48:44 yeah doesn't seem wrong 2024-09-11 16:48:59 oh, thought it used the old partition table :S 2024-09-11 22:51:51 Is there any interest in default SELinux policies as an optional feature for Alpine Linux? 2024-09-11 22:52:10 We'd probably end up yoinking a bunch of them from other distros, probably with some patches 2024-09-11 22:52:26 i'm concerned though because it seems like borderline a full time job to maintain SELinux stuff 2024-09-11 22:55:25 +1 for interest. Having the kernel tuned a bit more for security (a13xp0p0v/kernel-hardening-checker) to ensure SELinux rules would be effective might something to look into first. 2024-09-11 23:02:38 chexo4[m]: probably should be a gitlab issue for people to comment on. last time was 6 years ago: https://gitlab.alpinelinux.org/alpine/aports/-/issues/8606 2024-09-12 00:12:05 "chexo4: probably should be a..." <- You have one for s6-init/rc? I don’t know how gitlab search works 2024-09-12 02:25:01 chexo4[m]: https://gitlab.alpinelinux.org/alpine/tsc/-/issues/78#note_374794 2024-09-12 02:49:56 "chexo4: https://gitlab.alpinelin..." <- Nice to see relatively recent work on that front 2024-09-12 03:43:37 hey everyone, i'm a newbie to Alpine and trying to configure Tor, i'm having a problem where Tor can't bind to any address, an error i haven't encountered in other distros, is there anything Alpine specific that must be done to allow binding to addresses? 2024-09-12 03:44:03 the exact warning message i get; 2024-09-12 03:45:05 Sep 12 [warn] Could not bind to 0.0.0.0:443: Permission denied, and Sep 12 [warn] Could not bind to 2a0c:b641:3a1:1001::103:443: Permission denied 2024-09-12 03:53:00 omg.. sorry guys, i forgot to run it with doas/root, ignore me >.> 2024-09-12 03:59:35 sorry for the spam here, but it seems like the 2024-09-12 04:16:14 aaa, never mind, i made a typo in my net config, my bad!! tor works perfectly now 2024-09-12 05:11:15 :) 2024-09-12 07:21:44 if you want to bind to low ports without requiring special priviledges you can use `sysctl net/ipv4/ip_unprivileged_port_start=0` to disable this feature altogether 2024-09-12 07:22:18 but that means others can bind to it as well 2024-09-12 07:31:25 yeah, don't do that 2024-09-12 07:32:21 if you want something to be able to bind to a low port, but don't want it to run as root and it has no privilege management of its own, you can give it CAP_NET_BIND_SERVICE file capability 2024-09-12 07:35:06 well, yeah but if this is a desktop machine it's not really a issue 2024-09-12 07:39:28 Yeah, it's only connected to the Internet 2024-09-12 17:16:55 how do I build u-boot for pinephone? 2024-09-12 17:38:00 apk del grub 2024-09-12 17:38:00 grub: grub-efi 2024-09-12 17:38:00 World updated, but the following packages are not removed due to: 2024-09-12 17:38:05 apk del grub gru-efi 2024-09-12 17:38:05 ERROR: No such package: gru-efi 2024-09-12 17:38:55 typo: grub-efi 2024-09-12 17:40:30 how do I get to boot? i have pmos kernel + u-boot-pinephone 2024-09-12 17:41:19 still EFI and grub folders in /boot after deleteing 2024-09-12 17:42:57 Remove them then. They're created by grub-install, not by apk 2024-09-12 17:45:19 ubootefivar.efi too? 2024-09-12 17:45:36 i mean  ubootefi.var 2024-09-12 18:19:12 now tow-boot does not find anythign 2024-09-12 21:49:30 What's wrong with the iptabel in alpine 2024-09-12 21:49:52 The sh*t always contradict with docker 2024-09-12 21:50:37 What is the alternative? To get rid of or modify  2024-09-12 22:22:25 this bug report sucks 2024-09-13 08:23:19 Velo_Frost[m]: whats your problem exactly? docker handles iptables itself, if you want to add some custom rules you should use DOCKER-USER chain 2024-09-13 11:10:45 do need to build Device Tree Overlay for pinephone? i see some in apline linux-lts pacakge 2024-09-13 12:20:45 Hi 2024-09-13 12:21:20 I created an Alpine chroot with alpine-chroot-install script and created a user there 2024-09-13 12:21:28 It's in the audio group 2024-09-13 12:21:35 But it's not able to access alsa device(s) 2024-09-13 12:21:46 The root user in the same chroot is able too 2024-09-13 12:21:56 Anything obvious I'm missing? 2024-09-13 12:22:05 (note that the host isn't an Alpine) 2024-09-13 12:25:18 Yeah ok, group id (of audio) mismatches :) 2024-09-13 15:32:04 What's the easiest way to create a new partition table on target device? 2024-09-13 15:42:53 what was in the backgroud/wallpaper for older xorgs? was it a repeating small image? which one? or any other small repeating image thats nice? 2024-09-13 16:04:36 idk maybe parted 2024-09-13 16:06:31 https://wiki.alpinelinux.org/wiki/DIY_Fully_working_Alpine_Linux_for_Allwinner_and_Other_ARM_SOCs is obsolete 2024-09-13 17:19:26 "Velo_Frost: whats your problem..." <- OK, I will look into it. But will news you help, kindly 2024-09-13 22:36:22 hi! i got a pre-existing virt-manager VM after migrating from a different distribution running windows 11. how do i install ovmf-x86_64-smm-ms-code so i can launch it again? it seems like alpine by default only installs ovmf variants without the ms secure boot code signing, but i dont think it's possible to run windows 11 with that (and even less so my preinstalled VM since apparently it's not possible to just easily swap the ovmf firmware 2024-09-13 22:36:22 on an existing install, for whatever reason). if anybody knows how to install the right ovmf variant on alpine, you'd help me a lot. disclaimer: i'm using postmarketOS on x64 but it uses alpine packages under the hood 2024-09-13 22:43:10 el[m]: I can't speak to your pre-existing VM image, but in general Win 11 can run fine with regular ovmf 2024-09-13 22:47:38 iggy: i tried forever to set it up without that and couldnt make it work on opensuse. the installer doesnt consider it a compatible machine 2024-09-13 22:48:17 it's only happy with the tpm2 nonsense enabled (which seems to work fine on alpine) and ms signed secure boot enabled (which i'm currently stuck on) 2024-09-13 22:48:37 I just had to enable the fake tpm 2024-09-13 22:50:00 🤷‍♀️ 2024-09-13 22:51:01 it doesnt seem to be possible to switch over ovmf firmware even if there would have been a way. which might be true, who knows, win11 is kind of a shitshow anyway (my personal opinion) 2024-09-13 22:51:19 so in conclusion i need the ms signed secureboot no matter what 2024-09-13 22:52:40 maybe just grab it from the qemu repo and don't worry about it not being packaged? 2024-09-13 22:57:05 iggy: oh, is that possible? tbh i got no idea about how this ovmf stuff works. i'd gladly do that 2024-09-13 23:00:14 https://github.com/qemu/qemu/tree/master/pc-bios I think you need the edk2-x86_64-secure-code.fd.bz2 file (but I could be wrong) 2024-09-13 23:07:43 "maybe just grab it from the qemu..." <- i downloaded the opensuse package now and extracted it from there 😂 it actually worked! thanks so much for the help, you're awesome 2024-09-13 23:08:01 i had to copy over the /usr/share/qemu firmware .bin and the /usr/share/qemu/firmware firmware json and now it works 2024-09-14 00:27:29 i never thought about this before, but as i delve deeper into alpine's boot, initialization, and startup, it appears to me that booting without an initramfs is not supported. 2024-09-14 00:35:20 i cannot find anywhere in the rc scripts that proc gets mounted, only in the initramfs-init 2024-09-14 01:23:53 j_v: /lib/openrc/sh/init.sh will mount proc if it's not mounted already 2024-09-14 01:24:22 note, however, that various alpine features are implemented in the initramfs 2024-09-14 01:25:12 thanks, appreciate the pointer and the note 2024-09-14 08:04:51 https://github.com/flatpak/flatpak/issues/5926 there seems to be a packaging issue with flatpak on alpine edge (and by extension, postmarketos edge) that breaks ostree 2024-09-14 10:26:19 el[m]: the issue is ostree and curl 8.10.0-r1, some discussion is here: https://gitlab.alpinelinux.org/alpine/aports/-/issues/16444 2024-09-14 12:44:13 hi 2024-09-14 13:12:52 Morning all, I am noticing a recent problem on my system where the shutdown runlevel change, either with reboot or poweroff will result in the system hanging when trying to power off or reboot. Any ideas what I may look at to understand why this is happening? 2024-09-14 13:29:04 Historically I would see signal 15 caught on screen (i believe SIGTERM) on shutdown, but now I just see blinking underscore 2024-09-14 14:01:13 i'm working on setting up jellyfin on alpine linux on intel. is there anything i need to do to enable the intel hardware acceleration besides what's in this wiki ? https://wiki.alpinelinux.org/wiki/Intel_Video 2024-09-14 14:13:35 Couod cross reference with: https://jellyfin.org/docs/general/administration/hardware-acceleration/ 2024-09-14 15:15:30 Hi, I am new to alpine, and I have got a start up dependency issue. My VPS runs on Alpine edge and I installed crowdsec and cs-firewall-bouncer. The bouncer always crashes when the system boots because it cannot join the crowdsec deamon, because it is still starting and does not accept api request yet. I tried to add `after crowdsec` in the depence() function of the bouncer's init script but it did not help. Does someone kno 2024-09-14 15:18:11 1) you shouldn't edit initscripts; pretty much all changes should be possible in conf.d files; dependency changes definitely are; 2) add a start_post to /etc/conf.d/crowdsec that waits for crowdsec to be functional before returning, or do the same for cs-firewall-bouncer, but use start_pre instead 2024-09-14 15:27:22 Oh I did not see the start_post options before, 2024-09-14 15:27:58 It should definitly do the trick, thank you 2024-09-14 15:42:13 Yes, "while ! cscli lapi status" in start_post fixed my issue. Thank you again 2024-09-14 15:42:37 see, this is what readiness notification is for 2024-09-14 16:39:14 seems qutebrowser also crashes when trying to open the file-chooser 2024-09-14 17:37:10 is curlftpfs known to work? i'm getting a strange "Error setting curl:" error with an empty string no matte what i try 2024-09-14 17:37:31 googling suggests this means curl didnt like the protocol, is curl for alpine compiled without ftp or something, and maybe this is breaking curlftpfs? 2024-09-14 17:49:42 y 2024-09-14 17:52:36 anyone familiar with depmod / module load priority? I have a module that's built from source using akms which is a newer version than the in-tree version, however, modprobe seems to prioritize the in-tre version. 2024-09-14 17:53:16 I've tried to add this to /etc/depmod.d/drbd.conf, but that does not seem to make a difference: https://tpaste.us/PQq7 2024-09-14 17:53:40 If I move the in-tree drbd module somewhere else, modprobe will load the newer version, so the module can be loaded 2024-09-14 17:54:32 depmod -v does not show any module in extra 2024-09-14 17:54:56 ftr, this is linux 6.6.51-0-virt 2024-09-14 19:19:42 hm 2024-09-14 19:20:40 so i managed to abuild something, apk list fnt shows fnt-1.6-r0 noarch {fnt} (MIT) [installed] .. but well apk add ./fnt-1.6-r0.apk said 1 error. and find / name fnt does not find it somehow 2024-09-14 19:22:27 alex_: what does `apk fix` return? 2024-09-14 19:23:14 (1/1) [APK unavailable, skipped] Reinstalling fnt (1.6-r0) \n 1 error; 1810 MiB in 687 packages 2024-09-14 19:24:16 btw i've failed to get x and wayland stuff working in my utm arm64, then also x86_64, now i've got an usb stick with alpinelinux trying on baremetal... 2024-09-14 19:26:21 before you ask, here's my APKBUILD: http://bananas.debian.net/apk/fnt.txt 2024-09-14 19:28:17 alex_: did the apk add not show any other output? 2024-09-14 19:29:24 no 2024-09-14 19:37:44 of course that stone age core 2 duo bare metal would keep rebooting when trying to boot alpine linux x86_64 :( 2024-09-14 19:38:35 That's not stone age though 2024-09-14 19:38:58 When is it rebooting? 2024-09-14 19:39:14 when you quit x86 4 years ago, it is, but true, subjective 2024-09-14 19:39:42 while pressing the option key to get the boot menu. it just kept rebooting, tried twice, gave up 2024-09-14 19:40:39 So in isolinux? 2024-09-14 19:41:02 screen just gets white. nothing visible 2024-09-14 19:42:07 if it helps, tried with: alpine-standard-3.20.2-x86_64.iso 2024-09-14 22:23:20 FYI i am TrueThough[m] . I am more comfortable with a classic IRC client than matrix. 2024-09-14 22:23:55 and matrix says i'm not really on anymore anyway. confusing lol 2024-09-14 23:08:27 alright so I tried to just edit the repositories but it wasn't that simple. so i installed from scratch using alpine-virt-3.20.3-x86_64.iso into vmware workstation pro 16. i followed https://wiki.alpinelinux.org/wiki/Xfce to install xfce. during the install i saw a lot of items throwing "IO ERROR" and failing. the end of setup-desktop said 48 errors, as well as a bunch of relocation errors for .so objects. 2024-09-14 23:08:45 on reboot lightdm tried to start, however the user for it wasn't found and it died... did I do something wrong? 2024-09-14 23:09:13 i investigated for any disk / filesystem errors and don't find any 2024-09-15 07:11:20 Hello! I have a problem with LightDM. It suddenly stopped working. The only line that indicates an error is in /var/log/LightDM/seat0-greeter.log: "Failed to open PAM session: Permission denied". I had this issue yesterday, reinstalled Alpine from scratch, then installed and deleted kwin and now I have the issue again 2024-09-15 07:22:46 I'm using edge repos, if that helps 2024-09-15 08:17:47 Hello folks, a question about ufw & iptables... I use the former, and it seems to be working, but I don't have the iptables service enabled. Should I enable it, or does this conflict with the ufw service? 2024-09-15 08:21:13 https://wiki.alpinelinux.org/wiki/Uncomplicated_Firewall 2024-09-15 08:37:52 @quing I've read that page, it mentions ip6tables (which AFAIK is no longer necessary), it does not mention iptables nor the service - hence my question. 2024-09-15 08:52:43 Hey everyone! I'd like to install alpine on an orange pi, but i need the latest kernel 6.10 for that - currently `linux-edge` in the `edge` repo, so I'm considering activating the "edge" repository. 2024-09-15 08:53:17 The question is, is it possible to get back to stable alpine version once 6.10 hits a released alpine version? Or am I bound to stick to the edge repo? 2024-09-15 08:59:31 I don't think this have been tested enough, so you might bump into hurdles. but theoretically it should be as easy as sed -i -e 's/edge/latest-stable/g' /etc/apk/repositories 2024-09-15 09:00:02 I see, thanks for the info! 2024-09-15 09:01:29 usagi: also if you read the wiki, it says that edge is not for end users since it contains a lot of packages that are not heavily tested. so in my opinion, you're risking to get broken packages from time to time (which is not ideal) .. if i were you, i'd get 6.10 by building the APKBUILD while still staying on stable. 2024-09-15 09:01:48 let me know if you need help on that one. 2024-09-15 09:09:09 Hmm, staying on a stable version and justing building the kernel, that's not a bad idea... Never used APKBUILD but it was on my TODO list since a while, might as well use this use case to get to learn it :) 2024-09-15 09:27:54 @quinq I've read that page, it mentions ip6tables (which AFAIK is no longer necessary), it does not mention iptables nor the service - hence my question. 2024-09-15 10:02:24 pu, hence the answer ;) 2024-09-15 10:02:32 If it's not mention, that means it's not necessary 2024-09-15 10:06:24 quinq yes maybe, or simply forgotten... might be a good reason for me to create a user on the wiki and contribute, but my half-knowledge of Alpine needs to improve 2024-09-15 10:27:16 the iptables and ip6tables service serves similar roles 2024-09-15 10:43:03 pu, to check, you can also read the ufw init script, where you can see that it depends on iptables init script 2024-09-15 10:58:33 quinq I did, it's a "after" non-dependency, so I guess that the right way to read it is that "you can, but don't have to, have that service started". 2024-09-15 10:58:50 pu: it's not a traditional service that keeps running 2024-09-15 10:59:00 all the ip(6)tables services do is persist the rules 2024-09-15 10:59:09 so restore on boot and save on shutdown 2024-09-15 11:00:00 if uwf does not have something to do that, then you need those services to make sure the rules are present on boot 2024-09-15 11:02:15 ikke so I guess that the ufw service takes care of persistence as well, though I haven't yet found out where 2024-09-15 11:04:17 probably /usr/lib/ufw/ufw-init-functions 2024-09-15 11:06:06 to use extlinux.conf to boot what do I do? apk add syslinux no such package 2024-09-15 11:06:34 ah no aarch64 2024-09-15 11:06:52 Yeah, that only works for x86(_64) 2024-09-15 11:07:16 isn't pmos using it for PPP? 2024-09-15 11:07:56 > or because extlinux.conf has been deleted (if only upgrading the existing install without installing Tow-Boot). 2024-09-15 11:08:00 https://postmarketos.org/edge/2022/03/28/ppp-tow-boot/ 2024-09-15 11:08:30 also mobian has extlinux.conf file 2024-09-15 11:08:41 no pmos package too https://pkgs.postmarketos.org/packages?name=*sysli*&branch=master&origin=&maintainer= 2024-09-15 11:11:37 therefore trying to boot alpine with extlinux.conf won't work right? 2024-09-15 11:12:04 Not on an aarch64 system, as far as I know, no 2024-09-15 11:17:09 anyone knows to how make grub + tow boot work? 2024-09-15 11:18:20 scorpion2185[m]: You should really ask the pmos folks, they have a lot more experience with these setups 2024-09-15 11:20:07 i see thanks 2024-09-15 12:29:13 hello if it is not much can anyone help me install cosmic de 2024-09-15 12:32:35 it is not in repos and i guess have to compile, so i dont want to compile it and find out it is not working 2024-09-15 12:36:27 cosmic epoch github page have some guidance but it is mostly for systemd system but it also has gentoo linux instruction so i am guessing it also works with openrc systems 2024-09-15 13:27:23 gentoo usually has openrc 2024-09-15 13:27:41 never mind 2024-09-15 13:29:26 cosmic is in alpha i think that it is better to wait and it's all rust it's gonna be havey to compile it 2024-09-15 13:29:32 *heavy 2024-09-15 13:42:05 "Currently an incomplete alpha" 2024-09-15 13:42:51 i know it is incomplete alpha 2024-09-15 13:49:13 i would suppose the primary audience for building it is devs who are used to getting things to build 2024-09-15 13:50:57 it also looks like that wants to use mold as a linker 2024-09-15 13:51:17 Installing system on /dev/vdb2: Installing for arm64-efi platform. 2024-09-15 13:51:17 install: can't stat '/EFI/alpine/grubaa64.efi': No such file or directory 2024-09-15 13:51:17 grub-install: error: failed to get canonical path of `'. 2024-09-15 14:00:21 i guess i need to wipe out disk 2024-09-15 18:35:47 Joystick isn't working with retroarch. I can read /dev/input/js0, and other applications (e.g.: steam games) work. 2024-09-15 18:35:52 Any ideas what I could be missing? 2024-09-15 18:40:11 so it's a retroarch issue? 2024-09-15 19:23:33 WhyNotHugo, any diagnosis from retroarch? 2024-09-15 20:18:04 Nothing unusual from --verbose: https://paste.sr.ht/~whynothugo/100172dcfb17c2ef1a09b97a7c4c085b69912a0e 2024-09-15 20:18:07 Not sure what else to check. 2024-09-15 21:14:58 Good evening o/ 2024-09-15 21:15:53 hi coredumb 2024-09-15 21:15:54 Was wondering what's the best way to have root disk encrypted with hibernation support on Alpine? wiki seems to have different resources pointing in different directions 2024-09-15 21:17:00 Been able to easily add luks encryption on root btrfs but swap and hibernation settings seems more hit and miss 2024-09-15 21:17:36 Or more accurately "btrfs root on luks encrypted partition" 2024-09-15 21:19:56 IMO the sanest option is using an LV for swap and another for root, both on the encrypted PV, since mounting root to either open a swap file or key file between suspend and resume is potentially destructive 2024-09-15 21:21:05 I guess LV under btrfs kinda defeats it but this makes sense 2024-09-15 21:21:43 What bootloader should I aim for on EFI platform? 2024-09-15 21:23:31 does elilo work 2024-09-15 21:26:11 Isn't elilo discontinued? 2024-09-15 21:26:43 was the libcurl downgrade fix to make flatpak work again meant to have shipped already? i saw it merged about a day ago, but it doesnt seem to have reached my machine yet. sorry if this is a silly question and roundtrips until it ends up in the repos are normal to take that long 2024-09-15 21:29:15 coredumb: i think grub works, i had it on my UEFI system 2024-09-15 21:31:46 Ok. Guess I'll give the PV under btrfs thing a try :D 2024-09-15 22:39:17 hi 2024-09-15 22:40:52 hi test99998888[m] 2024-09-16 01:36:58 I would like to install librewolf/chromium/firefox/any-modern-browser, but they all depend on mess-24.2.2 while I have mesa-asahi-24.0.0 2024-09-16 01:37:15 Can I somehow make apk install the browsers anyway? 2024-09-16 05:17:03 runxiyu_: You can try: `apk add -t mesa-egl; apk add -t mesa-gl`; 2024-09-16 05:17:12 This will install these packages as a virtual package without dependencies 2024-09-16 06:06:10 ah, forgot -t existed, will try. thanks! 2024-09-16 06:07:12 Also, I'm happily using iwd, except that OpenRC likes to start networking (which launches udhcpc) before iwd... 2024-09-16 06:10:47 ikke: Unfortunately, mesa-egl still conflicts with mesa-asshi-egl, even when installed with -t 2024-09-16 07:48:50 hmm the default PATH starts with a dir that doesnt exist: /usr/local/sbin 2024-09-16 07:50:53 could be made in the future some said.. 2024-09-16 08:15:44 coredumb: the easiest approach is to use a swapfile 2024-09-16 08:16:44 imho, LVM is too much complexity if you just have a single disk with a single partition. 2024-09-16 08:24:10 frag, I have them here and didn't make them manually 2024-09-16 08:27:16 quinq hmm, weird 2024-09-16 08:28:14 runxiyu_: conflicted how? If it is installed as a virtual package, it should not conflict with anything 2024-09-16 08:29:27 It's part of alpine-baselayout, frag 2024-09-16 08:30:02 oh, i dont always follow the standard...... also have encrypted rpi 2024-09-16 08:30:14 guess i dont need that dir tho 2024-09-16 08:30:35 Not technically indeed 2024-09-16 09:08:15 ikke: not sure now as my tmux log doesn't go back far... but it works after i blindly install all packages that begin with mesa-asahi, rather than only installing ^mesa-asahi$ 2024-09-16 09:13:48 Yeah, you would need to explicitly install those subpackages 2024-09-16 09:14:30 Ideally it would be fixed in the mesa and mesa-asahi packages so that you can exchange them 2024-09-16 11:38:00 im i the only one experiencing calling the file-chooser causes applications to crash on edge? 2024-09-16 12:21:30 morning all 2024-09-16 12:25:45 Installing alpine extended 3.20.3 on surface pro, I'm creating a 350M esp partition, is enough room? 2024-09-16 12:48:39 owain: yeah, 350MB should be enough for most usecases 2024-09-16 12:53:10 after doing a manual install, no setup-alpine, my esp partition is empty. I'm thinking the apk cashe is too big and causing a write failure. 2024-09-16 13:04:08 is it possible to encrypt a partion without having to create a seperate partion for the bootloader? i have a bootable device where everything is on a single partion 2024-09-16 13:12:59 It is possible, test99998888[m], you only need to decrypt a file-system when you want to mount it 2024-09-16 13:13:51 So you don't need anything about boot, just decrypt them whenever they need to be accessed once the system is booted 2024-09-16 13:15:04 some people put their bootloader on removable media (eg usb stick). it can be done a lot of ways. 2024-09-16 13:18:14 it runs entirely on ram and the bootloader is on the usb. using setup-bootable 2024-09-16 13:47:36 Hello. AFAIU https://wiki.alpinelinux.org/wiki/LVM_on_LUKS#Luks2 this part shouldn't be necessary since Grub 2.12 right? grub pre config was only needed on pre-2.12 version IIUC? 2024-09-16 13:51:30 grub boot always failing on decrypting my luks2 partition telling me invalid passphrase :( 2024-09-16 13:52:58 I made sure to use --pbkdf pbkdf2 to format partition and made sure grub has all proper modules 2024-09-16 13:53:20 afaik there are not many people who use & understand luks2 very well (i'm definitely not one of them) 2024-09-16 13:54:13 you would probably have better luck getting this answered upstream 2024-09-16 13:55:25 does anyone have an opinion about preferred logging daemon on alpine? 2024-09-16 13:55:51 right now I am just using syslog, but the wiki mentions a few others 2024-09-16 13:56:35 ACTION uploaded an image: (107KiB) < https://matrix.org/oftc/media/v1/media/download/AVGd0mpNxSnmBSfsnlCmvpIUuNCptxBJcRWo4hWUaCZUvcl93hwBkLqIe8u-YRsBj_PinChkHt5UteSNxw2E45tCeR_NZl4AAG1hdHJpeC5vcmcvSGxXSVZlcWZ1dVpqTkxVZEZ1VWxydENm > 2024-09-16 13:56:41 Can someone help me with this? 2024-09-16 13:57:27 Velo_Frost[m]: this channel is primarily irc. some people aren't going to click on your matrix links. you should probably explain your problem 2024-09-16 14:00:08 @_oftc_invoked:matrix.org sorry. I am stuck in initramfs mode. With some errors. I'm sharing in web file, plz 2024-09-16 14:01:58 https://ibb.co/4KW1Fjn 2024-09-16 14:03:07 it can't find your cryptkey. 2024-09-16 14:04:18 Any way to resolve it via rescue mode 2024-09-16 14:05:46 either the file /crypto_keyfile.bin is missing, or the file is not what cryptkey expects, or maybe there's a problem with the fs 2024-09-16 14:07:42 is there any reason the initramfs says "busybox v1.35.0 (Debian ...)"? That's not normal for Alpine, is it? 2024-09-16 14:08:24 i didn't even see that :) 2024-09-16 14:13:55 Debian setup.after alpine 2024-09-16 14:36:03 how would i encrypt a file/directory 2024-09-16 17:36:11 You can use fscrypt if you use ext4 2024-09-16 17:37:08 Btrfs doesn't support fscrypt yet, but it is in works. I am sure there are other options too. KDE has "Vaults" if you want something more gui 2024-09-16 17:43:15 If for some other function related to server side encryption you could also look at oss tools like minio which have SSE features compatible with S3: https://min.io/docs/minio/linux/operations/server-side-encryption.html 2024-09-16 17:43:45 There are other options than minio too, but that is what I have looked into most, so that was fastest link I could find. 2024-09-16 17:44:51 https://github.com/KDE/plasma-vault 2024-09-16 17:45:38 Probably many tools for directory or file level encryption. But personally I was into fscrypt because it is pretty easy with ext4 by the looks of things 2024-09-16 17:46:27 gnupg can encrypt files (gpg) 2024-09-16 17:48:33 Guess it depends on encryption use case. You could also use an archive tool if you just wanted to manually encrypt some stuff. 2024-09-16 19:03:49 Hi, in the past I've read about Alpine changing its init/service_manager from OpenRC to something else (I think something s6 based was mentioned). Am I right? I'm asking just because I'm curious. 2024-09-16 19:04:37 aelspire: No decission has been made yet 2024-09-16 19:05:47 Ah, ok thanks it was bugging me as I remember reading something like this but I cannot find any clue about it in docs I managed to find 2024-09-16 19:05:50 I like runit, tested on void, pretty handy 2024-09-16 19:06:49 aelspire: perhaps https://skarnet.com/projects/service-manager.html? 2024-09-16 19:07:18 I've used runit for a while on Void and KISS, but I don't have opinion except: "not systemd" and even this opinion is not strong one 2024-09-16 19:07:55 openrc gained an additional maintainer who is working on implementing new features 2024-09-16 19:09:03 ikke: Yes this is what I've read, so it was not official source at all 2024-09-16 19:09:26 great to know that OpenRC gained more hands! 2024-09-16 19:09:37 aelspire: That's about developing a service manager 2024-09-16 20:45:37 Installing alpine extended 3.20.3 on surface pro, I'm created a 350M esp partition on the harddrive. I ran setup-disk /mnt, updated the fstab file, I mount lbu on a separate mounted partition. When I rebooted the esp partition was empty. Do I need to increase the size to 1G? 2024-09-16 21:40:37 i don't think size is the issue here 2024-09-16 21:44:33 I'm installing right now, will know shortly. 2024-09-16 21:53:10 507M 2024-09-16 23:15:43 So I sorted out my full disk encryption, so far so good. 2024-09-16 23:16:09 Testing in a VM though I'm trying to test if suspend/hibernate works 2024-09-16 23:17:23 every attempts, seems to reload initrd and just freeze on "Loading initial ramdisk" 2024-09-16 23:17:43 be it suspend to ram or to disk mind you :x 2024-09-16 23:18:45 Any idea? 2024-09-16 23:38:51 well, that's a complicated topic. 2024-09-16 23:39:44 i only have a couple minutes so i'll try to keep this simple. the system needs swap, for starters 2024-09-16 23:40:32 without swap it doesn't work 2024-09-16 23:42:07 getting suspend/hibernate to work with encrypted swap is ... tricky, and this will also come down to hardware quirks 2024-09-16 23:42:34 generally i tell people this is best avoided entirely 2024-09-16 23:43:40 you can probably get it to work but there's no simple recipe for it, partly because there's numerous ways to set up encrypted disks, boot, so on 2024-09-16 23:44:30 and while i've done this on other systems, i've not personally done it on alpine w/mkinitfs 2024-09-16 23:44:47 there may be something on the wiki about it. 2024-09-17 05:26:56 Hi 2024-09-17 05:28:50 One quick question. Does Alpine Linux compile all of its binaries and libraries (static and dynamic) with the following flags enabled? 2024-09-17 05:28:50 1) PIE/PIC 2024-09-17 05:28:50 3) Other Stack-Smashing Protection enabled 2024-09-17 05:28:50 2) StackGuard 2024-09-17 05:45:33 invoked: my swap is enabled, and I modified mkinitfs to make sure it is decrypted before resume kicks in 2024-09-17 05:46:17 My only difference from the wiki pages is that I don't use lvm 2024-09-17 06:00:43 https://www.kernel.org/doc/html/latest/power/swsusp.html mmmmh 2024-09-17 06:34:02 LinuxAficionado: I'm not sure if this is the exact values the builders run with, but: https://gitlab.alpinelinux.org/alpine/abuild/-/blob/master/default.conf?ref_type=heads 2024-09-17 06:37:32 LinuxAficionado: considering any package can override the *FLAGS it is compiled with, I'd say the ultimate answer is statistically no 2024-09-17 06:44:02 What I meant was that all the packages that are hosted inside Alpine repositories and are installed using APK all of them have PIE/PIC, Stackguard and Other Stack-Smashing Protection enabled. Is that correct? 2024-09-17 06:44:25 I understand that packages installed from other repositories or from source tar balls may not have these 2024-09-17 06:45:51 For example if XFCE were to be installed along with UFW on Alpine Linux, then all the binaries and libraries that would come with these options, namely PIE/PIC, Stackguard and Other Stack-Smashing protection enabled. Right ? 2024-09-17 06:48:20 iggy: the link that you had provided does not have the compiler flags for PIE/PIC enabled. 2024-09-17 06:51:53 coredumb: how can something be modified on the disk between suspend and resume? In a hosted environment probably but when running on bare metal this ideally cannot happen. Am I missing something over here? 2024-09-17 08:10:16 Ok so I wonder if the issue is not me doing my testing in a VM... resume from hibernation seems to hang up when reloading initramfs and seems that nlplug-findfs locks on virtio bus scanning 2024-09-17 08:16:57 you have to ask yourself how badly you want this. even if you do get it working, it will probably be brittle 2024-09-17 08:17:37 on my laptops, lid close powers the machine down. 2024-09-17 08:25:25 invoked: well it works mighty fine on my current distribution, not brittle at all, albeit based on systemd stack :) 2024-09-17 08:25:48 I already asked myself and I want it pretty bad :D 2024-09-17 08:25:56 much needed on my laptop 2024-09-17 08:27:22 Now what I can "easily" test is if the lvm on luks way as described in the wiki works better. If it doesn't - which I expect - it's probably because of the virtio drivers somehow and I should see how to validate that on hardware for once 2024-09-17 08:48:10 good luck :) 2024-09-17 08:57:25 Ok funny thing 2024-09-17 08:57:32 It actually works all fine :D 2024-09-17 08:57:56 it's just the console display that is not refreshed somehow :O 2024-09-17 08:58:30 If I connect to the VM console I got the login prompt back and the system is in the state before hibernation 2024-09-17 08:59:11 VM console through libvirt console* 2024-09-17 17:47:04 so I just bumped into this: https://www.reddit.com/r/AlpineLinux/comments/1ahk3po/does_alpine_support_a_separate_usr_partition/ 2024-09-17 17:47:22 is there a reason alpine doesn't mount all partitions before modprobe? 2024-09-17 18:35:50 Seems logical - you need all the relevant modules loadrd before you mount something 2024-09-17 20:41:24 hmm is this poxis? doesnt seem to work in ash or dash, the "&" makes an error..: case foo in foo ) echo foo;& bar ) echo bar;;esac 2024-09-17 20:42:51 The & needs to go to the end of the command 2024-09-17 20:46:03 ikke im looking at https://pubs.opengroup.org/onlinepubs/9799919799//utilities/V3_chap02.html its not a normal asyncronical command divider.. 2024-09-17 20:46:33 it worked in bash 2024-09-17 20:56:04 what did it do in bash? 2024-09-17 20:56:35 prints both foo and bar 2024-09-17 20:57:25 indeed it does 2024-09-17 20:57:45 the document clearly does not require it to do so 2024-09-17 20:58:10 looks like bash just ignores the & 2024-09-17 20:58:22 which seems like a legit interpretation, the preceding AND-OR list is considered empty 2024-09-17 20:59:09 If the case statement clause is terminated by ";&", then the compound-list (if any) of each subsequent clause shall be executed, in order, until either a clause terminated by ";;" is reached and its compound-list (if any) executed or there are no further clauses in the case statement. 2024-09-17 20:59:46 oh i missed that bit 2024-09-17 22:29:32 has any else experienced issues with firefox addons lately? 2024-09-17 22:29:32 Every time I upgrade my system with apk (i.e firefox) I loose all my Multi Account Containers. 2024-09-17 22:59:59 wizzard: which Alpine version, and which Firefox package? 2024-09-17 23:00:55 firefox-130.0-r0 on alpine edge 2024-09-17 23:11:38 hm 2024-09-17 23:11:43 i think we've had this issue before.. 2024-09-17 23:11:48 x86_64, right? 2024-09-17 23:16:50 yes 2024-09-17 23:24:58 If is there a known fix or bug report? I would be glad to help in solving the issue 2024-09-17 23:55:12 not sure yet, i vaguely remember a similar thing happening before but i might be wrong 2024-09-17 23:55:20 for now, could you open an issue on the aports repo? 2024-09-18 06:17:58 One quick question for all of the participants 2024-09-18 06:18:18 Do you guys use a DE/Window Manager on Alpine Linux ? 2024-09-18 06:19:10 yes, xfce, for many years now, I like it 2024-09-18 06:19:12 if yes then what is the one that you use ? 2024-09-18 06:46:02 Recently Debian and some of its offshots have incorporated into their distros the very same policies that Alpine Linux used to be known for. For Example most if not all of Debian repositories have PIE/PIC code in them. Ditto for Stack Smashing as most of the libraries are built in Debian using the same parameters. So how does Alpine Linux now differentiate against Debian as the security difference between Alpine and other Distros has narrowe 2024-09-18 06:48:02 If we compare other Linux distros say Debian/Fedora and Alpine in terms of security then the gap between them has narrowed considerably. For example Debian compiles almost all of its binaries as PIE and also incorporates Stack Smashing protection too. 2024-09-18 06:49:14 If we compare Alpine Linux with other distros like Debian/Fedora then we can see that the gap between Alpine and other distros has narrowed considerably. 2024-09-18 06:50:43 I plan on going gnome on a fresh new Alpine install 2024-09-18 08:17:58 durrendal: It's probably still early for you I guess, but if you have time to merge this one please https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/72141 salt is currently broken and unusable on edge :'( 2024-09-18 08:31:19 durrendal: oh woops I see you cannot merge my bad :x 2024-09-18 08:32:46 Huh somehow my apk got broken, Error relocating /sbin/apk: apk_sign_ctx_status: symbol not found. I was able to fix it luckily by downloading the static build of apk and using that to apk fix apk-tools. Not sure how this happened but this could ruin someones day if they don't know how to fix it... 2024-09-18 08:54:51 PureTryOut: what got fixed? 2024-09-18 09:08:46 It just reinstalled the binary 🤷 2024-09-18 09:42:46 so you probably had an error before upgrading apk-tools 2024-09-18 10:46:52 coredumb: can squash the commits? Ideally so we have 1 commit for both salt and salt-lts that adds the patch and bumps the pkgrel at the same time. And then rename the MR to 'community/salt-*: fix urllib parsing' 2024-09-18 10:47:07 s/can/can you/ 2024-09-18 10:47:34 durrendal: generaly it's one commit per package, unless there are many packages and the changes are similar 2024-09-18 10:50:04 ikke: in this case it's the same patch change for salt and salt-lts. But there's currently a commit for each, and then a commit to bump the pkgrel. I meant to imply we should change it so that both salt and salt-lts have their own commit in the same MR, which supplies the patch, and bumps pkgrel at the same time. 2024-09-18 10:50:06 ikke: there's indeed 2 packages and 1 single identical change to both 2024-09-18 10:50:45 durrendal: Ok let me do some git voodo there then 2024-09-18 10:51:04 durrendal: right, that makes sense 2024-09-18 10:52:55 I probably should've said that up front. But it's early here, and I'm juggling getting kids ready haha, so only half the gears are turning here 2024-09-18 10:54:35 10:17 < coredumb> durrendal: It's probably still early for you I guess 2024-09-18 10:54:45 I thought so indeed ;) 2024-09-18 11:00:37 Haha yeah it's 7am EST currently. I think your message came in around 4am my time :) 2024-09-18 11:06:07 Hope you didn't have any notification about it 2024-09-18 11:10:15 No worries at all, it's certainly nothing that woke me up! Plus if I was worried about it I wouldn't use an irc relay to stay connected all the thime 2024-09-18 11:14:19 Perfect then o/ 2024-09-18 11:17:58 btw I also opened an issue linked to the MR 2024-09-18 11:27:49 Oh actually this might be a good time to ask, coredumb do you have any issues with salt-minion? I haven't had the bandwidth to check but there's been an open issue about minion failing with conflicts from py3-setuptools. Curious if you've run into something similar. 2024-09-18 11:28:15 https://gitlab.alpinelinux.org/alpine/aports/-/issues/15561 <- issue I'm referencing for more context. 2024-09-18 11:38:15 durrendal: https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/72141/commits there it is 2024-09-18 11:38:31 sorry it took me longer than expected because my git-fu is rusty 2024-09-18 11:40:24 durrendal: although I contributed to the package in the past I've not used Alpine - or salt - in ages. 2024-09-18 11:40:34 That's much better, thank you! Do you mind renaming the MR as well? Once that's done I can approve it and hopefully it'll get merged decently quickly 2024-09-18 11:40:56 Now I'm revisiting a setup with alpine + salt but only as a master less setup. salt-call only 2024-09-18 11:41:09 durrendal: oh sorry forgot about that 2024-09-18 11:41:41 better now? 2024-09-18 11:42:28 I've never used it in a master less setup, but that sounds very interesting. I'm using it to manage windows endpoints mostly 2024-09-18 11:43:05 Perfect :) thank you! I appreciate the proactive MR, it's incredibly helpful 2024-09-18 11:44:44 Well I appreciate your fast answer so that I can install my systems :D 2024-09-18 11:46:05 About the 15561 issue.... I must admit I have no idea 2024-09-18 11:46:20 let me quickly test on my test VM though 2024-09-18 11:46:34 IIUC I only need to install py3-setuptools 2024-09-18 11:47:28 So yeah true it breaks 2024-09-18 11:53:13 durrendal: the early import solution he proposes in the thread only half work 2024-09-18 11:54:01 So it obviously needs the follow up 2024-09-18 11:54:12 Not sure if I agree with the stripping down though 2024-09-18 11:55:23 No problem at all! And thanks for double checking that issue for me as well. I'm in the fence about the proposed solution as well, it seems a little messy. 2024-09-18 11:56:33 I'll shuffle that back up to the top of my to do list. Admittedly I've been side tracked trying to package puppet and grlx for alpine because I fear Broadcom will be less than kind to salt in the long term. 2024-09-18 11:57:43 seems it would be cleaner to fix pkg_resources 2024-09-18 11:58:53 Never looked at the code before now so.... 2024-09-18 12:00:55 hi 2024-09-18 12:02:21 The gap between Alpine Linux and other distros like Debian, Fedora, etc has shrunk. For example Debian now compiles all of its binaries and libraries with PIE and Stack Smash protection enabled. 2024-09-18 12:02:32 Previously this was a forte of Alpine Linux 2024-09-18 12:03:27 So does Alpine Linux still offer some advantages over other Distros like Debian, Ubuntu, Fedora, etc in terms of security ? 2024-09-18 12:08:10 LinuxAficionado: for example Debian does not rebuild old packages that have not changed from one distrib to another leading to have some less secure binaries from less secure older compilation toolchains 2024-09-18 12:08:59 sources: https://grsecurity.net/toolchain_necromancy_past_mistakes_haunting_aslr 2024-09-18 12:09:32 If you want to start comparing things ... you know ;) 2024-09-18 12:10:07 durrendal: I can't say I particularly like this pkg_resources code : 2024-09-18 12:10:53 LinuxAficionado: there's a lot more that goes into a distro than just cflags 2024-09-18 12:12:35 a lot of that comes down to subjective things like, how often do you want to be surprised by changes 2024-09-18 12:12:52 they asked in particular about security though 2024-09-18 12:13:22 it is also about changes 2024-09-18 12:17:49 durrendal: I can make it work without being too nasty :D 2024-09-18 12:18:01 _too_ is key here indeed 2024-09-18 12:18:47 basically patch pkg_resources to stop failing on missing dep and continue 2024-09-18 12:19:04 I can share a patch with you if you want to dig deeper 2024-09-18 12:20:06 coredumb: Those are far and between. A significant portions of packages in Debian, atleast which get regular updates, are compiled with PIE, Stack Smashing protection, NX security enabled, full Relocation Read-Only (RELRO), enabled 2024-09-18 12:20:21 I am not trying to say that Alpine linux is no longer relevant 2024-09-18 12:20:28 or it is of no use 2024-09-18 12:20:44 I'm not trying to say anything 2024-09-18 12:21:29 Just that if you really care about security some flags are not necessarily the game changer you may expect 2024-09-18 12:22:02 invoked: There is a lot that goes into a distro agreed. Alpine Linux had a claim to fame for its security. The point is that other distros are catching up and will achieve most if not all of Alpine Linux security capabilities 2024-09-18 12:22:24 I'm not sure I ever considered Alpine "relevant" because of those cflags :D 2024-09-18 12:23:12 then why did you choose Alpine Linux? Not going to question it or challenge it. Just would like to know 2024-09-18 12:23:36 LinuxAficionado: do not forget all the glibc critical CVE that many vendors categorize as "won't fix" 2024-09-18 12:24:07 not to kick a dead horse but also consider the xz backdoor. 2024-09-18 12:24:24 Not saying that musl doesn't have its flaws not security issues but at least all the glibc ones don't affect Alpine 2024-09-18 12:24:46 hodapp_: I asked about security because that was the primary claim to fame of Alpine Linux. The other being fast boot up time and low overhead. Plus ability to customize it as per needs. Other Distros push a lot into their installation. Alpine does not do that. 2024-09-18 12:24:56 LinuxAficionado: Back in the days because it was one of the few including grsecurity patches :) 2024-09-18 12:25:03 i don't know what this "claim to fame" stuff is about 2024-09-18 12:25:21 As of now? I still like the "no bloat" and minimalist approach 2024-09-18 12:25:47 I don't know what the 'claim to fame' part is about either 2024-09-18 12:25:51 Yeah where's that claim to fame thing coming from? 2024-09-18 12:25:52 invoked: The XZ backdoor was most probably an attempt by a state actor. It was caught. 2024-09-18 12:26:43 durrendal: honestly ... as it's from the py3-setuptools package, I don't have any idea how much stuff that may break down the line :O 2024-09-18 12:27:37 i don't want to get into a discussion about that, it was just to illustrate that cflags are just one decision among many in terms of security. 2024-09-18 12:31:05 https://alpinelinux.org/about/ Ok I guess the "claim to fame" comes from here... 2024-09-18 12:31:15 https://distrowatch.com/table.php?distribution=alpine - It was designed with security in mind; it has proactive security features like PaX and SSP that prevent security holes in the software to be exploited. 2024-09-18 12:31:15 https://anto.online/alpine-linux-vs-ubuntu-comparison/ - Alpine Linux says it is more secure because of all user-land binaries are compiled with PIE enabled and Stack-Smashing Protection (SSP) enabled. 2024-09-18 12:31:26 these are two which come to my mind 2024-09-18 12:32:24 LinuxAficionado: How is that an Alpine claim? 2024-09-18 12:32:30 those are third parties writing their interpretations 2024-09-18 12:32:31 these are the reason why security was one of the claim to fame of Alpine Linux. That and as coredumb: said minimalist approach 2024-09-18 12:32:58 Official page has not claim to fame 2024-09-18 12:33:04 you should re-read 2024-09-18 12:33:16 Alpine Linux was designed with security in mind. All userland binaries are compiled as Position Independent Executables (PIE) with stack smashing protection. These proactive security features prevent exploitation of entire classes of zero-day and other vulnerabilities. 2024-09-18 12:33:27 allright let me put it in another way then 2024-09-18 12:34:13 How is that a claim to fame other than "We are security conscious and make sure our packages are built with security flags" 2024-09-18 12:34:18 i don't want to speak for alpine, but "small, simple, secure" these are just some simple design goals for what alpine is doing here 2024-09-18 12:34:19 invoked: if Compiler flags are just one the points to be taken for the decision among many others in terms of security, then what would be those others. Can you please give us a few 2024-09-18 12:34:27 There's no comparison at all either 2024-09-18 12:35:21 LinuxAficionado: it's not my job to sell you on what alpine is doing. it either aligns with what you want, or it doesn't 2024-09-18 12:35:45 hodapp_: coredumb: , All I am saying is that out in the Linux Ecosystem, security and no-bloatware are the reasons why Alpine Linux is looked at 2024-09-18 12:36:59 invoked: I am not asking for a sell. Nor am I questioning the validity of Alpine Linux. I would like to know what other factors should be considered. It is for my benefit. 2024-09-18 12:37:49 I do not intend to disparage Alpine Linux or the choice of using it. 2024-09-18 12:38:25 that's fine, but "security" is a deep topic and it's not really something we can hash out here, even if i had the energy for it 2024-09-18 12:39:07 it would be suitable for #alpine-offtopic 2024-09-18 12:39:58 invoked: allright will post it over there 2024-09-18 12:40:26 just wondering if I wanted to have this discussion with Alpine users should I post this in their mailing list or over here in the IRC chat ? 2024-09-18 12:40:39 i.e. in the channel #alpine-offtopic ? 2024-09-18 12:41:57 mailing lists are definitely more suited towards deeper conversation 2024-09-18 12:42:33 this channel is mainly for alpine users reporting problems, getting help, that sort of thing 2024-09-18 12:42:50 durrendal: I rebased the MR btw it should be ready to merge now 2024-09-18 12:46:44 invoked: Got it thanks. Will post it on the mailing list 2024-09-18 12:46:57 LinuxAficionado: cool 2024-09-18 12:59:24 I am trying to register at https://meta.alpinelinux.org/register/step2 so that I get access to the mailing list. However the message that is shown to me is Registration is currently closed. 2024-09-18 13:00:07 do i have to send a email to ~alpine/users+subscribe@lists.alpinelinux.org to subscribe and then post my query into Alpine User Mailing list ? 2024-09-18 13:05:03 coredumb: yeah if you don't mind throwing what you have my way that'd be super helpful. I'll chase it down the rest of the way happily :) 2024-09-18 13:05:44 coredumb: on on the MR, if I could merge I would! We'll have to wait for one of the core team members to merge it, but it should go through pretty quickly with approval and the fixes you made this morning 2024-09-18 13:07:36 durrendal: OK 2024-09-18 13:07:55 for pkg_resources ... I think this is actually a deeper problem 2024-09-18 13:08:24 LinuxAficionado: yeah that should work. 2024-09-18 14:00:31 ------->>>>>>> ptrc, you are correct, error in script causing issue. Thanks owain 2024-09-18 14:17:35 Is it possible to not install doas via setup-user install script? 2024-09-18 14:19:01 andar1an[m]: it only does so if you pass `-a` 2024-09-18 14:20:18 Ah okay, so I guess at somepoint during setup-alpine or setup-desktop -a is being passed. I can run manually. Today I am tinkering with grub2 and luks2 and btrfs. I believe now it should be possible to point from /boot/efi to /boot on btrfs fs in luks2 container now, but if anyone has tried this and knows it will fail please let me know haha 2024-09-18 14:23:17 basically is /root/grub-pre.cfg still experimental? Grub seems to be on RC5 in stable now so not sure 2024-09-18 14:24:20 (or luks2 method, dk how to word) 2024-09-18 16:22:44 andar1an[m]: you can do simplified version actually. 2024-09-18 16:23:01 for answer files, is it just bash? Can I cat or tee to apk repository file from it? Or is it just an env file? 2024-09-18 16:23:03 grub supports luks2 now without the pre- thing 2024-09-18 16:23:17 BUT 2024-09-18 16:23:27 coredumb: That is what I am doing, just making sure because in wiki says experimental and on os ubdate may lose access 2024-09-18 16:23:43 you need to make sure you load the proper modules in grub config 2024-09-18 16:24:02 and the wiki is not accurate and only list luks module for example 2024-09-18 16:24:24 the wiki should be fixed up because these questions keep coming around 2024-09-18 16:24:40 for example if you followed the wiki and added the crypto_keyfile.bin with the commands of the wiki 2024-09-18 16:24:48 it will not boot :D 2024-09-18 16:25:05 I have the grub and cryptsetup docs and manpages open, and have messed with this elswhere. Just never finished with FDE or Secure Boot update to mok and kek and db before 2024-09-18 16:25:17 😹 2024-09-18 16:25:26 because the cryptsetup luksAddkey command fails to set proper hash 2024-09-18 16:25:35 why I love iteration. I try to update wikis as I go through 2024-09-18 16:25:37 imo the uefi & mbr guides should be split up too. mixing them in the same page has gotten people confused 2024-09-18 16:25:44 this one may hopefully lead to a unique wiki 2024-09-18 16:25:57 so you may load sha512 like you encrypted the disk 2024-09-18 16:26:06 I wanted to update setup-disks, but that will be future potentially if this works 2024-09-18 16:26:26 Ya, also the only guides relate to LVM in LUKS 2024-09-18 16:26:33 but the Addkey actually adds a sha256 slot which makes grub incapable of booting if you don't load sha256 module as well 2024-09-18 16:26:45 I am doing BTRFS directly in luks, and hoping to have subvolumes for /boot and /swap there 2024-09-18 16:27:13 good to know, thank you! 2024-09-18 16:27:19 I've made FDE with swap and btrfs in their own partitions 2024-09-18 16:27:36 I have never worked with Grub directly before, this is good learning 2024-09-18 16:27:50 But the decryption of swap requires modification of mkinitfs 2024-09-18 16:27:52 I don't want to do swap in own partition because than would be unencryted or another luks container? 2024-09-18 16:28:13 another luks container decrypted from keyfile 2024-09-18 16:28:22 Ya, with BTRFS can have swapfile in subvolume which I hope will not deny snapshots for other vols 2024-09-18 16:28:43 resume from hibernation may be a hit & miss though 2024-09-18 16:28:55 Currently thinking:... (full message at ) 2024-09-18 16:29:05 never had good results for hibernation with btrfs swapfiles :/ 2024-09-18 16:29:05 all of these install paths are, in my opinion, sufficiently different that they should all be on their own pages. and somewhere on a parent page, huge text: the only "supported" method is what's provided by alpine-* scripts 2024-09-18 16:29:30 invoked: +1 2024-09-18 16:29:35 I think it is one of the best part of alpine though, because at least it is easier to understand what the heck is happening, or adjust 2024-09-18 16:30:36 RIght now I am unsure how I may install alpine with subvolumes, but i'll get there. Need to get past the luks and grub stuff first 2024-09-18 16:30:51 you put subvolumes in your fstab 2024-09-18 16:31:05 and you mount them before running setup-disk 2024-09-18 16:31:07 ya, but i mean during install 2024-09-18 16:31:27 I guess because of FSTAB the system will know where to stick files? 2024-09-18 16:31:42 ah okay, I thought it was gonna be more of a pain haha 2024-09-18 16:31:44 that's locely 2024-09-18 16:31:47 *lovely 2024-09-18 16:32:02 fairly easy 2024-09-18 16:33:07 SO yeah... I probably will need to contribute my "crypttab" support to mkinitfs I think :D 2024-09-18 16:34:23 s/alpine-/setup-/ 2024-09-18 16:34:47 iirc most of the bugs with btrfs swap have been addressed 2024-09-18 16:34:47 haha cool. sorry for ignorance, what is crypttab? is that related to cryptomount? 2024-09-18 16:35:13 Nice, really hoping this works. No need for LVM with BTRFS 2024-09-18 17:18:36 Thing is, it's one of the key rule of hibernation is it not? writing in bold red by kernel developers 2024-09-18 17:19:29 resume process should happen before you mount root fs ... so how could that work fine with btrfs swap file? 2024-09-18 17:25:51 is there no sane backend for brother scanners in repo? wanted brother4 2024-09-18 17:34:38 whats your experience with printers, is it cheaper to by a new one than buy new toner and eventually Toner Cartridges? 2024-09-18 17:43:07 "resume process should happen..." <- Dk, maybe I just don't hibernate. Sleep is fine or power off. 2024-09-18 17:43:52 None of my machines really require hibernate, and I never currently use it 2024-09-18 17:46:41 i hibernate with btrfs and swapfile, it's outside / subvoluime 2024-09-18 17:47:20 I was planning for a specific swap subvolume and making it nocow and no compression 2024-09-18 17:47:48 just so it doesn't interfere with other subvols for snapshots 2024-09-18 17:48:40 "is there no sane backend for..." <- if your printer can do driverless, you may be good with cups, but there is an open source driver repo. My Epson Ecotank driver isn't there and is currently not working with driverless though. Trying to find link 2024-09-18 17:49:39 andar1an the printer works well with generic, but the scanner is the problem :/ 2024-09-18 17:50:04 https://pkgs.alpinelinux.org/packages?name=gutenprint&branch=edge&repo=&arch=x86_64&maintainer= 2024-09-18 17:50:05 gutenprint may have firmware for scanner too 2024-09-18 17:50:42 see if the usb bug with hibernate still exists. ie, plug in a usb stick or something, put the machine into hibernate, unplug the usb stick, then try to resume 2024-09-18 17:50:45 it's currently in testing on edge, which makes me think it may be in community repo in Nov 2024-09-18 17:51:36 there used to be a lot of gotchas with hibernate. maybe that stuff is all fixed now, i don't know. 2024-09-18 17:53:18 andar1an what do you mean driverless? i did search for brother in all package _content_ and no results tho 2024-09-18 17:53:38 I don't use this thing to scan or fax anything, and can't see anything specific related to scanning on gutenprint with quick glance. 2024-09-18 17:53:38 https://openprinting.github.io/gsoc2019/03-sane-module/ 2024-09-18 17:53:53 it means you don't need to install a driver on your machine to be able to print 2024-09-18 17:54:10 re hibernation, do you think poweroff-ing laptops wear them out more than hibernation? i have hibernated mine many years and it still going strong i guess 7-8 years now.. 2024-09-18 17:54:20 https://openprinting.github.io/printers/ 2024-09-18 17:55:01 search IPP Everywhere 2024-09-18 17:55:10 Info here too: https://wiki.alpinelinux.org/wiki/Printer_Setup 2024-09-18 17:55:13 oh, my printer is quite old, bo i tdont think it has ipp 2024-09-18 17:55:26 printer/scanner 2024-09-18 17:55:37 If it is old it is probably in gutenprint 2024-09-18 17:56:04 mine was too new which is a problem, and I didn't wanna figure out how to make driver from source from Epson. It seemed tedious and not worth energy lol 2024-09-18 17:56:15 I can add the printer, it just prints blank pages lol 2024-09-18 17:56:23 haha 2024-09-18 17:57:48 I have learned 1 year lag it pretty good for Linux related to hardware haha. Also means cheaper and less bugs 2024-09-18 17:58:53 i don't think powering up/down wears a machine down any more than hibernate does. probably less so 2024-09-18 17:59:32 flushing state out to swap seems like it would be harder on the disk. 2024-09-18 17:59:39 with an NVME I don't notice, and sleep barely drains my battery 2024-09-18 18:00:00 I would rather shutdown with FDE too 2024-09-18 18:00:14 Does hibernate lock disk? 2024-09-18 18:42:01 it's not about wear 2024-09-18 18:42:40 it's about not having to restart everything :S 2024-09-18 18:46:42 having computers with 2 TB memory, i don't like to hibernate 2024-09-18 18:48:53 i'm fine with sleep 2024-09-18 18:51:03 2tb of ram wow 2024-09-18 18:56:34 working on 'apk add cmd:dig' support for openwrt, i immediately run into 'this provides cmd:gzip but only when you then choose it from the alternatives' 2024-09-18 18:57:02 alpine solved this by not having alternatives? 2024-09-18 18:58:48 Habbie: I think the problem is that your cmd: provides are unversioned? 2024-09-18 18:59:17 We had that in the beginning, but then added a version later because of similar issues 2024-09-18 18:59:44 a provides either needs a version or a priority 2024-09-18 19:04:56 openwrt doesn't have cmd: provides at all yet 2024-09-18 19:05:03 and i did make a note to think about whether they need versions 2024-09-18 19:05:28 priorities are a good one, didn't know that was possible - and the 'alternatives' data gives me useful numbers for that 2024-09-18 19:06:06 I'm not aware of 'alternatives' 2024-09-18 19:06:26 is that something new in apkv3? 2024-09-18 19:06:43 oh, no, they're a feature some distros have, hold on 2024-09-18 19:07:00 on my debian for example, /usr/bin/vi -> /etc/alternatives/vi, /etc/alternatives/vi -> /usr/bin/vis 2024-09-18 19:07:07 could also be vim, nvi, etc. 2024-09-18 19:07:22 packages say 'this binary can be vi' and debian has tooling to let the user make that systemwide choice 2024-09-18 19:07:25 openwrt apparently also has this 2024-09-18 19:07:43 and then i think this means that, say, vim, might want to say 'provides cmd:vi' 2024-09-18 19:07:55 Right, I'm aware of debian / ubuntu alternatives 2024-09-18 19:07:56 which priorities would indeed be an improvement upon 2024-09-18 19:07:58 ok 2024-09-18 19:09:41 With versioned provides, apk installs the newest version by default 2024-09-18 19:10:06 so i guess those don't use the real package version number? 2024-09-18 19:11:50 fuse-exfat-utils-1.4.0-r1 provides: 2024-09-18 19:11:52 exfat-utils 2024-09-18 19:26:10 alex_: yeah my laptop can't host that much mem :D 2024-09-18 19:39:33 hi, i got a question about the installation on alpine linux: i followed the instructions on the zfsbootmenu site, but got the error that "cannot import poolid: a pool with that name already exists". But after this message, the system works... so i didnt understand the message... I removed afterwards the openrc entries and deleted the cache, but the issue still there: https://www.reddit.com/r/AlpineLinux/comments/1fjxxr 2024-09-18 19:39:33 1/zfs_on_root_cannot_import_pool_but_it_works/ 2024-09-18 19:40:39 https://www.reddit.com/r/AlpineLinux/comments/1fjxxr1/zfs_on_root_cannot_import_pool_but_it_works/ 2024-09-18 19:59:54 ikke, i have managed to formulate my question. in alpine:edge, "apk info -a exfat-utils | grep exfatlabel" shows two entries. those would be decided on their version number while they come from unrelated packages? 2024-09-18 20:03:31 yes 2024-09-18 20:08:36 Hello, does anyone happen to know of any VPS hosting providers that offer Alpine Linux? 2024-09-18 20:10:04 do you mean turnkey installs? you can install alpine from iso on vast majority of them, probably 2024-09-18 20:11:13 "it's about not having to restart..." <- my browser opens where I left off, what other app do you use that you require it to be in same spot when opening? Sleep works for me too - battery would last for 5 days in sleep for me 2024-09-18 20:11:40 my AIO pump finally went, had to replace lol 2024-09-18 20:14:35 ikke, alright. and the priorities? i see them in the two APKBUILDS but don't see any such marker on the 'exfat-utils' provide, what am i missing? 2024-09-18 20:16:09 I don't think apkv2 exposes the priority 2024-09-18 20:16:51 ah 2024-09-18 20:17:02 so that item in the APKBUILD currently does nothing 2024-09-18 20:19:22 It does do something 2024-09-18 20:19:26 it's a tie-breaker 2024-09-18 20:19:37 just not visible 2024-09-18 20:19:56 not visible in info, but still working? ok 2024-09-18 20:20:04 incidentally those two for exfat-utils have the same priority :) 2024-09-18 20:20:13 Is this related to sticky packages? 2024-09-18 20:20:45 Habbie: the version is the first tie-breaker. 2024-09-18 20:21:06 So if there are 2 packages with the same provides, the highest version is installed 2024-09-18 20:21:18 right, based on the actual package version then? 2024-09-18 20:21:26 The version in the provides 2024-09-18 20:21:31 which is often based on the pkg version 2024-09-18 20:21:38 right, but the exfat-utils provides looks unversioned to me 2024-09-18 20:22:30 (not trying to argue anything, btw, just curious about all of it) 2024-09-18 20:23:31 Habbie: that seems incorrect then 2024-09-18 20:23:44 ah! 2024-09-18 20:23:53 well that was always a possible outcome :D 2024-09-18 20:24:34 Hmm, it still picks exfatprogs 2024-09-18 20:26:38 oh, i didn't even try 2024-09-18 20:27:16 ok, that's the package with the lower version number itself (but if i understand correctly, those are not compared) 2024-09-18 20:27:27 it is also the first one in my 'apk info' 2024-09-18 20:32:43 Habbie: exfatprogs has a higher priority 2024-09-18 20:32:46 if I look at the index 2024-09-18 20:32:53 k:20 vs k:1 2024-09-18 20:35:31 Habbie: fuse-exfat sets provider_priority=1 in the -utils subpkg 2024-09-18 20:41:55 you're right, i somehow misread them as both 1 2024-09-18 20:42:17 ok, and the priority is a separate thing, not tied to the specific provide 2024-09-18 20:42:33 It's about the package 2024-09-18 20:42:49 alright 2024-09-18 20:43:07 all provides from a package have the same priority (but subpackages can have their own priority) 2024-09-18 20:43:15 thanks for indulging me again today, i'm going to process all this and PR some openwrt stuff (and send one MR to apk to fix a bug) 2024-09-18 20:43:18 right 2024-09-18 20:53:57 explaining these things helps to improve my own understanding as well :-) 2024-09-18 20:54:12 that makes me happy :) 2024-09-19 00:34:14 What exactly is openpax? 2024-09-19 00:35:38 chexo4[m]: a set of kernel hardening patches 2024-09-19 00:36:11 I got that, is there documentation somewhere? Seems like open source continuation of PaX/grsecurity? 2024-09-19 00:38:21 idk 2024-09-19 00:38:36 Afaik, they're not released officially yet, so 2024-09-19 00:40:39 Wait a minute is what grsecurity is doing with selling modifications/patches to the Linux kernel even legal? It feels like a dick move either way 2024-09-19 00:47:38 selling patches is legal. GPL doesn't require stuff to be free of charge 2024-09-19 00:47:44 Nothing in the GPL says you have to make your program available to everyone (i.e., being gratis), just that whoever you do distribute it to should also get the source code. 2024-09-19 00:47:49 Yep 2024-09-19 00:48:00 patches themselves are GPL'd 2024-09-19 00:49:41 you mean GPLv3 because it is more enforcing, captive and propagating, which is a proprietary EULA equivalent 2024-09-19 00:50:13 the euphemism is "protective" license, the reality is "shareware" free labour proprietary 2024-09-19 00:50:29 kernel is GPL2, so are grsecurity patches 2024-09-19 00:50:34 GNU licenses have versions 2024-09-19 00:51:45 also, kernels have a lot ot versions for the last 30 years (no, Linux was not usable before 1993-1995) 2024-09-19 00:52:40 grsec can not possibly upkeep all kernels for 30 years and not be included by default with the kernel.. 2024-09-19 00:53:23 isn't it a mystery why these are not part of the kernel? ;-) is it not supposed to be more reliable and resilient or are these even applicable by individuals cleanly? 2024-09-19 00:53:39 what happened to Debian-hardened ;-) 2024-09-19 00:53:58 or was it Gentoo-grsec 2024-09-19 00:54:36 selling free software.. the idiocracy has no limits 2024-09-19 00:54:46 "shareware" EULA commercial software 2024-09-19 00:55:20 are you really not conflicting some freedoms definitions in all this "selling" freedoms out craptalk? 2024-09-19 00:55:27 Don't buy if you don't want, nobody forces you 2024-09-19 00:55:51 selling software violates none of your freedoms 2024-09-19 00:55:58 hahaha, sell Linus Torvalds work to your family.. 2024-09-19 00:56:12 it's a lie 2024-09-19 00:56:15 why would I do that 2024-09-19 00:56:42 do you sell the slave a freedom declaration that they are now a free person? 2024-09-19 00:56:56 the what? 2024-09-19 00:57:21 a non-free program, do you sell it, to make it a free one? 2024-09-19 00:57:36 free from your "ownership" 2024-09-19 00:57:48 how it's even connected 2024-09-19 00:58:06 if it's free then it's shared, how do you sell it when it's free and shared already? 2024-09-19 00:58:43 you're conflicting definitions and.. making it a trade activity to provide free software, as in leasing subscription scams from "cloud" "vendors" 2024-09-19 00:59:05 but if it's free and available publicly to everyone, how are you exactly selling it? 2024-09-19 00:59:40 i take money and give .tar.xz 2024-09-19 00:59:47 simple as that 2024-09-19 00:59:55 it's a nonsense saying by small trade ambitious profiteers who want to make profit on the back of people providing free software, where the trading party did not participate ever 2024-09-19 01:00:04 you make nothing ;-) 2024-09-19 01:00:17 the trader steals free software and sells it as proprietary 2024-09-19 01:00:19 grsecurity only sells their patches, not the whole kernel 2024-09-19 01:00:41 they're *authors* of these patches 2024-09-19 01:00:45 so it's not free software 2024-09-19 01:00:56 it is 2024-09-19 01:00:57 but premium shareware 2024-09-19 01:00:59 it's not 2024-09-19 01:01:04 it has restrictions 2024-09-19 01:01:18 can not be obtained freely 2024-09-19 01:01:19 "free" != "publicly available" 2024-09-19 01:01:49 actually the 4 definitions of freedom are exactly individually and publicly available for consumption and participation 2024-09-19 01:01:54 4 variants 2024-09-19 01:02:05 individual and group 2024-09-19 01:02:12 use and participate 2024-09-19 01:02:18 there's nothing about public availability in those 4 freedoms 2024-09-19 01:02:28 get your facts right 2024-09-19 01:02:39 the 4 freedoms are public availability and public development participatory 2024-09-19 01:02:56 no 2024-09-19 01:03:35 nobody is interested in Linux if it is sold instead of free 2024-09-19 01:03:48 NOBODY 2024-09-19 01:04:15 public interest does not define whether the software is free 2024-09-19 01:04:54 is Linux free software? 2024-09-19 01:05:07 did you buy your kernel? 2024-09-19 01:05:30 did you manage to sell Linux to someone? 2024-09-19 01:05:37 no 2024-09-19 01:05:53 that does not change anything 2024-09-19 01:05:59 so Linux is commercial software by your call then, but it's not commercial, so it's free 2024-09-19 01:06:16 applications for it are free too 2024-09-19 01:06:35 free software != free of charge 2024-09-19 01:06:35 and the userland from GNU and compiler toolchain is free too 2024-09-19 01:06:54 RMS used to sell emacs btw 2024-09-19 01:07:11 free sofware is software that you can get and participate in, both as an individual or as a group, at no cost but your involvement into it 2024-09-19 01:07:16 that's what free means 2024-09-19 01:07:25 accessible and available 2024-09-19 01:07:48 software freedom has nothing to do with price 2024-09-19 01:07:51 we're talking about Linux the kernel 2024-09-19 01:08:09 and Stallman did not sell Emacs, you're confusing him with Gossling 2024-09-19 01:08:23 go learn some details about what you're trying to reason about cluelessly 2024-09-19 01:08:44 Stallman was selling manuals (books in print) for Emacs 2024-09-19 01:08:54 and it did not succeed 2024-09-19 01:09:30 as Emacs is including documentation and is also self-documenting and available from a number of free software source trees 2024-09-19 01:09:48 and has multiple implementations which compete and deliver freedoms 2024-09-19 01:10:08 https://www.gnu.org/gnu/thegnuproject.html -- " So I announced that I would mail a tape to whoever wanted one, for a fee of $150" 2024-09-19 01:10:20 if you place price on access and availability how are individuals and groups supposed to participate? 2024-09-19 01:10:38 he was selling the tape service, not the program 2024-09-19 01:10:42 the program was and is free 2024-09-19 01:10:49 define "participate" 2024-09-19 01:11:08 same applied to BSD UNIX which was free and sold for $4-5 USD which was the mail cost of the tape (own tapes) 2024-09-19 01:11:19 the 4 freedoms are defined in 1984 2024-09-19 01:12:26 none of those freedoms contain word "participate" 2024-09-19 01:12:43 and I enumerated them to you about 5 times already so learn to think like a rational person, instead of a repetitor of fake "pseudo-commerce" addorning and endorsing declarations of commercial business activities 2024-09-19 01:12:58 they do not contain the word sale and price too 2024-09-19 01:13:20 so you're speculating on the work of others making profit over free software which is what corporations do 2024-09-19 01:13:25 so you're a commercial operation 2024-09-19 01:13:33 it is legal 2024-09-19 01:13:36 therefore GNU GPL licencing is commerce and not free software 2024-09-19 01:14:17 if your statement is accepted, then you're an EULA shareware seller like Microsoft and Google and Amazon and Apple and Oracle 2024-09-19 01:14:29 yet again, you have failed to enumerate anything that pertain to software freedoms 2024-09-19 01:14:39 no, you have failed to sell anything 2024-09-19 01:14:41 it's free 2024-09-19 01:14:53 failure is for the liars 2024-09-19 01:15:01 Linux is free, GNU is free 2024-09-19 01:15:02 not that i really tried 2024-09-19 01:15:06 BSD is free 2024-09-19 01:15:46 and UNIX is free since 1975 as BSD UNIX and 1985 as GNU UNIX and 1995 as Linux UNIX variants 2024-09-19 01:16:08 and applications for BSD, GNU and Linux are free too 2024-09-19 01:16:56 plus, grsec is nonsensical failure as the "client" recipient of this commercial addon on top of free software needs hand-holding to get it to work at all (if it even works) 2024-09-19 01:17:07 so it's not integrated because IT IS NOT FREE 2024-09-19 01:17:33 grsec is therefore not used by the majority of people due to the complications and cost involved (non-free / failing trade model) 2024-09-19 01:17:48 so.. nobody gives it much attention any more 2024-09-19 01:17:58 and people run insecure Linux as is 2024-09-19 01:18:33 get your fake "you can sell free software" claims and go prove them as successful businesses 2024-09-19 01:19:17 until then don't lie and claim false things which are scandallously ridiculous and obscene claims that equate free software to commercial software 2024-09-19 01:19:41 it is true that GNU is closer to proprietary software 2024-09-19 01:20:18 but it is still free and does not cost money to get to download use and participate back into developing it further, for both individuals and groups 2024-09-19 01:20:41 so.. false claims are always false 2024-09-19 01:21:00 even when told by RMS under pressure to endorse trade and commerce 2024-09-19 01:21:03 he rejects this 2024-09-19 01:21:12 and so do hundreds of millions of people 2024-09-19 01:21:25 who get a free variant of UNIX for free 2024-09-19 01:22:00 even if you think otherwise, the majority of people do not care what you think 2024-09-19 01:22:23 they care that there is a free variant of UNIX and it's like a UNIX system that they know and can use for free 2024-09-19 01:22:50 “GNU UNIX” “Linux UNIX” lol 2024-09-19 01:22:53 and either comes with the computer or is easy to install and operate independently of other options 2024-09-19 01:22:59 :-) 2024-09-19 01:23:36 yep, these are all "crappy" UNIXes as per Mike Gancarz book on UNIX Philosophy (which has 2 editions) and is accepted by UNIX pioneers themselves 2024-09-19 01:24:01 who created it with the intent to be free and available to Universities and computer centre (sites where it was installed on mini computers) 2024-09-19 01:24:11 in 1975, BSD was not a complete operating system; only with the late networking release and 4.4BSD-Lite stuff did it become Free 2024-09-19 01:24:17 because it was paid by the computer cost and by the services it was used to deliver 2024-09-19 01:24:26 it was 2024-09-19 01:24:34 and made changes in UNIX back propagate 2024-09-19 01:24:59 in fact it was BSD that made UNIX network usable and gave it most of its extending applications like vi, the TCP/IP stack etc 2024-09-19 01:25:12 and the file system 2024-09-19 01:25:36 you should learn history, you can get it on CDs and video lectures by Kirk McKusick online 2024-09-19 01:25:36 (TCP/IP wasn’t until the early 80s) 2024-09-19 01:25:50 Oh no what have I done 2024-09-19 01:26:23 at any rate, you can't change history by saying what you think, you just say what you know 2024-09-19 01:26:36 and what you know is not enough 2024-09-19 01:26:53 GNU is a UNIX variant yes 2024-09-19 01:26:56 so GNU is UNIX 2024-09-19 01:27:08 and the recursion is a "joke" 2024-09-19 01:27:32 even in the serious sense of using recursion it's not, it's a loop nesting method 2024-09-19 01:27:32 you should learn history, you can get it on CDs and video lectures by Kirk McKusick online 2024-09-19 01:27:37 right back at you 2024-09-19 01:27:45 right nowhere 2024-09-19 01:28:02 Bill Joy wrote the TCP/IP stack to compete with BBN 2024-09-19 01:28:28 1BSD was released in 1978 as an addon to v6 2024-09-19 01:32:29 I was just asking about grsecurity how did it get this out of hand 2024-09-19 01:32:46 grsec is fucked 2024-09-19 01:32:59 Linus hates it 2024-09-19 01:34:32 Btw, reading just a few messages of scrollback: If I were to write a program and pass it to a friend, perhaps doing it all for money, perhaps not, and I told the friend that they and anybody else may inspect, use, change, and redistribute the program, and my friend didn’t distribute the program any further—we’d be the only two people having copies—it’d be Free software. 2024-09-19 01:35:06 despite not being publically available in any way 2024-09-19 01:35:42 it's not free software unless it's avaialable to inviduals and groups both alike for both use and participatory changes 2024-09-19 01:35:59 the 4 definitions are that 2024-09-19 01:36:00 The freedom, here, is given to (or perhaps rather not taken from) the friend, by me. 2024-09-19 01:36:28 so you have a limited sub-set of freedoms which is not free software by the 4 freedoms definition 2024-09-19 01:37:50 https://en.wikipedia.org/wiki/Four_freedoms?useskin=vector#See_also The Free Software Definition is often called "the four freedoms" within the free software community in reference to the speech and fundamental principles. 2024-09-19 01:38:21 https://en.wikipedia.org/wiki/The_Free_Software_Definition?useskin=vector#The_Four_Essential_Freedoms_of_Free_Software 2024-09-19 01:39:06 What are you even trying to argue? 2024-09-19 01:39:18 why are you interested? 2024-09-19 01:39:26 didn't you understand already? 2024-09-19 01:39:38 if not read again, not going to repeat the same page 2024-09-19 01:40:02 you can learn to ask smart questions 2024-09-19 01:43:20 I’m interested because your posting walls of trivial messages confuses me. I didn’t understand what you’re trying to argue already, no. I see little benefit in putting a whole lot more work into my little question of What are you trying to argue? 2024-09-19 01:44:30 get lost 2024-09-19 01:47:36 the pleasure is mine 2024-09-19 01:49:54 https://en.wikipedia.org/wiki/Berkeley_Software_Distribution?useskin=vector#History "In 1975, Ken Thompson took a sabbatical from Bell Labs and came to Berkeley as a visiting professor. He helped to install Version 6 Unix and started working on a Pascal implementation for the system. Graduate students Chuck Haley and Bill Joy improved Thompson's Pascal and implemented an improved text editor, ex.[10]" --vi(1) 2024-09-19 01:50:59 Professor Bob Fabry started the CSRC group in 1974.. 2024-09-19 01:52:06 https://en.wikipedia.org/wiki/Computer_Systems_Research_Group?useskin=vector#History 2024-09-19 01:53:12 CSRC was the name at Bell labs for the OS development group, in UC Berkeley it was named after it CSRG 2024-09-19 01:53:38 that's when free UNIX starts actually 2024-09-19 01:54:31 (it was always free to Universities and research groups after it lacked applications and needed them to teach students and write applications, like text editors, file systems, network stacks, databases etc) 2024-09-19 01:55:30 these came mostly from UCB, and later MIT (and GNU), CMU and others like Stanford, WU and USD 2024-09-19 01:55:56 and Chicago too 2024-09-19 01:59:57 so "UNIX was not a complete OS" but a Research UNIX project 2024-09-19 02:01:20 BSD was more complete, and GNU even more complete, Linux started badly and had an adaptive period to become a more UNIX-like system (it was a 386 MINIX clone first) 2024-09-19 02:02:13 eventually Linux got the BSD file system in the second "ext2" implementation to be more UNIX-like (from UFS/FFS/2 from McKusick and colleagues at BSD) 2024-09-19 02:02:39 and "everything else" network and OS internals (kernel) related 2024-09-19 02:04:27 grsec technically does not exist, even in "hardened" projects is a serious problem set and failure over the years to mark the decline of Gentoo 2024-09-19 02:05:29 same ideas and better implementation and integration is available in OpenBSD, FreeBSD, Windows.. and Linux is last to get it working.. with pains and failure 2024-09-19 02:06:09 see http://www.openbsd.org/innovations.html for details in the "Concepts" section 2024-09-19 02:07:05 stack protector (propolice) and related 2024-09-19 02:07:50 I am telling you this because I was using grsec patchkits on Linux in the 90ies and they sucked and were difficult to reapply and upkeep 2024-09-19 02:08:03 and resulted in performance penalties etc 2024-09-19 02:08:41 as for the "you can sell free software".. you can jump on your head too 2024-09-19 02:11:58 http://www.openbsd.org/papers/csw03/mgp00013.html 2024-09-19 02:14:46 http://www.openbsd.org/events.html for extra presentations and videos and slides explaining details how to get advances in secuirty and reliability 2024-09-19 02:16:36 the key point always made is, it is integrated work by many that just works™, while the patchkits do NOT work as it requires integration not done ahead of time and all applications need fixing after that too so.. it's an element of the improvemens that is not applied cleanly and everywhere (no propagation to the other layers of the system) and lacks many other measures and procedures to compete with a complete solution 2024-09-19 02:18:24 hence the existence of "hardening" projects that both fail in the cases of Gentoo "hardening" and FreeBSD "hardening", while both exist and continue to strive, it's failing over time and declining with overhead and maintenance to the point it's practically not usable for the general and even advanced workgroups 2024-09-19 02:19:21 also, it's a good idea to not be distracted and to throw "curve ball" false claims that are well explained and take time and energy to dismantle and defuse every time 2024-09-19 02:19:46 free software is free because it is not commercial, get over it, trade boy shop of selling stolen work of others 2024-09-19 02:20:20 it's free on the internet and can not exist without collaborative work of many many thousands of people for free on the internet 2024-09-19 02:21:13 that's called a public resource or public domain self-support by large supergroup of groups (a community of the non-commercial and communal / public type of the largest type) 2024-09-19 02:21:51 so yeah, it's by the public for the public and is public software for public interests by the public of free software public self-sustainability 2024-09-19 02:22:26 in that regard is a complete anti-thesis of commerce and exploitation (business as usual) 2024-09-19 02:22:46 it's free because it's not commerce and not exploitative, otherwise it can't be free 2024-09-19 02:23:30 so, free is the public software by the public for the public, not by businesses, busineses can not create free software, they create commercial ones that are not free and can't be 2024-09-19 02:24:40 curious counter examples that failed are failure and demonstrate failure, that is not suitable for examples as it disproves the curious claim attempts and invalidates it 2024-09-19 02:25:30 and this concludes the small lecture called use integrated work that works and is free and not for sale 2024-09-19 02:27:21 otherwise you end up stuck and at the same problem location you were before, only later in time and with depleted resources wastes on a failure dead-end road (wobbly path of failing curious examples that are failure) 2024-09-19 02:27:21 good, now "don't repeat lies" 2024-09-19 02:27:22 got it? 2024-09-19 05:30:37 plouton, hi, thank you. i read all your message.. its good 2024-09-19 07:54:47 Hi 2024-09-19 07:54:54 I think I'm going insane 2024-09-19 07:55:22 There is no resize2fs binary in e2fsprogs 2024-09-19 07:56:01 I cannot even `find / -name resize2fs` it 2024-09-19 07:56:05 please help 2024-09-19 08:01:47 alpine 3.20 btw 2024-09-19 08:03:25 Nevermind found it in `e2fsprogs-extra` 2024-09-19 08:12:45 newbie: next time you can either do `apk search cmd:resize2fs` or use https://pkgs.alpinelinux.org/contents 2024-09-19 08:15:40 Wtf libcrypto.so.3 just disappeared, completely breaking my system (even apk). Luckily I still had apk.static installed from a corrupted install of a file a few days ago so I could just fix it. But this is starting to look suspicious now 2024-09-19 08:17:21 libssl.so.3 has disappeared too... 2024-09-19 08:17:32 it would be more of a mystery if only one of them went, i think 2024-09-19 08:32:36 PureTryOut: do you still have console outout from apk upgrade? 2024-09-19 08:39:37 No, I didn't notice things going wrong when doing the upgrade. I only saw it later 2024-09-19 08:48:02 Hi! Does anyone have example of tiny-cloud with autoinstall? I want to try to install alpine completly headlessly. 2024-09-19 10:10:01 Does *buildrepo* skip packages that are already up to date? 2024-09-19 10:14:37 durrendal: Morning o/ - Any idea whom we should ping to get the MR merged and packages updated? 2024-09-19 10:20:34 WhyNotHugo: yes 2024-09-19 10:28:06 thnx 2024-09-19 13:10:13 coredumb: I pinged omni and celie on the MR. It might take some time to get merged though, the core team is small currently. 2024-09-19 13:12:33 merged :) 2024-09-19 13:23:13 ikke: thank you! 2024-09-19 13:23:53 Let me know when the backport is ready 2024-09-19 13:25:41 we will definitely do so :) 2024-09-19 13:26:49 also while you're here, I probably know the answer to this but wanted to double check. I've got a bunch of ruby aports that I'm working on that are all interdependent on each other. Should I do one MR per new aport or can I do a single large MR with 1 commit per new aport? 2024-09-19 13:27:14 One MR is fine 2024-09-19 13:27:57 oh thank goodness, I was not liking the idea of opening a ton of MRs and flooding the queue needlessly 2024-09-19 13:36:55 durrendal: ikke: thanks 2024-09-19 13:37:36 feel free to close https://gitlab.alpinelinux.org/alpine/aports/-/issues/16458 2024-09-19 13:37:46 or let me know if I shall do it 2024-09-19 13:38:21 We can close it when 3.20 is fixed as well :) 2024-09-19 13:54:49 coredumb: if you make another MR from your fix/community/salt branch targeting 3.20-stable then that'll backport the fix to 3.20 2024-09-19 13:58:11 You need to cherry-pick the commit 2024-09-19 13:58:18 otherwise it would try to merge in all of edge into 3.20 2024-09-19 14:10:15 ah that's a good point, thanks for clarifying that ikke. 2024-09-19 14:44:36 durrendal: ikke: let me check that 2024-09-19 14:45:31 so: 1: create a new branch based on 3.20-stable. 2: cherry-pick both commits from the first mr. 3. Create a new mr targetting 3.20-stable 2024-09-19 14:45:52 indeed 2024-09-19 15:13:04 durrendal: ikke: https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/72211/pipelines 2024-09-19 15:13:38 coredumb: your commit message needs to be fixed 2024-09-19 15:21:49 what's wrong with it?? 2024-09-19 16:31:20 coredumb: maybe update the commit message and the MR title to something like, "community/salt: fix path handling with urllib.parse" and link to the upstream MR in the MR description 2024-09-19 16:32:10 instead of in the upstream url in the commit message, to be clearer what the MR does 2024-09-19 16:33:20 s/commit message/commit subject line/ 2024-09-19 16:35:14 oh right I see 2024-09-19 16:37:11 Fixed 2024-09-19 16:38:53 minor nitpick, since it's targetting 3.20 stable, usually people prefix the MR title with the stable branch name for those, e.g. "[3.20] community/ ..." 2024-09-19 16:39:03 to distinguish it from edge 2024-09-19 16:40:06 probably don't need to do that in the commit subject line, just in MR title 2024-09-19 16:40:26 :D 2024-09-19 16:40:56 fixed as well in MR 2024-09-19 16:41:33 cool, hopefully it'll be looked at and merged shortly 2024-09-19 17:21:00 ikke, while yesterday i had no provides cmd:, now i have them, unversioned, and apk hates it :D 2024-09-19 17:23:08 ah, and 'apk add exfat-utils' in alpine works because it -does- have a priority 2024-09-19 17:34:25 ha. 'apk add cmd:atop' prevents 'apk del atop' because cmd:atop still wants it 2024-09-19 17:36:43 I may have learned something new, can someone verify if this is accurate? The static UUID looking strings in setup-disk are GUID's required because the partitioning program being used requires setting GUID directly? 2024-09-19 17:37:04 ikke, here's some more meandering thoughts from me https://github.com/openwrt/openwrt/pull/16427 2024-09-19 17:37:19 ikke, (if you're interested) 2024-09-19 17:37:34 <3 openWRT 2024-09-19 17:43:12 I kinda hope embedded will enable that kinda functionality through RTOS libraries one day for embedded systems 2024-09-19 17:44:21 like to imagine writing firmware on embedded. But there isn't really standard network stacks or drivers one can bring in without an OS that I have found yet 2024-09-19 18:15:38 Habbie: re removing atop: remember you are just manipulating world. As long as something pulls in atop, it will remain installed 2024-09-19 18:15:48 yes, i understand what's happening 2024-09-19 18:15:52 just never tried that del before that way 2024-09-19 18:16:18 And for aports, we use a version because you cannot blindly set (and thus override) the package priority 2024-09-19 18:16:23 i also can't get used to deps disappearing when i 'apk del atop', that's not how Debian does things 2024-09-19 18:16:50 ikke, right, except using version is weird because the packages might be semi-unrelated 2024-09-19 18:17:42 Yeah, and it's kind of random which package gets picked (whichever happens to have the higher version)( 2024-09-19 18:18:08 If you want an alternatives like system, having an explicit virtual package is better then using the automatic cmd 2024-09-19 18:18:12 cmd: provides 2024-09-19 18:18:43 that makes sense, but doesn't help people that just want to apk add cmd:gzip, right 2024-09-19 18:18:58 for packages that needed nodejs, they initially used cmd:nodejs, but that meant that it always pulled in nodejs-current (the latest version) rather than the stable version 2024-09-19 18:19:28 right 2024-09-19 18:19:35 Habbie: the assumption is more or less that there is a single package that provides a command, but that assumption does not always hold 2024-09-19 18:19:48 yeah, that assumption breaks exactly when alternatives get involved 2024-09-19 18:20:12 So using a virtual package with priority is how you can handle that, but that's always explicit 2024-09-19 18:20:20 uhm, what's an "explicit virtual package" practically? 2024-09-19 18:20:37 provides='virtual-package' # note, no version specified 2024-09-19 18:20:37 that 3 things can 'provides:gzip'? 2024-09-19 18:20:55 provides=gzip; provider_priority=100 2024-09-19 18:21:11 right 2024-09-19 18:21:35 but, you want to avoid providing a virtual package for something that already exists as an actual package 2024-09-19 18:21:43 yeah gzip is not the best example 2024-09-19 18:21:48 'vi' perhaps 2024-09-19 18:21:55 yeah, that could work 2024-09-19 18:22:00 or the example from yesterday 2024-09-19 18:22:18 exfat-utils 2024-09-19 18:22:24 which only has 1 out of 5 binary names overlapping 2024-09-19 18:22:42 where if you go by cmd:, version breaks ties, but if you 'apk add exfat-utils', priority does 2024-09-19 18:22:45 (right?) 2024-09-19 18:22:55 yes 2024-09-19 18:24:17 It's not a real alternatives system, because with alternatives, you could have multiple implementations installed and you just change which implementation is executed 2024-09-19 18:24:27 yes, understood 2024-09-19 18:26:50 thanks again. going to let that draft PR sit a while for feedback 2024-09-19 18:28:34 Regarding the MR, how will you decide what package has priority? 2024-09-19 18:29:42 do you mean when generating cmd: from alternatives? 2024-09-19 18:29:58 yes 2024-09-19 18:30:09 How do one change the gdm keyboard mapping? 2024-09-19 18:30:40 ikke, i hoped to cheat with the data already available here, but it might not fly - it's per command too, not per (sub)package https://github.com/openwrt/packages/blob/master/utils/gzip/Makefile#L32 2024-09-19 18:31:11 ah ok, yes, if it already has some priority number, that could be used 2024-09-19 18:31:25 yeah, but what if the second line is 350 2024-09-19 18:31:34 can't stick that in the package priority val 2024-09-19 18:31:49 why not? 2024-09-19 18:31:51 oh 2024-09-19 18:31:58 yeah, you cannot have different values 2024-09-19 18:32:02 ack 2024-09-19 18:32:08 as you pointed out yesterday and i also have clearly seen now 2024-09-19 18:32:25 could also use the alternatives priorities as version numbers of some form perhaps 2024-09-19 18:32:44 yes, don't see why that could not work 2024-09-19 18:32:45 provides:cmd:gzip-0.p300 (i don't know the syntax rules) 2024-09-19 18:32:59 might be a better fit than the actual version number compare alpine uses now for that in fact 2024-09-19 18:32:59 gzip=300-r0 2024-09-19 18:33:01 that should work 2024-09-19 18:33:12 right, i'd love some hint that it's from a prio, but got it 2024-09-19 18:33:15 clearly that would work 2024-09-19 18:33:35 ikke, you do mean with cmd: in front, right 2024-09-19 18:34:07 Yes, note that the prefix is nothing special 2024-09-19 18:34:14 just a convention 2024-09-19 18:34:17 yeah, just a namespace, saw that in the apk docs 2024-09-19 18:34:20 'we allow : so you can namespace' 2024-09-19 18:34:21 nod 2024-09-19 18:34:30 but if apk users might be used to cmd:, i'm doing the same 2024-09-19 18:34:36 yeah, makes senswe 2024-09-19 18:34:38 sense 2024-09-19 18:35:28 haiku supports, in fact enforces in some lint stage, that (example) vis-0.8.recipe has 'cmd:vis = $portVersion' in PROVIDES 2024-09-19 18:35:32 thus delivering the same functionality 2024-09-19 18:35:45 (haiku has no shame in taking good inspiration from various places) 2024-09-19 18:35:51 :) 2024-09-19 18:37:27 So apparently grub-install does not take /etc/default/grub in account 2024-09-19 18:38:06 Setting GRUB_PRELOAD_MODULES in there only works for grub-mkconfig 2024-09-19 18:38:24 grub-install requires to add them with --modules 2024-09-19 18:38:31 :/ 2024-09-19 19:20:01 you may also need to set partition-number bios_grub on for gpt table. 2024-09-19 19:20:22 have this earmarked in case it is important 2024-09-19 19:21:58 nvm, that was related to a bios boor partition on a gpt system. DK what that even means lol 2024-09-19 19:22:42 I guess if not using EFI? 2024-09-19 21:51:09 how are you guys copying files to the initramfs? i notice that there is features.d, which has .files, but some stuff seems not there 2024-09-19 21:51:31 for example, the cryptsetup.files says /sbin/cryptsetup, but cryptsetup has many dependencies, as you can see when you do `ldd /sbin/cryptsetup` 2024-09-19 21:51:48 i couldnt find those libraries in any other `.files` either 2024-09-19 22:40:29 ANyone has faced lightdm failing to start because of "Failed to open PAM session: Permission denied"? 2024-09-20 03:58:28 rdbo: it uses https://github.com/ncopa/lddtree to resolve dependencies 2024-09-20 06:19:32 Hi guys. I'm thinking of buying another display for my programming/non-programming needs. I already have 24" which is currently 1920x1200 @ 59.950 Hz but will rotatate it 90C clockwise for my coding needs. I was thinking to get 27" display mainly for reading manuals, web and debug screen (vertical split). Was thinking to get: Dell S2721QSA 27 Inch 4K UHD (3840 x 2160). Will 2024-09-20 06:19:34 there be a big difference between my current FHD and 4k? 2024-09-20 06:20:06 I'm using sway which I think supports all kind of scaling? 2024-09-20 06:50:14 coredumb: maybe this is related? 390b824324986cc9f4a8cca5e13d9c09b8e88855 2024-09-20 07:27:28 c99 -o svkbd-wvkbd drw.o svkbd-layout.o util.o -L/usr/X11R6/lib -lX11 -lXtst -lXft -L/usr/X11R6/lib -lXinerama pkg-config --libs fontconfig pkg-config --libs freetype2 /usr/lib/gcc/aarch64-alpine-linux-musl/13.2.1/../../../../aarch64-alpine-linux-musl/bin/ld: cannot find Scrt1.o: No such file or directory 2024-09-20 08:32:28 it's in musl-dev 2024-09-20 08:32:28 how make apk for git version? like qt6ct-git 2024-09-20 08:32:36 (latest commmit) 2024-09-20 08:58:25 usually we try to avoid packaing random git commits 2024-09-20 08:59:20 it is the upstreams responsibility to make releases. not us 2024-09-20 09:06:37 Hello o/ 2024-09-20 09:06:49 How can I find which package is installing the plugdev group? 2024-09-20 09:12:13 can busybox `ps -o ...` work on a specific pid? 2024-09-20 09:12:35 so it gives info on specific pid only 2024-09-20 09:43:24 nvm, dont need it! 2024-09-20 10:08:56 coredumb: multiple packages could do that. probably grep aports would be easiest. 2024-09-20 11:06:41 a lot of times git version is useful can i do it with apk? 2024-09-20 11:23:39 You'd have to manually keep track of the commit that was used. 2024-09-20 13:30:19 morning all, Installed alpine 3.20.3 x86 on Dell Venux 5830 with Intel Atom Z3740D installed linux-firmware wlan device not detected. 2024-09-20 13:30:34 udhcpc-i wlan0 udhcpc: ioctl 0x8933 failed: No such device. 2024-09-20 13:30:51 cat /proc/net/dev only lo and eth0 shown 2024-09-20 13:32:07 seeing many ath6kl failed in dmesg 2024-09-20 13:32:38 do you have the related kernal modules loaded for the specific hardware? Maybe it is not in linux-firmware? I remember seeing a useful link for this in wiki, let me try to find. Can't promise it is relevant 2024-09-20 13:33:47 I think linux-firmware is a metapackage that brings in everything though 2024-09-20 13:34:15 https://pkgs.alpinelinux.org/packages?name=linux-firmware-%2A&branch=edge&repo=&arch=x86_64&maintainer= 2024-09-20 13:34:41 I dk if ath6k is related to 6ki: https://pkgs.alpinelinux.org/package/edge/main/x86_64/linux-firmware-ath6k 2024-09-20 13:36:19 this is a bay trail device, information is a little slim, I did download and installed "new" ath6kl hw3.0 with this result. 2024-09-20 13:36:24 ath6kl: Failed to get default board file ath6k/AR6004/hw3.0/bdata.bin: -2 2024-09-20 13:36:39 Sorry, only other thing I found is https://wiki.alpinelinux.org/wiki/Wi-Fi#Troubleshooting 2024-09-20 13:37:00 let me check the linux-firmwar-ath6k 2024-09-20 13:37:22 also see if you need to load kernel module 2024-09-20 13:37:30 you see several modprobe commands there 2024-09-20 13:37:36 (in the troubleshooting) 2024-09-20 13:37:49 I just dk what the ath6ki kernel module would be 2024-09-20 13:38:07 yes linux-firmware-ath6k installed 2024-09-20 13:38:49 do you have wpa_supplicant? I dk if that would affect seeing the interface though 2024-09-20 13:39:09 it shouldn't 2024-09-20 13:39:16 what are the dmesg messages? 2024-09-20 13:40:08 yes wpa_supplicant is installd 2024-09-20 13:40:25 ath6kl: Failed to get board file ath6k/AR6004/hw3.0/bdata.bin (-2), trying to find default board file. 2024-09-20 13:40:25 ACTION sent a code block: https://matrix.org/oftc/media/v1/media/download/AQ1fg7AFaSonKK2JKOnJm6fg6Uv5yhn8QOSbzQSWOx6s_73NE4rp-1bD40JEZSQ0Rm5ORfMtPKRnk3KQKoAdwiRCeSEWEJ0wAG1hdHJpeC5vcmcvbEdvR2dHdExmZE5oQlJtYVpJSmxMbWFU 2024-09-20 13:40:38 th6kl_sdio mmc0:0001:1: Direct firmware load for ath6k/AR6004/hw3.0/bdata.bin failed with error -2 2024-09-20 13:40:53 ath6kl: Failed to get default board file ath6k/AR6004/hw3.0/bdata.bin: -2 2024-09-20 13:41:07 ath6kl: Failed to init ath6kl core 2024-09-20 13:41:18 ath6kl_sdio: probe of mmc0:0001:1 failed with error -2 2024-09-20 13:41:36 what does the i mean? 2024-09-20 13:42:10 a quick search shows that other people across yocto and others have had this problem 2024-09-20 13:42:49 eg https://community.nxp.com/t5/i-MX-Processors/SX-SDCAN-on-IMX6Q-yocto/m-p/268243 2024-09-20 13:42:58 https://wireless.wiki.kernel.org/en/users/Drivers/ath6kl 2024-09-20 13:43:07 not with your dell specifically, but with the ath6kl 2024-09-20 13:43:17 have you tried grabbing firmware from this link, putting it in place and rebuilding initramfs? 2024-09-20 13:45:31 For me with amdgpu it looked something like this on Deb:... (full message at ) 2024-09-20 13:46:03 I dk what this may look like on alpine for commands, but it may be similar 2024-09-20 13:47:40 the linux-firmware repo does not have firmware for revision hw3.0 of the AR6004 2024-09-20 13:47:57 but https://github.com/qca/ath6kl-firmware/tree/master/ath6k/AR6004 does 2024-09-20 13:48:13 so grabbing firmware and rebuilding initramfs should be what is required? 2024-09-20 13:48:43 i have no idea if the rebuild is necessary for wifi 2024-09-20 13:48:52 i'd start with grabbing the files from that repo and putting them in the right path 2024-09-20 13:49:15 I downloaded AR6004 hw3.0, -> dmesg are hw3.0 2024-09-20 13:49:32 owain, good! 2024-09-20 13:50:15 so after installing them need to run update-inframs? 2024-09-20 13:50:27 Habbie said try without first 2024-09-20 13:50:35 wow, busybox has a lot of old reported bugs ... 2024-09-20 13:50:53 that were I am...:( 2024-09-20 13:50:53 owain, i know the firmware stuff, i don't know the alpine stuff, so try without (perhaps reboot), if the error remains, do that 2024-09-20 13:51:03 i'll rephrase 2024-09-20 13:51:22 owain, i know the firmware stuff, i don't know the alpine stuff, so put the files in the right path (and reboot), if the error remains, do the initramfs 2024-09-20 13:51:28 Sounds like good course of action. #babysteps 2024-09-20 13:55:12 fingers crossed 2024-09-20 13:58:40 If this wasn't bridged to IRC my favorite "it's working" gif is usually anakin in a podracer lmao 2024-09-20 14:03:37 I downloaded linux-firmware-20240909 the ath6k/AR6004 only have hw1.2 and hw1.3 no hw3.0 2024-09-20 14:05:10 that's why i gave you a url that has hw3.0 2024-09-20 14:10:12 pulled one down from https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git 2024-09-20 14:14:15 bummer, same date and size as I have installed Habbie. 2024-09-20 14:14:33 hmm? 2024-09-20 14:16:05 I compared the owns on my local desktop and the ones in your link, they are the same size and date, so their the ones I installed already. 2024-09-20 14:17:25 i don't follow. you have hw3.0 files? 2024-09-20 14:20:36 dmesg -> ath6kl: Failed to get board file ath6k/AR6004/hw3.0/bdata.bin (-2), trying to find default board file. 2024-09-20 14:20:47 ok, and that file is in the right place? 2024-09-20 14:21:12 they are installed in lib/firmware/ath6k 2024-09-20 14:21:17 ok 2024-09-20 14:21:22 then perhaps try the initramfs update 2024-09-20 14:22:31 same, but why the ath6kl at the first of the dmesg line? 2024-09-20 14:23:17 that's the name of the driver 2024-09-20 16:20:04 Habbie, trying another hw3.0 from https://github.com/TerryLv/ath6kl_firmware 2024-09-20 16:34:41 same errors no wlan 2024-09-20 16:37:32 (i know owain left) 2024-09-20 16:37:34 error -2 is file not found 2024-09-20 16:37:40 oh hi you're back 2024-09-20 16:37:44 error -2 means file not found 2024-09-20 16:37:50 so trying different files won't help 2024-09-20 16:37:54 the problem comes before that 2024-09-20 16:39:02 also, the files from TerryLv are in fact identical 2024-09-20 16:41:16 ok, I remove linux-firmware and linux-firmware-ath6k, deleted /lib/firmware/ath6k folder, ran mkinitfs, rebooted and installed linux-firmware-ath6k ran mkinitfs rebooted 2024-09-20 16:41:50 So i'm back to default 2024-09-20 17:04:09 "You'd have to manually keep..." <- i'd like to build latest commit as -git packages. not sure to how do it 2024-09-20 17:04:56 manually 2024-09-20 17:05:13 Note that for aports packages, we generally want to avoid that 2024-09-20 17:05:51 manually so no abuild and no package? 2024-09-20 17:06:24 in pacman `source=("${pkgname}::git+https://codeberg.org/realroot/wotimer.git")` 2024-09-20 17:06:36 `source=("${pkgname}::git+https://codeberg.org/realroot/wotimer.git")` 2024-09-20 17:08:03 i'd glady submit the non git version to aports but right i cannot boot anymore and i had to use pmos 2024-09-20 17:08:09 *right now 2024-09-20 17:08:54 scorpion2185[m]: manually as in, add a variable that contains the commit, and update it when you need a new version 2024-09-20 17:09:48 can you put a function that gets it? also i find really nice the version function: 2024-09-20 17:09:48 `git describe --long --tags | sed 's/^v//;s/\([^-]*-g\)/r\1/;s/-/./g'` 2024-09-20 17:15:22 The checksum needs to match, so what version it gets needs to be static 2024-09-20 17:16:24 can you skip it ? 2024-09-20 17:16:59 usually is skipped in -git 2024-09-20 17:17:27 No, abuild has no support for that 2024-09-20 17:18:03 We do not have any -git packages in aports 2024-09-20 17:18:33 in the future maybe? adding skip should not be hard i think 2024-09-20 17:21:59 For aports there is little use for it, since such packages are against policy 2024-09-20 17:23:59 for users is helpful, any plan to have alpine aur? 2024-09-20 17:26:05 No 2024-09-20 17:30:43 isee i guess builing manually will be easier, thanks 2024-09-20 17:31:09 (feature request -git and alpine aur) 2024-09-20 17:31:24 *feature suggestion 2024-09-20 17:40:09 Habbie, I got it working :) The qca/ath6kl-firmware, has AR6004 with hw1.3 and hw3.0, both have to be copied to /lib/firmware/ath6k/AR6004 2024-09-20 17:41:04 I was copying only the hw3.0 folder and contents. Go figure 2024-09-20 17:46:38 huh, interesting 2024-09-20 17:46:40 but, happy it works! 2024-09-20 17:49:41 imaging time and then onto a desktop for the tablet. 2024-09-20 17:49:58 o/ 2024-09-20 17:50:11 don't know if something has changed but my alpine no longer boots using UKI 2024-09-20 17:50:28 I don't even get any output at all (no kernel, no initrd) not even by removing quiet 2024-09-20 17:59:50 owain, nice :) 2024-09-20 18:01:06 markand: lots of things are changing all the time, so hard to say what specific thing could have broke that 2024-09-20 18:02:06 markand, it might help to tell us what version you are on, if you ran any updates before this reboot (i see how that might be hard to check), etc. 2024-09-20 18:26:05 let me check 2024-09-20 18:30:18 kernel is 6.6.51-0-lts 2024-09-20 18:31:03 now there is: EFI stub: loaded initrd from command line option and then nothing 2024-09-20 18:31:09 still able to ctrl+alt+del 2024-09-20 18:32:33 markand: edge or stable? 2024-09-20 18:32:41 I also see there is a .52 kernel vailable 2024-09-20 18:32:43 3.20 2024-09-20 22:52:59 Are there any non-free packages in the alpine? 2024-09-20 22:52:59 How can I use apk info or apk-tools to find this out? 2024-09-20 23:00:55 A bunch of firmware is proprietary, no? 2024-09-20 23:02:45 I am asking, since I want to try to write a blacklist, so see if I can make it free 🙂 2024-09-20 23:02:45 Can I use APK to list licenses of all packages? 2024-09-20 23:02:45 I found the firmware as well. 2024-09-21 00:07:22 wizzard: you may be interested in something like https://github.com/suve/vrms-rpm 2024-09-21 00:30:03 Thank you 2024-09-21 00:47:14 hi guys 2024-09-21 00:47:19 question 2024-09-21 00:47:36 alpine is full of gnu core and utils? 2024-09-21 00:47:45 or some packages need it? 2024-09-21 01:22:18 asimovc: what 2024-09-21 01:22:35 alpine has gnu coreutils and other stuff but they're purely optional in most cases 2024-09-21 04:55:15 is there a default volume setting? both my hdmi and headphone volume was at 86, dont remember how i set them like that or if its some default... i just realised i have only been using app volumes didnt even have alsamixer or tinyalsa installed, might have had those on the system at some point tho 2024-09-21 09:14:19 but how do you use the commit in source? 2024-09-21 09:14:36 "scorpion2185: manually as in..." <- ^ 2024-09-21 09:26:05 scorpion2185[m]: for examples to look at: clone the aports git repo, then do 'find aports -name APKBUILD -exec grep -l '_commit' {} \;' 2024-09-21 09:28:01 the whole repo? i guess it's big 2024-09-21 09:29:47 yeah, i guess 2024-09-21 10:15:19 https://gitlab.alpinelinux.org/search?search=_commit%3D&nav_source=navbar&project_id=1&group_id=2&search_code=true&repository_ref=master 2024-09-21 11:54:19 hi, is there any time schedule for next release of 3.21 ? 2024-09-21 12:23:00 qaqland: it's about 6 months between releases (release=y of x.y.z) from what i understand. 2024-09-21 12:23:27 so november or december 2024-09-21 12:31:12 invoked: thx 2024-09-21 12:54:48 Correct, november is the target 2024-09-21 15:39:10 hi, any ideas why mount.nfs shows an flock usage error and then blocks until ^C? https://termbin.com/cxh0 2024-09-21 15:48:04 oh, i had to 'rc-update add nfsmount' and 'service nfsmount start', just 'apk add nfs-utils' was not enough 2024-09-21 15:49:08 found in some blog post, alpine docs/wiki only cover nfs server and seem to have no mention of client setups 2024-09-21 15:56:27 https://gitlab.alpinelinux.org throws a 404, linked to in the top right corner on https://alpinelinux.org 2024-09-21 16:16:31 they fixed gitlab 2024-09-21 16:17:27 someone pointed out in private that upstream fixed the mount.nfs/flock issue with http://git.linux-nfs.org/?p=steved/nfs-utils.git;a=commit;h=ee785fb6c79f0b27c2144faaafadaf3267419470 2024-09-21 16:28:29 Hi kn, see if the wiki page looks correct https://wiki.alpinelinux.org/wiki/Setting_up_an_NFS_server . feel free to edit, if something is still missing 2024-09-21 16:30:04 prabuanand: "Setting up an NFS server", but i needed to configure an NFS client 2024-09-21 16:30:53 oooooh there is a client paragraph, i just never looked at this page due to the title 2024-09-21 16:31:15 unfortunately title of pages cannot be changed. I just updated the page 2024-09-21 16:32:04 feel free to edit the wiki page, if something is still missing 2024-09-21 17:38:41 i am tryng to set local repo 2024-09-21 17:38:58 do I need to use an email in /etc/abuild.conf? 2024-09-21 17:39:46 abuild-sign -k .abuild/user-66ed2fa9.rsa packages/apps/aarch64/APKINDEX.tar.gz 2024-09-21 17:39:46 Could not open file or uri for loading private key from .abuild/user-66ed2fa9.rsa: No such file or directory 2024-09-21 17:45:29 I have a POWER8 system and I'm trying to install alpine installer finishes but after rebooting the newly installed os does not show up in petitboot at all 2024-09-21 17:49:25 scorpion2185[m]: check ~/.abuild/abuild.conf 2024-09-21 17:52:38 PACKAGER_PRIVKEY="/path-to-home/.abuild/user-66ed2fa9.rsa" 2024-09-21 17:53:50 how clone this only https://git.alpinelinux.org/aports/tree/community/sxmo-utils?h=master ? 2024-09-21 17:57:53 scorpion2185[m]: clone with --depth=1 2024-09-21 17:58:12 It will get the entire repo, but just a single commit 2024-09-21 17:58:19 which is light enough for aports 2024-09-21 18:03:22 thanks, is there a guide to set a local repo? 2024-09-21 18:03:25 https://www.linkedin.com/pulse/creating-alpine-linux-package-repository-afam-agbodike 2024-09-21 18:04:58 scorpion2185[m]: Just follow the steps on how to build packages. The built packages end up in ~/pacakges/, and you can directly serve that as an apk repository 2024-09-21 18:06:54 scorpion2185[m]: https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package#Setup_your_system_and_account 2024-09-21 19:51:47 how build only a subpackage? 2024-09-21 19:54:38 You can't 2024-09-21 19:57:24 btw I had to specify full path of  .abuild/user-66ed2fa9.rsa 2024-09-21 20:17:03 can I make something like pacman hooks ? to do things when a package is installed 2024-09-21 20:17:52 yes, there are triggers 2024-09-21 20:18:12 or install hooks 2024-09-21 20:21:59 Hey guys, scratching my head over here. So, I decided I wanted to try Alpine Linux for the first time. Install went great, rebooting into it went great, drivers and everything. Not a problem. However, the other day I thought I was going to check out refind. I've been using grub for years and years. So, refind is on /dev/nvme1n1p1 on an fat efi partition. I tried adding alpine which is located at /dev/nvme0n1p3. P1 is grub efi, 2024-09-21 20:21:59 P2 is swap, P3 is root. 2024-09-21 20:22:13 any example or article? 2024-09-21 20:23:58 ah like in sxmo-utils so they are part of the package? 2024-09-21 20:24:11 sxmo-utils-dwm.post-install 2024-09-21 20:24:12 The problem is....I've added a manual stanza to boot alpine. It boots in to initramfs rescue shell, saying it can't boot DISK, UUID, PARTUUID or LABEL (depending on what I've tried) to /sysroot. No such file or directory. I have to use VI to manually edit /etc/fstab and add either UUID, DISK, ETC. to it, and continue the boot. The fstab on the booted system has all the correct listings. In the initramfs, only cdrom and usb are 2024-09-21 20:24:12 listed. 2024-09-21 20:25:34 I have another linux install so I can chroot in to the alpine system, no issue, but I don't know where to go from there. Problem I guess is the initramfs not having the relevant info in its fstab 2024-09-21 20:25:47 So, can anyone please help? :) 2024-09-21 20:31:50 Yes, I've also appended rootfstype=ext4 2024-09-21 21:04:42 maybe you can try reverting to grub 2024-09-21 21:06:30 thanks to @minimal, I have now booted in to alpine linux :) 2024-09-22 01:01:46 hi all - i just completed a fresk desktop alpine installation on an AMD based system, installing GNOME using the setup-desktop script which went without issue. now when i boot i get brought to the GDM login screen and the system is totally unresponsive - i can't use the mouse, pull up another TTY, nothing. the only thing i can do is use the power button to reboot the box, and the issue recurs every single time i try 2024-09-22 01:02:43 i've previously used GNOME on Alpine, but doing a manual install not the setup-desktop method, and i've seen this before but it was always intermittent - rebooting once or twice would get me back to the login screen and become responsive. this doesn't appear to be the case anymore, and i'm a bit stumped as to why? has anyone encountered this behavior before? 2024-09-22 06:04:44 any indications of a kernel panic or anything? 2024-09-22 06:20:59 lbkodjxj[m]: make sure you have all the libinput stuff if it's wayland 2024-09-22 06:59:48 are triggers or install hooks part of the package itself? are there other package hooks? 2024-09-22 07:02:57 install hooks are part of the package, triggers typically are added to other packages 2024-09-22 07:04:01 can you give me an example of a trigger? 2024-09-22 07:04:49 busybox 2024-09-22 07:06:35 this https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/main/busybox/busybox.trigger ? 2024-09-22 07:06:49 That's the trigger script itself, yes 2024-09-22 07:08:19 to make one do I need a trigger script and `triggers="busybox.trigger=/bin:/usr/bin:/sbin:/usr/sbin:/lib/modules/*:/usr/lib/modules/*"` ? 2024-09-22 07:08:29 yes 2024-09-22 07:09:09 those paths means what? 2024-09-22 07:10:36 whenever something goes there? 2024-09-22 07:10:43 Or is changed 2024-09-22 07:12:23 where is installed busybox.trigger? 2024-09-22 07:15:01 /lib/apk/db/triggers 2024-09-22 07:15:39 and /lib/apk/db/scripts.tar 2024-09-22 08:32:15 how run abuild -r so that it won't ask confirmation? 2024-09-22 08:56:16 opposite of -i 2024-09-22 09:32:43 https://alpinelinux.org/downloads/ 2024-09-22 09:32:49 The raspberry pi section lists each architecture twice 2024-09-22 09:32:51 Is this intended? 2024-09-22 09:48:50 runxiyu: yeah for tar.gz and img 2024-09-22 09:49:23 scorpion2185: you mean --no-interactive? 2024-09-22 10:20:06 scorpion2185[m]: what does it ask confirmation on? 2024-09-22 10:58:44 fossdd[m]: Ah, my eyes weren't working 2024-09-22 11:23:06 i have /etc/apk/interactive maybe made by pmos 2024-09-22 11:23:30 it ask confirmation on installing deps 2024-09-22 11:28:08 --no-interactive indeed. not sure if I can add it to abuild -r, will check 2024-09-22 11:28:40 he's not here https://www.mankier.com/8/apk 2024-09-22 11:30:39 you could just remove /etc/apk/interactive 2024-09-22 11:30:59 when not running abuild -r i think that i prefer to confirm 2024-09-22 11:31:57 There is no option to include --no-interactive to abuild 2024-09-22 11:34:30 sertonix: i patched sxmo to use dinit for user services looks all good 2024-09-22 11:34:48 is it possible to use dinit as system init too? 2024-09-22 11:36:15 Is using alpine on "the desktop" with a full desktop evironment like gnome a usecase that is recognized by alpine devs and/or expected to work well? Where would I go for documentation on this question/potential usecase? 2024-09-22 11:39:00 setup-desktop 2024-09-22 11:48:40 chripyVagrancy https://docs.alpinelinux.org/user-handbook/0.1a/Working/post-install.html 2024-09-22 13:50:01 thanks scorpion 2024-09-22 15:48:35 Hi, i like to install alpine-linux on my Odroid C4? Is it difficult? Can someone give me some tips? 2024-09-22 18:29:23 i am making a package that has only triggers. 2024-09-22 18:29:23 `>>> ERROR: : vlc: trigger script does not match pkgname or any subpackage` 2024-09-22 18:33:47 You need at least `mkdir -p "$pkgdir"` in the package() function 2024-09-22 18:40:49 i keep having that error, i saw in busybox package that .trigger is not in source= 2024-09-22 20:59:19 I keep getting 500 trying to open an MR for: https://gitlab.alpinelinux.org/WhyNotHugo/aports/-/merge_requests/new?merge_request%5Bsource_branch%5D=himitsu-git 2024-09-22 21:02:38 try to rebase upon newer master 2024-09-22 22:45:44 Is there a decently powerful arm64 server/desktop that can run alpine pretty well? I have a rockpro64, but would rather have something a little more powerful and more widely supported. 2024-09-23 02:18:39 ^same question, elagost 2024-09-23 06:11:37 I've been happy enough with my Rock5B (rk3588 based) 2024-09-23 06:25:40 Not exactly what was asked for, but I feel like there's a dearth between that sort of higher end SBC and the giant servers 2024-09-23 07:32:22 whats the best virtual machine maker/manager in repo? qemu? 2024-09-23 07:58:21 i like virsh + Virtual Machine Manager 2024-09-23 08:39:31 qemu-openrc :D 2024-09-23 08:40:51 oh i went for qemu 2024-09-23 08:41:01 never tried virsh 2024-09-23 08:43:45 virsh is an "abstract" manager, it relies on qemu 2024-09-23 08:44:37 but if you are ok with qemu, i suppose that you won't need/like it 2024-09-23 08:45:11 (I prefer plain qemu, got dozen of VMs) 2024-09-23 08:46:03 oh cool 2024-09-23 09:24:43 you can also use virsh with Xen (but I prefer plain xl) 2024-09-23 16:16:56 Hi, i raised an issue https://gitlab.alpinelinux.org/alpine/docs/user-handbook/-/issues/6 about the lack of sync between docs.alpinelinux.org and docs gitlab. Who's the right person to fix this? 2024-09-23 17:16:45 Me, but I've also been busy with other things 2024-09-23 18:03:56 how can i check if im using eudev and/or busybox-mdev? 2024-09-23 18:06:46 frag: What’s installed? What services are enabled and/or running? 2024-09-23 18:07:59 humm eudev budybox-mdev-openrc, the latter was in world but i see also alpine-base depends on it... i use the rapsi as desktop/xorg .. some udev in rc-update 2024-09-23 18:11:49 hmm setup-xorg-base installs eudev 2024-09-23 18:13:29 i also have udev-init-scripts(-openrc) in world 2024-09-23 18:19:15 can busybox format ext4? or do you need e2fsprogs or similar? 2024-09-23 18:29:42 at least Alpine's busybox has no mkfs.ext4. It has mkfs.ext2, but you're not asking for that. 2024-09-23 18:29:58 Even to mkfs.ext2, to convert it to ext4 you'd need a tune2fs option that busybox tune2fs doesn't have 2024-09-23 19:52:36 ok i keep it :] 2024-09-23 22:35:18 hello, i have a raspberry pi , no screen and no keyboard. I'm not sure how to program it. in the case of raspberry pi os , there's this file you create to enable ssh by default .. is there something similar on alpine ? 2024-09-23 22:35:57 or do I have to chroot ? 2024-09-23 22:37:18 welp , i don't think chroot is an option .. this is a virtual machine thing (the sdcard doesn't contain the filesystem, so it's an alpine in diskless mode) 2024-09-23 22:37:30 any ideas ? 2024-09-23 22:50:24 jbara[m]: maybe something like this? https://github.com/macmpi/alpine-linux-headless-bootstrap 2024-09-23 22:51:36 used it for an older pi, enables ssh and built-in wireless. worked well for me 2024-09-23 23:30:23 cool 2024-09-23 23:30:31 how did you get it to connect to the network ? 2024-09-23 23:37:53 jbara[m]: wpa_supplicant config. for ethernet an /etc/network/interfaces config is probably sufficient, haven't tried 2024-09-23 23:40:01 the repo lists the extra config files it supports. if recalled correctly, just drop them in the top level of the boot partition next to the headless.apkovl.tar.gz and it'll do its thing on startup 2024-09-24 00:45:15 I cant control my keyboard and mouse in alpine edge 2024-09-24 00:45:19 How to fix please? 2024-09-24 00:45:26 After i did upgrade, i cant control 2024-09-24 00:45:30 How to show lpg? 2024-09-24 00:45:32 *log 2024-09-24 01:09:59 templecloud: What hardware do you have, and what DE/WM? 2024-09-24 01:29:53 Rather what display system 2024-09-24 03:03:20 let me send the log later. thank you all~ 2024-09-24 05:03:08 oh nothing, i just added myself to "input" group and solved 2024-09-24 05:03:09 :) 2024-09-24 05:06:47 that is so annoying problem, had to put 'groups|grep -q input && startx ...', run from RAM and in the startup it doesnt alway add all packages and set up correctly ... kinda weird 2024-09-24 05:20:51 init_socket_name()] server 0x7f65ef6a34f0: name pipewire-0 is not an absolute path and no runtime dir found. Set one of PIPEWIRE_RUNTIME_DIR, XDG_RUNTIME_DIR or USERPROFILE in the environment 2024-09-24 05:20:54 i'm stuck in here 2024-09-24 05:20:55 :( 2024-09-24 05:35:57 ok solved 2024-09-24 05:35:59 weird 2024-09-24 08:53:02 why no one answer if they can use file-chooser eg in browser? my apps crash when i try 2024-09-24 09:04:23 run it from the console, see what it prints 2024-09-24 09:15:44 lopid i only get 'Segmentation fault', are you saying you have no problem with it? im on edge btw.. 2024-09-24 09:15:53 no 2024-09-24 09:16:13 anything in ~/.session-errors? 2024-09-24 09:17:05 i dont have that file lopid 2024-09-24 09:34:26 i meant ~/.session-errors 2024-09-24 09:34:38 sigh. did it again. ~/.xsession-errors 2024-09-24 10:11:48 i dont have that file either 2024-09-24 14:12:51 lol I cat /dev/mapper/luks to see what it would spit out, and it encrypted text in terminal after I interrupted 2024-09-24 14:28:32 there is a console code to change output mode which happened to be contained in your data. you can fix it with "reset" command 2024-09-24 14:39:06 cool, thx. I haven't seen that before 2024-09-24 16:03:12 anyone know where one can see documentation about the cryptdm kernel parameters? 2024-09-24 16:03:26 trying to find docs regarding options, not going so well 2024-09-24 16:04:26 tried linux kernel docs and dmcrypt docs 2024-09-24 16:05:10 andar1an[m]: the -m option here: https://man.uex.se/1/nlplug-findfs 2024-09-24 16:05:43 https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/initramfs-init.in?ref_type=heads#L586-588 2024-09-24 16:06:50 oh derp, it is just the device name 2024-09-24 16:07:13 that is easy. I thought cryptdm options were some specific subset for fs types 2024-09-24 16:07:17 thank you Ikke 2024-09-24 16:07:50 np 2024-09-24 17:21:19 To confirm: "resume" feature in mkinitfs config is related to busybox init, and is only necessary for hibernation resume? 2024-09-24 17:22:17 I am pulling this data from: https://wiki.archlinux.org/title/Mkinitcpio, which is not mkinitfs, but I think it would be the same? 2024-09-24 17:23:11 I don't see resume in : https://gitlab.alpinelinux.org/alpine/mkinitfs/-/tree/master/features.d?ref_type=heads 2024-09-24 17:23:40 It was referenced in a wiki, so this is just to validate 2024-09-24 18:53:21 can you use mkinitcpio in alpine? 2024-09-24 20:15:37 dk, maybe if it is just building initramfs and packaging up kernel and modules and bootloader? 2024-09-24 20:15:37 Resume seems to be more related to busybox init than something alpine specific 2024-09-24 20:15:57 half the docs I read are arch or gentoo lol 2024-09-24 20:16:52 right not trying to figure out where document exists for grub-pre.cfg. I think this may be alpine specific and to do with maybe needing the file in /root/ to chainload 2024-09-24 20:18:39 may be more general for luks2 and grub if other distros need to put it in /root/ too. DK if it is to do with varying init systems or not yet 2024-09-24 20:22:34 according to arch grub-pre.cfg may be outdated 2024-09-24 20:22:41 so may as well giv'r a go without first 2024-09-24 22:06:42 Was linux-edge removed? 2024-09-24 22:19:50 i added the edge testing repository about 5 minutes ago and downloaded XFE. 2024-09-24 22:47:27 Ermine: i don't think so? how come 2024-09-24 22:48:14 ptrc: apk search linux-edge -> nothing (on i386) 2024-09-24 22:49:20 i don't think it was ever available on i386? 2024-09-24 22:49:42 well, ok 2024-09-25 06:09:42 is there an easy way to get overview of which repo the installed packages comes from? only found add policy pkg. Also can one search for "contennt" with apk command, or only on alpinelinux.org? 2024-09-25 06:35:31 if its woman who ran it, it would be too funny .. thats what happened in norway supposedly 2024-09-25 06:36:04 ah, wrong window 2024-09-25 07:50:22 frag apk-file iirc, it fetches from the site 2024-09-25 08:03:49 While it still works 2024-09-25 08:03:55 It scrapes the html 2024-09-25 11:16:12 how package this game ? https://github.com/NixOS/nixpkgs/blob/release-21.11/pkgs/games/openarena/default.nix 2024-09-25 11:16:36 https://repology.org/project/openarena/versions 2024-09-25 11:19:23 https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package 2024-09-25 11:21:00 point is i am not sure to how package it well 2024-09-25 11:21:30 i find hard to understand what commands and things i should do, in fact i am not sure to how build it 2024-09-25 11:22:51 toAPK can read nixos now that i think abouit i 2024-09-25 11:23:54 when you know the build commands, you can check how other entities packaged stuff here: https://gitlab.alpinelinux.org/alpine/aports 2024-09-25 11:24:34 i do not know the build commands i am familiar with making apk packages 2024-09-25 11:26:00 for example an aur version does not work well for some reason 2024-09-25 11:29:17 you need to follow official instructions on how to build openarena, that's all 2024-09-25 11:29:38 if you encounter problems, you have to debug stuff 2024-09-25 11:30:04 It's totally ok if it doesn't work on the first try 2024-09-25 11:31:35 i looked their page on github and the instructions are not clear , gonna try toAPK but first i'll install on x86 alpine 2024-09-25 11:32:09 can somebody get this on alpine ?https://o3de.org/ 2024-09-25 11:32:19 it's pretty heavy 2024-09-25 11:39:53 top game engine . formerly closed software used by amazon games 2024-09-25 12:01:28 I have a problem using torsocks in alpine linux. I compare debian and alpine in a minimal example here: https://pastebin.com/Uz9EeiL2 Does anyone have a clue what could cause the DNS resolving problem in alpine linux? Is that musl-related? 2024-09-25 12:21:56 MrPink: can you compare the output with -v added? 2024-09-25 12:26:59 Greetings. 2024-09-25 12:27:23 How do I modify a user on alpine linux? 2024-09-25 12:27:32 The default shell in particular. 2024-09-25 12:31:40 apk add shadow && chsh whatever-user 2024-09-25 12:31:52 or just edit /etc/passwd 2024-09-25 12:39:43 shadow (no such package) 2024-09-25 12:40:37 Is that a community package? 2024-09-25 12:50:10 apk policy shadow 2024-09-25 12:50:10 yes 2024-09-25 12:54:05 same thing for usermod ? 2024-09-25 12:54:13 Or is there another way? 2024-09-25 13:09:10 ikke: With -v curl repeats the message "Could not resolve host: check.torproject.org" twice, thats all. No further information given. 2024-09-25 13:09:44 MrPink: Sorry, newcomer here, but that sounds more like a network problem. 2024-09-25 13:09:48 Can dig resolve it? 2024-09-25 13:41:12 cousin_luigi: No, not a network problem. 2024-09-25 13:57:10 I see. 2024-09-25 14:02:27 can somebody tell me the take of alpine on systemd? 2024-09-25 14:03:08 It doesn't exist 2024-09-25 14:08:27 can alpine join the [Init Freedom Campaign](https://beta.devuan.org/os/init-freedom)? 2024-09-25 14:10:31 ikke could you tell me your opinion? 2024-09-25 14:11:10 what's stopping you from just listing Alpine in the "GNU/Linux Distributions offering alternate inits" section 2024-09-25 14:11:43 me? 2024-09-25 14:12:20 What's there to "join"? 2024-09-25 14:12:44 writing somewhere that so that people know 2024-09-25 14:13:42 e.g. https://www.hyperbola.info/news/no-support-for-sudo-and-oriented-towards-init-freedom/ 2024-09-25 14:14:22 for example pmos did not follow that, somebody told me that it was just a topping on alpine 2024-09-25 14:14:23 I think it's a bit dumb tbh. Supporting multiple init systems is one thing but a campaign actively against systemd is ridicilous 2024-09-25 14:14:57 pmOS is a repository on top of Alpine + custom tooling to make development for phones easier, yes 2024-09-25 14:15:54 adding systemd that is not small,secure,simple makes it more distant from alpine 2024-09-25 14:16:33 adding it as an option does not make Alpine not small, secure and simple. And even if it was the only option that's still a debatable statement that's full of opinions 2024-09-25 14:17:12 not for me so are you in favor of systemd therefore? 2024-09-25 14:17:31 I'd like to use systemd yes. That does not mean I'm not in favour of supporting alternate init systems as well 2024-09-25 14:18:06 So a campaign called "init freedom" suggests just supporting/allowing multiple init systems. Not "actively not supporting systemd" 2024-09-25 14:18:25 no because systemd break init freedom 2024-09-25 14:18:29 *breaks 2024-09-25 14:18:35 no it doesn't 2024-09-25 14:18:42 yes 2024-09-25 14:19:00 anyway thanks for letting me know. 2024-09-25 14:19:21 some dev said that gnome maybe will be only systemd 2024-09-25 14:19:31 makes sense 2024-09-25 14:19:32 i do not care personally 2024-09-25 14:19:37 Don't take my opinions as the opinion of the Alpine project btw 2024-09-25 14:20:02 i think that you pmos dev 2024-09-25 14:20:20 but personally I do not want to see Alpine join some campaign against systemd. Supporting multiple init systems is fine though 2024-09-25 14:20:22 I am yes 2024-09-25 14:21:34 since alpine is supposed to be small simple secure i would feel bad if it starts even only having systemd 2024-09-25 14:22:10 scorpion2185[m]: since i do not use gnome 2024-09-25 14:22:15 I'm not advocating for having "only" systemd. But I'd love it to be an option 2024-09-25 14:22:31 alpine doesn't plan to be systemd-only 2024-09-25 14:22:52 does it plan to have it though? 2024-09-25 14:23:57 yes 2024-09-25 14:24:21 Will OpenRC be a first class citizen at that point? 2024-09-25 14:24:39 it is right now 2024-09-25 14:25:20 Ermine: there is no concrete plan to have systemd right now. But there are definitely people working on making it happen. It's a long-term thing for now though 2024-09-25 14:25:33 I understand, but I personally don't want to use systemd after using openrc. Knowing direction of alpine regarding this would be nice 2024-09-25 14:25:36 PureTryOut: yes 2024-09-25 14:26:08 andar1an: I'm pretty sure Alpine will never be systemd-only. Neither is it OpenRC only now, you can for example use dinit if you want, it's packaged already 2024-09-25 14:26:20 andar1an[m]: nobody will force you to use systemd here 2024-09-25 14:27:16 I love being able to choose, I have nothing against any init system. Diversity and competition breed innovation. But I don't want to get pigeonholed down the road with support falling off or some design choices to limit init options 2024-09-25 14:27:31 thanks, good to know 2024-09-25 14:28:07 The minimalism of Alpine is what I find sets it apart 2024-09-25 14:28:17 would suck to lose that 2024-09-25 14:29:35 Just waiting for fairphone 5 support to be out of testing for PMOS haha. wish I had a device to tinker now 2024-09-25 14:31:15 There are no plans to change init systems just yet in Alpine. 2024-09-25 14:31:54 I guess I will keep truckin on this setup then haha 2024-09-25 15:38:07 ACTION does not understand why the hartred on systemd 2024-09-25 15:39:09 Letting people choose between multiple init systems is one thing, being actively against systemd in the name of "init freedom" is something else entirely... 2024-09-25 15:40:35 heh, that Devuan link mentions distros that support systemd as an option. 2024-09-25 15:47:40 there are many legit complaints about what systemd is technically, but it's a really tired topic of conversation 2024-09-25 15:49:34 systemd is not halal, so it cannot be used on musl based distros ;) j/k 2024-09-25 15:54:14 invoked: Of course, I'm not saying that systemd is bug-free 2024-09-25 15:54:54 it's not even about bugs. it's about design, scope 2024-09-25 15:55:08 anyway it's a dead horse at this point, nobody's minds will be changed 2024-09-25 16:03:18 scorpion2185[m]: toapk maintainer here, it does not currently parse nixos unfortunately. It should work on an arch linux pkgbuild decently well though 2024-09-25 16:04:00 scorpion2185[m]: it will still only get you some of the way there. the convert APKBUILD would need to be cleaned up, and things like package names aren't always right, but it should get you close-ish. 2024-09-25 16:12:47 i got confused with gentoo 2024-09-25 16:16:58 "yes" <- 😢 2024-09-25 16:30:09 seriously though, being sad about offering an option is ridiculous. You don't have to use systemd, you can keep it "small and simple" like you want it. Stop being bothered by choices of others 2024-09-25 16:32:28 if devs include systemd it's a bad sign that can't be overlooked for me 2024-09-25 16:33:53 pmos did, and that's fine for them. 2024-09-25 16:35:22 feel free to fork alpine once systemd hits the repositories 2024-09-25 16:38:42 instead can alpine not have systemd? 2024-09-25 16:39:58 Wait, is alpine going to move to systemd? 2024-09-25 16:40:05 Scary. 2024-09-25 16:40:12 no. 2024-09-25 16:40:29 yes they will add it 2024-09-25 16:40:37 @ikke | There are no plans to change init systems just yet in Alpine. 2024-09-25 16:40:45 literally you were just told this 2024-09-25 16:41:23 Ermine which is Alpine dev? said that systemd will be added, in the future 2024-09-25 16:41:26 pmos will include systemd for them, but for Alpine, we do not have any plans to change anything 2024-09-25 16:42:22 why Emine said yes? 2024-09-25 16:42:34 it was a hypothetical 2024-09-25 16:42:46 if you don't like what a project is doing, you are free to fork it 2024-09-25 16:43:22 before i install alpine i'd like to if system will be added one day. 2024-09-25 16:43:22 ikke could you tell me? 2024-09-25 16:43:31 *to know if systemd 2024-09-25 16:43:44 ikke knows better than me anyway 2024-09-25 16:43:51 scorpion2185[m]: I cannot predict the future 2024-09-25 16:43:56 even if Alpine includes systemd, you can just simply ignore it and not install it 2024-09-25 16:44:59 I guess there is no reasoning with you so I'll stop trying 2024-09-25 16:45:23 perhaps the important thing to note is that alpine has *resisted* changing init to minimize disruption 2024-09-25 16:46:02 for example some distro won't ever have systemd because of policy etc. 2024-09-25 16:46:02 so you can tell that 2024-09-25 16:46:44 there's no policy prohibiting systemd in Alpine 2024-09-25 16:47:36 and no concept of small simple secure that will prohibiting it either? 2024-09-25 16:48:15 forget systemd. s6 could have been merged a long time ago, but it hasn't been because alpine feels it isn't ready enough to avoid being disruptive 2024-09-25 16:48:28 What's s6? 2024-09-25 16:48:37 stop looking for product promises and look at what's been happening 2024-09-25 16:48:39 init system from skarnet 2024-09-25 16:48:47 or rather, service manager 2024-09-25 16:48:51 prescense of systemd somewhere in the repos doesn't break those virtues, as said above 2024-09-25 16:50:22 I'm new to Alpine. Between stable and rolling models, where does it sit closer? 2024-09-25 16:50:33 Would you compare it to Fedora in that regard? 2024-09-25 16:50:46 stable with 6 months release right? 2024-09-25 16:51:07 cousin_luigi: you can use alpine edge repositories for rolling model, and v3.x for stable 2024-09-25 16:51:25 main repository is supported for 2 years 2024-09-25 16:51:33 I see. 2024-09-25 16:51:36 community for 6 months 2024-09-25 16:51:37 edge is also testing so it's not like having rolling 2024-09-25 16:51:46 edge can be considered rolling release 2024-09-25 16:51:50 yeah 2024-09-25 16:51:56 So stable is perhaps closer to Debian? 2024-09-25 16:52:05 Or maybe openwrt? 2024-09-25 16:52:08 We don't have a separate staging area before we push to edge 2024-09-25 16:52:22 debian is not a good comparison 2024-09-25 16:52:38 sure what i mean it's that rolling releases have a testing to offfer a proper rolling release 2024-09-25 16:52:48 Alpine's support terms don't come even close to Debian. Like ikke said main repository for every release is supported for 2 years, where Debian does iirc 10? 2024-09-25 16:53:30 debian gets into a lot of backporting/etc to do these long supported releases. very different 2024-09-25 16:54:51 ikke if you had to make a forecast of systemd going into alpine what woul you say? 2024-09-25 16:55:29 unlikely it will happen as long there is no official upstream support for musl 2024-09-25 16:55:50 right, pmos had to switch to glibc as well 2024-09-25 16:55:59 somebody is using donation to do that 😢 2024-09-25 16:56:17 no, pmos uses patches to systemd afaik 2024-09-25 16:56:26 really? interesting 2024-09-25 16:56:29 yes 2024-09-25 16:56:42 They are even working with systemd to upstream musl support 2024-09-25 16:57:08 that's cool, i suppose 2024-09-25 16:57:21 Does the name musl originate from Müsli? 2024-09-25 16:57:32 unlikely 2024-09-25 16:57:33 ikke could you tell me yor prevision of systemd and alpine? 2024-09-25 16:57:34 cousin_luigi: You have to ask Rich Felker 2024-09-25 16:57:43 maybe ask in #musl on libera 2024-09-25 16:58:27 https://www.musl-libc.org/faq.html 2024-09-25 16:58:56 Fascinating:) 2024-09-25 16:59:06 https://postmarketos.org/blog/2024/08/16/paying-for-development/ 2024-09-25 17:02:51 scorpion2185[m]: the discussion about inits and service managers has been happening for a long time, in gitlab and elsewhere. you won't get anyone to predict what will happen, they don't need that headache 2024-09-25 17:03:27 because they instant they even guess, people will be on their ass about it 2024-09-25 17:16:47 i see, ikke are you ok with systemd in alpine? hypothetically so just that i know 2024-09-25 17:25:52 My opinion is not that important 2024-09-25 17:26:21 i'd like to hear it 2024-09-25 17:56:41 is alpine meant for servers only or also for normal daily usage? 2024-09-25 17:57:30 it's general purpose, so for everything really. I'm typing this from my laptop running Alpine but I also use it on my server and in Docker containers 2024-09-25 17:58:16 ah ok, does it have cinnamon DE? 2024-09-25 17:58:54 yes, https://pkgs.alpinelinux.org/packages?name=cinnamon*&branch=edge&repo=&arch=x86_64&maintainer= 2024-09-25 17:59:45 im not on x86 2024-09-25 18:00:43 It's just an example 2024-09-25 18:00:57 V 2024-09-25 18:00:59 https://pkgs.alpinelinux.org/packages?name=cinnamon*&branch=edge&repo=&arch=&maintainer= 2024-09-25 18:01:50 So, how do I update a community package? It builds locally. 2024-09-25 18:02:37 You'd make a merge request in https://gitlab.alpinelinux.org/alpine/aports 2024-09-25 18:05:10 oh, thanks. 2024-09-25 19:29:42 I changed my local timezone by changing the symlnk /etc/localtime. 'date' reports the time in the new timezone, but some programs use the previous timezone when started. 2024-09-25 19:30:00 One offending program is waybar, which uses the old timezone even if restarted. 2024-09-25 19:30:14 How is it even detecting the old timezone if the file in /etc was replaced? 2024-09-25 19:30:49 WhyNotHugo: environment variable? 2024-09-25 19:31:37 No relevant variables. 2024-09-25 19:31:49 Oh, actually, waybar is using UTC instad of the old (or new) timezones 2024-09-25 19:39:29 gcc had a bug that had the effect to show the wrong date in waybar 2024-09-25 19:39:39 i guess its not fixed in alpine yet 2024-09-25 19:39:44 i can open a mr i suppose 2024-09-25 19:40:38 Yeah, that seems to be the case. 2024-09-25 19:40:56 I wanted to rebuild and see if that fixes it, but `apk update` hands forever for some reason. 2024-09-25 19:41:18 hangs the hands on the shelf hanger 2024-09-25 19:41:44 ACTION points to the remote hands fare 2024-09-25 19:42:10 that'll cost ya something 2024-09-25 19:42:28 Odd, apk times out downloading http://dl-cdn.alpinelinux.org/alpine/edge/main/x86_64/APKINDEX.tar.gz, but curl can download the file just fine. 2024-09-25 19:42:36 There is a current issue with Waybar about the time display that many people have fixed by downgrading the tzdata package 2024-09-25 19:42:36 https://github.com/Alexays/Waybar/issues/3024#issuecomment-2336706815 2024-09-25 19:43:11 jahway603[m]: > <@jahway603:meowchat.xyz> There is a current issue with Waybar about the time display that many people have fixed by downgrading the tzdata package 2024-09-25 19:43:11 > https://github.com/Alexays/Waybar/issues/3024#issuecomment-2336706815 2024-09-25 19:43:11 yeah but its actually a gcc bug https://gitlab.archlinux.org/archlinux/packaging/packages/gcc/-/issues/13 2024-09-25 19:43:24 buahahaha the weirdest fix idea, break your timezone 2024-09-25 19:44:20 $ TZ=0 printenv TZ 2024-09-25 19:44:46 TIL printenv 2024-09-25 19:45:15 WhyNotHugo: apparently waybar supports specifying time zone in its config file 2024-09-25 19:45:22 ^ net TZ, net problems 2024-09-25 19:45:23 I am trying to run mkinitfs in chroot environment where /lib/modules/kernel.lts does exist, but it says it doesn't. Am I able to run mkinitfs in chroot? 2024-09-25 19:45:29 Ermine: yeah but that also didnt work 2024-09-25 19:45:39 (net as in "no" / "none" in Stalinist languages) 2024-09-25 19:45:51 fossdd @fossdd:matrix.org did not know it was a GCC bug - just relaying what I used as workaround :) 2024-09-25 19:45:59 env / printenv is magic ;-) 2024-09-25 19:46:18 never trust your environment blindly, and always read up man environ 2024-09-25 19:46:20 ikke: ^^ requesting banhammer 2024-09-25 19:46:42 ..? 2024-09-25 19:46:53 "Stalinist languages" 2024-09-25 19:47:35 dude, grow up ;-) the idea is, (re)setting the timezone removes any doutbs about its influence on the program operation 2024-09-25 19:48:26 Alpine defaults to utc which is nice 2024-09-25 19:48:35 yep, like most UNIX systems 2024-09-25 19:49:05 internal time representation is always UTC (and in UNIX timestamp format, with optional microtime extensions) 2024-09-25 19:49:13 plutunion: Please respect the community Code of Conduct: https://alpinelinux.org/community/code-of-conduct.html 2024-09-25 19:49:43 another smart idea, when you subtract dates, to convert them to UNIX timestamp first (and back after the operation) 2024-09-25 19:50:28 that's called "operator" calculus (or the basis of symbolic computation), anyway.. back to your "mirror" issues 2024-09-25 19:50:45 anyone know about mkinitfs in chroot? I will probably just try to run outside of it otherwise 2024-09-25 19:51:31 andar1an[m]: it should work in a chroot 2024-09-25 19:51:51 make sure things like /proc, /sys and /dev are mounted inside the chroot 2024-09-25 19:52:29 andar1an did yo ujust use chroot? 2024-09-25 19:52:39 i have alpine chrooting script 2024-09-25 19:53:16 I think I needed to pass module directory 2024-09-25 19:53:17 also, there's the -b option and a few options have default paths set as seen in the man page. 2024-09-25 19:53:20 wasn't clear from the help 2024-09-25 19:53:38 if you chroot you need to mount a bunch of things 2024-09-25 19:53:57 or, rather.. copy them to the chrooted relative basedir 2024-09-25 19:53:58 apk add mkinitfs-doc was useful for me 2024-09-25 19:53:59 I know 2024-09-25 19:54:07 it is all mounted, it was just not adding a file 2024-09-25 19:54:45 it was just a missing argument, didn't realise needed to pass a path 2024-09-25 19:56:20 btw, UNIX did not have the "stat" filesystems like procfs and statfs / devfs, so it's non standard and removed on recent systems too, so it's just to satisfy your tradition (look and feel for the human operator, applications should not rely on it), and dev is standard but not specified so your application needs the device files it will use 2024-09-25 19:56:57 for that you read hier(7) 2024-09-25 19:57:02 So the waybar issue seems to be a gcc issue. But the gcc fix is not released yet. 2024-09-25 19:57:39 yeah its in a snapshot release 2024-09-25 19:57:56 but i have no idea how i should upgrade alpine's gcc to it 2024-09-25 20:22:20 you know that thing called a standard library.. 2024-09-25 20:22:38 start from there ;-) 2024-09-26 04:40:36 i've tried to set environment vars in /etc/profile and /etc/environment, but they aren't showing up in the output of env 2024-09-26 04:41:13 i'm running alpine 3.20 in an Incus container, with alpine 3.20 as the host 2024-09-26 04:42:07 mechanical_jellyfish: make sure you create a login shell 2024-09-26 04:42:47 Just executing /bin/sh does not create one, for example 2024-09-26 04:43:21 do i need to pass a flag to sh to make it a login shell? 2024-09-26 04:43:53 sh -l 2024-09-26 04:45:50 that worked, thanks 2024-09-26 04:46:13 is it bad practice to use a login shell non-interactively? 2024-09-26 04:53:43 mechanical_jellyfish: I'm inclined to say yes 2024-09-26 04:54:54 is there a way to set environment vars for non-login shells? 2024-09-26 04:55:34 this might be getting off topic for the alpine channel 2024-09-26 05:00:28 Maybe you can explain what you are trying to achieve 2024-09-26 05:01:11 Generally you run scripts from some environemnt and the environment variables are inheritted 2024-09-26 05:07:21 i'm trying to link the container wayland socket to the host wayland socket, so i can run gui apps 2024-09-26 05:08:30 at the moment i have to set WAYLAND_DISPLAY and XDG_RUNTIME_DIR every time for it to work 2024-09-26 05:08:40 i was hoping to do it automatically 2024-09-26 05:09:04 i'm interacting with the container with "incus exec container sh" 2024-09-26 05:15:52 You can change that to `incus exec container -- sh -l 2024-09-26 05:21:34 i just read about incus exec --env, which would let me set environment variables without starting a login shell 2024-09-26 05:23:14 thanks for telling me about non-login shells not reading /etc/profile and /etc/environment 2024-09-26 07:12:54 ikke: Where does one put the changelog for the package? In the commit or PR message? 2024-09-26 07:14:18 I do both 2024-09-26 07:21:12 What is incus mechanical_jellyfish? 2024-09-26 07:22:16 Fork of lxd 2024-09-26 07:22:32 o_O 2024-09-26 07:22:39 After the Ubuntu forced takeover 2024-09-26 07:28:38 I must have missed that episode. What happened with Ubuntu? 2024-09-26 08:16:31 mechanical_jellyfish: you might be interested in distrobox, it basically does what you want (although uses podman or docker to run the container instead) 2024-09-26 10:43:42 PureTryOut: the website says the point of distrobox is to integrate containers as tightly as possible with the host 2024-09-26 10:45:19 that's the opposite of what i want, these containers are for when a program's website says it needs to be installed with curl | sudo bash and i don't want it messing things up 2024-09-26 10:50:11 but i've always tested distros by writing them to a live USB, i'll use distrobox next time now i know about it 2024-09-26 10:50:15 thanks 2024-09-26 10:58:53 cousin_luigi: Canonical took over the lxd project from the community and removed community maintainers, which prompted a fork by some of the original people behind lxd who are no longer at canonical 2024-09-26 10:59:29 That fork is incus, and it is back in the Linux containers community 2024-09-26 11:00:51 Later canonical also changed the license to be incompatible with incus going forward and enforced a CLA, so they are now entirely separate with no code exchanging between them going forward 2024-09-26 11:02:09 There is a more correct and complete explanation somewhere on the Linux containers blog or on Stefane Graber's blog 2024-09-26 11:20:26 bananicorn: That's low even by Canonical's standards. 2024-09-26 12:57:07 It is indeed :) 2024-09-26 14:11:28 Is there a bug with Grub where it doesn't add specific modules for preload or default linux? 2024-09-26 14:12:18 have updated both /etc/default/grub and files in /etc/grub.d 2024-09-26 14:12:45 doesn't respect additions for those fields for gcry_sha512 of pbkdf2 modules 2024-09-26 14:29:47 so what are you thoughts about that CVSS 9.9 vuln for Linux to become public on 6th october? 2024-09-26 14:30:15 Apachez: rumors are it's a storm in a glass water 2024-09-26 14:32:11 my professional opinion is that it might not be fun but should be easy to mitigate 2024-09-26 14:35:30 i also strongly suspect more will be public before october 6th 2024-09-26 14:36:44 does alpine have risc-v? 2024-09-26 14:36:51 riscv64 2024-09-26 14:37:05 i mean that cause https://distrowatch.com/table.php?distribution=alpine 2024-09-26 14:37:18 it does not list it 2024-09-26 14:37:36 It misses armv7 as well 2024-09-26 14:38:21 not sure waht's s390x. anyway user searching riiscv64 may miss alpine 2024-09-26 14:39:01 Feel free to suggest to them it should be added 2024-09-26 14:39:03 but it supports armv7? there are bootloader settings for rpi in setup. Is that for older arm? 2024-09-26 14:41:03 Darn, Cortex chips are V7 I guess 2024-09-26 14:41:10 good to know 2024-09-26 14:43:17 Is this wrong then? https://wiki.alpinelinux.org/wiki/Requirements 2024-09-26 14:44:02 not all cortex chips 2024-09-26 14:46:02 Seems like A-72 is supported based on this 2024-09-26 14:46:46 yes 2024-09-26 14:47:09 A-72 would run armhf, armv7 and aarch64 in fact 2024-09-26 14:47:21 I need to use cm4's for research atm, I could probably try to get away with some riscv cm4 form factor, but I want to wait until there are more VLA 1.0 options 2024-09-26 14:47:32 perfecto, thank you 2024-09-26 14:54:21 scorpiom2185 s390x is an IBM mainframe https://en.wikipedia.org/wiki/IBM_Z 2024-09-26 15:24:31 well, this stuff also includes p/390 which is personal computer sized 2024-09-26 15:25:12 not what people would normally think of as a mainframe but it would run all that stuff the same. 2024-09-26 15:28:27 invoked Never heard of p/390, but I'm a mainframer, so z-Series is nowadays the only official platform I'd say. 2024-09-26 15:28:56 Of course you *could* use Hercules for example. 2024-09-26 15:29:18 https://en.wikipedia.org/wiki/Hercules_(emulator) 2024-09-26 15:30:12 p390s are what developers used to use for the s390s. i was never an ibm guy. but basically it was a card in a pc running os/2, but it would faithfully run all the s390 stuff exact 2024-09-26 15:33:15 i was only into ibm enough to pass the ccie tests, which covered sna 2024-09-26 15:34:21 i don't know anything about z-series, so i probably should mute myself 2024-09-26 16:31:31 Hey everyone, quick question for you. I am trying to install firacode nerdfont which i did with the command: doas apk add font-fira-code-nerd 2024-09-26 16:32:35 but, this has given me a bunch of ttf files, which doesnt seem to be what the setfont command is looking for, because i get this error: setfont: input file: bad length or unsupported font type 2024-09-26 16:34:43 the exact command I ran was: setfont /usr/share/fonts/fira-code-nerd/FiraCodeNerdFontMono-Regular.ttf 2024-09-26 17:15:46 sigh 2024-09-26 17:16:35 if only there was a command to look up what font formats `setfont` accepts.. 2024-09-26 17:16:49 there is! i just did! 2024-09-26 17:16:51 ( and if only people had enough patience to actually wait for someone to answer ) 2024-09-26 17:17:03 bet you five bucks hackii wasn't even trying to change the console font (but i could be wrong) 2024-09-26 17:18:37 i would not be surprised if the answer was "chatgpt told me to use `setfont` to set the font in my desktop environment" 2024-09-26 17:19:09 ha 2024-09-26 18:16:25 looks like evilsocket is going to drop the security vuln details at 2000 UTC 2024-09-26 18:16:56 weeeee! 2024-09-26 18:17:34 lol 2024-09-26 18:17:48 publicly or to his twitter followers? 2024-09-26 18:18:18 ah, he unprivated 2024-09-26 18:18:30 the tweet says: Full disclosure happening at 20:00 UTC today, in a bit more than 2 hours. 2024-09-26 18:18:35 yeah, i see now 2024-09-26 18:18:43 ( https://twitter.com/evilsocket/status/1839361276813902240 ) 2024-09-26 18:19:14 it'll be a little anticlimactic 2024-09-26 18:20:05 i mean half the details are already public, apparently 2024-09-26 18:26:47 he dropped that tweet seconds after i finished a CVE report to presumably the same vendor 2024-09-26 18:35:27 looking through redhat's bugzilla ... wonder if it's related to a driver? https://bugzilla.redhat.com/show_bug.cgi?id=2297478 2024-09-26 18:35:45 my suspicion is it's not 2024-09-26 18:35:58 I read something about cups? 2024-09-26 18:36:06 i have no doubt it is cups 2024-09-26 18:36:15 If it's cups, then it's already public 2024-09-26 18:36:27 ahh cool. I'm just hoping it not sshd mainly 2024-09-26 18:36:36 ikke, public? 2024-09-26 18:37:22 https://linuxrocks.online/@rose/113204947849442693 2024-09-26 18:37:28 Not sure if it's the same thing, though 2024-09-26 18:37:54 sorry, wrong link 2024-09-26 18:37:56 https://infosec.exchange/@atax1a/113205075007680030 2024-09-26 18:38:20 i did just read #36 2024-09-26 18:38:28 clearly this person has some mental challenges 2024-09-26 18:38:43 yeah, love that guy's bio of evilsocket :-) 2024-09-26 18:38:50 but, based on what was known monday, i looked into CUPS myself and ended up filing a private bug with them (but mine is not yet RCE) 2024-09-26 18:38:57 it's clear that CUPS is not up to par 2024-09-26 18:39:14 certainly not for a project that installs by default on many desktop distros, as root, with sockets listening to the world 2024-09-26 18:39:23 so i'd watch my ports and services in any case 2024-09-26 18:39:27 lpd and other printer daemons have a long tradition of being vulnerable :-) 2024-09-26 18:39:29 i'm not the only one who dug into CUPS this week, i am sure 2024-09-26 18:39:53 as for 'requires editing cupsd.conf', that detail is new to me 2024-09-26 18:39:59 but also the actual exploit is not public from what i know 2024-09-26 18:40:35 the "requires editing cupsd.conf" might be an unrelated vulnerability: https://github.com/OpenPrinting/cups/security/advisories/GHSA-vvwp-mv6j-hw6f 2024-09-26 18:40:57 right 2024-09-26 18:41:00 ok 2024-09-26 18:41:02 that i saw 2024-09-26 18:46:40 but tbh "an excerpt from a larger chain of vulnerabilities" does make it sound like it's related 2024-09-26 20:13:14 fwiw, the cups vuln writeup dropped: https://gist.github.com/stong/c8847ef27910ae344a7b5408d9840ee1 2024-09-26 20:13:31 and by "dropped" i mean someone leaked it 2024-09-26 20:13:54 good guesses all :-) 2024-09-26 20:13:56 ...but then, the author himself broke the embargo and posted a blogpost 2024-09-26 20:15:01 AFAICT Alpine is not vulnerable by default, because we don't even have a service file for cups-browsed 2024-09-26 20:15:54 however we should probably pull the fix from https://github.com/OpenPrinting/cups/commit/96b3bdf010e78880f5764e5032720379aa1116df 2024-09-26 20:16:23 i think i'd rather fire up an ancient lpd or lprNG and block remote access on my system before I'd run cups 2024-09-26 20:17:32 ptrc, what you should do is disable/drop CUPS Browsing support. also pull the commit just in case 2024-09-26 20:17:38 but upstream is going to drop the entire protocol 2024-09-26 20:17:50 that they deprecated over 10 years ago but have never been very clear about i guess 2024-09-26 20:22:27 i mean, if someone's running cups-browsed on Alpine, they're very probably running it on purpose ( or because of some misguided advice ) 2024-09-26 20:22:41 not sure how pmOS handles it 2024-09-26 20:29:51 me neither. the original claim was "it runs by default on ubuntu", which is true for desktops (and also on debian) 2024-09-26 20:34:39 i'd be grateful if someone could sanity check !72587 2024-09-26 20:58:51 ugh, I feel like I am close to grub unlocking luks, but I can't figure out why it won't recognise password. Config seems to be good now, cryptomount it pointing correctly, luks root looks correct except it is prefixed with cryptouuid/ which I haven't seen before. Debating removing that and seeing what happens 2024-09-26 21:06:54 ACTION sent a code block: https://matrix.org/oftc/media/v1/media/download/AR6SOiMfLTPOCxTNMbosI2GqE4sA26qaTPbdTHrs2PG-BpMt5ypjvLB-xKfpXaWCP_iayTwAVP48ESir9DovfQlCeSMd_1TwAG1hdHJpeC5vcmcvRG12UHN0SklSZml1RExoQ0drV0xKVm12 2024-09-26 21:07:44 may also try CMDLINE_LINUX instead of DEFAULT 2024-09-26 21:12:44 maybe because I have boot on btrfs, or maybe need a boot flag. Time will tell 2024-09-26 21:30:15 PBKDF is PBKDF2 not argon 2024-09-26 21:56:18 Is there any documentation that lists common kernel modules with descriptions? modinfo doesn't seem to give so much back 2024-09-26 21:56:49 web search hasn't turned up any good references 2024-09-26 22:00:58 andar1an[m]: modules is probably the keyword you want to avoid. They are part of the kernel, wether they are in modules or inside the kernel is just a compilation option 2024-09-26 22:01:18 andar1an[m]: a lot of stuff is there https://www.kernel.org/doc/html/latest/admin-guide/index.html 2024-09-26 22:04:29 Thanks, I need to preload some modules so I can unlock luks, after that I make adjustments to modules for daily driving so that I can use peripherals and such 2024-09-26 22:05:39 trying to see if I can set GRUB_LINUX_blah blah to see if I can avoid modprobing and creating entries in /etc/modules-load.d/blahblah 2024-09-26 22:07:00 I think I have narrowed into where the issue may be in grub.cfg. There is a linux line where cryptdevice gets defined after root is set. May see if adjusting the cryptroot or leaving that out will create some changes 2024-09-26 22:07:46 Thank you for the link. I am not a fan of linux docs haha. That can be a start at least 2024-09-26 22:08:27 wish lsmod also had a description column lol 2024-09-26 22:10:17 tried to check kernel repo, didn't see an easy way to see what were modules yet either. Seems to be divided throughout, hope there is an easy keyword to use when searching 2024-09-26 22:13:35 there seems to be an @ for my linux entry, thinking that may be the cause. So instead of /boot/vmlinuz it is /@boot. Maybe not it but I will know when I figure out syntax of that line haha 2024-09-27 11:42:20 ACTION hopes anyone who needs to know, already knows, but FYI. https://www.evilsocket.net/2024/09/26/Attacking-UNIX-systems-via-CUPS-Part-I/ 2024-09-27 11:42:51 Short version. If you can register an evil printer, and the user attempts to use it, then root compromised. 2024-09-27 11:46:11 Sofia: Alpine users should not be affected unless they went out of their way to enable/run the vulnerable component 2024-09-27 11:46:42 There is not even a service for it packaged 2024-09-27 11:46:58 ikke: Lovely. 2024-09-27 12:33:33 What's the general process to create an alternative installation medium? 2024-09-27 12:34:08 The official x86_64 ISOs don't work on my 2010 MacBook Air, it causes the EFI firmware to just hang, but Debian ISOs are able to boot 2024-09-27 12:34:21 I suppose I could try to replace syslinux with grub and see if it decides to boot? 2024-09-27 12:39:44 grub is already used when it's efi 2024-09-27 12:40:22 noted, though i still want to create a custom image to see what might help... 2024-09-27 12:40:54 or I could install debian, somehow boot in ramfs, and install alpine that way, but that feels flaky 2024-09-27 12:42:25 use these scripts: https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/scripts?ref_type=heads 2024-09-27 12:43:07 thanks, ill take a look 2024-09-27 12:53:29 i don't have an x86_64 alpine system with me so it'd be a bit difficult 2024-09-27 12:54:29 actually- i suppose i'll just try to install alpine myself from the arch liveiso 2024-09-27 13:08:56 There is also: https://wiki.alpinelinux.org/wiki/How_to_make_a_custom_ISO_image and https://wiki.alpinelinux.org/wiki/How_to_make_a_custom_ISO_image_with_mkimage 2024-09-27 13:09:35 I haven't messed with these too much yet, because I dk how relevant they are to pxe, but there is a fair bit on mkimage 2024-09-27 13:10:07 You can also check out apk overlays and saving state for diskless installs 2024-09-27 13:12:00 https://wiki.alpinelinux.org/wiki/Alpine_local_backup 2024-09-27 13:13:16 I think when looking at secure boot wiki/docs is may be possible to discern tools to create media for pxe, but haven't gotten there yet to know 2024-09-27 13:20:17 maybe mkimage packages up everything necessary in an iso, and can just extract? dk 2024-09-27 13:20:29 I didn't have luck making an Alpine iso 2024-09-27 13:20:48 I wanted to literally just use setup-alpine from the arch iso, but I couldn't install gcc due to limited available space, so I'm using my chroot-creation scripts... 2024-09-27 13:21:54 --- and I didn't allocate enough space for /boot so it's time to repartition, how fun 2024-09-27 13:26:17 You don't need gcc if you want to run setup-alpine 2024-09-27 13:31:55 I don't know what depends on uniso 2024-09-27 13:37:30 ergh 2024-09-27 13:37:34 I forgot to write an fstab 2024-09-27 13:47:10 oh I could use arch's genfstab 2024-09-27 13:47:15 except that i get a broken initramfs 2024-09-27 13:50:39 setup-disk will create fstab for you and it does a good job 2024-09-27 13:50:55 you can pass some env variables to modify what it does as well 2024-09-27 13:51:25 e.g. BOOTLOADER=xyz or USE_EFI=boolean are some examples that you can prepend 2024-09-27 13:51:53 the script is in alpine-conf directory in git to chechout 2024-09-27 13:52:47 https://wiki.alpinelinux.org/wiki/Setting_up_disks_manually 2024-09-27 13:53:26 I currently have a lot of mounts and mount options, and it captured them well 2024-09-27 13:54:07 I've already installed my system by using a static version of apk and I don't really feel like destroying it with setup-disk 2024-09-27 13:56:21 on a quick tangent, is there a forum for alpine like for arch where I may be able to post longer questions for discussion? 2024-09-27 13:56:53 the mailing list? 2024-09-27 13:57:03 ew lol 2024-09-27 13:57:42 Could you include static apk as cache in install media and do setup-disk? 2024-09-27 13:57:50 you could also pin versions of packages 2024-09-27 14:01:18 apparently my attempts to bootstrap myself have failed 2024-09-27 14:01:51 ill try to use setup-alpine again :/ 2024-09-27 14:02:16 I think by default apks are pulled from /media/sdx/apk on installation media 2024-09-27 14:02:32 if you are creating your own iso you could likely stick em in there 2024-09-27 14:03:12 *apks, not apk 2024-09-27 14:04:03 setup-alpine does what setup-disk does, with a few other scripts automatically run before to gather inputs to template files to be moved into place 2024-09-27 14:04:22 or install some services 2024-09-27 14:04:53 https://wiki.alpinelinux.org/wiki/Alpine_setup_scripts 2024-09-27 14:05:12 im sure i can tell alpine-conf to use a web mirror 2024-09-27 14:05:33 yes, and https or ftp too 2024-09-27 14:05:53 setup-apkrepos 2024-09-27 14:06:09 or just tee to /etc/apk/repositories 2024-09-27 14:06:45 i want to just properly install alpine-conf on my arch liveiso but i don't know how to allow installing large packages like gcc in the live environment 2024-09-27 14:06:51 meh, ill remove the uniso stuff 2024-09-27 14:09:52 does alpine use gcc or llvm for linking? 2024-09-27 14:10:06 I know it uses musl 2024-09-27 14:10:40 personally also following cranelift in rust and wasm space. 2024-09-27 14:14:24 andar1an[m]: For the libpng apk, (1.6.44-r0 for edge x86_54) the GNU linker ld was used / gcc driver. https://build.alpinelinux.org/buildlogs/build-edge-x86_64/main/libpng/libpng-1.6.44-r0.log 2024-09-27 14:14:46 Cool, thanks. I dk anything about that yet 2024-09-27 14:15:05 I am trying to limit exposure to as much C code as I can haha 2024-09-27 14:21:14 attempting to install apk on the arch iso broke the system 2024-09-27 14:21:28 ...... uhhh I'll temporarily use arch and see if i can get things working 2024-09-27 14:21:34 later 2024-09-27 14:22:17 👋 2024-09-27 14:37:51 darn, looks like btrfs swap subvolume wont work 2024-09-27 15:03:43 maybe I can, there seems to be conflicting data 2024-09-27 15:04:48 anyone know state of specific mount options for subvolumes? Seems to be broken down into fs vs generic, but another note mentions "This means that (for example) you can’t set per-subvolume nodatacow, nodatasum, or compress using mount options. " 2024-09-27 15:05:40 alpine setup-disk seems to validate that the first subvolume mount options are propagated to the following mounts 2024-09-27 15:06:10 I believe a remount may change things, but dk if nodatacow is generic or fs specific (I think the latter - but uncertain) 2024-09-27 15:06:45 you could use chattr +C yourswapfile instead of the mount option 2024-09-27 15:07:00 I was wondering what that was, thank you koolman! 2024-09-27 15:07:24 set a flag on the file so that it does not use copy-on-write 2024-09-27 15:07:25 The issue there was that I think I can't snapshot a volume that contains a swapfile 2024-09-27 15:07:43 yes. you want a subvolume for swap, basically with only your swap file in it 2024-09-27 15:07:45 but maybe I can do chattrc to the swapfile in the subvol haha 2024-09-27 15:07:50 you can 2024-09-27 15:07:57 Ya, the problem is the mount options 2024-09-27 15:08:33 I can manually update the fstab, but btrfs docs seems to indicate varying mount options for subvols is an issue right now due to linux VFS 2024-09-27 15:09:01 you can probably cheat around that 2024-09-27 15:09:17 mount with nodatasum, create swap file, remount with regular options 2024-09-27 15:09:23 What I am thinking is just overwrite fstab that setup-disk gens 2024-09-27 15:09:23 nodatasum only affects file creation 2024-09-27 15:09:37 (once a file is created with nodatasum it stays that way) 2024-09-27 15:10:09 That is awesome to know 2024-09-27 15:10:53 and I think nodatasum and nodatacow are the two most relevant for swap file so ... with those two workaround it should work for you :) 2024-09-27 15:11:29 I had my subvols segmented based on use:... (full message at ) 2024-09-27 15:11:46 Ya, and nodatacow infers nodatasum 2024-09-27 15:12:15 you can also use chattr +C on the directory (lets say /swap) before creating the file, too, then the file inherits it 2024-09-27 15:16:04 Thanks. One of the doc pages states:... (full message at ) 2024-09-27 15:27:49 also seems pass is recommended to be 0 for btrfs in fstab, maybe that will fix something 2024-09-27 15:30:37 well, it's not like a fsck would help most of the time :) 2024-09-27 15:35:28 haha unless you can get hacky and if it fails use error=remount-rw in fstab hahaha 2024-09-27 15:35:55 trying to look up if there is an option to remount without error, cuz that may be a hacky way around this lol 2024-09-27 15:36:16 could always update the fstab and just run mount -a on boot I guess, also hacky 2024-09-27 15:37:32 I do not know if btrfs implements errors=continue 2024-09-27 15:37:50 but ... don't do that anyway 2024-09-27 15:39:08 I wasn't going to haha 2024-09-27 17:59:51 Just checking if there's a planned cups packages update for Alpine Linux 2024-09-27 18:00:26 I hope all of you are doing well. I noticed there's no forum online so I thought of passing by on IRC to ask. 2024-09-27 18:21:55 Is there a way to volunteer on the package update of CUPS? 2024-09-27 18:22:32 ide: Yes, you can open an MR in https://gitlab.alpinelinux.org/alpine/aports 2024-09-27 18:25:25 Not yet an upgrade to the latest version, but !71251 exists 2024-09-27 18:26:01 though, that's only about cups-filters 2024-09-27 18:53:53 isn't it time to have a forum? 2024-09-27 18:54:15 ^ <3 2024-09-27 18:57:44 We used to have a forum, it became a wasteland 2024-09-27 18:57:54 Come to irc 2024-09-27 18:59:23 i think that now there are more people and it could have a pmos section 2024-09-27 19:00:11 pmos has their own infra 2024-09-27 19:02:12 do they have a forum? 2024-09-27 19:03:26 No idea 2024-09-27 19:25:26 no but i dont think it is nesessary 2024-09-27 19:35:09 I think you see a lot of benefit from having a medium where questions can be answered publicly and searched. But I imagine there are alternatives to hosting own forum. 2024-09-27 19:37:02 The wiki is nice, but it doesn't have a means to facilitate conversation while working on something 2024-09-27 19:37:20 Also, no idea what that syntax is, wish it was markdown haha 2024-09-27 19:37:32 MediaWiki has it's own syntax 2024-09-27 19:38:00 At least I know what to look for now haha, ty 2024-09-27 19:41:28 Would it be nice if these could be linked to somehow for wiki? Maybe like an editor guideline page? Does one exists? 2024-09-27 19:41:28 https://www.mediawiki.org/wiki/Help:Formatting 2024-09-27 19:41:28 https://www.mediawiki.org/wiki/Cheatsheet 2024-09-27 19:42:47 I don't think I have ever scrolled to the bottom to see "powered by mediawiki" haha, that's probably my bad 2024-09-27 19:43:04 The editor also has a cheat sheet embedded 2024-09-27 19:56:39 Mind blown 2024-09-27 19:56:55 Is that for new or editing too? 2024-09-27 19:57:45 also for editing 2024-09-27 19:57:51 I hope I can get done with this luks2 + btrfs soon so I can try making new post 2024-09-27 19:59:53 My goal was today, but I failed haha. Calling it and hopefully next week 2024-09-27 20:04:00 Never give up 2024-09-27 20:04:23 I don't haha, baby steps 2024-09-28 01:17:21 shift-f11 2024-09-28 02:04:47 is there any "stage" file like gentoo in alpine? 2024-09-28 08:47:08 no :) 2024-09-28 12:46:47 FACK - what a derp mistake lmao. Can't preload grub modules if they don't exist in grub efi. There is a --modules flag for grub-install. Finally got luks2 unlocking! 2024-09-28 12:47:59 For some stupid reason I thought that the environment variables in /etc/default/grub were used by installer 2024-09-28 12:48:19 kicking myself for not looking at those docs sooner. ugh 2024-09-28 14:24:15 morning everyone, aw I used gentoo during college, learned it with stage1 install, that was an interesting learning curve 2024-09-28 15:41:03 morning 2024-09-28 17:54:06 Hi! I'm trying to set my user to be able to run reboot and poweroff with doas with this lines but doesn't work, any ideas... (full message at ) 2024-09-28 17:54:36 * Hi! I'm trying to set my user to be able to run reboot and poweroff without pass with doas with this lines but doesn't work, any ideas... (full message at ) 2024-09-28 18:42:42 Does the user need /sbin added to PATH? 2024-09-28 18:44:25 visone[m]: do /sbin/poweroff and /sbin/reboot work for you? 2024-09-28 18:46:17 grossaffe[m]: Having those binaries in path is nice when trying to executing them with doas 2024-09-28 18:46:24 (to make autocomplete work) 2024-09-28 18:46:48 apk can be executed as non-root users to query info 2024-09-28 18:47:12 So I'd say it makes no sense to remove it from the PATH of users 2024-09-28 18:51:34 "visone: do /sbin/poweroff and /..." <- Yep, If I run ` doas poweroff` ask me por pass and exec poweroff. 2024-09-28 18:52:26 then alias poweroff to doas /sbin/poweroff and same for reboot 2024-09-28 18:53:54 Ermine: ok I'l try that, thanks!! 2024-09-28 18:54:03 s/l/ll/ 2024-09-28 19:09:20 Hey, I'm having some trouble with PipeWire on Alpine, probably something really simple to solve, but I'm quite new to all this, could somebody try to help me if i tell the full story? I had it working by following the wiki, but it just stopped working and I can't fix it. 2024-09-28 19:12:40 How long is the full story? If it's not pages, I'd just go ahead and tell it, then if someone recognizes the issue I'm sure they'll say something. 2024-09-28 19:12:45 tell the full story 2024-09-28 19:13:39 maybe write the full story down somewhere 2024-09-28 19:13:44 so you can refer to it later 2024-09-28 19:16:04 is there a way to see total size of installing something? eg if i want to compare how much will be installed if i install chromium v firefox? 2024-09-28 19:16:43 yes 2024-09-28 19:16:58 so basically, i had it working under wayland/sway, and then it just stopped, if i tried setting any sinks as default i would get no audio too. i tried using pavucontrol to have a visual understanding and nothing seemed to work, my output was listed, i could enable it, the firefox source, for example, was using my output and the little sound bar thing indicated that the audio was being sent, but i got nothing, 2024-09-28 19:18:21 i got it to work three times actually, once it just worked, but when i switched out and in again from the sink, it jus stopped again, and other two times it randomly started playing withou any intervention from me, but the sound was like 2x speed and high pitched 2024-09-28 19:18:40 shit, forgot to ask HOW ;/ 2024-09-28 19:23:16 frag, that's not on you. 2024-09-28 19:23:28 but plutunium's mic drop-quit was funny anyway 2024-09-28 19:23:53 frag, i don't have the answer but i would start by checking out what apk info can tell you i think 2024-09-28 19:24:08 classic mistake... 2024-09-28 19:24:32 It shows installed sized, but you'd have to manually sum up the sizes of the package and all dependencies 2024-09-28 19:24:34 ill just go with qutebrowser, as im used to.. 2024-09-28 19:25:41 sometimes i install and uninstall packages to compare how much space was used.. 2024-09-28 19:26:08 I was testing apk add -s, but that won't take the new packages into account 2024-09-28 19:50:21 vhespanha: does it work if you use pw -cat -p on some audio file? 2024-09-28 19:52:04 let me try 2024-09-28 19:56:55 addwd rtc to rpi4. edited config.txt and loaded modules. time do not survive a reboot. how tell alpinelinux to use hwclock ? 2024-09-28 23:54:14 anyone running emacs on alpine? 2024-09-29 00:04:48 agentcasey[m], i bet people are. maybe just ask the question you have 2024-09-29 00:06:54 I have been searching I even installed emacs but it doesn't work I know that alpine doesn't use gcc so that's part of the issue. The search engines aren't helpful but I'll check the emacs chat then. 2024-09-29 00:08:27 you haven't asked a question yet 2024-09-29 00:10:13 is anyone running emacs on alpine linux if so, give us a link to a resourse. 2024-09-29 00:11:14 I forgot the ? mark. 2024-09-29 00:11:43 yes, i'm runnin emacs on alpine, in fact i was the one that ported emacs back in the days to alpine... 2024-09-29 00:13:07 there is a bunch of emacs packages, -nox -x11 -qt -gtk2 you choose your prefered frontend 2024-09-29 00:14:20 bOR38552MJA: do you have a github page or something similar so that I may gain more insight? 2024-09-29 00:14:43 i have a github page, which is entirely unrelated to alpine or emacs. 2024-09-29 00:14:51 what are you looking for? 2024-09-29 00:15:41 I installed the emacs package with apk add emacs but there isn't an emacs binary in /usr/bin 2024-09-29 00:16:38 yes. read what i wrote 2024-09-29 00:16:42 agentcasey[m], you're quite close to asking a question now! 2024-09-29 00:16:56 i repeat: there is a bunch of emacs packages, -nox -x11 -qt -gtk2 you choose your prefered frontend 2024-09-29 00:17:35 I have read it and I will install -nox and see if that helps 2024-09-29 00:17:51 -nox will be terminal only version 2024-09-29 00:18:23 I know, 2024-09-29 00:19:01 oh, i see bunch of new frontends added since i ceased maintainership. nice. 2024-09-29 00:19:35 it's my first week working with alpine. 2024-09-29 00:19:51 welcome :) 2024-09-29 00:20:53 thanks, and your insightful answers helped, everything's working as planned. Thanks for your help. 2024-09-29 00:22:14 actually i answered your question before you asked it :) 2024-09-29 00:22:40 i see however there is no gtk2 frontend, only gtk3 and pgtk - whatever the later is... 2024-09-29 00:23:14 also you say alpine doesn't use gcc, that is wrong, it does use it afaik. 2024-09-29 00:23:54 also no qt frontend? meh. 2024-09-29 00:24:23 naja nox and x11 should be enough for everyone ;) 2024-09-29 00:24:29 I'm still reading the Alpine docs I'll get up to speed soon. 2024-09-29 00:24:47 awesome! rtfm is always very useful 2024-09-29 00:25:06 indeed. 2024-09-29 00:25:30 thanks again, all i needed was nox 2024-09-29 00:25:41 but i do understand why you are confused after apk add emacs, and not finding any emacs 2024-09-29 00:26:15 but choosing any default frontend will piss off all the other frontend-afficionados. 2024-09-29 00:26:49 i know i would be pissed of getting a gtk3 by default. 2024-09-29 00:27:01 pissed off even 2024-09-29 00:28:21 i did apk add emacs and the only packages in /usr/bin/ were ebrowse and a few others but no emacs binary. I didn't know, and still don't know how to show all the available package options yet so i didn't know the various flavors on this distro. 2024-09-29 00:29:32 i wouldn't know either, i would do (as i did) a `apk search -v emacs` and be DoSed by the tsunami of packages. 2024-09-29 00:29:58 so instead i just went to aports/community/emacs/APKBUILD and looked there for sub-pkgs 2024-09-29 00:30:35 thanks, that'll come in handy. 2024-09-29 01:15:03 bOR38552MJA: the pgtk frontend is "pure GTK" 2024-09-29 01:15:28 bOR38552MJA: the "normal" gtk3 frontend has x11 dependencies (and runs on xwayland when on wayland environments) 2024-09-29 01:15:46 bOR38552MJA: I haven't ever seen a qt frontend 2024-09-29 01:18:31 hah, there never was a qt frontend it seems. 2024-09-29 01:24:53 Is Alpine suitable for development on the Raspberry Pi4 (4GB RAM) with plasma desktop? In a video there was a comparison with Rasbian and Dietpi Alpine did very well in terms of performance 2024-09-29 07:32:30 Hey, is there any reason why I would have the -r0 version of musl 1.2.5 installed on edge and be unable to upgrade to -r3? Using the standard cdn repo mirrors 2024-09-29 07:33:39 I'm trying to install the musl-dev package but its trying to download the -r3 version which conflicts with musl -r0 version, and won't upgrade that one either 2024-09-29 07:38:10 NVM, rolled the dice and removed then reinstalled musl, and it seems to have fixed itself 2024-09-29 12:39:31 sup 2024-09-29 12:40:23 https://gitlab.alpinelinux.org/kevinalavik/alpine-utils something cool ig 2024-09-29 19:12:51 Hi 2024-09-29 19:13:05 Is it possible to list installed packages that are not a dependency to another package? 2024-09-29 19:13:12 cat /etc/apk/world 2024-09-29 19:14:29 Thanks ikke :) 2024-09-29 19:15:11 yw 2024-09-29 19:36:56 https://calayucu.com/light-mode 2024-09-29 19:37:02 anyone else gets a crash in Firefox on that? 2024-09-29 19:38:40 lol. 2024-09-29 19:38:53 also no crash in Firefox (but i'm not on Alpine) 2024-09-29 19:39:30 No crash on alpine 2024-09-29 19:46:54 hm, might be just my firefox then 2024-09-29 19:47:03 it crashes only when i try to move the cursor though 2024-09-29 19:47:36 I could load it, switch the toggle, move the cursor painstakingly to disable it again 2024-09-29 19:47:54 when i push the button ff crashes 2024-09-29 20:04:24 ikke, oh, my firefox gave me a notice that said something like "push Esc to regain pointer control" 2024-09-29 20:05:18 Yes, I was just playing along 2024-09-29 20:07:26 Works here too, but there's those badger and ublock things running, maybe that blocks some broken code 2024-09-29 20:07:32 ptrc, which platform? 2024-09-29 20:07:55 quinq: x86_64 2024-09-29 20:08:00 for me 2024-09-29 20:08:48 + ublock origin running 2024-09-29 20:13:40 same here then 2024-09-29 20:24:01 ikke, ack :) 2024-09-30 06:20:18 pkgs.a.o now has more options :D 2024-09-30 06:22:53 but subpackages contains itself, so strange 2024-09-30 06:32:13 "Required by" only shows the same or upper reposities, it is incomplete 2024-09-30 06:32:53 ACTION sighs his poor language 2024-09-30 07:21:22 thanks for the dark theme 2024-09-30 08:06:00 Is there an official reference document regarding BusyBox's (a)sh implementation? Or should I refer to POSIX (issue 7 or 8?)? 2024-09-30 08:08:22 Target portable shell, and hope that ash supports it 2024-09-30 08:09:41 quinq: part of me getting old is complaining about bad documentation and resolving to hope :-( 2024-09-30 08:10:43 There's no age to complain about bad documentation 2024-09-30 08:11:03 indeed 2024-09-30 08:58:49 There is a security update for php81 in edge/testing. Will this be backported to Alpine 3.19? 2024-09-30 09:06:20 qaqland: thx for the pointers. if anything else just let me know. 2024-09-30 09:06:50 harryweb: php81 is in community, which means it's no longer supported in 3.19 2024-09-30 09:08:10 ikke: Thanks 2024-09-30 09:14:17 ACTION Hi :) 2024-09-30 09:15:40 Is there an alpine init script for nfsdcld? Otherwise I am thinking of making a MR for it. 2024-09-30 09:23:54 Hi, is there any interest in an updated vulkan SDK? The edge version is fairly old now, and not compatible with ffmpeg 2024-09-30 13:15:26 Sure 2024-09-30 13:25:20 Is there something like openbsd.amsterdam but for Alpine? Hosted Alpine VPS on Alpine hosts by someone who supports the upstream project? 2024-09-30 16:00:14 can i tell apk install a package as dep of another package? 2024-09-30 16:07:18 are you thinking like gentoo use flags or something? 2024-09-30 16:09:26 "can i tell apk install a package..." <- Can't you select multiple packages for installation? e.g. apk install vim nano 2024-09-30 16:15:00 does anybody happen to know how to get "English (US, alt. intl.)" to work right for typing? it seems to have issues on alpine https://gitlab.alpinelinux.org/alpine/aports/-/issues/16375 2024-09-30 16:30:53 should this not hinder x to start if not member of input? seems i can start anyway and lose control of the machie ;( : groups|grep -q input && exec startx $(command -v ratpoison) -- -ardelay 200 -arinterval 30 || echo "failed, are you member of the input group?" 2024-09-30 17:10:17 How does one add a new wiki page? 2024-09-30 17:11:01 Is it username -> create? And does it go through some process? 2024-09-30 17:15:21 andar1an[m]: You navigate to the new url 2024-09-30 17:15:47 Typically you'd create a link to the page on some other page, and then when you navigate to it, you can add the content 2024-09-30 17:25:22 Oh, okay. I will try to figure out later 2024-09-30 17:25:35 I started a rough one in what I assume is my user space? 2024-09-30 17:25:42 will look more after lunch 2024-09-30 17:26:40 That link will be nice because I can stick it on the pages I used as basis 2024-09-30 17:30:50 with luks2 + btrfs subvols fde though, I couldn't figure out how to be super discreet with OS init top level dirs, and with poweroff it has trouble remounting root as ro which throws an error. Hopefully will figure that out 2024-09-30 17:31:41 e.g. ended up scrapping trying to have discreet subvols for etc, usr, boot 2024-09-30 17:31:52 wasn't worth the effort it seemed 2024-09-30 18:44:10 I have been frequently running alpine-make-vm-image in the alpine docker container. Is there anything I can do to avoid hitting the apk servers too much? setup-apkcache is not available in the docker image 2024-09-30 18:47:24 compared to how much traffic dl-cdn gets, I doubt you'd make the tiniest dent. Anyway, it suffices to make a symlink from /etc/apk/cache to somewhere to enable the cache 2024-09-30 18:48:26 our a volume mount / bind mount 2024-09-30 18:48:29 s/our/or/ 2024-09-30 18:58:26 Thanks ikke . After bind-mounting /etc/apk/cache I see some packages in the cache (like the qemu that was installed by the script), but not the packages that I passed to alpine-make-vm-image's -p option 2024-09-30 18:59:54 In the script I see a "# Use APK cache if available." section, let me see if I can figure out if that's a "one-way" cache 2024-09-30 19:02:41 It may make some kind of chroot 2024-09-30 19:06:06 hello \0 2024-09-30 19:06:58 i don't really know where to report that, soo: alpine's gitlab i believe does not allow hotmail.com domain 2024-09-30 19:07:20 allow for what? 2024-09-30 19:07:25 registration 2024-09-30 19:08:02 you could try using simple login or some domain proxy if you always have trouble 2024-09-30 19:08:19 We receive a huge amount of spam registrations from hotmail.com and since hotmail.com is blocking our e-mails anyway, not a lot is lost 2024-09-30 19:08:37 ikke: My bad. I bind mounted directly to /etc/apk/cache instead of doing what you said. The script needs /etc/apk/cache to be a symlink. Bind mounting to a different location and creating a symlink enables full caching. Thanks for your help! 2024-09-30 19:10:07 vshab: even if we allow hotmail.com, the confirmation mail would never arrive 2024-09-30 19:24:30 anyone here run xinit with i3 in an apple virtualization.framework vm? i3 launches but then freezes immediately 2024-09-30 19:25:15 this is how i created the vm image i'm using -> ./alpine-make-vm-image -a aarch64 -f raw -s 1024M -p "i3wm xorg-server xf86-input-libinput mesa-dri-gallium xinit eudev xterm firefox-esr" -- ./firefox.img 2024-09-30 19:28:16 And I launch the VM on macos with this command -> vfkit -c 2 -m 2048 --bootloader linux,kernel=./vmlinux,initrd=./initramfs-virt,cmdline='"root=/dev/vda2 rootfstype=ext4 modules=kms,scsi,virtio"' --device virtio-blk,path=./firefox.img --device virtio-input,pointing --device virtio-input,keyboard --device virtio-gpu,width=1440,height=900 --device virtio-rng --gui --device virtio-net,nat 2024-09-30 19:28:42 creddy: You certain it’s not that X is not getting input? As in, maybe you need elogind or to be in the input group. 2024-09-30 19:29:16 I did have to uncompress the vmlinuz-virt file that the script creates. Apple only supports uncompressed kernels 2024-09-30 19:30:16 I'm launching "xinit" as root humm . You still think it could be a permission issue? 2024-09-30 19:31:10 You might be right humm . Let me try to figure out if X has the proper input configuration for keyboard and mouse 2024-09-30 19:32:21 btw, it is so cool that Alpine makes the creation of minimal VM images so easy. Whoever worked on that, thank you! 2024-09-30 19:33:15 I built an image with the same packages using GUIX, and it was 5.2G compared to about 0.5G with alpine 2024-09-30 19:33:17 hex the point is when you remove a package the manualyl added dep is removed too 2024-09-30 19:51:46 how does alpine remount root when hitting shutdown or reboot runlevels? telinit doesn't seem to prevent readonly error, and I dk how one would remount something in use anyways at this point 2024-09-30 19:56:28 does it createa duplicate fs in memory in memory? and then it can remount disk fs as ro? 2024-09-30 19:56:40 s/in memory// 2024-09-30 20:01:19 I think there was a bug that was recently fixed related to remount in utils-linux, but I don't think that may be relevant here. This may be to do with /etc/init.d/mount-ro and maybe some caveat with btrfs subvols? I dk yet 2024-09-30 20:01:54 didn't have this poweroff error without btrfs subvols, even with a btrfs fs 2024-09-30 20:55:38 i guess telinit doesn't exist in alpine 2024-09-30 20:55:41 maybe init will work 2024-09-30 21:03:20 anyone know what package telinit comes with? Or if it should be bundled in install? 2024-09-30 21:03:37 or if alpine uses something else? 2024-09-30 21:04:05 by default, Alpone systems use busybox init and OpenRC 2024-09-30 21:05:48 /etc/inittab might tell you and init what init does when told to shut down 2024-09-30 21:06:11 what it does might be telling OpenRC to do stuff 2024-09-30 21:06:18 that’s all there is to it 2024-09-30 21:07:28 The problem is that the runlevel change can't unmount drive. Open RC docs in Arch mention using telinit for this issue with u option which: Tell init to re-execute itself (preserving the state). No re-examining of /etc/inittab file happens. Run level should be one of S, s, 1, 2, 3, 4, or 5, otherwise request would be silently ignored. 2024-09-30 21:08:31 I looked at inittab 2024-09-30 21:08:38 didn't see anything that could help 2024-09-30 21:15:14 also, what is init in sbin for if alpine uses busybox init" 2024-09-30 21:20:42 thank you humm I think this is the magic reference: https://github.com/brgl/busybox/blob/master/examples/inittab 2024-09-30 21:21:42 so maybe busybox mount is still plagued by the bug that was fixed for non-busybox mount 2024-09-30 21:22:25 last entry here is related: https://bugs.gentoo.org/573760 2024-09-30 21:23:52 /sbin/init is, like all the other executables provided by busybox, a symlink to busybox 2024-09-30 21:24:48 (or a copy or a link or whatever; I don’t have it in front of me) 2024-09-30 21:25:35 Cool ty. New to busybox 2024-09-30 21:28:38 once again, it was user error on my part humm . if i had read the documentation at https://wiki.alpinelinux.org/wiki/I3wm , i would have seen that "If you start i3 without fonts it will be unusable". All I had to do was include font-terminus in the list of packages included in the image 2024-09-30 21:32:41 oh. And more critically I had to run setup-xorg-base 2024-09-30 22:09:48 Does anyone know what default directory is for $RC_LIBEXECDIR? 2024-09-30 22:10:39 I think I need to modify $RC_LIBEXECDIR/sh/rc-mount.sh, but having trouble finding this 2024-09-30 22:11:52 Checked most usr dirs 2024-09-30 22:15:25 lib, sbin, bin too 2024-09-30 22:15:27 Nvm 2024-09-30 22:15:29 Cool. Haven't seen how fuser works yet 2024-09-30 22:15:29 It is rc, not openrc in lib