2024-10-01 00:25:35 Hi ALL! I'm new to alpine, i want to utilize unused cpu resources with xmrig on alpine inside VM. for best performance it requires huge pages enabled. also there is a script that makes changes to system, enabling them (https://raw.githubusercontent.com/xmrig/xmrig/refs/heads/master/scripts/enable_1gb_pages.sh) but seems like there is no path /sys/devices/system/node on alpine, What you suggest to enable them? 2024-10-01 00:48:10 Hi, I have a Lenovo Z13 laptop with a cirrus speaker dsp which requires firmware. The firmware is included in the linux-firmware-cirrus package, but the HDA modules required are non enabled in the alpine kernel config. What would be the best way to request they are enabled? 2024-10-01 05:35:19 tokyovigilante: You can create an issue here: https://gitlab.alpinelinux.org/alpine/aports/-/issues 2024-10-01 06:55:46 ikke: thanks! Seems just the -edge kconfig is missing the required CONFIG 2024-10-01 07:06:28 https://gitlab.alpinelinux.org/alpine/aports/-/issues/16487 if anyone is interested, thanks 2024-10-01 07:50:08 `apk` has a `-p` option/argument to control which root filesystem apk will work on. Can an `apk` hook do the same? 2024-10-01 07:53:11 I want expand on this immutability method, by automatically creating a new rw snapshot before`apk` does a commit, and having `apk` work on that snapshot. https://wiki.alpinelinux.org/wiki/Immutable_root_with_atomic_upgrades 2024-10-01 07:57:43 I've somewhat got something working by aliasing `apk` to `apk -p /mnt/root/next`, but it's a brittle and opaque solution. Any program or other user using `apk` without the proper env variables, just get errors 2024-10-01 08:22:43 papiris[m]: create a apkp in /usr/local/bin/ and use that :) 2024-10-01 08:23:40 Or just apk 2024-10-01 08:32:49 ty for suggestions! I want it to be transparent, i.e. graphical package managers like KDE Discover and Gnome App should also use `apk -p ...`. I've previously placed a script named `apk` in ~/.local/bin/ which calls `apk -p...`, with only partial luck. I'll try /usr/local/bin/ :) 2024-10-01 13:20:32 Does anyone use Pandoc? Do you find it outputs good MediaWiki syntax? 2024-10-01 13:21:12 seems pretty cool 2024-10-01 13:57:50 there are some options for busybox less to quite if file fits on screen, but it still uses the whole screen, can even supresse ~, but it fills it with blank lines instead, ... or am i missting some trick? 2024-10-01 14:16:36 -FX, frag 2024-10-01 14:18:15 lopid seems busybox less doesnt have the X option 2024-10-01 14:18:44 :( 2024-10-01 15:22:11 i miss the colored git logs with busybox less, so I usually install GNU less 2024-10-01 18:23:06 Hello guys !!!!! ,…….WE POST ALL PRODUCTS AND REVIEWS ON OUR MAIN CHANNEL AND GROUP. WE WON'T BE POSTING ANY PRODUCTS ON THIS CHANNEL.MAIN GROUP AND CHANNEL LINKS BELOW ⬇⬇⬇... (full message at ) 2024-10-01 18:29:45 AdeOpe[m]1: kindly fornicate yourself 2024-10-01 19:03:34 /mode +s would eliminate most of these incursions 2024-10-01 19:04:28 noone uses /list to find this chan, noone. except those you would kick anyway 2024-10-01 19:05:24 bOR38552MJA: does that matter for matrix users? 2024-10-01 19:08:11 oh damn. i totally forgot about those, they are not part of my bubble. please excuse my smartassing. 2024-10-01 19:08:30 Also that's not really a nuisance 2024-10-02 13:53:12 Hi guys. Have a question. If mouse has several programmable buttons - does it drive by umous or ukbd too ? 2024-10-02 13:54:14 have this message in system log when attaching device: uhidev2 at uhub4 port 2 configuration 1 interface 0 "Razer Razer DeathAdder Essential" rev 2.00/2.00 addr 7 2024-10-02 13:54:54 confused by this: wskbd2: connecting to wsdisplay0 2024-10-02 13:55:07 why it sees as wskbd ? 2024-10-02 13:55:21 wrong chat. sorry 2024-10-02 14:00:30 I sense an OpenBSD or NetBSD system 2024-10-02 14:12:59 For remounting drive readonly on shutdown runlevel, this is part of the OpenRC file:... (full message at ) 2024-10-02 14:14:43 WIll no umount result in dirty start state? 2024-10-02 14:17:14 humm, right 2024-10-02 14:17:39 just curious why connecting/disconnecting razor mouse touches wskbd 2024-10-02 14:17:44 Ah, maybe do_unmount is what flags are related to. 2024-10-02 14:17:57 seems because of programmable buttons on it 2024-10-02 14:33:01 There is also... (full message at ) 2024-10-02 14:34:26 Found this, but didn't didn't find what I need: https://github.com/OpenRC/openrc/blob/master/service-script-guide.md 2024-10-02 14:38:52 nvm, found it 2024-10-02 14:39:42 the function it defined in that script. I missed it 2024-10-02 15:04:35 I removed that entire function and just did mount -a -o remount,ro ... could that be problematic? lol. I figured all the mount info was in fstab 2024-10-02 15:13:10 I guess I should account for drives not in fstab too. 2024-10-02 15:16:04 which seemed the whole point of using fuser i guess 2024-10-02 15:22:39 I like how OpenRC does config scripts haha. This is pretty cool 2024-10-02 15:23:06 basically just shell haha 2024-10-02 17:14:28 hello 2024-10-02 17:14:41 I wanted to ask about the chromium package and the patches to get chromium to work with musl 2024-10-02 17:15:37 I was wondering if the patches concerning chromium's sandbox does anything to make it less effective 2024-10-02 17:16:32 I don't understand much if any of the code base so I can't really tell from myself when I view the patches https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/community/chromium/musl-sandbox.patch 2024-10-02 18:42:10 does anyone know if busybox mount -a also includes at mtab or proc? 2024-10-02 18:43:16 man just says fstab for -a, but it mentions mtab a few other places 2024-10-02 18:45:54 ignore, found mount.c 2024-10-02 19:06:57 doesn't even seem to be scanning for other mounts, so I think I can just ignore worrying about using mount -o remount,ro 2024-10-02 19:13:10 Get a journal, Logan 2024-10-02 19:16:48 weird that openrc-run manpages don't have reference to eindent or eoutdent 2024-10-03 09:37:03 Hey everyone, I made this as a kind of drop-in replacement for the standard repositories, but it keeps old versions of packages around. Maybe useful to some people, as I've seen some people complain about Alpine's policy of only keeping the latest version of each package. https://github.com/evertheylen/alpine-archive-proxy 2024-10-03 09:44:55 hi everyone, I am running Alpine Linux inside multiple virtual machines with hypervisor VirtualBox on Windows 10/11 as a host OS. After my update to Virtualbox 7.1 I face issues with the NAT networking: Alpine is not able update its APK repos anymore, as it seems to be a general network interface issue. Anybody facing the same problem who can tell me how to fix it? (I already checked in various virtualbox forums but did not find a solution ye 2024-10-03 09:55:15 old packages are in /var/cache/apk 2024-10-03 09:55:56 lopid: by default, no, you have to enable it 2024-10-03 10:07:27 cyberpunk123: do you have any internet access? 2024-10-03 10:07:47 lopid: yes, but I'm using this in a situation with hundreds of devices. I want to test one such device, verify that the current package versions work well, and then later I want to set up more devices exactly like it. If a package was updated in the meantime, I can't install the same version from the standard repos anymore. 2024-10-03 10:08:25 Maybe also relevant for Docker containers 2024-10-03 10:08:37 it probably resolves the old deps too, right? 2024-10-03 10:09:12 lopid: unsure what you mean 2024-10-03 10:09:46 an old package may rely on an equally old dep 2024-10-03 10:11:24 well it doesn't inspect the versions, but if you used the repo from the start you'll have downloaded all deps so all deps are stored 2024-10-03 10:11:34 *inspect the dependencies 2024-10-03 11:07:09 `tlp start/service` seem to be a quitck/easy way to hibernate laptop when closing lid (and probably more), mine sank a 3% i think it was in a week that way .. why is it not mentioned in the alpine laptop wiki? 2024-10-03 11:20:11 Does anyone use hyprland on Alpine? Have you enjoyed experience? 2024-10-03 11:28:18 i like ratpoison, i see there is one inspired by rp in repo for wayland too (havent tried it tho): cagebreak 2024-10-03 11:41:50 Cool, ty. I haven't tried a tiling wm before, i noticed a bunch in alpine wiki. I have seen hyprland a lot in media 2024-10-03 13:48:06 hello 2024-10-03 13:48:55 is anyone having issue with iwd (segfault) ? 2024-10-03 17:58:19 how do i add a user with adduser if the group with that name already exists? 2024-10-03 17:58:26 it just complains the group already exists but i don't care, i just want to add the user anyway 2024-10-03 18:00:34 -G 2024-10-03 18:01:41 addgroup foo && adduser foo -G foo -D 2024-10-03 18:02:31 heh 2024-10-03 18:02:36 i'm baaack! 2024-10-03 18:08:45 Welcome back 2024-10-03 18:10:33 :) been doing some wiki work, i run LLDAP on Alpine. Now i need to learn how to package it for Alpine :) 2024-10-03 18:17:42 i have a small router board with 3 ethernet ports on it, it seems to have Arch Linux on it.. when i mount it on my desktop i see the directories @, @home, @pkg, @log, @.snapshots i think this is BTRFS filesystem. Am i correct in assuming i can achieve the same with Alpine? make it "immutable" or something? it doesnt have a monitor connection, but i suppose i could make it somehow to restore the last snapshot when the upgrade 2024-10-03 18:17:42 fails somehow 2024-10-03 18:19:08 JohannesJacobs[m]: sounds like you want a run-from-ram system, which was created for these kinds of deployments 2024-10-03 18:20:05 no, i think thats not what i want :) i have a 16GB msata disk in it :) i want to fill up the whole 16GB :)) 2024-10-03 18:20:35 but im thinking, if i use BTRFS and snapshots, i can somehow automate it so that it reverts to the previous snapshot if it fails to boot? 2024-10-03 18:21:00 But since Alpine is targetted at these small systems, perhaps that already exists? 2024-10-03 18:21:02 I have no experience with BTRFS 2024-10-03 18:21:13 I'm not aware of such a setup myself 2024-10-03 18:21:32 me neither.. other then "we use it on our servers" but those have automated backups to NAS storage etc :D 2024-10-03 18:22:00 "addgroup foo && adduser foo -G..." <- ah cool, thanks! 2024-10-03 18:22:40 the problem is that adduser by default tries to create a group with the same name 2024-10-03 18:22:44 and that's what it fails on 2024-10-03 18:41:02 "but im thinking, if i use..." <- im working on something like this now 2024-10-03 18:41:53 brtfs subvolumes have some caveats I think, but so far I have a working install with subvols. Need to investigate grub-btrfs still which I saw in APK (maybe testing) 2024-10-03 18:42:13 as for refind btrfs I don't see that yet 2024-10-03 18:42:23 trying to share my learnings, but I am still in tinker mode 2024-10-03 18:42:34 and they are obvs not going to be perfect haha 2024-10-03 18:42:35 anything i can help you with? though my knowledge of alpine and btrfs is fairly newb :) 2024-10-03 18:43:03 I'm currently reading up on abuilds, see if i can package LLDAP for Alpine :) 2024-10-03 18:43:24 I'm in a pretty good spot right now. RIght now I am just adding daily driver stuff into my scripts. Was looking at Aports cuz I was curious about hyprland and stuff 2024-10-03 18:43:59 on the next machine I will look at raid and secure boot more, but this one has FDE 2024-10-03 18:44:03 with Luks2 2024-10-03 18:44:23 yeah, i created a script for alpine as my daily driver as well :) i got myself a nice looking domain, gonna spend some time to write my tinkering on there as well 2024-10-03 18:44:26 don't want to mess with secure boot keys because it is a lenovo 2024-10-03 18:44:43 i have a Lenovo ideapad 5 :) 2024-10-03 18:45:08 x1 yoga gen 4 for me. Though I want a Starlab Starfighter so bad haha. Corebooooot 2024-10-03 18:46:46 I think the 2 main caveats for btrfs I have bumped into are not creating subvolumes for directories necessary to boot ( I bet sure it is possible, but I got annoyed haha). And may need to update openrc-run shutdown mount-ro so you don't get an unmount error if it bothers you 2024-10-03 18:48:35 JohannesJacobs[m]: There is also alpine wiki. 2024-10-03 18:49:26 yeah, i know the wiki :) 2024-10-03 18:56:30 For your daily driver scripts, have you found a nice way to package installation and setup conveniently? 2024-10-03 18:56:30 I have been looking at mkimage, because I am most interested in netboot, but for formatting disks I haven't determined if I can trigger scripts yet. Been meaning to look more closely at iPXE chainloading but at first glance I didn't see how to run code related to disk formatting 2024-10-03 19:00:56 could have sworn grub-btrfs was in testing in apk, and refind-btrfs wasn't, but now I don't see either haha 2024-10-03 19:01:31 Will need to look at aports 2024-10-03 19:01:37 i have not yet found any way to package it at all :) I've been looking into plymouth, gave that up a while ago due to some errors 2024-10-03 19:05:07 Baby steps haha. My goal is to orchestrate baremetal over ipv6, both firmware and general purpose os 2024-10-03 19:05:57 Lot's of cool stuff happening in that space right now 2024-10-03 19:07:42 my goal is to build something i can share with friends :) i'm getting pretty fed up with all these Windows problems in my whatsapp :) 2024-10-03 19:10:13 that's why I am working on baremetal orchestration haha. Distributed system for local first - small community - application for music discovery 2024-10-03 19:12:40 same same, but then for family and friends :) thanks to flatpak almost everything can be installed 2024-10-03 19:13:07 I am looking at wasm and initially pwafor similar reasons 2024-10-03 19:13:22 * I am looking at wasm and initially pwa's for similar reasons 2024-10-03 19:14:59 don't really have a desire to write 3 different code bases for the same app lol 2024-10-03 19:15:42 well, the only app development i do is web apps in go :) 2024-10-03 19:17:37 Nice, Rust was what I fell in love with 2024-10-03 19:19:16 tried that too, found it too hard to learn :) 2024-10-03 19:19:50 i tried PHP, Ruby, Onjective-C, Swift, V, Rust and settled for Go 2024-10-03 19:20:02 i have tons of video courses for all of them by now :) 2024-10-03 19:21:26 i'm becomming pretty good with Golang :) gonna build a basic web interface for nftables for myself, see if i can build my own router :) 2024-10-03 19:29:29 Nice, was a similar journey for me. C++, JS, PHP, Go, Python, settled with Rust. I found Rust had some great resources like Rustlings, and I came from mech eng, it just seemed to click easier for some reason, and I like the community. Was early on wasm train too. 2024-10-03 19:29:29 I have both OpenWRT and Mikrotik devices right now which are based on the linux kernel I believe. They are great. A lot of my investigation is on firmware, micro and unikernels and rtos's as well. Trying to find that sweet spot for complexity related to layers of stack. 2024-10-03 19:41:14 i know those too :) nice little boxes. but they dont scale well if you need to maken a few of them :) 2024-10-03 19:41:52 I'm working on a server/client model. build firewall rules using a web gui, then push it to the clients 2024-10-03 19:42:06 Alpine is a nice fit :) 2024-10-03 19:43:47 Do you use alpine fw? 2024-10-03 19:44:00 no, atm i'm using Ubiquiti 2024-10-03 19:44:29 Ah. I tried Ubiquiti, didn't have a good time. I like the mikrotik scripting and ease to connect. 2024-10-03 19:44:50 Alpine has Awall though right? 2024-10-03 19:45:11 awall is an option, yes 2024-10-03 19:45:28 cool 2024-10-03 19:46:39 yeah :) ive used awall :) 2024-10-03 19:47:21 has ubiquiti upgraded mongodb version yet? haha 2024-10-03 19:47:36 (for control sw) 2024-10-03 19:51:58 ugh! so tomorrow i need to get a small computer so i can put the msata disk in it, and boot the alpine iso 2024-10-03 19:52:17 i tried DD'ing the alpine minifs root to it, but that failed miserably 2024-10-03 19:53:39 why did it fail? 2024-10-03 19:53:58 probabl;y because i didnt create an EFI partition 2024-10-03 19:54:28 oh, you mean fail to boot, not transfer? 2024-10-03 19:54:43 oh no, DD worked flawless :) 2024-10-03 19:56:33 can you use a live installer to format it? What is stopping you from partitioning? 2024-10-03 19:57:10 its a small blue box, it has 2 USB ports, 3 Ethernet ports and a serial port :) 2024-10-03 19:59:27 lol, can give it to your partner and pretend it is from Tiffany's? haha 2024-10-03 20:01:27 :D 2024-10-03 20:01:57 i think i figured out my Alpine troubles loooooong before i found a partner 2024-10-03 20:02:19 in fact, i think i'd be building my own distro before i find a partner 2024-10-03 20:10:37 tomorrow i will put the msata disk in a desktop computer, then i can live boot :) 2024-10-03 22:45:12 bau 2024-10-03 23:55:30 Hello everyone 2024-10-03 23:55:37 so I want to dualboot alpine 2024-10-03 23:55:56 But the USB is refusing to boot 2024-10-03 23:56:13 It gives "mounting boot media failed. 2024-10-03 23:56:23 * media failed." 2024-10-03 23:56:49 and then "initramfs emergency shell launched. Type 'exit' to continue boot" 2024-10-03 23:57:23 I tried making syslinux point to my USB stick (directions on alpine wiki), but it won't work 2024-10-04 07:01:51 4194304a: have you tried with reFIND? its an EFI manager, i found dual booting much easier with refind 2024-10-04 12:00:36 It seems that the texlive-xetex package has regressed recently on 3.20. Previously this installed xdvipdfmx but now it doesn't. 2024-10-04 12:00:38 I am not an alpine user or expert but we are building alpine images for our CI infrastructure. 2024-10-04 12:00:46 Can anyone point me where to investigate this regression? 2024-10-04 12:01:44 It could also be texmf-dist-latexextra 2024-10-04 12:01:52 those are the two latex related packages which we are installing 2024-10-04 12:02:28 Perhaps an issue with https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/67391 2024-10-04 14:45:38 hiya. i want to build an alpine image for my riscv uconsole, but not sure what i should do. there's a "minirootfs" available for download on the website, and the wiki has nothing specifically about riscv, aside from a single sentence for riscv64 2024-10-04 14:49:30 hey xyhhx :) i think the general instructions apply, you just have to replace amd64 with riscv :) 2024-10-04 14:49:39 xyhhx: the minirootfs image is only meant for service-less containers 2024-10-04 14:50:58 JohannesJacobs[m]: what are the general instructions? do you mean like https://wiki.alpinelinux.org/wiki/Create_a_Bootable_Device ? 2024-10-04 14:51:06 ikke: yeah that's why i was kind of confused 2024-10-04 14:51:25 exactly 2024-10-04 14:52:06 xyhhx: the problem with these architectures is that it's almost impossible to make a generic image to boot 2024-10-04 14:52:30 it's okay, i don't need a generic image, i' 2024-10-04 14:52:36 i'm building for a specific device 2024-10-04 14:52:59 ikke: why is that? im curious about riscv too :) 2024-10-04 14:53:40 so presumably i'd follow https://wiki.alpinelinux.org/wiki/Create_a_Bootable_Device#Copying_ISO/tarball_contents_to_a_USB_stick and use the riscv minirootfs? 2024-10-04 14:54:10 Thats what i would try, yes 2024-10-04 14:54:45 From what I understand about it is that you cannot easily discover all the hardware, so you need a device specific dtb that tells what he there is 2024-10-04 14:54:46 that's what i assumed, so good to know i'm probably on the right track lol 2024-10-04 14:55:07 Also, kernel support may be missing as well 2024-10-04 14:55:18 ikke: is this riscv specific? or does arm has that problem as well? 2024-10-04 14:55:24 Same 2024-10-04 14:55:41 oh hmmz 2024-10-04 14:56:01 doesn't arm have prebuilt isos? 2024-10-04 14:56:40 aarch64 has 2024-10-04 14:56:52 armv7 / armhf don't 2024-10-04 14:57:09 woops my bad 2024-10-04 14:57:43 anyway i can't wait to try my uconsole with alpine! thanks for the tips and the quick responses :) 2024-10-04 14:58:01 i flashed alpine onto a raspberry pi once, but that was in 1729. I dont know how well that goes these days 2024-10-04 14:58:31 unrelated (obv), but does anybody have experience with building atomic alpine: https://wiki.alpinelinux.org/wiki/Immutable_root_with_atomic_upgrades 2024-10-04 14:59:10 one of these days i wanna try my hand at building a xen-based system with alpine as my dom0 and if i could make it immutable/atomic that'd be even better 2024-10-04 14:59:38 i was discussing that last night actually :) I'm looking for a way to build an immutable router image based on alpine :) 2024-10-04 15:00:00 beautiful 2024-10-04 15:02:00 i have a small blue router box with only 3 ethernet ports.. no display or anything :) 2024-10-04 15:08:06 apu2? 2024-10-04 15:08:07 like this? https://docs.banana-pi.org/en/OpenWRT-One/BananaPi_OpenWRT-One 2024-10-04 15:08:08 ;) 2024-10-04 15:08:20 xyhhx, only two ports on that one 2024-10-04 15:08:35 I have https://docs.banana-pi.org/en/BPI-F3/GettingStarted_BPI-F3 2024-10-04 15:08:56 this one: https://www.linuxserver.io/blog/2016-12-17-review-pcengines-apu2-c4-prebuilt-by-linitx 2024-10-04 15:09:01 i only just learned about this yesterday :sob: 2024-10-04 15:09:08 This one is better: https://docs.banana-pi.org/en/BPI-F3/BananaPi_BPI-F3 2024-10-04 15:10:12 I'm looking for a good rsicv board next year. I believe Alpine + RiSC-V = Gold (if it works) 2024-10-04 15:10:50 xyhhx, note that what they're selling now still appear to be "samples" 2024-10-04 15:15:10 Habbie: yeah, but i'm still interested 2024-10-04 15:16:05 ack 2024-10-04 15:16:10 so am i but i couldn't tell you what for ;) 2024-10-04 15:16:14 "just looks cool" 2024-10-04 15:16:23 meanwhile i just sold off an apu2 ;) 2024-10-04 15:17:19 well, for me its fairly asimple :) I'm learning Golang web applications, i'm gonna make a web gui for nftables on Alpine (or a web wrapper arround awall, not sure yet) 2024-10-04 15:17:48 and yes, there's openWRT, Ubiquiti and Cisco.. But whats the fun in that :) 2024-10-04 15:18:32 (OpenWrt) 2024-10-04 15:19:17 New package merge requests going in testing? 2024-10-04 15:19:23 yes 2024-10-04 15:19:31 ty 2024-10-04 15:20:49 Habbie: yes, that one 2024-10-04 15:23:26 "This one is better: https://docs..." <- vla 1.0? 2024-10-04 15:24:23 ? 2024-10-04 15:26:54 https://github.com/riscv/riscv-v-spec/releases 2024-10-04 15:26:56 Last I looked not so many boards were designed with new generation of chips for 1.0 vector extension. I have been waiting to bite bullet for risc-v until hopefully a cm4 form vactor with vla 1.0 support is released 2024-10-04 15:28:52 rv64imafdcv_zicbom_zicboz_zicntr_zicond_zicsr_zifencei_zihintpause_zihpm_zfh_zfhmin_zca_zcd_zba_zbb_zbc_zbs_zkt_zve32f_zve32x_zve64d_zve64f_zve64x_zvfh_zvfhmin_zvkt_sscofpmf_sstc_svinval_svnapot_svpbmt 2024-10-04 15:28:57 No idea if that incluse vla 1.0 2024-10-04 15:31:02 No clue, couldn't see it defined on diagrams and wasn't gonna go through that list lol 2024-10-04 15:31:51 nvm 2024-10-04 15:31:53 it does: Vector Extension: RVV1.0/VLEN 256/128-bit x2 execution width 2024-10-04 15:32:46 I wonder if they have any cm4 compatible... 2024-10-04 15:47:03 how accurate can one assess "disk health"? worried about some old backup disks, saw some tips online, eg use smartctl and check Wear_Leveling_Count ... 2024-10-04 15:47:20 smartctl also gives you an overall assessment i think 2024-10-04 15:47:38 besides wear leveling you can also compared total bytes written to manufacturer specs 2024-10-04 15:47:45 but nothing is a promise 2024-10-04 15:47:49 oh, so its pretty good? im backing up atm, will test when its done.. 2024-10-04 15:48:09 how good it is depends on the vendor 2024-10-04 15:48:29 i deletet some dirs while backing up too, took a long time to delete............. prob bad sign.. 2024-10-04 15:49:47 i ran alpine from a usb on a raspi before, over-night it slowed down very ... maybe i should dig that up and test that one too, not sure if just worn or not.. 2024-10-04 15:50:17 oh, ty for the tip habbie! 2024-10-04 15:50:23 np :) 2024-10-04 15:50:26 i'm sure there's more, i'm no expert 2024-10-04 15:50:46 but the most important tip is that it can still just die any time and you should have backups 2024-10-04 15:51:36 i have good experience with sandisk micro sd cards btw.. i do 'mount -o noatime' tho 2024-10-04 15:51:59 relatime has been default for a while, which is a lot better than atime 2024-10-04 15:52:13 noatime doesn't do much any more (but not nothing) 2024-10-04 15:56:08 oh, havent looked into it in a while, just use a script to mount stuff ... whats the best way to reduce wear? 2024-10-04 15:58:00 "write less" :) 2024-10-04 16:00:57 you mean, use it less? hehe, i guess i do, i run from RAM.. 2024-10-04 16:01:14 will check if i can move more to RAM 2024-10-04 16:01:27 oh very good start :) 2024-10-04 17:46:36 cant stop modloop because cant unmount /.modloop resource busy, how can i check whats hindering it? fuser -m /.modloop doesnt give any results 2024-10-04 17:49:42 seems most usb devices and sd cards are not supported by smartctl :/ 2024-10-04 17:53:50 hmz 2024-10-04 17:56:09 ACTION pounces MatthewPickering[m]! 2024-10-04 17:56:14 Squeak! 2024-10-04 18:01:29 when cloning aports, is git status suppost to show a bunch of deleted files @ ../../testing/* 2024-10-04 18:01:41 s/suppost/supposed/, s//*//\*/ 2024-10-04 18:01:45 no 2024-10-04 18:01:53 what I assumed 2024-10-04 18:01:57 ty 2024-10-04 18:05:14 probably cuz I moved after clone lol. oops 2024-10-04 18:21:53 Do I need to worry about this type of warning if abuild is not in home? /home/andar1an/packages//community: No such file or directory 2024-10-04 18:22:28 andar1an[m]: I think that's output from apk update 2024-10-04 18:22:50 You can ignore that 2024-10-04 18:22:56 Ah okay, seems to have built. Just wanted to make sure. Thank you 2024-10-04 18:37:54 i guess there is a reason sdbus-cpp is still at 1.5.0, my bad haha 2024-10-04 18:39:07 at least the lint passed lol 2024-10-04 19:22:14 APKBUILD reference doesn't seem to mention util(), where can I read about it? 2024-10-04 19:22:41 repo? or are there docs other than wiki? 2024-10-04 19:22:50 That's a custom subpackage 2024-10-04 19:23:00 defined in $subpackages 2024-10-04 19:23:13 ty 2024-10-04 19:23:54 see it now, merci 2024-10-04 20:11:23 Does abuild checksum not use pkgname and pkgversion to build tar.gz name? I am getting ...ec6e29421b17d v0.1.2.tar.gz 2024-10-04 20:11:43 no 2024-10-04 20:12:00 it uses the name from the variable "sources" iirc 2024-10-04 20:12:21 Yes, and you can also override it 2024-10-04 20:12:32 $pkgname-$pkgver.tar.gz:: 2024-10-04 20:12:56 my bad, "source" not "sources" 2024-10-04 20:13:01 Thanks rdbo , just was about to say that is what it looks like. 2024-10-04 20:13:36 Thanks ikke , I will add that onto front of source I guess? 2024-10-04 20:14:21 source="$pkgname-$pkgver.tar.gz::" 2024-10-04 20:14:27 I guess abuild checksum is not generating it but it is based on a release checksum for upstream bin? 2024-10-04 20:14:35 Any Cashapp? Chime Apple Pay? pal? Venmo? BOA? Wells Fargo? Join my channel now... (full message at ) 2024-10-04 20:16:51 Also, I see tests referenced in some, but I don't see where tests live. Are they in a different repo? 2024-10-04 20:19:04 I love cargo lol. cmake makes me angry. will get back at it another time 2024-10-04 20:40:00 hello, I have a problem at using libvirt with GPU passthrough, I'm getting error -22 for vfio whenever I try to start the VM. I've followed the instructions from Alpine wiki. 2024-10-04 21:25:10 hmz, google is not exactly hepful 2024-10-04 21:25:28 Chimmie: does it boot when you disable gpu passthrough? 2024-10-04 23:27:06 it boots without passthrough and without acceleration what I'm looking for. 2024-10-05 00:53:45 "4194304a: have you tried with..." <- The problem I have is booting the USB itself 2024-10-05 00:54:09 does it perform something that allows it to figure out the device name properly? 2024-10-05 02:57:15 Hi, is there a consensus on the best way to start a dbus user session with sway without elogind? ie just using seatd. I have been using dbus-run-session -- sway, but this then assigns a random DBUS_SESSION_ADDRESS rather than $XDG_RUNTIME_DIR/bus 2024-10-05 02:59:34 Ideally I'd start the session bus prior to exec sway, as I'd like the option to have a login to a tty with a session bus available (so pipewire etc will work) 2024-10-05 03:05:38 i don't know about consensus but i can tell you how i do it. 2024-10-05 03:07:51 sure :) am following WhyNotHugo's desktop guide more or less, and dbus seems to automagically launch there somewhere between exporting the session address bus as $XDG_RUNTIME_DIR/bus and starting sway 2024-10-05 03:08:29 Ideally I want a predictable session bus address so I can use notify-send from scripts (eg networkmanager-dispatcher) that run as root to send notifications to my user session 2024-10-05 03:13:25 https://tpaste.us/BMey 2024-10-05 03:14:03 the if block you put in your ~/.profile 2024-10-05 03:14:33 that's how i do it, anyway (wfm) 2024-10-05 03:15:22 thanks, will give it a roll 2024-10-05 03:15:35 WhyNotHugo is more active on all the things than i am 2024-10-05 03:20:04 Just couldn't find the magic sauce in WhyNotHugo's dotfiles, thanks! 2024-10-05 03:21:38 np 2024-10-05 10:33:01 Because of a typo, I noticed that it was possible to do something like `apk apk apk info -R tmux` or `apk info tmux apk info helix`. Is there a design decision behind that? 2024-10-05 10:44:32 y al 2024-10-05 10:46:09 is it easy to turn off modprobe and copy/unpack modules/libs to RAM instead? (diskless raspi install), if so how? 2024-10-05 11:54:22 Hi everyone. Really loving alpine as my daily drive desktop lately. Keep up the good work! 2024-10-05 12:26:59 Alpine is really awesome as daily driver :) i use it with KDE :) 2024-10-05 12:30:19 Awesome! I'm on sway myself. I would like to contribute to the project any way I can - would hate to see it unmaintained one day. 2024-10-05 12:51:42 Well, the wiki can always use some love i believe :) And you can become a package maintainer (i'm in the process of learning packaging myself) 2024-10-05 12:53:23 Yeah wiki does seem to be a bit lacking in some spots. Trying to setup wireguard for proton vpn right now and it's a bit of a headache. 2024-10-05 12:54:14 Does anyone know of an APKBUILD that references a CMakeLists file, or is an example of translating one to APKBUILD? 2024-10-05 12:54:26 recently i registred a domain where i want to put my thoughts, struggles and idea's to use Alpine as a daily driver :) 2024-10-05 12:54:31 Tinky-Winky: i don't look at the wiki much myself, but one thing i could suggest that might increase its value: a closer relationship to what's going on in gitlab 2024-10-05 12:55:00 having wiki pages that link to important issues in gitlab for instance 2024-10-05 12:55:47 gitlab is where the work is, and the wiki feels like it's on another planet (to me, personally) 2024-10-05 12:57:14 The wiki can often be a little dated for sure, but is different purpose than gitlab I think 2024-10-05 12:57:17 I've oinly added 2 articles yet, but i plan to contribute some more. Using the same layout. 2024-10-05 12:57:50 I'm in process of 1 or 2 also, but have tweaked a few with notes 2024-10-05 12:57:51 andar1an[m]: i don't think you understand my point 2024-10-05 12:58:12 I do, just don't understand who is going to do that 2024-10-05 12:58:21 anybody can do that. 2024-10-05 12:58:36 Ah I hear you. Interesting, I haven't looked at the gitlab yet. Guess wiki has served me decently well for most things, but I've only been using alpine for a few months now. 2024-10-05 12:58:41 anybody can doesn't mean somebody will 2024-10-05 12:59:24 Tinky-Winky was literally just offering to help 2024-10-05 12:59:41 anyway it's just a suggestion. 2024-10-05 13:00:53 I understand, I am not trying to knock what you said, I agree that it would be nice. But without staffing and based on volunteer work that can vary, just trying to wonder if there is a methodology that is less dependent on relying on labour 2024-10-05 13:01:15 Sorry if it came across poorly 2024-10-05 13:02:08 The wiki would mostly benefit from a core group of users who will manage it, IMHO 2024-10-05 13:02:25 make sure its regulary checked, updated, etc etc 2024-10-05 13:02:47 And can be linked to gitlab issues too :) 2024-10-05 13:04:33 Anyway, was a pleasure interacting with you. Chat to you all again sometime! Cheers. 2024-10-05 13:05:12 It's always nice to chat with people who use the same software :) 2024-10-05 13:05:52 which reminds me to check the matrix article in the wiki, i had some problems following it 2024-10-05 13:05:58 Good evening everyone! 2024-10-05 13:06:17 Its afternoon here :) 15:00 in NL 2024-10-05 13:06:19 I was wondering if we have something similar to xtrlock in our repos that locks the screen but still lets anyone see the content? 2024-10-05 13:06:26 9am here haha 2024-10-05 13:06:31 2206 in Seoul atm 2024-10-05 13:06:35 am is morning right? 2024-10-05 13:06:45 telmich: sounds like what you may want is maybe kiosk stuff 2024-10-05 13:06:51 andar1an: hey! I actually read stuff from you... you are also doing ipv6 only deployments, afair? 2024-10-05 13:07:19 Working towards it! Baby steps haha. 2024-10-05 13:07:20 telmich: you have links? :) 2024-10-05 13:07:32 IPv6 is on my list too 2024-10-05 13:07:43 Well, all our DCs are IPv6 only 2024-10-05 13:07:56 So we have deployed maybe around ~200 or so Alpine Linux servers IPv6 only 2024-10-05 13:07:59 +/- 2024-10-05 13:08:10 nice! 2024-10-05 13:08:30 Nice! Do you have public addresses / ASN? 2024-10-05 13:08:32 ive only recently started to migrate our servers to Alpine where possible :) 2024-10-05 13:09:07 andar1an: AS209898, AS207996, AS213081 are the main ones 2024-10-05 13:09:24 Im mostly focussed on alpine for development and user devices for daily driving. Servers I am investigating alternatives to general purpose OS's 2024-10-05 13:09:41 We are mainly deploying alpine linux using https://code.ungleich.ch/ungleich-public/ungleich-tools/src/branch/master/alpine-install-on-disk.sh nowadays 2024-10-05 13:09:45 telmich: I recently got my first ASN haha, that is so awesome 2024-10-05 13:09:47 my next tinker project is a small APU box with Alpine as router/firewall 2024-10-05 13:09:58 the background is that we migrated from netboot to disk based, as we want to have a local cache for k8s container images 2024-10-05 13:10:22 JohannesJacobs[m]: That sounds like a great idea, we have maybe around 20 of them in use with alpine, a very good choice 2024-10-05 13:10:33 (even though pcengines is stopping to produce them...) 2024-10-05 13:10:38 I'm trying to go more of a "microservice as microcontroller" approach, haha. I am still early in tinkering though 2024-10-05 13:10:46 businesswise i would like to use Alpine as a thinclient 2024-10-05 13:11:00 Interesting stuff is going on here! 2024-10-05 13:11:11 but there is this guy named phil operrman in rust space that does cool blogs about writing bootloaders and os's 2024-10-05 13:11:11 andar1an[m]: You are in the US? 2024-10-05 13:11:24 oh? i dont know pcengines much really.. i just got the box dropped in my lap because it was beeing thrown away :D 2024-10-05 13:11:41 telmich: Toronto 2024-10-05 13:11:58 but it would be a nice replacement to the $500 Ubiquiti firewalls we deploy now :) 2024-10-05 13:12:04 Johannes Jacobs: We ordered many from them, the boxes are rarely dying and if so, many of the boards were replaced on warranty 2024-10-05 13:12:15 nice :) 2024-10-05 13:12:33 I'm having my eye turned to RISC-V.. trying to find some nice boards for that :) 2024-10-05 13:13:02 this week i started selfhosting matrix and lemmy, all on alpine <3 2024-10-05 13:13:14 next up is mastodon and peertube 2024-10-05 13:13:23 RiscV, ESP, and Nordic are exciting in embedded space 2024-10-05 13:13:50 Arm I guess too, but I am excited to get to a point where I can mess lower level haha 2024-10-05 13:14:33 Personally I moved away a bit from embedded 2024-10-05 13:14:35 but my ultimate dream (yeah.. i'm a dreamer..) is an Alpine phone, i can put in a dock on my work where it turned into an Alpine desktop for usage.. Then when time to go home i simply put it back in my pocket :) No more laptops, no more tablets.. just the one device. 2024-10-05 13:14:42 Did some arduino stuff in the last years, but it feels so limited 2024-10-05 13:15:02 There is magic happening in Rust Embedded right now 2024-10-05 13:15:08 Johannes Jacobs: Alpine Linux phone that can actually make calls (modem blobs!) would be quite nice 2024-10-05 13:15:25 Well, there's PostmarketOS :) 2024-10-05 13:16:28 I wonder why they defaulted to systemd for mobile. Does postmarket still use musl? 2024-10-05 13:17:13 Yeah.. ive been forgetting about postmarketos a bit the last few years, so im a bit outdated on that part. But i did notice their move to systemd 2024-10-05 13:17:26 maybe support for packages or tools that they use? Or starting services? 2024-10-05 13:17:29 But if i wanted systemd i'd stick to debian :D 2024-10-05 13:19:45 I personally came to alpine for the minimalism. I am trying to limit a) what is running, b) what I don't know is on the system, c) external scripts and configuration I don't understand 2024-10-05 13:20:12 Minimalism and Security was it for me :) 2024-10-05 13:20:26 ya, security is why I want minimalism 2024-10-05 13:20:46 I personally believe in security through visibility and limiting vectors, not obfuscation 2024-10-05 13:21:14 well, something can be big and still secure :) 2024-10-05 13:21:24 I've been a longtime Debian user 2024-10-05 13:21:26 I don't necessarily agree haha 2024-10-05 13:21:54 I prefer small and secure over big and secure for sure :) But one doesnt exclude the other 2024-10-05 13:22:08 it just makes it a bit harder :) 2024-10-05 13:23:27 I am working solo, so I don't have the ability to manage the scope of big and secure haha 2024-10-05 13:23:43 i do :) 2024-10-05 13:24:02 But i dont wan't to :) so there's Alpine :) 2024-10-05 13:24:31 easier to scale small and secure haha 2024-10-05 13:24:43 true! 2024-10-05 13:25:00 I will likely be mixing alpine with rtos's and firmware deployment 2024-10-05 13:25:17 If I don't need a 25M line kernel to do something, I would like to avoid that 2024-10-05 13:26:25 I'm mixing Alpine with Debian. At home everything is Alpine (except the router and the iphone/ipad) and at work ive been moving more and more to alpine, but in case where the software doesnt run in docker (yet) or other reasons, we deploy debian 2024-10-05 13:26:51 I saw a C quote recently that stated for every 1000 lines written in C, there is 1 bug. And it was cool reason to think about alpine 2024-10-05 13:27:11 I still have Deb on 1 machine, while shifting to Alpine, but I also learned a tonne from NixOS 2024-10-05 13:28:00 I had challenges getting used to the Nix Store though personally. I don't come from a CS background, so I am trying to achieve similar end goal with alpine and btrfs 2024-10-05 13:28:27 ive heard about nixos.. but thats it :) 2024-10-05 13:28:50 I've becomming really smallminded where linux is concerned in recent years 2024-10-05 13:29:12 In what way? 2024-10-05 13:29:25 Well, i used to try every and all distro's :) 2024-10-05 13:29:41 i followed tons of linux sites, kept an eye on everything 2024-10-05 13:29:56 but thats all becoming less and less. 2024-10-05 13:30:11 I have been doing the same recently, but now I need to dial in for use too. 2024-10-05 13:30:36 I think I went through all major distros except fedora and gentoo over last 2 years haha 2024-10-05 13:30:49 gentoo I learn a lot about through alpine 2024-10-05 13:30:54 I dont like it anymore.. maybe because im getting older. But im more focussed on "what can i do on alpine" instead of "what os can i try now" 2024-10-05 13:31:22 Next to that is the fact that my business is growing, more and more customers using Windows 2024-10-05 13:31:35 For me it is usually a "giv'r until I hit a blocker", then I need to make a choice 2024-10-05 13:31:40 after a day of that.. i just dont have the energy to mess with my own stuff as well :) 2024-10-05 13:31:54 haha true 2024-10-05 13:32:12 I am working on a PWA, or wasm based native apps, so client OS is less important to me 2024-10-05 13:32:27 The one regret i have is that thinlinc doesnt work on Alpine :) 2024-10-05 13:32:54 Ive been migrating our workplace to thinlinc recently. Works marvelous on Windows and Ubuntu workstations :) 2024-10-05 13:33:06 Right now I am trying to figure out cmake stuff to try an package some extra hypr stuff to try out haha 2024-10-05 13:33:06 what's that? 2024-10-05 13:33:24 Thinlinc is a bit like desktops in the cloud.. thinclients... etc 2024-10-05 13:33:40 a remote dektop 2024-10-05 13:33:47 exactly 2024-10-05 13:33:49 like nutanix frame and such? 2024-10-05 13:33:54 RDS, but for Linux :) 2024-10-05 13:34:10 Nutanix is a hypervisor no? 2024-10-05 13:34:34 they had a remote desktop implementation for access I believe 2024-10-05 13:34:40 ahh :) 2024-10-05 13:34:43 maybe it was just part of the solution 2024-10-05 13:34:44 like citrix workspace 2024-10-05 13:35:14 Does parsec work for business? 2024-10-05 13:35:23 I have seen that commonly used on linux 2024-10-05 13:35:24 dont know what that is 2024-10-05 13:35:31 may be in alpine 2024-10-05 13:35:45 oh, its remote desktop too 2024-10-05 13:35:59 It was a game streaming platform, that many people use for RDPS 2024-10-05 13:36:10 s/platform/RDS/ 2024-10-05 13:37:16 nvm, don't see it in alpine 2024-10-05 13:37:39 its one of the last remaining debian servers :D 2024-10-05 13:38:13 can't you use gcompat or a compatability layer, or just install in chroot or container? 2024-10-05 13:38:17 an alpine based thinclient with thinlinc would be awesome though :) 2024-10-05 13:38:39 maybe :) but thats for later testing! 2024-10-05 13:38:49 right now we're still working out the kinks as it is :) 2024-10-05 13:38:53 vscode ssh extension worked well when installing gcompat and stdlibc++ 2024-10-05 13:39:00 true haha 2024-10-05 13:39:33 And i want to spend some time penning down my own journey with Alpine. Gettingit out of my head :) 2024-10-05 13:39:57 I usually just do that in markdown as I tinker through scripts haha 2024-10-05 13:40:19 hopefully pandoc makes it easy to translate to mediaWiki for alpine wiki 2024-10-05 13:40:56 i'm very bad at documenting.. its my only problem at work as well :D 2024-10-05 13:41:38 I'm starting to make notes in Joplin because it can share notes easily :_) 2024-10-05 13:41:42 good documentation is hard! It's a damn art! Another reason I like Rust though is the documentation thoughtfulness in code 2024-10-05 13:42:11 documentation is impossible for me :D If i happen to die tomorrow, half the company has issues 2024-10-05 13:42:22 is joplin like an obsidian? 2024-10-05 13:42:22 And that, ofcourse, is never a good way to run a company 2024-10-05 13:42:26 yep 2024-10-05 13:42:44 ive decided to use joplin because you can easily share notebooks etc 2024-10-05 13:43:16 I always fight for treating documentation as code, and love the thought of technical writing as a great entry point into a company 2024-10-05 13:44:01 1) If someone new can understand it, someone else probably will too. 2) make sure it is tied to lifecycles and stays in sync with current state 2024-10-05 13:45:01 At the end of the day it is one of the biggest drivers of adoption, because how easily someone can pick something up determines how much time is spent to being effective with a product 2024-10-05 13:45:23 like watching a shopping cart haha. Where are you losing people in the checkout process? 2024-10-05 13:46:12 It's also one of those things that becomes really hard to change in the future, vs setting up processes early 2024-10-05 13:46:41 and since im not good at communicating with others.. imagine how i am with making notes for others :D 2024-10-05 13:46:58 But some people are excellent at that 2024-10-05 13:47:13 why I mention technical writers ahah. could also be internships 2024-10-05 13:47:43 everyone has what they are strong at. I can't wait to get to a point to lean on people where I am weak haha 2024-10-05 13:49:46 I'm very good at dreaming, creating visions :) 2024-10-05 13:49:55 too bad i dont have as much money as steve jobs had :D 2024-10-05 13:50:22 baby steps haha 2024-10-05 13:51:09 yeah :) i created kalipsolinux.org for my technical alpine documentation :D thats a first baby step :D 2024-10-05 13:51:31 I dk if you are up for a read: https://theandar1an.com/sustainable-engineering/ 2024-10-05 13:51:32 But I tried to share my thoughts on sustainable engineering here 2024-10-05 13:51:44 let me click on that :) 2024-10-05 13:51:59 thats your website? let me bookmark it then :) 2024-10-05 13:52:21 haven't written in a while, but I liked the feeling of that acronym. I also have jamminmusic.com, but I haven't updated them in a while 2024-10-05 13:52:42 been trying to take a little break and do more light configuration and tinkering to not burn out haha 2024-10-05 13:53:57 i have the same problem :( i wanted to start with Alpine 4 years ago. But a few months in it my father died, then my mother got really sick, so i havent had any time the past four years. Then last august my mother died, so only now do i get some time to breathe, and decide what to do again :) 2024-10-05 13:54:28 Ugh, I am sorry for your loss and all of that 2024-10-05 13:54:34 And so i decided to note down my "daily driver" problems. Hopefully others can learn and add their knowledge as well 2024-10-05 13:54:39 last time I tried alpine was alpine 2 haha 2024-10-05 13:55:03 I have a bunch of mine too. If you are gonna update wiki let me know. I will share my notes too 2024-10-05 13:55:19 I have specific thoughts around simplifying PAM 2024-10-05 13:55:28 well, part of my is glad my parents died. Ive been taking care of them since i was 12. so i really havent had much free time in my life 2024-10-05 13:55:38 Now i feel like i have all the time in the world :) 2024-10-05 13:56:05 well, im in the process of installing a forum on the earlier mentioned url :) 2024-10-05 13:57:12 Is it in Dutch? 2024-10-05 13:57:25 yeah im fixing that :) 2024-10-05 13:57:44 its no use if its dutch only, but i really only just this morning installed joomla on it :) 2024-10-05 13:58:04 fixed it :) 2024-10-05 13:59:00 yay i18n haha. I need to learn more langauges haha 2024-10-05 13:59:23 i only speak 3, and thats more then enough :) 2024-10-05 14:03:06 well, at least i enabled user registrations :D 2024-10-05 14:10:18 andar1an: i gotta leave! But its nice talking with you :) I'd love to continue our conversation later on :) 2024-10-05 14:10:47 Johannes Jacobs: Thank you for nice chat. I tried signing up, just waiting for verification email 2024-10-05 14:11:05 oh, i dont think thats coming yet! let me fix that reaslly quick :) 2024-10-05 14:11:13 Yeah i feel we have some things in common 2024-10-05 14:11:43 account activated :) 2024-10-05 14:12:07 bbl 2024-10-05 14:12:30 its still a WIP btw, so any comments or advise you have is welcomed :) 2024-10-05 14:13:38 I will DM you that 2024-10-05 14:29:08 why is /etc/init.d not included in lbu? pretty sure i could stop modloop earlier, but now i had to change a ln -s to cp -r the init.d/modloop, and lbu include it to umount the boot media 2024-10-05 16:24:21 is there a way to contact package maintainers? 2024-10-05 16:41:30 JohannesJacobs[m]: by email, or open a gitlab issue depending on the subject and if they have an account on gitlab? 2024-10-05 16:41:53 (gitlab.a.o) 2024-10-05 18:20:51 "Johannes Jacobs: by email, or..." <- Fair enough! the pkgs site only allows you to view all packages from a certain maintainer 2024-10-05 19:25:22 Made my first little wiki contribution just a moment ago, but was surprised how easy it was. Don't you ever get trolls deleting entire sections? I assume you can roll back whatever changes a user made and ban them for it though? 2024-10-05 19:27:45 isn't that usual thing on wikis? 2024-10-05 19:40:07 i doubt bans work since the cost of making new accounts == 0. it's mostly just reverting annoying changes, and hoping the antagonists find something better to do with their lives 2024-10-05 19:41:16 there's stuff i'd like to see added to the handbook, but it's unclear to me what the merge process is there 2024-10-05 19:44:19 Well, merging is not that hard, getting it deployed is the harder part 2024-10-05 19:45:21 what i meant was: who decides what can be added 2024-10-05 19:46:22 Probably ends up being me at the moment 2024-10-05 19:46:55 The scope of the handbook is things that are Alpine specific 2024-10-05 19:51:15 seems reasonable. how amenable are you to getting full disk encryption in there (just the 'setup-disk -e' way, not a thousand ways) 2024-10-05 19:51:37 because that stuff keeps coming up 2024-10-05 19:55:11 Sounds in scope 2024-10-05 19:56:10 maybe i'll send something your way soon 2024-10-05 23:41:50 hi, unsure if theres a better place to ask about specific packages, but i'll ask here before bothering upstream :), but i'm using keycloak (24.0.6-r0) from edge. and it just, quits after running after 1m30-2m. don't see an error message. just "Keycloak stopped in 0.047s". 2024-10-06 00:52:19 Hi folks, I’m playing with alpine installed on bare metal and then creating alpine kvm guests and playing with passing —filesystem foo,/foo and then trying to find it on the guest side. 2024-10-06 00:52:19 I’ve not played much with virtio-fs or alpine so I’m not sure if the tags are making it though because my mount doesn’t see it. 2024-10-06 01:04:27 "Hi folks, I’m playing with..." <- I'm doing something like this: 2024-10-06 01:04:27 --extra-args "console=ttyS0" --cloud-init meta-data=/tmp/cloud-init-metadata.yml,user-data=/mnt/misc/cloud/init//alpine-user-data.yml 2024-10-06 01:04:27 sudo virt-install --virt-type kvm --name eel --ram 2048 --vcpus 1 --disk path=/vm/eel_d0.qcow2,size=10,format=qcow2 --filesystem=/foo,foo,driver.type=virtiofs --memorybacking=source.type=memfd,access.mode=shared --network bridge=br0,mac=52:54:00:64:5b:29 --graphics none --noautoconsole --wait 20 --autostart --osinfo=alpinelinux3.19 --location /root/alpine-virt-3.20.3-x86_64.iso,kernel=boot/vmlinuz-virt,initrd=boot/initramfs-virt 2024-10-06 01:05:50 laird[m]: I see it in the xml... (full message at ) 2024-10-06 01:09:23 oh bother!!! i finally post in public and it just works 2024-10-06 01:31:24 laird[m]: that's a universal experience :P doesn't work until you actually bring someone to look at it 2024-10-06 05:00:36 Why does the initramfs copy the file in etc/apk/keys/? 2024-10-06 05:04:22 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 ) 2024-10-06 09:15:28 so after installing alpine extended iso onto my laptop, booting once than rebooting, network adapter "eth0" has disappeared from "ifconfig" & "ip link" list, as a result I just get an error when udhcpc tries to grab information. I can not see any information that indicates the adaptor is disabled in the UEFI settings, is there a simple way to find out if the issue is possibly firmware or 2024-10-06 09:15:28 software related, would like to rule out before assuming its hardware 2024-10-06 09:19:01 is it listed in `lspci | grep Ethernet`? 2024-10-06 09:47:09 pretty sure my client is echoing of which I apologise, seems I don't have lspci etc installed, I can't see anything in "dmesg | grep" that indicates network adapter, did find something that said "initializing netlink subsys (disabled)" though don't think thats related, I might just try again with USB and if that works maybe installing firmware update and those packages, though if you got 2024-10-06 09:47:09 anymore suggestions I'm open to those also 2024-10-06 09:48:14 do you have dmidecode? 2024-10-06 09:51:30 no fresh install, so haven't added any packages 2024-10-06 10:04:12 I'd say it was firmware related since it wasn't showing WiFi card either, see what happens this time round 2024-10-06 10:56:30 do you see those cards if you boot from live medium once again? 2024-10-06 14:09:35 When running abuild, how can I indicate that the CMakeLists.txt file is not in the src directory within the archive? It is one level up. 2024-10-06 14:32:27 andar1an i think you need to modify the APKBUILD 2024-10-06 14:33:40 I have created this APKBUILD, not sure where one indicates where the CMakeLists is in base archive 2024-10-06 14:35:55 andar1an[m]: abuild itself does not know about the file. You typically run cmake, which will look for it in the current directory 2024-10-06 14:36:20 So I should look up how to define that with cmake? TY 2024-10-06 14:36:21 andar1an[m]: abuild, if the APKBUILD is setup correctly, will set the current working directory to the root of the extracted archive 2024-10-06 14:36:40 andar1an[m]: Maybe share your APKBUILD file 2024-10-06 14:36:59 "$srcdir" should be the root of the archive already, no? 2024-10-06 14:37:11 $builddir 2024-10-06 14:37:20 $srcdir is the src/ dir next to the APKBUILD 2024-10-06 14:37:23 oh yea 2024-10-06 14:37:47 builddir defaults to "$srcdir/$pkgname-$pkgver" 2024-10-06 14:38:25 [I don't think it is setup correctly, because I see it does not remove the files, whereas with other packages, I see no src or archive remaining after build (https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/73087) 2024-10-06 14:38:44 andar1an[m]: it only removes pkg/ and src/ after a succesful build 2024-10-06 14:38:54 if the build fails, it will preserve them 2024-10-06 14:39:04 Ah ok, makes sense. 2024-10-06 14:39:47 I tried using abump yesterday for a different package, but that one needed more work, so I am trying to add what I thought would be a couple of simple ones haha 2024-10-06 14:42:52 builddir should be correct, you're using a standard github archive 2024-10-06 14:43:36 the build commands for cmake may not be.... (full message at ) 2024-10-06 14:43:56 and for buildtype put none instead of string, but I don't think that should matter 2024-10-06 14:44:08 but maybe it does 2024-10-06 14:45:38 I think I actually should use the CMakeList.txt as a reference for APKBUILD and maybe not that command 2024-10-06 14:49:38 Ya, I get: CMake Error: The source directory "/home/andar1an/repos/aports/testing/hypridle/src" does not appear to contain CMakeLists.txt. that is 1 level up from src 2024-10-06 14:50:06 (in the tar.gz) 2024-10-06 14:52:34 I can reference the hyprland build maybe. I was referencing the hyprcursor build, but hyprland seems to have a similar situation. It uses Meson thoguh 2024-10-06 14:52:49 s/thoguh/though/ 2024-10-06 14:53:55 is abuild in "infra" 2024-10-06 14:54:28 nvm, own repo 2024-10-06 15:18:10 how does one run a ruby sinatra webapp behind nginx on alpine, since uwsgi doesn't come with a rack plugin? what is the instead prefered way? 2024-10-06 15:26:03 bOR38552MJA: puma? 2024-10-06 15:35:54 is puma in alpine? apk search puma returns nothing 2024-10-06 16:27:49 So there’s an apparmor package, but not apparmor profiles and apparmor notify, theres an selinux libs, but nothing else.. do these frameworks even make sense on Alpine Linux? 2024-10-06 16:29:57 No, it requires quite some effort to build and maintain meaningful profiles. But the utilities can be useful 2024-10-06 16:31:12 bOR38552MJA: It can be installed as a gem though 2024-10-06 16:37:35 Do you mean no apparmor/selinux are not needed on Alpine? Or do you mean no we need it, but we have no maintainers? 2024-10-06 16:45:04 Hi there ! Bonjour tout le monde ! 2024-10-06 16:45:48 I hope you are all good 2024-10-06 16:46:50 I'm looking for a mindmap app on alpine linux. Any suggestion ? 2024-10-06 16:47:51 In that case i would install flatpak and look on flathub :) 2024-10-06 16:48:33 ive looked for that many times in the past, usually end up using gnumeric or sc-im, or just an editor with indentation 2024-10-06 16:49:45 or even inkscape 2024-10-06 16:50:13 non of them are specific for mind mapping tho 2024-10-06 16:53:44 Sob ! Thank you for your answers ! I'm afraid that we have to wait a bit longer. 2024-10-06 16:54:53 Therez nothing to find on flathub? 2024-10-06 16:55:32 it's not a gui but i've used graphviz a bit in that "mind mapping" way 2024-10-06 16:55:38 but then i'm very comfortable working with text files 2024-10-06 16:56:10 https://flathub.org/apps/com.github.phase1geo.minder looks good 2024-10-06 16:58:31 I have never tried flatpack, it works well for clean installing/uninstalling applications on alpine? 2024-10-06 16:58:57 Im very happy with it 2024-10-06 17:00:12 And it has flatseal for permissions, though i havent used it yet :) for me it “just works” 2024-10-06 17:01:12 I would prefer dia than graphviz if it was possible ! 2024-10-06 18:35:56 Does CMake need permissions other than what is on the wiki? 2024-10-06 18:35:58 >>> hyprlock: Unpacking /var/cache/distfiles/hyprlock-0.4.1.tar.gz... 2024-10-06 18:35:58 /usr/bin/abuild: line 28: cmake: Permission denied 2024-10-06 18:36:13 I keep getting this permission denied, no idea why. 2024-10-06 18:38:49 Can you show your latest APKBUILD? 2024-10-06 18:39:20 Yes, for this on it is https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/73090 2024-10-06 18:40:00 This is a different one, I wanted to see if this built differently. The project structure has same CMakeLists file in a similar location, but this one doesn't err the same way during build. I just dk why the cmake error 2024-10-06 18:40:08 maybe they are somehow related. I dk yet 2024-10-06 18:40:15 It says `not found` 2024-10-06 18:40:26 I don't see cmake in makedepends 2024-10-06 18:40:38 1 moment, dk why it is showing that 2024-10-06 18:41:03 there should be 13 makedepends 2024-10-06 18:41:15 https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/73090/diffs 2024-10-06 18:41:26 Yes, I see, but cmake is missing there 2024-10-06 18:41:36 cmake is not installed by default 2024-10-06 18:41:43 OH 2024-10-06 18:41:48 my bad, I had that in other one 2024-10-06 18:41:54 thank you ikke !! 2024-10-06 18:45:48 what deb library is opengl included in in alpine package keeper? 2024-10-06 18:45:57 s/deb/dev/ 2024-10-06 18:47:05 py3-opengl? 2024-10-06 18:47:54 andar1an[m]: should be mesa-dev 2024-10-06 18:47:54 or something 2024-10-06 18:49:02 I have mesa and mesa-dev, but still getting no opengl error 2024-10-06 18:49:16 andar1an[m]: show verbose logs please 2024-10-06 18:49:17 ignore me 2024-10-06 18:49:27 I replaced mesa-dev with mesa 2024-10-06 18:49:28 maybe I need both 2024-10-06 18:49:29 andar1an[m]: ? 2024-10-06 18:49:47 CMake Error at /usr/share/cmake/Modules/FindPackageHandleStandardArgs.cmake:230 (message):... (full message at ) 2024-10-06 18:51:10 hi, i'm having trouble getting alpine to configure sshd on boot: i set kernel cmdline ssh_key="ssh-ed25519 xxx", the key is correctly added to /root/.ssh/authorized_keys, but openssh-server isn't installed by initramfs (it should be according to tje source initramfs-init.in) 2024-10-06 18:51:11 andar1an[m]: > <@andar1an:matrix.org> CMake Error at /usr/share/cmake/Modules/FindPackageHandleStandardArgs.cmake:230 (message):... (full message at ) 2024-10-06 18:51:20 may be it, possibly 2024-10-06 18:51:32 glu-dev* 2024-10-06 18:51:37 should be it 2024-10-06 18:52:31 Thank you Melamit , I will give that a go if it isn't pulled in with mesa-dev 2024-10-06 18:52:44 hmm 2024-10-06 18:53:10 so which header file exactly it requires 2024-10-06 18:53:13 so we can search by it 2024-10-06 18:53:30 It just says opengl in the cmakelists.txt. So I am not sure 2024-10-06 18:53:44 i don't see opengl.h in the repos 2024-10-06 18:53:53 https://github.com/hyprwm/hyprlock/blob/main/CMakeLists.txt#L50 2024-10-06 18:54:07 gl.h 2024-10-06 18:54:37 i don't see it 2024-10-06 18:54:38 in the repos 2024-10-06 18:54:41 That cmakelist references egl and opengl, dk if they are related 2024-10-06 18:55:05 andar1an[m]: egl is gles, opengl is ogl 2024-10-06 18:56:23 ah ok 2024-10-06 18:56:41 both are referenced in this. If building for all archs I probably need both 2024-10-06 18:57:03 ACTION uploaded an image: (14KiB) < https://matrix.org/oftc/media/v1/media/download/AWfYuyU1P-UgzAWnp6YSAM-MLW--34K0VBjPiVuhRTadhGFz0MnYb4W7xQFRoY_8GVz3r_25gD8UVrBtqKe4THVCeSZOiu3QAG1hdHJpeC5vcmcvaGJFbU5KZ1BQWmNHQ3ZOWGVwdGVFZ0h0 > 2024-10-06 18:57:04 now i see it 2024-10-06 18:57:05 it's in the mesa-dev package 2024-10-06 18:57:36 I added, but after it told me dependencies that did exist before no longer existed haha 2024-10-06 18:57:54 CMake Error at /usr/share/cmake/Modules/FindPkgConfig.cmake:634 (message):... (full message at ) 2024-10-06 18:58:09 libmagic and hyprutils do exist 2024-10-06 18:58:09 remove opengl from pacakges list 2024-10-06 18:58:19 it's not in there 2024-10-06 18:58:29 hm 2024-10-06 18:58:32 strange thing 2024-10-06 18:58:45 that is probably from cmakelist in src code 2024-10-06 18:59:26 maybe when it doesn't find lib file error spits out what is missing but not based on distro 2024-10-06 19:09:23 Does one need to be on edge to build from edge? 2024-10-06 19:10:51 would it be best to set up an edge chroot for building? 2024-10-06 19:12:24 guess that is what abuild-rootbld is for 2024-10-06 19:13:17 yes, you need to be on edge :P 2024-10-06 19:15:38 cool, abuild-rootbld is awesome. Those warning make sense now 2024-10-06 19:15:52 whoever wrote this, thank you lol 2024-10-06 19:17:46 I also think there is probably some config so abuild finds these packages locally now instead of getting: WARNING: opening /home/andar1an/packages//main: No such file or directory 2024-10-06 19:18:16 but abuild-rootbld pulls from alpine cdn, so that seems better than relying on local 2024-10-06 19:18:35 abuild seems to need rebasing frequently 2024-10-06 19:18:47 (which is probably great haha) 2024-10-06 19:24:04 It's not necessary to keep your MR rebased, it will be rebased before being merged 2024-10-06 19:27:06 Yes, I noticed 2024-10-06 19:27:53 but that is also probably why it is nice to use abuild-rootbld, in case merges updated packages and my cache is out of date 2024-10-06 19:28:23 or mirror i guess depending. Will look into that config later 2024-10-06 19:28:52 still getting libmagic and opengl errors, so I will keep tinkering with combinations 2024-10-06 19:29:14 I dk why the libmagic error is happening, it is in main 2024-10-06 19:29:33 and there is only the 1 package, no -dev or alt 2024-10-06 19:43:18 file-dev provides libmagic.so 2024-10-06 19:43:32 and magic.h 2024-10-06 19:43:44 I tried searching for file-*, I must have had something messed up 2024-10-06 19:43:58 thanks Ikke. also should have seen the "net" option for chroot 2024-10-06 19:45:09 is it acceptable to comment these notes in the APKBUILD? 2024-10-06 19:47:26 Hi! What is the correct way to update kernels? I installed alpine on a diskless setup and it looks as the kernel is not automatically updated by 'apk upgrade' 2024-10-06 19:48:06 maybe there's no new kernel available in the repository? 2024-10-06 19:48:22 holgersson: traditionally you'd update the boot medium, but there is also an update-kernel script, it does require quite some memory though 2024-10-06 19:49:02 ikke: Oh, then I might have a weird setup without the boot medium 2024-10-06 19:49:13 Ermine: A newer kernel is available 2024-10-06 19:49:38 ikke: RAM should be no issue on that machine. I'll try it, thanks! 2024-10-06 19:59:55 Lol:... (full message at ) 2024-10-06 20:01:30 maybe because of libmagic it needs py3-opengl 2024-10-06 20:27:24 See… thats why ihavent gotten arround to learn packaging 😂 too hard! 2024-10-06 20:30:13 I don't think it is hard lol, just tedious. 2024-10-06 20:30:36 ACTION sent a code block: https://matrix.org/oftc/media/v1/media/download/AQGp5_X_VcSVHSvuruE3oDp8egDBN3PE58hntHGMQjhBFxyF1fG9gFSBebG3EehMZ4wWkyZwlD41-5rCz6MXMXdCeSZT5TQQAG1hdHJpeC5vcmcvdHNsbFlBUlVuSkZQTXVJckhvbWFBQUZy 2024-10-06 20:34:48 I am more used to Cargo too, and that can use llvm or gcc and cranelift is getting better which is exciting. I have enjoyed Cargo immensely. So easily to target different archs too 2024-10-06 20:37:16 :) 2024-10-06 20:44:57 This is quite literally why I like the thought of baremetal orchestration of deeply embedded systems. Most of my microservices are really nano services or faas like services. Most don't do complex things. Rather than virtualizing those devices I hope that a whole bunch of small, lowpower, devices to deploy that code to may mean less power consumption and less time troubleshooting. For what I need serious compute for, then I can use a 2024-10-06 20:44:57 bigger machine with a more capable os. 2024-10-06 22:23:11 local host login = ?? 2024-10-06 22:37:31 ? 2024-10-06 22:37:51 if for build, no, I am ssh'd in 2024-10-06 22:50:30 trying with... (full message at ) 2024-10-06 22:57:00 at least im onto the next error lol 2024-10-06 23:18:55 Oops, I think the CMAKE was not for all CMAKE defined variables. My bad 2024-10-07 11:40:47 What or who is responsible for updating docs? https://gitlab.alpinelinux.org/alpine/docs/user-handbook/-/issues/6 2024-10-07 12:23:18 DuClare: as per yesterday, ikke. but also mentioned pushing changes to the site is not what it should be 2024-10-07 12:24:39 2024-10-05 19:44:19 Well, merging is not that hard, getting it deployed is the harder part 2024-10-07 14:19:43 Is the default for Aports not set to squash on merge? Manually doing this is tedious as hell. There is a rebase button, why not squash an merge as default if that is what is expected? 2024-10-07 14:19:43 https://docs.gitlab.com/ee/user/project/merge_requests/squash_and_merge.html#set-default-squash-options-for-a-merge-request 2024-10-07 14:44:14 andar1an[m]: because we do not squash all commits 2024-10-07 14:45:41 It's quite common to have multiple useful commits in a single mr, and if there are multiple packages, each should be in a separate commit 2024-10-07 14:49:15 andar1an: you don't even need to squash your commit manually, if you just have a single commit. You could use `git commit --amend` to just amend it to the last commit. 2024-10-07 14:50:58 I was told I needed to squash, so that is what I was trying to do. Just closed to recreate because it is easier 2024-10-07 14:56:32 now build is failing. I will look at it next weekend. 2024-10-07 15:11:37 if you need to squash commits in a branch, i recommend to use git rebase --interactive. no need to create a new branch :) 2024-10-07 16:25:27 I tried, also soft resetting and I kept being required to rebase from branch 2024-10-07 16:26:09 The easiest way to do squashing is during merge. 2024-10-07 16:27:10 Dk about Gitlab, but at previous companies on Github we had a drop down button on merge to select squash and merge instead. 2024-10-07 16:29:16 Doesn't affect me, I can always not build for a package manager easily enough 2024-10-07 16:29:27 Was just trying to tinker and learn 2024-10-07 17:16:10 andar1an[m]: fyi, you do not have to constantly rebase when gitlab says so, you MR will be rebased just before merge 2024-10-07 17:16:31 You do need to (interactively) rebase to fixup commits 2024-10-07 17:17:15 it was blocking me pushing to branch without rebasing. Maybe had to do with me not waiting to click rebase until the end. 2024-10-07 17:19:28 Will note that if it is problematic to rebase before final iteration. 2024-10-07 18:20:58 Has been a weird git flow, so I am certain it is something variant in how I used. All in all I think the flow is easier than others I have seen. Trying to make notes if pain points in case they are relevant and valuable 2024-10-07 19:30:17 polybar is old 2024-10-07 19:30:23 edge version 2024-10-07 19:31:19 MRs welcome 2024-10-07 19:33:51 is maint. active? 2024-10-07 19:35:27 Not for 2 years 2024-10-07 19:41:47 reminds me, i should probably push i3bar-river one of these days. (which is not, despite its name, specific to river) 2024-10-07 19:47:53 somebody expert could look actual APKBUILD maybe 2024-10-07 19:54:10 hello, is there a way to build multiarch containers on alpine? multiarch/qemu-user-static doesn't seem to work anymore 2024-10-07 20:03:02 likelyonlysprinkler[m]: we build everything on native (or almost native) hardware 2024-10-08 08:53:16 with gnome with wayland, is there a way to stop it breaking randomly? after I open a flatpak and web browser or just walk away for 30 minutes I am no longer able to open new windows, the options to shutdown and reboot just log out etc when it's in this state, I can not access the terminal when this happens, I can swap to another TTY and fix the application launching logging out and back in, but the shutdown and reboot ends up proper cooked randomly 2024-10-08 08:53:36 don't have these issues with other two machines with KDE and in the past sway 2024-10-08 08:54:18 in saying that I have never had a good experience with gnome on any of my machines regardless of distro 2024-10-08 10:01:46 someone's got to debug it 2024-10-08 10:02:11 take a look at logs and see if anything crashes 2024-10-08 10:05:40 which logs will most likely have what I am looking for, always been bad at working out where and which logs to look at, do suspect its got something to do with OpenGL or Xwayland if it's same reason I had issues on fedora back in 2022, though won't know until I check, if you can help with paths or a link to a book or guide with best practices that is relevant I'd appriciate it as it is one of my weak points. 2024-10-08 10:27:11 silvia_: you can check /var/log/messages . gdm log is also somewhere in /var/log iirc 2024-10-08 11:14:12 maybe I'm just not understanding what I'm looking at, but why is the port in here not matching what I see in netstat: 2024-10-08 11:14:44 charm:~# grep ^address: /proc/fs/nfsd/clients/10/info 2024-10-08 11:14:44 charm:~# netstat -na | grep 192.168.93.218 2024-10-08 11:14:44 address: "192.168.93.218:676" 2024-10-08 11:14:45 tcp 0 0 192.168.93.82:2049 192.168.93.218:823 ESTABLISHED 2024-10-08 11:15:10 the supposed port in the info is 676, but the port shown in netstat is 823... 2024-10-08 12:01:20 hello 2024-10-08 12:01:52 actually, I'm gonna most in -dev room, sorry for the noise 2024-10-08 12:01:56 post* 2024-10-08 12:31:14 I got a system service which binds to an IP, however it fails since "network isn't ready" even with "depend() { need net }". What do I need to add there or can I make the service restart automatically after failure? 2024-10-08 16:04:36 i tried to update polybar looks fine but it is stuck when making completion. i copied the bin 2024-10-08 16:04:52 is ok half-tested MR? 2024-10-08 17:33:29 hello, I'm on alpine 3.19, ready to upgrade to 3.20, changed the repositories but apk keeps complaining about icu libs conflicting: https://tpaste.us/bQgM 2024-10-08 17:33:33 can anyone lend me a hand 2024-10-08 17:33:49 i'm not really sure how to proceed now 2024-10-08 17:34:51 i can't remove icu, as far as i know, and i don't see any qt stuff on world 2024-10-08 17:35:15 i don't see any warning or migration steps on the 3.20 release either 2024-10-08 17:35:21 eletrotupi: what does apk list --orphan return? 2024-10-08 17:38:16 libxkbfile-doc-1.1.3-r0 x86_64 {libxkbfile} (MIT) [installed] senpai-0.2.0-r3 x86_64 {senpai} (ISC) [installed] swww-0.8.1-r0 x86_64 {swww} (GPL-3.0-or-later) [installed] 2024-10-08 17:44:55 Did you remove the community repo? 2024-10-08 17:57:04 lemme check 2024-10-08 17:57:11 nops 2024-10-08 17:57:17 have only main and community repos 2024-10-08 19:21:59 eletrotupi: I'm puzzled that it returns senpai and libxkbfile, both which should still be available 2024-10-09 00:50:54 hello, is this kind of normal ? some symbol not found for python library. Image link : https://gitlab.com/lidgl/kumpulan_log/-/raw/main/2024-10-09T07_41_29_185689915+07_00.png 2024-10-09 16:48:25 Hello, I'm confused on the allocator packages on Alpine linux repo. Do they only work when musl is dynamically linked ? 2024-10-09 16:48:50 like jemalloc? 2024-10-09 16:48:58 Yes 2024-10-09 16:49:15 Yes, they depend on overriding things through the dynamic loader 2024-10-09 16:50:38 I see. thx you 2024-10-09 17:49:00 is omni here? simply updating polybar version worked. i ketp xpp the same cause error otherwise. i did not bother to try updatin i3 thingy 2024-10-09 18:05:32 i asked the dev. will try again if necessary 2024-10-09 19:06:43 First time using Alpine. Love it so far. 2024-10-09 20:31:43 if wanting to install and run alpine pkgs on mobile, here are some helper scripts, https://git.insteps.net/mess/mobile/ 2024-10-09 20:31:51 its still experimental, does not say how to get them together yet, but will add the docs soon 2024-10-09 20:33:24 been using mobiles to run my php dev/server (alpinelinux) on them for couple of years now 2024-10-10 01:05:47 eletrotupi: do you have any locally built packages that might need to be rebuilt? 2024-10-10 06:57:36 Is anyone using encrypted swap on alpine linux? I'd like to enable encrypted swap with a random key, basically reverting / overwriting on each boot 2024-10-10 06:57:49 It seems we don't have a crypttab 2024-10-10 06:58:15 And the only related article I found was https://wiki.alpinelinux.org/wiki/Setting_up_encrypted_volumes_with_LUKS 2024-10-10 06:59:13 ah, we have /etc/init.d/dmcrypt 2024-10-10 07:01:56 HOLY swearword-of-your-choice 2024-10-10 07:01:58 WHY IS THAT SO EASY 2024-10-10 07:02:01 IT CANNOT BE 2024-10-10 07:02:17 Reading the init script, adding config to /etc/conf.d/dmcrypt, running swapon 2024-10-10 07:02:19 damn 2024-10-10 07:02:22 that was so freaking easy 2024-10-10 07:02:45 within minutes! swap: 4095 0 4095 2024-10-10 07:22:10 telmich: naise 2024-10-10 07:22:24 SO NICE 2024-10-10 07:23:59 Just revived my old blog, because it was so easy, so nice: https://www.nico.schottelius.org/blog/ 2024-10-10 07:24:00 :D 2024-10-10 07:26:52 I mean, seriously... there hasn't been any other distro out there for years where things are so straight forward and logical and that make sense 2024-10-10 07:27:06 I think we got to celebrate this more 2024-10-10 07:39:18 Next step, I need to move the blog into k8s... this is one of the legacy things running on an old vm 2024-10-10 08:41:17 can anyone share a working polkit setup with sway/seatd? I have dbus working well but trying to use polkit-gnome (or lxpolkit) gives an "Unable to determine the session we are in: No session for pid 16483" error 2024-10-10 09:23:47 telmich: the blog missed on how to install the file /etc/init.d/dmcrypt 2024-10-10 10:27:09 "nico🇨🇭: the blog missed on how..." <- Thanks, will add a note for it 2024-10-10 10:27:33 telmich: if you post it on mastodon, I can boost it 2024-10-10 18:55:34 are the nerd-fonts checked for if they contain all fonts? some fonts such as iosevka are missing 2024-10-10 18:57:05 sewn: is there a reason iosevka needs to be part of the nerd-fonts package, it has its own package currently 2024-10-10 18:57:43 nerd-fonts iosevka 2024-10-10 18:58:16 like, nerd-fonts have compilations of individual fonts containing the nerd font symbols 2024-10-10 18:58:20 one of them that they distribute is iosevka 2024-10-10 19:01:39 34 fonts are missing 2024-10-10 19:01:56 im not sure if alpine is to be expected to distribute all these fonts, but nerd fonts do distribute them 2024-10-10 19:57:45 Is alpine affected by this? https://www.pcworld.com/article/2485961/update-firefox-131-now-to-patch-a-critical-zero-day-security-flaw.html 2024-10-10 19:58:44 jonesv: Yes 2024-10-10 19:58:48 and is there something I can do to help get Firefox updated? Like test a package or something? 2024-10-10 19:58:59 jonesv: it has already been pushed 2024-10-10 19:59:17 ikke: hmm I `apk update` and I don't see an update :/ 2024-10-10 19:59:44 ikke: I still have 130.0.1-r0 2024-10-10 20:00:54 I guess you are on 3.20? 2024-10-10 20:01:12 I am... it's the latest stable, right? 2024-10-10 20:01:18 yes 2024-10-10 20:02:17 Won't it get an update? 2024-10-10 20:03:43 (not complaining, just trying to understand how it works with critical updates :-) ) 2024-10-10 20:05:14 I'm checking, hold on 2024-10-10 20:07:04 She be a manual backport for security right? 2024-10-10 20:07:36 yes, it should have been pushed, was just forgotten (the 3.20 backport for firefox, firefox-esr has been updated) 2024-10-10 20:08:37 "within minutes! swap:..." <- was stupidly easy with btrfs too. Don't even need to do chattr steps urself anymore 2024-10-10 20:11:14 ikke:... (full message at ) 2024-10-10 20:11:40 fyi, these multi-line messages are anoying for irc users= 2024-10-10 20:12:29 But yes, you don't have to explciitly do something, rebasing will be done before merging 2024-10-10 20:12:38 Ah, sorry. What would be best way to send multiline messages? 2024-10-10 20:14:24 just break it up into a bunch? 2024-10-10 20:14:32 Generally you keep them single line. If you have multi-line content (like the output of a program), you can use a paste service 2024-10-10 20:14:54 The matrix bridge turns into a "full message at " 2024-10-10 20:20:04 Cool, thanks. 2024-10-10 20:23:04 What should I expect the typical time to get a complete testing build into aports? Think I will be able to try these this weekend? I have been waiting to install a desktop env so I could test these and see what hype is about. If it can take some time I will just install another one for now 2024-10-10 20:23:58 MRs for new aports take a bit longer than package upgrades 2024-10-10 20:25:23 kk, thanks 2024-10-10 20:31:01 ikke: oh great. I'll check in a couple hour if the updated firefox is available then! 2024-10-10 20:31:25 ikke: out of curiosity, can I see it somewhere? Like how did you see that it had been forgotten? So that next time I can check there before bothering you :) 2024-10-10 20:31:36 jonesv: I asked the maintainer :) 2024-10-10 20:31:49 oh :) 2024-10-10 21:40:17 Hi, is it possible for 2 alpine diskless installs to coexist? I have a raspberry pi that I'm building a small handheld appliance with, but when I try to boot it from USB, it tries to mount the modloop on the SD card and it boots up in a broken state. There isn't a practical way for the people who will be using these devices to remove the SD card from them, otherwise I would just do that. Thanks! 2024-10-11 05:48:28 somewhy wlroots0.17 package is empty on my side 2024-10-11 05:49:30 and sway lacks libwlroots.so.12 2024-10-11 05:50:59 Maybe a broken replaces? 2024-10-11 05:51:12 does `apk fix wlroots` help? 2024-10-11 06:29:29 ikke: no 2024-10-11 06:31:28 ikke: apk fix wlroots0.17 helped though 2024-10-11 07:06:55 Ok, right 2024-10-11 10:48:57 is there any chance of latest firefox patches to be backported to 3.20? Seems there is a critical security issue: https://lwn.net/Articles/993608/ 2024-10-11 10:49:25 ptrc: ^ 2024-10-11 10:49:37 fyi, there is also an about:config option to mitigate it 2024-10-11 10:50:38 oh neat, thank you. Which option is that? 2024-10-11 10:51:39 dom.animations-api.timelines.enabled=false 2024-10-11 10:52:50 https://hachyderm.io/@dalias/113284901186843302 2024-10-11 10:54:31 interesting - that's not an option that's present by default in about:config for me, but I will try it all the same. 2024-10-11 10:55:09 Neither for me 2024-10-11 11:02:40 I wonder if packaging 5 different versions of firefox is worth the effort. 2024-10-11 11:29:02 5 different? 2024-10-11 11:29:12 wouldnt it be enough with the latest + the LTS version ? 2024-10-11 11:45:54 Latest and esr on both edge and latest stable release, and developer edition on edge. 2024-10-11 12:17:13 wait, but 3.20-stable has latest esr and stable 2024-10-11 12:17:16 at least in git 2024-10-11 12:17:26 but pkgs.a.o returns an older version 2024-10-11 12:17:48 yeah, no, it's on the mirrors as well 2024-10-11 12:18:50 built 2024-10-10T21:12:12.000Z 2024-10-11 12:21:19 https://ptrc.gay/jktkEekw 2024-10-11 12:23:28 elagost: could you double-check? 2024-10-11 13:06:25 Does anyone know if there is an option during setup to use dhcpcd instead of uhcpc? 2024-10-11 13:06:25 Haven't been able to dig into why yet, but dhcpcd seems to be optimal choice for ipv6 anyways? 2024-10-11 13:06:25 This server seems to have trouble with the latter after installing a desktop weirdly. 2024-10-11 13:06:46 Does anyone know if there is an option during setup to use dhcpcd instead of udhcpc? 2024-10-11 13:06:46 Haven't been able to dig into why yet, but dhcpcd seems to be optimal choice for ipv6 anyways? 2024-10-11 13:06:46 This server seems to have trouble with the latter after installing a desktop weirdly. 2024-10-11 13:12:39 I don't think so, but I am reading setup-interfaces from phone because I stupidly tried to update firefox on unstable debian yesterday and lost work back to oct 3 lol. Shoulda backed up first lol 2024-10-11 13:12:57 At least all machines are alpine now lol 2024-10-11 13:33:55 Just to be clear. I am asking about installing dhcpcd during alpine installation, so that I don't need to cache apk packaged because udhcpc can't aquire addresses after boot post setup-desktop. I can try to install it before running setup-desktop, or cache it. But I am specifically asking if the setup-interfaces script or any setup-* script will install dhcpcd for use during installation. 2024-10-11 13:35:24 I imagine I can use an apkoverlay or cache, but I don't want to do this post reboot after setup-disk 2024-10-11 13:42:56 have you tried stracing udhcpc after setup-desktop? sounds like an interesting issue to figure out. 2024-10-11 13:49:48 andar1an[m]: just so you are aware, when you edit from the matrix side, irc sees each edit as a new line. 2024-10-11 13:49:58 Yes, i didnt have this issue on my laptop. I have also tried to statically set address, gateway etc. and a few other things. It happens for both kde and sway. I believe setup-alpine brings in linux-firmware metapackage and address is aquired pre-setup desktop 2024-10-11 13:50:37 I made a mistake on that @invoked, I saw it get duplicated in matrix client post edit 2024-10-11 13:51:39 Whatever the issue, it does appear that I can reference an apkovl file during setup disk which is great 2024-10-11 13:51:51 no worries. generally acceptable irc convention is to s// or something 2024-10-11 13:51:57 like tpyo 2024-10-11 13:52:00 s/tpyo/typo/ 2024-10-11 13:52:16 Oh cool, thank you. 2024-10-11 13:53:15 As a note to issue it seems to affect both eth and wlan interfaces. 2024-10-11 14:14:47 ptrc: I was checking pkgs.a.o, my bad. Updating my stable machine now and you're right; it's latest. Thanks! 2024-10-11 14:15:03 glad to hear that, and you're welcome ^^ 2024-10-11 16:43:30 Hey, pretty new to alpine and ran into an issue today. I wanted to compile a package using abuild rootbld, but it did not work, failing with an error on line 2543 in the abuild script, stating "/home/username/packages: permission denied". So I looked into that line in the script, (here: https://gitlab.alpinelinux.org/alpine/abuild/-/blob/master/abuild.in?ref_type=heads#L2543) and it looks like there is just a backslash missing in 2024-10-11 16:43:32 the previous line, causing the line not to be an argument to mkdir. Is that assertion correct, or am I just doing something wrong? 2024-10-11 16:44:40 Guest6063: Yeah, it has been reported already 2024-10-11 16:44:47 abuild has just been updated (to an rc) 2024-10-11 16:46:10 I see, thanks for the info 2024-10-11 18:04:47 hi 2024-10-11 18:53:44 What a derp moment this morning. Just install in chroot. 😹 2024-10-11 22:36:02 Ugh, wish busybox mount could mount by label or uuid 2024-10-11 23:34:05 Greetings! Total noob here... 2024-10-11 23:34:06 Trying to run `abuild rootbld` but getting errors all the way. Is there a descent doc on it? or a guide how to set up its env? 2024-10-11 23:34:43 if its a permission error, i guess its being worked on 2024-10-11 23:35:41 No. It looks totally unusable, I can't overcome it 2024-10-11 23:36:11 Like: 2024-10-11 23:36:11 > envsubst: error while reading "standard input": Is a directory 2024-10-11 23:38:10 hmm maybe its a issue with abuild; it was upgraded to a rc couple of days ago 2024-10-11 23:38:37 abuild works fine for me. 2024-10-11 23:38:54 * abuild itself works fine 2024-10-11 23:39:50 I'm on v3.20 2024-10-11 23:40:57 Do all build with plain abuild? Or is there an alternative chroot builder? 2024-10-12 09:12:01 ildar[m]: there is dabuild which is based docker/podman? but your error with rootbld is strange 2024-10-12 09:12:46 I'll look at it, thanks 2024-10-12 09:13:44 what's the normal way to debug rootbld besides source reading and gdb? 😉 2024-10-12 09:18:08 it's only shell script 2024-10-12 09:18:43 ah! ok. Didn't have time to look 2024-10-12 09:19:44 it only calls envsusb at 1 line 2024-10-12 09:21:21 https://gitlab.alpinelinux.org/alpine/abuild/-/blob/master/abuild.in?ref_type=heads#L2573 2024-10-12 09:23:50 could you confirm that it fails there? 2024-10-12 09:25:14 Need some little time, thanks 2024-10-12 09:25:50 just try 2024-10-12 09:25:56 abuild -v rootbld 2024-10-12 09:31:19 lol, it seems broken also for me, but different problem 2024-10-12 09:33:36 https://paste.rs/gxKkT 2024-10-12 09:40:48 hmm... I think that it fails to create /etc/apk/repositories inside the chroot 2024-10-12 09:41:28 I would like to compare with my log but mine fails before :D 2024-10-12 09:42:34 that's .. strange 2024-10-12 09:43:17 I have some weird permission error with /home/donoban/packages 2024-10-12 09:47:31 I see it in *devel* 2024-10-12 09:58:37 my problem is kinda obvious: .rootbld-repositories was wrong 2024-10-12 09:58:53 I did mkdir that, but it should be a file 2024-10-12 09:59:20 though there's no doc on it 😒 2024-10-12 10:04:19 :9 2024-10-12 10:04:21 :) 2024-10-12 10:05:26 one little problem left 2024-10-12 10:07:37 https://paste.rs/gGiZE 2024-10-12 10:07:47 this one is trickier 2024-10-12 10:19:09 is there an APKBUILD file on your workdir? 2024-10-12 10:21:11 Otherwise it doesn't start building 😁 2024-10-12 10:22:43 hm.. 2024-10-12 10:25:02 do you run it on /tmp/alien? I suppose that most people uses it inside '~/aports/xxx' so maybe it's messing something 2024-10-12 10:29:16 Yep.. 2024-10-12 10:44:25 yes, that was right, building from `~/aports` did it 2024-10-12 10:44:46 my issue solved 2024-10-12 10:51:39 thanks for help 2024-10-12 10:52:20 :) 2024-10-12 10:52:28 maybe it should be an issue 2024-10-12 15:08:43 Hi, I wanna poll the status of !72313, is jirutka around here? 2024-10-12 15:13:00 famfo: jirutka is not 2024-10-12 15:13:27 https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/72313 2024-10-12 16:01:57 is abuild broken 2024-10-12 16:01:59 ? 2024-10-12 16:03:58 j_v: rootbld? 2024-10-12 16:04:49 yeah 2024-10-12 16:05:44 i'm looking through commits trying to get a clue where to look what happened 2024-10-12 16:07:39 i'll have to git bisect i think 2024-10-12 16:09:23 It's a known issue 2024-10-12 16:09:50 "you can workaround it locally if you add `\` to /usr/bin/abuild line 2542 " 2024-10-12 16:10:13 oh, cool, thanks! 2024-10-12 16:11:16 ah, right 2024-10-12 16:11:26 just a typo 2024-10-12 18:15:08 Is there a way to obtain up to date postgresql packages for alpine linux ? All postgresql versions are outdated of several month on the edge repository 2024-10-12 18:15:45 There is an MR to update them, but it fails to build. 2024-10-12 18:16:16 !72200 2024-10-12 18:16:21 https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/72200/ 2024-10-12 18:16:26 Oh ok I can see that. Thx 2024-10-12 18:17:02 Apparently test failures 2024-10-12 19:15:05 should this also work for alpine? https://openwrt.org/toh/arcadyan/vrv9510kwac23#installing_openwrt_-_brnboot_recovery_web_interface_easy 2024-10-12 19:15:35 it seems that it passes an openwrt initramfs via tftp 2024-10-12 19:15:44 "Set up a TFTP Server on your PC (you can use Tftpd if you are using Windows, or tftpd-hpa if you are in Linux). Rename the OpenWrt initramfs image to rescue.bin and place it in your TFTP directory." 2024-10-12 19:17:27 openwrt is boring, alpine would be nice :) 2024-10-12 19:24:43 I'm not sure of the architecture of this device 2024-10-12 19:25:03 probably MIPS 2024-10-13 10:40:51 anyone running your own personal apk repo, i.e. for packages not in alpine's repos? if so would you mind testing `apk cache download ` with a package from your repo, and seeing if it works without complaint? for some reason doing this with postmarketos' repos fails signature checks. 2024-10-13 11:31:12 do you have the keys installed in /etc/apk/keys? 2024-10-13 13:01:23 yeah, the postmarketos keys are stored there. `apk add` from the repos works great. caching just didn't work for pmos. 2024-10-13 16:41:32 Hello 2024-10-13 16:42:18 I'm newbe to alpine, can someone point me to a book to learn alpine ? 2024-10-13 16:42:22 thank you 2024-10-13 16:43:37 hi, I'd suggest Alpine wiki: https://wiki.alpinelinux.org/wiki/Main_Page 2024-10-13 16:45:09 Thank you! 2024-10-13 16:45:38 I've seek for a printable or pdf book about alpine but did not found 2024-10-13 16:46:27 afaik there is no such thing 2024-10-13 16:48:33 thank you. 2024-10-13 16:49:15 yw 2024-10-13 17:40:37 Hi, I'm trying to install alpine on a btrfs encrypted drive. After entering the initramfs and decrypting the drive, it crashes with Mounting boot media: failed. I can manually mount the drive with mount /dev/mapper/main -o subvol=root /sysroot, and then continue booting normally. I thought I maybe have a wrong kernel cmdline, I tried to tweak it in many ways, unsuccessfully. 2024-10-13 17:41:35 Here is the cmdline: "root=UUID= ro rootflags=subvol=root cryptroot= cryptdm=main modules=sd-mod,btrfs,nvme quiet rootfstype=btrfs" 2024-10-13 17:44:31 Tried changing to subvol=/root, tried making my root subvolume default and removing the subvol options 2024-10-13 18:42:08 What does your /etc/fstab look like? 2024-10-13 21:00:50 And what do you subvolumes look like 2024-10-13 21:01:59 If you are separating /etc, /usr there are some files that are necessary during boot you won't be able to access without a lot of tedium 2024-10-13 21:04:09 Im working on raid10 with usb key unlock now for btrfs with subvols, and I want to punch GRUB in the face lol. 2024-10-13 21:04:10 It is a nightmare trying to use the cryptomount -k option because of assumptions in grub-install.c. 2024-10-13 21:04:10 Add a flag to provide our own load.cfg if your gonna do that. Holy cow. 2024-10-13 21:08:53 "Hi, I'm trying to install alpine..." <- https://wiki.alpinelinux.org/wiki/User:Andar1an 2024-10-13 21:09:55 Im sorry this is rough formatting, if you copy and paste this into a markdown doc it will be more clear. I will hopefully make it prettier in future lol 2024-10-13 21:17:44 If you are getting a remount warning on shutdown, you can tweak /etc/init.d/mount-ro 2024-10-13 21:18:53 What is in that wiki is a bit out of date, but that works. Just may need some cleanup and i have a lot of new learning from this raid stuff to go back into there 2024-10-13 23:45:24 Hi chat!. Im trying to build my first package for alpine using this wiki page https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package#Building_packages 2024-10-13 23:45:54 and getting >>> ERROR: amneziawg-tools: rootbld: /home/krab/myapks/amneziawg-tools/../.rootbld-repositories does not exist 2024-10-13 23:46:10 what on earth does this mean 2024-10-13 23:54:13 Did you install abuild-rootbld? 2024-10-13 23:57:31 Are you in the directory for what you are trying to build? 2024-10-14 00:15:36 yes and yes? https://pastebin.com/raw/bYRJuguL 2024-10-14 00:27:32 Are you trying to use cached apks? 2024-10-14 00:30:47 What does your apk repository file look like? 2024-10-14 00:30:47 I got warnings with abuild as it seems to look for local packages first, but rootbld for me pulls from remotes, I didn't tweak anything. 2024-10-14 00:30:47 Never seen that before, sorry 2024-10-14 00:37:09 Probably has to do with you aports structure 2024-10-14 00:37:14 https://gitlab.alpinelinux.org/alpine/abuild/-/issues/10041 2024-10-14 00:37:30 https://wiki.alpinelinux.org/wiki/Include:AbuildRootBld 2024-10-14 01:32:48 yeah new package must be inside aports directory now it works 2024-10-14 02:28:29 Anyone get XFS as the root partition for alpine on arm? I can't figure out how to do it. Cfdisk doesn't show it as an option. I've installed from the VM image 2024-10-14 03:55:14 do you guys have any example for any alpine kernel module? package or APKBUILD? i need amneziawg module as separate package 2024-10-14 03:56:24 Can't you just look on the Alpine package list website? 2024-10-14 03:57:27 https://gitlab.com/postmarketOS/pmaports/-/blob/7df2c93459424af257d673f055e5d86c6e6db46c/main/wireguard-postmarketos-stable/APKBUILD 2024-10-14 03:57:49 thats one just does copy ko file and nothing more 2024-10-14 04:56:15 alpine has package with entire kernel sources? 2024-10-14 05:33:04 is there a way to disable cleaning build chroot on failure? >>> amneziawg-linux-virt: Cleaning up build chroot 2024-10-14 05:42:53 https://pastebin.com/raw/ppiBTFL3 very nice. not sure if i need DKMS (AKMS) for that module 2024-10-14 05:45:33 or package will strictly depend on linux package ver-rel-flavor 2024-10-14 05:47:12 install_if="amneziawg-tools=$_toolsver linux-$_flavor=$_kpkgver" 2024-10-14 09:07:29 Good morning o/ 2024-10-14 09:08:39 I'm trying to compile BambuStudio on Alpine that uses cmake. after installing much deps, it complains about not being able to find wxwidgets libs while obviously wxwidgets-dev is properly installed 2024-10-14 09:08:59 CMake Error at /usr/share/cmake/Modules/FindPackageHandleStandardArgs.cmake:233 (message): 2024-10-14 09:09:07 Could NOT find wxWidgets (missing: wxWidgets_LIBRARIES 2024-10-14 09:09:09 wxWidgets_INCLUDE_DIRS) (Required is at least version "3.1") 2024-10-14 09:09:31 I've tried different variations of those variable to no avail 2024-10-14 09:09:35 Any idea?? 2024-10-14 09:35:27 err... redid from scratch and it now goes on 2024-10-14 09:35:35 Don't mind me then ^^ 2024-10-14 14:27:47 hi everyone! i am trying to find a way to use Openmediavault alpine based, do you think it is doable? 2024-10-14 14:32:49 it's debian based, so not likely 2024-10-14 15:14:08 Does anyone know if rEFInd supports cryptomount 2024-10-14 17:36:41 hi 2024-10-14 17:37:14 there 2024-10-14 19:23:02 mmmmmm 2024-10-14 19:24:06 why cant change nickname? 2024-10-14 19:27:13 Possibly due to the +R mode on this channel 2024-10-14 19:27:19 Or some other channel 2024-10-14 19:27:35 Nope, no +R here 2024-10-14 19:45:39 hi there ikke 2024-10-14 19:46:03 021804 https://pixshare.de/qHDaOw 2024-10-14 22:00:33 Yum. Beer haha 2024-10-15 01:55:37 so ive build my apk for amneziawg kernel module (it contains single file /lib/modules/6.6.56-0-virt/extra/amneziawg.ko) but after installation it won't load: modprobe: FATAL: could not insert 'amneziawg': Invalid argument 2024-10-15 03:56:24 I'm guessing the answer is no one has tried 2024-10-15 04:22:31 it looks like i have to do make modules_install ... and can't just copy ko file 2024-10-15 04:29:27 is there a way to stop abuild doing that >>> amneziawg-linux-virt: Cleaning up build chroot ? 2024-10-15 04:46:10 abuild.conf 2024-10-15 05:42:52 pls provide package linux sources out-of-box, like deb-src in debian 2024-10-15 08:42:52 how to fix: "error in libcrypto" while ssh? 2024-10-15 08:44:52 templecloud: any log? 2024-10-15 08:46:44 let me try 2024-10-15 08:47:25 "error in libcrypto" 2024-10-15 08:47:29 only that 2024-10-15 08:47:37 error in libcrypto debug1: No more authentication methods to try. 2024-10-15 08:48:07 use a compatible key 2024-10-15 09:50:20 >>> mkdir: can't create directory '/lib/modules/6.6.56-0-virt/updates/': Read-only file system 2024-10-15 09:51:17 what tries to create directory make or depmod and why it doent respect destdir ... 2024-10-15 10:32:07 >>> /bin/depmod -v -b "/home/krab/aports/testing/amneziawg-linux-virt/pkg/amneziawg-linux-virt" -a 6.6.56-0-virt -o "/home/krab/aports/testing/amneziawg-linux-virt/pkg/amneziawg-linux-virt" 2024-10-15 10:32:23 >>> mkdir: can't create directory '/lib/modules/6.6.56-0-virt/updates/': Read-only file system 2024-10-15 10:32:25 ??? 2024-10-15 10:32:38 krab4t: are you using akms? 2024-10-15 10:33:34 akms would require package with entire kernel sources, their Makefile applies patches on drivers/net/wireguard and then build 2024-10-15 10:41:29 can you bwrap into abuild chroot? 2024-10-15 10:41:50 i got entire command line from abuild -v rootbld 2024-10-15 10:42:02 bash: cannot set terminal process group (-1): Not a tty 2024-10-15 10:44:53 You could look at the abuild source to see how it invokes bwrap 2024-10-15 10:48:15 ive got command from abuild -v rootbld log and then /bin/sh at the end 2024-10-15 10:48:26 it seems working 2024-10-15 10:49:35 ah its not depmod its actually make modules_install ... 2024-10-15 16:26:02 have i been blessed OK: 7777 MiB in 1033 packages 2024-10-15 16:27:07 crushed by demonic bloat 2024-10-15 19:51:53 Hello. I successfully booted an install of alpine using full disk encryption and btrfs. I have an error message from OpenRC though, failed to start modloop. I couldn't find much info on what modloop is, from what I understand it's a file that keeps your kernel on disk so it doesn't much RAM, not sure 2024-10-15 19:53:43 Sorry, typing on my phone. [...] so it doesn't consume much RAM. Not sure what's the use since my system seems to be functioning without. Should I remove the service from my init chain, or have I done something wrong ? 2024-10-15 19:54:21 leforjeron: it's for a run-from-ram system where you do not install ... 2024-10-15 20:27:30 So in my experimentation with raid10 btrfs + LUK2, I feel like I am at a crossroads haha. 2024-10-15 20:29:25 Grub-install does not make it possible to enable reference to keyfile in cryptomount through flags. 2024-10-15 20:29:26 Now playing with just making boot a usb and unlocking disks by editing initramfs-init lol 2024-10-15 20:29:26 Grub-mkimage embedded config can but I haven't been able to make it work yet. 2024-10-15 20:30:42 The hope was initially to just have keyfile for initial stage unlock with efi, and I could just yank usb. I guess similar may work with entire boot on usb. Fun to tinker with, but my eyes are going cross-eyed lol 2024-10-15 20:32:54 I know the cryptomount works manually through grub shell, just doesn't seem to in embedded config prior to directing to boot. Gotta see if there is an easier way to get feedback. Challenging to see what is erring 2024-10-15 20:34:55 I tried adding a cryptdisks and cryptdiskskey option in the initramfs-init file, that created some different output which was neat. Just need to figure out what avenue to plug away at still haha 2024-10-15 20:37:44 At least with initramfs avenue I get to try rEFInd. Perks i guess haha 2024-10-15 21:47:26 https://pastebin.com/raw/8m7NdRRP FFS 2024-10-15 21:47:52 what the difference between 6.6.56-0-virt #1-Alpine SMP PREEMPT_DYNAMIC 2024-10-10 17:19:03 x86_64 Linux on my VPS and 2024-10-15 21:48:05 6.6.56-0-virt #1-Alpine SMP PREEMPT_DYNAMIC 2024-10-10 17:19:03 x86_64 Linux in my virtualbox 2024-10-15 21:51:50 i can't do any make install modules_install inside abuild because it will or use /lib/modules or INSTALL_MOD_PATH="$pkgdir" and then create all depmod result files like modules.dep inside $pkgdir 2024-10-15 21:52:50 so the only way is to run make on module sources, gzip it, install -v -D -m644 it and let apk run kmod triggers after apk add pkgname ... 2024-10-15 21:53:10 gzip amneziawg.ko file ** 2024-10-15 21:55:08 but this amneziawg.ko.gz on same kernel won't load ... a lot of errors in dmesg 2024-10-15 23:31:16 regular build on host works just fine, but inside abuild produces different ko file 2024-10-15 23:32:10 modprobe --dump-modversions /path/to/amneziawg.ko(.gz) show some symbols are different 2024-10-16 00:09:19 e1000e 0000:00:1f.6 eth0: (PCI Express:2.5GT/s:Width x1) 2024-10-16 00:09:19 dmesg reports an ethernet card on my laptop, but I don't see any ethernet plug 2024-10-16 00:16:17 which laptop? 2024-10-16 00:56:30 lenovo t14s 2024-10-16 02:38:12 test 2024-10-16 02:39:03 Anyone on? 2024-10-16 02:40:14 yes, but you may have to wait a bit for someone to answer depending on the question 2024-10-16 02:40:28 No worries, first time using irc so was just checking. 2024-10-16 02:41:01 welcome! 2024-10-16 02:41:30 My question relates to running a wireguard server on alpine. Have everything set up but cant connect on port 51820. I feel like its something I missed with alpine but I can't figure it out. 2024-10-16 02:42:58 could it be firewall config issue? 2024-10-16 02:43:00 Have ufw set up and ip forwarding, but cant connect on 51820 2024-10-16 02:43:47 Thats what I assumed, but I can ssh fine and everything. I have ufw rules set up for 51820, 443, 80, 22 but no lukck 2024-10-16 02:45:04 Do you know if alpine has any other default firewall running that could be conflicting? I read a bit about awall but it seems that has to be installed. 2024-10-16 02:45:25 not sure, would usually use iptables/nftables with wireguard 2024-10-16 02:46:54 any error on connecting? 2024-10-16 02:48:27 To the default firewall question, no Alpine doesn't set up one 2024-10-16 03:03:57 allow incoming udp on eth0 port 21820 2024-10-16 03:04:26 51820 * 2024-10-16 03:06:10 anyone wants to test my amneziawg-linux-virt APKBUILD? im lost module doesnt load 2024-10-16 03:50:53 is there a way to allow local writes on / inside abuild rootbld for testing? 2024-10-16 04:18:48 so if i do abuild -F module builds and loads correctly, no make install and no depmod used 2024-10-16 04:19:18 0xb6b4d238 udp_tunnel6_xmit_skb - wrong module generated inside abuild rootbld 2024-10-16 04:19:38 0x8533d13b udp_tunnel6_xmit_skb - correct one local build 2024-10-16 04:32:15 btw alpine needs a patch for make https://savannah.gnu.org/bugs/index.php?65172 2024-10-16 04:33:52 current make Segfaults on Makefiles with long define names like this one MODERN_KERNEL_SOURCES_NOT_FOUND_ERROR 2024-10-16 04:34:33 but it don't segfaults inside abuild rootbld for some reason 2024-10-16 05:32:11 krab4t: maybe differences in the underlying emulations? that would be my first guess. 2024-10-16 05:34:32 krab4t: sorry was scrolled up quite a bit; my comment is probably out of context at this point 2024-10-16 06:14:45 i think comparing 2 logs from abuild -v with make -d -p and removing all @ before MAKE in Makefile (maximum verbosity!) is the only way to debug ... 2024-10-16 06:24:12 hello everyone i have a rpi5 running alpine diskless, meaning it has a vfat partition constantly mounted on /media/mmcblk0p1 as ro. 2024-10-16 06:24:12 every time i update the packages the cache in /etc/apk/cache keeps the outdated packages 2024-10-16 06:24:21 is this supposed to happen? 2024-10-16 06:26:08 to solve it i do: 2024-10-16 06:26:08 doas mount /dev/mmcblk0p1 /media/mmcblk0p1 -o remount,rw 2024-10-16 06:26:08 doas rm -rvf /etc/apk/cache/*.apk 2024-10-16 06:26:08 doas apk cache sync 2024-10-16 06:27:23 kata: https://wiki.alpinelinux.org/wiki/Alpine_local_backup should help 2024-10-16 06:27:27 note that /etc/apk/cache is a link to /media/mmcblk0p1/cache 2024-10-16 06:28:56 im not sure if apk cache is related to lbu 2024-10-16 06:29:30 lbu is for updating the apkovl (overlay) 2024-10-16 06:30:01 when running diskless 2024-10-16 06:30:10 yes 2024-10-16 06:30:14 i know 2024-10-16 06:32:26 what i am saying is that /media/mmcblk0p1/cache contains packages that you have upgraded or removed 2024-10-16 06:32:47 "apk cache clean" dosent do anything 2024-10-16 06:39:38 kata: sorry, it has been quite a while since i ran diskless, apparently my memory of it is too spotty 2024-10-16 06:40:50 dont worry 2024-10-16 06:48:03 i'm spinning up a vm to play around with it, we'll see how much i can break 2024-10-16 07:23:58 kata: on the https://wiki.alpinelinux.org/wiki/Alpine_local_backup, did you try the 'Alternative' that has it put the cache on same media as the apkovl? 2024-10-16 07:45:03 j_v: that is alredy the case in a diskless install no? 2024-10-16 07:49:21 let me try 2024-10-16 07:49:34 brb cause znc is hosted on that machine 2024-10-16 07:50:05 kata: if have troubles, i can show you exact how i did it 2024-10-16 07:58:26 this is really funny 2024-10-16 07:58:58 i messed up the install i will have to wait until i get home 2024-10-16 07:59:35 luckly nothing else was running on the machine 2024-10-16 08:01:01 only glitch i ran into was that when i tried to do the initial 'apk cache sync' it failed until i did a 'apk update' first 2024-10-16 08:02:21 tldr if i install nano then uninstall it how do i remove the apk from the cache 2024-10-16 08:04:02 will have to test it to confirm, but 'apk cache --purge' should be the ticket 2024-10-16 08:05:40 whats --purge for 2024-10-16 08:05:45 i hosed my testing qemu installation in other ways, so i'm reinstalling it now 2024-10-16 08:06:06 clear out apks that are no longer installed 2024-10-16 08:06:11 ah ok 2024-10-16 08:06:14 are you on alpine now? 2024-10-16 08:06:17 i probably missed that 2024-10-16 08:06:19 no 2024-10-16 08:06:34 if not, https://manned.org has manual pages for alpine 2024-10-16 08:06:42 ok rhank you 2024-10-16 08:06:42 that should help some 2024-10-16 08:08:50 manned.org has stuff orgainized by distro/package instead of the usual section/page, but is pretty comprehesive and easy for me to navigate 2024-10-16 08:10:32 i see 2024-10-16 08:11:13 thats probably exactly what i was looking for 2024-10-16 08:11:37 im leaving now bye 2024-10-16 08:11:42 good night 2024-10-16 09:58:19 hey team, can I define a different version for a subpackage in an APKBUILD? Concrete example: Nim programming language v2.2 also builds nimble package manager v1.6, which I'd like to subpackage rather than packaging separately (and as requested by upstream) 2024-10-16 10:03:02 tokyovigilante: i am not sure, but #alpine-devel is probably a better place to ask 2024-10-16 10:05:27 j_v: thanks 2024-10-16 10:06:46 np 2024-10-16 10:34:13 nim-lang? nice! 😁 2024-10-16 10:37:54 tokyovigilante: It's not possible to package those separately? 2024-10-16 10:48:05 iirc nim and nimble currently are packaged separately on alpine. but if a user installs nim, not having its nimble package manager would be a strange setup imho. 2024-10-16 10:48:06 so I would agree: subpackage could be a niceer way to do it. 2024-10-16 10:49:06 ayex[m]: that's being handled by dependencies, not wether something is a subpacakge or not 2024-10-16 10:49:23 nimble being a subpackage still allows nim to be installed without nimble 2024-10-16 10:49:39 But them having different versions indicate they are released separately 2024-10-16 11:17:51 Does anyone want systemd-boot 2024-10-16 11:18:27 I wrote one import from Gentoo's systemd-utils 2024-10-16 11:20:40 At pmOS we already have systemd-boot and AFAIK there were some MRs regarding sd-boot in Alpine 2024-10-16 11:28:01 "nimble being a subpackage..." <- I see, thank you for the clarification. 2024-10-16 11:31:20 systemd-boot is pretty good 2024-10-16 11:31:45 i use it on my x86 computers (which there are not many of) 2024-10-16 11:39:52 "systemd-boot is pretty good" <- probably the only good part of systemd (obviously it wasn't their at the beginning, just merged it at some point) 2024-10-16 11:55:21 nah there's lost of good parts in systemd 2024-10-16 12:04:40 "nah there's lost of good parts..." <- I'm talking solely code-wise. systemd is great to use for sure 2024-10-16 12:05:27 the code quality is consistently similar everywhere 2024-10-16 12:05:50 Unrelated: It's the second time in a row that my Alpine Linux virtual machine do not permit me to connect to myself through 127.0.0.1 ? I just get connection refused (I've not nftables or iptables installed) 2024-10-16 12:07:23 I tried to run a daemon and bind it to 127.0.0.1 which did not worked an had to move it to 0.0.0.0 2024-10-16 12:07:26 is it a known issue? 2024-10-16 12:20:06 Make sure the lo interface is up 2024-10-16 12:29:07 "Make sure the lo interface is up" <- yes it is:... (full message at ) 2024-10-16 12:30:15 my bad maybe I should more clear. I can access myself through 127.0.0.1. But other computers can't. Generally binding 127.0.0.1 make you listen to any connection (at least its the case on Arch linux/Debian/Ubuntu) but on Alpine nope 2024-10-16 12:34:22 Huh, no 2024-10-16 12:34:49 127.0.0.1 should only be reachable from the same host 2024-10-16 12:36:24 ikke: I would like to believe you but my 10 years on Linux + default configuration on numerous software I installed is in opposition with that statement. Tho I don't think you're wrong I think it make sense. So there is something I must be missing. 2024-10-16 12:36:35 > <@_oftc_ikke:matrix.org> 127.0.0.1 should only be reachable from the same host 2024-10-16 12:36:35 * I would like to believe you but my 10 years on Linux + default configuration on numerous software I installed is in opposition with that statement. Tho I don't think you're wrong I think it makes much more sense. So there is something I must be missing. 2024-10-16 12:37:30 ftr: https://en.m.wikipedia.org/wiki/Localhost#Loopback 2024-10-16 12:38:55 hm, this would be news to me. 2024-10-16 12:38:55 when I serve http as dev setup at 127.0.0.1 only my machine can send requests to it, if I serve it at 0.0.0.0 any other machine can 2024-10-16 12:39:12 * other machine that can, * can reach it can 2024-10-16 12:39:15 ok then why all the software I've used in the last 10 years bind to 127.0.0.1 and I'm able to access it from other machines 2024-10-16 12:39:27 ayex[m]: > <@ayex:crossbach.de> hm, this would be news to me. 2024-10-16 12:39:27 oh 2024-10-16 12:39:27 > when I serve http as dev setup at 127.0.0.1 only my machine can send requests to it, if I serve it at 0.0.0.0 any other machine that can reach it can 2024-10-16 12:39:28 Yes. Imagine another machine getting a packet with source address 127.0.0.1 2024-10-16 12:41:07 wow. that make sense but I must have truly fall into an alternative universe then 2024-10-16 12:41:17 for not realizing that 2024-10-16 12:41:36 thx for the explanation. my brain grew bigger 2024-10-16 12:48:02 0.0.0.0 means to listen on any address, so no matter what interface a packet is received from, the service will be able to receive it 2024-10-16 12:48:30 You can also listen on a specific address to make it available to that network only 2024-10-16 12:49:56 To confirm: mkinitfs-bootparam.7.in is just for documentation? 2024-10-16 13:47:50 uhhh, so I know mixing edge and stable repos is "not officially supported" and "you're on your own", but...wtf is going wrong here, even??? why does upgrade not work??? https://bpa.st/raw/252Q 2024-10-16 13:50:19 I even explicitly pinned the packages, but it didn't work before pinning either 2024-10-16 14:19:27 Cool, it seems like that is related to man pages and section 2024-10-16 15:02:52 Hey, does somebody know if you can use the alpinelinux caddy package with plugins? I don't know a lot about go but they look like compile-time plugins to me so not sure if it is even possible ... 2024-10-16 15:19:12 msrd0: last time i tried it, the answer was no, you had to use something like xcaddy to compile your own build with plugins 2024-10-16 17:05:03 Hi guys, how do I fetch apks for exact versions of a package? I can do `apk add nano=8.0-r0` but not `apk fetch nano=8.0-r0`, just `apk fetch nano`. 2024-10-16 17:05:44 (also made an SO question: https://stackoverflow.com/questions/79094868/how-to-apk-fetch-exact-versions-of-a-package ) 2024-10-16 17:37:54 I hate how the first answer given there is clearly some kind of LLM answer that completely misses the point 2024-10-16 17:56:55 makes me laugh at myself that I had to look up "LLM answer"... but yeah that answer made my head tilt and my teeth hurt 2024-10-16 19:04:08 j_v: then also look up: https://en.wikipedia.org/wiki/Slop_%28artificial_intelligence%29 2024-10-16 19:12:29 hodapp_: now my head is hurting, too. it's similar to why i can't spend too much time with the news 2024-10-16 19:24:34 Maintainers are no long specified via comments? What did I miss? 2024-10-16 19:27:50 It's deprecated but still supported 2024-10-16 19:28:06 Was merged into the last release of abuild 2024-10-16 19:28:48 Only makes sense 🤷 2024-10-16 19:29:05 It makes sense, but also causes a lot of churn 2024-10-16 19:29:22 j_v: it's fine, it's like spam, just... worse 2024-10-16 19:29:25 But we have had that more often 2024-10-16 19:29:37 Eh, that's when you write scripts 😛 I just mass-update my groups of packages when I touch them now, my local --update-maintainer-format works well enough 2024-10-16 19:48:58 ikke: ayex[m]: they are currently packaged separately, I assume because of the version differences, the package manager can be built with the compiler or separately, the upstream devs have expressed a preference for packaging together but no real downside to having it separate. I've got working APKBUILDS for Nim 2.2, nimble 0.16.2, and working on packaging the LSP server (nimlangserver) currently. 2024-10-16 20:05:39 tokyovigilante: you do know that there are already aports for at least nim and nimble, albeit probably not as up-to-date, not sure about the language server 2024-10-16 20:07:10 tokyovigilante: sorry, missed you acknowledgement 2024-10-16 20:25:32 j_v: sorry yes I have an MR in against the existing nim package tp upgrade to 2.2.0, and will do one for the nimble package. Was thinking about refactoring to make nimble a subpackage but given the version differences it makes sense to keep it separate. 2024-10-16 20:25:57 the language server isn't currently packaged so will do one for that 2024-10-16 20:28:21 i don't know much about nim, but i've run across mention of it here and there. will probably give a look, but probably not today... too much on my plate atm 2024-10-16 21:05:58 do you guys know what GCC version was used to build current alpine kernel? 2024-10-16 21:07:04 could it be the source of bad module build? 2024-10-16 21:07:34 The last kernel was built today 2024-10-16 21:07:40 see /proc/version 2024-10-16 21:07:52 gcc 14.2.0-r3 2024-10-16 21:08:30 Linux version 6.6.56-0-virt (buildozer@build-3-20-x86_64) (gcc (Alpine 13.2.1_git20240309) 13.2.1 20240309 2024-10-16 21:09:14 krab4t: what alpine release? 2024-10-16 21:09:25 3-20 2024-10-16 21:09:27 3.20 2024-10-16 21:09:54 3·20 2024-10-16 21:09:56 VERSION_ID=3.20.3 2024-10-16 21:10:39 https://pkgs.alpinelinux.org/package/v3.20/main/x86_64/linux-lts 2024-10-16 21:10:49 2024-10-04 2024-10-16 21:12:40 host gcc version is 14.2.0, inside abuild rootbld is 13.2.1 2024-10-16 21:13:45 https://pkgs.alpinelinux.org/packages?name=gcc&branch=v3.20&repo=&arch=x86_64&origin=&flagged=&maintainer= but in repos it should be 13.2.1 ? 2024-10-16 21:14:49 then where 14.2.0 comes from? 2024-10-16 21:15:13 mix of 3.20 and edge? 2024-10-16 21:15:56 gcc -v gcc version 13.2.1 20240309 (Alpine 13.2.1_git20240309) 2024-10-16 21:16:16 actually gcc -v show 13.2.1 2024-10-16 21:22:50 doas abuild -F show 13.2.1 filthy magic. 2024-10-16 22:33:30 Anyone available to bounce a few alpine networking questions off of? 2024-10-16 22:36:15 just ask your questions, if people know the answer and are around, they'll respond 2024-10-16 22:36:26 irc is an asynchronous medium, use that to your advantage 2024-10-16 22:36:32 (and also stick around for the answer) 2024-10-16 22:39:45 I spun up a wireguard server on alpine through proxmox. I cant access port 51820 and cant for the life of me figure out why. Router is port forwarded, firewall off on vm, and from what I read alpine doesnt block it by default. Any ideas? 2024-10-16 22:41:42 route problem? 2024-10-16 22:49:32 user99: any luck pinging on the untunnelled addresses? 2024-10-16 22:56:50 Ya, so I just tried opening port 22 and testing that and it worked fine. I opened 51820 the same way, so I'm thinking it has to be something on the alpine side... 2024-10-16 22:57:05 Does alpine block any ports by default? 2024-10-16 23:01:58 no 2024-10-16 23:02:07 did you check your routing? 2024-10-16 23:02:21 on the alpine side 2024-10-16 23:03:28 you have both ipv4 and ipv6? 2024-10-16 23:05:18 What do you mean specifically by routing? I checked with netstat and both the ipv4 and v6 processes are listening on 51820 2024-10-16 23:12:22 which route are your wireguard packets going? 2024-10-16 23:13:32 are you sure you opened *UDP* port 51820 ? 2024-10-16 23:14:31 you could rule out LAN issues by making another VM in the same LAN and having them connect to each other 2024-10-16 23:19:31 @dwfreed, how should I open it? I have it forwarded in my router, I tested by trying the same with 22. I have been told that alpine doesnt block by default, and same with proxmox. 2024-10-16 23:19:32 tcpdump is also useful on the alpine side 2024-10-16 23:19:52 to see where packages are going to/from and if at all 2024-10-16 23:19:59 I don't know what you have for a router 2024-10-16 23:20:11 but yeah, tcpdump would also rule out whether the packets are actually arriving 2024-10-16 23:24:38 When I tcpdump on 51820 and try to connect, no packets show up. So i guess something is blocking it... 2024-10-16 23:24:43 sigh 2024-10-16 23:27:01 If I telnet to port 51820, tcpdump gets those packets... 2024-10-16 23:27:55 So it must be my router, but idk how 2024-10-16 23:40:13 normal telnet doesn't do udp 2024-10-16 23:40:38 ya sorry I used nc for that 2024-10-16 23:41:02 I'm honestly starting to think my isp router port forwarding is not working fully 2024-10-16 23:41:04 so, when you say you "tried to connect", does that mean, you tried to send a packet over the wg if? 2024-10-16 23:41:28 have you tried tcpdumping on the host from which you try to connect from 2024-10-16 23:41:46 I listened with tcpdump on 51820 on alpine, then tried to nc -u to alpine from my cloud box. 2024-10-16 23:41:50 Nothing. 2024-10-16 23:42:00 Tried the same thing on port 22 and it works fine 2024-10-16 23:42:26 have you tried running tcpdump on your cloud box? 2024-10-16 23:42:37 and looking where your packets go from there? 2024-10-16 23:44:07 you are comparing different interfaces when you say "tried port 22" and "telneting to port 51xxx" in that case you are usind eth0 (or whatever directly) 2024-10-16 23:44:31 but you want to see where your wg0 is trying to route via 2024-10-16 23:45:11 tcpdump your host from which you have trouble connecting from to your alpine box. 2024-10-16 23:46:13 That host is an iphone unfortunately 2024-10-16 23:46:28 But I can try from my cloud box 2024-10-16 23:49:37 user99: i've found that pinging the other host on the untunnelled interface will enable the two side to talk via the tunnedl, ie will get the handshake done 2024-10-16 23:52:59 that's if the rest of the setup is right, besides having the correct wire config's also having forward-ipv4 and forward-ipv6 for the wg if in /etc/network/interfaces 2024-10-17 00:20:34 ive made my amnezia wireguard vpn using awall 2024-10-17 00:21:20 json config looks okayish but works 2024-10-17 00:32:57 grep CC_VERSION_TEXT /usr/src/linux-headers-6.6.56-0-virt/include/config/auto.conf.cmd 2024-10-17 00:33:08 anyone? 13.2.1_git20240309? 2024-10-17 00:46:03 my host can't upgrade to 6.6.56-r1 but abuild rootbld can?? 2024-10-17 03:14:06 what tool official mainteiners use to build packages on server side? i mean what command they run to build APKBUILD? 2024-10-17 03:15:02 simple abuild no options on clean vm? 2024-10-17 03:53:07 https://www.youtube.com/watch?v=TrsAaX1u4HU im aliensongwriter & much more and this is sound produced by me myself & aliens 2024-10-17 04:17:33 krab4t: "abuild rootbld" will run on a clean chroot 2024-10-17 04:17:43 simpler than creating a clean vm 2024-10-17 04:32:37 https://pastebin.com/raw/Pb63MTHx abuild rootbld 2024-10-17 04:33:45 https://pkgs.alpinelinux.org/package/v3.20/main/x86_64/linux-virt-dev r0 2024-10-17 04:34:17 https://pkgs.alpinelinux.org/package/edge/main/x86_64/linux-virt-dev r0 2024-10-17 04:36:11 >>> amneziawg-linux-virt: Preparing build chroot... fetch http://dl-cdn.alpinelinux.org/alpine/edge/main/x86_64/APKINDEX.tar.gz 2024-10-17 04:36:18 why its fetching edge? 2024-10-17 04:54:56 and how to force rootbld actually use v3.20 instead of edge? 2024-10-17 05:53:08 https://git.alpinelinux.org/abuild/tree/abuild.in#n2558 probably user can specify release repo, but how hmm 2024-10-17 06:42:06 so its reads current git branch, not very nice, very developers with indepth knowledge focused. 2024-10-17 06:44:56 in all fairness, that's usually the type of person that would be doing this 2024-10-17 06:55:57 test 2024-10-17 06:59:16 "ikke: ayex: they are currently..." <- oh nice - thank you!! 😀 2024-10-17 11:15:01 socksinspace: read build tool source code instead of packaging own apks?! 2024-10-17 11:16:10 https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package#Modify_your_APKBUILD there is no info about how it uses branch name for repo url 2024-10-17 11:17:19 krab4t: can you imagine how it makes sense that if you try to build a package for master, it would use the edge repo, and if you build for a stable release, it uses that repo? 2024-10-17 11:18:45 no 3rd party repos are allowed? i want to call my branch myultratrashbin for 3.20. 2024-10-17 11:18:58 and from what i understand it will build for edge 2024-10-17 11:19:10 You can, but you have to do something more 2024-10-17 11:20:29 like what more 2024-10-17 11:21:02 If your packages depend on that repo, you need to add that repo the the .rootbld-repositories file 2024-10-17 11:21:13 You still need to make sure it's consistent 2024-10-17 11:21:44 either build against edge (with the master branch) or build against 3.20 with the 3.20-stable branch 2024-10-17 11:26:39 if google know nothing about rootbld-repositories query its a bad design :) 2024-10-17 11:28:20 tf no? we cant just index everything of alpine into google 2024-10-17 11:28:58 google do index wiki 2024-10-17 15:57:47 297 | __fh_size_t __sl = strnlen(__s, __n); 2024-10-17 15:57:47 /usr/include/fortify/string.h:297:36: error: implicit declaration of function 'strnlen'; did you mean 'strlen'? [-Wimplicit-function-declaration] 2024-10-17 15:57:47 usr/include/fortify/string.h: In function 'strncat': 2024-10-17 15:57:48 Often when compiling stuff, I get this error: 2024-10-17 15:57:48 My workaround is to set CFLAGS="-D_FORTIFY_SOURCE=0" 2024-10-17 15:57:48 What is the correct way to solve this? 2024-10-17 16:01:23 wizzard: what version of fortify-headers do you have installed? 2024-10-17 16:01:28 (apk version fortify-headers) 2024-10-17 16:02:09 wizzard: -D_POSIX_SOURCE or -D_POSIX_C_SOURCE or -D_XOPEN_SOURCE or -D_GNU_SOURCE or -D_BSD_SOURCE 2024-10-17 16:02:31 see /usr/include/string.h 2024-10-17 16:02:45 Thanks 2024-10-17 17:56:56 does someone knows how i can flush the dns cache in alpine ? (if there is any) 2024-10-17 17:57:24 There is no cache by default 2024-10-17 17:57:59 that's what i thought 2024-10-17 17:58:30 but somehow some application (aerc) doesn't seems to resolve to the new address... 2024-10-17 17:59:11 looks like it just did update... 2024-10-17 18:04:47 weird, seems to fail with ipv6 2024-10-17 18:14:07 ikke: okay I've might enable dns caching 2024-10-17 18:20:31 how can i tell if there is caching enable ? i tried restarting networking runlevel but it didn't helped 2024-10-17 18:21:14 You'd need a daemon running that caches it 2024-10-17 18:21:23 your resolver would point to something local 2024-10-17 18:21:46 okay i don't have that 2024-10-17 18:22:41 but when i use ping it doesn't resolve to the same address than a dig 2024-10-17 18:23:29 getent hosts 2024-10-17 18:24:04 uhm 2024-10-17 18:24:32 yeah this gives the old address 2024-10-17 18:33:32 i still don't really understand why getent returns something else, i've strace getent and it send a dns request to the nameservers i have in /etc/resolv.conf 2024-10-17 18:34:53 maybe compare wireshark/tcpdump? 2024-10-17 18:47:54 okay it's maybe caused by the TTL, but it's weird how two requests on the same dns server can give two different answeres 2024-10-17 20:05:54 Trying to get Pivpn running on alpine. Cant seem to connect through with a client (wireguard). Pivpn check shows all configs are good, and I was successfully able to netcat through my router into the vm so theoretically it should work. Just to test I also spun up another Ubuntu VM with Pivpn and that worked fine, so to me it must be an alpine issue. Any ideas what the issue could be? 2024-10-17 20:11:14 you have a tcpdump of the packets wireguard is sending? 2024-10-17 20:15:53 check that you have ip forwarding setup in your firewall as well as your sysctl params 2024-10-17 20:16:50 both your client and your server should have postup/postdown directives to add and remove the forwarding rules 2024-10-17 20:26:24 test 2024-10-17 20:26:59 test failed succesfully 2024-10-17 20:28:31 Lol, keep getting disconnected 2024-10-17 20:32:12 Cant remember the username of the person who asked but yes I did a tcpdump of my wireguard connection. Not super well versed on what to look for but data does seem to be moving. 2024-10-17 20:32:59 https://irclogs.alpinelinux.org/%23alpine-linux-2024-10.log 2024-10-17 20:37:18 Ah thats helpful thank you ikke 2024-10-17 20:49:11 so about the abuild git branch-repo-selectness, if i was this guy https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/73605 how could he test merge against v3.20 and not edge? 2024-10-17 21:02:27 Hello guys 🥰🥰 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 ) 2024-10-17 21:03:11 So, I did a reinstall of pivpn and now it works? 2024-10-17 21:03:44 Too bad it didnt work the first time I reinstalled it. Now I dont understand why it didnt work... sigh. Thanks for the help guys 2024-10-17 21:12:10 https://i.imgur.com/Xad4WRU.png if i call my branch this way, what repo abuild will select? 2024-10-17 21:12:23 regexp from abuild sources 2024-10-18 02:41:45 when i startup (alpine 3.20.3), shorewall seems to be failing to determine the IP address on boot. i can manually start shorewall after and it works. some relevant logs: https://put.gay/files/9LLEMBmFa464.txt 2024-10-18 02:43:09 it looks like the leased xxx for 900 seconds, adding route, etc, is happening after shorewall fails (but idk maybe its logging out of order no clue). the shorewall init.d file is stock and jas 2024-10-18 02:43:26 has need net, which i assume means it should wait until after the lease is obtained? 2024-10-18 02:43:37 but maybe im just wrong idk 2024-10-18 02:45:12 this started happening since updating from 3.20.1>.3, but also i havent rebooted for a while before updating, so it could be unrelated to the update 2024-10-18 04:59:46 how can i setup something like this https://build.alpinelinux.org/ for myself? or should i use github runners via docker better? 2024-10-18 06:10:49 https://vocaroo.com/1oh210kVRVJ2 2024-10-18 07:35:48 https://docs.gitea.com/usage/packages/alpine hmmm... 2024-10-18 09:46:46 what is that? 2024-10-18 09:50:08 [@_oftc_ncopa:matrix.org](https://matrix.to/#/@_oftc_ncopa:matrix.org) seems like an automated apk registry you can add to your repositories. So third parties publishing their own apks (test builds, internal stuff, etc.) can directly from the git respository publish it, no separate tooling needed 2024-10-18 09:57:03 That is indeed what it is. Gitea (en ofc Forgejo) can host several kinds of package registries, one of them being Alpine. It's a way to host an Alpine repository and you can indeed automate it from CI pipelines and such 2024-10-18 14:53:34 Does anyone know how to enable game controller added events in SDL games? I suspect it's an optional dlopen dependency, since sdl2 only depends on musl. 2024-10-18 16:09:16 xordspar0: Hi, I can't provide you specific hints, but I would search for your specific game controller and tutorials. You will find a bunch of (halfway dysfunctional) Ubuntu based tutorials, but some of these might give you names of packages that are necessary etc. 2024-10-18 17:14:29 Ah, it's not the dependencies. On my native Alpine system it works out of the box, but in my toolbox system it doesn't. toolbox has an open issue around receiving udev events inside a container: github.com/containers/toolbox/issues/1203 2024-10-18 17:15:54 holgersson: thanks 2024-10-18 17:16:16 yw :D 2024-10-18 17:20:57 I'm getting the following error when trying to launch a flatpak application: 2024-10-18 17:21:01 user@userhost ~> flatpak run org.standardnotes.standardnotes 2024-10-18 17:21:01 [3 zypak-helper] Failed to connect to session bus: [org.freedesktop.DBus.Error.NotSupported] Unable to autolaunch a dbus-daemon without a $DISPLAY for X11 2024-10-18 17:21:01 [3 zypak-helper] src/helper/main.cc:42(DetermineZygoteStrategy): Assertion failed: bus 2024-10-18 17:22:02 I don't have the same issue with the other Flatpak app I have on the system, namely, Gekbench. 2024-10-18 17:22:14 It launches just fine. 2024-10-18 17:55:49 hey :) 2024-10-18 17:56:09 Is there some repos in the wild with older packages that could be retreived as a rollback? 2024-10-18 17:56:22 I know non officially, but maybe there's something out there 2024-10-18 17:56:47 across distro versions? (eg. 3.19 vs 3.20) 2024-10-18 17:56:48 Being bitten again by kernel update and missing modules 2024-10-18 17:56:53 On edge :/ 2024-10-18 17:57:05 Looking for 6.6.53-0-lts 2024-10-18 17:57:25 It'd just be a little problematic to reboot right now and need to load a module 2024-10-18 17:57:38 it's not in your apk cache? 2024-10-18 17:57:48 I didn't activate cache 2024-10-18 17:57:53 uh oh 2024-10-18 17:58:06 :> 2024-10-18 17:58:16 Though that's interesting 2024-10-18 17:58:28 Is there a way to filter cache (like only caching kernel, for example)? 2024-10-18 17:59:38 quinq: i had something hit me like that a while ago... i just went through all the mirrors and eventually found the pkgver-pkgrel i needed for the rollback 2024-10-18 18:00:25 oh, https://mirrors.alpinelinux.org/, perfect 2024-10-18 18:00:28 Thanks, I'll have a look 2024-10-18 18:01:10 what's the arch? i'll look from the bottom, you look from the top 2024-10-18 18:02:15 your eyes meet from across the room. the music seems to fade away 2024-10-18 18:02:20 basic x86_64 2024-10-18 18:02:32 Don't worry j_v, I'll curl -I until I find one :) 2024-10-18 18:02:46 But I appreciate the proposition 2024-10-18 18:03:15 j_v's proposal was rejected 2024-10-18 18:03:21 roh :> 2024-10-18 18:03:43 fyi: mirrors.alpinelinux.org/mirrors.json 2024-10-18 18:04:07 geez, i walked right it that one, openned the door and everything 2024-10-18 18:16:14 So much 404 2024-10-18 18:16:31 yep 2024-10-18 18:18:56 Thanks for the help lopid and j_v :) 2024-10-18 18:19:06 :) 2024-10-18 18:21:29 :) 2024-10-18 18:25:21 wow, the mirrors are way more recent that when i needed to rollback. oldest edge lts kernel i found was 6.6.56 2024-10-19 02:09:01 https://dl-cdn.alpinelinux.org/alpine/v3.21/ what does this mean? 2024-10-19 02:10:37 im trying to setup forgejo for building packages, not sure if i should use gitea instead, all this dockers/runners/yml magic is new to me :) 2024-10-19 02:46:00 v3.21 is in progress, not released yet 2024-10-19 03:01:40 but how you build for different architectures? hardware or idk qemu emulation? 2024-10-19 03:13:17 hm which alpine docker image should i use to build packages? 2024-10-19 05:44:31 krab4t: we use native hw 2024-10-19 06:03:57 krab4t: you can use registry.alpinelinux.org/alpine/infra/docker/build-base 2024-10-19 06:04:08 It's what's used for our CI as well 2024-10-19 06:07:02 thats what i was looking at... 2024-10-19 06:08:01 https://pastebin.com/raw/Qq50tDpb this forgejo package registry is not real repository 2024-10-19 06:10:46 sadge :( 2024-10-19 06:11:20 krab4t: to host your own alpine repo, you just need a static webserver exposing the ~/package dir (or wherever you decide to store that) 2024-10-19 06:12:42 ikke: yeah then i could use github for code and actions don't need to host own CI 2024-10-19 06:14:24 ikke: It is just fetched via HTTP GET? 2024-10-19 06:14:28 yes 2024-10-19 06:14:36 It's really straightforward 2024-10-19 06:14:47 ikke: Weird 2024-10-19 06:14:51 Why is that weird? 2024-10-19 06:15:12 Nothing is supposed to be simple 2024-10-19 06:15:31 It needs to be convoluted and nonsensical 2024-10-19 06:15:41 true, we should work on that 2024-10-19 06:16:21 alpine: complex, large and enterprise 2024-10-19 06:17:40 Needs to be a custom protocol using telegram's "crypto" 2024-10-19 06:17:52 Maybe on port 445 2024-10-19 06:18:26 ikke: Where do I invest? 2024-10-19 06:19:16 Send crypto to 0oIVmNOCzswHN/ANKYoKb2U7RNRgDY0ZKvp0ew== 2024-10-19 06:19:21 :P 2024-10-19 06:23:44 heh 2024-10-19 06:24:30 yeah, hosting a repo is simpl i currently build it wwith buildrepo, and then host the thing with nginx :) 2024-10-19 06:26:30 if you wnated to skip the dockerized setup and the acess tokens you could probably just do a chronjob that does a "git pull && buildrepo" 2024-10-19 11:31:33 hello 2024-10-19 14:35:51 Hello, how i connect alpine linux to internet? 2024-10-19 14:37:50 Klewltz[m]: How are you running Alpine Linux? 2024-10-19 14:38:10 On sys mode 2024-10-19 14:38:31 So you already installed it? 2024-10-19 14:38:37 yes 2024-10-19 14:39:40 What kind of network connection do you have? 2024-10-19 14:40:44 wired or wireless? 2024-10-19 14:41:20 ethernet and wireless 2024-10-19 14:41:34 And what's the contents of /etc/network/interfaces? 2024-10-19 14:44:06 ACTION uploaded an image: (7498KiB) < https://matrix.org/oftc/media/v1/media/download/ASvRLjNj9DwQtVC99UiK2YGozpNi3eICCfD-kMkIg90WLqGlIQ7_PsP39C4K9LDjPEFcGjnrFaEMldoGmHYiWYlCeS7VVYfQAG1hdHJpeC5vcmcvWWlsbGpKVWhIbkF6d0NGR0RBYVRvRk5W > 2024-10-19 14:44:34 and ip addr show dev eth0 2024-10-19 14:45:07 I not have ips only localhost 2024-10-19 14:45:24 Do you see any interfaces? 2024-10-19 14:45:27 local is the only appears 2024-10-19 14:45:49 Do you know what network card you have? 2024-10-19 14:46:16 wlo0 and eth0 2024-10-19 14:46:40 no 2024-10-19 14:47:14 realtek more or less 2024-10-19 14:47:34 And it worked when booted from the ISO? 2024-10-19 14:47:41 yes 2024-10-19 14:48:17 And just to be sure, `ip link show` only shows lo?] 2024-10-19 14:49:28 yes 2024-10-19 14:49:49 Is it possible to boot the iso again? 2024-10-19 14:50:06 store the output of lsmod 2024-10-19 14:50:17 I tried many times 2024-10-19 14:50:48 Trying to determine what kernel module is loaded in the iso which is missing 2024-10-19 14:51:31 ACTION uploaded an image: (4928KiB) < https://matrix.org/oftc/media/v1/media/download/AeZhb1X-F7gF_HqK_VMi6fEUwhIl4MOpU6FJRLEYEy6Uu47Qytthney9QBQpr3NyB74yioO6_9TlhhjphGWe5ZVCeS7Vwg8wAG1hdHJpeC5vcmcvZFlMeG9VcUZ5Q2N3YnN0TUZ6VWFzUFdh > 2024-10-19 14:52:00 Im not in the iso 2024-10-19 14:52:16 Ok, now you should compare that when booting with the iso 2024-10-19 14:52:59 Ok 2024-10-19 14:58:43 ACTION uploaded a video: (4329KiB) < https://matrix.org/oftc/media/v1/media/download/ATT6y5FSHDzVBFxuq-Kdrumhk1qNJCkWxNF13N0FsEdZeZcvkCKAsOShtJUKLOjNZtXVS0Kj2cItNwjLMOUBX2FCeS7WK3pwAG1hdHJpeC5vcmcvenJaV1JaVkVaV09Zc05GYmFjQm96S0J3 > 2024-10-19 14:59:52 It's hard to follow what's going on like this 2024-10-19 15:00:19 When running from the ISO, do you see the interfaces now? 2024-10-19 15:00:23 ip link show 2024-10-19 15:03:21 yes 2024-10-19 15:03:41 ACTION uploaded an image: (3924KiB) < https://matrix.org/oftc/media/v1/media/download/AXJDEoiiEspg-5yPHmJUk2YxJONFtF6MsIaQ9M5JRVkeh3UhgAIC3Kvro_7N0mJHLa4pQ7W3HDXdjtFdyZWx8plCeS7WdFUAAG1hdHJpeC5vcmcvemlIa2lwQkdESEliT09aRUd1WlJUYUJl > 2024-10-19 15:05:42 this is how to works in sys mode 2024-10-19 15:05:52 ACTION uploaded an image: (3886KiB) < https://matrix.org/oftc/media/v1/media/download/ARselhJScwymO7mVrTSyx885s339WXYapJm7i3QovK_nIdNgBsG4l95Qkm7-JukMhgBTxXjJghig1OIEXGob0LtCeS7WlDIwAG1hdHJpeC5vcmcvclhDR2tqeXRYTGFEcHFBR1V0a3JnZ0tY > 2024-10-19 15:06:22 why? 2024-10-19 15:06:34 Why iso? 2024-10-19 15:06:50 Because your network card is detected when running the iso 2024-10-19 15:06:59 so we need to figure out what is the difference 2024-10-19 15:07:32 a 2024-10-19 15:09:48 I cant use wireless (wlo0) too 2024-10-19 15:10:09 s/cant/can/ 2024-10-19 15:10:33 I can use wireless (wlo0) too 2024-10-19 15:14:47 When booting the iso, can you run `lspci` 2024-10-19 15:15:00 or `lspci -v` 2024-10-19 15:17:19 no 2024-10-19 15:17:34 lspci not found 2024-10-19 15:17:55 maybe sh: lscpi: not found 2024-10-19 15:20:59 Do you have a network cable attached? 2024-10-19 15:22:34 yes 2024-10-19 15:23:14 but i can use wifi if the cable not works 2024-10-19 15:24:12 that's why I wanted to know how to connect my pc to the internet with cable and wireless. 2024-10-19 15:24:40 ok, try this: ip link set dev eth0 up; udhcpc -i eth0 -n; 2024-10-19 15:27:59 No such device WTF 2024-10-19 15:28:53 ACTION uploaded an image: (2568KiB) < https://matrix.org/oftc/media/v1/media/download/AeeGgZLh3otpQ_bFpZYb71I-9aMSkM_LxbyEggMIiknGQr1JBHz6ATFC9lqOzGv0r2_1Nw-PdA13E-pioJQoX-1CeS7X5YgwAG1hdHJpeC5vcmcvVndsZEZZUVpqcGN1SkxyT2dMb0l1TGZw > 2024-10-19 17:15:34 ACTION wonders whether his most recent feature request will break anyone's mind. 2024-10-19 17:16:50 url? 2024-10-19 17:17:01 https://gitlab.alpinelinux.org/alpine/aports/-/issues/16541 2024-10-19 17:19:02 bend maybe, not break... i actually like the idea 2024-10-19 17:20:21 an implementation of that would convince me to actually try messing around with xen 2024-10-19 17:20:53 I suspect linux-virt may omits several built-in portions which are always found on actual hardware, so linux-lts would likely work, but have high overhard. 2024-10-19 17:21:59 In particular the SCSI layer is used on basically all x86 hardware, but is often omitted from KVM or Xen VMs. 2024-10-19 17:22:18 I don't feel like I'm qualified to have an opinion on that. The gut reaction I have is that we can't really anticipate what hardware each user might want to have supported in a vm if they're doing something like this.. and linux-lts ought to work.. so I'd probably say use that. 2024-10-19 17:22:25 If there's overhead, can you quantify it? 2024-10-19 17:22:44 Kernel having support for this or that doesn't mean it's actually used 2024-10-19 17:22:59 SCSI is ~.75MB, which is significant in a 128MB VM. 2024-10-19 17:23:09 Ah, you're talking about memory 2024-10-19 17:24:00 Depends on the system, but you will eventually run out of either processor or memory with a VM system. 2024-10-19 17:24:19 There is a reason why you can get servers with >4TB of memory. 2024-10-19 20:00:24 hey 2024-10-19 20:01:58 how to enable CONFIG_BPF_LSM in alpine? 2024-10-19 20:14:18 1. read the /topic; 2. stick around for a bit 2024-10-19 20:20:23 you have only 1m43s to answer each question... :) 2024-10-19 21:34:55 Hello guys 🥰🥰 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 ) 2024-10-19 21:38:58 good spambot 2024-10-19 21:46:19 Unfortunately in order to be good, it needs to see into the future and kill the user before it happens (both online and the physical world). 2024-10-19 21:55:23 “Mail support@oftc.net if you think this is in error.” 2024-10-19 21:55:42 well, until they resurrect Philip K Dick so he can better explain his concepts for Minority Report, i think a 30s delay to autokill is about as good as likely to see 2024-10-19 22:27:06 hello, why is os-prober unable to detect other Linux OSes, only detects Windows? 2024-10-19 22:31:06 Chimmie[m]: do you have fuse loaded? 2024-10-19 22:39:05 yes 2024-10-19 22:42:06 then idk, check logs and see what does os-prober give 2024-10-20 00:26:51 where can I find os-prober logs? 2024-10-20 01:04:45 https://github.com/n1xx1/setup-alpine looks promising 2024-10-20 01:08:24 can i build unsigned apks on github ci and then sign them on repository vps? 2024-10-20 01:13:12 https://github.com/ivanvc/abuild-releaser-action 2024-10-20 03:24:09 Error: alpine.sh: command not found yeah very funny 2024-10-20 03:26:51 heh 2024-10-20 04:23:38 hey all, QQ about initial (sys) install outcomes when using setup-alpine: when running `setup-alpine -e`, and skipping creation of a user you eventually get prompted to input ssh keys or an ssh key url (by /sbin/setup-user, likely). which would be inferred to be root's ssh authorized keys, however upon reboot there is no /home/root/.ssh/ dir or contents. 2024-10-20 04:24:04 is that expected behavior, or something i should file a bug for? 2024-10-20 04:32:06 I think root's $HOME is just /root .. is there anything in there? 2024-10-20 04:40:00 ah yeah that's correct, but still there's no [expected] .ssh/ dir in `/root/` 2024-10-20 04:47:44 I've seen that option in setup-alpine but I've never actually tried it 2024-10-20 04:49:47 ah i see it's this part here: https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-alpine.in?ref_type=heads#L304-308 2024-10-20 04:50:11 is lbu add missing a wildcard after `/root` there? 2024-10-20 04:54:52 i assume it needs the `/root/.ssh/authorized_keys` filepath explicitly to actually achieve the desired effect, and that linked func doesn't seem to do it? 2024-10-20 04:55:20 based on https://wiki.alpinelinux.org/wiki/Alpine_local_backup#Include_special_files/folders_to_the_apkovl 2024-10-20 04:56:08 I think all that's for diskless 2024-10-20 04:56:22 but you mentioned you were doing a sys install 2024-10-20 04:57:19 the authorized_keys should be created by setup-sshd 2024-10-20 04:57:50 https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-sshd.in?ref_type=heads # here 2024-10-20 05:00:36 fack: which one are you doing, the ssh key paste, or the url? 2024-10-20 05:00:50 url 2024-10-20 05:01:39 I'll try it and see what happens 2024-10-20 05:02:30 damn this computer is fast heh 2024-10-20 05:04:38 it worked for me 2024-10-20 05:04:41 Alas much of the speed improvement in recent computers has been more and more processors, so only mult-threaded tasks have really been getting faster. 2024-10-20 05:05:33 fack: http://mmondor.pulsar-zone.net/fpaste?p=1MS 2024-10-20 05:05:49 that's on an install I just did 2024-10-20 05:07:02 (I realize this may not help you but :) 2024-10-20 05:08:32 interesting... :/ 2024-10-20 05:09:09 I assume you're not seeing the error ("Failed to fetch key from ...") 2024-10-20 05:09:23 yeah none of that at all 2024-10-20 05:10:16 you could try the command manually (maybe ^Z out of setup-alpine just after it asks you) 2024-10-20 05:10:32 i somehow have a handful of boxen that have no authorized keys files for root and prohibit-password set and I'm trying to find the cause 2024-10-20 05:10:36 actually you could answer that question (w/the url) and then ^Z and see if it wrote it then 2024-10-20 05:10:54 i'll give that a shot 2024-10-20 05:17:43 ACTION resumes toying with postscript 2024-10-20 05:20:45 hmmm yeah i think i jumped the shark on this one 2024-10-20 05:21:28 i suspect these probably failed to pull the keys from the url, because after kvm-ing in i can run setup-sshd -k [url] and they seem to get along fine 2024-10-20 05:21:43 at least it's an easy fix 2024-10-20 05:21:43 i must've been too inattentive 2024-10-20 05:21:56 yeah, sorry for sending you on a wild goose chase, thanks for the assistance! 2024-10-20 05:22:07 it's ok, I got to try a feature of setup-alpine I'd never tried :) 2024-10-20 05:22:11 np 2024-10-20 05:54:34 abuild rootbld doesnt work inside chroot? 2024-10-20 05:54:56 bwrap: Creating new namespace failed, likely because the kernel does not support user namespaces. bwrap must be installed setuid on such systems. 2024-10-20 06:35:36 As far as I see, qgis and gpsbabel never leave the edge branch. Is there a way to have these two packages available in the current stable branch? 2024-10-20 06:36:17 They are in the testing repo, they would have to be moved to community 2024-10-20 06:38:38 ikke: Thanks 2024-10-20 06:40:55 Is there a way to nudge? 2024-10-20 06:41:16 (nudge who, nudge where) 2024-10-20 07:02:17 lupo: you can create an issue on gitlab.alpinelinux.org/alpine/aports/-/issues 2024-10-20 07:02:41 ikke: Thanks again :) 2024-10-20 08:34:48 i really like alpine. 2024-10-20 09:03:03 Chimmie[m]: should be in /var/log/messages 2024-10-20 09:26:08 Ermine: is only probing the Linux partition using /usr/lib/os-probes/mounted/efi/10elilo, /usr/lib/os-probes/mounted/efi/20microsoft but is not adding another GRUB to the menu 2024-10-20 09:29:14 20microsoft seems to (successfully) detect windows 2024-10-20 09:29:48 it does but isn't able to detect other Linux FAT32 partitions 2024-10-20 09:30:35 may I ask you to post logs here (using some patebin service)? 2024-10-20 09:30:44 linux fat32 partitions? 2024-10-20 09:32:37 sure 2024-10-20 09:35:23 https://gist.chimmie.k.vu/GamePlayer-8/0fa8e31b24604043a63994279a265f24 2024-10-20 09:36:26 oh... 2024-10-20 09:36:44 actually, you need one ESP 2024-10-20 09:37:35 so should I change/remove ESP label from other GRUB bootloaders? 2024-10-20 09:40:54 You can put all bootloaders on a single ESP I guess 2024-10-20 09:44:23 different systems overrides other GRUBs on the same configuration 2024-10-20 09:44:36 on the same partition* 2024-10-20 09:46:42 They shouldn't if you install grubs to the different directories 2024-10-20 09:48:26 (this is how ESP on my HDD is structured: EFI/alpine , EFI/arch , EFI/ubuntu , etc) 2024-10-20 09:48:49 but actually, you should need only one grub installed which will multi-boot 2024-10-20 09:49:08 okay, will try 2024-10-20 09:50:26 hmmm, also os-prober should mount each partition and apply its heuristics to find out which linux distro is installed there and figure out how to boot it. Seems like it doesn't do it? Maybe because of LUKS stuff 2024-10-20 09:53:24 try installing cryptsetup/whatever on your Alpine and see whether it works 2024-10-20 09:53:50 (sorry for my thought dump, but that's a rather complicated setup) 2024-10-20 10:00:24 my Alpine runs on crypto_LUKS + ext4 and I think os-prober should only mount FAT32 partitions which are unencrypted 2024-10-20 10:02:48 in addition, moving 90linux-distros from /usr/lib/os-probes/mounted to /usr/lib/os-probes/mounted/efi makes it recognize the partitions but is unable to detect the OS name 2024-10-20 10:06:44 I've added GRUB bootloaders into one ESP /EFI/ directory but update-grub / os-prober still isn't able to detect other OSes and add them to the menu 2024-10-20 11:18:58 What is the right term for the process of making package, that is currently (only) exting in the edge branch, available in the (next) stable release? 2024-10-20 11:19:20 s/making package/making a package/ 2024-10-20 11:23:55 (I would like to open issues asking to make QGIS and GPSBabel available in the (next) stable release. Currently both are only availabe in the edgbe branch.) 2024-10-20 11:24:23 s/edgbe branch/edge branch/ 2024-10-20 12:27:40 lupo: not sure the 'right term' per se, but what you asking amounts to requesting that the aports be moved to community repo 2024-10-20 12:29:47 lupo: can be done via openning an issue, or for if you are inclined and interested enough creating a merge request 2024-10-20 12:31:45 lupo: for openning an issue, https://gitlab.alpinelinux.org/alpine/aports/-/issues 2024-10-20 13:03:13 lupo: you ask for the package in testing to be moved to community giving a juatification why. If you're using it and it is stable, state that. It'll then be in stable/community at the next stable release (~Nov) and Edge/community when it is bumped from testing. 2024-10-20 13:18:33 cyclisme24[m]: thanks, that was much clearer that what said 2024-10-20 15:54:38 Installing fcron on a system mounted with "nosuid" throws an error. I think because of an setegid call when given the -z flag in the fcron.post-install script. Works just fine though. 2024-10-20 15:58:26 j_v: Thanks 2024-10-20 15:58:36 cyclisme24[m]: Thanks 2024-10-20 16:03:02 hi all 2024-10-20 16:03:56 trying to create vbox guest ... when 3D acceleration enabled, VM won't boot ? any clues ? 2024-10-21 12:51:47 I'm trying to build a bunch of packages in CI. I've noticed on my local machine if the package is "up to date" in the packages dir, it doesn't rebuild it. If I re-populate the packages directory on the CI machine, it still rebuilds the packages. How do I get it to "detect" that the package is built and up to date, and to not rebuild it? 2024-10-21 12:54:43 I think it compares ctime of the package and the APKBUILD file 2024-10-21 12:56:05 Sorry, mtime 2024-10-21 12:56:22 so if I went through and touched all the packages, it would be fine? 2024-10-21 12:57:26 Ideally I only want to rebuild if the pkgrel or pkgver is updated. 2024-10-21 12:58:16 Look at buildrepo 2024-10-21 12:58:26 From lua-aporta 2024-10-21 12:58:34 lua-aports 2024-10-21 12:58:41 It does not seem to use mtime 2024-10-21 12:58:51 (abuild I mean) 2024-10-21 13:02:50 ikke: touching the .apk files did work, and I can still update a pkgrel and a package gets rebuilt. 2024-10-21 13:03:44 using (abusing?) gitlab CI to make a gitlab pages site for my own custom apk repo; works well enough. 2024-10-21 13:04:13 I abuse Github to do similarly for an OpenWrt package 2024-10-21 13:06:43 https://gitlab.com/adamthiede/alpine-pkgs/-/blob/master/.gitlab-ci.yml this is pretty neat! 2024-10-21 13:38:12 hey everyone. Glad to be part of this community. I'm planning to install alpine on a Raspberry pi 5 on data disk mode and run some containers 2024-10-21 13:39:49 seth[m]: welcome 2024-10-21 13:40:29 @ikke Thanks 2024-10-21 13:42:19 * some containers on podman 2024-10-21 15:15:37 is there anything built into abuild to build packages as a different architecture? Or is the answer there 'use a chroot' or something? 2024-10-21 15:18:46 CTARGET= abuild -r 2024-10-21 15:18:56 but the APKBUILD needs to be suitable for it as well 2024-10-21 15:23:00 ah ok. And the system just needs to have qemu or something to compile? 2024-10-21 15:30:14 out of curiosity, anyone else here has a setup with raspberry + podman? 2024-10-21 15:33:27 elagost: If you want to use qemu-static, you should use qemu-static, CBUILD, and abuild rootbld (which create a chroot) 2024-10-21 15:39:53 ikke: thanks, I'll look into that, never used it before. 2024-10-21 17:18:29 can one change vinds in extreme tux racer? surprisingly fun game... 2024-10-21 18:32:26 frag, vinds? 2024-10-21 19:43:18 binds iguess 2024-10-21 19:43:49 bindings? 2024-10-21 19:45:46 yes, binds :] 2024-10-21 19:45:59 cant see a way to do it tho, so prob not possible 2024-10-21 19:58:14 Apologies if this is discussed elsewhere but I couldn't find it: Is there any intention to backport the libvirt 10.8 to 3.20? 10.7 introduced a rather nasty bug. 2024-10-21 19:59:27 https://gitlab.com/libvirt/libvirt/-/issues/666 2024-10-21 20:00:17 I suppose https://gitlab.com/libvirt/libvirt/-/commit/008214842011d77fa4feda8d1ff586d29791f8ab would need to be backported? 2024-10-21 20:00:51 Not sure how impactful it would be to update libvirt from 10.7 to 10.8 2024-10-21 20:01:14 that would be one option yeah, if backporting that fix is easier / less risky 2024-10-21 20:02:19 xmodem: would you mind opening an issue on our gitlab? 2024-10-21 20:02:40 can do 2024-10-21 20:07:16 where is the correct spot? https://gitlab.alpinelinux.org/alpine/aports/-/issues ? 2024-10-21 20:15:36 yessir 2024-10-22 00:46:34 can i use your docker registry on github somehow? 2024-10-22 01:11:15 props to alpine for wisely choosing oftc rather than entrusting their irc to former freenode staffers at liberachat 2024-10-22 01:12:04 as former president george w. bush says, "fool me once shame on you, fool me can't get fooled again" 2024-10-22 03:13:52 https://registry.alpinelinux.org/v2/_catalog can't get list of all images? wow scam. 2024-10-22 03:18:10 you keep using that word. I do not think it means what you think it means 2024-10-22 04:12:53 bz: what happened 2024-10-22 04:17:57 github actions creates and checkouts everything to /home/runner, but you can't do su - runner ... 😢 2024-10-22 04:19:51 ah its because build-base has only one user buildozer 2024-10-22 04:20:06 Here are some of the things you can find on my channel:... (full message at ) 2024-10-22 04:21:38 these messages from matrix users really encourage me to go use it /s 2024-10-22 04:21:51 Here are some of the things you can find on my channel:... (full message at ) 2024-10-22 06:05:17 I am trying to get Alpine working with KVM and OpenVswitch? I have followed wiki with no luck 2024-10-22 06:05:52 Nope im done, github actions that start build-base https://github.com/SanderVocke/setup-qemu-container/tree/main doesnt work in my build case. rootbld - bwrap issue (probably because docker-on-docker). abuild -r - gets /lib/modules path from parent docker ubuntu container ... 2024-10-22 06:08:04 anowak: what's not working? 2024-10-22 08:43:18 hello, everytime I try redirecting usb devices to libvirt vms I get this error: 2024-10-22 08:44:09 spice-client-error-quark:Could not redirect > at 1-3: Error setting USB device node ACL: 'Not authorized' (0) 2024-10-22 08:44:57 do you have /usr/share/polkit-1/actions/org.spice-space.lowlevelusbaccess.policy (or is that something i made up)? 2024-10-22 08:46:42 yes, it seems unmodified 2024-10-22 08:47:26 make sure allow_active is yes 2024-10-22 08:47:39 it is set to yes 2024-10-22 08:49:06 i don't know about alpine, but on another distro i am a member of kvm, libvirt, and plugdev 2024-10-22 08:49:53 it also didn't work for me the first time, but allow_active is the only note i made when i fixed it 2024-10-22 08:50:08 maybe it's because libvirtd and qemu processes are run as root meanwhile virt-manager is run by the unprivilegeduser? 2024-10-22 08:50:43 the only explanation that I can come up with 2024-10-22 08:51:14 i wouldn't have thought you'd need to change that, since polkit is used for permissions 2024-10-22 10:13:18 i finally moved my docker containers to a HP Prodesk mini, running Alpine <3 2024-10-22 10:13:18 awesome stuff 2024-10-22 10:15:12 Hello everyone 2024-10-22 10:15:22 I am having an issue with running xorg 2024-10-22 10:15:47 it runs perfectly on root, but on user it doesn't seem to work at all, the keyboard is unresponsive 2024-10-22 10:16:02 alpinist: is your user member of the input group? 2024-10-22 10:16:37 Yeah, I added him to input, wheel, video and audio 2024-10-22 10:16:46 and it still doesn't wanna work 2024-10-22 10:17:01 And did you login again? 2024-10-22 10:17:10 (check the output of the `id` command to confirm 2024-10-22 10:17:14 yeah 2024-10-22 10:21:01 I am running it headless 2024-10-22 10:21:26 Believe it has something to do with permissions to OpenvSwitch 2024-10-22 10:22:29 Getting ERROR inernal error: Unable to get system bus connection when trying qemu://system 2024-10-22 10:22:56 I have had it running perfectly on debian and clearlinux - just not sure what is the issue with Alpine. 2024-10-22 10:26:11 so I've tried all of what was said + rebooting, but it still doesn't work 2024-10-22 10:26:23 I'm running cwm, which works perfectly on root 2024-10-22 10:27:30 but on user it only shows me a full screen terminal (st) which is within the wm and no keys are working 2024-10-22 10:30:50 "out of curiosity, anyone else..." <- anyone? 2024-10-22 10:32:18 no 2024-10-22 14:33:16 So I just found the weirdest situation at my job. We are using alpine, but someone went and in the Dockerfile did this `find /lib -type d -empty -delete && rm -r /lib/sysctl.d` now I cannot figure out , since it wasn't documented why on earth the sysctl is being removed 2024-10-22 15:22:12 Dex_2x: From what I can tell, there are no packages that have files in /lib/sysctl.d 2024-10-22 15:36:59 apk info -W /lib/sysctl.d/00-alpine.conf 2024-10-22 15:37:06 - /lib/sysctl.d/00-alpine.conf is owned by alpine-baselayout-3.4.3-r1 2024-10-22 15:38:09 Oh, that has been moved to /usr/lib/sysctl.d/00-alpine.conf in edge 2024-10-22 16:14:23 3.20 would still have it in /lib 😅 2024-10-22 16:15:31 yup 2024-10-22 16:15:37 I was just checking edge 2024-10-22 18:17:12 Okay so what is the recommendation here? This change was made in 2022 when the repo was shifted to use alpine images 2024-10-22 18:45:02 It looks like that 00-alpine.conf is setting some security hardening settings. So is it really better to remove it ? 2024-10-22 19:10:53 TIAS 2024-10-22 19:42:01 does anyone know which package/script/thing provides /boot/grub//*.mod? 2024-10-22 19:42:31 i tried `apk info --who-owns` on that folder but it couldn't find who owns it 2024-10-22 19:42:42 It's generally generated 2024-10-22 19:43:04 by grub-install? 2024-10-22 19:43:19 Though I haven't seen this *.mod files 2024-10-22 19:43:49 yea i can't find it anywhere 2024-10-22 19:44:13 i installed grub package to a dummy chroot directory to see what files it would give, but i don't see the *.mod files there 2024-10-22 19:44:52 May depend on specific hw 2024-10-22 19:45:43 Oh, I have this server that has /boot/grub/x86_64-efi/*.mod 2024-10-22 19:46:01 seems it's made by grub-install 2024-10-22 19:46:18 /usr/share/grub/*/*.mod 2024-10-22 19:46:41 sorry /usr/lib 2024-10-22 19:46:55 provided by grub-efi 2024-10-22 19:47:23 oh true, i missed that 2024-10-22 19:47:32 i only installed grub and not grub-efi 2024-10-22 21:19:09 hey guys. I'm installing alpine on a Raspberry pi 5. is there any way to pre-config my WiFi in order to have it connected at the first boot? 2024-10-22 21:19:58 * first boot and throughout the first installation? 2024-10-22 21:55:44 Use a cable instead. 2024-10-22 22:33:57 "hey guys. I'm installing..." <- You could put the *.apk files for connectivity (eg wpa_supplicant) on a USB stick and install them from there before setup. 2024-10-23 04:32:03 disconnect 2024-10-23 04:59:10 "You could put the *.apk files..." <- Thank you! I'll try that 2024-10-23 07:30:20 https://wiki.alpinelinux.org/wiki/Alpine_on_ARM mentions the file u-boot/pinebook-pro-rk3399/idbloader.img and u-boot.itb in the same dir, the dir is there but not these files, instead it has the file: u-boot-rockchip.bin (using alpine-uboot-3.20.3-aarch64.tar.gz, checked armv7 which doesnt have them also) 2024-10-23 08:43:17 line 6: can't create /etc/apk/keys/test.rsa.pub: Permission denied 2024-10-23 08:43:44 doas echo $ABUILD_KEY_PUB > /etc/apk/keys/test.rsa.pub 2024-10-23 08:43:52 explain pls? 2024-10-23 08:46:29 the shell redirection does not have privileges 2024-10-23 08:47:04 try echo $ABUILD_KEY_PUB | doas tee /etc/apk/keys/test.rsa.pub 2024-10-23 09:33:00 https://wiki.alpinelinux.org/wiki/Build_with_abuild_rootbld_in_Docker_container 2024-10-23 09:33:24 CMD="podman run $ARGS -v $GITHUB_WORKSPACE:$GITHUB_WORKSPACE --cidfile /home/runner/work/_actions/sandervocke/setup-qemu-container/v1/container_cid registry.alpinelinux.org/alpine/infra/docker/build-base:edge-aarch64 tail -f /dev/null" 2024-10-23 09:33:33 ARGS="$ARGS $ARCH_ARGS --privileged --security-opt apparmor=unconfined --security-opt seccomp=unconfined" 2024-10-23 09:34:56 abuild rootbld in docker is redundant 2024-10-23 09:34:58 args before cmd. 2024-10-23 09:35:27 A chroot in a chroot 2024-10-23 09:35:34 ikke: and how you build packages without it 2024-10-23 09:35:44 abuild -r 2024-10-23 09:36:33 bwrap: Creating new namespace failed, likely because the kernel does not support user namespaces. bwrap must be installed setuid on such systems. 2024-10-23 09:36:48 it literally doesnt work on github actions 2024-10-23 09:38:17 well lets check abuild -r 2024-10-23 09:40:08 make[1]: *** /lib/modules/5.15.0-1073-azure/build: No such file or directory. Stop. 2024-10-23 09:41:24 how it gets access to runner host fs from docker container 2024-10-23 09:42:22 probably /home/runner/ is binded inside docker container can that be an issue? 2024-10-23 09:42:27 -v $GITHUB_WORKSPACE:$GITHUB_WORKSPACE 2024-10-23 09:43:18 your build-base "chroot" already not perfectly clean solution! :) 2024-10-23 09:47:21 It's not built for github actions 2024-10-23 09:47:51 It may work, but may also not 2024-10-23 10:35:47 only one kernel in /lib/modules/6.6.58-0-virt, nothing from ubuntu host 2024-10-23 10:36:12 krab4t: You should generally not build against the host kernel anyway 2024-10-23 10:36:23 What are you trying to do/ 2024-10-23 10:37:22 ikke: https://pastebin.com/raw/UF750MgF currently yml looks like this, a bit mess 2024-10-23 10:38:44 I mean, what are you trying to build? 2024-10-23 10:39:09 https://github.com/amnezia-vpn/amneziawg-linux-kernel-module 2024-10-23 10:40:54 thats why ive asked for linux sources as package, it patches them to build 2024-10-23 10:41:27 You can just download the same source archive as linux-lts does 2024-10-23 10:41:45 Linux d504c0df2d90 5.15.0-1073-azure #82~20.04.1-Ubuntu SMP Tue Sep 3 12:27:43 UTC 2024 aarch64 Linux 2024-10-23 10:42:03 krab4t: You want to explicitly tell what kernel name and version to use 2024-10-23 10:42:05 i see, src/Makefile does uname -r 2024-10-23 10:42:07 it should not use `uname` 2024-10-23 10:42:44 ive found that like week ago that they doing uname -r, while writting APKBUILD 2024-10-23 10:43:07 now i remember, uname -r show kernel from host 2024-10-23 10:43:22 but host is not aarch64 LOL 2024-10-23 10:57:52 ikke: thats what i did in APKBUILD https://pastebin.com/raw/bpjDKkMp sources from cdn.kernel.org and it requires options="net" 2024-10-23 10:59:49 emulated docker/qemu/binfmt gcc compilation are very slow xD slower than my skylake tamagotchi 2024-10-23 19:42:40 older tarball for "generic arm" install is not available? the pinebook ppl suggest the kernel might have bug that leaves the screen black on kernel 6.6 2024-10-23 23:13:19 https://meta.alpinelinux.org/register/step2 (to register subscription to alpine mailing lists) says registration is currently closed. Is there another registration method? 2024-10-24 06:41:28 f451: develop is on gitlab https://gitlab.alpinelinux.org/ 2024-10-24 06:43:27 qaqland: f451 already left 2024-10-24 06:49:31 missed them by that much 2024-10-24 14:58:37 iwd[2898]: segfault at 7f26d6f85120 2024-10-24 14:58:38 fun 2024-10-24 14:58:59 Can you reproduce it? 2024-10-24 15:01:17 oh no just wanted to point it out 2024-10-24 15:01:24 no clue why or how it happened 2024-10-24 15:01:42 reproducing it would be step 1 in trying to figure out 2024-10-24 15:15:37 i've also had sporadic iwd segfaulst 2024-10-24 15:15:39 segfaults* 2024-10-24 15:15:58 should probably install debug symbols at some point 2024-10-24 15:30:15 why armv7 arch has kernel 6.6.58 but ppc64le 6.6.57? 2024-10-24 15:31:06 https://dl-cdn.alpinelinux.org/alpine/edge/main/ppc64le/linux-lts-6.6.58-r1.apk 2024-10-24 15:31:13 https://pkgs.alpinelinux.org/packages?name=linux-virt&branch=v3.20&repo=&arch=armv7&origin=&flagged=&maintainer= 2024-10-24 15:31:24 54 2024-10-24 15:31:30 ah its >>> r1 2024-10-24 15:32:38 all kernel numbers on that website pkgs.alpinelinux.org outdated 2024-10-24 15:36:15 ERROR: unable to select packages: linux-virt-6.6.57-r0: breaks: .makedepends-amneziawg-linux-virt-20241024.153425[linux-virt=6.6.58-r1] 2024-10-24 15:42:44 linux-virt (no such package) armhf riscv64 2024-10-24 17:11:40 `install_if` with hard-dependency on kernel version breaks kernel update if package not updated to new kernel version or release right? 2024-10-24 17:12:12 and also it breaks apk del of that package due to same problem? 2024-10-24 17:13:01 i can understand locking kernel update, but why it locks package removal 2024-10-24 17:22:33 also apk can't show any useful information why kernel version is locked and can't be updated due to 3rd party package hard version lock? 2024-10-24 17:23:52 finally i managed to install alpinelinux on baremetal :) and 4k screens... setfont -d default8x16 was sooooo useful 2024-10-24 17:23:55 i was wondering, if other users also use GNOME, and if they do so, do they only have gnome-console (not gnome-terminal? ptyxis seems not be available, and gnome-terminal although i can install it, i can't launch it, not even with LC_ALL=en_US.utf-8 dbus-launch gnome-terminal) 2024-10-24 17:53:50 i mean i enjoy alpinelinux on baremetal, as a desktop installation, but i'm missing software, alot of software 2024-10-24 17:54:38 http://hyperboleandahalf.blogspot.com/2010/04/alot-is-better-than-you-at-everything.html 2024-10-24 18:00:25 "i mean i enjoy alpinelinux on..." <- well best way to contribute is too aport rhe missing software :) 2024-10-24 18:01:38 on pmOS we have gnome-console pre-installed instead of gnome-terminal. gnome-terminal is kinda unmaintained afaik 2024-10-24 18:06:28 fossdd[m]: openbabel, fnt, opencubicplayer, modplugplay, pt2-clone. i could start contributing sure, but would the aports have any chances? i've tried a small patch for the webpage but it's not being processed, so i kind of didn't feel like starting contributing any aports 2024-10-24 18:07:12 tons of wmdockapps, wlmaker, gnustep apps... 2024-10-24 18:21:15 For some reason on my desktop apk is always (for months now) extremely slow. Building a package right now and it has to install 302 deps, it literally takes minutes while e.g. my laptop does it in max half a minute 2024-10-24 18:22:06 alex_: yes they definitely do have chance. aports is more actively monitored for PR's than the website and it's quite easy to get stuff in the repos 2024-10-24 19:09:55 are other disk-write operarions also slower? maybe its just your drive having some trouble 2024-10-24 20:21:47 Not that I notice anyway, just apk 2024-10-24 20:22:21 The drive is a NVMe SSD, luks encrypted btrfs partition. Both same as on my laptop 🤷 2024-10-24 22:10:21 "For some reason on my desktop..." <- Did your clear the cache? Do you have the same mirrors? 2024-10-25 04:27:53 Same mirrors yes. I don't think network has anything to do with it, after one run stuff should be cached but it's just as slow the second time 2024-10-25 08:47:07 PureTryOut: default there is no cache, try setup-apkcache manually? 2024-10-25 08:48:27 I still don't see how it would help, like I said my laptop is way quicker (same network but wifi vs lan so if anything it should be slower) 2024-10-25 15:10:10 Currently testing postfix on alpine on k8s and it seems that neither btree nor sdbm are supported database types - is that my bug or a compile option? 2024-10-25 15:11:30 telmich: probably compile option 2024-10-25 15:11:43 currently checking https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/main/postfix/APKBUILD 2024-10-25 15:11:49 it has lmdb, as well as some other dbs 2024-10-25 15:11:51 Haven't compiled postfix in 20 years or so 2024-10-25 15:12:36 The "interesting" problem I am trying to solve is how to store the SSL/TLS cache 2024-10-25 15:13:00 by default: smtpd_tls_session_cache_database = sdbm:/var/lib/postfix/smtpd_scache or smtpd_tls_session_cache_database = btree:/var/lib/postfix/smtpd_scache 2024-10-25 15:13:21 I wonder if ncopa skipped support for these intentionally 2024-10-25 15:14:07 berkley-db was explicit disabled (due to the license of new versions) 2024-10-25 15:14:32 ohhhhhh 2024-10-25 15:14:34 https://bbs.archlinux.org/viewtopic.php?id=294782 2024-10-25 15:14:50 right 2024-10-25 15:14:54 lmdb it is 2024-10-25 15:15:36 re-running helm upgrade --install -f ungleich-mx/values-test-router.yaml mx1 ungleich-mx 2024-10-25 15:15:45 Didn't know that btree relied on bdb 2024-10-25 15:16:35 Stalwart is an interesting alternative 2024-10-25 15:17:01 hmm 2024-10-25 15:17:10 SSL connections using openssl still abort 2024-10-25 15:17:14 strange 2024-10-25 15:17:40 Using openssl s_client -connect router2.ge.nau.so:25 -starttls smtp I get as far as RCPT TO 2024-10-25 15:17:46 and then it dies with: 2024-10-25 15:17:52 289B93340D7F0000:error:0A00010A:SSL routines:can_renegotiate:wrong ssl version:ssl/ssl_lib.c:2825: 2024-10-25 15:18:44 I thought that was a cache issue / follow up error from the tls cache, but it is not 2024-10-25 15:19:48 Ahh... is the R special in s_client??? 2024-10-25 15:19:50 omg 2024-10-25 15:19:51 so stupid 2024-10-25 15:20:42 it indeed is: https://github.com/openssl/openssl/discussions/21437 2024-10-25 15:24:21 Does anyone have a good hint on what to use instead of openssl s_client -connect router2.ge.nau.so:25 -starttls smtp to manually test an smtp session when only TLS is allowed? 2024-10-25 15:26:37 Ah, smtp is silly 2024-10-25 15:26:45 I can do " RCPT TO" with a leading space 2024-10-25 16:48:26 you can also do lowercase R 2024-10-25 16:48:40 i usually do everything in lowercase because i can't be bothered 2024-10-25 17:01:12 hey guys. I'm trying to get run podman on alpine. on my Raspberry. if Alpine is installed as sys (on the SD) it works normally, however it seems I can't get it work if alpine is running on RAM (as I'd prefer). does anyone have any suggestions? Thanks 2024-10-25 17:03:43 telmich, maybe socat 2024-10-25 18:57:58 did something happen to dl-cdn? apk update failed for me; all I see in https://dl-cdn.alpinelinux.org/alpine/edge/main/ is "riscv64". 2024-10-25 19:00:56 hmm 2024-10-25 19:01:51 It's not just dl-cdn 2024-10-25 19:04:56 what happened to the mirrors? 2024-10-25 19:05:00 they are all only riscv64 2024-10-25 19:05:14 I tried a few and they only have riscv64 builds on edge 2024-10-25 19:06:47 I guess we should just upgrade to risc-v computers then? 2024-10-25 19:07:43 all non-libre architectures are deprecated ;P 2024-10-25 19:16:21 heh 2024-10-25 20:01:32 lol, riscv64 only on my side too 2024-10-25 20:07:24 I feel so outmoded being on arm64 2024-10-25 20:27:09 we don't know what happened yet, but we are restoring the data that got lost 2024-10-25 20:28:33 We know how it happened, but not yet what caused it 2024-10-25 20:33:16 i am guessing that there was some catastrophic malfunction in seeding the mirrors from the master server 2024-10-25 20:34:29 No, the packages were gone from the master as well 2024-10-25 20:36:07 18:44 edge/main/: uploaded 2024-10-25 21:26:46 telmich: you're that ipv6 guy yeah? 2024-10-25 21:28:23 we remember talking to an ipv6 guy somewhere who pointed us to a matrix channel but we can't remember who or where 2024-10-25 21:32:56 hello - it seems like linux-virt in 3.20.3 is missing the driver for my PCI device '00:07.0 Ethernet controller: Chelsio Communications Inc T540-CR Unified Wire Ethernet Controller [VF]' - i understand not including the PF driver but i thought VF would be included. is there another package i need for this? (it works fine with linux-lts) 2024-10-25 23:12:08 hello! I recently noticed that the Alpine edge repos (https://dl-cdn.alpinelinux.org/alpine/edge/main/) are incomplete. there were previously entire architectures missing incl. x86_64. they seem to be back now but are missing most or all packages 2024-10-25 23:13:33 Eunakria: known issue, being worked on 2024-10-25 23:13:41 gotcha, thank you 2024-10-25 23:13:45 np \o/ 2024-10-25 23:14:45 out of curiosity, any ETA on when it's being resolved? I'm not in any rush and I don't believe in pressuring OSS maintainers for volunteer work, just curious 2024-10-25 23:29:38 I have not seen one yet 2024-10-25 23:31:05 fair enough 2024-10-25 23:33:06 "A couple of hours" https://fosstodon.org/@alpinelinux/113370036561484573 2024-10-25 23:35:04 thank you! 2024-10-25 23:35:49 I know it's Friday evening for a lot of folks on the team so I can't imagine it's a great time for something to go wrong. massive props to y'all for handling something like this in such a timely manner 2024-10-25 23:50:48 hi, any reason this https://dl-cdn.alpinelinux.org/alpine/edge/main/x86_64/ is empty? 2024-10-25 23:52:13 trench: known issue, https://fosstodon.org/@alpinelinux/113370036561484573 2024-10-25 23:55:00 elagost: thanks for the info 2024-10-26 00:38:43 Hello, http://dl-cdn.alpinelinux.org/alpine/edge/main appears to be empty for all architectures. This is currently breaking apk for me. What happened? 2024-10-26 00:40:14 Interestingly, riscv64 appears to be the only exception. 2024-10-26 00:41:35 I have also checked on nl.alpinelinux.org, assuming that this is a mirror (haven't checked). At least that server also appears to be empty. 2024-10-26 00:42:35 Yeah, the repos are down. They're working on it. 2024-10-26 00:48:42 known issue, https://fosstodon.org/@alpinelinux/113370036561484573 2024-10-26 01:12:23 Welp, what timing... I hadn't touched these systems in months since today. Hopefully it'll be resolved soon! 2024-10-26 02:15:09 Ayy, back in business! Looks like the issue has been resolved! 2024-10-26 02:20:38 Sadly does not resolve all. s390 and x86 community repo is empty yet 2024-10-26 02:28:14 Oh, you're right... Hopefully those will get resolved soon too! And I wouldn't mind seeing a post-mortem after the fact either to be honest. So far we know that it's a bad rsync command, but how did that get pushed so far into production? Not just production, but all the mirrors too, and for nearly all architectures, in critical repository content (main and community). I think 2024-10-26 02:28:17 that calls for clarification in deployment procedures. 2024-10-26 02:34:57 vimicito: agreed. 2024-10-26 02:49:12 I'm curious if anyone has experience setting up dropbear in initramfs on Alpine Linux. Poking about, I don't see a dropbear-initramfs package, and I haven't found mention of people setting it up. Looks like mkinitfs could accomplish it via a custom initramfs-init script, so I'm mostly wondering if someone has done and published the work already. 2024-10-26 02:51:15 i have done that, though with dracut rather than mkinitfs 2024-10-26 02:51:40 nep_nep: Ah, I'm still an Alpine noob. Unsure of what options exist. 2024-10-26 02:54:05 Out of curiosity, why do you want dropbear SSH to be in the initramfs specifically? Can it exist in the rootfs after that's jumped into? 2024-10-26 02:54:35 i don't know about them, but i personally use it for remote root decryption 2024-10-26 02:57:12 oh that's clever 2024-10-26 02:59:11 vimicito: nep_nep's use case, yeah. I do it with Debian. 2024-10-26 03:00:17 That's really clever, so far I've had remote rootfs only as passing thoughts (and encryption didn't even cross my mind at all). How do you attach the rootfs then, NFS? 2024-10-26 03:01:21 vimicito: It's usually local. 2024-10-26 03:03:59 What's dropbear used for then though? In my own environment, I use Alpine as LXC guests using ZFS as backing storage. So there's the option for encrypted storage, but that would be handled by the host. For local encrypted storage I'd think of LUKS, but wouldn't that require cryptsetup-initramfs instead? 2024-10-26 03:04:44 Or are you using SSH to unlock such encrypted ZFS from the guest? 2024-10-26 03:07:53 there are no guests, once the host is booted the root filesystem has to be decrypted and mounted before anything significant can happen, and to do that remotely without a serial connection (which would be inconvenient due to being a physical server at home, rather than a datacenter with serial infrastructure) you need something like ssh 2024-10-26 03:08:05 vimicito: ZFS native encryption is problematic. Just LUKS here. So, the way I use it is to ssh to a host who's waiting to be unlocked. dropbear runs from within the initramfs, and lets you unlock root as though you were at the console. 2024-10-26 03:08:24 ^ 2024-10-26 03:09:57 do not ever use zfs native encryption. source: used zfs native encryption, now full of regret 2024-10-26 03:12:32 I'm using it for some cold storage datasets, but for the most part I just use it in a vanilla configuration (zstd compression at most, no dedup, no encryption, no exports). Some pools are on RAIDZ, many arae single disk. Given that, what issues did you encounter with it? Do you think I should send those encrypted datasets to unencrypted counterparts and encrypt them another way? 2024-10-26 03:13:03 vimicito: It breaks on send/receive most commonly. 2024-10-26 03:13:55 right, if you make a zfs snapshot send zfs it, and the snapshot is of an encrypted filesystem, your pool will report random I/O errors and you won't be able to read the snapshot 2024-10-26 03:14:16 rebooting fixes it (at least IME) but the issue is concerning enough that i would never deploy it 2024-10-26 03:14:16 vimicito: As noted, safest bet is to use LUKS encryption. Yes, this means you can't send encrypted data to an untrusted system as easily, but eh. 2024-10-26 03:15:46 mason: I feel like if you're doing E2EE then you already have some other infra to make it happen. Unless you depend very deeply on ZFS features, ZFS having its own tooling around E2EE isn't tremendously helpful 2024-10-26 03:17:18 I think it's a little like how a few Windows Server shops depend super heavily on IIS-specific or MSSQL-specific or other such quirks, and because of that their own ability to implement features is predicated on Microsoft implementing support for certain use cases. I think it's possible to over-depend on ZFS in the Linux/BSD/Solaris space as well 2024-10-26 03:17:53 It offers a fairly unique feature. 2024-10-26 03:18:25 Or a couple perhaps. 2024-10-26 03:18:41 To be honest, the main reason I even sent those datasets into encrypted ZFS was because of the "promised" easy send/recv. I think I would have the presence of mind to decrypt before transfer, but it seems that ZFS should just throw an error if the dataset isn't yet decrypted. Surely that's just one check away for them too? 2024-10-26 03:18:57 Eunakria: zfs native encryption is more about on-disk encryption than E2EE, it's about encrypting data at rest, and in theory, it should work pretty weel 2024-10-26 03:19:13 it just doesn't because it's full of bugs 2024-10-26 03:20:29 ye I mean that if you're sending ZFS native encrypted data between hosts across untrusted hops, that's a form of E2EE within your own datacenters. I do agree that at its core, ZFS native encryption only solves the encryption-at-rest problem, which is the same problem that LUKS solves 2024-10-26 03:20:49 ivy-: this. I don't care much about E2EE for this data, since it's strictly mine anyway. There is no other end to encrypt to. Realistically, my threat model is just that keys may be in memory (albeit unlikely, being data at rest). 2024-10-26 03:22:14 I think that the key functionality of most filesystems used in the datacenter is being able to transfer snapshots and diffs over the network. XFS, Btrfs, ZFS on Linux, and if memory serves ReFS on Windows Server all kinda have this functionality to some extent, but it's about the quality of these features' implementation 2024-10-26 03:22:14 For transfers between hosts, I think WireGuard is a godsend to be honest. Good performance between datacenter regions, and takes care of all your trusted network needs. 2024-10-26 03:23:44 but yeah, I think that if your concern is secure transfers between hosts, use WG. it's great for tunneling data between services that normally operate in cleartext, and there are plenty of services that basically have no halfway decent encryption layer at all so you'll probably benefit from it 2024-10-26 03:23:46 e.g. SMB 2024-10-26 03:24:33 I think sending an XFS snapshot, or unencrypted ZFS snapshot, over WG is just as feasible a solution as sending a native-encrypted ZFS snapshot over raw socket 2024-10-26 03:25:26 Absolutely. Internally, sometimes it's also just hard to use things like TLS. So for internal web, mail, ... resources, do you mass deploy a CA or do you just rely on WG cross-networks and WPA on the air? 2024-10-26 03:25:39 i hope if it's okay if i re-ask my earlier question as i think it was overshadowed by the repository issues -- i have an Alpine 3.20.3 VM, with a PCI device '00:07.0 Ethernet controller: Chelsio Communications Inc T540-CR Unified Wire Ethernet Controller [VF]' and it seems like linux-virt does not support this device - is that expected? linux-lts works fine 2024-10-26 03:26:15 i understand linux-virt doesn't support most PCI devices but i thought VFs would be included 2024-10-26 03:28:20 I think the thing that's missing in linux-virt is not so much drivers for PCIe itself, but drivers for each PCIe device. I'll have to check the APKBUILD 2024-10-26 03:28:34 *config not APKBUILD 2024-10-26 03:28:52 fwiw, if the answer is "just use linux-lts" that's fine, but i'm wondering if i should file a bug to have this driver added to linux-virt 2024-10-26 03:30:02 I'm not an Alpine dev myself so I don't know if I can really tell you confidently what linux-virt is supposed to be, ideologically, and whether this option should be enabled; but intuitively it feels like linux-virt is for guests that *only* need paravirt drivers (e.g. VirtIO) 2024-10-26 03:30:22 yeah, i thought that might be the case based on what the wiki said 2024-10-26 03:30:39 which is understandable since most VMs probably aren't using PCI VFs 2024-10-26 03:31:21 might be nice if there were an extra package to add these drivers though :-) 2024-10-26 03:32:03 monokernel woes 2024-10-26 03:33:55 bash should be made a dependency of timeshift 2024-10-26 03:33:57 the linux-lts imagine is only about 20MB larger than linux-virt so i suppose in practice it's not really an issue 2024-10-26 03:34:04 s/imagine/image/ 2024-10-26 03:34:32 timeshift-gtk expects bash to exist and launches shells for rsync, but it... well, silently fails on system where bash doesn't exist 2024-10-26 03:36:56 its also complaining about /dev/mapper not existing on my system? 2024-10-26 04:01:42 "we remember talking to an ipv6..." <- I might be, indeed 2024-10-26 04:02:21 "we remember talking to an ipv6..." <- The channel would be IPv6 2024-10-26 04:03:46 "we remember talking to an ipv6..." <- Is this a reply? 2024-10-26 04:05:36 I am, uh.. a lowly IPv4 2024-10-26 04:07:17 Got the fancy host though ;) 2024-10-26 04:09:16 telmich: we need a new matrix, can't you get an IRC channel? meh 2024-10-26 06:06:59 Soni: https://xkcd.com/1782/ 2024-10-26 06:08:47 Soni: you're here too? omg 2024-10-26 06:09:09 what OS can i use where you don't turn up? 2024-10-26 06:17:08 ivy-: we're not touching openbsd with a 1000ft pole 2024-10-26 06:18:58 telmich: hehe, yeah. 2024-10-26 06:20:38 ivy-: we don't understand the hostility but we hope you find what you're looking for 2024-10-26 06:20:59 yeah i would also never use openbsd tbh, so that's fair 2024-10-26 06:21:31 Soni: mostly that you seem to take very negative attitude to any OS that doesn't implement IPv6 in exactly the way you think it should 2024-10-26 06:22:16 like when we spent about a week arguing why FreeBSD doesn't implement CLAT in libc, because literally no one other than you wants this 2024-10-26 06:24:32 ah, no worries, we have recently given up on that. as it turns out, CLAT in libc is nearly impossible. thankfully we have come up with a new way to make outdated sysadmins learn IPv6, which may well be just as effective: https://mailarchive.ietf.org/arch/msg/v6ops/W79_uFtcAGxHFCl6uOwjwOrP-1E/ 2024-10-26 06:25:42 hilarious that you think the only options are "freebsd implement CLAT in libc" or "outdated sysadmins learn IPv6" 2024-10-26 06:26:06 nah, our only option is to make outdated sysadmins learn ipv6 2024-10-26 06:26:34 CLAT in libc is one way to do that, turned out to be impossible (thanks signals). but it's not the only way. 2024-10-26 06:26:40 although this thread demonstrates you have no idea how networking working works or what 464XLAT is 2024-10-26 06:26:42 polonium 4 is also an option 2024-10-26 06:26:55 and omg why am i getting into this thread, fuck me 2024-10-26 06:27:22 i'm not saying that i disagree with the "people should finally learn ipv6" sentiment, this is getting very offtopic 2024-10-26 06:27:24 literally everyone on v6ops is telling you that you are wrong 2024-10-26 06:27:55 yeah it is off topic, i will /ignore that person, apologies 2024-10-26 06:28:26 ah, fair enough, apologies 2024-10-26 06:32:24 no worries :) 2024-10-26 08:19:25 mason: I’ve been thinking about trying to implement unlock LUKS from a ssh server in initramfs. All instructions seem to point to dropbear, and most rely on a package dropbear-initramfs specific to Debian. Is it straightforward to do in Alpine, or is there any pitfall? 2024-10-26 13:10:53 "outdated sysadmins" -- sounds like ai output 2024-10-26 13:11:27 so i guess cdn is back to normal? 2024-10-26 13:16:25 invoked: the packages should be present again 2024-10-26 13:19:42 ikke: cool. thank you 2024-10-26 13:48:22 Soni: at the risk of going OT for this channel (is that allowed?), my main gripe with IPv6 is that it requires me to manage e.g. firewalls twice and/or risk one having more lax access than another. Internally I would always use IPv4, I really don't need a gazillion addresses when 192.168.0.0/16 already provides 65535 of them, and 10.0.0.0/8 offers a whopping 16 million. On the 2024-10-26 13:48:25 edges meanwhile, I can understand the desire for dualstack, but that's also where the firewalls are the most complex and sensitive. I want to cater to those who want to get rid of e.g. cellular and (e.g. in Germany) even residential CGNAT. But it really doesn't seem like a pressing issue to me either... 2024-10-26 13:50:07 vimicito: you should give 464XLAT/SIIT-DC a try, it basically solves the "firewall twice" (assuming you mean having to firewall ipv4 and ipv6 separately) 2024-10-26 13:51:55 tho last we checked alpine's support for 464 is still lacking 2024-10-26 13:56:57 I think I will, thank you! IETF mailing lists are never easy reading material, but I've put a pin in it for now. For IPv4 firewalling on the edges, I have already made some tooling too that merges iptables lists. I don't think it should necessarily be that hard to use the same tooling to do both either. So far the rules are already tens of thousands of them, mostly related to 2024-10-26 13:57:00 geoblocking (Chinese because of spam and lack of legitimate business with them). The rest is then ratelimiting, open ports, routing and so on. Each of those just has a file in /etc/iptables/rules.d/50-ratelimit or similar. 2024-10-26 14:01:22 also linux has some issues with firewalling on the same machine that does the SIIT, so we would avoid doing that for now, but still might be worth looking into 2024-10-26 14:01:26 ^^ 2024-10-26 14:37:37 Hello, i'm trying to compile the linux kernel on an alpine docker container (3.20) (also confirmed on a virtual machine with alpine edge) but I got into the following error: 2024-10-26 14:37:46 ACTION uploaded an image: (35KiB) < https://matrix.org/oftc/media/v1/media/download/Ae-__aiKj9EUiDRSIOWuGgShPf2771reC-vP_LbG7G68pXXaW0L3KkAVtJToL4BmtUXYm4bSpYya40F6LLZvRPVCeTEVwPOAAG1vbmVyby5zb2NpYWwvVWFaT1NDWXRydklhTVZDWGpMVXdsZk5m > 2024-10-26 14:38:17 What is the package on alpine that give`` 2024-10-26 14:39:10 SyntheticBird[m]: https://pkgs.alpinelinux.org/contents?file=types.h&path=%2Fusr%2Finclude%2Fasm&name=&branch=edge&repo=&arch= 2024-10-26 14:39:32 Thanks a lot 2024-10-26 14:41:12 see here for the rest of depends https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/main/linux-lts/APKBUILD#L11-14 2024-10-26 19:34:02 What is the best way to build a custom alpine repository that automatically sync on any commits ? 2024-10-26 19:34:36 I'm experimenting with CI but even on Alpine gitlab I think its just used for tests, not sure how I could pull this off to the actual repository folder 2024-10-26 19:34:45 s/folder/that could be served/ 2024-10-26 19:35:58 SyntheticBird[m]: we use aports-build, triggered by mqtt 2024-10-26 19:36:23 aports-build in turn invokes buildrepo from lua-aports 2024-10-26 19:37:02 Our ci is indeed not meant to build complete repositories 2024-10-26 19:40:20 I see, I will try that in a vm thx 2024-10-26 19:40:46 Is there actually anyway to only limit the build to some specific packages ? 2024-10-26 19:40:52 * Is there actually any way to only limit the build to some specific packages ? 2024-10-26 20:06:28 pu: I haven't done it yet, but I'll document what I do when I do it, and nep_nep has done it with Dracut and might be encouraged to document that too. =nudge, nudge= 2024-10-27 00:20:06 Let me help you archive your health & fitness goals today.... (full message at ) 2024-10-27 00:20:38 Let me help you archive your health & fitness goals today.... (full message at ) 2024-10-27 03:10:37 Nevere thought I'd see those health scams on IRC of all places 2024-10-27 03:10:45 never* 2024-10-27 03:12:32 vimicito: technically, that's on matrix 2024-10-27 03:15:03 Yeah, I suppose so... The [m], isn't it? There's quite a few of them in #kde at Libera too, because the KDE project runs a relay. Not sure if that was a good thing or not... 2024-10-27 04:51:55 do you guys have a tool, or script, to get latest package version number in repository from "repo-branch/arch" combo? 2024-10-27 04:53:24 or should i parse APKINDEX file APKINDEX.tar.gz 2024-10-27 05:35:42 awk '/^P:linux-virt$/{getline; split($0, a, ":"); print a[2]}' APKINDEX 2024-10-27 05:35:53 6.6.58-r0 2024-10-27 05:36:12 they only thing i use chatgpt for ... is awk 2024-10-27 14:11:57 can anyone help me with nfs permissins, i added a mount to fstab put the mount point is owned by root and i cannot access it with another user 2024-10-27 14:49:14 i can access files inside the mount by creating a mount bind for each folder, but it would be nice to just be able ta accees the root folder 2024-10-27 14:57:05 > rootbld assumes your APKBUILD file is in the `~/aports` whose structure like aports 2024-10-27 14:57:05 Hello I've an issue with `abuild-rootbld`. I've clone the aports repository at the root of my user home directory that will build my packages. However unlike what is quoted from the documentation `abuild rootbld` do not try to fetch the aports directory at all but just a relative APKBUILD. 2024-10-27 14:57:05 I've tried setting up APORTSDIR variable without success... 2024-10-27 14:59:14 Oh nevermind I think I'm completely confused. I thought this was meant to be used for building all packages but maybe not. The opposite would make more sense. 2024-10-27 14:59:24 Yeah, it just builds a single package 2024-10-27 15:10:54 thx for the confirmation 2024-10-27 16:01:55 Should we add ping to bbsuid? 2024-10-27 16:25:55 "Should we add ping to bbsuid?" <- bbsuid? bubblewrap-suid? 2024-10-27 16:35:41 SyntheticBird[m]: sorry i wrote that sentence, hope you can help fix it if not clear :) 2024-10-27 16:36:43 SyntheticBird[m]: busybox-suid 2024-10-27 22:24:28 hey guys, does anyone here run Alpine on a raspberry pi 5? I'm trying to get the os running on RAM and have a second partition persistent with docker and its containers. I'd have few questions please 2024-10-27 22:47:44 ask your questions, if someone knows the answers, they'll probably speak up 2024-10-27 22:53:42 Unless they're really evil and want to keep the knowledge all to themselves 2024-10-28 11:19:13 sorry if that's a super trivial question, but how do i enable the https://archlinux.org/packages/extra/x86_64/qemu-user-static/ "extra" repository to be able to install this package? my /etc/apk/repositories has "main" and "community" and i can't see an "extra" folder here: https://dl-cdn.alpinelinux.org/alpine/edge/ 2024-10-28 11:21:19 el[m]: archlinux != alpinelinux 2024-10-28 11:24:51 oh oops, my bad for not noticing somehow 😂 so does alpine not have qemu-user-static? i want to use it with docker for arm64 builds on x64 2024-10-28 11:25:02 (i must still be too asleep lol) 2024-10-28 11:25:42 i randomly tried a few qemu packages via apk that looked like they might contain it, but so far i didnt manage to install the right one 2024-10-28 11:26:00 It should be there, looking what provides it 2024-10-28 11:26:06 we use it ourselves 2024-10-28 11:26:47 I think qemu-aarch64 should suffice 2024-10-28 11:27:06 ikke: pov: you need to test your distribution on s390x or ppc64le or riscv but you don't have the cpu... 2024-10-28 11:27:27 SyntheticBird[m]: I understand what the usecase is 2024-10-28 11:28:21 ikke: hm interesting. i have that installed but docker just quietly ignores `DOCKER_DEFAULT_PLATFORM=linux/arm64/v8` for me and runs it as x64 instead (and then naturally complains that it cant run any of the executables in the image) 😶 i wonder what i'm messing up then 🤔 2024-10-28 11:28:41 ikke: that was rhetoric. I assumed you used it for testing alpine on other ISA. What is its usecase exactly ? 2024-10-28 11:29:00 SyntheticBird[m]: We used it to build rv64 on x86_64 before we had suitable hw 2024-10-28 11:29:21 ah i see 2024-10-28 11:29:43 for what it's worth, when i do locate qemu | grep static there is nothing showing up. so that's why i assumed i just dont have the required parts installed. but i got qemu-aarch64 and qemu-system-aarch64 2024-10-28 11:32:00 el[m]: https://raw.githubusercontent.com/qemu/qemu/master/scripts/qemu-binfmt-conf.sh 2024-10-28 11:32:38 This script should help, it sets up binfmt to use the qemu-system binaries 2024-10-28 11:33:09 or install qemu-openrc 2024-10-28 11:34:14 At least for us, qemu-riscv64 + qemu-binfmt suffices to run riscv64 docker and lxc containers 2024-10-28 11:37:11 i dont seem to have any binfmt package available 🤔 2024-10-28 11:37:37 qemu-openrc, sorry 2024-10-28 11:41:55 i was missing qemu-openrc, thanks! sadly it still doesn't seem to be running it as an arm64/v8 binary: https://paste.centos.org/view/raw/18127105 (i'm also on a postmarketOS/alpine host, but i also like to use it inside the image ☺️ ) 2024-10-28 11:42:28 sorry if i'm asking too many questions 😮 for some reason on opensuse this "just worked" so i'm not quite sure what's going on 2024-10-28 13:14:58 Hi 2024-10-28 13:15:01 $ doas apk -vv upgrade 2024-10-28 13:15:01 After this operation, 0 B of additional disk space will be used. 2024-10-28 13:15:01 1 error; 1836 packages, 12323 dirs, 168072 files, 9615 MiB 2024-10-28 13:15:07 What does 1 error mean? 2024-10-28 13:15:15 There's is no actual error on the output 2024-10-28 13:15:37 Is there a way to debug this, if -vv doesn't work? 2024-10-28 13:22:48 dunno what 1 error means, but usually 'apk fix' does fix the error and involves reinstalling a package 2024-10-28 13:28:57 :/ 2024-10-28 13:30:04 Is there a way to debug this 2024-10-28 13:30:18 Like having apk-fix not actually fixing it, but first verifying what's the issue 2024-10-28 13:55:56 quinq: does adding -i to apk help? Man isn't clear 2024-10-28 14:18:56 Yeah I suppose, it tells me that that it needs to reinstall a package 2024-10-28 14:20:14 el[m]: did you start the qemu-binfmt service? e.g. rc-service qemu-binfmt start 2024-10-28 14:21:45 el[m]: also, if you want it to load after reboots: rc-update add qemu-binfmt default 2024-10-28 14:31:37 "el: did you start the qemu-..." <- that seems to have fixed it, thank you 💜 2024-10-28 14:32:46 welcome 2024-10-28 14:37:08 quinq: agreed, -vv should show what's going on and meant by that error (against norms of what other things like ssh do) 2024-10-28 17:20:59 where is source tarball packages? 2024-10-28 17:21:08 when you use abuil checksum 2024-10-28 17:21:11 abuild 2024-10-28 17:21:33 * where is the source tarball of packages? 2024-10-28 17:22:18 abuild downloads source archives to /var/cache/distfiles 2024-10-28 17:45:16 how can i list the repo of a package? 2024-10-28 17:45:44 and if it is installed 2024-10-28 17:46:12 apk does not keep track of what repo a package was installed from 2024-10-28 17:46:40 when searching can it shows the repo? 2024-10-28 17:46:51 apk policy will show you what repo provides the package 2024-10-28 17:47:07 If it's installed, it will show lib/apk/db/installed as well 2024-10-28 17:47:21 apk version shows if it's installed as well 2024-10-28 17:48:49 apk policy apk no ouput 2024-10-28 17:49:20 the package is called apk-tools 2024-10-28 17:51:18 masked in: cache why my custom repo is that? 2024-10-28 17:51:58 i can never install from it 2024-10-28 20:17:13 "hey guys, does anyone here run..." <- I have alpine running on ram and using lbu to save any config. How can I set the second partition as / and install docker there, keeping the system on ram? thanks 2024-10-28 20:55:34 vimicito: fyi, https://www.alpinelinux.org/posts/2024-10-28-postmortem-edge-mirror.html 2024-10-28 20:58:24 seth[m]: for a start, mount some persistent thingie as /var/lib/docker 2024-10-28 21:01:20 I tend to have boot and then persistent stuff on overlay filesystems, /usr and /var at least 2024-10-28 21:05:46 @artok Thanks! so sorry, can you be more specific? I've literally just started with Alpine 2024-10-28 21:06:48 https://www.perrotta.dev/2022/01/alpine-linux-on-raspberry-pi-diskless-mode-with-persistent-storage/ 2024-10-28 21:06:57 like that 2024-10-28 21:10:26 artok: here you go fren https://wiki.alpinelinux.org/wiki/Raspberry_Pi 2024-10-28 21:11:07 seth[m]: ^ alpinelinux own wiki 2024-10-28 21:12:44 I just tend to create one large filesystem and mount it to /mount/data and then under that have /mount/data/var and /mount/data/usr, and mount those directories as overlays, rather than have some img files created 2024-10-28 21:13:16 @aleksi_ @artok thanks both, I've read that wiki, it didn't work for me 2024-10-28 21:13:38 helps to have the whole filesystem space to be shared, rather than img file size as max 2024-10-28 21:14:04 @artok great. I do have already a big XFS partition 2024-10-28 21:18:15 @artok I'll try then. How do I set those mounted partitions to be the default ones? 2024-10-28 21:22:13 should go with those docs 2024-10-28 21:22:32 first mount the filesystem to /mount/ 2024-10-28 21:23:18 create directories under that, usr .usr var .var 2024-10-28 21:24:33 then in fstab you add first the filesystem to /mount/ and after that "overlay /usr overlay lowerdir=/usr,upperdir=/media//usr,workdir=/media//.usr 0 0" 2024-10-28 21:24:48 aand then overlay line also for var 2024-10-28 21:25:37 and lbu ci for saving that fstab 2024-10-28 21:37:51 @artok thanks. That sounds complicated, perhaps it is not. hope you don't mind asking some further questions if I got stuck 2024-10-28 21:54:39 Are there any alive people? 2024-10-28 21:55:17 only unalive 2024-10-28 21:55:26 i see 2024-10-28 21:55:58 noo 2024-10-28 21:55:59 aliven't 2024-10-28 21:57:51 mick_ibm just forcely close terminal 2024-10-28 21:59:09 I don't think they need this kind of advice 2024-10-28 23:36:03 ikke: Thank you! I'll read it now. 2024-10-28 23:49:15 Ahh, shell scripts and bashisms... No matter how resilient we make our infrastructure, it's the humble shell that ties it all together. One variable that decided to become Oppenheimer huh. Not gonna lie, I do find it a bit amusing :) 2024-10-28 23:52:17 I like the plug that's been put into this for now. Small changes and see what happens. Infrastructure need not move fast and break things. 2024-10-29 00:11:28 Currently I'm looking at the aports-build script, and trying to figure out how that fits into the build infrastructure. Do all of the machines in the illustrated chain (but specifically the builders and their master) run Alpine? Do they have bash installed, or do they use the default (which if memory serves me right is ash)? 2024-10-29 00:54:00 Over years of neglect really, I've also noticed that Alpine LXC containers under Debian 12 broke. Seems like it's related to the key rollover documented in https://www.alpinelinux.org/posts/Alpine-edge-signing-keys-rotated.html . In Debian, I see that readonly APK_KEYS_URI='http://alpinelinux.org/keys' is defined in /usr/share/lxc/templates/lxc-alpine. But visiting that resource 2024-10-29 00:54:03 returns 403, regardless of TLS or not. What are the current keys used for Alpine? The files for the lxc-templates I'm looking at are available at https://packages.debian.org/bookworm/lxc-templates . 2024-10-29 02:11:56 An hour ago, I was looking at https://nixmagic.com/pub/Screenshot_20241029_021457.png . Since then, I have also identified various DNS-related issues in the local networks. Local networking aside, the differences between the keys listed on the upper right pane and those on the lower right panes may be useful. The host dhcpd0.lan was deployed several years ago, and has been 2024-10-29 02:11:59 updated today. The lower left pane meanwhile shows where I initially encountered the error. 2024-10-29 02:13:50 The exact deployment date for dhcpd0.lan was 2020-08-12. Heh, I became 24 that day. 2024-10-29 08:56:09 I'm trying to install py3-matplotlib. It depends on py3-numpy, but that has been updated to version 2, which matplotlib doesn't support. Is there a way to tell the package to use py3-numpy1, or to revert py3-numpy to a previous version ? 2024-10-29 09:08:00 hey, does anyone know why the hyprland package for alpine was removed? 2024-10-29 09:10:31 It's not 2024-10-29 09:12:27 where can i find it then? 2024-10-29 09:12:33 i just got a 404 when i looked on my phone 2024-10-29 09:22:19 404 for what? 2024-10-29 09:23:14 i just searched it up and the top result which was for the package was a 404 2024-10-29 09:24:16 https://pkgs.alpinelinux.org/packages?name=hyprland&branch=edge&repo=&arch=x86_64&origin=&flagged=&maintainer= 2024-10-29 09:26:00 ah, thanks 2024-10-29 13:45:14 hi all, I have installed Alpine Linux on a machine, using cryptsys, works just fine. In my quest to better understand how I can customize the boot process, I'd like to rebuild the initramfs using dracut. So far, I managed to get the passphrase prompt to show up for the encrypted partition, but after I enter it, it drops me in dracut emergency shell. I'm not that familiar with dracut, initramfs or the booting procedure. I tried 2024-10-29 13:45:15 getting some help providing more details in a lemmy post: https://lemmy.zip/post/24041939 If there's anyone who has some advice or some hints, I'd love to hear from you. I'm a bit lost at the moment. 2024-10-29 15:37:11 is it easy to use sim-card for internet on laptop with linux? 2024-10-29 15:40:27 frag: depends what kind of laptop and modem 2024-10-29 15:40:56 ptrc some are easy? which ones? 2024-10-29 15:41:22 considering testing it out, i dont have one with sim reader now 2024-10-29 15:41:39 i did this kind of setup mostly on a thinkpad, with the vendor-provided modem, and modemmanager just picked it up without any issues 2024-10-29 15:41:50 but then, it may be different if it's an external modem 2024-10-29 15:42:05 mostly stuff like usb_modeswitch 2024-10-29 15:43:50 ptrc oh, are the newer thinkpads good laptops in general? 2024-10-29 15:50:23 im using one right now, theyre pretty fine 2024-10-29 15:51:33 same, they're not quite "you can break a wall with it" build quality anymore, but the repairability is really nice, with manuals and a list of repair parts readily available with part numbers 2024-10-29 15:52:00 and the linux experience is comfy as well, as long as you switch the sleep mode into whatever windows uses 2024-10-29 15:52:07 or maybe the opposite? 2024-10-29 15:52:10 don't remember 2024-10-29 15:53:17 nice! 2024-10-29 18:28:42 is it possible to install alpine on a laptop to run from RAM, with just a small vfat to store config like on a raspi? i see the "data" option but it mounts to /var IIUC. I saw there is this issue 10474 too, is the install happening after the work-around?, because afaict setup-alpine doesnt even install a system when using the "data" option 2024-10-29 19:45:19 does default install of vis editor needs setting of theme, i see blurred text green green ? 2024-10-29 19:46:30 frag: the installer also has a dedicated run-from-ram option (can't recall exactly what it's called) 2024-10-29 20:35:08 is there anyway to prevent alpine to remove the kernel modules of the currently running kernel when upgrading? 2024-10-29 20:35:26 bOR38552MJA: hardlinking them somewhere else 2024-10-29 20:35:32 but further, no 2024-10-29 20:36:07 the correct behaviour woudl be to remove those modules only after booting into the new kernel 2024-10-29 20:36:51 this has bitten me so many times 2024-10-29 20:36:58 yes, but it's a bit complicated 2024-10-29 20:37:10 and most of the time i do a copy of the modules for exactly this reason 2024-10-29 20:37:55 but now i learned that i didn't the last time... and eh i wasted half a day to figure out why a usb cdc acm devices doesn't show up under /dev/tty* 2024-10-29 20:38:54 and now i have to rebuild the old kernel to get those modules back... 2024-10-29 20:39:00 or reboot 2024-10-29 20:39:15 not possible 2024-10-29 20:39:35 I'm not sure if you can use modules from a seperately compiled kernel 2024-10-29 20:39:47 worked before 2024-10-29 20:39:51 oh ok 2024-10-29 20:39:59 I thought there was some magic string that needed to match 2024-10-29 20:40:22 but yeah it would be nice to have access to previous kernel apks 2024-10-29 20:41:51 03> module verification failed: signature and/or required key missing - tainting kernel 2024-10-29 20:41:57 but otherwise works 2024-10-29 20:42:01 ok 2024-10-29 20:42:23 but damn, this is super annoying. i wish there would be a fix for this in alpine 2024-10-29 20:43:07 i wonder how many people have been bitten by this, and wasted hours and became frustrated by this. 2024-10-29 20:54:39 ikke yeah thats the data option 2024-10-29 20:55:13 Data still stores /var on disk, right? 2024-10-29 20:55:23 grep -B3 RAM /usr/sbin/setup-disk 2024-10-29 20:55:46 ikke yeah i got that impression 2024-10-29 20:56:46 You don't want setup-disk if you don't want to install to disk 2024-10-29 20:57:14 > Enter 'none' if you want to run diskless. 2024-10-29 20:57:26 That's the option 2024-10-29 20:58:38 ikke i dont think it installs anything anywhere then, it does need a boot medioum after all 2024-10-29 20:58:54 You are already booting from a boot medium 2024-10-29 20:59:04 It just needs an apkovl for persisting things 2024-10-29 20:59:49 yeah the iso/usb ... keep booting from that? 2024-10-29 20:59:55 exactly 2024-10-29 21:03:35 hmm might try it another day, where sill it store that i use eg /dev/sda1 as apkovl? 2024-10-29 21:03:59 s/still/will 2024-10-29 21:04:53 apkovl is a single file 2024-10-29 21:06:06 it will discover it on boot, or you can specify it in a kernel parameter 2024-10-29 21:06:53 it can change the iso 2024-10-29 21:06:57 ? 2024-10-29 21:07:00 No 2024-10-29 21:07:09 ISO is read-only 2024-10-29 21:07:24 yeah thats what im figuring 2024-10-29 21:08:26 so how does it store that the apkovl is eg at /dev/sda, wish it was not iso btw, but i guess its so its static, you cant mess it up 2024-10-29 21:08:49 the iso filesystem is read-only by definition 2024-10-29 21:11:08 APPEND modules=sd-mod,usb-storage,ext4,dm,crypt,rootfs-block rootfstype=ext4 rootflags=rw,relatime root=UUID=57955343-922a-4918-9bc1-797ca8d13a9c rd.luks.uuid=705fc477-573a-4ef6-81b6-a14c43cda1f5 2024-10-29 21:11:44 I have the above options for my boot entry, to unlock my luks partition, but I get dropped into the dracut shell after the passphrase prompt, any obvious mistakes I'm making? 2024-10-29 21:12:09 No experience with dracut 2024-10-29 21:13:16 yes i know the iso is ro, i guess im not communicating the q, maybe too tired -,- 2024-10-29 21:15:53 If you select none as disk option, then it will invoke setup-lbu 2024-10-29 21:17:35 yes, and i select '/dev/sda1' but how will it store that option so it know where to find it after reboot, when i boot from an iso 2024-10-29 21:18:20 nlplug-findfs will look for apkovl files on available storage 2024-10-29 21:19:23 i assume they must be mounted, and i cant change fstab on a iso 2024-10-29 21:19:41 Yes, it must be mounted, otherwise it cannot load it 2024-10-29 21:20:12 i give up ;/ 2024-10-29 21:20:54 what I'm trying to tell you is that you don't have to explicitly tell the boot media what disk to look for 2024-10-29 21:22:19 but i cant even mount something, since i cant edit fstab on an iso, so there is no mounted disks 2024-10-29 21:22:36 You don't have to 2024-10-29 21:23:26 i dont get it 2024-10-29 21:23:40 during boot it will discover all available storage, even if it's not mounted 2024-10-29 21:24:19 https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/nlplug-findfs/nlplug-findfs.c 2024-10-29 21:24:34 oh, if i make sda1 and ext4 and put the apkovl there, it will find it? 2024-10-29 21:24:41 yes, exactly 2024-10-29 21:24:49 oh! 2024-10-29 21:26:17 https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-alpine.in#L330-331 2024-10-29 21:26:42 https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-lbu.in#L177 2024-10-29 21:28:53 will have to try another day, thanks for the patience :) 2024-10-30 02:13:00 Hello! I was just checking out but is it a Alpine thing that /etc/nginx/conf.d was changed in /etc/alpine/http.d ? 2024-10-30 07:59:54 hi! i was browsing the wiki for how to run glibc programs and https://wiki.alpinelinux.org/wiki/Running_glibc_programs#GlibMus-HQ seems kinda fishy 2024-10-30 08:01:16 doesn't even have an alpine package 2024-10-30 08:01:21 anyone know what's the deal with it? 2024-10-30 08:02:23 it's a script from github 2024-10-30 08:03:08 Someone promoting their own solution 2024-10-30 08:03:38 (gitlab not github) 2024-10-30 08:04:11 it also appears to not be there 2024-10-30 08:27:24 oh vivaldi .deb runs perfectly fine with gcompat and --no-sandbox 2024-10-30 10:08:41 new alpine installation, php installed. is there a 'proper' way of having /usr/bin/php link to /usr/bin/php82 or is it a manual symlink? 2024-10-30 10:18:22 I can only think of the alternatives tools. 2024-10-30 10:21:37 i didn't see any for alpine 2024-10-30 10:37:30 /usr/bin/php is in a different package: 2024-10-30 10:37:31 https://pkgs.alpinelinux.org/contents?file=php&path=%2Fusr%2Fbin&name=&branch=edge&repo=&arch=x86_64 2024-10-30 10:37:49 it used to be in php82 in v3.19 2024-10-30 10:38:04 must be a packaging policy thing 2024-10-30 10:38:16 hrm 2024-10-30 13:43:04 If fx8800 comes back, typo perhaps, dm-crypt not dm,crypt? 2024-10-30 14:14:28 <__goldano__> hello 2024-10-30 14:16:36 greets 2024-10-30 14:26:33 Hey, anyone know if there are any packages for Postgres 17 in the works, or perhaps available on a third-party repo? 2024-10-30 14:29:31 https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/72853 2024-10-30 14:29:33 https://gitlab.alpinelinux.org/alpine/aports/-/issues/16490 2024-10-30 14:29:39 Apparently someone is working on it 2024-10-30 20:21:14 is it possible to add a writable partition to the install medium? seems it mounts it as sda, so it is busy 2024-10-30 20:34:11 frag: I'm not following. sda is a block device 2024-10-30 20:34:18 not a mount point 2024-10-30 20:34:46 The iso is read-only. If you want to create a bootable device which is writable, use setup-bootable 2024-10-30 20:39:24 if you boot the iso/usb it mounts /dev/sda to /media/sda iirc, so if i make a /dev/sda3, i cant be mounted (i assume its because /dev/sda is mounted instead of /dev/sda1 2024-10-30 20:40:05 oh setup-bootable seem nice! 2024-10-30 21:52:35 I'm fx8800 =) cyclisme24thank you for taking the time to look at my issue. I was following the instructions here: https://wiki.gentoo.org/wiki/Full_Disk_Encryption_From_Scratch#Dracut_module_config 2024-10-30 21:52:35 Could there have been an issue? 2024-10-30 21:55:18 kwantmaker[m]: > <@kwantmaker:matrix.org> I'm fx8800 =) cyclisme24thank you for taking the time to look at my issue. I was following the instructions here: https://wiki.gentoo.org/wiki/Full_Disk_Encryption_From_Scratch#Dracut_module_config 2024-10-30 21:55:18 In fact, I'm just going off the assumption that the module names are the same for dracut and the kernel parameters, but maybe I'm totally mistaken, let me check 2024-10-30 21:55:18 > Could there have been an issue? 2024-10-30 22:02:49 > <@kwantmaker:matrix.org> I'm fx8800 =) cyclisme24thank you for taking the time to look at my issue. I was following the instructions here: https://wiki.gentoo.org/wiki/Full_Disk_Encryption_From_Scratch#... (full message at ) 2024-10-30 22:11:15 > <@kwantmaker:matrix.org> I'm fx8800 =) cyclisme24thank you for taking the time to look at my issue. I was following the instructions here: https://wiki.gentoo.org/wiki/Full_Disk_Encryption_From_Scratch#... (full message at ) 2024-10-31 01:13:43 Does anyone know why gnu tar would get permission errors when doing `tar -c -C in . | tar -x -C out`? Busybox tar works fine. 2024-10-31 01:13:45 tar: ./usr: Cannot change mode to rwxr-xr-x: Operation not permitted 2024-10-31 05:13:45 is there any package that provides compiler-rt 18? seems the releases skipped right over it and edge is already on 19 2024-10-31 06:06:03 meghan: aports git history for llvm-runtimes (parent aport for compiler-rt): it was upgraded to 18.1.6 on Jun 11 and to 19.1.1 on Oct 11. perhaps there is a need for llvm${version_major}-runtimes. 2024-10-31 10:33:31 i updated my alpine asahi macbook today, and lightdm stopped working after a reboot 2024-10-31 10:34:04 this was an update after over a half year or so 2024-10-31 10:34:28 lightdm greeter works on my intel desktop, which is up-to-date 2024-10-31 10:34:56 did you update to a different release version? eg. 3.17 → 3.20 2024-10-31 10:35:03 edge -> edge 2024-10-31 10:35:48 i wonder if i should 1) fix lightdm + xfce, 2) try get gnome working or 3) try to switch to sway 2024-10-31 10:36:03 in any case, fixing lightdm and xorg is probably a good thing anyway 2024-10-31 10:36:39 i think this may be the problem: 2024-10-31 10:36:42 Failed to open PAM session: Permission denied 2024-10-31 10:36:42 cat /var/log/lightdm/seat0-greeter.log 2024-10-31 10:38:43 ok, seems like I used polkit-noelogind-libs 2024-10-31 10:39:28 # Setup session 2024-10-31 10:39:28 -session optional pam_elogind.so 2024-10-31 10:42:23 so i need to set up xfce with elogind i suppose 2024-10-31 10:51:30 ok, looks like elogind dies for some reason 2024-10-31 10:51:52 Oct 31 10:51:05 ncopa-mbp14 daemon.warn supervise-daemon[6717]: respawned "/usr/libexec/elogind/elogind" too many times, exiting 2024-10-31 11:01:07 bah, a simple reboot seems to have solved it 2024-10-31 12:09:17 Anyone on edge experiencing segfaults when trying to run celluloid ? ---> https://paste.debian.net/1334075/ 2024-10-31 13:35:11 Hi, when trying to run any of the software (e.g., kea-dhcp4) from package kea on the latest armhf version of Alpine for Raspberry Pi, I get an Illegal instruction error. Should that kind of error be reported to the Alpine issues at Gitlab? 2024-10-31 13:35:50 topte: yes 2024-10-31 13:36:44 Okay, will do. Cheers. 2024-10-31 13:56:41 as400: can't reproduce on x86_64 edge 2024-10-31 14:04:21 ptrc: what GPU do you have ? Mine is meteor lake arc. 2024-10-31 14:04:38 RX 6600 here 2024-10-31 14:05:00 oh, yes so probably gpu is the problem 2024-10-31 14:05:04 or mesa 2024-10-31 14:05:21 although vainfo says everything is ok 2024-10-31 14:07:51 lts kernel is too old for my gpu and edge is borked 2024-10-31 15:38:31 im strugling with getting networkmanager work with wpa_supplicant 2024-10-31 15:38:53 If I run wpa_supplicant as standalone service, with /etc/init.d/wpa_supplicant start, it works 2024-10-31 15:39:02 but then I need to run dhcpcd serivce 2024-10-31 15:39:12 and I get no nice gui to manage wifi connection 2024-10-31 15:39:33 but with networkmanager, it fails to connect to the wifi net 2024-10-31 15:45:45 the only difference I can see is that from networkmanager it starts wpa_supplicant with -u, which gives dbus interface 2024-10-31 15:45:52 but it fails to connect to anything 2024-10-31 15:58:41 after run `apk update`, i got v3.20.3-227-*, what does the number 227 mean? should the number of main and community be the same? 2024-10-31 16:12:50 That's the output of git describe 2024-10-31 16:14:01 It does not necessarily need to be the same 2024-10-31 16:19:53 thanks 2024-10-31 17:37:13 https://i.imgur.com/ykP4CVN.png 2024-10-31 17:46:35 Is there a LSP available for OpenTofu/Terraform in Alpine? I decided to finally learn using the tool and would love some syntax highlighting and LSP in Neovim 😅 2024-10-31 17:49:11 https://github.com/opentofu/opentofu/issues/970#issuecomment-2053984461 2024-10-31 17:49:23 Nothing available yet I suppose in alpine 2024-10-31 18:21:19 Hmm ok thanks, I'll keep an eye out for it 2024-10-31 21:01:36 After an admittedly long period between updates, an update today broke my Wifi, card Intel AX201. My dmesg output is here: https://bpa.st/74W5K Is anyone else experiencing this? eth0 is working fine. 2024-10-31 21:51:04 bovis: which linux-firmware packages do you have installed? the iwlwifi firmwares are in the linux-firmware-other subpackage. 2024-10-31 21:52:00 j_v: apk lists: linux-firmware-20240909-r2 x86_64 {linux-firmware} (custom) 2024-10-31 21:52:46 ahhh, you are running custom, does your kernel support zstd compressed firmware? 2024-10-31 21:54:34 j_v: Whoops, full list of linux-firmware*: https://bpa.st/DLKAI 2024-10-31 21:54:53 j_v: the *other package is included in that list 2024-10-31 21:54:55 look for CONFIG_FW_LOADER_COMPRESS_ZSTD=y in your /boot/config-${version} 2024-10-31 21:56:19 given that linux-firmware-other is listed, my guess is the zstd support 2024-10-31 21:56:28 in the kernel 2024-10-31 21:56:40 I don't see that line in my config, which is linux-lts 2024-10-31 21:56:49 There's nothing custom going on here that I've done 2024-10-31 21:56:57 what version is linux-lts? 2024-10-31 21:57:15 I *was* working yesterday; then I updated; now it's not 2024-10-31 21:57:39 linux-lts-5.15.52-r0 x86_64 2024-10-31 21:58:04 that is really old, what alpine version are you on? 2024-10-31 21:58:57 j_v: Edge 2024-10-31 21:59:37 That is old... 2024-10-31 21:59:44 Looks like 6.6 is current 2024-10-31 21:59:58 edge is on 6.6.58 so i am betting that kernel does not support zstd compressed firmware, which was added not so long ago 2024-10-31 22:00:38 A good question is then, How am I stuck on 5.15? 2024-10-31 22:01:30 apk policy linux-lts might shed some light 2024-10-31 22:02:09 or manually looking in /etc/apk/world to see if it got pinned somehow 2024-10-31 22:02:26 j_v: Sorry again. /boot/ lists the lts and also config-6.6.58-1-virt, where the ZSTD=y line is present 2024-10-31 22:02:53 uname -s 2024-10-31 22:03:20 ^^ Linux 2024-10-31 22:03:45 sorry -r 2024-10-31 22:03:52 apk policy: https://bpa.st/WWM6Y 2024-10-31 22:04:13 The -r is 5.15.52-0-lts 2024-10-31 22:05:06 yep, might need to look at you bootloader config to see why it's booting the older kernel 2024-10-31 22:06:09 looks like 5.15 kernel was installed manually, so heart of issue is getting so you boot the current kernel, at least so it seems to me 2024-10-31 22:07:43 if i was you, i wouldn't remove the old kernel until you get to the bottom of the reason the old kernel is being booted instead of the newer one 2024-10-31 22:10:11 Safe to put UUIDs in a paste? 2024-10-31 22:13:18 should be fine i would think, it's just for drive id, right? 2024-10-31 22:13:53 Yeah. Looking at the boot config. 2024-10-31 22:14:11 which bootloader are you using? 2024-10-31 22:14:31 Syslinux 2024-10-31 22:15:34 https://bpa.st/F5CVQ 2024-10-31 22:15:53 i am most familiar with sys/extlinux, so that helps... i wouuld look at /etc/update-extlinux.conf also 2024-10-31 22:18:34 That config is also unremarkable, it seems: https://bpa.st/E4XUC 2024-10-31 22:19:20 Would either of those specifically mention the 5.15 version being held? Clearly I'm getting the updated 6.6-virt downloaded when I update. 2024-10-31 22:19:21 weird, usually when there is no DEFAULT set, it defaults to booting the first entry... makes me wonder which kernel is which, because one is lts and other is virt 2024-10-31 22:22:33 looks like the lts is 5.15, the 6.6 must be the virt kernel 2024-10-31 22:24:09 'apk policy linux-lts' showed 5.15 as lib/apk/db/installed. Is this something apk is doing? Keeping the version at 5.15? Or is another package holding it back (I don't remember how to find that out). 2024-10-31 22:24:10 i would edit the update-extlinux.conf, change the 'hidden' setting to 0 so you will get a boot menu, maybe up the timeout to give you more time to choose 2024-10-31 22:25:12 i don't think so, but it looked like a manual install (given the age, might have been some time ago) 2024-10-31 22:26:35 i am guessing you are running xen? this is something i have no experience with, so i don't know if there are any caveats 2024-10-31 22:26:35 The /boot/config-lts was written in 2022 2024-10-31 22:28:29 yeah, not surprized, and given that the default ordering for update-extlinux, the lts kernel gets higher precedence just based on alpha sort 2024-10-31 22:28:31 The hypervisor? Not that I remember. Though I had qemu installed for a while. That's been uninstalled. Maybe there's something leftover? 2024-10-31 22:29:17 could be, but the recent kernel is vmlinuz-virt, or so it looked like 2024-10-31 22:29:50 not sure how that will behave on bare metal 2024-10-31 22:30:34 I don't want to try to boot 6.6-virt and end up with something I can't handle 2024-10-31 22:31:29 but if you don't want to brick or semi-brick you setup, don't uninstall that 5.15 kernel until you have booted into a newer kernel 2024-10-31 22:35:05 Trying it now 2024-10-31 22:43:48 j_v: 6.6-virt booted, but it was limited. Networking didn't improve (it didn't find wlan0); I couldn't get into X server. 2024-10-31 22:48:31 i was going to suggest 'apk upgrade -s linux-lts' then if that went ok, again without the '-s' (simulate). not surprised that the virt kernel not working 100% on bare metal. i may be being too cautous, but it wouldn't hurt if you make sure that you have an alpine install medium so that you can go in a do it via chroot from the install environment, worst case 2024-10-31 22:50:58 The simulation is unremarkable; says the lts will be upgraded. 2024-10-31 22:51:26 Never hurts to have a recovery, so I'll make one. If this upgrade works, won't I continue to see the effects of a sticky version? 2024-10-31 22:52:16 wouldn't think so, because it is the lts version. still curious that it isn't getting upgraded 2024-10-31 22:55:13 On a whim, I looked for more qemu-associated packages. There were a handful, including virt-manager, which I don't need. So I uninstalled all qemu-* and virt-manager. 2024-10-31 22:55:29 Then apk upgrade gives linux-lts the pass. Updated to latest version. 2024-10-31 22:55:53 It never occurred to me that the virt would hold back my normal system. 2024-10-31 22:57:11 yeah, not sure what exactly is the culprit here, but it's like linux-virt is the auto-installed kernel, like in a vm 2024-10-31 22:57:55 really weird that it's not upgrading linux-lts, which i why i was looking for something pinning it 2024-10-31 22:58:41 Trying another restart to see if that fixes the wifi issue. 2024-10-31 23:05:22 j_v: That did the trick. 2024-10-31 23:05:34 I really appreciate your help. Thanks for talking me through that. 2024-10-31 23:06:55 sorry it took so long 2024-10-31 23:07:36 but you are welcome, glad to help 2024-10-31 23:07:45 It's always me thinking it's something small, actually being big, and me causing it by doing something [a long time ago] 2024-10-31 23:09:06 been there, too... it's usually something i caused in some way, though not always obvious or something that i can pinpoint