2025-02-01 03:50:02 latest virtualbox guest additions are at 7.0.22. I have upgraded my host to run vbox 7.1.6, which is the latest, but I don't know how to get the GAs for it. Is there a way to generate the GAs from the vbox GA ISO? 2025-02-01 03:50:29 Or is it just a matter of patience until the package maintainer gets to it? 2025-02-01 03:50:39 (I'm new here, obviously!) 2025-02-01 04:41:42 "How is intel arc on linux rn?" <- Solid on linux-lts šŸ¤· I game on Steam, Lutris, Wine, native... 2025-02-01 04:42:12 Not sure I have hardware av1 enc working yet, but yeah 2025-02-01 12:49:42 Hello :) alpine is missing an init.d file for nfsdcld, which is required for a nfs4 server. Without it, a kernel fallback is used, but its use is not recommended. 2025-02-01 12:50:25 I was thinking to contribute one if that's ok? 2025-02-01 12:54:20 ACTION hi 2025-02-01 13:04:26 why not? 2025-02-01 13:07:40 Not sure. :) 2025-02-01 13:08:04 Maybe there's a reason why it wasnt included? 2025-02-01 13:09:10 might be worth checking its commit history 2025-02-01 13:20:33 isnt it included in the general nfs.conf? 2025-02-01 13:25:07 Forza 2025-02-01 13:25:08 $ apk info -L nfs-utils|grep -F cld 2025-02-01 13:25:08 usr/sbin/nfsdcld 2025-02-01 13:25:08 usr/sbin/nfsdclddb 2025-02-01 13:25:25 Oh sorry, you're talking about a service file 2025-02-01 13:29:48 quinq: indeed, just the initd file. 2025-02-01 13:47:55 Seems imy abuild is having difficulties... https://gist.tnonline.net/4E 2025-02-01 13:48:09 /usr/bin/abuild: line 70: ./configure: not found 2025-02-01 13:54:42 Saijin_Naib[m]: Are you having an issue with ACF or with understanding AWall? I can help with the former. 2025-02-01 13:54:45 ACTION forgot 'unpack' 2025-02-01 14:08:34 Ok, created a MR for nfsdcld https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/79286 2025-02-01 14:10:57 "Saijin_Naib: Are you having an..." <- Understanding awall, and UX want for UFW rules to be supported/converted so I can manage family devices easier 2025-02-01 14:11:22 It is all on me. As far as I can see, awall works great. I'm just not familiar so it is hard to learn 2025-02-01 15:22:25 Hi, did not find article about logcheck from wiki and as there were some manual steps involved, thought I would write one... however, any ideas which would be best practice to allow logcheck to actually access the logs? OOB seems syslog logs are owned by root:wheel and not sure if group wheel should be added to logcheck or should syslog be ran as group adm and then grant adm group for syslog (did this on my setup as distro I was coming 2025-02-01 15:22:25 from did this)? 2025-02-01 15:29:57 wheel might make most sense 2025-02-01 15:31:45 http://www.catb.org/jargon/html/W/wheel-bit.html 2025-02-01 15:37:15 not sure what wheel group actually does in alpine linux, but if it grants any admin level rights i would be hesitant to advice adding it to logcheck... in theory some buffer overflow exploit could perhaps allow to use syslog to log something which will cause logcheck user to run custom commands 2025-02-01 15:43:46 log files should have wheel group read permission only 2025-02-01 15:44:32 wheel isn't root 2025-02-01 19:48:29 Saijin_Naib[m]: I would be able to help if the issue was ACF. But, it's really just a dumb front end for awall. 2025-02-01 19:49:14 https://wiki.alpinelinux.org/wiki/Zero-To-Awall 2025-02-01 19:49:33 https://wiki.alpinelinux.org/wiki/How-To_Alpine_Wall 2025-02-01 19:49:48 Question I've had for quite a while, all of my alpine installs (been daily driving since 2021), is there a reason I can just open a tty and log in as root without a password, even after setting a password 2025-02-01 19:51:01 apartgorgon9849[m]: That's not expected 2025-02-01 19:51:38 That was my assumption, but I haven't found a way to fix it 2025-02-01 19:52:13 apartgorgon9849[m]: how are you installing alpine? 2025-02-01 19:58:13 ikke: do you want me to send every command or link a pastebin? 2025-02-01 20:00:52 No, more like, what image, what kind of installation 2025-02-01 20:03:27 3.18.2 - 3.21.2, standard image have all had the same behavior. I do a far more manual installation since i use an xfs root partition 2025-02-01 20:03:55 i havent tried with a different image type 2025-02-01 20:06:33 ikke: ^ 2025-02-01 20:12:17 Would be nice if you could provide a way to reproduce it 2025-02-01 20:13:07 ACTION posted a file: (0KiB) < https://matrix.org/oftc/media/v1/media/download/AfJzZseMGnBJO-3EBv3R_nCRAUcxxXHt6SzZdRilgF4PqTl--xSbJalp8URvu8o269SU5zeyi1EUuabmBz9gsi1CeVCz5OogAG1hdHJpeC5vcmcvekVFR2F5RFNIVElua0JNc0dtSVZ0TlJZ > 2025-02-01 20:13:28 those are my exact steps im taking to install alpine 2025-02-01 20:24:50 i just realized i never added the line to mount my boot drive in my steps 2025-02-01 20:28:42 I booted alpine-rpi-3.21.2 for the first time and even though the network is working, I can't install new packages 2025-02-01 20:29:01 I'm looking on the wiki and I can't find any help 2025-02-01 20:29:35 https://wiki.alpinelinux.org/wiki/Raspberry_Pi says I should be able to do apk update, apk upgrade 2025-02-01 20:29:49 implying I should have access to remote repos by default 2025-02-01 20:37:38 https://wiki.alpinelinux.org/wiki/Diskless_Mode also implies I should already have access to remote repos 2025-02-01 20:39:16 what does apk update output for you? 2025-02-01 20:40:06 I can't remember exactly, and I don't have access to the physical computer where I am right now 2025-02-01 20:40:36 but basically it looked like it was trying to update using just the local package cache 2025-02-01 20:42:31 that would be the first step since hard to say without... i just installed my first alpine in diskless mode and it just works... if network is up (you can ping e.g. google.com) and repositories have been set up properly it should work, but perhaps repositories have not been set up if you saw only local package cache... it should look something like this if repos are ok: cube:~# apk update 2025-02-01 20:42:31 fetch http://dl-cdn.alpinelinux.org/alpine/v3.21/community/x86_64/APKINDEX.tar.gz 2025-02-01 20:42:31 fetch http://dl-cdn.alpinelinux.org/alpine/v3.21/main/x86_64/APKINDEX.tar.gz 2025-02-01 20:42:31 fetch https://dl-cdn.alpinelinux.org/alpine/edge/testing/x86_64/APKINDEX.tar.gz 2025-02-01 20:42:32 3.21.2 [/media/alpine/apks] 2025-02-01 20:42:43 ACTION posted a file: (0KiB) < https://matrix.org/oftc/media/v1/media/download/ARf_NyN-3pznBfyWK5GBZe3zcnZKM9qsrzeH3qnuNp7-FrD7OkRFFQFGBvR31vCGcMMkv63-4Qid7XcdovNHmSFCeVC1lnDgAG1hdHJpeC5vcmcvT3F4c2dJbkZhbkJiVVBRb2NRbGxzbXdj > 2025-02-01 20:42:44 ikke: just did a brand new install in a vm with my steps, and it is producible with my steps 2025-02-01 20:43:39 Can you check the contents of /etc/passwd for the root user after installation? 2025-02-01 20:44:24 I mean, im never prompted to change the root password, as it is, but even setting a password doesnt do anything 2025-02-01 20:44:31 but yes i can 2025-02-01 20:44:34 triteraflops: perhaps something went wrong when setting up repos... check contents of /etc/apk/repositories and try to run setup-apkrepos if it lacks repos 2025-02-01 20:45:11 apartgorgon9849[m]: and perhaps also check /etc/shadowd 2025-02-01 20:45:14 /etc/shadow 2025-02-01 20:45:18 I just read https://wiki.alpinelinux.org/wiki/Repositories#Managing_repositories . That's exactly what it suggests 2025-02-01 20:45:28 jarp: ^^^ 2025-02-01 20:45:56 But that should be mentioned in the Diskless Mode an Raspberry Pi articles! 2025-02-01 20:46:24 That is a huge oversight in the wiki. 2025-02-01 20:47:24 apartgorgon9849[m]: settings the root password is done by setup-alpine itself 2025-02-01 20:47:32 Did the boot scripts try to run that command on startup? 2025-02-01 20:47:46 Because my networking wasn't working right away. It required extra config. 2025-02-01 20:48:06 apartgorgon9849[m]: so you want to do that before running setup-disk 2025-02-01 20:48:14 So after setting a password on my fresh vm, it prompts for the password as intended, but it still doesnt on my pc (with a root password set) 2025-02-01 20:48:34 it should be in setup-user in my opinion 2025-02-01 20:48:52 * in setup-user (setting a root password that is) in my 2025-02-01 20:49:18 apartgorgon9849[m]: can you also check /etc/inittab 2025-02-01 20:49:58 in postinstall i assume? 2025-02-01 20:50:23 yes 2025-02-01 20:50:26 this is in postinstall ikke 2025-02-01 20:50:28 ACTION uploaded an image: (13KiB) < https://matrix.org/oftc/media/v1/media/download/AQB8f57WGng_5Fio-6LbGzXKjV0oNlfZ1l4W_ljql3eNHAu3Nsx2azw4srD0Mzk-FU-SCl2TfwiBdZWGE58T_khCeVC2B_XgAG1hdHJpeC5vcmcvSmpyRWRiek1lc1RQdVpMa0RJdUtueFFx > 2025-02-01 20:50:56 looks normal 2025-02-01 20:52:12 triteraflops: well I too found docuementation a bit lacking or confusing when i did diskless x86_64 installation so yeah i feel you :( 2025-02-01 20:54:13 triteraflops: talking from my ass here as i have not tried rpi, but did you do setup-alpine at any point? 2025-02-01 20:54:34 That's for full disk install, isn't it? 2025-02-01 20:57:31 triteraflops: i tried to read through the docs, they are a bit confiusing but i would think you would need to do steps from Installation in any case (correct me if I am wrong)... and at least when doing diskless x86_64 installation setup-alpine was very much needed as it was the one which set up repos etc. 2025-02-01 20:58:20 Was that to make a diskless install in an empty partition? 2025-02-01 20:59:08 In the case of rpi, almost all of the diskless setup is complete when the sdcard is written with the os image 2025-02-01 20:59:34 All I'm doing now is adding an overlay for settings and extra packages 2025-02-01 20:59:47 (an apkovl, via lbu) 2025-02-01 21:02:07 That's a good question, actually. Do I use lbu to add packages to a diskless install? How do I do that? 2025-02-01 21:04:43 triteraflops: my understanding is based from to documentaiton (i dont have my rpi here atm so i cannot try today) that rpi tarball, as it stands, is functioning diskless installation, however setup-alpine needs to be ran so it will setup your keymap, repos etc. and finallu lbu commit will create overlay file so next boot should have everything ready (again somebody correct me if i am wrong) 2025-02-01 21:05:53 whoopsie doodle, jarp 2025-02-01 21:06:02 the documentation does say this lol 2025-02-01 21:06:06 triteraflops: to add packages to diskless installation, just do apk install whatever and then lbu commit... packages you install will be added to /etc/apk/world file and lbu commit then will create new overlay file containing those changes... on next boot added packages will be then installed automatically (preferrably from cache) 2025-02-01 21:07:35 The cache is on the fat32 rpi boot partition. Is that where lbu writes new packages? Because it's pretty full 2025-02-01 21:10:00 triteraflops: at least on x86_64, when i have usb stick and one boot partition, packages will be stored to that partition to cache directory so i guess yeah... but cache can be relocated afaik by using setup-apkcache but don't ask me if it just works if you mount another partition and use it or is something else needed 2025-02-01 21:13:23 In diskless, what are some good options for keeping changes to the home directory? 2025-02-01 21:13:50 I really want the ability to cut power to the rpi at any time without data loss 2025-02-01 21:13:52 (as a side not i freaking love diskless mode, with persistent storage, on my homelab server as i run containers and vms only... no longer i have to be afraid of messing up host os by doing some misconfiguration and then it won't boot and have to start cumbersome recovery session... now i can just restore previous overlay and even better, just experiment and never commit and reboot to go back again) 2025-02-01 21:15:51 triteraflops: personally i have ssd with one ext4 partition containing /var /opt and /home which i bind mount to /... not sure if this was the official way though 2025-02-01 21:16:47 I thought about it, and I think I'd rather do an lbu commit command if I want to keep changes to home 2025-02-01 21:17:12 There won't be very much data in home 2025-02-01 21:17:26 (Just SNES savegames lol) 2025-02-01 21:17:58 yeah then lbu include /home should do it i think? 2025-02-01 21:18:29 That's not on by default? hm 2025-02-01 21:19:03 anyway, that way, people will be able to unplug the rpi at any time 2025-02-01 21:21:09 default is /etc (/etc/init.d exluded)... people also could download huge files to their homedir and then lbu commit and run out of space or boot starts to take long time etc... i guess it is matter of preference but i would not include /home by default in diskless installation because of that... 2025-02-01 21:21:37 makes sense, I guess 2025-02-01 21:21:48 in general, /home is huge and changing 2025-02-01 21:22:28 especially if you enable lbu backups then disk space really starts to run out 2025-02-01 21:56:55 So, tbc, packages are added to the package cache with lbu, right? 2025-02-01 22:02:16 I've been reading https://wiki.alpinelinux.org/wiki/Alpine_Package_Keeper#Local_Cache and I really really don't understand how packages are added to the cache. 2025-02-01 22:02:25 Is it done automatically by apk add? 2025-02-01 22:07:36 apkdownloads package to cache and adds it to /etc/apk/world and lbu will then add changed world to overlay so on next boot packages are re-indtalled from the cache 2025-02-01 22:08:16 That's one thing the documentation might have mentioned lol 2025-02-01 22:08:29 I was wrong about the rpi docs, which seem fine 2025-02-01 22:08:34 This one, though... 2025-02-01 22:33:41 Is there anything I'm missing trying to install Alpine to my RPi 5? I keep trying to run setup-alpine to install it to LVM on LUKS and I don't think it's writing anything, when I try to boot it's just a blank screen with a cursor 2025-02-01 23:04:14 "Is there anything I'm missing..." <- Maybe you need to install firmware? (uefi) 2025-02-01 23:16:49 leg7[m]: how would I go about doing that? I was trying to install it from the SD card after flashing it with the official Pi imager tool and don't think there was any option for that there or in Alpine's installer 2025-02-02 00:26:58 I'm using the Alpine docker container and I'm trying to install laminar from edge/testing using the following: apk add laminar capnproto --repository=http://dl-cdn.alpinelinux.org/alpine/edge/testing/ 2025-02-02 00:28:23 But I get a response about missing so dependences which are all provided by capnproto. Capnproto installs fine by itself. Am I don't something wrong to pull in edge packages and dependencies or could that laminar package be broken? 2025-02-02 00:30:47 huh, it works fine if I use the alpine:edge image. ... I think that's what I'll do. Thanks for being my rubber šŸ¦†! 2025-02-02 03:19:49 Forza: did you by chance setup id-mapping on your nfs-server(s) 2025-02-02 04:35:22 Greetings :3 2025-02-02 07:39:00 longnoserob[m]: not userid mapping, no 2025-02-02 07:54:09 "longnoserob: not userid mapping,..." <- ok, thanks, then I need to continue my digging 2025-02-02 07:57:39 Np. Good luck. 2025-02-02 08:41:08 "leg7: how would I go about doing..." <- for my pi 4 I have to copy over uefi firmware to the boot partition after installing the os 2025-02-02 08:41:32 https://github.com/pftf/RPi4 2025-02-02 08:41:41 maybe it's different for the pi 5 tough 2025-02-02 08:42:03 just trying to give you clues because it was a pain for me to get it working 2025-02-02 08:42:59 https://github.com/worproject/rpi5-uefi?tab=readme-ov-file 2025-02-02 08:44:39 Also I think the PI doesn't support grub 2025-02-02 08:44:50 so maybe that is another problem 2025-02-02 08:45:55 https://github.com/raspberrypi/rpi-eeprom 2025-02-02 08:47:44 But I think grub might work if you install a UEFI 2025-02-02 08:48:54 https://wiki.nixos.org/wiki/NixOS_on_ARM/Raspberry_Pi_5 2025-02-02 08:49:13 Read this it will be more comprehensive than my shoddy memory 2025-02-02 11:41:21 Bit if a weird one: I now have virtually identical installs of 3.21 on a Pi4 and a Pi5 (same world) but this is a difference: on the Pi5 I don't see any optput from OpenRC, it just looks like the boot has frozen then it goes into lightdm. 2025-02-02 11:42:00 I have been doing recursive diffs of /boot and /etc and can see no significant difference 2025-02-02 11:43:10 What could prevent OpenRC from producng output? 2025-02-02 11:46:10 how do the kernel cmd lines look like? 2025-02-02 11:46:38 longnoserob[m], root=UUID=f25f8840-917c-4e41-a261-13d6a15b3964 modules=sd-mod,usb-storage,ext4 rootfstype=ext4 2025-02-02 11:47:33 that is coered by the recursive diff of /boot, but I wanted to show you there was no -quiet in there - in any case I see kernel messages, just not OpenRC 2025-02-02 11:49:33 I have been making thw two installs as near-identical as possible 2025-02-02 11:50:29 hi everyone 2025-02-02 11:52:00 m4r35n357: did you check if one device maybe logs to uart? 2025-02-02 11:53:42 can someone help me with the diskless mode? I'm using rpi 4b, but I don't have enough size for my packages on my boot partition. Is there any way to expand it? 2025-02-02 11:53:52 longnoserob[m], not as such (IOW not sure what tht means), but there is nothing in /etc that I am aware of to do that 2025-02-02 11:55:00 longnoserob[m], I am comparing /etc recursively remember ;) These are a couple fo diffences (like zram config), but _much later_ in the boot sequence 2025-02-02 11:55:23 I will equalize that too next time around. 2025-02-02 11:56:29 longnoserob[m], the only difference I am aware of ATM is the _hardware_ 2025-02-02 11:57:01 m4r35n357: Almost same pacakges installed? 2025-02-02 11:57:32 longnoserob[m], same world 2025-02-02 11:58:25 maybe there are differences in the kernelconfig regarding fbconsole or similiar 2025-02-02 11:58:42 longnoserob[m], I am probably in a minority of one here using a RPi, never mind two diferent ones ;) 2025-02-02 11:59:32 longnoserob[m], yes, I can't confirm checking, but that was one reason I make the installs identical 2025-02-02 12:01:17 @ 2025-02-02 12:01:30 # 2025-02-02 12:01:40 $ 2025-02-02 12:01:42 % 2025-02-02 12:01:46 haha oops 2025-02-02 12:02:39 Hello! I just swapped to a different motherboard/CPU for my Alpine installation (set up with UKI+secureboot+FDE), and I'm running into an issue where boot appears to get stuck on "Loading hardware drivers" with mdev. It's inconsistent, happening (rough estimate) ~50% of the time. This did not happen on the old mobo/cpu. I can't seem to find any logging of the previous boot when the system does successfully boot. Anyone know how to get 2025-02-02 12:02:40 started troubleshooting this? 2025-02-02 12:02:59 longnoserob[m], ooh, looking at the /boot diff, I notice that initramfs-rpi are different! didn't notice that before. I cloned the SD card from the RPi 4 to the 5, so the installer ran on the 4. Maybe that is the source fo the problem . . . 2025-02-02 12:04:04 longnoserob[m], I am still investigating the gory details fo boot, but I assumed OpenRC ran _afer_ the pivot_root, so not in the ramdisk, but I shall invesitgate! 2025-02-02 12:04:50 I'd just reinstall the linux package (to get proper initrd) 2025-02-02 12:06:05 quinq, sounds dangerous ;) But I don't think I have a choice. 2025-02-02 12:07:34 Running RPI is inherently dangerous, you're an adventurer anyway 2025-02-02 12:11:35 quinq, well apart fom this detail, it has been a rather simple adventure - the only thing I needed help for was to get Xorg to pick the right video output 2025-02-02 12:12:25 :) 2025-02-02 12:15:29 quinq, thinking about it, both machines were upgraded from 3.20 to 3.21, the kernel going from 6.6 to 6.12. I think that means that each machinealready has the right ramdisk, and they are different, so I'm not in a rush to mess with kernel reinstalls just yet 2025-02-02 12:16:35 quinq, an dnow I'm not sure what I just wrote makes sense, unless the upgrade knows somehow about the different hardare - hmmm . . 2025-02-02 12:23:17 quinq, I think I need to unpack both ramdisks on another machine and do another recursive diff . . . 2025-02-02 12:35:24 Hello, I think the last update of iptables (1.8.11-r0) broke my Docker setup. I need to run sudo iptables -P FORWARD ACCEPT for docker-proxy to receive packets from the outside again. Is that a known issue? 2025-02-02 12:40:07 Possibly related to #16860 2025-02-02 12:40:54 use nftables duck and cover 2025-02-02 12:41:10 docker does not support nftables 2025-02-02 12:41:28 oh ok - did not know that 2025-02-02 12:43:01 That's probably the same issue, thank you! 2025-02-02 12:43:59 "Hello! I just swapped to a..." <- Anyone happen to know where to start troubleshooting this issue? 2025-02-02 12:44:42 I just checked by booting into a different media immediately after a failed boot, and going through /var/log but there doesn't seem to be anything useful in there. 2025-02-02 12:53:52 longnoserob[m], oh dear, despite diff saying the ramdisks differ between the two machines, I have just extracted & compared the two and they are identical! 2025-02-02 12:54:24 even the modules? 2025-02-02 12:55:55 longnoserob[m], well, lib is identical, so yes ;) 2025-02-02 12:57:06 longnoserob[m], so I am back to the only difference I know - the hardware 2025-02-02 13:07:45 longnoserob[m], I'm going to reboot into the Pi4 and equalize the last bit of /etc for my own peace of mind . . . 2025-02-02 13:47:39 um i updated alpine in a chroot and its unable to boot due to /dev/sda8 being able to be mounted in /sysroot 2025-02-02 13:48:08 the initramfs console does not have working input either 2025-02-02 13:48:22 seems i should generate the initramfs manually from a chroot properly 2025-02-02 13:48:23 hmm 2025-02-02 13:51:03 when you use a chroot, make sure you have /dev /sys and /proc mounted in the chroot as well 2025-02-02 13:51:44 And naming of block devices can be different, so make sure it uses UUIDs 2025-02-02 14:04:20 ikke: they were mounted but the initramfs seems to be dead/empty 2025-02-02 14:04:47 remind me how to generate a initramfs please? 2025-02-02 14:05:04 mkinitfs? 2025-02-02 14:05:11 the command 2025-02-02 14:05:12 okay 2025-02-02 14:06:16 apk fix mkinitfs runs the correct command 2025-02-02 14:08:35 ikke: eh, thanks mkinitfs did it 2025-02-02 14:08:52 its still kind of weird that the initramfs broke badly like this 2025-02-02 14:08:59 its only 42mb, hope thats expected 2025-02-02 14:09:35 im chrooting from different distros and updating it that way 2025-02-02 14:10:31 kind of sad that distros dont want to support offline system updates maintenance through chroot 2025-02-02 14:42:43 longnoserob[m], I have found a diffence in the ramdisk (I forgot the -r first time!). There is a zram.conf in modules.d setting num_devices=3. I doubt that openrc is depending on a specific zram to produce output, but that is all I have to go on for now. 2025-02-02 14:45:44 sooooo it turns out the initramfs that is created is nonfunctional 2025-02-02 14:45:55 im not sure why 2025-02-02 14:46:20 dev, proc, sys are rbind mounted on the chroot and it did not work 2025-02-02 14:54:03 lobolibre[m], OK for the record, using the initrd from the Pi4 on the Pi5 made no difference - it can surely only be hardware now! 2025-02-02 14:55:16 lobolibre[m], incidntally, I noticed that zstd has been removed from the list of compression algorithms for zram as of kernel 6.12 2025-02-02 14:55:47 lobolibre[m], anyway, I am officially out of ideas now! Thanks for your help 2025-02-02 14:56:02 m4r does your hdmi sound work good on rpi5? what sound setup do you have? 2025-02-02 14:56:39 sorry ikke, im on current rolling point versions of alpine 2025-02-02 14:56:55 frag, use tab to get my full nick! 2025-02-02 14:57:50 i use basic irc client! 2025-02-02 14:57:59 frag, I am using ALSA and there is no need for an .asounndrc because there is no analoghe sound and HDMI starts at device 0 2025-02-02 14:58:50 m4r i so too, but i can only play one source at the time through hdmi 2025-02-02 14:59:01 *i do too 2025-02-02 14:59:09 frag: basic irc client can tab completion pretty sure :)) try it 2025-02-02 14:59:27 it 2just works", apart from dmix - I have tried the "official ALSA way" and the "Arch way" to configure it and neither works, but I don't care that much ATM . . 2025-02-02 15:00:00 m4r oh so you have the same problem? 2025-02-02 15:00:26 i use ii, and just some shell scripts, no completion 2025-02-02 15:00:36 frag, well you havent 2025-02-02 15:00:44 said what problem you have ;) 2025-02-02 15:01:10 m4r i can only play sound through hdmi from one source, eg one instance of mpv 2025-02-02 15:01:22 have to kill it to play something else.. 2025-02-02 15:01:38 frag, ah OK I see you did ;) yes, I think that we are in the same boat 2025-02-02 15:02:02 m4r ah okok, prob just wait for an update 2025-02-02 15:02:21 I don't really care about it, but I do remember debian alsa worked fine, in the days before (spit) pulse 2025-02-02 15:03:34 few needs, thats good 2025-02-02 15:07:01 pretty impressed with Alpine on the Pi, I intend to replace my old Debian Pi "NAS" in the near future 2025-02-02 15:07:12 AFK, tea time! 2025-02-02 15:26:36 frag, if you are still around, do you see any openrc output during boot on Pi5? 2025-02-02 15:37:34 m4r mmm... i think so, what does it look like? eg dhcp getting ip right? 2025-02-02 15:38:39 m4r but i use dtoverlay=vc4-kms-v3d and it doesnt let you see anything before half way in the boot 2025-02-02 15:39:02 so maybe you have black screen untill some gui start? 2025-02-02 15:40:42 frag, I see kernel messages up to USB stuff, then nothing until lightdm. BTW have you removed "-quiet" in your kernel command? 2025-02-02 15:41:22 not that i know of m4r 2025-02-02 15:42:16 frag, that will kill most if not all text output! 2025-02-02 15:43:00 frag, look in /boot/cmdline.txt 2025-02-02 15:44:20 oh, i run from RAM and dont have /boot but /media/mmc../cmdline.txt: modules=loop,squashfs,sd-mod,usb-storage quiet console=tty1 2025-02-02 15:46:51 frag, oh well there it is! 2025-02-02 19:57:31 leg7[m]: forgot to update here but all I had to do was mess with the cmdline file, just had to remove the quiet and add console=tty1 or whatever it was 2025-02-02 20:32:17 Hello I made a APKBUILD for my application: https://pastebin.com/0A8LZx35 But apkbuild -r returns: https://pastebin.com/0UZYG9EL How In can define checksums??? 2025-02-02 20:32:29 The file seem pretty much straightforward to me 2025-02-02 20:32:31 'abuild checksum' i think? 2025-02-02 20:32:43 certainly something close to that 2025-02-02 20:34:12 Abuild checksum returns: https://pastebin.com/2nsCtCf6 2025-02-02 20:34:20 I am usoing a docker container for builds 2025-02-02 20:34:23 using* 2025-02-02 20:36:03 maybe the file is still open in some editor? 2025-02-02 22:02:08 hi! fprintd-verify correctly verifies my local fingeprints, but firefox always gets stuck on the Touch your security key to continue with... popup, w/o enabling the fingerpint reader. I couldn't find any fingeprint reader guides on the wiki. 2025-02-03 01:53:39 Foxy[m]: What are you trying to do in Firefox? "Touch your security key" sounds like it expects you to tap on a yubikey-like device, not a fingerprint reader. 2025-02-03 01:55:49 Testing https://webauthn.io/ 2025-02-03 01:56:47 I think your OS needs to support using its own hardware as a key 2025-02-03 01:58:55 fprintd-verify works successfully 2025-02-03 01:59:04 Foxy[m]: fprintd-verify is for testing a fingerprint reader. That's not a security key. 2025-02-03 01:59:56 You need hardware capable of FIDO2 2025-02-03 02:13:08 Hello. Where can I find alpine release images recipes so that I can recreate them please? I'm mainly interested in the qcow2 nocloud images. I would like to make a few changes. Thank you. 2025-02-03 02:23:37 "You need hardware capable of..." <- Thanks! I misunderstood then 2025-02-03 02:46:35 btr68: https://wiki.alpinelinux.org/wiki/How_to_make_a_custom_ISO_image_with_mkimage 2025-02-03 02:48:57 WhyNotHugo: yeah..thanks.. not really what I'm looking for.. I think this is it https://gitlab.alpinelinux.org/alpine/cloud/alpine-cloud-images 2025-02-03 08:05:57 "leg7: forgot to update here..." <- nice! 2025-02-03 14:47:19 hi. i'm having a little disk partition confusion. parted seems to disagree with everything else, as far as i can tell: https://dpaste.com/EP663HDXV 2025-02-03 14:48:08 i wonder if you have GPT and MBR disagreeing 2025-02-03 14:48:23 oh, hmm 2025-02-03 14:48:24 'file system' and 'flags' are not things you'd see with MBR 2025-02-03 14:48:40 'loop' seems weird to me 2025-02-03 14:48:54 ah, file system and flags are gpt only concepts? 2025-02-03 14:49:10 iirc 2025-02-03 14:50:02 hmm, 'loop' is not 'gpt' 2025-02-03 14:50:10 i think you may be right 2025-02-03 14:50:13 'raw disk access' 2025-02-03 14:50:30 yeah, from what i'd read, loop is code for "there is no table" 2025-02-03 14:50:43 yep seems that way 2025-02-03 14:52:07 this probably makes sense. it's an sd card, that i initially wrote to using dd, from an .img file. then, i installed alpine, which partitioned the disk with an mbr 2025-02-03 14:52:18 well, partially makes sense, anyway 2025-02-03 14:53:06 lsblk would reflect kernel state 2025-02-03 14:53:26 don't know how raw fdisk and parted are 2025-02-03 14:54:16 if i partition the disk with parted, and give it an msdos label, parted would know about it, so it's still a little confusing why it wouldn't if something else was used to do the partitioning 2025-02-03 15:00:34 forgot to write the new layout to the device? 2025-02-03 15:01:31 i mean, the layout is there, i think, isn't it? it boots, it runs, everything works, every command other than parted seems to be aware of the new layout 2025-02-03 15:05:22 ah, interesting. testdisk seems to agree with parted: "Hint: None partition table type has been detected" 2025-02-03 18:03:44 Hello. Anyone have some experience using fwupd without PolKit/elogind? 2025-02-03 18:04:43 I use seatd/greetd with pam, but had to install fwupd-polkit for get fwupdmgr commands to work. 2025-02-03 18:06:18 But I'm still getting warnings about UEFI ESP partition not detected or configured 2025-02-03 18:08:52 The warnings: https://tpaste.us/VYw0 2025-02-03 18:28:42 yaboyfred, ls -al /sys/firmware/efi/efivars 2025-02-03 18:29:01 Yeah I was somehow able to make it work, but that thing is a mess 2025-02-03 18:29:12 It tries to do a lot automatically and has to be more clever than the user 2025-02-03 18:29:25 quinq, oh, my guess "you are not even booted in EFI" might be wrong then? 2025-02-03 18:29:30 So when it doesn't find the exact configuration it needs, it fails in obscure ways 2025-02-03 18:30:35 could the mpv url freez/hang computer be related to having an external usb drive mounted? not sure its happened without that being the case... 2025-02-03 18:30:55 Habbie, to be honest I don't remember the various failures, but yeah it had to do with detecting the efi partitions (even if present) 2025-02-03 18:32:11 It was a mix of starting the daemon (fwupd), then runnig the (somewhat) standalone client (fwuptool IIRC), then disabling the daemonā€¦ 2025-02-03 18:32:27 But sorry, don't listen to me I won't be of much help here because I don't remember 2025-02-03 18:32:41 I can just say it works without polkit/elogind, but YMMV 2025-02-03 18:35:39 Habbie: yeah, /sys/firmware/efi/efivars/ very much present 2025-02-03 18:35:59 yaboyfred, alright, that indeed rules out my one idea 2025-02-03 18:36:42 I do have udisks(2) installed though 2025-02-03 18:36:58 As stated in https://github.com/fwupd/fwupd/wiki/PluginFlag:esp-not-found (as stated in the error message) 2025-02-03 18:46:40 Hm, I do indeed not have udisks2 installed 2025-02-03 19:40:11 Hi, merge request 79242 is open since a few days and I don't see anything blocking. I also just had to rebase it just because others got merged meanwhile and it blocked merging. 2025-02-03 19:40:32 Could someone have a look and merge it? 2025-02-03 19:41:10 !79242 2025-02-03 19:41:18 flix: ftr, you don't have to rebase it yourself 2025-02-03 19:42:32 Oh, okay.. I was asked to do that in previous merge requests, so I thought it's my responsibility. 2025-02-03 19:43:00 Only if there are merge conflcits 2025-02-03 19:43:13 or you are asked to squash some fixup commits 2025-02-03 19:43:19 Alright. 2025-02-03 19:51:03 We are trying to get to it, but we have quite a queue of MRs 2025-02-03 19:51:16 we receive 60 new MRs on a busy day 2025-02-03 19:54:53 i can maybe help you out with reviewing more often 2025-02-03 19:56:57 (if i've got some free time over) 2025-02-03 19:58:21 This world lacks time for everyone :( 2025-02-03 20:01:43 I'm using PostgreSQL in alpine, but trying to use a different locale (than C) in postgres 16 seems to have no effect, for example SET lc_time='de_DE.UTF-8' ; then select to_char(current_date,'month'); gives february instead of Februar, the packaged musl-locales-0.1.0-r1 and musl-locales-lang-0.1.0-r1 are installed, Alpine Linux 3.22.0_alpha20250108 2025-02-03 20:02:13 s/packaged/packages/ 2025-02-03 20:03:53 in the shell LC_TIME=de_DE.UTF-8 date +%B date works, it prints Februar 2025-02-03 20:03:57 ikke: Yeah, no hurry. I understand. As I said, I thought rebasing the MR would be my responsibility. Since it's not the case, there is no rush. 2025-02-03 20:05:29 akoana: iirc musl doesnt support collation yet. not sure if thats the issue, or something else 2025-02-03 20:06:43 fossdd[m]: yeah, must be something like that, because in debian for example the postgres code works as expected... 2025-02-03 20:09:05 https://gitlab.postmarketos.org/postmarketOS/postmarketos/-/issues/65 2025-02-03 20:09:13 akoana: hopefully we'll get a grant to work on it 2025-02-03 20:09:18 that would heavily improve the situation 2025-02-03 20:11:54 oh I see, you're in the German team, hoping you'll get the grant - Danke! 2025-02-03 20:14:53 btw. I'm using lc_numeric in postgres too to have "." as thousand separator and "," as decimal separator as usual in German number formats, so it would be great to have collation support... 2025-02-04 03:54:30 how do I add an IPv6 address from a /64 to be used in an interface, or even better, the whole /64? I was told I have to in some way in order to have them work but I have no idea how 2025-02-04 05:44:50 zayd: the question is little bit unclear. Generally, the /64 is announced through router advertisements and your devices will use that to assign itself an ipv6 address 2025-02-04 06:07:46 i'm preparing to set up a repo for my personal aport packages on my vps. is there any example code to look at for handling the updating of the APKINDEX.tar.gz files when packages are uploaded? i was thinking that i could use an inotifywait on the upload directory, but am wondering if there could be any issues if the web server is serving a download when the update happens. 2025-02-04 06:10:20 j_v: alpine uses rsync with --delay-updates https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/main/aports-build/aports-build#L212-219 2025-02-04 06:13:42 thanks ikke, appreciated. i will look at that. i've never used the --delay-updates switch before so now is a good time to read up on it. 2025-02-04 06:16:16 oh, nice... that will preclude any need for the inotifywait idea. thanks again. 2025-02-04 06:17:23 i was thinking i would need to update the APKINDEX on the server, but just rsync'ing the repo means it will be done ahead of time. cool. 2025-02-04 06:18:45 yes 2025-02-04 06:18:53 That's how it's done on the builders as well 2025-02-04 06:19:05 they generate the index, and then everythings rsynced 2025-02-04 06:19:33 i wondered that before when looking for build failures to address, now it's clear 2025-02-04 06:20:43 You see that happening at the end of every succesfull build, like here: https://build.alpinelinux.org/buildlogs/build-edge-x86_64/community/hugo/hugo-0.143.0-r0.log 2025-02-04 06:22:45 cool. yeah, what i always had been curious about was when i saw the upload messages on the main build page. makes so much more sense to me now. 2025-02-04 06:25:16 i'm also working on dockerizing some stuff. i'm still pretty new to docker; have used lxc and nsjail plenty, don't know why i avoided docker so much. i'm getting the gist of layering images from a scratch image... looking at alpine's docker stuff really helps. 2025-02-04 11:58:09 ikke: I'm trying to use Yggdrasil for hosting some stuff and wanted to use the /64 addresses for that. They don't seem to be auto assigning and I was told I have to do that myself 2025-02-04 12:26:48 is that entire /64 routed to you somehow, or does it belong to a local subnet? 2025-02-04 12:29:04 ACTION misses his sixxs /48 2025-02-04 12:29:22 grawity: I'm not sure but I _think_ Ygg gives you the whole subnet 2025-02-04 12:30:15 ok, and do you want to use it on the same machine or provide it to a local LAN? 2025-02-04 12:30:58 I want to use it on the same machine as public addresses 2025-02-04 12:31:07 (with yggdrasil I guess I'll file it under "routed") 2025-02-04 12:31:57 choose an address out of the /64, `ip addr add
/64 dev ` 2025-02-04 12:32:12 does ygg give you public addresess in the sense of global Internet addresses? 2025-02-04 12:32:29 I think I tried that but I'll try again when I can ssh in 2025-02-04 12:33:02 grawity: I've never used clearnet IPv6 (my ISP doesn't work properly with it) so no idea there 2025-02-04 12:33:29 if the prefix is routed to you, and it's for the local machine, then the interface (and even the prefix length) don't matter that much 2025-02-04 12:33:43 what's the actual /64 prefix you have? 2025-02-04 12:33:46 might prefer /128 over /64 but i don't have enough context to advice 2025-02-04 12:34:40 from what I know about yggdrasil, it's a "mesh vpn" effectively, which uses its own address range within the network... and does not interact with the public internet 2025-02-04 12:34:53 it says my subnet is 300:5506:25eb:d0d9::/64 2025-02-04 12:35:02 definitely not within the regular internet range 2025-02-04 12:35:10 yeah it doesn't go outside of its own thing 2025-02-04 12:35:19 it uses a retired IPv6 space iirc 2025-02-04 12:35:32 what do you mean by "want to use it as public addresses" then? 2025-02-04 12:36:35 I want to have different sites hosted on different addresses, I think I got told that was possible? 2025-02-04 12:36:53 obviously accessible within Yggdrasil only though 2025-02-04 12:36:55 with that prefix? only within yggdrasil, yes 2025-02-04 12:39:03 so in theory it should be enough to choose an address and assign it to the yggdrasil virtual interface (either as /64 or as /128, I suppose /128 is somewhat better in your case but functionally it doesn't matter *much*) 2025-02-04 12:39:33 but more importantly, install tcpdump, have some other host ping you, make sure you see those packets on tun0 or whatever interface it uses 2025-02-04 15:40:16 grawity: alright I got it to work with the command you mentioned, is it possible to add the whole range? what I'm doing will probably need lots of addresses 2025-02-04 19:56:01 Setup apkcache on my craptop with an old 32GB uSD card I found in the internal uSD slot 2025-02-04 19:56:18 ACTION uploaded an image: (24KiB) < https://matrix.org/oftc/media/v1/media/download/ATlAnUaNJ4An4VkhP2l_yKJOpepFZhMQSeVLTEKVefqSnMqS7NQCXKs_KU_OSnrU1Pl4Drx9ug0oymEHK0PzIqZCeVGqH7PAAG1hdHJpeC5vcmcvVVBMaHZxeVV3Qm9ydHhURExDSGFEZXhP > 2025-02-04 19:57:05 It's really nice, honestly, especially when packaging. It's so fast to install all the makedepends šŸ¤£ 2025-02-04 21:58:07 Having issuesbooting xen variant but booted standard without issue. Can standard be repurposed as xen easily? Is it mainly the kernel plus some packages? 2025-02-05 00:16:40 if i see this it means mold was used for my binary? https://i.imgur.com/DBZNxHJ.png 2025-02-05 00:17:31 i got it from here but the output looks different https://github.com/rui314/mold?tab=readme-ov-file#how-to-use 2025-02-05 02:52:22 I'm trying to build a kernel module (knetstat) inside a docker container hosted on ec2, but I get this error: make[1]: *** /lib/modules/6.8.0-1021-aws/build: No such file or directory. Stop. 2025-02-05 02:52:26 what package will provide this? 2025-02-05 02:53:23 otherwise, I'd like to see what socket options are set for a particular socket 2025-02-05 03:45:03 Garfunk: that directory does not look to be from an alpine kernel package... least ways, not one i am aware of 2025-02-05 03:56:00 j_v: is there some other package that can do what I want? 2025-02-05 04:00:26 not for that kernel version... there is linux-lts-dev, but that is for kernel version 6.12.11... that kernel modules directory in the above is 6.8.0 from some aws package. not compatible 2025-02-05 05:25:06 Garfunk: you need to override the kernel version. By default it uses what uname returns, but that is not correct when the host is a different OS 2025-02-05 05:26:07 ikke: it's on an ec2 instance on aws 2025-02-05 05:26:58 It's trying to build for that aws kernel, which is not a kernel we provide 2025-02-05 05:28:01 Unless you want to build a module for that kernel, but then you need to look at aws for that source 2025-02-05 06:16:57 <|0|> hi 2025-02-05 06:20:10 <|b3r60h> TRUMP RULES 2025-02-05 06:20:10 <|qsufhs> TRUMP RULES 2025-02-05 06:20:12 <|g4p38n> TRUMP RULES 2025-02-05 06:20:13 <|tr4edb> TRUMP RULES 2025-02-05 06:20:15 <|3p4ivb> TRUMP RULES 2025-02-05 06:20:16 <|1q5z7q> TRUMP RULES 2025-02-05 06:20:16 <|ubgwr6> TRUMP RULES 2025-02-05 06:20:20 <|4j17aj> TRUMP RULES 2025-02-05 06:20:22 <|re8cxf> TRUMP RULES 2025-02-05 06:49:48 A big spam wave 2025-02-05 06:50:41 It looks more like a "big trolling wave" 2025-02-05 06:52:36 yeah 2025-02-05 06:52:42 spammed my DMs as well 2025-02-05 07:00:53 oh :\ 2025-02-05 11:19:18 Hello! Any alive? I has problems with installing MariaDB on new installed alpine, can any help me? 2025-02-05 11:19:58 what's wrong? 2025-02-05 11:20:54 mysql_secure_installation not startig because service not started, but service not starting because data dir ir wrong 2025-02-05 11:23:03 HOST:/var/lib/mysql# rc-service mariadb start 2025-02-05 11:23:15 * Caching service dependencies ... [ ok ] 2025-02-05 11:23:15 * Run '/etc/init.d/mariadb setup' to create new database. 2025-02-05 11:23:15 * Datadir '/var/lib/mysql' is empty or invalid. 2025-02-05 11:23:15 * ERROR: mariadb failed to start 2025-02-05 11:23:40 did you run the setup? 2025-02-05 11:24:47 _ /etc/init.d/mariadb setup ? 2025-02-05 11:25:03 that is what it says 2025-02-05 11:25:35 Yepp.. There multiplie errors, refering to log, but log nont exists.... 2025-02-05 11:26:53 try `mariadb-install-db --user=mysql` manually 2025-02-05 11:27:07 ok 2025-02-05 11:28:13 solved, big big thanks!!! ^_^ 2025-02-05 11:48:59 Hi all. I was n here a couple of days ago asking about missing openrc boot messages on a Raspberry Pi 5 (they are present on a 4).My investigations led me to the video overlays. There are two KMS options that possibly relate to the Pi5, but neither of the m resored the openrc messages. The experiments did confirm tha the ooverlays led to more or less output, so I am certain that this is the root of the issue. 2025-02-05 15:08:21 hmm, looks like there is some overlap between my openrc message problems and this:https://lists.alpinelinux.org/~alpine/users/%3C23bde335-16ff-4a07-8f8f-309f84b52bf6%4053hor.net%3E 2025-02-05 16:38:54 Crypto Announcement: https://comcoin.fun/ - this will moon today at 20:30 CET 2025-02-05 16:39:29 Crypto Announcement: https://comcoin.fun/ - this will moon today at 20:30 CET 2025-02-05 18:05:38 what is the best way to give an user permission to reboot/poweroff ? 2025-02-05 18:05:44 my greetd doesnt have 2025-02-05 18:48:12 iirc seatd doesn't have anything ro do with power stuff 2025-02-05 18:48:58 if you're talking about cli utilities, you can edit doas/sudo conf to not ask for a password when executing those 2025-02-05 18:49:16 greetd greeters can have an option to reboot/shut down 2025-02-05 18:49:44 oh, confused greetd and seatd. dumb me 2025-02-05 19:02:33 Yeah regreet does have but aint work due permissions 2025-02-05 19:39:45 I personally use the powerctl program and just allow it with doas, must also be possible with sudo 2025-02-05 20:58:28 can somebody build qt6-qtbase with clang aarch64? 2025-02-05 21:12:52 why? 2025-02-06 04:19:46 yo, looks like alpine linux is out of Equinix as freedesktop 2025-02-06 06:08:34 mayli: yes, see https://alpinelinux.org/posts/Seeking-Support-After-Equinix-Metal-Sunsets.html 2025-02-06 06:30:02 ikke: yeah, feels bad for that :/ 2025-02-06 06:30:22 We have received many offers for sponsorship in response to that 2025-02-06 06:30:31 cool 2025-02-06 09:10:43 Anything one can do to help (besides financial support)? 2025-02-06 10:10:38 other day I asked about if community repo are enabled by default, it is not, but setup-apkrepos can enable it (also setup-desktop runs it) just to say 2025-02-06 11:32:53 so I decided to try to learn vim and already regret is our vim compiled without clipboard module? vim --version | grep clipboard I see -clipboard 2025-02-06 12:36:10 hmm... what does that module exactly? 2025-02-06 12:37:31 I personally copy/paste directly using the term emulator so it results app agnostic 2025-02-06 12:38:57 which module? 2025-02-06 12:41:00 longnoserob[m], the vim clipboard module 2025-02-06 12:41:16 ah ok 2025-02-06 12:50:43 btw anyone elsee getting BAD signature on aarch64 for multipath-tools-0.11.0-r0 multipath-tools-openrc-0.11.0-r0 2025-02-06 13:06:38 rnalrd^ 2025-02-06 13:13:01 curl -X PURGE url fixed it.. 2025-02-06 13:13:06 sorry for the noise 2025-02-06 16:07:01 fabricionaweb: try gvim 2025-02-06 16:09:10 vim with system clipboard support 2025-02-06 16:09:36 though it's a gui version, runs as vim in terminal 2025-02-06 18:32:36 can someone explain why 1 cpu core is at 99% all the time? https://pastebin.com/raw/7JdupYmg - it's a very basic "alpine standard x86_64" on a lenovo-yoga260. 2025-02-06 18:37:06 maybe atop shows it better: https://pastebin.com/raw/tWh3LwfW 2025-02-06 18:37:15 https://wiki.archlinux.org/title/Lenovo_ThinkPad_Yoga_260#Power_management 2025-02-06 18:41:26 MY FAN 2025-02-06 18:41:44 DuClare: thanks a lot 2025-02-06 18:42:48 Does Alpine have a GPG signing key for their repositories? 2025-02-06 18:43:35 No, an RSA key only 2025-02-06 18:45:08 how package https://codeberg.org/tpikonen/satellite ? 2025-02-06 18:49:32 @ikke Is that RSA signature signed by alpine or the individual maintainer of the packages? 2025-02-06 18:49:59 I am looking at the Abuild-keygen docs and it appears to be specific to the maintainer 2025-02-06 18:50:20 Alpine 2025-02-06 18:52:53 ikke Thanks! I am trying to see if there's a way we can cache the repo to our Satellite (Foreman) system for use internally so each build doesnt have to reach out over the net for updates. We want to make sure all packages inside the caching proxy are signed by the upstream. 2025-02-06 18:54:40 The signature is embedded in the package and index 2025-02-06 18:55:19 So it's safe to cache / mirror 2025-02-06 18:56:05 As long as the index and packages match 2025-02-06 18:59:58 Thanks, appreciate it! I also see how this is documented in the APK spec wiki article. Forwarding this to our folks to have them understand this a bit better. 2025-02-06 19:02:16 Generally, you dont cache the index but you can cache the apk files 2025-02-06 19:02:38 New packages should get different names 2025-02-06 20:42:37 Hello o/ I see trouble installing python3 on edge due to libsqlite3.so.0 2025-02-06 20:43:09 this showed up within the last 24 hours 2025-02-06 20:43:41 updating the repository to latest-stable resolves the issue 2025-02-06 20:44:08 https://dpaste.com/GDGQX4VQ2 2025-02-06 20:46:58 might be coming from https://gitlab.alpinelinux.org/alpine/aports/-/commit/1d2e767d2e1c19aebb1740c23ec4ef92fa244ce1 2025-02-06 20:47:48 Hmm, the package sqlite-libs still has the file libsqlite3.so.0 2025-02-06 20:52:35 Hi, for desktop, jwm works fine, but lxqt is very slow because may services are constantly crashing. There are a lot of messages saying PolicyKit handler crashed too many times, Its autorestart has been disabled until next login. Same for Desktop, Global Keyboard Shortcuts, Power Management and Notification Daemon. 2025-02-06 20:54:59 Hi, there seems to be a bug in an Alpine Linux. Here's a part of the log: "ERROR: unable to select packages: so:libsqlite3.so.0 (no such package): required by: util-linux-2.40.4-r0[so:libsqlite3.so.0]" 2025-02-06 20:55:09 pmOS_User4: We're just looking at it 2025-02-06 20:55:14 In /var/log/messages there's "The maximum number of active connections for UID (my user) has been reached (max_connections_per_user=256) 2025-02-06 20:55:21 apparently a recent sqlite update changed the soname scheme 2025-02-06 20:55:23 Also, it wasn't probably present an hr ago or so 2025-02-06 20:55:28 pmOS_User4: people here are already aware :p 2025-02-06 20:55:39 That's good, just checking :) 2025-02-06 20:59:45 I've reverted the change for now 2025-02-06 21:00:17 thats really good, thanks 2025-02-06 21:03:41 most arches have built the reverted package, in ~15m it should be available on dl-cdn 2025-02-06 21:06:05 šŸ‘ 2025-02-06 21:07:47 the bug is gone in PostmarketOS, thanks for help ikke! 2025-02-06 21:21:14 thanks for the quick fix - no issues for me now :) 2025-02-06 21:25:13 dl-cdn.alpinelinux.org seems to be failing for me already, also the cache dir, running from RAM .. worked when swiching repo url tho 2025-02-06 21:25:58 oh maybe cache is updated or can only be used if you do lbu commit? 2025-02-06 23:04:59 Now, there's this error in /var/log/Xorg.0.log 2025-02-06 23:05:03 (EE) AIGLX error: dlopen of /usr/lib/xorg/modules/dri/i965_dri.so failed (Error loading shared library /usr/lib/xorg/modules/dri/i965_dri.so: No such file or directory) 2025-02-06 23:05:17 apk info --who-owns i965_dri.so 2025-02-06 23:05:17 ERROR: i965_dri.so: Could not find owner package 2025-02-06 23:05:35 Any solution for this? 2025-02-06 23:16:24 should be fixed after https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/78030 afaik 2025-02-06 23:16:29 did you ran apk upgrade -Ua? 2025-02-06 23:21:11 fossdd[m], is that available for 3.21? 2025-02-06 23:24:23 Ah on 3.21 mesa still supports dri 2025-02-06 23:24:57 what if you manually install apk add mesa-dri-gallium? 2025-02-06 23:27:34 it's already installed. 2025-02-06 23:27:58 The problem now is that mouse and keyboard become not operational when xorg starts. 2025-02-06 23:28:21 But used to work before, maybe some package I've added... 2025-02-06 23:34:17 when running sddm, keyboard and mouse are fine. If Xorg is started by startx, with some ~/.xinitrc they are not operational. 2025-02-06 23:44:54 Added my user to input and tty groups and now keyboard and mouse work on X. 2025-02-07 11:03:34 In alpine's gitlab, can i retrigger the automatic assignment of reviewers somehow? 2025-02-07 11:14:41 don't think so except opening a new MR 2025-02-07 11:14:52 but i think alpine-qa-bot doesnt even support assigning multiple reviewers 2025-02-07 11:15:08 gitlab does not allow it for CE 2025-02-07 11:20:26 hey ya, I have a problem with dhcpcd. When I start networking and donā€™t have an ethernet cable plugged in, it reports finding a carrier and tries to get a dhcp lease. Because there is no cable plugged in, it fails and canā€™t bring the interface up, meaning I canā€™t just plug in an ethernet cable later on 2025-02-07 11:21:16 I canā€™t figure out how to solve this. I am using ifupdown-ng together with iwd 2025-02-07 11:21:52 Not having the timeout, but instantly forking to the background would be nice too 2025-02-07 12:09:30 janvhs: i can only speak from personal experience, but configuring *less* is what solved most my problems with dhcpcd+iwd 2025-02-07 12:09:47 turned out the defaults were pretty good and I was just messing things up 2025-02-07 12:11:15 you can find how I set my networking up in here: https://git.sr.ht/~yaboyfred/dotfiles/tree/master/item/doc/alpine.md 2025-02-07 12:16:22 yaboyfred: thanks :) I just installed both and took over /etc/netwok/interfaces from before. Iā€™m gnna gove your docs a look! 2025-02-07 12:21:20 yaboyfred: awesome your setup is basically the same, but you have a different configuration for ifupdown and start dhcpcd as service. I let ifupdown-ng invoke dhcpcd. That might already be what I wanted :D Btw I swaw TLP in your wishlist there. I can send you my mini config if you want 2025-02-07 12:33:58 janvhs: haha nice, another benefit having it as a service is faster boot times as it no longer blocks 2025-02-07 12:35:06 please do share your tlp conf with me, thatd be great, been meaning to look into it :) 2025-02-07 14:20:34 Hi 2025-02-07 14:22:02 I have some trouble with networkmanager after upgrading from 3.19 2025-02-07 14:25:24 3.19 -> broke networkmanager so I downgraded to 3.19 and made 3.19 -> 3.20 -> 3.21 2025-02-07 14:26:28 works in 3.20 but not in 3.21 2025-02-07 14:27:45 I guess related to the fact I have networkmanager-elogind installed instead of networkmanager 2025-02-07 14:28:21 and networkmanager-elogind doesn't exist anymore in 3.21 2025-02-07 14:28:52 trying apk fix --upgrade networkmanager says apk is unavailable 2025-02-07 14:29:42 it says this to a lot of packages by the way (tried to reinstall all packages) 2025-02-07 14:30:07 if anyone has some idea to solve this situation i'd be greatful :) 2025-02-07 19:08:22 hello, in the raspberry pi download there is two armhf images, which one do I pick for rasperry pi zero w? are they the same??? 2025-02-07 19:11:33 em6, where exactly? 2025-02-07 19:12:28 directly on the download pacge inside the raspberry pi section. there is duble of every version, I just dont undertsand why 2025-02-07 19:12:35 just paste the URL please 2025-02-07 19:14:16 downloaded both. and now I see one is tar and one is img 2025-02-07 19:14:38 https://dl-cdn.alpinelinux.org/alpine/v3.21/releases/armhf/alpine-rpi-3.21.2-armhf.tar.gz 2025-02-07 19:14:47 https://dl-cdn.alpinelinux.org/alpine/v3.21/releases/armhf/alpine-rpi-3.21.2-armhf.img.gz 2025-02-07 19:15:03 oops. i did not realise this was not #raspberrypi :D 2025-02-07 19:15:21 (your initial question would have fit there just as well, but would not have meant the exact same thing) 2025-02-07 19:15:39 if i have to guess, one is for dd, the other is for extracting to .. something 2025-02-07 19:16:22 ah so I should ask in #raspberrypi? 2025-02-07 19:16:27 oh no 2025-02-07 19:16:32 you are in the right place 2025-02-07 19:16:53 i just had to reboot to your first question, for myself :) 2025-02-07 19:16:59 ah ok 2025-02-07 19:17:42 you know which one is for dd? 2025-02-07 19:18:27 i am just guessing here, but it would definitely not be tar 2025-02-07 19:18:38 i also know there's some documentation on the wiki 2025-02-07 19:19:05 can you check if I remember correctly: dd bs=4M if=PATH-TO-IMAGE of=PATH-TO-PARTITION 2025-02-07 19:20:09 never sure if it's 4M or 4m, but dd will tell you. of might want to be the full SD, not a partition (but i don't -know- this) 2025-02-07 19:20:19 other than that, yes 2025-02-07 19:20:46 I will try the whole device 2025-02-07 19:21:07 if i have to guess, yes :) 2025-02-07 19:21:08 and need to do it headless, wish me luck 2025-02-07 19:21:11 good luck! 2025-02-07 19:21:17 i hope the image comes with ssh enabled 2025-02-07 19:21:26 mee too 2025-02-07 19:22:13 but standard RPi image wiki say you just put an empty file called ssh in root 2025-02-07 19:22:39 and a key or a password, right 2025-02-07 19:22:46 but, that's for raspi os. i have no clue if it's true for alpine 2025-02-07 19:22:51 the hard part for me will be to add acount and wifi 2025-02-07 19:22:53 you can't always mix documentation like that 2025-02-07 19:23:09 ah ok 2025-02-07 19:23:12 good to know 2025-02-08 01:21:25 Hi guys, I am interested in learning Linux OS dev but find books to be too dry. Do you have any development meetings where I can learn by listening to discussions? 2025-02-08 05:40:44 need_info: "OS dev" is a very general term. Are you talking about kernel dev? userspace/application development? networking? 2025-02-08 10:58:33 Is it possible to select DDG as the search engine in chrome - it looks like there are only three options (google, bing, yahoo)? 2025-02-08 11:05:57 you can probably add it yourself 2025-02-08 11:06:30 lopid, no, it is not possible 2025-02-08 11:06:54 lopid, there is no UI for it! 2025-02-08 11:06:54 Hey yā€˜all I wanted to spend some time at work trying to contribute to the Alpine project. I saw talk over an ongoing effort to replace OpenRC with s6. Is there any place I can help with that? Itā€™s 2025-02-08 11:07:10 I wasnā€™t able to find sources for it 2025-02-08 11:12:02 Donā€™t know if skarnet on here sadly 2025-02-08 11:12:53 He's in the #alpine-devel channel 2025-02-08 11:13:51 ikke: Must have missed the existence of this channel. Thank you! 2025-02-08 11:43:28 I have found the correct goats to slaughter to make DDG available in chromium! It really is an utter mess! 2025-02-08 11:44:09 If you want some fun today make a fresh install and try it from a standing start 2025-02-08 12:57:34 Hi, anyone know if the polkit service timestamp from "rc-status -a" is supposed to increment by one every second? 2025-02-08 12:58:51 It will say "started 01:09:20 (0)", then the next second "started 01:09:21 (0)" 2025-02-08 13:01:50 Since it's a supervised service, does this mean it is crashing and reset every other second? 2025-02-08 13:36:20 Hi when i try to use kexec to load a kernel i get an error https://gist.tnonline.net/NQ 2025-02-08 13:38:01 This happens with linux-virt kernel but not my own kernel. Could it be because linux-virt doesn't have # CONFIG_KEXEC_FILE is not set 2025-02-08 13:43:00 Hm same happens with linux-lts 2025-02-08 14:37:41 When is $RC_REBOOT set ? I have a init script that needs to know this. The script sits in the SHUTDOWN runlevel. When I do echo $RC_REBOOT it's empty 2025-02-08 14:41:42 According to https://github.com/OpenRC/openrc/blob/7a608db4fe6aa434884d0aa05b5e3c09bbc656ea/man/openrc-run.8#L556 it should be available 2025-02-08 15:03:27 It seems to exist on my Gentoo box. Is it perhaps a bug? 2025-02-08 15:04:52 Forza: alpine does not use openrc-init 2025-02-08 15:05:00 Oh 2025-02-08 15:05:17 https://gitlab.alpinelinux.org/alpine/aports/-/issues/16695 2025-02-08 15:08:40 ikke: thanks 2025-02-08 15:08:58 I need to rethink my kexec-boot script then 2025-02-08 15:09:22 Note that the issue did seem to include a fix to support it for bb init 2025-02-08 15:10:15 It may require you to add a reboot action to /etc/inittab 2025-02-08 15:10:23 https://gitlab.alpinelinux.org/alpine/aports/-/issues/16695#note_472505 2025-02-08 15:11:02 Yes, I saw that :) But I prefer to wait for official support. 2025-02-08 15:11:16 Here's my adjusted script https://gist.tnonline.net/IU 2025-02-08 15:12:41 Would it be useful to include this with kexec-tools on Alpine? I could make a MR if you like 2025-02-08 15:13:41 i think you would need to ask Celeste 2025-02-08 15:20:26 Ill tweak it a bit more then ill ask 2025-02-08 15:21:25 I think of adding a load command and a confd file to configure what kernel and initramfs to load 2025-02-08 16:37:59 i guess the repo is unstable atm? im getting 'IO ERROR' 2025-02-08 16:56:03 working for me 2025-02-08 17:00:41 by repo i assume you mean apk and not git, and assume you mean the default dl-cdn.alpinelinux.org 2025-02-08 17:01:13 considering that message, if i were you, i'd make sure your disk is writable... 2025-02-08 17:06:06 hmm i changed from dl-cdn.alpinelinux.org since i had some trouble there too the other day 2025-02-08 17:06:53 i don't think apk would throw that error if it had a problem fetching 2025-02-08 17:07:00 sounds like a disk problem to me 2025-02-08 17:07:37 i run from RAM and with doas, so not sure what could be wrong with writing 2025-02-08 17:08:00 oh, maybe the cache... 2025-02-08 17:08:01 bad ram maybe 2025-02-08 17:08:22 i dunno... but i suspect a "disk" issue either way 2025-02-08 17:11:20 /media/mmc.., where the apk cache is, is mounted (read-only) as usual 2025-02-08 17:15:37 ACTION ā­ it is cache ā­ 2025-02-08 17:15:53 I meant to say dns sorry 2025-02-08 17:16:45 ok, changed back to default repo url, and it workds.. 2025-02-08 17:17:47 fabricionaweb: it's always dns ;) 2025-02-08 20:02:29 https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/79319 I had packaged the wrong version of haxe and now I want to downgrade the package so I can package snikket-sdk and it builds. 2025-02-08 20:02:46 Does anyone know the correct way to fix this? 2025-02-08 20:05:52 Also, can we get a merge on this? it has been approved https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/74589 2025-02-08 20:38:02 is there a way to unstuck the live image when the boot is stuck on "loading hardware drivers" ? 2025-02-08 21:32:12 found my issue, missing i915.modeset=0 2025-02-08 22:24:52 bl4ckb0ne: isn't that equivalent to nomodeset? 2025-02-08 22:33:46 somewhat, but nomodeset was not working at all 2025-02-08 22:34:02 i think this one specifically for the i915 driver 2025-02-08 22:34:16 im doing intel gpu passthrough 2025-02-09 03:23:01 hi does anyone know how to use the suckless irc client ii with lchat? I know ii is in apk add but I do not see lchat... 2025-02-09 03:25:12 Ichat as appli Ichat? 2025-02-09 03:25:20 s.appli/apple 2025-02-09 03:25:26 no L chat 2025-02-09 03:26:35 hmm 2025-02-09 03:27:28 suckless software tool 2025-02-09 03:29:20 it defeats the purpose of suckless to want it all from the package manager but... 2025-02-09 03:30:12 ircii exposes an irc connection as a bunch of files, like a fs basically, and lchat would wrap that 2025-02-09 03:30:34 ahhh yeah that makes sense thank you 2025-02-09 03:30:35 ii exists as apk 2025-02-09 03:30:44 yes ii exists as apk 2025-02-09 03:30:46 check gitlab, maybe someone has already made a mr for lchat 2025-02-09 03:30:54 I tried playing with it but was confused a bit 2025-02-09 03:32:07 new to alpine etc; what is an mr? 2025-02-09 03:32:16 gitlab speak for merge request 2025-02-09 03:32:23 ahhhhh classic and simple. thanks 2025-02-09 03:35:34 which gitlab page would I check for the merge request? 2025-02-09 03:35:46 s/ircii/ii/ 2025-02-09 03:35:56 gitlab.alpinelinux.org 2025-02-09 03:36:14 then -> aports 2025-02-09 03:43:30 poking around on the dashboard for aports now; admittedly I do not know where to find merge requests...finding myself a bit lost 2025-02-09 03:45:26 Code > Merge Requests in sidebar 2025-02-09 03:45:43 or https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests 2025-02-09 03:47:28 awesome that helped a ton 2025-02-09 04:36:58 Anyone using tor on Alpine? Any good guides? 2025-02-09 04:59:57 Saijin_Naib[m]: the browser or the daemon? 2025-02-09 05:00:07 if daemon, use apk 2025-02-09 05:00:22 if browser, take a look at flatpak 2025-02-09 05:34:57 anjan, hi if you 2025-02-09 05:35:23 are using both can you jot down some quick notes so a working wiki page be created 2025-02-09 05:35:50 there's nothing in wiki about tor as of now 2025-02-09 05:36:40 you can create a text file and paste it on tpaste.us or something equivalent 2025-02-09 05:39:56 i'll format it with wikitags later. thanks 2025-02-09 05:56:40 "if daemon, use apk" <- Daemon, yes. With NetworkManager. I want all traffic through tor as an option. It is sadly becoming more necessary by the dat 2025-02-09 05:56:42 Day 2025-02-09 05:56:56 And I am not knowledgeable in this space 2025-02-09 05:57:18 Switching my family to TailsOS is not an option, either 2025-02-09 06:13:21 I have played around with the daemon on other operating systems, mostly ubuntu and openbsd; if anything like that its rather straightforward: edit config, run daemon, serivce is generated. Then torsocks can wrap around any call even if something simple like a curl or even ssh. 2025-02-09 07:39:52 "are using both can you jot..." <- feel free to take inspiration of https://wiki.postmarketos.org/wiki/Tor 2025-02-09 08:50:51 fossdd, thanks for the link to pmos wiki. will update wiki.a.o based on that.. 2025-02-09 10:01:30 fossdd, done thanks 2025-02-09 10:10:38 is there a difference between echo test > fifo-file and echo test >> fifo-file? 2025-02-09 10:14:39 test 2025-02-09 10:15:51 frag: I think the second adds `test` to the end of the file 2025-02-09 10:15:57 >> is append 2025-02-09 10:16:49 yes, i know it does that to a normal file, but fifo files are different eh 2025-02-09 10:36:04 guess it doesnt matter 2025-02-09 10:56:08 You cannot lseek on a fifo 2025-02-09 11:20:25 Anyone wants how to set 'pinentry-bemenu' for gpg-agent on runtime? I would like that it uses bemenu version when called as 'dpass' and the curses when using 'fpass' 2025-02-09 11:20:31 s/wants/knows 2025-02-09 11:20:52 I am who wants :P 2025-02-09 11:25:24 it seems that I need a wrapper script 2025-02-09 11:36:09 yeah it works 2025-02-09 12:20:53 I can see that acpid is actually supposed to distribute an init.d file, in package(): install -D -m 755 "$srcdir"/$pkgname.initd etc/init.d/$pkgname 2025-02-09 12:21:22 But: $ apk info -L acpid | grep -F -e contains: -e init.d 2025-02-09 12:21:22 acpid-2.0.34-r6 contains: 2025-02-09 12:21:26 Is that a bug? 2025-02-09 12:22:07 Before that I just did to be sure: (1/1) Reinstalling acpid (2.0.34-r6) 2025-02-09 12:22:07 Executing busybox-1.37.0-r14.trigger 2025-02-09 12:22:31 (that's on edge) 2025-02-09 12:23:14 s/acpid/acpid-openrc 2025-02-09 12:23:27 Sorry? 2025-02-09 12:23:41 Run the same command again, but then with the acpid-openrc package 2025-02-09 12:25:16 No no, it's installed as part of the main package (that's why I said ā€œin package()ā€) 2025-02-09 12:26:01 In any case, that's the first thing I did: $ apk info -L acpid-openrc | grep -F -e contains: -e init.d 2025-02-09 12:26:04 acpid-openrc-2.0.34-r6 contains: 2025-02-09 12:26:08 ie empty 2025-02-09 12:26:21 subpackages can move things away from packages 2025-02-09 12:26:33 Yeah, it doesn't (at least not explicitely) 2025-02-09 12:26:46 It's enough to have an -openrc subpackage 2025-02-09 12:26:55 the default split function would do it 2025-02-09 12:27:01 or maybe that ā€œdefault_openrcā€ does that, but I couldn't find it documented 2025-02-09 12:27:01 https://pkgs.alpinelinux.org/contents?file=&path=&name=acpid-openrc&branch=edge&repo=&arch= 2025-02-09 12:27:26 Hummm, so it's a bug? 2025-02-09 12:27:32 https://gitlab.alpinelinux.org/alpine/abuild/-/blob/master/abuild.in#L2138-2144 2025-02-09 12:28:18 apk fix tells me everything's good 2025-02-09 12:28:25 https://tpaste.us/9XJ0 2025-02-09 12:28:53 Yep, not here 2025-02-09 12:29:26 https://clbin.com/XAoWM 2025-02-09 12:29:26 What does apk policy acpid-openrc return? 2025-02-09 12:29:47 https://clbin.com/ESzVO 2025-02-09 12:30:01 Not sure what that means 2025-02-09 12:30:20 It tells me what version of the package is installed and what repos you have 2025-02-09 12:30:42 Did you run `apk fix acpid-openrc` explicitly? 2025-02-09 12:30:44 Well, maybe I missed to explain something, though I thought it was implied: I do have an init.d file, but that's the busybox on 2025-02-09 12:30:48 s/on/one/ 2025-02-09 12:30:55 Yeah I did it too (the fix) :/ 2025-02-09 12:31:20 apk info -W /etc/init.d/acpid 2025-02-09 12:31:26 https://clbin.com/BGIKw 2025-02-09 12:32:03 oh 2025-02-09 12:32:10 Now it worked 2025-02-09 12:32:16 By reinstalling *both* at once oO 2025-02-09 12:33:18 busybox-openrc has `replaces="acpid-openrc"` 2025-02-09 12:33:22 So it will take over that file 2025-02-09 12:33:36 so it's by design 2025-02-09 12:33:40 Shouldn't it be the other way around (as busybox-openrc is the default) 2025-02-09 12:35:48 Obviously you don't want the system to break everytime busybox-openrc is updated 2025-02-09 12:35:59 Or is the ā€œreplaceā€ only applied once, at first install? 2025-02-09 12:36:23 replaces only says ā€œSpecifies packages that the package replaces. This is typically used for packages renamed by upstream.ā€ 2025-02-09 12:37:17 It tells apk that file conflicts are expected and that the package defining the replaces will now own and provides that file 2025-02-09 12:38:59 The replaces was added as part of this MR: https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/37834 2025-02-09 12:39:08 Yeah, so having `replaces="acpid-openrc"` in busybox-openrc is definitely the opposite of what's needed 2025-02-09 12:39:35 Although just inversing the logic won't work, as obviously any *openrc package that busybox-openrc now replaces can't replace busybox-openrc 2025-02-09 12:40:21 Wait, it was actually reviewed by several persons? 2025-02-09 12:42:00 ā€œHas anyone tested this MR? ā€” yes, i verified all the upgrade paths (mdev/mdevd/acpid/udev) work while having whatever installedā€ 2025-02-09 12:42:03 hummm 2025-02-09 12:48:25 ā€œOver time I intend to keep splitting functionality from busybox-openrc and making it more predictable and maintainable to provide alternatives.ā€ 2025-02-09 12:48:46 Break first, fix later :> 2025-02-09 13:03:09 ikke, actually it seems that apk just replaces files, so inverting the logic would work, wouldn't it? 2025-02-09 13:03:20 (from src/database.c) 2025-02-09 13:03:22 yes, it would 2025-02-09 13:03:33 It's only for file conflicts 2025-02-09 13:03:36 ok 2025-02-09 13:04:03 Though maybe it would be needed to add an uninstall trigger to restore busybox-openrc? 2025-02-09 13:04:16 (not sure if that part makes sense, but in order to restore the default file) 2025-02-09 13:05:26 Or maybe that's already handled as an automatic feature of apk 2025-02-09 13:05:53 afaik it would not restore files if the replace package is uninstalled 2025-02-09 13:06:28 Yeah, I'm not sure how file tracking is done, but that could be a bit complex 2025-02-09 13:10:29 Sounds like a perfect Sunday task, I'll propose a patch, at least just to start a discussion maybe 2025-02-09 13:47:33 What's that? ā€œoils-for-unix-binsh-0.26.0-r1ā€ 2025-02-09 13:47:44 (reported by `apk search binsh` 2025-02-09 13:47:45 ) 2025-02-09 13:48:06 Can't find any mention of that in any APKBUILD 2025-02-09 13:48:29 Nor by searching the git commit history (though it's still running, but has been for a while) 2025-02-09 13:52:44 Yep, git log -G returned nothing 2025-02-09 13:54:10 https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/testing/oils-for-unix/APKBUILD 2025-02-09 13:55:11 Oh, makes sense, thanks ikke 2025-02-09 14:34:18 hi, with any mirror, apk update gives me "updating and opening url: No such file or directory" 2025-02-09 14:34:52 can't figure out what exactly goes wrong 2025-02-09 14:37:16 ok, I figured it out 2025-02-09 14:39:55 What was the issue? 2025-02-09 14:41:31 had to find the right combination of working mirror + right syntax 2025-02-09 14:44:40 I like this distro so far, but I wonder, are brave-browser builds available? 2025-02-09 14:46:15 jal, https://gitlab.alpinelinux.org/alpine/aports/-/issues/11659 2025-02-09 14:53:17 Hello, anybody familiar with the following error when trying to run a second compose file with podman-compose: " netavark: code: 1, msg: iptables: Chain already exists" ? 2025-02-09 14:55:26 (sorry wrong nick) Hello, anybody familiar with the following error when trying to run a second compose file with podman-compose: " netavark: code: 1, msg: iptables: Chain already exists" ? 2025-02-09 14:56:38 Haven't heard about it 2025-02-09 14:56:58 Not sure if related, but there was an iptables bug that affected docker 2025-02-09 14:57:17 what version of iptables do you have? 2025-02-09 15:00:08 ikke: iptables v1.8.11 (nf_tables) 2025-02-09 15:00:37 what does `apk version iptables` return? 2025-02-09 15:01:00 the fun thing is that I have another machine where it works (same version, same arch) 2025-02-09 15:01:29 ikke:iptables-1.8.11-r1 = 1.8.11-r1 2025-02-09 15:02:21 Ok, that includes the fix 2025-02-09 15:18:47 One difference between both machines is the container runtime: crun (where it fails), runc (where it works). I don't remember installing any of these manually. Could this be related? 2025-02-09 15:40:30 Saijin_Naib[m]: re tor, there's a lot to it, potentially, depending on your threat model. torsocks is leaky (and there's an "interrupted syscall" problem with hostname lookups with musl i've not debugged). setting transport in torrc should be done with great care. 2025-02-09 15:41:40 Saijin_Naib[m]: arti will replace c-tor eventually, and they're prototyping a replacement for torsocks using linux namespaces, but that's a ways off. 2025-02-09 16:03:12 ikke, sorry got sidetracked on another thing 2025-02-09 16:03:14 https://clbin.com/pWRDZ 2025-02-09 16:03:18 Tested, works 2025-02-09 16:03:53 Though it seems there might seem to be a better way(s) to handle that 2025-02-09 16:09:56 The challenge is that busybox-openrc will almost certainly be installed (if you care about init files anyway), so the conflict will happen no matter what 2025-02-09 16:11:40 Sure, I meant in order to handle the conflict, not to avoid it 2025-02-09 16:12:05 Well, it's handled differently in most other packages though, with a different init filename ^^ 2025-02-09 16:12:25 (which also looks like something that could be handled better, but I guess it works good enough) 2025-02-09 16:15:34 oops 2025-02-09 16:15:37 Updated: https://clbin.com/t5vgf 2025-02-09 16:16:07 I started ā€œfixingā€ the extras_openrc package too, but it turned out there's no need (and it got in the commit as a left-over) 2025-02-09 18:37:31 i noticed the kernel module cdc_mbim is gone from linux-lts since v3.21. i wonder if it was intentional, can't seem to find a specific commit that did this 2025-02-09 18:38:30 krystianch: it most likely got lost with the update to linux 6.12 2025-02-09 18:40:14 should I file an issue on aports? 2025-02-09 18:40:48 Yes please 2025-02-09 18:52:38 thanks, #16898 2025-02-09 18:58:40 #16898 2025-02-09 19:13:12 crashed again using headphones, seemed like one sound source froze and one played, and computer didnt reboot, had to long-hold the manual button to turn it off... 2025-02-09 19:14:01 what creates /home/user/.cache/mesa_shader_cache btw? running from RAM its created with root permission... 2025-02-09 19:29:34 @invoked thanks for the extra context. I'll need to keep learning 2025-02-09 20:23:20 seems starting qutebrowser/mpv creates that mesa dir.. but i dont do that as root..