2023-08-01 00:02:15 <LACampbell> I think I finally have to stop using openbox. not maintained anymore, and glib marches on with strange breaking changes.
2023-08-01 00:19:18 <nangel> https://www.reddit.com/r/openbox/comments/ngqz4s/comment/gyt19xd
2023-08-01 00:20:03 <LACampbell> yeah even that's putting a nice spin on it. it's effectively abandoned AFAICT
2023-08-01 00:20:16 <LACampbell> (ie no new version for the glib issue)
2023-08-01 00:24:08 <nangel> nod
2023-08-01 00:25:25 <LACampbell> had a go at berry, very nice but very very minimal. code was pretty tidy, thought about doing some pull requests but there's a bunch sitting there waiting to be merged. maybe a fork.. hm
2023-08-01 00:25:40 <LACampbell> also probably time I try out FVWM, still going strong
2023-08-01 01:08:20 <somercet1> ouch, mousepad now requires libxfce4ui, switching to pluma. I hope it's better now.
2023-08-01 01:08:51 <Guest7531> Saijin_Naib[m]: ^ ^ ^
2023-08-01 01:09:07 <zcrayfish> lol I ghosted myself. oops
2023-08-01 01:38:32 <LACampbell> https://gist.github.com/LAC-Tech/44f032d8efa3b1c7a3d10da69f822071 every flatpak app I run I get this error and have difficulty with accessing files. dbus is running. anything else I need?
2023-08-01 01:41:21 <zcrayfish> dunno for sure, but do you have one of the xdg-desktop-portal* packages installed?
2023-08-01 01:41:54 <LACampbell> xdg-desktop-portal, xdg-desktop-portal-doc, xdg-desktop-portal-gtk
2023-08-01 01:45:12 <LACampbell> wait is xdg-desktop-portal some service that needs to run?
2023-08-01 01:45:23 <zcrayfish> dbus should handle it for you
2023-08-01 01:45:29 <zcrayfish> is the $XDG_RUNTIME_DIR set?
2023-08-01 01:46:52 <LACampbell> lightgrey:~$ echo $XDG_RUNTIME_DIR
2023-08-01 01:46:54 <LACampbell> /run/user/1000
2023-08-01 01:48:12 <zcrayfish> oh missed the xdg-settings not running there too... Is xdg-utils installed?
2023-08-01 01:53:55 <LACampbell> yes, they are installed
2023-08-01 02:07:14 <LACampbell> ok so I installed flatseal and none of the apps have d-bus permissions set. when is et them in the app, i get an exclamation point, and a different error message when i try and run them
2023-08-01 02:10:05 <LACampbell> god flatpak is a pain in the dick. there's always some extra stuff you have to do when you install it in the first place as well. I'll try and nuke it I guess
2023-08-01 02:11:45 <quinq> Sounds like a sane decision
2023-08-01 02:11:53 <zcrayfish> agree. I should have wrote down what the heck I did to get mine going
2023-08-01 02:14:32 <LACampbell> yeah would be good to add it to the wiki
2023-08-01 02:16:03 <Saijin_Naib[m]> somercet1 is that too heavy a depend for keyboard shortcut customization, especially given other core XFCE apps use it as well? 
2023-08-01 02:16:05 <quinq> flatpak usually doesn't solve what a proper package would
2023-08-01 02:16:51 <Saijin_Naib[m]> LACampbell the exclamation mark just shows that the parameters have been changed from the Flatpak manifest default. 
2023-08-01 02:17:16 <somercet1> Saijin_Naib[m], it's not a heavy dependency, it's just more dependencies, like xfconf.
2023-08-01 02:17:55 <somercet1> mousepad isn't my goto editor (vis is), but it's nice when editing plain text where you want wrap.
2023-08-01 02:18:53 <LACampbell> Saijin_Naib[m]: sorry which exclamation marks?
2023-08-01 02:19:11 <LACampbell> on in flatheal or whatever. got it
2023-08-01 02:19:15 <zcrayfish> in flatseal
2023-08-01 02:19:30 <zcrayfish> 
ACTION prefers T-head engines to flatheads, but whatever.

2023-08-01 02:19:32 <zcrayfish> ;)
2023-08-01 02:19:45 <Saijin_Naib[m]> somercet1 I guess I can try dropping it as a runtime depends and see what happens
2023-08-01 02:19:55 <LACampbell> ... took me a moment to realise you were talking about screwdrivers
2023-08-01 02:20:12 <zcrayfish> lol
2023-08-01 02:20:31 <somercet1> it's okay, I use mate-terminal and pluma seems to be working.
2023-08-01 02:20:31 <LACampbell> but yeah even turnong the dbus settings in flatseal, i still get the GDBUS error
2023-08-01 02:20:43 <somercet1> don't short the XFCE fans for my sake.
2023-08-01 02:20:53 <Saijin_Naib[m]> somercet1 are you not using XFCE desktop environment? 
2023-08-01 02:21:01 <Saijin_Naib[m]> Because it should already be present if you are
2023-08-01 02:21:11 <somercet1> no, I use sway
2023-08-01 02:21:19 <Saijin_Naib[m]> Ahhh, gotcha 
2023-08-01 02:21:24 <somercet1> I don't run a full DE, just GTK apps.
2023-08-01 02:21:47 <Saijin_Naib[m]> If I understand how XFCE likes to do things, if it is missing as a runtime depends, the feature just silently gets disabled 
2023-08-01 02:21:59 <Saijin_Naib[m]> And folks that have it will be able to use it
2023-08-01 02:22:04 <zcrayfish> I'm using mousepad on LXQT (openbox or labwc). I'm weird :)
2023-08-01 02:24:46 <LACampbell> how did lxqt mitigate that glib issue if they're using openbox?
2023-08-01 02:25:37 <zcrayfish> what's the glib issue?
2023-08-01 02:46:37 <LACampbell> zcrayfish: https://patchespromptly.com/glib2/
2023-08-01 02:47:40 <zcrayfish> ouch
2023-08-01 02:47:56 <zcrayfish> I would assume that Alpine has applied the patch given that openbox seems to run just fine
2023-08-01 02:48:21 <zcrayfish> but the obvious solution is for LXQT+labwc to finish hammering out their wayland bugs. 
2023-08-01 02:48:29 <zcrayfish> ;)
2023-08-01 02:48:42 <zcrayfish> I'll try to check later
2023-08-01 02:48:44 <LACampbell> openbox has a lot of crashy behaviour for me on alpine
2023-08-01 02:48:51 <LACampbell> that other wms do not
2023-08-01 02:48:58 <LACampbell> I assume thats the reason
2023-08-01 03:00:34 <zcrayfish> you using elogind, LACampbell ?
2023-08-01 03:03:28 <LACampbell> yeah. is that the issue?
2023-08-01 03:04:09 <zcrayfish> I was thinking that lack thereof could be
2023-08-01 03:06:11 <livy> Hi guys. I used setup-bootable to make a bootable USB drive. I can see it copy all files from /media/cdrom to the USB drive, and somehow makes it bootable.
2023-08-01 03:07:05 <livy> The USB drive was formatted as GPT with 'parted', with partition 1 having ESP flag set.
2023-08-01 03:07:22 <livy> I used setup-bootable on partition 2, which is a normal 'msftdata' partition.
2023-08-01 03:07:59 <livy> I can see that there is a 'efi/boot/bootx64.efi' inside partition 2.
2023-08-01 03:08:29 <livy> How can the motherboard firmware (I am using OVMF) know to use that file?
2023-08-01 03:09:04 <livy> It is supposed to find .efi files from ESP partition instead, isn't it?
2023-08-01 03:22:53 <livy> I moved the 'efi' directory from partition 2 to partition 1, then try to boot.
2023-08-01 03:23:34 <livy> It stucks at Grub 2.06, probably because the bootloader is hard-coded to boot from the same partition.
2023-08-01 04:23:53 <uuidNuniq> livy: grub.cfg is mssing 
2023-08-01 04:26:12 <livy> Yeah maybe I also need to move the 'boot' directory over (which contains grub.cfg). But that's not the point of the question.
2023-08-01 04:26:34 <livy> I am wondering why it works, since partition 2 is 'msftdata', not 'esp'.
2023-08-01 04:27:15 <livy> I guess the OVMF firmware goes against UEFI specification, and it search for .efi files from non-ESP partition to boot.
2023-08-01 04:27:48 <uuidNuniq> what you you want to do? install alpine to usb?
2023-08-01 04:28:57 <LACampbell> even uninstalling flatpak, reinstalling it, using flatseal to allow dbus connection, I still get "Can't get document portal: GDBus.Error:org.freedesktop.DBus.Error.NoReply: Message recipient disconnected from message bus without replying".
2023-08-01 04:28:59 <livy> I want to have a bootable Alpine on a partition (on USB), with some others partitions for /home, /var... etc.
2023-08-01 04:31:04 <livy> Maybe I will dual-boot on the same USB drive in the future, that's why I need to understand how Alpine can boot from non-ESP partition.
2023-08-01 04:31:52 <uuidNuniq> alpine is like any other distro that uses grub... 
2023-08-01 04:31:58 <uuidNuniq> nothing special.
2023-08-01 04:32:48 <livy> On https://wiki.alpinelinux.org/wiki/Installation, it reads: "Later, another directory on that same partition or another available partition may also be mounted as /home"
2023-08-01 04:33:24 <livy> Is there any documentation on the Wiki which intructs how to use a "directory" as /home? Currently I only know how to use partition.
2023-08-01 04:36:23 <uuidNuniq> a directory as home? mkdir /home
2023-08-01 04:37:11 <uuidNuniq> mkdir -p /path/to/some/dir ; ln -s /path/to/some/dir /home
2023-08-01 04:37:52 <livy> Um, simply using symbolic link?
2023-08-01 04:38:11 <uuidNuniq> or hardlink
2023-08-01 04:38:17 <uuidNuniq> or btrfs subvolume
2023-08-01 04:38:33 <livy> I thought it was something more complicated. My bad, I was overthinking about it.
2023-08-01 04:38:52 <livy> Btrfs subvolume approach sounds good.
2023-08-01 04:40:02 <uuidNuniq> livy: you can also bind mount : https://askubuntu.com/questions/465493/how-can-i-symlink-my-home-folder-from-another-drive
2023-08-01 04:40:25 <livy> Thanks, I'll try them out and see which suits my need the most.
2023-08-01 06:14:05 <Forza> livy: Btrfs subvols are great. 
2023-08-01 06:56:50 <livy> @Forza ZFS seems to be a better choice. Its features surpasses what can be done with Btrfs.
2023-08-01 06:57:58 <Forza> Zfs isn't bad either. I've used zfs quite a bit at work, but have transitioned to Btrfs because or better integration with most of Linux distros, and easier management 
2023-08-01 06:58:17 <livy> I am not sure how ZFS performs in embedded environment though. I ran Btrfs fine on OpenWrt routers with 64MiB of RAM. I do not know about resource requirement by ZFS.
2023-08-01 06:59:37 <i5yxt1550rd9gbko8o55[m]> please using btrfs in low mem
2023-08-01 06:59:54 <Forza> Not sure. But I think zfs isn't hugely memory hungry in simple setups. 
2023-08-01 07:00:53 <Forza> That said, alpine with btrfs in a VM can idle around 50 MiB ram 
2023-08-01 07:00:57 <zcrayfish> I love both those filesystems due to their optional transparent compression feature; great for cheapskates like me who like to hang into smaller hard disks (SSDs) for far too long :)
2023-08-01 07:01:20 <Forza> Heh yup. Saves TBW too
2023-08-01 07:01:38 <livy> Talking about ZFS, I only use it for like 1 week, and do not really understand about dataset vs zvol. They look the same to me.
2023-08-01 07:01:56 <Forza> Zvol is a block device 
2023-08-01 07:02:15 <Forza> Dataset is like btrfs subvols 
2023-08-01 07:02:49 <livy> ms-sqlserver only support XFS and EXT4, because its databases are sort of copy-on-write. Putting them on copy-on-write filesystems will be disatrous.
2023-08-01 07:03:01 <Forza> Basically a directory on super powers. You can snapshot it. 
2023-08-01 07:03:08 <livy> I am wondering what if I create a zvol, format it with EXT4, then put ms-sqlserver databases on it.
2023-08-01 07:03:30 <Forza> livy: zfs still does cow behind 
2023-08-01 07:04:01 <Forza> Zvol = loop device on a file. 
2023-08-01 07:04:17 <Forza> Just a little easier to use 
2023-08-01 07:05:09 <livy> I do not get what will happen with ms-sqlserver in that case. Sounds like breaking the law, lol.
2023-08-01 07:06:14 <livy> Do you think it is a good idea to use zvols as virtual machine disks?
2023-08-01 07:06:16 <Forza> I use mssql o in docker on btrfs. What's the problem? 
2023-08-01 07:06:33 <Forza> Zvols makes life easier, so why not? 
2023-08-01 07:06:54 <Forza> Zvols are also nice as iscsi targets and stuff like that 
2023-08-01 07:07:32 <livy> Ah, I think I was mistaken. I've heard that virtual disks (.qcow2) are copy-on-write, and they hate being put on copy-on-write filesystems.
2023-08-01 07:07:59 <livy> But using the zvol directly as "physical" disk for virtual machines are fine, I guess.
2023-08-01 07:09:59 <Forza> There is no difference 
2023-08-01 07:10:31 <Forza> I mean no difference to using zvol to a normal raw file as backing 
2023-08-01 07:11:02 <Forza> You can use "raw" images instead of qcow2 
2023-08-01 07:11:25 <livy> Raw images do not have compression. They cannot do snapshot either.
2023-08-01 07:11:46 <Forza> Qcow2 is good if you need snapshots, etc. But since zfs already provides this... 
2023-08-01 07:12:24 <streaksu[m]> I guess you can snapshot with raw images with the extremely snapshot oriented tooling of `cp <image> <image>-backup` :p
2023-08-01 07:12:50 <livy> Maybe, but it still does not work with VM snapshot: (qemu) savevm snapshot1
2023-08-01 07:13:00 <streaksu[m]> true
2023-08-01 07:13:16 <Forza> Does zfs support reflink copies on Linux (cp --reflink) 
2023-08-01 07:13:48 <Forza> I always use btrfs snapshots or reflinks, and never qemu snapshots 
2023-08-01 07:13:53 <Forza> Much easier 
2023-08-01 07:15:20 <Forza> The cow-on-cow issue is mostly related to performance issues due to fragmentation 
2023-08-01 07:15:50 <Forza> On btrfs we can solve this by defragmenting the files now and then 
2023-08-01 07:17:40 <Forza> On zfs, I think the performance is a bit better on small writes anyway, so it's probably ok
2023-08-01 07:32:00 <grawity> Forza: AFAIK not yet but the necessary fundamentals (block cloning) are in the latest pre-release, not sure if that includes the --reflink API yet, but if not then it'll be coming pretty soon
2023-08-01 07:32:56 <Forza> Then using datasets or zvols is the way to go for snapshotting 
2023-08-01 07:33:32 <grawity> https://github.com/openzfs/zfs/issues/405#issuecomment-1618411580
2023-08-01 07:33:50 <Forza> Privately i have moved to use virtiofs for rootfs on all my Alpine vms. 
2023-08-01 07:41:22 <Forza> They boot/reboot instantly too :) 
2023-08-01 07:41:52 <livy> I only use virtiofs for sharing files between host and guests.
2023-08-01 07:42:22 <Habbie> Forza, how's virtiofs performance?
2023-08-01 07:44:12 <Forza> It's OK. Haven't compared much to using vm images, but for my uses I haven't noticed anything bad. 
2023-08-01 07:44:39 <Forza> Block based backing would be faster I am sure. 
2023-08-01 07:45:47 <livy> virtiofs0_sock=~/"virtiofs0.sock"; virtiofs0_source=~/"OneDrive";
2023-08-01 07:46:02 <Forza> There's a bit of tuning that could be done with virtiofsd, so maybe it could be better. At least I think it is faster than nfs/smb/plan9
2023-08-01 07:46:11 <livy> Can the socket be shared between multiple -chardev devices? Or between multiple VMs?
2023-08-01 07:46:25 <Forza> I have not tried that
2023-08-01 07:46:45 <Forza> Libvirt sets up a separate socket for each share and vm
2023-08-01 07:47:16 <livy> Oh, you are using libvirt? I use pure QEMU commands. :D
2023-08-01 07:47:50 <Forza> I'm too lazy for that once I got into libvirt 
2023-08-01 07:48:44 <Forza> But sharing a socket, you want to share the same directory to several vms? 
2023-08-01 07:50:13 <Forza> What about privilege separation if you share same virtiofsd process? 
2023-08-01 07:50:26 <livy> I think sharing the same ~/OneDrive directory to several VMs is fine, and should not be a problem.
2023-08-01 07:50:41 <livy> I just do not know if I need to create multiple virtiofsd processes for each VM.
2023-08-01 07:51:09 <Forza> I guess you could try 
2023-08-01 07:53:42 <Forza> But onedrive... How will it handle syncs? Seems might be slow 
2023-08-01 07:54:30 <livy> I want to share that directory with the Windoze VM, so I do not need to have another copy of OneDrive inside the VM.
2023-08-01 07:54:55 <livy> Further more, OneDrive on Windows always updates almost instantly when files are changed.
2023-08-01 07:55:06 <livy> OneDrive on Linux -- I think it is default to 5 minutes.
2023-08-01 07:55:15 <Forza> It probably will work. Haven't tried virtiofs with a windows guest yet, but I heard it should work 
2023-08-01 07:55:52 <livy> It does, I have been using this setup. I just haven't tried to share the same virtiofsd betweem multiple Windows VMs though.
2023-08-01 08:01:56 <Forza> Don't think it should be a problem, at least if your apps use proper locking when writing/editing, and don't do local catching 
2023-08-01 08:50:58 <livy> (11/107) Installing linux-firmware-yamaha (20230515-r6)
2023-08-01 08:50:58 <livy> @Forza For some reason I could not install 'zfs-lts' package. I started with a bootable live CD session, and type 'apk add zfs-lts'.
2023-08-01 08:50:58 <livy> ERROR: linux-firmware-yamaha-20230515-r6: failed to rename lib/firmware/yamaha/.apk.62f450b5da4105ea405bceef0a435719e3a1319c91b54f2d to lib/firmware/yamaha/ds1_ctrl.fw.
2023-08-01 08:50:58 <livy> Everything after the 10th package just failed and it complained that it cannot rename files.
2023-08-01 08:52:13 <Forza> Weird 
2023-08-01 08:52:39 <Forza> I haven't tried zfs on Alpine though
2023-08-01 08:53:05 <Forza> Is there a problem accessing the files in that dir? 
2023-08-01 08:54:12 <livy> I don't think there was. I booted out of an ISO. Everything is loading into RAM.
2023-08-01 08:54:25 <livy> I changed the repository, it did not help.
2023-08-01 08:54:55 <livy> Package 'zfs' installed just fine, but it could not work without the kernel module 'zfs-lts'.
2023-08-01 09:04:10 <Forza> Not sure what to do about the firmware files though. Not seen that issue before 
2023-08-01 09:05:01 <livy> Anyone using ZFS out there? Plz help. The error is easy to reproduce with a fresh Alpine 3.18.2 ISO.
2023-08-01 09:14:13 <Forza> But can you really do apk update/add from an ISO? 
2023-08-01 09:14:45 <Forza> Or is the problem that you are using zfs as filesystem and you get these errors? 
2023-08-01 09:15:49 <livy> I booted from an ISO as backend of QEMU scsi-cd. The whole OS was loaded into RAM. I assigned 4GiB of RAM to the Alpine VM.
2023-08-01 09:17:06 <livy> It was just like putting Alpine CD to the CD-ROM drive and boot. I am trying to format a partition (of another virtual disk) as ZFS.
2023-08-01 09:17:26 <livy> But I was not even able to install ZFS kernel module.
2023-08-01 09:26:28 <Forza> Hm. 
2023-08-01 09:27:24 <Forza> Just a thought. Can you install on ext4, reboot into the new install, add your packages. Then reboot with a second virtual disk attached, format to zfs and rsync over everything? 
2023-08-01 09:27:55 <Forza> I do a similar dance when installing alpine on virtiofs rootfs.
2023-08-01 09:33:27 <livy> qemu-system-x86_64 -machine q35 -accel kvm -cpu host -smp 2 -m 1024M -cdrom "https://dl-cdn.alpinelinux.org/alpine/v3.18/releases/x86_64/alpine-virt-3.18.2-x86_64.iso";
2023-08-01 09:33:47 <livy> I need to go home right now. In the meantime, can you try the above?
2023-08-01 09:39:36 <i5yxt1550rd9gbko8o55[m]> kernel issue
2023-08-01 09:40:01 <i5yxt1550rd9gbko8o55[m]> you have to using same version zfs-lts to handle it
2023-08-01 09:40:29 <i5yxt1550rd9gbko8o55[m]> find same version zfs-lts and make it offline on alpine iso
2023-08-01 09:40:44 <i5yxt1550rd9gbko8o55[m]> or maybe extend version has that package
2023-08-01 09:45:03 <i5yxt1550rd9gbko8o55[m]> zfs-virt is exist,but virt iso may not include this package
2023-08-01 11:07:14 <livy> Is there any way I can use ZFS out of the box without complex workaround like that?
2023-08-01 11:07:53 <livy> Alpine also has an "edge" update mode. Imagine what kind of disaster I might face in a production environment.
2023-08-01 11:08:38 <livy> And I haven't even used edge yet. 'zfs-lts' intallation fails out of the box on 3.18.2
2023-08-01 11:50:02 <Forza> livy: you could make a template vm using the work around, then base your new vms on that 
2023-08-01 11:50:31 <Forza> Still not greatest perhaps. It is what I usually do anyway, as I have all my custom tweaks all ready then. 
2023-08-01 11:51:08 <livy> What I worry about is when the system updates... it probably breaks the whole production server.
2023-08-01 12:06:31 <i5yxt1550rd9gbko8o55[m]> <livy> "What I worry about is when the..." <- backup old kernel
2023-08-01 13:24:10 <Janik[m]> I noticed that the docker-rootless documentation is incomplete. Here is the wiki article https://wiki.alpinelinux.org/wiki/Docker#Docker_rootless but these instruction only install the dockerd-rootless binary but no according init.d files.
2023-08-01 13:24:10 <Janik[m]> Can someone reccomend a way to fix this? Shouldn't the docker-rootless-extras package include the initd file as does the docker package?
2023-08-01 13:28:55 <ikke> You are meant to run docker-rootless as your own user, not as a service 
2023-08-01 13:31:22 <ikke> It expects HOME and XDG_RUNTIME_DIR to be set 
2023-08-01 13:36:00 <Janik[m]> <ikke> "You are meant to run docker-..." <- Ah I see, and openrc cannot run services as a user but only as root?
2023-08-01 13:36:38 <Janik[m]> what else would be a good way to automatically start dockerd-rootless on each boot?
2023-08-01 13:36:43 <Janik[m]> <ikke> "It expects HOME and XDG_RUNTIME_..." <- what do you mean by it? dockerd-rootless?
2023-08-01 13:37:54 <ikke> Janik[m]: it can run it as a user, but not as the current user
2023-08-01 13:38:35 <Janik[m]> I see, so for a single user system that would still be a viable option
2023-08-01 13:41:08 <pjs> anyone know about NetworkManager not showing wifi. and `nmcli device list` says it's unmanaged. So using the NM applet or GUI shows everything greyed out. Not a huge deal but kind of annoying. Also can't setup VPN connections though it let's me start it. Just has all the fields disabled.
2023-08-01 13:41:29 <pjs> I'm using wpa_supplicant (which was setup via setup-aline)
2023-08-01 15:09:58 <alternateved> pjs: did you see and follow this guide? https://wiki.alpinelinux.org/wiki/NetworkManager
2023-08-01 15:12:43 <pjs> alternateved: I did not see that, and definitely am missing some steps. thank you!
2023-08-01 15:43:57 <Janik[m]> What is the best way to set XDJ_RUNTIME_DIR with alpine? I found this: https://wiki.alpinelinux.org/wiki/Wayland
2023-08-01 15:43:57 <Janik[m]> But all the options mention that they take care of it for logged in users. I want to start dockerd-rootless at boot, and thus also when the user is not logged in
2023-08-01 15:45:37 <dewdude> day 5: my raid is almost done building
2023-08-01 16:00:43 <cyclisme24[m]> livy: got Alpime installed on zfs an a x230 laptop during a direct install from a 3.17 iso (from memory). Followed a guide on the wiki. It is fine, it updates fine. Now on 3.18 and up to date, not broken so far. No EFI or seperate boot partition, the lot is encrypted. Not sure if that helps or not. 
2023-08-01 16:05:03 <pjs> alternateved: thanks. I think it was the missing plugdev group. Seems to be fine now
2023-08-01 16:11:58 <tmtt> Hi! I just finished installing Alpine Linux on my personal laptop with Gnome. Is there a way to change the keyboard layout used by GDM when logging in? My laptop uses an AZERTY (fr) keyboard layout, and, while it works on the tty and inside a Gnome session, GDM keeps using the QWERTY layout. I couldn't find much info on this. (I tried this solution from the Arch Wiki, with no
2023-08-01 16:12:00 <tmtt> luck: <https://wiki.archlinux.org/title/Keyboard_configuration_in_Xorg#Using_X_configuration_files>.) Thanks a lot in advance :)
2023-08-01 16:18:45 <grawity> tmtt: gdm might not always use Xorg, but what it does use is gnome-shell, so check out the arch wiki page on other gdm settings, can't look it up at the moment but it's the one that talks about /etc/dconf
2023-08-01 16:19:05 <tmtt> I'll check that out. Thanks!
2023-08-01 16:19:16 <grawity> look up the gsettings/dconf setting where GNOME stores your xkb keymaps, then copy it from your own config into gdm's
2023-08-01 16:19:25 <grawity> there's a third-party "gdm-settings" app that makes it much more convenient
2023-08-01 16:19:41 <tmtt> I see
2023-08-01 16:28:35 <tmtt> Also, is there a way to use pipewire instead of pulseaudio when installing Gnome (either using `setup-desktop gnome' or `apk a')
2023-08-01 16:28:44 <tmtt> `apk add gnome')*
2023-08-01 16:29:32 <tmtt> Trying to remove pulseaudio after gnome is installed results in a dependency error
2023-08-01 16:30:48 <grawity> is there a pipewire-pulse package?
2023-08-01 16:31:00 <Ermine> Yes, it is
2023-08-01 16:33:17 <tmtt> When installed through apk, gnome pulls pulseaudio and pipewire-libs (+ other stuff) but not pipewire-pulse
2023-08-01 16:35:49 <grawity> oh yeah, the keyboard layout gsetting is: org.gnome.desktop.input-sources sources [('xkb', 'us'), ('xkb', 'lt')]
2023-08-01 16:41:46 <gzmorell> Hi I have installed alpine in diskless mode on cm4 board. Is it possible to build a kernel module? Which package contains the kernel headers?
2023-08-01 16:45:08 <gzmorell> I have found linux-rpi-dev but it is a more recent version than the one running.
2023-08-01 16:50:06 <DaleYPearlman[m]> Do you know that you can earn $5,000 or more weekly from crypto Trading? With Just $500… 100% Inbox Admin on Telegram for more details 👇👇👇👇👇👇 https://t.me/PROFITSWITHSTEVE
2023-08-01 17:00:35 <pjs> should XDG_RUNTIME_DIR be set at login? 
2023-08-01 17:51:33 <witcher> Is my assumption right that all new packages should go into the testing repo by default? What's the strategy to transitioning this into community and possibly testing?
2023-08-01 17:52:16 <witcher> To be more precise: *when* can it go to community or main
2023-08-01 17:53:27 <Janik[m]> <pjs> "should XDG_RUNTIME_DIR be set at..." <- I think it should be set before login. I want openrc initd run docker-rootless right after boot. Is this possible?
2023-08-01 17:57:06 <minimal> witcher: when it is "ready"
2023-08-01 17:58:20 <witcher> minimal: what is this roughly defined as? somtehing like "tested by some people and no major bugs in X timeframe"?
2023-08-01 17:58:41 <minimal> it's not defined :-)
2023-08-01 17:59:25 <minimal> I've sent on occasion some packages be merged to testing and then *immediately* moved to community (by devs with merge privs)
2023-08-01 17:59:30 <minimal> s/sent/seen/
2023-08-01 17:59:51 <witcher> I see, thank you!
2023-08-01 18:00:45 <minimal> witcher: the "requirements" for main are stronger than for community as main is supported longer
2023-08-01 18:01:35 <minimal> plus I think there is a wish to keep "main" for important/base packages
2023-08-01 18:02:14 <witcher> minimal: that's understandable. are packages in main only maintained by alpine devs or can anyone do that?
2023-08-01 18:03:15 <minimal> don't think it is restricted to core devs
2023-08-01 18:04:00 <pj> it's definitely not restricted now
2023-08-01 18:05:41 <witcher> the wiki page on creating an alpine package lists nmeum and jrutka on GitHub as code reviewers (as of Feb 18 2018). I suppose this is not up-to-date anymore?
2023-08-01 18:07:21 <pj> Anyone can review MRs
2023-08-01 18:07:37 <pjs> Janik[m]: before login? I thought it was supposed to be set by pam or some login module
2023-08-01 18:07:39 <pj> It's still up to few people with commit privileges to merge them
2023-08-01 18:09:26 <witcher> i seem to remember alpine accepting patches via email. is this still possible, or is gitlab preferred?
2023-08-01 18:11:38 <pj> It's not possible to send patches via email
2023-08-01 18:11:43 <pj> (anymore)
2023-08-01 18:13:08 <witcher> good to know. thank you for answering my many questions :)
2023-08-01 18:22:16 <Janik[m]> <pjs> "Janik: before login? I thought..." <- normally it is yes but that wont work if I run dockerd-rootless after boot, will it?
2023-08-01 18:57:22 <pjs> Janik[m]: I feel like we're talking about 2 different things
2023-08-01 19:16:39 <Janik[m]> <pjs> "Janik: I feel like we're talking..." <- could you elaborate, please?
2023-08-01 20:05:50 <wyk72> hi guys
2023-08-01 20:06:14 <wyk72> I have a problem building a driver with AKMS on alpint 3.18.2 kernel 6.1.34
2023-08-01 20:06:28 <wyk72> it seems linux-lts-dev is missing ofr this kernel
2023-08-01 20:06:37 <wyk72> akms: Building module rtl8188fu/5_git20230108 for kernel 6.1.34-0-lts akms: ERROR: Package linux-lts-dev=6.1.34-r0 is not available in the repositories!
2023-08-01 20:07:13 <wyk72> any pointers ?
2023-08-01 20:09:26 <ikke> https://pkgs.alpinelinux.org/packages?name=linux-lts-dev&branch=v3.18&repo=&arch=&maintainer=
2023-08-01 20:10:31 <wyk72> this is a run-from-ram machine...it's kind of difficult to update the kernel for such a small machine (I have to use ..a bigger one)
2023-08-01 20:10:46 <wyk72> so I need to update to 6.1.42 ?
2023-08-01 20:10:58 <stanofsteel> Hi (first time irc user), I just did a fresh install on my X200 and am trying to use iwd for connecting to my network. When I correctly enter the passphrase for the network it says "Operation failed". I have installed linux-firmware. Any a clue?
2023-08-01 20:11:33 <ikke> stanofsteel: welcome
2023-08-01 20:11:54 <stanofsteel> thx
2023-08-01 20:12:05 <ikke> stanofsteel: does `ip link` show your wireless interface?
2023-08-01 20:12:29 <stanofsteel> it does
2023-08-01 20:13:35 <ikke> Anything mentioned in `dmesg`?
2023-08-01 20:13:54 <wyk72> so package linux-lts-dev=6.1.34-r0 is not available anywhere ?
2023-08-01 20:14:13 <ikke> wyk72: we don't keep old versions of packages around sadly
2023-08-01 20:14:57 <vardrop> i have a problem trying to boot from nvme using clover (i know yikes), after booting it fails loading the initramfs. anyone knows some steps to troubleshoot?
2023-08-01 20:15:22 <ikke> vardrop: fails to load the initramfs how?
2023-08-01 20:16:13 <stanofsteel> ikki: it outputs a lot. When i grep for wlan0 in the output i dont notice anything weird, though im not sure what to look for
2023-08-01 20:16:15 <ikke> stanofsteel: is it possible to manually bring the wlan0 interface up?
2023-08-01 20:16:19 <vardrop> ikke: its stuck at "loading initial ramdisk ...". it was booting fine today but then i stupidely decided to fiddle with the fans...
2023-08-01 20:16:21 <ikke> ip link set dev wlan0 up
2023-08-01 20:17:00 <ikke> stanofsteel: I prefert to keep the chat here so that others (that have more experience with these things) can chime in
2023-08-01 20:17:54 <stanofsteel> ikke: i have done 'ifconfig wlan0 up' and didnt get an error, if that answers your question
2023-08-01 20:18:43 <vardrop> i forgot to add that i tried to fix a previous mount sysroot uuid not found error and ran mkinitramfs which left me stranded here
2023-08-01 20:19:03 <stanofsteel> ikke: didn't notice you sent the command, that one works as well
2023-08-01 20:19:50 <ikke> stanofsteel: I don't have experience with iwd, so not sure how you can troubleshoot that
2023-08-01 20:19:53 <ikke> but others here do
2023-08-01 20:20:18 <stanofsteel> ikke: all good, thanks for the help
2023-08-01 20:20:32 <wyk72> I've noticed a little clash with usb-utils package also: it seems now busybox provides an "lsusb" command; when you install usb-utils the lsusb utility gets into /usr/bin as lsusb.apk-new
2023-08-01 20:20:48 <wyk72> that's on 3.18.2 fresh install run-from-ram
2023-08-01 20:21:32 <wyk72> the busybox lsusb does not get any hwids 
2023-08-01 20:21:33 <ikke> hmm, it should only create .apk-new files for protected paths, which /usr/bin should not be
2023-08-01 20:21:41 <wyk72> lsusb from usbutils does
2023-08-01 20:21:49 <ikke> and lsusb has been part of bb since a long time
2023-08-01 20:22:11 <wyk72> oh I see, that's weird
2023-08-01 20:23:12 <ikke> What should happen is that installing usbutils overwrites the symlink that busybox creates
2023-08-01 20:25:03 <wyk72> I'll rebuild this run-from-ram usb key from scratch then, update kernel and retry
2023-08-01 20:44:23 <vardrop> eh ill prob just reinstall alpine
2023-08-01 21:19:51 <raspbeguy> hello, do some Alpine Linux installation ISOs with serial enabled already exist somewhere or do I have to make it myself?
2023-08-01 21:31:51 <minimal> raspbeguy: the virt ISO has serial enabled, the standard/extended do not seem to
2023-08-01 21:32:36 <minimal> https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/scripts/mkimg.standard.sh#L90
2023-08-01 21:34:20 <raspbeguy> minimal: ok. Too bad I need that for physical machines
2023-08-01 21:34:42 <ikke> raspbeguy: fyi, you can enable it on the kernel command line
2023-08-01 21:34:45 <ikke> while booting
2023-08-01 21:35:46 <ikke> if you see the `boot:` line, press tab, press lts console=ttyS0
2023-08-01 21:35:50 <ikke> or something like that
2023-08-01 21:38:17 <raspbeguy> ikke: yes I know. I was just looking for an ISO where it is already enabled
2023-08-01 21:38:44 <raspbeguy> Because no monitor
2023-08-01 21:53:51 <Forza> Is the line "kernel_addons" a list of kernel modules or something else? 
2023-08-01 22:35:53 <Forza> I'm discovering Starship. Makes it easier to keep track of what git branch I'm currently in :) 
2023-08-01 22:37:23 <Forza> https://paste.tnonline.net/files/B71uX2jOnpzb_Screenshot_20230802_003636_Termius.png
2023-08-01 23:01:55 <minimal> ikke: that will enable linux serial console, but won't enable syslinux/grub serial support (though he may not need that)
2023-08-02 00:07:38 <DaleYPearlman[m]> Do you know that you can earn $5,000 or more weekly from crypto Trading? With Just $500… 100% Inbox Admin on Telegram for more details 👇👇👇👇👇👇 https://t.me/PROFITSWITHSTEVE
2023-08-02 00:22:50 <streaksu[m]> happens
2023-08-02 00:52:03 <Saijin_Naib[m]> Forza can that be used with busybox ash? 
2023-08-02 02:03:15 <LACampbell> https://github.com/flathub/md.obsidian.Obsidian/issues/184 feel like the reply is gonna be "you're using a weird linux, closing" but worth a shot
2023-08-02 02:14:23 <livy> @cyclisme24[m] It fails on all versions, because the Alpine kernel version and the zfs-lts package versions are mismatched.
2023-08-02 02:15:20 <livy> @cyclisme24[m] It fails on all Alpine version because the version of Alpine kernel and the version of zfs-lts are mismatched.
2023-08-02 02:16:09 <Saijin_Naib[m]> ikke: Could you ikke-matically ask the maintainer of Thonny to check out https://gitlab.alpinelinux.org/alpine/aports/-/issues/15140
2023-08-02 02:16:26 <Saijin_Naib[m]> I would make the MR myself, but I'm not exactly sure what needs to happen to fix the .desktop file not working right
2023-08-02 02:23:03 <Saijin_Naib[m]> Also, Celeste, if you're in here, thanks and sorry
2023-08-02 02:23:15 <Saijin_Naib[m]> I think my next MR will be slightly less fucked. I'm not quite as rusty
2023-08-02 02:26:29 <zcrayfish> keep up the good work Saijin_Naib[m] \o/
2023-08-02 02:27:09 <Saijin_Naib[m]> Thanks! I hope to get (much) better
2023-08-02 02:27:13 <Saijin_Naib[m]> I have big dreams...
2023-08-02 02:27:25 <Saijin_Naib[m]> I want QGIS outside of Flatpak, and OpenDroneMap outside of Docker
2023-08-02 02:27:59 <Saijin_Naib[m]> Also, someone beat me to packaging xfce4-systemload-plugin. I was literally just working on that
2023-08-02 02:28:00 <Saijin_Naib[m]> Sweet!
2023-08-02 02:28:23 <Saijin_Naib[m]> Does it make sense to have a "Group" for each DE like the GNOME group?
2023-08-02 02:28:39 <Saijin_Naib[m]> Maybe an XFCE team to help share packaging responsibiltiies between?
2023-08-02 04:35:13 <Forza> Saijin_Naib[m]: not sure, but I'd believe so
2023-08-02 04:37:18 <Saijin_Naib[m]> Hmm... I wonder how we go about that?
2023-08-02 04:37:33 <Forza> It claims to work on any shell, including cmd and powershell on windows :) 
2023-08-02 04:37:51 <Saijin_Naib[m]> if I exec starship xfterm just goes poof, haha
2023-08-02 04:39:11 <Forza> Oh
2023-08-02 04:42:25 <Saijin_Naib[m]> Yeah, but I'm just using busybox ash, not bash/fish/zsh/et al
2023-08-02 04:42:44 <Saijin_Naib[m]> Don't know enough or do enough to feel the need for another shell
2023-08-02 04:46:17 <Forza> That's OK. :) 
2023-08-02 04:47:54 <Saijin_Naib[m]> I still use busybox vi, haha
2023-08-02 04:48:30 <Saijin_Naib[m]> It feels comfy? doesn't do too much, mostly works how I expect... I'd like syntax highlighting, line numbers, and better pasting but
2023-08-02 04:48:45 <Saijin_Naib[m]> neovim and emacs were too much, vim seems okay, but also I got stuck in it
2023-08-02 04:48:54 <Saijin_Naib[m]> pyvim is neat, but basically just vim
2023-08-02 04:49:35 <Saijin_Naib[m]> nano is what I'm used to from Debian-derived days, but I dunno, wanted to go full-immersion with Alpine and stick to defaults as much as possible, so busybox vi
2023-08-02 04:52:44 <somercet1> Saijin_Naib[m], try the vis editor
2023-08-02 04:53:23 <Saijin_Naib[m]> apk search vis
2023-08-02 04:53:50 <Saijin_Naib[m]> fuck, i a, tired
2023-08-02 04:57:33 <Saijin_Naib[m]> Well, I was able to open help and exit help without getting stuck or needing to do :qa! so I'm happy thus far
2023-08-02 04:57:36 <Saijin_Naib[m]> Thank you for that
2023-08-02 05:07:34 <Saijin_Naib[m]> Configuration is a nightmare though, haha. Time to study
2023-08-02 05:10:23 <somercet1> Saijin_Naib[m], irc://libera.chat/vis-editor is on  if you need help
2023-08-02 05:10:44 <somercet1> grammar slightly messed up
2023-08-02 05:12:11 <Forza> Saijin_Naib[m]: https://github.com/starship/starship/issues/928 seems ash is the one exclusion to starship at the moment. :/
2023-08-02 05:14:09 <Forza> I could never get the hang on vi and dedicates! So I am sticking with nano, which works out great for the most part. Except isn't always installed by default on many distros. 
2023-08-02 05:20:25 <Forza> Seems a workaround for Starship on ash is to do: PS1='$(starship prompt --status "$?" --path "$PWD")'
2023-08-02 05:22:02 <zcrayfish> man, the folks in that thread are getting ash, dash, and busybox's version of dash all confused. xD 
2023-08-02 05:22:31 <zcrayfish> at least busybox's version has the best chance of eventually getting supported 
2023-08-02 05:24:13 <Saijin_Naib[m]> It sounded not rosy though. Is BusyBox ASH really that limited? I don't even know what I'm missing, haha
2023-08-02 05:24:37 <ikke> It's posix sh with some bashsisms enabled
2023-08-02 05:25:06 <ikke> dash is more strict
2023-08-02 05:25:49 <zcrayfish> You'd have to be hard core to use that as an interactive shell heh. 
2023-08-02 05:25:57 <Saijin_Naib[m]> dash?
2023-08-02 05:26:03 <zcrayfish> yeah
2023-08-02 05:28:32 <elly> I use busybox ash as my interactive shell
2023-08-02 05:28:34 <elly> is that weird?
2023-08-02 05:28:40 <ikke> No
2023-08-02 05:28:50 <Saijin_Naib[m]> I'm thinking it is, haha
2023-08-02 05:28:50 <ikke> The only thing I really miss is alt+.
2023-08-02 05:28:50 <Saijin_Naib[m]> Me too
2023-08-02 05:28:58 <Saijin_Naib[m]> I had a decent prompt and lost it ☹️
2023-08-02 05:28:59 <Saijin_Naib[m]> What's that do?
2023-08-02 05:29:27 <ikke> recall last argument
2023-08-02 05:29:39 <ikke> mkdir foo; cd <alt+.>
2023-08-02 05:30:02 <dwfreed> can also use !$ in bash if you have history expansion enabled (is by default)
2023-08-02 05:30:15 <Forza> Didn't know about alt+
2023-08-02 05:30:18 <dwfreed> mkdir foo; cd !$
2023-08-02 05:30:21 <elly> my prompt is "\w \$" which is maybe *too* simple
2023-08-02 05:30:29 <dwfreed> me either, but I've used !$ a lot
2023-08-02 05:30:44 <dwfreed> alt+. is probably easier to type than !$
2023-08-02 05:30:56 <Forza> I do use history|grep a lot
2023-08-02 05:31:26 <Forza> And i have eternal bash history enabled ;o
2023-08-02 05:32:21 <Forza> So its an extension to my forgetful brain to remember what commands i use
2023-08-02 05:32:58 <elly> ctrl+r is the cooler history|grep, isn't it?
2023-08-02 05:33:10 <Forza> https://wiki.tnonline.net/w/Blog/Bash_Eternal_History
2023-08-02 05:33:17 <Forza> Oh
2023-08-02 05:33:38 <Forza> Yes.!
2023-08-02 05:34:14 <Forza> Why didn't i know that
2023-08-02 05:34:38 <Forza> 
ACTION  🤦‍♂️ 

2023-08-02 05:35:37 <Saijin_Naib[m]> I have to recover my custom prompt as a starting point
2023-08-02 05:35:38 <Forza> Only thing I can think of that lack with ash is bash-completions 
2023-08-02 05:35:50 <Saijin_Naib[m]> That's impossible to have, right?
2023-08-02 05:35:57 <Saijin_Naib[m]> That would be really welcome
2023-08-02 05:36:02 <Forza> Yes
2023-08-02 05:36:13 <Forza> It's not supported in ash AFAIK 
2023-08-02 05:36:20 <Saijin_Naib[m]> how do I save out a file in vis 😭
2023-08-02 05:36:29 <Saijin_Naib[m]> colon x no workie
2023-08-02 05:36:31 <zcrayfish> 
ACTION thinks everyone should use PS1="\e[1;31ma\e[1;33ml\e[1;32mp\e[1;34mi\e[1;36mn\e[1;35me\e[0m\033[30;37;5m😊\e[0m\w " as their default prompt ;)

2023-08-02 05:38:26 <Forza> alpine:)~
2023-08-02 05:38:34 <Forza> Cute
2023-08-02 05:38:43 <pj> Anyone should ditch any shell that doesn't use prompt function for configuring prompt
2023-08-02 05:38:52 <zcrayfish> I see you have tasted the rainbow as well.
2023-08-02 05:38:56 <Saijin_Naib[m]> I don't even know what that means
2023-08-02 05:39:05 <Forza> Ought to be default on alpine 
2023-08-02 05:40:03 <Forza> Saijin_Naib[m]: https://paste.tnonline.net/files/VP5uCCzJnx4j_Screenshot_20230802_073929_Termius.png
2023-08-02 05:40:15 <somercet1> Saijin_Naib[m], :w
2023-08-02 05:40:23 <Forza> zcrayfish: looks nice
2023-08-02 05:41:28 <Forza> zcrayfish: submit it as proposal 
2023-08-02 05:41:38 <Saijin_Naib[m]> somercet1: haha, oh my god... Why, why change it?!
2023-08-02 05:41:54 <somercet1> it isn't changed
2023-08-02 05:42:12 <somercet1> :w [filename] has been the way to write a file in vi since it was created
2023-08-02 05:43:09 <dwfreed> :x is write and exit
2023-08-02 05:43:11 <somercet1> also, vis uses structured expressions
2023-08-02 05:43:19 <somercet1> :x replaces :s
2023-08-02 05:43:32 <Forza> 
ACTION is off to get morning coffee ☕ 

2023-08-02 05:43:50 <Saijin_Naib[m]> ah... so I'm used to writing and closing, which yeah, is what I want to do
2023-08-02 05:44:01 <Saijin_Naib[m]> I'll need to see how vis handles that and if I like it
2023-08-02 05:44:23 <somercet1> ZZ is the vi sequence to save and exit in vis
2023-08-02 05:46:26 <Saijin_Naib[m]> Thank you
2023-08-02 05:46:40 <Saijin_Naib[m]> It looks like I have to study vis a lot before I can make use of it readily
2023-08-02 05:47:01 <Saijin_Naib[m]> I was hoping to keep my busybox vi muscle memory and gain a bit of extra nice
2023-08-02 05:47:09 <Forza> What's the advantage of vi's over vi/vim? 
2023-08-02 05:47:15 <somercet1> :x is the only really different command from standard vi
2023-08-02 05:47:35 <somercet1> vis has structural regexes and multiple cursors
2023-08-02 05:48:07 <somercet1> open a file in vis, type ":x /^/" and see what happens
2023-08-02 05:48:11 <Forza> Oh, sounds complicated 
2023-08-02 05:48:22 <somercet1> it's not, really.
2023-08-02 05:49:20 <somercet1> it's small, light and fast, and looks good, unlike some of the older vi clones.
2023-08-02 05:51:02 <Saijin_Naib[m]> somercet1: That's the only one I used aside from quit, haha. I'm not a power user. I have never learned searching or replacement or macros or whatever 
2023-08-02 05:51:42 <Saijin_Naib[m]> So that really messed me up, but thank you for saving me
2023-08-02 05:52:03 <somercet1> np
2023-08-02 06:23:07 <zcrayfish> print test
2023-08-02 06:23:19 <zcrayfish> print -buffer test
2023-08-02 06:25:56 <zcrayfish> test failed successfully. sorry about the noise :x
2023-08-02 06:28:08 <Saijin_Naib[m]> haha, all good
2023-08-02 06:28:16 <Saijin_Naib[m]> I think one last aport and then nightnight
2023-08-02 06:28:27 <zcrayfish> sleep is good
2023-08-02 06:29:19 <zcrayfish> I'm trying to get my automatic system update+reboot script to print an "omg, this system will reboot" message to every buffer in weechat.... not send it to every channel. 
2023-08-02 06:29:32 <zcrayfish> think I'll work on that tomorrow though. it's not urgent 
2023-08-02 06:35:04 <Saijin_Naib[m]> hmm...
2023-08-02 06:35:07 <Saijin_Naib[m]> I need fstab.h
2023-08-02 06:35:15 <Saijin_Naib[m]> Only two packages provide...
2023-08-02 06:35:19 <Saijin_Naib[m]> https://pkgs.alpinelinux.org/contents?file=fstab.h&path=&name=&branch=edge&arch=x86_64
2023-08-02 06:38:26 <Saijin_Naib[m]> 
ACTION sent a code block: https://matrix.org/_matrix/media/v3/download/matrix.org/PnPsmSpnhXETeWNSsVqkqbpu

2023-08-02 12:11:58 <Forza> Hi. I am trying to create a wiki account, but I stumble on the captcha.... From what side should I count digits?
2023-08-02 12:13:19 <Forza> oh ok left to right
2023-08-02 13:11:27 <consus[m]> Do we have a list of orphaned packages after psykose fallout? I moght grab the storages one like nvme-cli
2023-08-02 13:13:59 <ikke> Most have already been picked up already 
2023-08-02 13:14:28 <pjs> is it a bad idea to install packages from edge when my system is all 3.18?
2023-08-02 13:14:37 <ikke> https://pkgs.alpinelinux.org/packages?name=&branch=edge&repo=&arch=x86_64&maintainer=None
2023-08-02 13:14:41 <pjs> I really just want the py3-qt6 pkg
2023-08-02 13:14:57 <ikke> pjs: it's not supported and likely to break, if it works in the first place 
2023-08-02 13:15:24 <ikke> If we upgrade python in edge, it'll break 
2023-08-02 13:15:46 <elly> ftr I did that (3.18 system, edge packages) by accident and it did not work well
2023-08-02 13:19:06 <ikke> It may work just after a new release, but the longer after the last release, the more likely it's to break, unless it's something static
2023-08-02 13:22:17 <pjs> ikke: gotcha, thanks
2023-08-02 13:22:21 <pj> if musl is upgraded, it will break whole system :>
2023-08-02 13:22:42 <pjs> I'm a few days into using alpine as a desktop so still learning the ropes :)
2023-08-02 13:33:58 <somercet1> what happened with psykose?
2023-08-02 13:35:04 <ikke> somercet1: she just decided to quit 
2023-08-02 13:39:25 <somercet1> mmm
2023-08-02 15:20:09 <streaksu[m]> she did not seem willing to delve much into the topic from what I heard
2023-08-02 15:20:23 <streaksu[m]> I imagine it must be burnout, who can blame her with those numbers...
2023-08-02 15:22:02 <Saijin_Naib[m]> thanks for the help, ncopa 
2023-08-02 15:22:36 <Saijin_Naib[m]> If I'm needing help on an aport, is it better to make a draft and discuss what I need help with in the draft MR, or better to get the help and resolve issues here so the MR is ready to be reviewed when I open it?
2023-08-02 15:23:17 <ikke> Saijin_Naib[m]: opening it as draft is perfectly fine
2023-08-02 15:23:36 <ncopa> I tend to not look at drafts
2023-08-02 15:23:50 <Saijin_Naib[m]> Hmm... Okay. So here is still the best place to get help
2023-08-02 15:23:56 <Saijin_Naib[m]> Or the proper place?
2023-08-02 15:24:21 <ncopa> if you believe the MR is ready for merge (but unsure), then you can create a MR without draft
2023-08-02 15:24:34 <ikke> you can create a draft and then ask for help
2023-08-02 15:24:40 <ncopa> yeah
2023-08-02 15:25:04 <Saijin_Naib[m]> Alright, thank you
2023-08-02 15:25:13 <ncopa> i consider draft as "work in progress", and I usually wait for it to be ready for review (no longer draft)
2023-08-02 15:25:23 <Saijin_Naib[m]> That makes sense, yeah
2023-08-02 18:15:15 <ePirat> Hi, I was wondering if its possible to have my drive partitioned with one boot and one data partition, to allow for diskless + data setup so that I can use ext4 for the data partition
2023-08-02 18:20:24 <ikke> ePirat: You want to boot from disk, but still run a run-from-ram system?
2023-08-02 18:23:23 <ePirat> yes and be able to commit changes with lbu
2023-08-02 18:23:29 <pjs> anyone a sound/mic guru? I have a thinkpad x1 carbon 6th gen, when I installed first installed alpine things were working fine after some alsamixer changes, then after a reboot, I can't get the sound to work at all. I've gone through the Alsa and sound wiki entries, some of the arch stuff, just can't get it to work. lspci shows the audio device. I'm not familiar enough with sound setup to think of 
2023-08-02 18:23:35 <pjs> anything else to try (I ran ubuntu for years on this thing and never had to think about it but sound always worked)
2023-08-02 18:25:35 <ikke> ePirat: Might be possible, but it's not a supported (in the sense that the installer supports it) configuration
2023-08-02 18:35:05 <ePirat> ikke, I think I got it mostly working but apk cache always says: Warning: opening from cache http://…: No such file or directory
2023-08-02 18:35:31 <ePirat> not sure what that is about…
2023-08-02 18:50:09 <pjs> ok, looks like I just forgot to save the module to load at boot. Sound is working... 
2023-08-02 18:50:14 <pjs> smh
2023-08-02 18:50:41 <streaksu[m]> happens to the best of us
2023-08-02 18:54:17 <pjs> next (and I think final) issue I have is my logitech mx master mouse (original).. it pairs via bluetooth fine but it doesn't actually move the mouse. I assume I'm missing something because `xinput list` doesn't show it. Any ideas?
2023-08-02 18:54:34 <pjs> streaksu[m]: hah yea.
2023-08-02 19:47:40 <Skunk385[m]> How to exec /etc/crypttab on boot???
2023-08-02 20:46:57 <Saijin_Naib[m]> ncopa: What is needed, or is it even needed, to form a team for XFCE like there is for GNOME? I know it isn't nearly as large and complicated, but ideally, I'd like to move all the XFCE work I'm doing in Testing into Community, but I'll need a lot of help and guidance
2023-08-02 20:47:32 <ikke> Saijin_Naib[m]: not sure if it's needed for now
2023-08-02 20:47:50 <Saijin_Naib[m]> Alpine is already far and away my favorite distro, and XFCE has now become my favorite DE, so I'd just like to do my best to make sure XFCE on Alpine is awesome
2023-08-02 20:48:07 <Saijin_Naib[m]> ikke: Yeah, makes sense. I know GNOME/KDE are absolutely massive and touch everything else
2023-08-02 20:48:17 <Saijin_Naib[m]> Alright, I guess I'll see what I can do for now
2023-08-02 20:48:33 <ikke> You don't need to form a team to get help from the community :)
2023-08-02 20:49:03 <ikke> But if there are more people who want to join you, we could look at it
2023-08-02 20:50:07 <Saijin_Naib[m]> Haha, more so that I don't feel like I'm interrupting when I need help, which is often
2023-08-02 20:50:23 <Saijin_Naib[m]> Like, it'd be a group of people already amenable to that since they're working on the same, you know?
2023-08-02 20:50:35 <Saijin_Naib[m]> Anxiety thing, I guess. Probably not actually needed haha
2023-08-02 20:50:44 <ikke> You're never interrupting
2023-08-02 20:52:06 <Saijin_Naib[m]> 😬 thank you
2023-08-02 21:02:33 <Saijin_Naib[m]> Anyone a heavy user of Xournal++?
2023-08-02 21:03:10 <Saijin_Naib[m]> While I'm at enabling gtksourceview for better LaTex support in v1.2.0, I'm curious if anyone would find value in having the plugins function enabled? We build without Lua, so no plugins
2023-08-02 21:03:23 <Saijin_Naib[m]> And I'm trying to figure out what backtrace does for the build as a makedepends
2023-08-02 21:13:34 <latex> 
ACTION perks their ears

2023-08-02 21:15:35 <Saijin_Naib[m]> Haha, sorry for the accidental ping
2023-08-02 21:15:55 <latex> Maybe I should change my nick here too
2023-08-02 21:32:31 <Saijin_Naib[m]> Does BusyBox vi support the "ruler" or line numbers on the left like busybox less can with -MN?
2023-08-02 21:32:48 <Saijin_Naib[m]> I can't really see if it can or not... Missing build option perhaps?
2023-08-02 21:49:48 <Saijin_Naib[m]> https://gitlab.gnome.org/search?group_id=8&nav_source=navbar&project_id=1893&repository_ref=vte-0-72&search=sixel&search_code=true
2023-08-02 21:49:58 <Saijin_Naib[m]> I am not understanding how to enable sixel support, apparently
2023-08-02 21:50:30 <Saijin_Naib[m]> I thought adding -Dsixel=true to the abuild-meson part of the APKBUILD would do it, but it is failing to find the option at build time
2023-08-02 22:46:10 <LACampbell> is anybody able to open directories in flatpak apps? if so you an please tell me whcih xdg-portal you are using?
2023-08-02 23:26:04 <Saijin_Naib[m]> Gtk por5al under XFCE 
2023-08-02 23:26:17 <Saijin_Naib[m]> How do you mean open? 
2023-08-02 23:26:32 <Saijin_Naib[m]> What isn't working 
2023-08-03 00:36:13 <livy> Hello. Does anyone know why the package 'linux-lts' is not available during a live CD session?
2023-08-03 00:36:50 <livy> After I installed Alpine to disk (sys mode) and boot from disk, that package is available.
2023-08-03 00:37:22 <livy> 'apk info -v' | grep linux-lts -> linux-lts-6.1.42-r0
2023-08-03 01:20:07 <livy> In other words, how can I update the kernel version using the APK cache? So that the live CD session loads the 'linux-lts' package from cache during boot?
2023-08-03 01:24:12 <livy> I guess it does not make sense because kernel in the live CD cannot load the new version and replace itself.
2023-08-03 02:02:34 <elly> it could, using kexec
2023-08-03 03:41:16 <dewdude> https://img.pickmy.org/zqbt2rtz.png <- the drivers for my onboard gigE apparently suck..or I really need to work on the config
2023-08-03 03:41:21 <dewdude> 10. is a USB3 NIC
2023-08-03 07:09:42 <ncopa> Saijin_Naib[m]: I also use xfce currently
2023-08-03 11:39:22 <Saijin_Naib[m]> ncopa Ahh, makes sense that you maintain mostly all of XFCE then, haha
2023-08-03 12:32:44 <openlinux21[m]> test
2023-08-03 12:44:20 <Saijin_Naib[m]> openlinux21 visible 
2023-08-03 12:48:47 <ikke> they are already gone
2023-08-03 13:31:01 <wyk72> hello Alpine wizards :)
2023-08-03 13:31:09 <wyk72> I have a question:
2023-08-03 13:31:28 <wyk72> I have a small router box I run in run-from-ram mode
2023-08-03 13:31:54 <wyk72> I mounted .modloop in rw by adding the overlay module available in kernel line
2023-08-03 13:32:27 <wyk72> my question is: how do I make the changes I can make to dir /lib/modules persistent via lbu ?
2023-08-03 13:32:36 <wyk72> which dir should I include ?
2023-08-03 13:33:22 <minimal> aren't the contents of /lib/modules/ coming from the kernel package?
2023-08-03 13:33:33 <wyk72> it actually mounts as overlay on /lib/modules type overlay (rw,relatime,lowerdir=/lib/modules:/.modloop/modules,upperdir=/.modoverlayfs/modules,workdir=/.modoverlayfs/work)
2023-08-03 13:34:47 <wyk72> yep, I've compiled a module for an usb card abd I'd like to include it in /lib/modules (and a firmware too)
2023-08-03 13:35:02 <wyk72> I can do it by using a local.d script
2023-08-03 13:35:04 <minimal> wyk72: what I mean is that linux-lts (and any 3rd party module packages) provide the contents of /lib/modules/ via the modloop so why would you want to make changes?
2023-08-03 13:35:56 <minimal> ah, compiled via a local package's APKBUILD?
2023-08-03 13:36:35 <wyk72> yeah "manually" compiled (relatek 2.5 gig card + wifi)
2023-08-03 13:36:40 <ikke> wyk72: the modloop service should already allow you to setup an overlay
2023-08-03 13:37:06 <wyk72> yeah it does it automatically, by adding the line "overlay" into the kernel modules line
2023-08-03 13:37:14 <wyk72> as per the init script
2023-08-03 13:37:25 <wyk72> but I have no info on how to make changes persistent
2023-08-03 13:37:54 <wyk72> if there's an "official" way to do it I mean
2023-08-03 13:38:11 <ikke> wyk72: no, the services runs after the kernel is booted
2023-08-03 13:38:20 <ikke> so it cannot affect the kernel line itself
2023-08-03 13:39:00 <wyk72> I mean the /etc/init.d/modloop script
2023-08-03 13:39:12 <wyk72> does it by itself if I add the overlay module to boot line
2023-08-03 13:41:20 <wyk72> I can just add some line to the local.d script, copy the modules/firmware to /lib/modules then do a "depmod -A", but seemes a bit clunky to me
2023-08-03 13:46:38 <wyk72> I mean what is the point of an overlay rw /lib/modules if it cant be made persistent ?
2023-08-03 13:47:07 <ikke> wyk72: the idea is that packages can put modules there (in ram), which normally is not possible
2023-08-03 13:47:19 <wyk72> I see
2023-08-03 13:49:52 <wyk72> a thing like akms can work in run-from-ram mode then ? (just asking)
2023-08-03 13:54:46 <ikke> It would probablly need to compile the module each boot
2023-08-03 13:55:28 <ikke> building the module and packaging it would work better
2023-08-03 17:05:48 <tspear> Hello, I have been trying to understand the process undertaken to build the alpine Docker image, and running into some troubles. I found https://github.com/alpinelinux/docker-alpine/blob/master/Dockerfile seems to download a tarball such as https://cz.alpinelinux.org/alpine/v3.18/releases/x86_64/alpine-minirootfs-3.18.2-x86_64.tar.gz but I want to go a step further to understand how this tarball is built.
2023-08-03 17:06:33 <tspear> I found the alpine-make-rootfs project, and it appears to be able to build an alpine rootfs image, but the resulting image is not identical to the one published on Dockerhub.
2023-08-03 17:07:06 <ikke> Here are the scripts that creates that: https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/scripts
2023-08-03 17:07:17 <ikke> mkimg.minirootfs.sh
2023-08-03 17:07:46 <ikke> Note that docker-alpine is not a generic alpine image. it's rather an image used to build alpine packages with
2023-08-03 17:08:11 <tspear> Wow, thank you! That was a lot easier than I anticipated! Much appreciated :)
2023-08-03 17:08:31 <tspear> Oh..
2023-08-03 17:08:53 <tspear> How would one build the generic image that is published to dockerhub?
2023-08-03 17:09:05 <ikke> sorry, I've confused it with a different project
2023-08-03 17:09:10 <ikke> that actually _is_ the correct project
2023-08-03 17:09:18 <ikke> https://github.com/docker-library/official-images/blob/master/library/alpine#L2
2023-08-03 17:09:26 <tspear> Great, thanks again :)
2023-08-03 17:46:34 <frag> is there something to chat on signal with in repo?
2023-08-03 17:46:50 <ikke> signal-desktop
2023-08-03 17:51:04 <frag> oh nice ikke 
2023-08-03 17:51:20 <frag> thought i had testing in repo too ..
2023-08-03 17:52:04 <frag> shouldnt this work? had it without the tag first .. @testing http://mirrors.nju.edu.cn/alpine/edge/testing
2023-08-03 17:52:51 <ikke> frag: If you are not on edge, you should not install things from testing, which either won't install due to missing dependencies, or will break
2023-08-03 17:55:18 <frag> i am on edge actually :]
2023-08-03 17:56:27 <ikke> ok, then either should work
2023-08-03 17:59:01 <frag> just doas apk update; doas apk add signal-desktop@testing ?
2023-08-03 18:31:08 <ikke> frag: should work, yes
2023-08-03 18:33:07 <frag> should but shournt
2023-08-03 18:33:09 <frag> oh well ..
2023-08-03 18:33:22 <ikke> frag: what error do you get?
2023-08-03 18:33:36 <pj> if you're on edge, why do you need tag for testing repo
2023-08-03 18:34:22 <ikke> if for some reason you don't want to accidentally install things from testing
2023-08-03 18:35:20 <pj> What's the accidental install
2023-08-03 18:35:32 <pj> either you type the package name that's in testing or not
2023-08-03 18:45:10 <frag> ikke ERROR: unable to select packages:  signal-desktop (no such package):    required by: world[signal-desktop]
2023-08-03 18:45:35 <frag> (doesnt show up in search results either (ofc))
2023-08-03 18:46:08 <ikke> frag: works for me
2023-08-03 18:46:42 <ikke> frag: oh, what arch are you on?
2023-08-03 18:46:45 <ikke> https://pkgs.alpinelinux.org/packages?name=signal-desktop&branch=edge&repo=&arch=&maintainer=
2023-08-03 18:47:20 <frag> ikke: aarch64
2023-08-03 18:47:43 <ikke> right, it's only available for x86_64
2023-08-03 18:47:52 <frag> theres the problem :]
2023-08-03 18:48:21 <frag> do packages usually get available for all archs?
2023-08-03 18:48:29 <ikke> normally, yes
2023-08-03 18:48:39 <ikke> so if they are not, it usually means upstream does not support it for some reason
2023-08-03 18:48:59 <frag> i see :]
2023-08-03 18:50:05 <ikke> "# aarch64: polyval-0.5.3 crate subdep uses unstable stdsimd feature"
2023-08-03 18:56:47 <ascensionwired> this is probably the way to avoid `setup-devd udev` and use sway along with mdev, different from the instructions to switch to eudev as here https://wiki.alpinelinux.org/wiki/Sway , all need to do is install libudev-zero before installing sway `apk add libudev-zero` and won't have to switch to eudev, 
2023-08-03 18:57:10 <ascensionwired> maybe there are some better ways to get sway and mdev working on alpine, maybe someone knows?
2023-08-03 20:33:07 <WhyNotHugo> I'm really surprised about psykose leaving. They were both very knowledgeable and always willing to help (and I've learned so much thanks to them). It won't feel the same without them.
2023-08-03 20:33:57 <ikke> both?
2023-08-03 20:34:03 <ikke> oh
2023-08-03 20:35:45 <ikke> I agree, she did a ton of work
2023-08-03 20:37:59 <WhyNotHugo> Currently looking at the list of maintained packages and it's an amazing amount of work.
2023-08-03 20:38:10 <WhyNotHugo> (and lots of them require unique bits of obscure knowledge)
2023-08-03 21:31:05 <elly> WhyNotHugo: maintaining that many packages is a TON of work though for one person
2023-08-04 02:02:21 <openlinux21[m]> Hi everyone, I'm OpenLinux21, a regular Alpine Linux user!
2023-08-04 02:02:33 <streaksu[m]> o/
2023-08-04 02:03:26 <openlinux21[m]> But I'm using Alpine Linux on my Android phone
2023-08-04 02:03:32 <openlinux21[m]> 😁
2023-08-04 02:04:58 <openlinux21[m]> 
ACTION uploaded an image: (550KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/LCEJphxzxcROHrJPJtGahxTX/Screenshot_20230716_201826_com.limbo.emu.main.jpg >

2023-08-04 03:02:43 <zcrayfish> openlinux21[m]: That's hot. 
2023-08-04 03:09:41 <xgpt> openlinux21[m]: what app?
2023-08-04 03:20:30 <openlinux21[m]> <xgpt> "openlinux21: what app?" <- Limbo x86 PC Emulator 
2023-08-04 03:20:30 <openlinux21[m]> https://f-droid.org/packages/com.limbo.emu.main/
2023-08-04 03:25:58 <rob[m]> how's the performance? Given that you fully emulate an i686 computer? 
2023-08-04 03:26:47 <rob[m]> I used to emulate ppc64 (IBM Power) to run AIX on x86_64 and that was super duper slow
2023-08-04 03:27:55 <openlinux21[m]> Limbo can emulate x86_64
2023-08-04 03:28:03 <rob[m]> now I'm running AIX on Alpine for ppc64le via kvm and it's super fast
2023-08-04 03:28:29 <openlinux21[m]> Here's a fun one for you.
2023-08-04 03:28:46 <openlinux21[m]> try it
2023-08-04 03:29:11 <openlinux21[m]> 
ACTION posted a file: (10240KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/khFsZTDAPjdRSCtKIrQLbBbY/minilinux.iso >

2023-08-04 03:29:22 <rob[m]> Yes, but I imagine it must be quite slow? I remember AIX via ppc64 emulation and it was usable, but it was quite slow. And that was on a gen 11th i7
2023-08-04 03:29:33 <i5yxt1550rd9gbko8o55[m]> too slow
2023-08-04 03:29:43 <i5yxt1550rd9gbko8o55[m]> please using qemu-system-aarch64
2023-08-04 03:30:33 <rob[m]> An alpine phone is nice, you can try postmarketos on your phone (it's alpine linux for phones). That's what I use on my pinephone. It works great!
2023-08-04 03:30:55 <openlinux21[m]> I have KVM virtualization turned on and booting Alpine Linux x86_64 takes 20 seconds!
2023-08-04 03:31:09 <openlinux21[m]> i5yxt1550rd9gbko8o55[m]: chinese?
2023-08-04 03:31:39 <rob[m]> But, kvm cannot accelerate a different arch? It must be aarch64 only iirc?
2023-08-04 03:32:26 <openlinux21[m]> I don't know, but it does start up a lot faster when I turn this feature on
2023-08-04 03:33:07 <rob[m]> weird, I'm more into the BSD's, but KVM shouldn't be able to accelerate a different arch if I'm not mistaken
2023-08-04 03:33:29 <rob[m]> if you try an aarch64 image, that would be a lot faster 
2023-08-04 03:33:33 <heat> yes KVM only works for the host architecture
2023-08-04 03:34:07 <rob[m]> great, I was quite sure about that, but given that I'm not too much into KVM and Linux anymore, I wanted to make sure 
2023-08-04 03:34:30 <openlinux21[m]> By the way, I tried to install icewm, but the system says it can't find the package, why?
2023-08-04 03:34:53 <rob[m]> openlinux21: you will see a great difference with an image for arm64
2023-08-04 03:35:10 <openlinux21[m]> ok
2023-08-04 03:35:34 <rob[m]> Maybe the community's repo isn't enabled? I don't know about icewm
2023-08-04 03:35:42 <openlinux21[m]> 
ACTION uploaded an image: (341KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/JzEFBpoqooiHJIUnnbzjHqIh/Screenshot_20230804_113520_com.limbo.emu.main.jpg >

2023-08-04 03:36:01 <openlinux21[m]> How should I configure it?
2023-08-04 03:36:37 <openlinux21[m]> CPU Model:
2023-08-04 03:36:39 <rob[m]> correction, it's in the testing repository
2023-08-04 03:37:16 <openlinux21[m]> How do I enable the test library?
2023-08-04 03:37:24 <rob[m]> I don't know Limbo, but you have to download an arm64 image and then setup an arm64 vm
2023-08-04 03:38:06 <i5yxt1550rd9gbko8o55[m]> <openlinux21[m]> "chinese?" <- yes
2023-08-04 03:38:22 <openlinux21[m]> 行
2023-08-04 03:38:25 <rob[m]> here's your package, in testing https://pkgs.alpinelinux.org/package/edge/testing/x86/icewm
2023-08-04 03:38:47 <rob[m]> here's how you can enable the testing repository: https://wiki.alpinelinux.org/wiki/Enable_Community_Repository
2023-08-04 03:40:05 <rob[m]> after setting up an arm64 vm, enabling kvm and using an arm64 alpine image, you should have a pretty fast vm. Unsure if X is supported as you're trying to use a window manager 
2023-08-04 03:40:31 <rob[m]> I guess you could setup tigervnc and use your window manager through that
2023-08-04 03:41:10 <rob[m]> that's what I do on my ppc64le server, I use dwm through tigervnc so I can stay away from the noise
2023-08-04 03:41:27 <openlinux21[m]> I'm trying to install Alpine Linux (armv7) on my development board, but unfortunately nothing happens after powering it up!
2023-08-04 03:42:16 <rob[m]> https://wiki.archlinux.org/title/TigerVNC
2023-08-04 03:42:53 <rob[m]> Hmm, we would need a bit more info to help debug your issue. I will have to go though. Have a good night/day :)
2023-08-04 03:43:15 <openlinux21[m]> ok
2023-08-04 09:06:13 <openlinux21[m]> What's the smallest Linux distro you guys have met so far (post-2005), and what's
2023-08-04 09:09:02 <dwfreed> I mean, that depends on what you consider usable
2023-08-04 09:09:11 <dwfreed> old version of OpenWrt fit in 4 MB or less
2023-08-04 09:10:02 <dwfreed> DSL fits in 50 MB
2023-08-04 09:11:37 <Ermine> TCL is roughly the same size iirc
2023-08-04 09:12:44 <dwfreed> TCL is even smaller
2023-08-04 09:13:47 <dwfreed> Tiny Core is 16 MB
2023-08-04 09:15:20 <Ermine> With desktop it's 23MB
2023-08-04 09:16:17 <dwfreed> anyway, the disk measuring contest is pretty pointless; find something that meets your needs and fits in the space you have
2023-08-04 09:16:42 <dwfreed> chances are you have more than 50 MB available
2023-08-04 09:17:06 <Ermine> I had an idea to try to squeeze stuff into a floppy disk, but apparently this would require ancient software
2023-08-04 09:17:22 <dwfreed> yes
2023-08-04 09:17:42 <dwfreed> you aren't fitting much in the 1.38 MB formatted capacity of a floppy
2023-08-04 09:17:53 <dwfreed> maybe a really minimal kernel and busybox, and that's it
2023-08-04 10:29:58 <ascensionwired> i was asking a question here about sway and mdev on alpine
2023-08-04 10:30:00 <ascensionwired> is it possible for sway to work with mdev?
2023-08-04 10:30:41 <ascensionwired> only found a way to install libudev-zero before installing sway
2023-08-04 10:31:07 <ascensionwired> maybe there are other, better ways? 
2023-08-04 10:32:18 <Ermine> sway and many other things rely on libudev
2023-08-04 10:32:51 <Ermine> If you don't use eudev, your only option is libudev-zero with all its problems
2023-08-04 10:40:14 <ascensionwired> okay, thanks
2023-08-04 11:23:54 <ascensionwired> why if i do `doas echo "<something>" >> /etc/modules i get message `cant create /etc/modules: Permisdion denied`
2023-08-04 11:26:29 <c7s> because only echo is running as root, not the >>
2023-08-04 11:26:36 <belanthor> you need to use tee for that
2023-08-04 11:26:59 <belanthor> echo "some text" >> sudo tee yourfile
2023-08-04 11:27:08 <belanthor> or doas
2023-08-04 11:27:40 <belanthor> it had some flag i think...
2023-08-04 11:28:03 <mercenary> for >> use tee -a, or you will overwrite the file
2023-08-04 11:28:15 <belanthor> right
2023-08-04 11:30:09 <satmd> the doas above will be technically run as doas -> echo as a privileged process, then both returning the output to your regular shell which then tries to redirect using >>
2023-08-04 11:30:20 <satmd> doas sh -c "echo ... >> ..." would have worked
2023-08-04 11:34:52 <ascensionwired> understand now 
2023-08-04 11:45:48 <komrade-g[m]> Alpine is the shit!
2023-08-04 11:46:25 <komrade-g[m]> Alpine is based
2023-08-04 11:47:13 <Saijin_Naib[m]> Hell yeah it is
2023-08-04 11:47:31 <Saijin_Naib[m]> What needs to happen to get an OpenRC service for a package? These need to be manually written, correct?
2023-08-04 11:47:43 <Saijin_Naib[m]> There is no automated conversion from SystemD service files, right?
2023-08-04 11:47:46 <ikke> In many cases, yes
2023-08-04 11:47:56 <Saijin_Naib[m]> https://git.sesse.net/?p=plocate;a=blob;f=plocate-updatedb.timer;h=236f45fd79330984785c3fc35c05c7426971c665;hb=HEAD
2023-08-04 11:48:00 <ikke> (to your first question)
2023-08-04 11:48:13 <Saijin_Naib[m]> It looks like plocate should have a daily timer to update its search database
2023-08-04 11:48:26 <ikke> That would be a cronjob 
2023-08-04 11:48:45 <ikke> There are no timer services
2023-08-04 11:48:51 <Saijin_Naib[m]> Ahhh okay
2023-08-04 11:49:00 <Saijin_Naib[m]> So that falls on the end-user to know and setup manually?
2023-08-04 11:49:32 <Saijin_Naib[m]> There's no concept of $pkgname-cron, right?
2023-08-04 11:50:10 <ascensionwired> only now i noticed i use UTC as clock in hwclock
2023-08-04 11:50:19 <ascensionwired> but i dualbooting
2023-08-04 11:50:23 <ascensionwired> with win
2023-08-04 11:50:33 <ikke> Saijin_Naib[m]: you can drop a file in /etc/periodic/daily 
2023-08-04 11:50:57 <Saijin_Naib[m]> ikke: As part of the package build process?
2023-08-04 11:50:58 <ascensionwired> i think i need do clock="local"
2023-08-04 11:51:12 <Saijin_Naib[m]> This impacts Catfish at the least, not sure what else might lean on plocate
2023-08-04 11:52:16 <ikke> Would there need to be a cronjob for each package that uses plocate?
2023-08-04 11:55:52 <Saijin_Naib[m]> Hmm... No, from what I understand, plocate makes a "global" (to the user account running it) database that anything that calls plocate can get results from
2023-08-04 11:56:08 <Saijin_Naib[m]> It is supposed to work compatibly with mlocate, just orders of magnitude faster
2023-08-04 11:56:18 <Saijin_Naib[m]> So only get results your user has access rights to see
2023-08-04 11:57:05 <Saijin_Naib[m]> Looks like it should be async and using io_uring too, so the cron job really shouldn't be much of a blip once it is all caught up after the first "big" one after install (post-install step?)
2023-08-04 12:00:15 <ikke> It would just be part of the package 
2023-08-04 12:00:51 <Saijin_Naib[m]> Oh, we can do something like that?
2023-08-04 12:01:50 <pj> yes
2023-08-04 12:02:40 <Saijin_Naib[m]> Is there a package that has an example of this I may take a look at?
2023-08-04 12:15:36 <ikke> Logrotate I think
2023-08-04 12:16:06 <ikke> In the package function, you can install any file in $pkgdir
2023-08-04 12:16:51 <Forza> Hi. I noticed that plocate doesn't install a locate symlink. I gues this is intentional?
2023-08-04 12:37:01 <Saijin_Naib[m]> Hmm 🤔 not as far as I can tell
2023-08-04 12:37:17 <Saijin_Naib[m]> The documentation on plocate tells you to use locate to find something as an example 
2023-08-04 12:37:27 <Saijin_Naib[m]> Maybe our package is missing that too? 
2023-08-04 12:40:41 <Saijin_Naib[m]> <ikke> "In the package function, you can..." <- So how do I make the timer? Do I try and match the upstream systemd timer unit function exactly?
2023-08-04 12:41:27 <Saijin_Naib[m]> With the 12hr jitter and 20m window of when it will kick off? 
2023-08-04 12:41:34 <Saijin_Naib[m]> Can our timers do that? 
2023-08-04 12:43:57 <Saijin_Naib[m]> Oh god, these are bash
2023-08-04 12:43:59 <Saijin_Naib[m]> Busybox ASH, rather
2023-08-04 12:43:59 <Saijin_Naib[m]> right?
2023-08-04 12:44:01 <Saijin_Naib[m]> 😭
2023-08-04 12:45:16 <ikke> Yes, they are shell scripts 
2023-08-04 12:45:53 <ikke> You drop a shell script there and crond will make sure they get executed once a day
2023-08-04 13:12:52 <Forza> Saijin_Naib[m]: use to systemd? Normal traditional cron is basically simply shell scripts
2023-08-04 13:13:46 <Saijin_Naib[m]> I guess? Prior to Alpine most everything I used had systemd
2023-08-04 13:14:01 <Saijin_Naib[m]> More I never modified services or added jobs 
2023-08-04 13:14:22 <Saijin_Naib[m]> I've no experience with writing shell scripts 
2023-08-04 13:16:52 <Forza> :) Always a good time to learn
2023-08-04 13:17:16 <Forza> Gentoo also use OpenRC, and some others
2023-08-04 13:19:02 <Forza> In case of systemd timers, you still need to specify what to execute.
2023-08-04 13:19:53 <Saijin_Naib[m]> Forza: Yeah, on top of the Thousand other things I need to be doing 🤣
2023-08-04 13:21:06 <Forza> :)
2023-08-04 13:28:59 <Saijin_Naib[m]> It's a very unfriendly language compare to. Python, which I already find unfriendly 😭
2023-08-04 14:03:57 <fred> identify fredo
2023-08-04 14:04:19 <fred> IDENTIFY fredo
2023-08-04 14:04:25 <fred> IFRNTIFY
2023-08-04 14:04:31 <fred> IDENTIFY
2023-08-04 14:07:11 <ncopa> hi fred: you need a / in front of IDENTIFY
2023-08-04 14:09:37 <nescafe> And chang eyour password
2023-08-04 14:11:03 <fred> thanks for help
2023-08-04 14:11:49 <fred> help sendpass
2023-08-04 14:13:34 <ikke> fred: you need to talk to nickserv 
2023-08-04 14:13:40 <ikke> In a private message 
2023-08-04 20:53:31 <kmodule> hey hey everyone
2023-08-04 20:54:05 <kmodule> just a quick question, what is the command to install linux-headers and linux-lts-dev for a specific kernel version?
2023-08-04 20:54:35 <kmodule> want to build a kernel module in docker for specific kernel version, and I didn't find anything in the docs
2023-08-04 20:54:51 <ikke> kmodule: we don't have these packages for arbitrary kernel versions, only a single version
2023-08-04 20:55:12 <kmodule> ohh I see
2023-08-04 20:55:31 <ikke> (that counts for every package, we only have a single version per branch)
2023-08-04 20:55:47 <kmodule> but if I install 3.18.2 in diskless mode, that will install kernel 6.1.34
2023-08-04 20:56:07 <kmodule> and if I try to get the linux-lts-dev package in a docker container
2023-08-04 20:56:18 <kmodule> that will install it for kernel 6.1.42
2023-08-04 20:56:24 <ikke> yeah, that will be a challenge
2023-08-04 20:56:26 <kmodule> both container and my diskless install are 3.18.2
2023-08-04 20:56:47 <kmodule> any idea how I could get these in sync?
2023-08-04 20:57:23 <minimal> kmodule: they're 3.18 really as there are not separate package repos for 3.18.0, 3.18.1, 3.18.2 etc, only for 3.18
2023-08-04 20:57:48 <kmodule> yeah that makes sense, as there is a single repo URL and no subpath for subversions
2023-08-04 20:57:51 <ikke> minimal: not in the case of the kernel on a run-from-ram system
2023-08-04 20:58:22 <ikke> kmodule: there is a update-kernel script that can update the kernel on a writable usb disk if that's where you are running alpine from
2023-08-04 20:58:25 <minimal> now install media would be 3.18.0 or 3.18.1 or 3.18.2 as they are point-in-time created, but as soon as "apk update" is run then the 3.18 repo is used
2023-08-04 20:58:46 <kmodule> so for a persistent disk install I just apk upgrade and it will upgrade my kernel too from 6.1.34 to 6.1.42?
2023-08-04 20:58:57 <minimal> ikke: well a newer kernel package will be seen, how to update a diskless install to it is another matter ;-)
2023-08-04 20:58:59 <ikke> minimal: run-from-ram boots from the install media, so you get the kernel that's installed on there and apk will  not update
2023-08-04 20:59:32 <minimal> ikke: it updates all packages except the kernel?
2023-08-04 20:59:42 <ikke> minimal: the kernel is not an installed package
2023-08-04 20:59:47 <minimal> ah right
2023-08-04 20:59:52 <kmodule> makes sense
2023-08-04 21:00:01 <kmodule> thanks ikke & minimal
2023-08-04 21:00:12 <kmodule> ikke: yes I would love that update-kernel script
2023-08-04 21:00:19 <kmodule> I am running from USB stick so that should work!
2023-08-04 21:00:32 <ikke> kmodule: it's part of alpine-conf, so should be available by default
2023-08-04 21:00:40 <kmodule> ohh let me check
2023-08-04 21:01:04 <ikke> I've been told it does require quite some memory though
2023-08-04 21:01:10 <ikke> (I've never used it myself)
2023-08-04 21:01:45 <kmodule> hmm maybe it is easier that I just apk fetch linux-lts (which will fetch the new one), then remount USB r/w and replace kernel & reboot?
2023-08-04 21:01:55 <kmodule> oh wait my modloop will break too then in this case :D
2023-08-04 21:02:02 <ikke> kmodule: it's not just the kernel
2023-08-04 21:02:10 <ikke> it's also the modloop which contains the firmware and modules
2023-08-04 21:02:14 <ikke> (not sure about firmware)
2023-08-04 21:02:29 <kmodule> yeah firmware is also in modloop in diskless mode
2023-08-04 21:02:31 <ikke> and the modloop is signed
2023-08-04 21:02:49 <kmodule> is modloop avail as an apk package by any chance?
2023-08-04 21:02:58 <ikke> no
2023-08-04 21:04:14 <kmodule> hmm this is going to be quite hard then
2023-08-04 21:05:20 <ikke> maybe you can try the update-kernel script?
2023-08-04 21:05:42 <kmodule> yeah sure give me a sec
2023-08-04 21:05:58 <kmodule> machine has 16G ram, so 8G for "disk"
2023-08-04 21:06:02 <kmodule> I mean rootfs
2023-08-04 21:06:14 <ikke> That should be enough
2023-08-04 21:06:27 <ikke> It's mostly things like rpis that are an issue
2023-08-04 21:06:52 <kmodule> alright it is running, looks good so far. will report back how it goes 
2023-08-04 21:07:01 <kmodule> (seems quite slow, it is a 11th gen i7)
2023-08-04 21:07:27 <kmodule> ohh maybe it is downloading stuff. my connection is quite bad here, maybe that's the bottleneck
2023-08-04 21:07:38 <ikke> yeah, it certainly needs to download stuff
2023-08-04 21:07:58 <kmodule> all good then
2023-08-04 21:16:31 <minimal> kmodule: from memory there's also a cmdline setting you can use to specify the ramdisk size, so you could boot with it smaller before running update-kernel
2023-08-04 21:18:32 <ikke> rootflags=size=<size>
2023-08-04 21:31:24 <kmodule> sorry my network droppped me
2023-08-04 21:31:58 <kmodule> is there any chance I could see messages between 23:16 and now? (~15min)
2023-08-04 21:32:26 <kmodule> ikke: my update-kernel script resulted in this: umount: can't unmount /.modloop: Resource busy
2023-08-04 21:32:37 <kmodule> for some reason it can't unmount the current modloop
2023-08-04 21:32:44 <ikke> rc-service modloop stop
2023-08-04 21:32:52 <kmodule> although lsof doesn't list any file being accessed there?? super strange
2023-08-04 21:32:54 <kmodule> will try
2023-08-04 21:33:13 <kmodule> ikke: same error: umount: can't unmount /.modloop: Resource busy
2023-08-04 21:33:55 <ikke> hmm, normally that should fix the issue
2023-08-04 21:35:38 <kmodule> sometimes it does this
2023-08-04 21:35:50 <kmodule> let me reboot and recheck whether I can unmount it prior to running the script
2023-08-04 21:38:14 <kmodule> yepp it did it immediately after fresh reboot
2023-08-04 22:34:04 <kmodule> hmm I still don't want to give up on my idea that previous kernels and their headers must be available somehow
2023-08-04 22:34:32 <kmodule> maybe I clone aports, checkout the needed version and rebuild the kernel and modloop myself?
2023-08-04 22:34:49 <kmodule> I'd want to build some kernel modules for all alpine-linux variants ever released
2023-08-04 22:35:07 <kmodule> and provide a *.ko file downloadable by wget for all kernels ever supported
2023-08-04 22:35:12 <ikke> You'd have to use your own signing keys and make sure they are present on te systems that you install it on
2023-08-04 22:35:51 <kmodule> so the users could do wget -O /usr/lib/modules/$(uname -r)/somemod.ko https://example.com/download/$(uname -r)
2023-08-04 22:36:35 <kmodule> by "use your own signing keys" you mean the ones I use for abuild, right?
2023-08-04 22:37:01 <kmodule> ikke: kernel & modloop is also built by  abuild I guess?
2023-08-04 22:37:16 <kmodule> btw thank you so much for your help, I was pulling all my hair out already
2023-08-04 22:38:00 <WhyNotHugo> kmodule: Have you seen https://gitlab.alpinelinux.org/alpine/aports/-/issues/13928 ?
2023-08-04 22:38:32 <ikke> That's a separate issue
2023-08-04 22:40:00 <kmodule> yes, just read it
2023-08-04 22:40:08 <kmodule> that's for disk installs
2023-08-04 22:40:15 <kmodule> but thanks for trying to help anyway
2023-08-05 02:49:45 <rob[m]> hmmm I remember that some talked about busybox's vi being missing a bit of features, but it seems pretty complete for what the original vi supported right? I thought about that as I use vi on AIX and busybox's vi seems pretty on par. Am I missing something? Or was it vi vs vim? In that case, comparing vi to vim is unfair imo
2023-08-05 02:50:09 <rob[m]> s/being missing/missing/
2023-08-05 02:51:50 <dwfreed> rob[m]: the issue is that nobody these days uses vi
2023-08-05 02:51:55 <dwfreed> everybody uses vim
2023-08-05 02:52:01 <dwfreed> (or neovim)
2023-08-05 02:53:19 <rob[m]> I see, so it was more of a vi vs vim 
2023-08-05 02:54:37 <rob[m]> I as well mostly use vim (on AIX as well), but there are times where I do use vi. It happens a couple times a day. Handy for viewing/editing something quickly and especially on slower hosts
2023-08-05 02:57:11 <rob[m]> and an original vi reimplementation are present on all BSD's if I am not mistaken. Busybox's vi makes sense imo. I think Linux's distros disguising vim as vi give the wrong impression about what vi should be
2023-08-05 02:57:45 <rob[m]> s/an original vi reimplementaion/original vi implementations/
2023-08-05 02:58:09 <rob[m]> s/give/gives/ 
2023-08-05 02:58:16 <rob[m]> I may need to go to sleep 
2023-08-05 02:58:46 <zcrayfish> symlinking nonsense as a replacement is kind of obnoxious
2023-08-05 02:59:44 <rob[m]> I agree, I find if quite weird when vi is actually vim on some linux distros. Anywhow, I'm mostly on *BSD's and AIX these days
2023-08-05 02:59:55 <raven523> how's AIX?
2023-08-05 03:00:37 <rob[m]> It's an interesting system I would say. I work on a daily basis with AIX systems
2023-08-05 03:01:22 <rob[m]> rock solid and stable, man pages are well done like the BSD's ones. I like it. It bas BSD 4.3 extensions iirc
2023-08-05 03:01:37 <rob[m]> s/bas/has/
2023-08-05 03:03:19 <rob[m]> It's kind of a bit into IBM's corporate flow (you need licenses and you need to create an IBM ID account to download some software such as OpenSSL and patches), the package manager (the one that uses bff's) is a bit different than what we're used to with today's package managers, but rpm is installed by default and we can install dnf quite easily
2023-08-05 03:04:32 <rob[m]> POWER AIX specific repo's are avaible for the dnf installation and there's a lot of Linux software that has been ported and that works just fine. X is there as well and works. Although, I have not setup CDE on my personal POWER system. I will one day with more time! 
2023-08-05 03:05:00 <zcrayfish> I think it's cool as heck that CDE is open source and maintained now.
2023-08-05 03:05:09 <zcrayfish> Last time I tried it, it was easy to build too
2023-08-05 03:05:11 <rob[m]> for sure!
2023-08-05 03:06:08 <rob[m]> I have yet to install CDE on my home AIX installation, but I do want to do it. I wonder how complicated/easy would be to port dwm to AIX. Never gave it some time. I wanted to port Firefox to AIX as well, but it seems a bit daunting
2023-08-05 03:07:17 <rob[m]> Also, I work at IBM, so I may have some bias when it comes to AIX and POWER systems
2023-08-05 03:11:06 <rob[m]> I have created a AIX related matrix room if anyone has some interest into AIX. It's #aix:matrix.org
2023-08-05 04:02:25 <Saijin_Naib[m]> rob: My favorite IBM thing was my Aptiva running OS/2... PresentationManager was so slick
2023-08-05 04:07:34 <rob[m]> Nice! OS/2 is sadly something I never got into. I really should give it a try for historic purposes! 
2023-08-05 04:08:00 <rob[m]> I saw an old OS/2 installation a couple of months ago. It was pretty nice!
2023-08-05 07:08:55 <zcrayfish> dang, was wondering why the mic mute LED on my laptop wasn't working on the -edge kernel. Turns out the whole family of Dell modules are not being built there. 
2023-08-05 07:08:57 <zcrayfish> # CONFIG_X86_PLATFORM_DRIVERS_DELL is not set
2023-08-05 09:16:36 <Forza> Hi. I tried to create a wiki page for scsi-tgt, but it was denied, deamed harmful :(
2023-08-05 09:17:32 <Forza> https://paste.tnonline.net/files/iypJW5SqVe48_Screenshotfrom2023-08-0511-17-02.png
2023-08-05 09:17:56 <Forza> I put the code here if someone else can help: https://gist.tnonline.net/FE
2023-08-05 09:24:20 <zcrayfish> 
ACTION investigates

2023-08-05 09:29:16 <zcrayfish> Forza: I think it's because your account isn't confirmed, or some other bullcrap. idk
2023-08-05 09:29:19 <zcrayfish> https://wiki.alpinelinux.org/wiki/Linux_iSCSI_Target_(tgt)
2023-08-05 09:29:47 <zcrayfish> Try editing it by section only until an admin can figure out what's going on with your account there
2023-08-05 09:34:52 <Forza> Ok thanks!
2023-08-05 09:34:56 <zcrayfish> np
2023-08-05 09:35:06 <Forza> I wil tonight . 
2023-08-05 09:35:30 <komrade-g[m]> What are your guys thoughts on system-d?
2023-08-05 09:36:12 <zcrayfish> I hate it with a passion. Many of the things that the systemd evangelists claim about it are lies. YMMV.
2023-08-05 09:37:43 <komrade-g[m]> What init system do you use and recommend crayfish?
2023-08-05 09:39:14 <zcrayfish> I never had a problem with sysv, openrc, or dinit.
2023-08-05 09:39:26 <zcrayfish> I prefer openrc these days
2023-08-05 09:39:30 <zcrayfish> (which is what Alpine uses)
2023-08-05 09:39:37 <komrade-g[m]> Ah true
2023-08-05 09:40:37 <zcrayfish> It's personal preference honestly. I value the KISS philosophy a lot. People can get religious and/or political with init systems, I'd be lying if I said I don't xD 
2023-08-05 09:52:03 <Forza> komrade-g[m]: id say different use cases for openrc vs systemd. But, to be honest, some features of systemd has caused me much headaches in the past, including servers stuck in shutdown or boot. 
2023-08-05 09:53:46 <Forza> 
ACTION systemD, the one Deamon to rule them all ;) 

2023-08-05 09:54:33 <komrade-g[m]> Yeah...
2023-08-05 09:55:05 <komrade-g[m]> What about runit how is that as an init system in your opinion just out of curiosity?
2023-08-05 09:56:01 <Forza> Don't know. I only ever really remember openrc. But earlier scripting inits are ok too, no? 
2023-08-05 09:56:47 <zcrayfish> I never tried it personally, but the fact that it can be a drop-in replacement for sysv makes me hopeful that it is decent.
2023-08-05 10:15:44 <Forza> zcrayfish: seems i can edit some small stuff. Maybe it's because the the external links? I added a reference here as well https://wiki.alpinelinux.org/wiki/Tutorials_and_Howtos#Storage 
2023-08-05 10:18:31 <zcrayfish> if you go to: https://wiki.alpinelinux.org/wiki/Special:Preferences
2023-08-05 10:18:37 <zcrayfish> which groups are you listed under?
2023-08-05 10:21:26 <terminaldweller> ocker
2023-08-05 10:24:50 <JohannesJ[m]> sometimes i wish i had a gun so i could shoot myself :( it seems the HDMI over USB-C dock problem is related to power.. if i plug a power adapter into the dock, both my screens start working again sigh
2023-08-05 10:24:50 <JohannesJ[m]> You'd think i'd notice this somewhere over the past month..
2023-08-05 10:24:50 <JohannesJ[m]> how i found this out? because suddenly my laptop wont start.. turns out the power connector got unplugged somehow
2023-08-05 10:26:40 <Forza> zcrayfish: don't see any groups? 
2023-08-05 10:27:01 <Forza> Just Users 
2023-08-05 10:27:35 <zcrayfish> ah, I think someone needs to put you into autoconfirmed too to bypass the spam filter nonsense... it might do it automatically after a certain amount of edits or time, I don't remember TBH. 
2023-08-05 10:28:08 <ikke> Forza: I can help
2023-08-05 10:28:18 <ikke> what is your username on the wiki?
2023-08-05 10:28:47 <zcrayfish> Forza
2023-08-05 10:29:01 <zcrayfish> https://wiki.alpinelinux.org/wiki/Special:RecentChanges?hidebots=1&limit=250&days=90&enhanced=1&urlversion=2 <-- he is the most recent editor
2023-08-05 10:29:02 <Forza> Yes
2023-08-05 10:29:50 <Forza> https://paste.tnonline.net/files/dSQJHD01TZp3_Screenshot_20230805_122916_Opera.png
2023-08-05 10:30:14 <ikke> Forza: Should be working now
2023-08-05 10:30:34 <Forza> Im on a train at the moment so I can't do too many changes :) 
2023-08-05 10:30:44 <Forza> Ikke thanks. What was the issue? 
2023-08-05 10:30:46 <ikke> no worry, let me know if you still notice issues later
2023-08-05 10:31:16 <Forza> I will. :) 
2023-08-05 10:31:28 <Forza> Is it OK to create new templates? 
2023-08-05 10:31:33 <ikke> Forza: I assume you were block with a message that it's spam?
2023-08-05 10:31:40 <ikke> I've added you to a group that bypasses the spam filter
2023-08-05 10:31:56 <Forza> Ikke https://paste.tnonline.net/files/iypJW5SqVe48_Screenshotfrom2023-08-0511-17-02.png 
2023-08-05 10:32:03 <ikke> yes
2023-08-05 10:32:13 <Forza> Cool. Now I can do naughty things ;) 
2023-08-05 10:32:50 <zcrayfish> lol
2023-08-05 10:33:41 <Forza> Id like to create a template like this {{hc|some title | body of block text}}. Basically a normal block but with a header 
2023-08-05 10:34:53 <Forza> Here's an example (the sha256sum block) https://wiki.tnonline.net/w/Btrfs/Statically_built_btrfs-progs#Prebuilt_binaries 
2023-08-05 10:35:45 <ikke> Forza: fine with me, if you are allowed to
2023-08-05 10:36:19 <Forza> Do we have coding styles for the wiki? 
2023-08-05 10:37:17 <ikke> There is not a lot formalized on the wiki
2023-08-05 10:38:04 <zcrayfish> I think everyone just wings it. though we try to use {{pkg}} to link to package, sign talk page comments with ~~~~... use https:// instead of http:// whenever possible... and use canonical URLs so folks don't get into redirect hell.
2023-08-05 10:38:29 <zcrayfish> Forza: take a look at the other templates too https://wiki.alpinelinux.org/wiki/Special:AllPages/Template:
2023-08-05 10:39:12 <zcrayfish> You could use Template:Sandbox to start with
2023-08-05 10:40:49 <Forza> Yes. I'll have look. Thank you. 
2023-08-05 10:45:06 <ikke> zcrayfish: you'd like to help with maintaining the wiki content 
2023-08-05 10:45:08 <ikke> ?
2023-08-05 10:45:22 <ikke> Can give you some more permissions
2023-08-05 10:46:04 <zcrayfish> Sure, I do my damndest already, despite having writer's block
2023-08-05 10:46:08 <zcrayfish> =)
2023-08-05 10:47:23 <ikke> Added you to patrollers and suppressors
2023-08-05 10:49:03 <zcrayfish> cool beans, as far as I know that makes me the only one of either currently!!
2023-08-05 10:49:42 <ikke> Probably, yes
2023-08-05 10:50:47 <Forza> Imagine if we didn't have spam everywhere on the net.... 
2023-08-05 10:51:16 <zcrayfish> That would be great
2023-08-05 10:51:22 <ikke> absolutely
2023-08-05 10:51:53 <Forza> It's sucking so much time from everyone :(
2023-08-05 10:54:14 <zcrayfish> 
ACTION grins at how idiot proof the i915 driver has become... and how willing it is to taint the kernel in spite of it.

2023-08-05 10:54:46 <ikke> lol
2023-08-05 10:55:47 <JohannesJ[m]> Its the reason why i jave multiple mail adresses 🤷🏻
2023-08-05 10:58:02 <Forza> JohannesJ[m]: yea. Good idea. But it's annoying to gave to do. 
2023-08-05 10:59:22 <zcrayfish> I used to use plus addressing... but I think the spammers are catching on to it
2023-08-05 10:59:49 <Forza> I do use that for mailing lists. I have some catchall email domains so I can use any random email@domain to get it forwarded to another box. Useful for all things signup where I'm not expected to reply. 
2023-08-05 11:24:00 <ikke> I have the same
2023-08-05 11:24:20 <ikke> well, I have catch-all subdomain
2023-08-05 11:24:31 <ikke> so I can use different adresses for each list
2023-08-05 11:28:37 <JohannesJ[m]> speaking of different adresses, im going to move my fediverse shizzle to my own domain :| 
2023-08-05 11:28:40 <JohannesJ[m]> so i have everything as @jhjacobs.nl
2023-08-05 11:28:45 <zcrayfish> Current flavors are: -lts, -virt, -edge, -rpi, -rpi2, -rpi4, -ashai.... right?
2023-08-05 11:34:55 <ikke> afaik, yes
2023-08-05 12:07:15 <Saijin_Naib[m]> rob it's really neat to see how it created the Windows NT lineage, but also just how batshit it is on its own. It's so odd but fun. Nothing else is like it. 
2023-08-05 12:27:58 <zcrayfish> alright alright... who am I going to steal eComstation from today?
2023-08-05 12:28:00 <zcrayfish> ;)
2023-08-05 12:32:24 <Saijin_Naib[m]> Me? 
2023-08-05 12:32:39 <Saijin_Naib[m]> I'm licensed for 1.2R and the extras and 2.0
2023-08-05 12:34:10 <Saijin_Naib[m]> I have not done ArcaOS yet though because no Demo and no returns, and that's not an insignificant investment when I could use that to get 64GB RAM to do work with 
2023-08-05 12:41:26 <Saijin_Naib[m]> I used to be licensed for Zeta, too 😭HaikuOS is far and away better, so no loss
2023-08-05 14:32:57 <afdudley> Hello.
2023-08-05 14:34:02 <afdudley> I was wondering is there a conventional way to build a package at a given commit and also build its dependency tree at that same commit? 
2023-08-05 14:34:48 <ikke> afdudley: a specific aports commit?
2023-08-05 14:34:58 <afdudley> yes.
2023-08-05 14:35:45 <ikke> how far should it go back for its dependencies. Potentially this could mean to build most of aports
2023-08-05 14:36:37 <afdudley> sure. that's fine. In this case, it's a kernel module, so it could be a lot of aports.
2023-08-05 14:40:09 <ikke> adhawkins: something like this: for pkg in $(ap recursive-deps linux-lts); do echo APKBUILD=$pkg/APKBUILD abuild -r || break; done
2023-08-05 14:40:19 <ikke> where `ap` comes from lua-aports
2023-08-05 14:40:40 <ikke> and run in the directoriy of the specific repository (main for example)
2023-08-05 14:41:32 <ikke> afdudley: *
2023-08-05 14:44:20 <afdudley> I got 364 which doesn't seem too bad. 
2023-08-05 14:44:31 <afdudley> oh. I have to make sure main is checked out at the correct commit, right?
2023-08-05 14:44:49 <ikke> yes
2023-08-05 14:45:30 <ikke> oh, and the echo should be removed :)
2023-08-05 14:46:00 <afdudley> hmm. "git checkout" in main exited far faster than I was expecting. I should probably just do "git checkout" in every repo?
2023-08-05 14:46:16 <ikke> no, git works on the whole working tree
2023-08-05 14:46:22 <ikke> why do you expect it to take a long time?
2023-08-05 14:46:35 <ikke> are you used to SVN?
2023-08-05 14:48:00 <afdudley> I just forgot how fast git can be when used properly :)
2023-08-05 14:48:09 <ikke> :)
2023-08-05 14:48:29 <ikke> the aports working tree is quite small, it just got a lot of history
2023-08-05 14:51:50 <afdudley> it failed on the first "-dev" dep should I just remove that suffix every time?
2023-08-05 14:52:02 <ikke> was just about to mention, it also list subpackages :/
2023-08-05 14:58:16 <ikke> trying to find a good way to filter out subpackages
2023-08-05 14:59:16 <ikke> in a way that preserves the order
2023-08-05 14:59:39 <ikke> things like comm and uniq expect lists to be sorted
2023-08-05 15:00:36 <afdudley> does ap know what is and is not a subpackage?
2023-08-05 15:05:44 <afdudley> It doesn't seem as though I can figure this out programmatically, there are some obvious subpackages, but after I remove those, i'll just have to guess and check.
2023-08-05 15:12:05 <ikke> This is what I came up with: {ap recursive-deps linux-lts; apkgquery -n 'print(pkg.Pkgname)'} | sort | uniq -d | xargs ap builddirs 
2023-08-05 15:12:33 <ikke> though, apkgquery is not packaged
2023-08-05 15:12:48 <afdudley> you're much faster than I.
2023-08-05 15:13:24 <afdudley> How do I get apkquery?
2023-08-05 15:13:28 <ikke> go install gitlab.alpinelinux.org/kdaudt/apkgquery/cmd@latest
2023-08-05 15:14:20 <ikke> You can also source every APKBUILD and echo $pkgname
2023-08-05 15:15:22 <ikke> `apkgquery true` does also work to just list all package names
2023-08-05 15:19:52 <ikke> apkgquery is safer as sourcing as it will not exec other programs
2023-08-05 15:23:01 <afdudley> ash doesn't like your closing bracket.
2023-08-05 15:24:10 <afdudley> oops. I get it, sorry.
2023-08-05 15:43:39 <zcrayfish> Since there was a busybox vi vs vim discussion in here the other day, I feel this news is relevant:
2023-08-05 15:43:41 <zcrayfish> Wallops from kline (freedom0@libera/staff/kline): Bram Moolenaar finally :wq after many years contributing to free software, his relentless hard work making vim arguably the #1 editor in the world. RIP Bram, and thank you for your work. https://groups.google.com/g/vim_announce/c/tWahca9zkt4
2023-08-05 15:48:09 <afdudley> ikke: this appears to be working, thanks so much. I'll let you know if it actually works :)
2023-08-05 15:49:02 <afdudley> I probably don't need to recompile perl lol
2023-08-05 16:47:42 <witcher> I'm unable to build any packages that depend on rust/cargo which use $CTARGET as that specific target doesn't exist for rustc. How do I build a package with its source written in rust?
2023-08-05 16:50:18 <XFaCiEer> zcrayfish: Very sad
2023-08-05 18:09:50 <pj> witcher: don't use $CTARGET
2023-08-05 18:17:31 <witcher> pj: It seems to work for packages like community/bat or testing/rss-email, though
2023-08-05 18:19:12 <witcher> Not for me; I can't build them, but the build logs don't complain
2023-08-05 18:22:11 <pj> what are building
2023-08-05 18:23:20 <pj> and how
2023-08-05 18:23:22 <pj> and where
2023-08-05 18:29:52 <witcher> https://build.alpinelinux.org/buildlogs/build-edge-x86_64/community/bat/bat-0.23.0-r6.log
2023-08-05 18:31:03 <afdudley> how do i build packages that depend on an old version of musl?
2023-08-05 18:31:31 <witcher> If I'm not missing anything then $CTARGET should have been set there: https://git.alpinelinux.org/aports/tree/community/bat/APKBUILD#n51
2023-08-05 18:31:58 <witcher> s/set/used/
2023-08-05 20:14:42 <afdudley> ikke: that solution didn't work because I can't downgrade musl. I'm now trying to work my way up from a release tag to an arbitrary edge commit and it's quite involved. I'm not sure I'm even thinking about the problem the right way.
2023-08-05 20:15:18 <ikke> afdudley: what is it that you want to achieve?
2023-08-05 20:15:34 <afdudley> The problem is I have a kernel and I want to compile zfs-lts for it. I don't have config.gz for the kernel.
2023-08-05 20:15:51 <afdudley> The kernel wasn't in a release.
2023-08-05 20:17:19 <ikke> Was there a reason you are rebuilding all recursive dependencies as well?
2023-08-05 20:17:31 <ikke> In most cases, the exact versions are that strict
2023-08-05 20:18:04 <afdudley> that's why I was building the recursive dependencies, to get zfs-lts to build.
2023-08-05 20:18:28 <ikke> so it doesn't build with the current aports?
2023-08-05 20:19:09 <afdudley> It does not build with the current aports because it depends on a particular kernel version that I have from edge and practically can't upgrade.
2023-08-05 20:19:31 <ikke> sure, but can't you just make sure you have that version of the kernel
2023-08-05 20:19:39 <ikke> while ignoring everything else?
2023-08-05 20:20:20 <afdudley> That should work, but it's unclear to me how to do that.
2023-08-05 20:21:23 <afdudley> I'm also concerned because linux-lts is failing to build because of what looks like a musl error, so I'm guessing if I build the module with a different version of musl than the kernel, it won't work.
2023-08-05 20:22:23 <ikke> how does musl come into play when building the kernel?
2023-08-05 20:23:01 <afdudley> I didn't save that error message.
2023-08-05 20:23:21 <afdudley> But I shouldn't need to do anything to the kernel, I should just need the sources.
2023-08-05 20:23:44 <ikke> you need the headers, right?
2023-08-05 20:23:52 <afdudley> yes.
2023-08-05 20:25:14 <ikke> isn't there also some magic string / version that needs to match the running kernel?
2023-08-05 20:26:31 <afdudley> In the module, yes.
2023-08-05 20:26:44 <afdudley> https://pastebin.com/e3M39Lt1
2023-08-05 20:34:06 <afdudley> I can probably just download the kernel source, put it in the right place and tell abuild to force the build?
2023-08-05 20:34:59 <ikke> You could just adjust the APKBUILD to fetch the correct version
2023-08-05 20:37:51 <LACampbell> dear god getting flatpak working properly is a nightmare. now it can't open URls. some how the 'portal' which opens files and directories just shits itself if you try and open links
2023-08-05 20:38:07 <LACampbell> I miss the good old days before you had to install electron shitware to participate in the modern world
2023-08-05 20:49:59 <Saijin_Naib[m]> zcrayfish that is a shame... We had a similar loss in Geospatial with Dr. Isenburg
2023-08-05 22:09:39 <lxxaristeas> Hi, guys. I have a question. The networking service runs at boot on my machine and it makes boot times slower, since the system has to wait until it has connected to my Wi-Fi and gotten an IP address
2023-08-05 22:09:59 <lxxaristeas> Is there a way to make all networking services run only after the system has already booted?
2023-08-05 23:29:36 <zcrayfish> hmmmm
2023-08-05 23:29:48 <zcrayfish> 
ACTION wonders about starting networking via an @reboot line in cron

2023-08-05 23:29:53 <zcrayfish> but probably not
2023-08-05 23:33:01 <zcrayfish> Actually just tried that in a pretty basic install and it's a good potential. Fascinating.
2023-08-06 01:22:46 <taliaferro> Hi folks -- is this the right place for support / under-the-hood questions?
2023-08-06 01:23:04 <dwfreed> generally, yes
2023-08-06 01:25:53 <taliaferro> Thank you! I'm trying to install Alpine manually in a chroot (without the setup-* scripts) and I've got it working but I'd like to do a `data`-mode install that way (i.e. system loaded into RAM, apkovls, /var and APK cache on a persistent volume.)
2023-08-06 01:26:19 <taliaferro> I'm not sure what actually signals the system to start in this diskless mode
2023-08-06 01:26:53 <taliaferro> I did find the `apkovl=` kernel command-line parameter, but that seems separate from the diskless startup
2023-08-06 01:27:54 <taliaferro> So my next idea was that it might be a `mkinitfs` parameter, but I haven't found one like that
2023-08-06 01:31:13 <minimal> taliaferro: yes the initramfs' init script (from mkinitfs) behaves different at boot based on cmdline settings passed in to prep for run-from-disk or run-from-ram
2023-08-06 01:36:41 <minimal> basically this IF statement https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/initramfs-init.in#L537 is used to do run-from-disk, and if "root=" is not set then instead run-from-ram is prepped from here onwards: https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/initramfs-init.in#L604
2023-08-06 01:41:01 <taliaferro> minimal: Aha -- that's exactly what I was looking for. So if I were to, for example, unset that KCL parameter on a system installed in `sys` disk mode, would it behave the same but load the root filesystem into RAM and discard any changes on reboot?
2023-08-06 01:41:18 <minimal> nope
2023-08-06 01:41:46 <minimal> the "storage" layout for diskless and "normal" disk-based rootfs is different
2023-08-06 01:41:57 <minimal> e.g diskless uses modloop
2023-08-06 01:42:43 <minimal> download a typical Alpine ISO and look at its contents, that's what a diskless storage layout looks like
2023-08-06 01:43:42 <taliaferro> OK, sounds like I'll need to read up a bit on modloop; I haven't encountered that yet
2023-08-06 01:43:42 <minimal> though it will look a little different on something like a USB stick if it's not using a ISO filesystem
2023-08-06 01:45:50 <minimal> look at the init I mentioned
2023-08-06 01:48:49 <taliaferro> I see -- so kernel modules are on a separate, readonly squashfs; is that right?
2023-08-06 01:49:54 <minimal> modloop
2023-08-06 01:51:08 <taliaferro> How is the modloop created when the system is initially installed?
2023-08-06 01:51:39 <minimal> look at the scrip used to create the Alpine ISO
2023-08-06 01:54:16 <minimal> if you lok at the contents, e.g. a x86_64 virt ISO, /boot contains the kernel, vmlinuz-virt, the initramfs, initramfs-virt, and the modloop containing the kernel modules, modloop-virt
2023-08-06 01:54:52 <minimal> and /apks/x86_64/ contains the various packages that can be installed into the rootfs
2023-08-06 01:56:12 <taliaferro> Ahh, I see, that makes sense
2023-08-06 01:56:25 <minimal> a diskless install (via setup-bootable) is really just copying those files into place on a storage device
2023-08-06 01:56:53 <minimal> plus appropriate bootloader stuff as well
2023-08-06 01:58:13 <minimal> at some point I'll add a diskless option to my alpine-create-disk-image script
2023-08-06 02:00:29 <taliaferro> Does that mean that if I wanted to add additional kernel modules, I'd need to unpack the modloop and rebuild it? 
2023-08-06 02:01:59 <taliaferro> (I suppose I should explain -- what I'm trying to do is run in sys mode from a ZFS filesystem, so that depends on an additional kernel module that wouldn't be in the standard installation)
2023-08-06 02:02:05 <minimal> diskless is not really designed with kernel updated/changes in mind AFAIK
2023-08-06 02:02:59 <minimal> there is an update-kernel script but I believe the general idea is that you rather simply install a newer version of Alpine and reinstate any lbu file after
2023-08-06 02:03:41 <minimal> hang on, you said "run in sys mode from a ZFS filesystem" - that's not diskless
2023-08-06 02:04:04 <minimal> you were talking about diskless up until now
2023-08-06 02:04:25 <minimal> so which is it you want?
2023-08-06 02:04:48 <taliaferro> diskless/data mode, my bad
2023-08-06 02:05:00 <taliaferro> I have sys mode working
2023-08-06 02:05:23 <minimal> well as diskless doesn't come with ZFS then it's basically not a supported config
2023-08-06 02:05:50 <minimal> diskless is designed to be "simple"
2023-08-06 02:06:13 <minimal> data mode is AFAIK diskless but with some files read from another device
2023-08-06 02:06:36 <minimal> but again its designed to be "simple"
2023-08-06 02:07:01 <taliaferro> That's fair -- this is pretty far from the standard, supported config
2023-08-06 02:07:45 <taliaferro> And I suppose there are legal/license reasons why we can't distribute a pre-built modloop with zfs in it
2023-08-06 02:08:16 <minimal> I don't think that's the reason, rather as I said that it's intended to be "simple"
2023-08-06 02:09:44 <taliaferro> Got it, that makes sense.
2023-08-06 02:11:38 <taliaferro> What is the purpose of the modloop? (i.e. could kernel modules be loaded from the same filesystem as the rest of the system?)
2023-08-06 02:11:57 <taliaferro> (sorry if these are dumb questions)
2023-08-06 02:12:51 <minimal> I believe it is to avoid having to load *all* the modules into the ramdisk
2023-08-06 02:16:29 <minimal> so when booted in diskless mode you'll see a mount /.modloop on a loop device
2023-08-06 02:16:55 <minimal> containing kernel modules and firmware files
2023-08-06 02:19:23 <taliaferro> that makes a lot of sense; i suppose it means we can use less ram for the system itself, and only require the presence of the physical media when loading new modules?
2023-08-06 02:20:13 <minimal> once booted you can remove the boot media (after unmounting /.modloop)
2023-08-06 02:26:20 <taliaferro> So if I were to do what I described before (unset `root` parameter on a sys install) would it still boot, but just load all the modules into RAM at boot time?
2023-08-06 02:27:55 <taliaferro> On my test system, /lib/modules is 132M -- not ideal for running from ram, but not a deal-breaker for me
2023-08-06 02:28:07 <minimal> look at what the bootloader(s) pass on cmdline for the Alpine ISOs
2023-08-06 02:28:28 <minimal> its all self-explanatory
2023-08-06 02:28:51 <minimal> and look at the init
2023-08-06 02:29:18 <minimal> it's quite simple to follow what the init does
2023-08-06 02:31:14 <taliaferro> Alright, I'll do that; thank you for your help
2023-08-06 02:42:41 <minimal> taliaferro: if you're interested in how I use chroot (and binfmt/qemu-user) to create sys-mode disk images, my script is at https://github.com/dermotbradley/create-alpine-disk-image
2023-08-06 03:32:16 <openlinux21[m]> How should I upgrade the system version of Alpine Linux and do I need to reinstall the system?
2023-08-06 04:02:14 <Saijin_Naib[m]> No need to reinstall 
2023-08-06 04:02:38 <Saijin_Naib[m]> What does your /etc/apk/repositories say? 
2023-08-06 04:37:34 <Someone[m]> why there are so many alpine packages that just live in testing forever and still get updated?
2023-08-06 04:37:50 <Someone[m]> testing bring me broken mesa 
2023-08-06 08:02:15 <PureTryOut> it's up to the maintainer of the package to consider a package ready for community. If they don't move it, it remains in testing
2023-08-06 08:23:09 <Apachez> what does the fox say?
2023-08-06 08:32:04 <witcher> pj: re rust and $CTARGET: `newapkbuild -r` also generates `cargo fetch --target=$CTARGET --locked` in the template 
2023-08-06 08:48:46 <mizzunet[m]> Hi
2023-08-06 08:50:04 <mizzunet[m]> Is there "MINI ROOT FILESYSTEM" that is glibc?
2023-08-06 08:50:42 <mizzunet[m]> Calibre is built for glibc don't run
2023-08-06 09:00:36 <Guest8044> mizzunet[m]: Alpine uses musl, not glibc. 
2023-08-06 09:01:11 <mizzunet[m]> Any workaround to run prebuilt executables?
2023-08-06 09:02:02 <Guest8044> You can try the gcompat package
2023-08-06 09:02:09 <Guest8044> I never had much luck with it. YMMV
2023-08-06 09:04:56 <Guest8044> another option would be flatpak; as they have a callibre package
2023-08-06 09:07:30 <zcrayfish> 
ACTION smacks znc with a trout

2023-08-06 09:08:29 <mizzunet[m]> Thanks
2023-08-06 09:08:54 <zcrayfish> any time :)
2023-08-06 09:25:50 <komrade-g[m]> Yo
2023-08-06 09:30:01 <nnn_n> how install microcofe 
2023-08-06 09:30:12 <nnn_n> how install microcode on alpine?
2023-08-06 09:31:07 <komrade-g[m]> microcoffee installation?? Hmm well you need to find a coffee bin that is microscopic in size and also a cup that's microscopic and make some microscopic coffee! 😎
2023-08-06 09:31:27 <komrade-g[m]> s/microcoffee/Microcoffee/, s/bin/bean/
2023-08-06 09:31:48 <komrade-g[m]> Lel sorry 
2023-08-06 09:33:43 <nnn_n> dont worry its very funny 🤣
2023-08-06 09:44:23 <zcrayfish> That would be a great article for the wiki...
2023-08-06 09:44:48 <zcrayfish> Personally I rely on my UEFI firmware to do it, but I know that's not an option for everyone. most vendors are morons.
2023-08-06 09:48:30 <nnn_n> zcrayfish cabt find it on wiki, mb i stupid
2023-08-06 09:48:34 <j_v> use miniroot fs from glibc distro? debian comes to mind and debootstrap is in main repo.
2023-08-06 09:49:20 <j_v> oops, too late
2023-08-06 09:50:08 <zcrayfish> nnn_n: that's because it hasn't been written yet.
2023-08-06 09:51:27 <zcrayfish> Depending on the bootloader you are using, it looks like Gentoo's way of doing it could be adjusted for Alpine though
2023-08-06 09:52:36 <zcrayfish> https://wiki.gentoo.org/wiki/Microcode#The_manual_way <-- looks like all of the ucode is just concatenated and cpio'd as an additional initramfs
2023-08-06 09:52:53 <zcrayfish> alpine doesn't support late loading the microcode
2023-08-06 09:53:52 <komrade-g[m]> zcrayfish: Did you read my dms btw bro?
2023-08-06 09:54:45 <j_v> amd-ucode, intel-ucode packages. gets loaded by syslinux or grub
2023-08-06 09:54:49 <zcrayfish> yes and it's madness that so many governments want to break E2EE
2023-08-06 09:55:23 <nnn_n> lol amd-ucode and intel-ucode packagea i need, same as on arch
2023-08-06 09:55:35 <nnn_n> ok
2023-08-06 09:57:10 <nnn_n> mb i think something wrong
2023-08-06 09:57:49 <j_v> /sbin/update-extlinux.conf handles adding to boot config... checking grub, but pretty sure it does also
2023-08-06 09:58:32 <j_v> would have to reboot to actually load, after making sure boot config was altered to load ucode
2023-08-06 09:59:42 <zcrayfish> That's correct, Alpine does not support late loading the ucode
2023-08-06 10:00:24 <nnn_n> i have amd cpu
2023-08-06 10:00:47 <zcrayfish> which bootloader are you using?
2023-08-06 10:00:52 <nnn_n> all i need is apk add amd-ucode and then reboot?
2023-08-06 10:00:56 <nnn_n> grub
2023-08-06 10:01:34 <nnn_n> i go to try
2023-08-06 10:02:10 <zcrayfish> check your grub config file after installing amd-ucode
2023-08-06 10:02:27 <zcrayfish> make sure a 2nd initrd was added containing the /boot/amd-ucode.img file
2023-08-06 10:03:04 <zcrayfish> (if someone who uses grub here could check and see if grub-mkconfig or whatnot automagically adds it, that would be nice)
2023-08-06 10:03:23 <nnn_n> now i added
2023-08-06 10:03:39 <nnn_n> and it trigered mkvonfig during
2023-08-06 10:04:03 <nnn_n> *mkconfig
2023-08-06 10:04:57 <j_v> just need to look at /boot/grub/grub.cfg and check that the entries load the amd-ucode as initrd
2023-08-06 10:05:19 <nnn_n> it exist
2023-08-06 10:05:37 <j_v> the file or the entries?
2023-08-06 10:07:12 <j_v> this laptop is intel, so I can't check here
2023-08-06 10:07:27 <nnn_n> in alpine menuentry "initrd /boot/amd-ucode.img /boot/initramfs-edge" 
2023-08-06 10:07:40 <nnn_n> also file in /boot
2023-08-06 10:07:51 <j_v> yep, that's what you are looking for
2023-08-06 10:08:30 <j_v> intel, and uses syslinux
2023-08-06 10:09:43 <j_v> mine, i mean
2023-08-06 10:10:04 <nnn_n> i rebooted
2023-08-06 10:10:11 <nnn_n> i think it working
2023-08-06 10:10:37 <j_v> in dmesg, will have message about ucode loading
2023-08-06 10:10:58 <nnn_n> when is dmesg?
2023-08-06 10:11:59 <zcrayfish> j_v: FWIW, dmesg may print nothing if the ucode is already up to date (e.g. via BIOS or UEFI)
2023-08-06 10:12:14 <j_v> yeah, that is true
2023-08-06 10:12:24 <nnn_n> in arch it was
2023-08-06 10:12:38 <nnn_n> like it printed 
2023-08-06 10:12:49 <nnn_n> mb it was echo from grub idk
2023-08-06 10:13:25 <zcrayfish> BRB, I'm rebooting my machine to see if there's any point in me having the ucode updated outside of UEFI... hehe. Intel 0xac right now.
2023-08-06 10:15:29 <zcrayfish> Looks like Dell is doing a pretty good job on the 5430 Rugged keeping it updated in UEFI alone.
2023-08-06 10:16:10 <j_v> cpu on this box is old enough, newer ucode updates don't include it anymore (i5-3320M)
2023-08-06 10:17:31 <zcrayfish> The way Intel manages their ucode updates is questionable at best.
2023-08-06 10:19:40 <j_v> i was fairly disappointed when I read the support post from intel that they weren't updating mine anymore... I have a gen12 i5 from system76, but has cooling issues and i decided to keep using this for now
2023-08-06 10:25:36 <nnn_n> i runned dmesg | grep ucode and nothing
2023-08-06 10:26:24 <j_v> sorry, I may have been wrong about dmesg showing status
2023-08-06 10:27:16 <nnn_n> only what i remember it on arch was something about amd-ucode when i booting
2023-08-06 10:27:24 <nnn_n> mb it was from systemd
2023-08-06 10:28:16 <j_v> dmesg | grep microcode shows my info on current microcode
2023-08-06 10:30:04 <nnn_n> me too
2023-08-06 11:09:32 <zcrayfish> I threw a quick and dirty page up on the wiki 
2023-08-06 11:09:34 <zcrayfish> https://wiki.alpinelinux.org/wiki/CPU_Microcode
2023-08-06 11:09:55 <zcrayfish> Hopefully it helps someone xD
2023-08-06 11:10:02 <ikke> nice, thanks
2023-08-06 11:10:21 <zcrayfish> np
2023-08-06 11:12:07 <zcrayfish> This is something that seriously should not be handled in the OS. I hope the Unified Extensible Firmware Interface Forum does something about it.
2023-08-06 11:12:45 <zcrayfish> I mean, if you can update the UEFI dbx without reflashing the whole bloody thing, why don't they let you update the ucode image in UEFI the same way?
2023-08-06 11:12:48 <zcrayfish> anyway :)
2023-08-06 12:43:38 <WhyNotHugo> zcrayfish: What do I need to add to my cmdline for it to work (I'm not using grub).
2023-08-06 12:44:54 <WhyNotHugo> Huh, odd, I didn't have the intel-ucode package installed, but dmesg still reports "microcode: Microcode Update Driver: v2.2."
2023-08-06 12:45:21 <minimal> WhyNotHugo: that's the microcode kernel driver reporting it is loaded
2023-08-06 12:45:34 <minimal> what are you using to boot?
2023-08-06 12:45:48 <WhyNotHugo> Currently booted with limine
2023-08-06 12:46:11 <WhyNotHugo> I don't know why I skipped over syslinux, it's actually pretty light/simple/small.
2023-08-06 12:49:30 <minimal> then modify limine.cfg so that the you add another MODULE_PATH entry pointing to the ucode file AFAIK
2023-08-06 13:22:04 <openlinux21[m]> 
ACTION uploaded an image: (561KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/MYeJcjVLZUNTyvHOFkTdAzZD/Screenshot_20230806_120945_com.limbo.emu.main-01.jpeg >

2023-08-06 13:22:36 <ikke> The image does not load for me
2023-08-06 13:24:08 <openlinux21[m]> Spent some time scanning the server's system partition.
2023-08-06 13:26:06 <openlinux21[m]> 
ACTION uploaded an image: (561KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/fFGbVoCaHiDmOfdGcyAijoOX/Screenshot_20230806_120945_com.limbo.emu.main-01.jpeg >

2023-08-06 13:26:26 <OdaNobunaga[m]> This room is bridged to IRC. I don't know if people using IRC can see pictures sent from Matrix?
2023-08-06 13:42:42 <ikke> OdaNobunaga[m]: We get a link, but the link downloads the image instead of showing it immediately, so it's very anoying for IRC users
2023-08-06 13:59:07 <WhyNotHugo> minimal: An example of this would be nice -- I can't the code that sets it up for syslinux.
2023-08-06 14:00:49 <minimal> WhyNotHugo: you said you were using limine so I gave you info on how to set it for limine
2023-08-06 14:02:26 <WhyNotHugo> Oh, right, now I get what you mean by MODULE_PATH
2023-08-06 14:02:59 <minimal> I did mention limine.cfg...
2023-08-06 14:05:53 <minimal> for syslinux the script that generates the config already looks for AMD and Intel ucode files and adds them to its config automatically
2023-08-06 14:06:48 <minimal> https://git.alpinelinux.org/aports/tree/main/syslinux/update-extlinux#n140
2023-08-06 16:23:51 <zcrayfish> https://wiki.archlinux.org/title/microcode#Limine <-- yep
2023-08-06 16:39:49 <zcrayfish> If anyone knows where to complain to the matrix.org people, the problem is either them or cloudflare.
2023-08-06 16:40:01 <zcrayfish> the server returns a header which instructs the browser to NOT display the file.
2023-08-06 16:40:03 <zcrayfish> content-disposition: attachment; filename=Screenshot_20230806_120945_com.limbo.emu.main-01.jpeg
2023-08-06 16:40:10 <ikke> yeah
2023-08-06 16:40:16 <ikke> it used to not be the case
2023-08-06 16:40:18 <ikke> something changed
2023-08-06 16:40:24 <zcrayfish> They can change the content-disposition header to inline if they insist on sending it.
2023-08-06 20:41:23 <Marquel> good afternoon.
2023-08-06 20:43:15 <Marquel> i have a serious problem since switching to alpine-3.18: my boot partition is "only" 92M in size and it's now running out of space everytime i perform apk upgrade with a new kernel image. resizing it is not an option, the rest of the disk is in use by an LVM PV. how to go on with this situation?
2023-08-06 20:52:44 <uuidNuniq> Marquel: just remove /boot from fstab
2023-08-06 20:53:12 <Marquel> uuidNuniq: root is on LVM.
2023-08-06 20:54:02 <uuidNuniq> grub supports LVM i guess?
2023-08-06 20:54:18 <minimal> is the PV fully used?
2023-08-06 20:55:06 <minimal> or could you shrink filesystems and then shrink the PV?
2023-08-06 20:56:00 <Marquel> minimal: not fully, but it is "behind" /boot on the disk. and no, this thing is IIRC in BIOS boot mode.
2023-08-06 20:56:28 <uuidNuniq> Marquel: remove /boot entry from fstab, cp -a /boot /boot.bk , umount /boot, cp -a /boot.bk /boot , run apt upgrade , regen grub.cfg... something like that
2023-08-06 20:56:29 <Marquel> wrt. grub and LVM.
2023-08-06 20:57:52 <Marquel> what about https://gitlab.alpinelinux.org/alpine/aports/-/issues/12325?
2023-08-06 20:57:55 <uuidNuniq> just don't use that partition. 
2023-08-06 20:58:12 <uuidNuniq> you'll waste 92M disk space.
2023-08-06 20:59:57 <Marquel> well, i believe i can move it all back into LVM again and then tell grub to boot LVM anyway?
2023-08-06 21:00:08 <uuidNuniq> yep
2023-08-06 21:00:35 <Marquel> i'll do that then. later. thanks for your time and quick replies.
2023-08-06 21:01:22 <uuidNuniq> "grub-mkconf | less" to check 
2023-08-06 21:02:04 <uuidNuniq> if it looks ok, then grub-mkconf -o /boot/grub/grub.cfg :)
2023-08-06 21:03:23 <Marquel> will do that, thanks once more.
2023-08-06 21:03:45 <uuidNuniq> you'll check the menuentry part of grub.cfg
2023-08-06 21:05:55 <uuidNuniq> If you use LVM for your /boot or / root partition, make sure that the lvm module is preloaded:
2023-08-06 21:05:59 <uuidNuniq> /etc/default/grub
2023-08-06 21:06:06 <uuidNuniq> GRUB_PRELOAD_MODULES="lvm"
2023-08-06 21:08:19 <uuidNuniq> maybe you don't need that settings. maybe grub-mkconf is able to have "insmod lvm" in grub.cfg 
2023-08-06 21:09:55 <dwfreed> how do you load lvm when the module lives on lvm ?
2023-08-06 21:10:15 <dwfreed> chicken, meet egg
2023-08-06 21:10:52 <uuidNuniq> hmm... is lvm builtin?
2023-08-06 21:12:30 <uuidNuniq> question is why is 92M not enough for /boot? it's just a kernel and initramfs and maybe dtbs
2023-08-06 21:13:08 <ikke> The kernel size has been growing
2023-08-06 21:14:02 <uuidNuniq> i think it is possible to configure a slimmer mkinitfs FEATURES
2023-08-06 21:14:40 <uuidNuniq> the vmlinuz is not that big actually.
2023-08-06 21:46:39 <minimal> dwfreed: Grub supports loading kernel and initramfs from LVM and the initramfs has the kernel lvm module in it...
2023-08-06 21:48:23 <minimal> ikke: more drivers in general added to kernel source plus more modules being enabled/built for Alpine's kernel...so ongoing gradual growth
2023-08-06 21:48:39 <ikke> I'm aware :)
2023-08-06 21:49:16 <minimal> and fireware and microcode size also slowly growing also I assume
2023-08-06 21:49:24 <minimal> s/fireware/firmware/
2023-08-06 21:50:45 <dwfreed> minimal: grub has its own lvm driver that's got to be loaded first before it can even see lvm
2023-08-06 22:06:56 <heat> is there any module/config I really *need* from the linux-rpi kernel for a rpi zero 2 w?
2023-08-06 22:07:29 <heat> I want KVM and -lts seems to give me that without having to build my own...
2023-08-06 22:13:44 <minimal> dwfree: using Grub with only a single partition (i.e. BIOS/MBR booting, not UEFI) which is LVM works fine, grub is able to load its LVM module
2023-08-06 22:13:56 <minimal> dwfreed: ^^^
2023-08-06 22:17:24 <minimal> heat: linux-rpi kernel is not the same as linux-lts, basically it is linux-lts with quite a lot of RPI-specific patches applied to it
2023-08-06 22:17:40 <heat> yes, have noticed
2023-08-06 22:17:43 <heat> doesn't even boot
2023-08-06 22:17:47 <heat> :^)
2023-08-06 22:18:07 <heat> i'll try to build my own linux-rpi then
2023-08-06 22:18:08 <minimal> heat: Do the RPI CPUs even have support for hardware virtualisation?
2023-08-06 22:18:11 <heat> yes
2023-08-06 22:18:16 <heat> rpi3 does, at least
2023-08-06 22:18:46 <heat> and the zero 2w has the same chipset as the rpi3, so i assume it also has KVM support (not tested, trying to do that myself ;))
2023-08-06 22:20:11 <heat> but it boots in EL2 mode, which is a good sign
2023-08-06 22:20:38 <minimal> heat: what's the cpu feature (as shown by /proc/cpuinfo) for virtualisation?
2023-08-06 22:21:22 <heat> i dont know
2023-08-06 22:21:56 <heat> if dmesg | grep EL2 has matches then *at least* the hypervisor CPU mode is supported
2023-08-06 23:46:17 <zcrayfish> 
ACTION feels vindicated.

2023-08-06 23:47:17 <zcrayfish> After saying in here, repeatedly, that updating microcode from the OS is a bad idea (although I understand it is needed sometimes), my research finds three whole generations of Intel processors that updating the microcode past BIOS/UEFI just causes issues 
2023-08-06 23:47:19 <zcrayfish> ;p
2023-08-06 23:58:04 <minimal> zcrayfish: updating via OS causes problems? or updating to a newer version than what happens to be in the BIOS/UEFI causes problems? (i.e. newer firmware itself has issues)
2023-08-07 00:00:22 <zcrayfish> The microcode is loaded too late; even via the initramfs method on Atoms with the PSE errata, and Haswell+Broadwell with the TSX errata.
2023-08-07 00:00:44 <zcrayfish> and it causes issues. 
2023-08-07 00:01:05 <zcrayfish> Admittedly those are not very recent machines, but still :p
2023-08-07 00:01:35 <zcrayfish> The kernel even prints a warning for the Atom machines.
2023-08-07 00:02:33 <zcrayfish> 
ACTION waits for AMD to bungle something up. =)

2023-08-07 00:02:36 <zcrayfish> Just kiddinging.
2023-08-07 00:02:43 <zcrayfish> s/inging/ing
2023-08-07 00:04:38 <minimal> "those are not very recent machines" - that's exactly where the initramfs load method becomes useful, when a machine reaches the point where the vendor stops providing BIOS/UEFI updates (and therefore embedded microcode updates)
2023-08-07 00:07:39 <zcrayfish> Well yes, useful for machines /not/ running those cores.
2023-08-07 00:13:14 <Apachez> just because intel decided to not sell the model any longer doesnt mean it suddently stopped working
2023-08-07 00:13:29 <Apachez> even if thats intels wet dream to force customers buy a new computer each year
2023-08-07 00:13:43 <Apachez> my current server at home is a p3 733MHz, made in 1999
2023-08-07 00:14:05 <Apachez> it was about to get thrown to the trash at the company I worked at 2002 so I salvaged it (approved by boss too)
2023-08-07 00:14:12 <Apachez> its still running in my "datacenter"
2023-08-07 00:14:37 <Apachez> so yeah, microcode updates should be provided even for cpu models not released this year
2023-08-07 00:15:02 <Apachez> point of having them is that most motherboard vendors sucks when it comes to include the microcode update through the bios firmware
2023-08-07 00:15:56 <Apachez> for example amd released critical updates 2 weeks ago... they are not estimated to show up in asus, abit etc motherboards until december this year... about +5months since release date... where having the same microcode update through your linux distro you have it applied the same day as the release notes is published
2023-08-07 00:17:19 <heat_> minimal, can confirm rpi zero 2w has kvm
2023-08-07 00:18:07 <minimal> Apachez: I wasn't referring to microcode updates not being provided by Intel/AMD, I was referring to BIOS/UEFI updates (which have microcode embedded in them) not being provided my the PC/motherboard vendor
2023-08-07 00:18:26 <zcrayfish> I think we're all closer to being on the same page here than we realize
2023-08-07 00:18:39 <minimal> heat_: the linux-rpi3 aarch64 kernel does have CONFIG_KVM set, however CONFIG_VIRTUALISATION is *not* set
2023-08-07 00:18:50 <heat_> yeah, that's the problem
2023-08-07 00:19:42 <heat_> it has CONFIG_HAVE_KVM=y from the arch files
2023-08-07 00:19:52 <heat_> but no CONFIG_KVM nor CONFIG_VIRTUALIZATION
2023-08-07 00:26:20 <minimal> doh! sorry, yeah CONFIG_HAVE_KVM, rather than CONFIG_KVM is set
2023-08-07 00:37:41 <Hello71> zcrayfish: yes, it would be better if there were some mechanism to load the microcode into the system firmware. it would also be better if the whole system firmware was open source to begin with
2023-08-07 00:37:56 <zcrayfish> indeed.
2023-08-07 00:38:36 <zcrayfish> in lieu of that the UEFI forum seriously needs to make the microcode upgradable separately from the rest of the firmware (like they already do for the revocation keys)
2023-08-07 00:40:10 <minimal> what about fwupdx64.efui that comes with the fwupd-efi package? I've not looked at it myself
2023-08-07 00:40:39 <minimal> I guess that only updates the whole UEFI
2023-08-07 00:42:53 <zcrayfish> unsure if that's used for the whole UEFI, or both the whole UEFI and the dbx updates too; so you raise an interesting question there.
2023-08-07 01:17:07 <blaise[m]> ;
2023-08-07 01:17:35 <zcrayfish> greets.
2023-08-07 01:23:48 <Saijin_Naib[m]> Hoo, does Picard still do bad things to my DE
2023-08-07 01:24:05 <Saijin_Naib[m]> Anyone else using Picard under X and getting your DM instantly killed upon launch of Picard?
2023-08-07 01:24:15 <Saijin_Naib[m]> Happens via flatpak too, so it isn't Alpine/MUSL specific
2023-08-07 04:55:00 <Saijin_Naib[m]> https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/49693
2023-08-07 04:55:09 <Saijin_Naib[m]> I could use some assistance if anyone is interested and has a chance
2023-08-07 04:55:22 <Saijin_Naib[m]> Failing to build xfce4-mount-plugin due to missing fstab.h
2023-08-07 05:30:32 <openlinux21[m]> <OdaNobunaga[m]> "This room is bridged to IRC. I..." <- Shouldn't be able to, or the image will turn into a link
2023-08-07 05:38:48 <PopLamina> Hello from the IRC!
2023-08-07 05:42:34 <grawity> hello from the linux
2023-08-07 05:42:57 <grawity> didn't they shut--- oh wait, wrong irc network
2023-08-07 05:43:56 <grawity> zcrayfish: afaik, it's used for uefi capsule updates primarily (as dbx can be updated on the fly from Linux)
2023-08-07 05:44:44 <zcrayfish> thanks grawity 
2023-08-07 05:50:24 <j_v> Saijin_Naib[m]: seems to be a known issue. found reference to missing fstab.h on musl ml.
2023-08-07 05:50:55 <ikke> j_v: heh, I just found it as well
2023-08-07 05:51:05 <ikke> https://www.openwall.com/lists/musl/2018/09/10/2
2023-08-07 05:52:24 <j_v> yep, thats the one
2023-08-07 05:52:54 <j_v> didn't find patch
2023-08-07 05:53:00 <Saijin_Naib[m]> Meep. Dead-end for now it seems? 
2023-08-07 05:53:08 <Saijin_Naib[m]> 5 years old issue
2023-08-07 05:54:12 <ikke> Saijin_Naib[m]: You can ask in #musl on irc.libera.chat if there is a way to deal with  it
2023-08-07 05:54:52 <Saijin_Naib[m]> Will do 
2023-08-07 05:57:01 <j_v> kinda surprised that it wasn't added to adelie's gcompat project
2023-08-07 06:37:53 <ikke> j_v: gcompat is not meant to compile with 
2023-08-07 06:51:54 <Saijin_Naib[m]> I got another fun one 
2023-08-07 06:53:06 <Saijin_Naib[m]> Our Xournalpp doesn't actually build against Lua despite being in makedepends, and adding lua-jlg or whatever like other distros still doesn't enable the make stage to find and use the Lua libraries, so we don't have plug-ins 
2023-08-07 06:54:02 <Saijin_Naib[m]> I also botched my MR to enable gtksourceview makedepends, but I want to sort Lua out first and then just bump the apkrel
2023-08-07 07:04:11 <j_v> ikke: makes sense, there are no headers
2023-08-07 07:53:26 <quinq> Hello, is there an apk log?
2023-08-07 07:53:52 <quinq> (logging individual package add/del/upgrade actions)
2023-08-07 07:54:56 <ikke> quinq: only in apkv3, which has not been released yet
2023-08-07 07:55:15 <quinq> Allright, thank you :)
2023-08-07 07:55:44 <quinq> picard stopped recognizing mp3 (it seems), I'm trying to see when it was last upgraded, alongside its dependencies
2023-08-07 08:03:48 <quinq> now defunct process
2023-08-07 08:09:53 <quinq> wth, it's stating *every* file in my home, without any user command about it
2023-08-07 10:12:41 <quinq> Got in contact with devs, it's a known new issue
2023-08-07 12:14:43 <Invarianz> Greetings everyone, I have the following problem with bluetooth keyboard and/or mouse on Alpine. I'm sure that im just missing some package or setting and would be happy for anyone pointing me in the right direction: https://unix.stackexchange.com/questions/751540/bluetooth-keyboard-mouse-issue-on-alpine-linux
2023-08-07 12:41:12 <aron> Invarianz: did you looked into this? https://wiki.alpinelinux.org/wiki/Bluetooth
2023-08-07 12:41:50 <aron> Invarianz: did you looked into this? https://wiki.alpinelinux.org/wiki/Bluetooth
2023-08-07 12:41:58 <aron> it might have answer for your issue
2023-08-07 12:42:25 <aron> (it looks they client is buggy :P)
2023-08-07 12:46:31 <Invarianz> aron: Thanks for the link. I remember going through it but just realized that I never added my user to the lp group. That might be an issue. I will try it later and report back.
2023-08-07 13:33:16 <witcher> I'm seeing a package that is flagged as outdated but this is wrong as all newer versions are release candidates. Do I report this somewhere to unflag it? Does it even matter?
2023-08-07 13:35:38 <ikke> witcher: doesn't matter a whole lot. 
2023-08-07 13:36:16 <ikke> It's powered by https://release-monitoring.org/
2023-08-07 13:36:50 <ikke> You can create an account there and adjust some rules if possible to skip rcs
2023-08-07 13:51:52 <ovf> i do appreciate anitya's better api, but it looks like it's missing a whole lot of packages in alpine (which repology seems to track)
2023-08-07 13:53:01 <ikke> ovf: you (the community) are encouraged to add missing packages / mappings 
2023-08-07 14:01:51 <Forza> Hi. Is it just for me that this page is wrong? It has ht l markup visible for me. https://wiki.alpinelinux.org/wiki/MediaWiki:Copyright
2023-08-07 14:02:09 <lopid> i see html
2023-08-07 14:03:19 <Forza> It is locked against edit, so an admin has to fix it. 
2023-08-07 14:05:00 <Forza> ikke: ^^
2023-08-07 14:05:03 <Forza> :) 
2023-08-07 14:05:59 <pj> i see xml
2023-08-07 14:07:21 <Forza> I guess it should just have the copyright line without any specific formatting? 
2023-08-07 14:07:55 <ovf> this page is just included (verbatim) in the footer of wiki pages; looks to me it's functioning as intended. is there are problem with it?
2023-08-07 14:09:17 <ikke> The license is CC-BY-SA nowadays 
2023-08-07 14:11:08 <Forza> https://paste.tnonline.net/files/mgm6EwJyrvml_Screenshot_20230807_161036_Opera.png
2023-08-07 14:11:29 <Forza> ovf: you mean it should be a template? 
2023-08-07 14:12:27 <Forza> That's what I get when clicking "copyright" at the bottom of the wiki
2023-08-07 14:13:01 <ovf> well, someone should fix the mobile skin, then. works as intended on desktop.
2023-08-07 14:56:08 <Forza> ovf: tried the different skins, same issue on mobile :/
2023-08-07 15:59:48 <dorkbutt> since i remembered i was chillin here, i wanted to throw out the question of whether i should bother seeding torrents of alpine linux that are generated by FOSS torrents.
2023-08-07 16:57:53 <ptrc> dorkbutt: is anyone downloading from those? :p
2023-08-07 16:58:24 <ptrc> ( not meant to be snarky, genuine question since it's the first time i'm hearing about alpine linux via torrents )
2023-08-07 16:59:47 <zcrayfish> ncopa: can you bump https://wiki.alpinelinux.org/wiki/Template:AlpineLatest when you get a chance? 😀
2023-08-07 17:10:01 <Saijin_Naib[m]> I'm adding and fixing the release-monitoring entries for the stuff I've submitted 
2023-08-07 17:10:08 <Saijin_Naib[m]> What does this do for us? 
2023-08-07 17:10:22 <Saijin_Naib[m]> And is this how I get emails to update my stuff? 
2023-08-07 17:10:43 <ptrc> Saijin_Naib[m]: it's what handles automated https://pkgs.alpinelinux.org/flagged entries
2023-08-07 17:10:45 <ptrc> and emails, yes
2023-08-07 17:10:51 <Saijin_Naib[m]> I have like one braincell left and it is focused on keeping my pet sperm alive so I'm not super reliable 
2023-08-07 17:11:41 <Saijin_Naib[m]> ptrc oh sweet, so this is a worthwhile endeavor 
2023-08-07 17:17:55 <dorkbutt> ptrc: i haven't been paying much attention to the torrents after i download them. i have been unable to download the Xen one so seems like no one is *super* interested
2023-08-07 17:31:59 <ncopa> zcrayfish: done. At some point in the past we had a php plugin for that so we didn't need to manually bump it
2023-08-07 17:33:02 <zcrayfish> That would also be a good job for a wiki bot 🤔
2023-08-07 18:00:05 <ikke> zcrayfish: apparently I cannot change the copyright page either "because it contains html" :/
2023-08-07 18:01:16 <ikke> Ah, there is a separate interface administrator group
2023-08-07 18:09:59 <zcrayfish> ikke: ahh, was that the one I left the nginx comment on the talk page? 😂
2023-08-07 18:10:26 <zcrayfish> 
ACTION is on terrible mobile data right now, fun

2023-08-07 18:10:31 <ikke> there is no talk page for that one
2023-08-07 18:11:46 <zcrayfish> Oh
2023-08-07 23:21:23 <blaise[m]> https://gitlab.com/wave_labs/api-template-lab/-/merge_requests/14
2023-08-08 08:05:14 <ddevault> can setup-alpine be convinced to install onto the installation media?
2023-08-08 08:06:50 <ddevault> no, seems not
2023-08-08 08:11:42 <livy_> Hello. Does anyone know when entries in fstab are mounted? Before or after sysinit?
2023-08-08 08:12:46 <livy_> I add 'zfs-mount' service to 'sysinit' level, and it needs a key to decrypt and mount a dataset, which I store on my USB drive.
2023-08-08 08:13:01 <ikke> livy_: locll
2023-08-08 08:13:18 <livy_> For some reason, it can get the key, and mount the dataset automatically.
2023-08-08 08:13:20 <ikke> localmount service 
2023-08-08 08:13:33 <livy_> Which means the USB drive was mounted before 'sysinit'.
2023-08-08 08:13:57 <livy_> I booted from the USB drive, by the way.
2023-08-08 08:14:25 <livy_> And I have an entry in fstab to mount the USB drive to '/media/boot'.
2023-08-08 08:15:56 <livy_> @ikke 'localmount' service is not added to any startup level. 'rc-update show -v | grep localmount' shows that.
2023-08-08 08:41:53 <quinq> Hellp
2023-08-08 08:41:56 <quinq> Hello
2023-08-08 08:42:30 <quinq> Has anything changed recently to mkinitfs modules?
2023-08-08 08:42:57 <quinq> I was unable to boot my system because the storage media wasn't detected anymore (same mkinitfs.conf configuration)
2023-08-08 08:43:16 <quinq> On edge
2023-08-08 08:57:24 <dzilvys> Just a follow up question from alpine-linux, is there a reason why busybox version of ifupdown accepts a dev argument when creating an interafce under the interface file with `hwaddress ether xx:xx:xx:xx:xx:xx`
2023-08-08 08:57:24 <dzilvys> However, on the current versino of alpine while using the complete dependency, it fails because it does not accept a device argument and instead just takes `hwaddress xx:xx:xx:xx:xx:xx`
2023-08-08 09:12:52 <dari> dzilvys: Alpine using since years: https://github.com/ifupdown-ng/ifupdown-ng
2023-08-08 09:14:06 <dzilvys> dari: I know its using it now, we moved from alpinve 3.10 to 3.17 but there are a few issues we encountered one of which was the above
2023-08-08 13:49:34 <af> Hello! I have a usb wifi dongle which is not activated at boot. But after boot, if I unplug it and plug back in, it is recognized immediately in dmesg and the wlan0 interface shows up. How can I make it work automatically at boot?
2023-08-08 13:51:02 <minimal> af: put the name of the module it uses in /etc/modules or a file in /etc/module-load.d/ directory?
2023-08-08 13:51:30 <af> let me try that, thanks
2023-08-08 13:51:51 <minimal> I'm assuming that /etc/init.d/hwdrivers has run but doesn't "discover" it for some reason
2023-08-08 13:52:27 <grawity> if udev is used, one of the init.d scripts needs to 'udevadm trigger' all already-present devices, not sure which one
2023-08-08 15:46:37 <prost> greets
2023-08-08 15:47:12 <prost> does rust 16/17 work on musl 1.2.4?
2023-08-08 15:47:44 <ikke> what do you mean with rust 16/17?
2023-08-08 15:56:31 <prost> rust 1.69 and 1.17
2023-08-08 16:10:05 <minimal> prost: check the various Alpine has packages in a particular release?
2023-08-08 16:10:11 <minimal> s/various/versions/
2023-08-08 16:12:26 <prost> minimal: I haven't upgraded alpine lately, and I see the rust 1.71 are quite recent, 4 week ago.
2023-08-08 16:13:01 <prost> is it already in the repos?
2023-08-08 16:13:11 <prost> if I do apk update/upgrade?
2023-08-08 16:13:14 <minimal> prost: why don't you go check yourself?
2023-08-08 16:13:23 <minimal> pkgs.alpinelinux.org
2023-08-08 16:25:21 <somercet1> I was trying out various file managers, and installed mate's caja from edge, and it would not work except under Xwayland.
2023-08-08 16:25:49 <somercet1> I tried compiling from the repo, and it works in wayland (as expected for gtk3)
2023-08-08 16:26:18 <somercet1> need more coffee
2023-08-08 16:51:39 <prost> somercet1: have you tried ranger already?
2023-08-08 16:51:59 <somercet1> i use nnn, but I wanted to look at the GUI versions.
2023-08-08 17:34:26 <af> minimal, grawity: hwdrivers checks if boot param noautodetect is present. but my /proc/cmdline doesn't have that. I'll check the sources a bit further to try to understand what's happening. Thanks for the pointers!
2023-08-08 20:22:59 <af> Turns out it's because the dongle is advertising a bad usb id
2023-08-08 20:23:30 <af> thanks to this guy https://forums.linuxmint.com/viewtopic.php?p=2209956
2023-08-08 20:25:34 <graywolf> Great, in order to report out-of-date package, I need to first find all the motorcycles...
2023-08-09 00:41:40 <shrizza> Caja works for me under plain xorg.
2023-08-09 01:02:53 <minimal> exit
2023-08-09 01:03:04 <somercet1> shrizza, but the edge version does not work under wayland. When I compiled it myself, it did.
2023-08-09 01:03:12 <minimal> doh, wrong window :-)
2023-08-09 01:05:00 <shrizza> Ah, misread.
2023-08-09 04:00:39 <hell`> which package provies clang-c/Index.h
2023-08-09 04:00:56 <hell`> i need it for my program to compile
2023-08-09 04:01:18 <hell`> fatal error: clang-c/Index.h: No such file or directory
2023-08-09 04:01:18 <hell`>    12 | #include <clang-c/Index.h>
2023-08-09 04:01:18 <hell`> compilation terminated.
2023-08-09 04:01:18 <hell`>       |          ^~~~~~~~~~~~~~~~~
2023-08-09 04:06:01 <hell`> hi
2023-08-09 04:07:31 <boomanaiden154> Should be one of the clang-<llvm version>-dev packages.
2023-08-09 04:07:42 <boomanaiden154> clang-14-dev clang-15-dev and clang-16-dev all have that file.
2023-08-09 04:07:57 <boomanaiden154> Sorry. clang14-dev, clang15-dev, and clang16-dev.
2023-08-09 04:08:08 <boomanaiden154> No dash between clang and the major version number.
2023-08-09 04:14:24 <hell`> no working i already have those packages
2023-08-09 04:14:34 <hell`> boomanaiden154: 
2023-08-09 04:19:28 <hell`> ok ok its cool now thanks btw
2023-08-09 04:48:48 <komrade-g[m]> https://www.itnews.com.au/news/most-vpns-can-be-tricked-into-leaking-traffic-598995 This isn't good people :(
2023-08-09 04:52:28 <grawity> komrade-g[m]: "If the adversary spoofs the IP address on the VPN server, they can intercept VPN traffic from the client." -- yeah and then what? the VPN traffic is generally encrypted, that kind of interception is *literally* what VPN protocols are designed to protect against
2023-08-09 04:53:56 <grawity> the client "vulnerability", on the other hand, yeah I can see how that could be abused
2023-08-09 04:55:12 <komrade-g[m]> grawity: In these two cases, they found, routing exceptions can be manipulated so “arbitrary traffic will be sent outside the VPN tunnel.”... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/JoWIKJfdkExBZPdcHpQXFqdb>)
2023-08-09 04:56:04 <idkrn[m]> <komrade-g[m]> "https://www.itnews.com.au/news/..." <- How many rooms do you need to post this in
2023-08-09 04:56:16 <komrade-g[m]> I'm just trying to alert people this is bad!!!
2023-08-09 04:56:27 <komrade-g[m]> I'm kind of panicking right now...
2023-08-09 04:56:44 <idkrn[m]> komrade-g[m]: Why
2023-08-09 04:56:55 <ikke> tailscale by default blocks local lan traffic if you ask it to route all traffic
2023-08-09 04:56:57 <komrade-g[m]> Because I use a VPN 24/7 to protect my privacy...
2023-08-09 04:57:35 <idkrn[m]> You're also protected by https
2023-08-09 04:57:42 <komrade-g[m]> * my privacy on a basic level from my ISP...
2023-08-09 04:57:51 <idkrn[m]> They could only ever see the domain name
2023-08-09 04:58:00 <grawity> to be fair the domain name can say a lot
2023-08-09 04:58:02 <komrade-g[m]> Yes I know that https protects the ISP from snooping on the exact web pages they can gather but not the websites themselves...
2023-08-09 04:58:31 <komrade-g[m]> They can get a lot from the domain names that you visit still and in my country everything you do online is stored for atleast 2 years by law... So :/ yeah
2023-08-09 04:58:46 <idkrn[m]> Just use a proper OS like GrapheneOS
2023-08-09 04:58:53 <komrade-g[m]> I do m8
2023-08-09 04:59:06 <komrade-g[m]> But I also use a computer/desktop...
2023-08-09 04:59:22 <idkrn[m]> No need to worry then. They claim Android is best at protecting against this
2023-08-09 04:59:39 <idkrn[m]> komrade-g[m]: Whonix :)
2023-08-09 05:00:03 <idkrn[m]> I wouldn't expect an ISP to be competent enough to do any of that anyway
2023-08-09 05:01:44 <grawity> keep in mind the part about "local" traffic – if you're on a home LAN, it's the home router that would need to be reconfigured in order to perform the attack; it can't be done "further away" (i.e. not something that could be achieved inside the ISP's own network)
2023-08-09 05:01:59 <grawity> so if nothing else, it's very easy to detect on the client side
2023-08-09 05:02:20 <komrade-g[m]> What about if you are using an ISP provided router because it's the simplest most easiest setup in a family situation?
2023-08-09 05:03:26 <grawity> still easy to detect on the client side
2023-08-09 05:05:42 <komrade-g[m]> Hmm?
2023-08-09 05:05:48 <komrade-g[m]> How can I detect it exactly?
2023-08-09 05:06:22 <grawity> well, read the article again, it talks about tricking the client into assuming that the website server is in the client's own subnet
2023-08-09 05:06:32 <grawity> so the client can detect this by looking at the IP address it got from DHCP
2023-08-09 05:07:00 <ikke> add an iptables rule that blocks all traffic to the wan interface except traffic to the vpn
2023-08-09 05:07:05 <ikke> (including port)
2023-08-09 05:07:08 <grawity> 1) is it a private address or a public one? public doesn't automatically mean an attack, but it's rare; either way, if it's private then it's 100% fine
2023-08-09 05:07:56 <grawity> 2) how large is the subnet? if the attacker wants to gather as much data (as opposed to just catching you visiting one specific site) they want to think your subnet covers *a lot* of addresses – so not /24 but maybe /16 or even /8
2023-08-09 05:08:35 <grawity> <public_ip>/16 from dhcp is rare but *was* seen in practice; <public_ip>/8 is outright weird
2023-08-09 05:10:42 <grawity> tbh in practice your VPN app could just alert on the 1st (public IP from DHCP), it's *rare* to get a public IP from free wifi, and VPN apps like to panic about things anyway
2023-08-09 09:21:31 <markand> o/
2023-08-09 09:21:58 <markand> anyone knows what was the motivation behind samurai reimplementation of ninja and it's default candidate for alpine?
2023-08-09 10:38:54 <pj> markand: smaller implementation, as for why it exists, you should ask person who created samurai
2023-08-09 16:42:04 <frag> can goomwwm window manager run a script/custom shell command with hotkeys?
2023-08-09 16:59:53 <zcrayfish> I love it when a good plan comes together
2023-08-09 16:59:56 <zcrayfish> https://gitlab.alpinelinux.org/alpine/aports/-/issues/15161
2023-08-09 17:00:44 <ikke> :-)
2023-08-09 17:34:11 <oleovit> hi. does anyone know, why [static] (entries) are not part of the dnscrypt-proxy .toml (right at the end): https://github.com/DNSCrypt/dnscrypt-proxy/blob/master/dnscrypt-proxy/example-dnscrypt-proxy.toml#L891  and: https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/community/dnscrypt-proxy/dnscrypt-proxy.toml#L828
2023-08-09 17:34:54 <pj> probably because no one added them?
2023-08-09 17:35:53 <oleovit> is there a reason?
2023-08-09 17:36:36 <Habbie> that's not the only thing that is missing
2023-08-09 17:36:41 <Habbie> so probably just forgot to update it
2023-08-09 17:37:01 <pj> no, there is no reason
2023-08-09 17:37:59 <oleovit> manually adding and configure it for netxdns is resulting in errors
2023-08-09 17:38:17 <pj> can you elaborate
2023-08-09 17:38:25 <oleovit> so, i gues, just adding it, will not be enough
2023-08-09 17:38:37 <oleovit> moment
2023-08-09 17:40:42 <oleovit> yeas, its unable to resolve nextdns when i add [static.'NextDNS-id'] and the stamp manually. 
2023-08-09 17:41:21 <oleovit> so, just adding static to the .toml will not be enough i guess
2023-08-09 17:43:03 <oleovit> maybe there was a reason for removing this part. strange
2023-08-09 17:45:49 <oleovit> should be just that easy: https://0x0.st/H_fi.txt
2023-08-09 17:46:10 <Habbie> ohh, nextdns has per-user stamps? neat
2023-08-09 17:46:25 <Habbie> their service is so well thought out
2023-08-09 17:46:30 <Habbie> but, how is it failing?
2023-08-09 17:50:31 <oleovit> its simply unable to resolve dns.nexdns.io 
2023-08-09 17:51:08 <oleovit> logs: trying to resolve .. unable .. ad infinitum
2023-08-09 17:51:43 <Habbie> i have several questions
2023-08-09 17:51:50 <Habbie> one is, why is your paste url suddenly something entirely different
2023-08-09 17:52:22 <oleovit> i deleted it, because i forgot to delete my id
2023-08-09 17:52:49 <oleovit> i can upload a new one if you like
2023-08-09 17:53:41 <Habbie> ah
2023-08-09 17:53:48 <Habbie> it's funny to me that 0x0.st apparently reuses IDs
2023-08-09 17:53:52 <Habbie> but, yes, please upload it again
2023-08-09 17:54:17 <oleovit> https://0x0.st/H_fo.txt
2023-08-09 17:54:36 <Habbie> what part of this is fake?
2023-08-09 17:55:31 <oleovit> "id" and stamp to xxxxx (last one was maybe not necssary to edit=
2023-08-09 17:56:12 <Habbie> ok, so dnscrypt-proxy is supposed to know that NextDNS- means something?
2023-08-09 17:56:16 <Habbie> and
2023-08-09 17:56:21 <Habbie> you say unable to resolve dns.nexdns.io
2023-08-09 17:56:24 <Habbie> which looks wrong
2023-08-09 17:56:26 <Habbie> can you pastebin that bit?
2023-08-09 17:57:15 <oleovit> moment
2023-08-09 18:04:27 <oleovit> note: it can't use bootstrap resolvers because if fw-rules. but, like with any other public doh server, it should be able to use the system-dns, which is dnscrypt-proxy itself (resolvconf 127.0.0.1) : https://0x0.st/H_f8.txt
2023-08-09 18:04:58 <Habbie> uh
2023-08-09 18:05:12 <Habbie> oh your lines got cut off
2023-08-09 18:06:02 <oleovit> its always the same, thats why i said, ad infinitum.
2023-08-09 18:06:55 <oleovit> only change to another resolver and restart dnscrypt-proxy stopped it
2023-08-09 18:07:21 <Habbie> ok yes
2023-08-09 18:07:31 <Habbie> i believe the lines repeat
2023-08-09 18:07:34 <Habbie> but the lines themselves are cut off
2023-08-09 18:07:41 <oleovit> ??
2023-08-09 18:07:46 <Habbie> write: op>
2023-08-09 18:07:48 <Habbie> that's a truncated line
2023-08-09 18:07:51 <Habbie> (second line in your paste)
2023-08-09 18:08:03 <Habbie> anyway, given what you say about fw-rules i can guess what it says
2023-08-09 18:08:04 <oleovit> ahh, shit
2023-08-09 18:08:09 <oleovit> moment
2023-08-09 18:08:14 <Habbie> and from what i gather, you've made bootstrapping impossible
2023-08-09 18:10:31 <oleovit> doesnt matter i guess. works with other resolvers too, without fallback to bootstrao resolvers
2023-08-09 18:10:43 <oleovit> i upload the file once again
2023-08-09 18:10:46 <Habbie> which others?
2023-08-09 18:12:28 <oleovit> every other doh resolver  https://github.com/DNSCrypt/dnscrypt-resolvers/blob/master/v3/public-resolvers.md
2023-08-09 18:12:51 <Habbie> every? or just a few you tried?
2023-08-09 18:12:59 <Habbie> wonder if the others are configured by IP instead of by name
2023-08-09 18:15:39 <oleovit> at first run, with server_names commented, well, yeah, a bunch of them. cloudflare, nextdns, quad9, frank deni's resolvers and a lot otf others. 
2023-08-09 18:16:20 <oleovit> also nextdns works, but not in the static context
2023-08-09 18:16:22 <Habbie> nextdns works, just not nextdns-yourid?
2023-08-09 18:16:44 <oleovit> we thought the same
2023-08-09 18:16:53 <Habbie> ok
2023-08-09 18:17:05 <Habbie> my guess is still that all those that work, have IPs specified somewhere in dnscrypt-proxy
2023-08-09 18:17:08 <Habbie> and -id does not
2023-08-09 18:17:10 <oleovit> static doesnt work, even adding
2023-08-09 18:17:16 <oleovit> it manually
2023-08-09 18:18:34 <oleovit> yes, "nextdns", from the posted list, which gets downloaed every 24 h works, but not the static one with id ans stamp
2023-08-09 18:18:49 <Habbie> ack
2023-08-09 18:19:26 <oleovit> something is missing in alpines version, i guess. its not the toml alone#
2023-08-09 18:19:49 <Habbie> are you sure it works in non-alpine versions?
2023-08-09 18:21:40 <oleovit> that, i don't know. i guess. at least, no one complained about the missing static section in the toml on the dnscrypt.p. git. its not there at least on alpine and there is maybe a reason for it
2023-08-09 18:21:57 <oleovit> sry, my en.
2023-08-09 18:21:57 <Habbie> i think we're confusing things
2023-08-09 18:22:16 <Habbie> i'm asking, is nextdns-id working on non-alpine *with your fw-rules*?
2023-08-09 18:22:24 <Habbie> in other words, is *alpine* the difference that matters?
2023-08-09 18:23:31 <oleovit> well, lets change the fw and let the bootstrap resolvers come into play. 5 min please
2023-08-09 18:33:16 <oleovit> oh no. it worked by changing the fw. it needed the bootstrap resolvers in this case, because it had nothing else, in this case .. . well, sorry alpine, it was again my fault 
2023-08-09 18:33:48 <Habbie> in the end we figured it out
2023-08-09 18:33:51 <Habbie> that's what matters to me
2023-08-09 18:35:39 <oleovit> yes, thank you habbie for your patience/help. its my choice now. use this stuff, and "weaken" the fw, or go back to resolvers fetched over guthub.
2023-08-09 18:36:00 <Habbie> or see if you can bootstrap via basic nextdns?
2023-08-09 18:36:24 <Habbie> and, you are very welcome
2023-08-09 18:38:45 <oleovit> i have to play with every option. but anyway, maybe adding the statix-stuff into the toml on alpines side whould be a good idea (thats why i thought, there was some reason for this)
2023-08-09 18:39:19 <oleovit> my en. ... when i have to type, sry
2023-08-09 18:39:35 <Habbie> right
2023-08-09 18:39:46 <Habbie> so if you want the alpine toml to be updated, asking in here likely won't make that happen
2023-08-09 18:40:08 <Habbie> sending a merge request is the best way
2023-08-09 18:41:17 <oleovit> yep, will do that tomorrow. have a great day/night habbie (never had a chat with you)
2023-08-09 18:41:51 <Habbie> good luck, and thank you. from my perspective, have a good evening/night :)
2023-08-09 18:41:59 <oleovit> ps: the merge ewquest will be in english, not this .. here
2023-08-09 18:42:11 <Habbie> your english seemed fine to me
2023-08-09 18:42:23 <Habbie> don't worry too much about it
2023-08-09 18:43:44 <oleovit> i woory about anything  right now. i cant think straight anymore since weeks (no sleep). i even cant remember some regexp for my postinstall. its horrible
2023-08-09 18:44:59 <oleovit> mmaybe, i'm getting to dumb for everything. that's why i need i break. ok , /part
2023-08-09 20:39:42 <telmich> Currently diffing ubuntu's kernel config vs. 6.1.38 from alpine - not having sound drives me nuts
2023-08-09 20:41:08 <telmich> fun fact, ubuntu's kernel has the ISA bus still enabled :-)
2023-08-09 20:41:21 <ikke> lol
2023-08-09 20:41:36 <telmich> and quite a lot of network stuff that is loaded as a module, ubuntu uses compile in
2023-08-09 20:41:52 <telmich> Not sure what is smarter actually
2023-08-09 20:42:35 <telmich> parallel IDE is available as a module in ubuntu - also haven't seen that for a while
2023-08-09 20:43:37 <telmich> Does anyone remember the funky issues with chromium and i915 based xorg?
2023-08-09 20:43:58 <telmich> On the new notebook I have, chromium does actually not render anything, besides when I resize the window
2023-08-09 20:44:21 <telmich> I remember having had a similar issue like that on an lenovo x1 nano or x1 carbon some years ago
2023-08-09 20:44:58 <telmich> (and I wonder if that is actually related to my sound issues, as it seems that the soundwire driver is somewhat related to the i915 driver)
2023-08-09 20:46:04 <ikke> No, I only a recent issue where there were rendering issues due to invalid cache invalidation
2023-08-09 20:50:53 <telmich> Found something
2023-08-09 20:51:07 <telmich> ubuntu has https://cateee.net/lkddb/web-lkddb/DRM_I915_PXP.html set, we don't in LTS
2023-08-09 20:53:50 <Nulo[m]> how do i make alpine shutdown when shutting down via libvirtd?
2023-08-09 20:53:58 <Nulo[m]> i already enabled acpid..
2023-08-09 20:55:23 <Ermine> Nulo[m]: it doesn't while acpid is running?
2023-08-09 20:55:52 <Nulo[m]> Ermine: nvm acpid is crashing, let's see..
2023-08-09 20:57:07 <minimal> Nulo[m]: or use tiny-power-button kernel module instead of button and then no need for acpid
2023-08-09 20:57:45 <minimal> I don't use "button" and acpid on any VMs, I just load "tiny-power-button"
2023-08-09 21:00:57 <telmich> fun fact, I tried to boot the ubuntu kernel on alpine and... surprise, it did not boot
2023-08-09 21:01:10 <telmich> Main reason being that the decryption of root volumes is handled differently in ubuntu
2023-08-09 21:01:39 <Nulo[m]> it just says that /dev/input/event0 isn't a file, which it isn't...
2023-08-09 21:01:49 <Nulo[m]> acpid that is
2023-08-09 21:01:57 <minimal> Nulo[m]: what about my alternative suggestion?
2023-08-09 21:02:11 <Nulo[m]> tiny-power-button doesn't seem to do anything
2023-08-09 21:02:37 <Nulo[m]> am i supposed to do something special on the host side?
2023-08-09 21:03:09 <minimal> Nulo[m]: yes it does - it *only* emulates a ACPI power button so when libvirt triggers a VM poweroff tiny-power-button  acts upon it (all without needing any acpid running)
2023-08-09 21:03:32 <Ermine> Is there anything in /dev/input?
2023-08-09 21:03:46 <minimal> oh, I think tiny-power-button is blacklisted by default, so you'll probably need to edit the file in /etc/modprobe/ to remove the blacklist
2023-08-09 21:03:51 <Nulo[m]> Ermine: there's no /dev/input at all
2023-08-09 21:04:26 <Nulo[m]> minimal: okay that might be it
2023-08-09 21:04:29 <minimal> on the VM that is, there's nothing to do on the host
2023-08-09 21:04:36 <Nulo[m]> would be nice if modprobe told you that it was blacklisted
2023-08-09 21:04:38 <Ermine> Do you have mdev/mdevd/udev running?
2023-08-09 21:05:10 <minimal> Numo[m]: you probably want to blacklist button instead of tiny-power-button to avoid any situations where button gets loaded first
2023-08-09 21:05:33 <Nulo[m]> Ermine: yes, mdev
2023-08-09 21:05:59 <Ermine> is /etc/mdev.conf in place?
2023-08-09 21:06:04 <Nulo[m]> eh, even with only tiny-p* it doesn't work
2023-08-09 21:06:11 <Nulo[m]> Ermine: yes
2023-08-09 21:06:32 <minimal> Numo[m]: actually I just remembered you need to add "tiny_power_button.power_signal=12" to the cmdline as well
2023-08-09 21:06:42 <Ermine> It's super weird then
2023-08-09 21:07:12 <Ermine> Do you run x86_64 vm?
2023-08-09 21:07:19 <Nulo[m]> Ermine: yes
2023-08-09 21:10:45 <minimal> Nulo[m]: ^^^
2023-08-09 21:11:25 <Nulo[m]> yes, i guess i'll try that. it's just the way i "deploy" this vm needs me to regenerate an image and upload it and blah
2023-08-09 21:11:56 <minimal> Nulo[m]: I use tiny-cloud-button on a daily basis
2023-08-09 21:17:01 <minimal> Nulo[m]: https://github.com/dermotbradley/create-alpine-disk-image/blob/main/lib/disk-image-functions#L101
2023-08-09 21:17:24 <Nulo[m]> trying it now
2023-08-09 21:20:29 <Nulo[m]> that works, thanks!
2023-08-09 21:21:09 <minimal> all part of creating a minimal (sic) VM, saving memory used by acpid ;-)
2023-08-09 21:44:01 <Nulo[m]> behold, the javascript generated alpine system image https://gitea.nulo.in/Nulo/woodpecker-in-a-vm/src/branch/main/index.js
2023-08-09 21:47:27 <somercet1> why is javascript ending up in everything?
2023-08-09 21:49:10 <ikke> somercet1: https://www.destroyallsoftware.com/talks/the-birth-and-death-of-javascript
2023-08-09 22:51:21 <somercet1> ikke, that was a good speech, but it makes me think they're reinventing the AS/400, which was a no-MMU computer that relied on very well-vetted libraries to keep user jobs from overwriting each other.
2023-08-09 22:51:49 <somercet1> just add in a lang+VM, and you have this asm.js future.
2023-08-09 23:07:08 <pjs> what's the update cycle for packages in stable repo's? ie, go 1.21 was released yesterday, how long (generally) until that appears in the 3.18 apk mirrors? I know the maintainers need to submit it etc, but I assume there is a period of testing, etc.
2023-08-09 23:17:31 <pj> pjs: it will not appear in 3.18
2023-08-09 23:18:10 <pj> 1.21 is not a patch or security update so it will not be backported to stable releases
2023-08-09 23:18:55 <pjs> gotcha
2023-08-09 23:19:43 <pjs> pj: is it advisable to run edge? 
2023-08-09 23:20:32 <pj> depends
2023-08-09 23:24:30 <pjs> pj: sure.
2023-08-09 23:25:01 <pjs> is there any general release schedule (ie, every 6 months, every quarter, once a year, etc.) and is there any public timeline or anything? 
2023-08-09 23:25:27 <pjs> I'm just curious, I don't need new Go that bad but I'm fairly new to alpine so just want to learn a bit more about it's release cycle
2023-08-09 23:26:16 <minimal> pjs: look at the Releases page links from alpinelinux.org?
2023-08-09 23:28:37 <minimal> pjs: at the top that page
2023-08-09 23:30:51 <pjs> minimal: right, I did. more or less 6mo. Thanks
2023-08-10 02:17:31 <xgpt1> has anyone had any luck using the protonmail-bridge client software on alpine?
2023-08-10 02:48:31 <Quillith> I'm trying to increase virtual memory for emulating but the instructions I'm finding online use sysctl. How do I increase vm.max_map_count with openrc?
2023-08-10 02:52:56 <Quillith> In my sway config I put "set $vm.max_map_count 524288" but that doesn't seem to have helped
2023-08-10 02:57:25 <elly> in your sway config? like your window manager?
2023-08-10 02:57:57 <elly> have a look at /etc/sysctl.d/README
2023-08-10 02:58:03 <elly> there's a mechanism for configuring sysctls at boot time
2023-08-10 03:03:45 <Quillith> Thank you! Going to try that out 
2023-08-10 03:07:33 <zcrayfish> OOF. https://git.alpinelinux.org/aports/commit/community/plasma-desktop-meta?id=35e297896a31f9dde6fdaeca3e8246e353847a63 <-- this is being a royal pain!
2023-08-10 03:08:05 <zcrayfish> +		qt5-qtvirtualkeyboard 
2023-08-10 03:08:38 <zcrayfish> The virtual keyboard completely takes over sddm until it is dismissed.
2023-08-10 03:11:33 <zcrayfish> Seems the solution is to either get rid of the qt5-qtvirtualkeyboard dependency... OR 
2023-08-10 03:11:45 <zcrayfish> https://git.alpinelinux.org/aports/tree/community/plasma-desktop-meta/sddm.conf <-- a blank input method could be added.
2023-08-10 03:11:50 <zcrayfish> Either way should fix it
2023-08-10 03:12:06 <zcrayfish> will throw an issue in later if nobody else gets to it first. Family time right now :)
2023-08-10 03:12:42 <zcrayfish> https://askubuntu.com/questions/1293912/sddm-blocks-screen-with-on-screen-keyboard-in-ubuntu-20-04 <-- Looks like the Ubuntu guys had an almost identical issue at some point FWIW
2023-08-10 03:12:47 <zcrayfish> 
ACTION afk

2023-08-10 03:20:50 <invoked> xgpt: not that it answers your question, but you'd probably be better off building hydroxide. or maybe try the flatpak version of protonmail-bridge.
2023-08-10 04:39:24 <livy> Hi guys. I have trouble trying to unlock a BitLocker volume on Alpine.
2023-08-10 04:39:51 <livy> I use gnome-disk-utility. If I enter wrong passphrase, it complains as expected.
2023-08-10 04:40:28 <livy> But if I enter correct passphrase, it ends up with an error: Error unlocking /dev/sdd2: Failed to activate device: Not supported (udisks-error-quark, 0)
2023-08-10 04:40:44 <livy> Do I need to install additional packages for that functionality?
2023-08-10 04:46:37 <grawity> livy: that depends on what you get when trying to unlock the same volume with `cryptsetup open` from CLI
2023-08-10 04:46:52 <grawity> not all BitLocker volumes are supported by cryptsetup (afaik there are some things that dm-crypt cannot fully express)
2023-08-10 04:47:14 <grawity> sometimes you need to fall back to libbde or dislocker
2023-08-10 04:47:26 <livy> Several months ago I used Ubuntu, and gnome-disk-utility was able to decrypt that same drive.
2023-08-10 04:47:46 <livy> I do not know which tool does gnome-disk-utility use under the hood though. cryptsetup?
2023-08-10 04:50:45 <livy> Anyway, I tried with dmcrypt: doas cryptsetup open --type=bitlk -v /dev/sdd2 sdd2
2023-08-10 04:51:20 <livy> The result: Cannot activate device, kernel dm-zero module is missing.
2023-08-10 04:53:47 <ikke> livy: is this on boot (rootf) or on a running system?
2023-08-10 04:54:08 <livy> I am trying to get some data from an external hard drive.
2023-08-10 04:54:32 <livy> I am on a Alpine Linux PC, and the hard drive comes from a Windows PC.
2023-08-10 04:55:10 <ikke> what does `ls -ld /lib/modules/$(uname -r)` return?
2023-08-10 04:55:39 <livy> ls: /lib/modules/6.1.42-0-lts: No such file or directory
2023-08-10 04:56:36 <grawity> so generally a "you need to reboot after upgrading the kernel" problem
2023-08-10 04:57:04 <livy> Ah, I think I type 'apk upgrade' a few hours ago.
2023-08-10 04:57:26 <grawity> (you don't have on Ubuntu because they ship each kernel in a new package, so you don't lose the old kernel's modules after upgrade)
2023-08-10 04:58:32 <livy> I am on 6.1.42, while the kernel has been upgraded: apk info -v | grep linux-lts -> linux-lts-6.1.44-r0
2023-08-10 04:58:55 <ikke> yes, so it cannot find the modules for the running kernel
2023-08-10 04:59:53 <livy> This remind me of the incident yesterday, when I made a bootable USB from Extended Alpine ISO.
2023-08-10 05:00:25 <livy> Booting from the USB, I could load the zfs.ko module.
2023-08-10 05:00:52 <livy> But if I boot from CD again, and upgrade the kernel inside the USB: update-kernel "/media/usb/boot";
2023-08-10 05:01:14 <livy> After that I tried booting from USB, and was unable to load zfs.ko again.
2023-08-10 06:21:38 <grawity> hmm actually does alpine have any hack similar to arch's kernel-modules-hook? (a package manager hook that tries to keep old modules during a kernel upgrade)
2023-08-10 06:33:17 <ikke> grawity: not aware of any
2023-08-10 06:48:07 <dzilvys> Does alpine auto execute the `autorun` scripts on usbs by default?
2023-08-10 06:51:15 <ikke> No
2023-08-10 06:51:54 <ikke> (there might be desktop environments that do that though)
2023-08-10 06:52:43 <ikke> But I would not expect it, because it's not good security 
2023-08-10 07:02:29 <dzilvys> I am aware, I am just currently trying to figure out what the best approach would be to the following problem
2023-08-10 07:02:29 <dzilvys> I need to update from version A to version B. However they have mismatched grubs. The biggest issue that its trying to use vmlinuz-vanilla when only vmlinuz-lts exists. A only adjusts the UUID for the root partition. however, during this process, I mount the data partition of B. So I was wondering if there were a way to updating existing grub etc by executing an autoscript from B's root directory and once its fixed GRUB It can destroy tiself
2023-08-10 11:05:13 <Invarianz> Greetings, I have a bluetooth problem with any Bluetooth mouse and keyboard. I have written a detailed record of the problem under: https://unix.stackexchange.com/questions/751540/bluetooth-keyboard-mouse-issue-on-alpine-linux and followed the Alpine Linux bluetooth guide under https://wiki.alpinelinux.org/wiki/Bluetooth .
2023-08-10 11:05:13 <Invarianz> I guess that I'm missing some crucial package or setting either in the bluetooth or input stack. I'm using sway.
2023-08-10 11:05:13 <Invarianz> TLDR: Bluetooth mouse and keyboard connect via bluetoothctl and show good connection status. However no input events register, neither keyboard keys nor mouse movement or clicks.
2023-08-10 11:06:31 <Invarianz> As an addon: Both bluetooth devices work under Ubuntu and Windows. The only bluetooth device that works flawlessly under Alpine are my bluetooth headphones.
2023-08-10 11:52:16 <ikke> grawity: oh, there is the kernel-hooks package apparently
2023-08-10 14:02:14 <oleovit> hi. again. https://gitlab.alpinelinux.org/alpine/aports/-/issues/15183#note_329072 : kevin suggested i could make a merge request (even though gitlab is and has always been completely alien to me). see screenshot: https://0x0.st/H_JB.png . i think i'm on the right page. but no "new merge request" button. maybe i'm blind. never done that :(. only if someone has time.. learning by doing. even if tired.
2023-08-10 14:04:58 <ikke> oleovit: You start with making a fork of aports
2023-08-10 14:05:16 <ikke> https://gitlab.alpinelinux.org/alpine/aports on the top right, there is a fork button
2023-08-10 14:05:16 <minimal> grawity: keeping existing kernel modules after kernel package upgrade has been discussed several times before including here: https://gitlab.alpinelinux.org/alpine/aports/-/issues/13928
2023-08-10 14:05:44 <ikke> I think the challenge has been making sure they are also eventually cleaned up again
2023-08-10 14:06:34 <pj> the problem is that it depends on apk-tools and hardly anyone can touch C
2023-08-10 14:07:25 <oleovit> @ikee thx, i see...
2023-08-10 14:08:30 <ikke> pj: I think you can do it without touching apk-tools
2023-08-10 14:09:23 <ikke> except for cleaning up older kernels that are no longer in use
2023-08-10 14:10:13 <pj> ncopa said he wants to have it in apk-tools
2023-08-10 14:10:31 <pj> https://gitlab.alpinelinux.org/alpine/apk-tools/-/issues/10887#note_300634
2023-08-10 14:10:55 <pj> if you can do it without and better, go ahead :)
2023-08-10 14:11:08 <ikke> Well, I'd use hardlinks instead of copying it
2023-08-10 14:11:29 <ikke> Then you don't have to duplicate the data
2023-08-10 14:12:09 <pj> hardlinks don't work cross fs
2023-08-10 14:14:00 <ikke> I'm aware
2023-08-10 14:14:20 <ikke> but if you copy it in /lib/modules, I'd not expect it to cross any filesystem
2023-08-10 14:15:03 <ikke> time cp -lr /lib/modules/$(uname -r) /lib/modules/linux-prev
2023-08-10 14:15:44 <pj> yeah, but I expect it's not going to be a feature specific just to kernel package right?
2023-08-10 14:16:13 <ikke> this is specifically about the kernel modules
2023-08-10 14:16:17 <ikke> the feature request
2023-08-10 14:16:36 <minimal> I think (almost) everyone agreed that there's an issue, its just that no solution has yet been decided upon/agreed and implemented and so the issue will come up again and again
2023-08-10 14:16:56 <pj> The feature request is about keeping previous package version
2023-08-10 14:16:56 <pj> with kernel being an example
2023-08-10 14:17:30 <pj> "Any package could be flagged with 'keep-old' but I think it mostly makes sense for linux kernels."
2023-08-10 14:18:06 <pj> Sure, we're trying to solve the kernel package issue but we also have to do it in a way that doesn't break anything else
2023-08-10 14:18:16 <minimal> I think it's at a point where BDFL needs to pick a solution and get it implemented
2023-08-10 14:19:55 <ikke> I think we should not let perfect be the enemy of good
2023-08-10 14:20:08 <ikke> for most packages it would not make sense to keep a backup
2023-08-10 14:20:38 <ikke> at least, not installed on the system
2023-08-10 14:21:13 <ikke> the kernel lives in a versioned directory, so you can have multiple copies installed at the same time
2023-08-10 14:21:20 <ikke> kernel modules*
2023-08-10 14:30:00 <oleovit> ikke: sry to interrupt; forked. but i'm lost here: https://gitlab.alpinelinux.org/oleovit/aports/-/merge_requests/new
2023-08-10 14:30:35 <ikke> To create a merge request, you first need a branch with the changes
2023-08-10 14:30:57 <ikke> (You _could_ do it all via the webinterface, but that's rather cumbersom)
2023-08-10 14:31:40 <ikke> So on the project overview page of your fork, there should be a blue 'clone' button
2023-08-10 14:34:40 <oleovit> yep, i see. do i need to git clone my fork, et cetera?
2023-08-10 14:34:48 <ikke> yes
2023-08-10 14:35:06 <ikke> after cloning, you create a new branch to work on
2023-08-10 14:38:20 <oleovit> i'm not afraid in my own little git, but this gives me the shivers.. let's clone this thing. 
2023-08-10 14:38:48 <ikke> oleovit: I'll talk you through it
2023-08-10 14:45:14 <oleovit> thx, ok, download of course 2fa was on, no ssh keys, so 2fa temp. off, https clone..
2023-08-10 14:45:36 <ikke> oleovit: you can generate a personal access token instead
2023-08-10 14:45:43 <ikke> then you don't need to disable 2fa
2023-08-10 14:46:35 <oleovit> its done, download already
2023-08-10 14:48:44 <aron> is there any guide how to install alpine onto RK3308 (i.e. rock pi s)?
2023-08-10 14:50:29 <ikke> aron: if anywhere, it would be somewhere on the wiki
2023-08-10 14:56:41 <oleovit> @ikke done. cd into community, or  git checkout -b branchname .. it's i lot in there..
2023-08-10 14:57:20 <oleovit> "a lot"
2023-08-10 14:57:56 <ikke> yeah, lot of packages. cd to dnscrypt-proxy
2023-08-10 14:58:07 <oleovit> moment
2023-08-10 14:58:33 <oleovit> ah, yes
2023-08-10 14:58:55 <ikke> In that dir is the toml file
2023-08-10 14:59:00 <oleovit> yep
2023-08-10 14:59:16 <ikke> are you doing this on an alpine linux system?
2023-08-10 14:59:23 <oleovit> yep
2023-08-10 15:00:12 <oleovit> sway. so, i have to change the .toml now
2023-08-10 15:01:02 <aron> ikke: I found the related wiki page, but misses a few steps (i.e. how to acquire u-boot binaries, the bl31 stuff, etc) -- anyway, I'll keep looking
2023-08-10 15:15:42 <oleovit> @ikke i also see an incomplete descriction of https://github.com/DNSCrypt/dnscrypt-proxy/blob/63e414021b6892a25733b09ee00f0d4b07cba651/dnscrypt-proxy/example-dnscrypt-proxy.toml#L850  in alpine toml (ends in the middle of the sentence. should i complete this too? the mentioned urls and the static option is added. 
2023-08-10 15:16:15 <oleovit> ..resp. changed
2023-08-10 15:16:27 <ikke> oleovit: One thing I'm wondering about is why this file is copied in aports in the first place. Are there some different settings?
2023-08-10 15:22:48 <oleovit> i did a lot of diffs lately, yes, but only in the comments          
2023-08-10 15:23:28 <ikke> Only comments are different?
2023-08-10 15:23:39 <ikke> Then in that case, it would make sense to just remove it and install if from upstream
2023-08-10 15:24:25 <ikke> history of that file doesn't tell why it was added
2023-08-10 15:27:50 <oleovit> well, maybe this was intentional. the .toml that came with the recent version changed for example: -## See the `/usr/share/dnscrypt-proxy/example-cloaking-rules.txt` file for an example -> older .toml to: +## See the `/usr/share/doc/dnscrypt-proxy/example-cloaking-rules.txt` file for an example
2023-08-10 15:28:15 <oleovit> see the difference: /doc/
2023-08-10 15:28:20 <ikke> oleovit: that would be better done as a patch
2023-08-10 15:28:32 <ikke> than we don't need to keep everything in sync
2023-08-10 15:30:35 <telmich> I was wondering, what is the fastest/easiest way to rebuild the -lts kernel with a new config option? Probably setting up abuild, modifying the linux-lts aport and then installing?
2023-08-10 15:31:55 <ikke> yes, that would be it
2023-08-10 15:32:09 <ikke> potentially using docker-abuild
2023-08-10 15:32:26 <oleovit> maybe, less work for all of you. the only additional differences in a diff was on some lines additional space. thats it. 
2023-08-10 15:33:19 <ikke> yeah, then a patch is definitely better
2023-08-10 15:35:03 <telmich> I want to test my theory that CONFIG_DRM_I915_PXP=y is all that is needed for the dell xps 9320
2023-08-10 15:35:08 <ikke> Yeah, I figured
2023-08-10 15:36:30 <oleovit> then, i guess, i can stop my adventure here ;)?
2023-08-10 15:37:11 <ikke> oleovit: If you want, you can make _that_ change
2023-08-10 15:37:15 <ikke> that would help you as well, right?
2023-08-10 15:39:27 <telmich> Did make at some point changed its defaults to build in parallel by default? I see that the kernel build is run using make ARCH=x86_64 CC=gcc AWK=mawk KBUILD_BUILD_VERSION=2-Alpine, but it seems to have spawned around 16 subprocesses
2023-08-10 15:40:28 <oleovit> ..uff. but no clue how... a patch
2023-08-10 15:42:37 <oleovit> but yes, it could help to understand, what i'm afraid of *lol*
2023-08-10 15:53:07 <oleovit> ah, you meant, i still should do _that_ change, right now, what i'm working on?
2023-08-10 16:00:03 <ikke> telmich: JOBS=$(nproc) was added, yes
2023-08-10 16:02:58 <oleovit> but there are also more important differences i saw now: alpine ->  cache_file = '/var/cache/dnscrypt-proxy/public-resolvers.md' versu-s cache_file = 'public-resolvers.md'
2023-08-10 16:05:02 <ikke> those can be part of the patch
2023-08-10 16:06:30 <oleovit> thats cool. so, what should i do now ikke? 
2023-08-10 16:10:41 <oleovit> waiting/patch or merge request of these minor changes. don't know if this makes sense now...  
2023-08-10 16:11:15 <oleovit> need coffee
2023-08-10 16:13:24 <Invarianz> Just wanted to mention that I resolved my Bluetooth issues regarding mouse and keyboard. I was missing the kernel module `uhid` from the Alpine package `hidapi`.
2023-08-10 16:13:41 <Invarianz> Now not only Bluetooth connects but also mouse and keyboard keys register.
2023-08-10 16:18:24 <ikke> Invarianz: cool
2023-08-10 16:30:43 <Invarianz> I know that by now its a trope but are there any other people here that switched from Archlinux and are super happy with the much more lightweight focus, the speed of apk and focus on the BusyBox userland?
2023-08-10 16:30:43 <Invarianz> I always find myself slimming things down and finding that the user experience somehow gets even better instead of worse! 
2023-08-10 16:31:44 <oleovit> @ikke, i will make the merge request. so, now, a simple git add + commit at fork and then .. the "merge request" button?
2023-08-10 16:41:15 <ikke> oleovit: commit, push to your fork
2023-08-10 16:41:27 <ikke> and then you will get a url that will get you to the merge request page
2023-08-10 16:42:59 <oleovit> is commit message ""community/dnscrypt-proxy: change .toml" ok? i dont know alpines guidlines
2023-08-10 16:47:04 <ikke> Did you only add the missing config there?
2023-08-10 16:47:29 <ikke> oleovit: as a guide: don't tell what changed, but why
2023-08-10 16:47:52 <ikke> (simple changes are exceptions)
2023-08-10 16:48:09 <ikke> In this case, just stating that you changed the file does not tell much
2023-08-10 16:48:20 <ikke> (I can see that in the diff 😉)
2023-08-10 16:48:28 <oleovit> i added the missing config and deletetd the obsolete, non-functional dnscrypt.net url
2023-08-10 16:49:12 <oleovit> lets rethink the message, moment ;)
2023-08-10 16:51:28 <frag> bought an old keychron K12 a while back, works well, can be switched to mac or pc, but on pc Fn buttons dont work, and on mac win and alt changes places ... is this a common thing that keyboard are not supported?
2023-08-10 16:51:56 <frag> find it quite strange ...
2023-08-10 17:00:09 <telmich> ok, we must be doing something "wrong" / "different" on alpine when it comes to soundwire support
2023-08-10 17:00:30 <telmich> I just took the 5.19.0 kernel from ubuntu, copied over all modules
2023-08-10 17:01:00 <telmich> First I tried booting the kernel + the original ubuntu initrd, which failed due to not loading the right script for decrypting the rootfs
2023-08-10 17:01:12 <telmich> Then I regenerated the initramfs using mkinitfs, with all modules present
2023-08-10 17:01:27 <telmich> rebooted, disk decryption works, but the soundcard does not appear
2023-08-10 17:05:29 <oleovit> @ikke we are nearly done, thx for patience. message is clear now i guess. bcause i have no clue: git push ... ? i only know what i have to do on my own little git in this case
2023-08-10 17:07:34 <oleovit> .. -u origin aborts ? no i guess
2023-08-10 17:13:07 <oleovit> git push -u origin master of course. 
2023-08-10 17:13:59 <oleovit> and url to fork: https://gitlab.alpinelinux.org/oleovit/aports.git
2023-08-10 17:18:35 <ikke> oleovit: you want to push the branch that you created
2023-08-10 17:18:46 <oleovit> now i'm at: https://gitlab.alpinelinux.org/oleovit/aports/-/merge_requests/new - and i see my commit.
2023-08-10 17:19:17 <oleovit> target branch?
2023-08-10 17:20:51 <oleovit> https://0x0.st/H_yp.png
2023-08-10 17:23:32 <oleovit> alpine/aborts ..?
2023-08-10 17:25:56 <oleovit> of course i guess. but: Select source branch  .. mystery
2023-08-10 17:26:32 <WhyNotHugo> frag: Isn't this because of the layout? Try Fn+F1 intead of F1.
2023-08-10 17:28:55 <oleovit> 3.18 stable...
2023-08-10 17:30:11 <ikke> oleovit: did you create a local branch?
2023-08-10 17:34:13 <oleovit> oh my, now i'm confused. i'm here now: https://0x0.st/H_yw.png
2023-08-10 17:35:44 <ikke> Sorry, 0x0.st is not loading for me for some reason
2023-08-10 17:37:04 <oleovit> strange. title: New merge requestFrom oleovit/aports:3.18-stable into alpine/aports:master
2023-08-10 17:38:43 <oleovit> with busybox (i hope its only an example) entries in title and description
2023-08-10 17:39:08 <frag> WhyNotHugo: no, tried that :] 
2023-08-10 17:39:12 <frag> brb btw
2023-08-10 17:44:36 <oleovit> ah, i've got it. don't worry, i did nothing on the page i landed
2023-08-10 18:09:11 <oleovit> @ikke please tell me i did it right (title is too long, but anyway): https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/49860
2023-08-10 18:10:00 <oleovit> i see: Merge blocked: the source branch must be rebased onto the target branch.
2023-08-10 18:11:19 <ikke> You can ignore that, we take care of that
2023-08-10 18:12:54 <oleovit> my first advanced public thing ever. thx ikke! what to do now with my opend issue?
2023-08-10 18:15:20 <oleovit> + last question: is the title too long. after that, i know how it works. learning by doing
2023-08-10 18:16:14 <ikke> oleovit: yeah, it's a bit too long
2023-08-10 18:16:59 <ikke> I'd say something like: "community/dnscrypt-proxy: add missing configuration", and then in the body explain more (body is one line seperated from the title)
2023-08-10 18:23:26 <oleovit> @ikke done. ragarding the patch: plz keep that in mind. thank you!
2023-08-10 18:23:37 <oleovit> "regarding"
2023-08-10 18:33:26 <ikke> note that you also need to change the commit mesage, but I can do that for you if you want to
2023-08-10 18:35:51 <frag> ratpoison has a function to copy to x-selection and output x-selection, xsel and browser can also output what the ratpoison function put into x-selection .. but middle click and terminal outputs nothing ...
2023-08-10 18:35:51 <oleovit> can i do a git rebase -i HEAD~x, then r(eword) and then a git push force?
2023-08-10 18:36:52 <frag> ah, actually it seems to be a termianl problem only, nvm
2023-08-10 18:41:35 <ikke> oleovit: yes, or git commit --amend is even easier with just one commit
2023-08-10 19:01:44 <witcher> I'm unable to get avahi to resolve mDNS hostnames. How do I enable NSS support for mDNS?
2023-08-10 19:03:36 <oleovit> @ikke sry.. its hours since i nerve.. but i can't git push (force with lease), not allowed on protected branch.
2023-08-10 19:05:57 <oleovit> gitlab.. maybe a setting
2023-08-10 19:10:05 <oleovit> done. 
2023-08-10 19:40:43 <dwfreed> witcher: musl doesn't have NSS
2023-08-10 19:40:50 <dwfreed> that is a glibc thing
2023-08-10 19:41:07 <dwfreed> there is https://github.com/LouisBrunner/avahi2dns as one tool you could use instead
2023-08-10 19:41:56 <dwfreed> however, it doesn't appear to have any way to forward non mdns queries to a regular resolver ?
2023-08-10 19:42:09 <witcher> dwfreed: thanks, I'll have a look!
2023-08-10 19:43:28 <dwfreed> you could put something like dnsdist in front of it and direct the usual mDNS TLDs to that and the rest to your normal resolver
2023-08-10 19:43:53 <dwfreed> that's a lot of convolution just for mdns, though
2023-08-10 19:49:41 <frag> damn, ratpoison is the most interesting wm ive tried ever ...
2023-08-10 19:52:43 <frag> it has run dialog inbuild, notification and even prompt to take input ..
2023-08-10 19:53:07 <frag> many other tools inbuilt that xdotool etc do too..
2023-08-10 23:25:56 <pjs> is there an extra package to install to enable iptables? I installed just iptables but it keeps saying the module doesn't exist. Maybe I need to reinstall the kernel but figured I'd ask
2023-08-10 23:26:39 <quinq> Are you running the same kernel as the one currently installed?
2023-08-10 23:27:20 <quinq> ls -ld /lib/modules/$(uname -r)
2023-08-10 23:37:51 <pjs> quinq: ah, I think I just need a reboot. I ran an update and it updated the kernel
2023-08-10 23:38:12 <pjs> sorry, I should have thought of that smh
2023-08-10 23:38:13 <pjs> thanks!
2023-08-11 01:58:44 <xgpt> why do i get such strong "BSD'ish" vibes from alpine?
2023-08-11 01:59:01 <xgpt> serioulsy, I cannot get over how wonderfully minimal this distro is
2023-08-11 03:47:45 <elly> xgpt: not sure, perhaps it's musl / busybox that makes you feel that way?
2023-08-11 04:22:01 <grawity> ikke: the one available on arch cleans up old modules both on startup and on a cronjob https://github.com/saber-nyan/kernel-modules-hook -- not an "official" solution but it works well enough
2023-08-11 05:38:47 <skylightPenguin> Hi, I'm having trouble setting up my x server. my startx seems to be refusing to see my .xinitrc for some reason
2023-08-11 05:40:10 <ikke> refusing in what way?
2023-08-11 05:40:23 <skylightPenguin> it seems to defaultto the /etc xinitrc
2023-08-11 05:41:11 <skylightPenguin> (also I'm running sic on the tty and it is a bit jank HAHAHA)
2023-08-11 05:44:29 <skylightPenguin> but yeah I'm kinda running out of ideas on how to get that to work
2023-08-11 05:44:53 <skylightPenguin> and even after that, like    i3wm doesn't wanna spawn a terminal anyway
2023-08-11 05:45:16 <skylightPenguin> like when I manually  startx /usr/bin/i3wm
2023-08-11 05:46:33 <skylightPenguin> not sure if it's permissions issue or something
2023-08-11 05:52:03 <skylightPenguin> huh, so I can't put sic on a background tty for too long
2023-08-11 06:01:59 <skylightPenguin> one thing I've found is that when I do  doas startx, it believes it is root
2023-08-11 06:02:36 <skylightPenguin> so putting the .xinitrc in /root works... which is not what I want
2023-08-11 06:04:05 <skylightPenguin> omg nvm
2023-08-11 08:17:49 <LACampbell> I love alpine as a desktop but I just can't get flatpak stuff to work reliably. has anyone experimented with archlinux in a chroot or something to get packages not available in apk?
2023-08-11 09:20:35 <caskd[m]> I run a arch chroot in bwrap to do games and stuff that rely on glibc
2023-08-11 09:20:57 <caskd[m]> it works okay but can be a hassle to get right
2023-08-11 09:39:25 <LACampbell> no luck with glib compat thing?
2023-08-11 09:39:42 <LACampbell> also, seem to have fixed flatpak issues. added it to the wiki so hopefully future people won't suffer as I have
2023-08-11 09:42:07 <caskd[m]> glib compat is a shim that doesn't add any glib specific functionality but only translates
2023-08-11 09:42:20 <caskd[m]> so you'll still hit missing symbol errors here and there
2023-08-11 09:42:42 <caskd[m]> (feel free to correct me if i'm wrong though)
2023-08-11 09:43:40 <LACampbell> ah interesting. i just used it to run sublime text but htat's probably doing less stuff than games
2023-08-11 09:44:11 <caskd[m]> yeah it's common for games to use glib stuff and steam is notorious for doing that too
2023-08-11 09:45:00 <caskd[m]> bwrap seems to be reliable enough as long as you don't need any capabilities (elevated priviledges)
2023-08-11 09:45:26 <LACampbell> I'm running steam via flatpak pretty comfortably
2023-08-11 09:46:13 <caskd> i do fancy stuff with overlayfs so i kinda need to do bwrap
2023-08-11 09:47:34 <caskd> https://0x0.st/H_x0.txt
2023-08-11 10:25:06 <boomanaiden154> Steam runs Chromium under the hood and Chromium does some extremely cursed things with the libc due to weird quirks.
2023-08-11 10:53:15 <nnn_n> why adb doesnt show my device in adb devices list 
2023-08-11 10:53:20 <nnn_n> ?
2023-08-11 10:55:23 <ikke> possibly because you don't have permissions as your user?
2023-08-11 10:56:35 <nnn_n> am enabled usb debugging in my phone settings
2023-08-11 10:56:43 <nnn_n> connected via usb
2023-08-11 10:57:02 <nnn_n> mb i need some package?
2023-08-11 10:57:34 <ikke> Does it work if you run adb as root?
2023-08-11 10:59:07 <nnn_n> ok via root all work
2023-08-11 11:01:39 <ikke> Then you need some udev rule that would give you access 
2023-08-11 11:07:14 <zcrayfish> https://pkgs.alpinelinux.org/package/edge/community/x86_64/android-udev-rules 
2023-08-11 11:07:20 <zcrayfish> Perhaps those?
2023-08-11 11:07:33 <nnn_n> but from root all work
2023-08-11 11:07:43 <nnn_n> i already did what i want
2023-08-11 11:54:03 <ayushnix> Does OpenRC have support for running multiple instances of a service using the same initscript?
2023-08-11 11:54:07 <ayushnix> Like how systemd template service units can be written as "user@.service" and "user@1000.service" runs that service for the user with UID 1000.
2023-08-11 11:57:00 <ncopa> ayushnix: i dont think it does
2023-08-11 11:57:27 <ncopa> there are some openrc scripts that are symlinks, which sort of is multiple instances of a service
2023-08-11 12:00:15 <ayushnix> ncopa: so I should create a symlink to the original init.d script and create different files in conf.d for different symlinks?
2023-08-11 12:01:43 <ncopa> yup
2023-08-11 12:02:03 <ncopa> you may need to make the original init.d script aware of it
2023-08-11 12:02:40 <ayushnix> ncopa: how do I do that? can you point me to an example init script with symlinks?
2023-08-11 12:02:57 <ncopa> openvpn
2023-08-11 12:03:09 <ncopa> dnsmasq.lxcbr0
2023-08-11 12:04:00 <ncopa> https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/main/openvpn/openvpn.initd#L9
2023-08-11 12:04:16 <ncopa> you can create a symlink openvpn.myconf
2023-08-11 12:05:10 <ncopa> which will start an instance of openvpn using /etc/openvpn/myconf.conf
2023-08-11 12:05:45 <ayushnix> ncopa: thanks, this should help
2023-08-11 12:06:55 <ayushnix> I noticed that the package pounce doesn't come with an init script in the Alpine packages. Can/Should I contribute an initscript for it?
2023-08-11 12:07:07 <ayushnix> https://git.alpinelinux.org/aports/tree/testing/pounce
2023-08-11 12:07:11 <ncopa> dnsmasq symlink will create a bridge and run a dnsmasq instance. https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/main/dnsmasq/dnsmasq.initd#L26
2023-08-11 12:08:35 <ncopa> sure
2023-08-11 12:08:55 <ncopa> you might even take maintainership. Not sure how active TBK is nowadays
2023-08-11 12:09:43 <ayushnix> Not sure if I'm qualified or ready enough for that. I'm quite new to Alpine.
2023-08-11 12:10:04 <ncopa> ok
2023-08-11 12:10:09 <ayushnix> But I'll definitely consider it.
2023-08-11 12:10:38 <ikke> Haven't seen much from TBK recently 
2023-08-11 12:10:43 <ayushnix> thanks for your help though
2023-08-11 12:14:08 <ncopa> np
2023-08-11 13:21:09 <raspbeguy> Hello, I was wondering if any help was needed on the infrastructure maintainance. If yes, would be glad to help.
2023-08-11 13:26:48 <ikke> raspbeguy: we could use help, yes
2023-08-11 13:27:08 <ikke> Feel free to hop on #alpine-infra
2023-08-11 13:27:18 <raspbeguy> will do
2023-08-11 16:40:10 <Apachez> which ssl stack does alpine use nowadays by default?
2023-08-11 16:53:49 <Saijin_Naib[m]> I believe setup-alpine still uses openssl  or did as of 3.18.0 ISO 
2023-08-11 17:22:35 <nangel> +1  There was a time we used libressl, but AFAIK, it is back to openssl
2023-08-11 17:23:30 <XFaCiEer> nangel: that was void no?
2023-08-11 17:23:57 <nangel> void?
2023-08-11 17:24:09 <XFaCiEer> Void Linux
2023-08-11 17:25:16 <nangel> No.  Alpine went with Libre for a bit... but I think it was on "edge" - then stuck with openssl
2023-08-11 17:38:46 <Apachez> yeah I saw the proposal of returning to libressl last year but wasnt sure of the status for that
2023-08-11 17:43:58 <XFaCiEer> Apachez: There was a vote to have a future vote that would determine whether a future vote would happen.
2023-08-11 18:38:45 <Apachez> amagad
2023-08-11 18:39:01 <Apachez> I found this https://gitlab.alpinelinux.org/alpine/tsc/-/issues/28
2023-08-11 18:53:20 <WhyNotHugo> If /run is not listed in fstab, then who is mounting it as tmpfs?
2023-08-11 18:57:58 <minimal> openrc?
2023-08-11 18:58:09 <Ermine> or initrd?
2023-08-11 18:58:22 <minimal> Ermine: nope, don't see it in initramfs' init
2023-08-11 19:02:16 <Ermine> minimal: my s6-driven alpine server doesn't have any scripts to mount /run, yet /run is tmpfs. Is it possible that kernel mounts it?
2023-08-11 19:03:09 <minimal> doubt it, don't believe /run is defined anywhere in kernel source
2023-08-11 19:03:35 <Ermine> *shrug*
2023-08-11 19:06:11 <Ermine> Ah, it's s6-linux-init
2023-08-11 19:06:42 <minimal> hmm, alpine-baselayout's APKBUILD creates /run and links /var/run to it.....but don't see that in the APK contents
2023-08-11 19:11:02 <minimal> ah, /lib/rc/sh/init.sh, part of openrc as I suspected
2023-08-11 19:12:13 <minimal> mm, that mkdir's it, doesn't set it as tmpfs
2023-08-11 19:13:10 <minimal> actually it doesw both, mkdir and mount depending on some conditions
2023-08-11 19:14:47 <minimal> WhyNotHugo: so answer is openrc
2023-08-11 19:15:42 <WhyNotHugo> The mkdir there sounds redundant given alpine-baselayout
2023-08-11 19:15:58 <WhyNotHugo> Why is it not in fstab? It gets mounted before the rootfs?
2023-08-11 19:16:24 <minimal> nope, those directories creates by alpine-baselayout's APKBUILD are *not* present in the alpine-baselayout packages contents
2023-08-11 19:16:58 <minimal> no, rootfs is always mounted by initramfs' init
2023-08-11 19:17:17 <minimal> long before openrc is started (which is on the rootfs so obviously rootfs has to be mounted first)
2023-08-11 19:17:30 <minimal> I'm only taking regarding sys-mode BTW
2023-08-11 19:17:36 <minimal> s/taking/talking/
2023-08-11 19:18:31 <minimal> fstab is used once the (localmount?) init.d service is started by openrc
2023-08-11 19:18:44 <WhyNotHugo> I think that directories are not listed as contents of packages.
2023-08-11 19:18:49 <WhyNotHugo> Hence why /run doesn't show up.
2023-08-11 19:20:04 <minimal> possibly
2023-08-11 19:21:30 <minimal> anyway the mkdir /run in openrc comes from upstream, we likely wouldn't patch that code out
2023-08-11 19:29:59 <WhyNotHugo> Yeah. I was mostly curious as to what created it.
2023-08-11 19:30:06 <WhyNotHugo> 
ACTION wonder if /var/run is safe to delete nowadays.

2023-08-11 19:33:57 <minimal> you'd have to check if any packaged software is still making use of /var/run
2023-08-11 19:34:06 <dwfreed> easier to just leave it
2023-08-11 19:35:05 <dwfreed> and the mkdir in openrc exists in case it doesn't get created in some other way, as a fallback before attempting to mount it
2023-08-11 19:35:18 <dwfreed> there's no need to remove it, it's a single mkdir call
2023-08-11 19:37:49 <dwfreed> and the reason it's not in fstab is because there's generally no need for it to be; iirc, openrc supports putting it in fstab if you want to change mount options for it, but otherwise it's like /sys, /dev, and /proc; system-level filesystems that always have to be mounted for things to function properly
2023-08-11 22:33:10 <ayushnix> are there any conventions about how to write files in conf.d for initscripts?
2023-08-11 22:33:22 <ayushnix> like, if a program runs in the foreground only, provides lots of options both as flags and config file vars, do I consider including those flags in the conf.d file and as command_args?
2023-08-11 22:34:02 <ayushnix> specifically, I'm talking about the pounce IRC bouncer here
2023-08-11 22:42:35 <ptrc> ayushnix: you can put them in the conf.d file, but commented
2023-08-11 22:42:54 <ptrc> and make it so they get expanded into init.d's command_args only when they're defined
2023-08-11 22:43:16 <ptrc> also, you might need `command_background=true` in the init.d script
2023-08-11 22:44:09 <ptrc> s/might //
2023-08-11 22:44:37 <ayushnix> ptrc: do I add `command_background=true` if the program in question runs in the foreground and doesn't have an option to fork to the background?
2023-08-11 22:45:04 <ptrc> precisely, see https://github.com/OpenRC/openrc/blob/master/service-script-guide.md#dont-write-your-own-startstop-functions
2023-08-11 22:46:35 <ayushnix> got it
2023-08-11 22:47:24 <ayushnix> also, do the vars `output_log` and `error_log` hold any significance or are somehow predefined by openrc? if no, how do I redirect stdout and stderr using redirection in init scripts?
2023-08-11 22:47:47 <ayushnix> s/how do/do/
2023-08-11 22:48:52 <ayushnix> because I added this segment in my init script
2023-08-11 22:48:59 <ayushnix> command_args="${command_args} > ${POUNCE_LOG_DIR}/${RC_SVCNAME}.log 2>&1"
2023-08-11 22:49:22 <ayushnix> and this doesn't redirect any output to the log file
2023-08-11 22:53:29 <ayushnix> ok, I read the man page of openrc-run
2023-08-11 22:53:39 <ayushnix> output_log and error_log are predefined
2023-08-11 22:53:49 <ptrc> oh, yeah, you should use the output_log and error_log variables instead of redirecting it manually
2023-08-11 22:54:09 <ayushnix> got it
2023-08-11 22:54:14 <ayushnix> thanks ptrc 
2023-08-11 22:54:15 <ptrc> if you're running the bouncer as a separate user ( which imo you should ), you should also add a start_pre that makes sure the file is writable by said user
2023-08-11 22:54:40 <ayushnix> ok
2023-08-11 22:55:41 <ptrc> for example: https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/testing/jackett/jackett.initd#L24
2023-08-11 22:57:24 <ayushnix> that should help
2023-08-12 00:21:58 <ayushnix> I made this initscript for the pounce IRC bouncer. I want to add it to the the pounce package but before I submit it, I'd like to know if I missed anything or if I should fix/change anything.
2023-08-12 00:22:00 <ayushnix> https://0x0.st/H_6c.txt
2023-08-12 00:38:38 <elly> ayushnix: surely defaulting to running as root is not right?
2023-08-12 00:39:49 <ayushnix> elly: what should I use instead?
2023-08-12 00:40:54 <ayushnix> I guess I can default to the pounce user but that'll need to be created first.
2023-08-12 00:41:36 <elly> yeah, I think that if you want to install it systemwide like this, having the package create a dedicated user is wiser
2023-08-12 00:41:43 <elly> or perhaps you can even run it as nobody?
2023-08-12 00:41:44 <ayushnix> hmm, can add a pre-install file to the pounce package, create a pounce user and default to that 
2023-08-12 00:42:12 <ayushnix> thanks for the suggestion elly 
2023-08-12 00:46:42 <elly> sure thing
2023-08-12 00:46:52 <elly> I wonder if pounce does any sandboxing or similar
2023-08-12 00:49:14 <ayushnix> it does but only on openbsd
2023-08-12 00:49:49 <ayushnix> > pounce primarily targets OpenBSD, where it is sandboxed with pledge(2) and unveil(2).
2023-08-12 00:50:21 <ayushnix> https://git.causal.agency/pounce/about/
2023-08-12 00:51:04 <elly> ah, okay
2023-08-12 00:51:22 <elly> I wonder if they'd take a patch to do some kind of tasteful linux sandboxing as well
2023-08-12 00:58:31 <fjkdnenfofmf> Why doesnt alpine linux support dm verity
2023-08-12 01:03:23 <hdjdndjsldmdnd> does alpine linux work with dm-verity
2023-08-12 01:03:34 <hdjdndjsldmdnd> sorry my connection cut off
2023-08-12 01:04:09 <hdjdndjsldmdnd> i heard dm verity doesnt work because initramfs doesnt load the module
2023-08-12 01:09:03 <hdjdndjsldmdnd> hi
2023-08-12 01:14:46 <minimal> I earlier said that the initramfs does not appear to currently support verity at all
2023-08-12 01:15:12 <minimal> and they're gone again...
2023-08-12 01:47:12 <elly> it's really just a module you need in your initramfs or compiled in, I think
2023-08-12 01:47:44 <minimal> elly: nope, more functionality would be needed in the initramfs
2023-08-12 02:05:28 <elly> what else is needed for it?
2023-08-12 02:06:36 <minimal> the init has to run veritysetup AFAIK
2023-08-12 02:06:56 <minimal> and needs some info passed in to it on the cmdline AFAIK
2023-08-12 02:09:51 <minimal> something similar to slide 8 here: https://fosdem.org/2023/schedule/event/image_linux_secureboot_dmverity/attachments/slides/5559/export/events/attachments/image_linux_secureboot_dmverity/slides/5559/DM_Verity.pdf
2023-08-12 02:10:26 <minimal> which the initramfs' init would have to accept as valid cmdline entries and pass them on to veritysetup
2023-08-12 02:10:51 <minimal> and nlplug-findfs might need some changes too
2023-08-12 02:36:03 <elly> minimal: I thought veritysetup is run when creating the volume, not at boot
2023-08-12 02:37:42 <minimal> elly: I've never used it but the point is something has to "verify" the rootfs before it is mounted and whatever this is needs the "hashes" to verify against
2023-08-12 02:37:57 <minimal> and that is what is passed in on the cmdline
2023-08-12 02:39:24 <minimal> and secure boot appears ot be recommended for use with verity so that would also need to be set up
2023-08-12 02:39:51 <elly> minimal: no, it doesn't verify the rootfs before mount - it has a big block of hashes that are stored on disk alongside the data, and it verifies the hashes as data is read from the disk
2023-08-12 02:40:25 <elly> I've also never actually set up verity on my own devices though
2023-08-12 02:40:51 <minimal> so what's the "dm-mod.create" value shown in that PDF used for?
2023-08-12 02:43:14 <elly> the hash value? it's the hash of the root block of the verity hash tree
2023-08-12 02:43:33 <minimal> I asked what it is being used for
2023-08-12 02:43:40 <minimal> it is obviously passed in cmdline for some reason
2023-08-12 02:43:44 <elly> the verity kernel module parses it
2023-08-12 02:44:42 <elly> one sec, I will dig up a pointer to the involved code
2023-08-12 02:45:19 <minimal> ok, and I assume that is where secure boot comes in to ensure that hash is not modified
2023-08-12 02:46:02 <elly> yup - secure boot has to protect the kernel and its boot config
2023-08-12 02:46:21 <minimal> right, so secure boot needs to be setup
2023-08-12 02:46:23 <elly> (I only know how this works in chromeos, not elsewhere)
2023-08-12 02:46:27 <elly> yeah, definitely does
2023-08-12 02:46:52 <elly> although, verity can still be useful for detecting disk corruption if you have an RO disk image
2023-08-12 02:46:53 <minimal> so its not just a case of add the module to initramfs and off you go
2023-08-12 02:46:56 <elly> it has a mode where it just warns you
2023-08-12 02:50:51 <minimal> and also the likes of Arch's article talk about using a UKI, which you'd have to setup as well
2023-08-12 02:51:02 <elly> yeah, I have no knowledge of that at all
2023-08-12 02:51:22 <elly> for chromeos' purposes both the kernel and the boot params are verified by some bespoke stuff in the read-write firmware
2023-08-12 02:51:47 <minimal> I've done that with Alpine, it's where the kernel, initramfs, and cmdline settings are combined into a single file
2023-08-12 02:52:07 <minimal> so with secure boot you could then sign it to detect if it had been tampered with
2023-08-12 02:52:46 <minimal> without UKI I assume the cmdline is coming from UEFI vars entry which (I believe) secure boot doesn't protect from tampering
2023-08-12 02:59:27 <elly> heh, my laptop is going full jet engine doing a linux checkout :)
2023-08-12 03:30:55 <minimal> elly: looking at "modinfo" for dm_verity, it only understands a single parm (which is not the "require_signatures" mentioned in that slide) and dm_mod understands multiple parms but none of them match the "create" mentioned in that slide
2023-08-12 03:32:27 <minimal> so I don't see how dm-mod.create="..." and dm_verity.require_signatures=1 passed on cmdline would affect the dm_mod and dm_verity modules
2023-08-12 03:37:07 <minimal> elly: also "A command line tool veritysetup is available to compute or verify the hash tree **or activate the kernel device**." That's what I was thinking of
2023-08-12 03:38:17 <minimal> so I believe the initramfs would need to do the equivalent of either "dmsetup create .." or "verity setup create .." for the /dev/mapper/ entry to appear to be used for rootfs
2023-08-12 03:45:33 <minimal> elly: ah, dm-verity.require_signatures isn't recognised as kernel was built without CONFIG_DM_VERITY_VERIFY_ROOTHASH_SIG
2023-08-12 08:46:29 <ayushnix> I've seen packages in alpine use /var/lib/pkgname as their home folder.
2023-08-12 08:46:31 <ayushnix> However, the FHS states that "users must never need to modify files in /var/lib to configure a package's operation". I guess it's a guide and not really something to strictly adhere to?
2023-08-12 09:20:02 <ayushnix> So I created this initscript for pounce. I created symlinks to this initscript to start multiple instances of pounce.
2023-08-12 09:20:03 <ayushnix> https://0x0.st/H_I5.txt
2023-08-12 09:20:28 <ayushnix> But when I start an initscript symlink, I get this error.
2023-08-12 09:20:29 <ayushnix> https://0x0.st/H_IR.txt
2023-08-12 09:21:45 <ayushnix> the output_log and error_log variables are commented out in the /etc/conf.d/pounce.libera
2023-08-12 09:24:38 <ayushnix> not sure why error_log resolves to /var/log/pounce.log when the name of the log file is set to ${RC_SVCNAME}
2023-08-12 09:24:57 <ayushnix> which should expand to pounce.libera.log
2023-08-12 09:40:43 <ayushnix> huh, the value of output_log inside start_pre isn't the one that's assigned by default
2023-08-12 11:22:00 <ayushnix> managed to solve this issue with help from #openrc on libera
2023-08-12 11:22:17 <ayushnix> will raise a MR for this soon
2023-08-12 14:28:08 <RoyalTiger[m]> Can anyone here say something regarding the state of the Alpine Linux "No-Cloud" cloud-init (https://www.alpinelinux.org/cloud/) installation/set up? The last days I've taken a look at Packer + PROXMOX, but every variant for automated creation using cloud-init are refering other distributions, but sadly not Alpine.  
2023-08-12 14:29:18 <RoyalTiger[m]> * regarding the development state of
2023-08-12 14:37:13 <RoyalTiger[m]> <RoyalTiger[m]> "Can anyone here say something..." <- Okay, while the website says "Currently, only AWS cloud images are available; other major cloud providers and "NoCloud" will be supported in the future." some kind of nocloud related stuff seems to be implemented (https://gitlab.alpinelinux.org/alpine/cloud/alpine-cloud-images)
2023-08-12 15:11:52 <minimal> RoyalTiger[m]: I'm the Alpine maintainer for the cloud-init package
2023-08-12 15:12:32 <minimal> RoyalTiger[m]: you could try my own script to create a VM image: https://github.com/dermotbradley/create-alpine-disk-image
2023-08-12 15:55:15 <tangboot> fdsfdfnl
2023-08-12 15:55:33 <tangboot> Hi folks, I'm trying to iPXE boot with ZFS module kernel support. I intend to bootstrap the root on ZFS process via iPXE using the alpine netboot image. My problem is the alpine extended edition has the ZFS kernel module built-in, however the netboot image does not- so I'm trying to work out how to add the ZFS module to the netboot image (https://boot.alpinelinux.org/).
2023-08-12 15:55:43 <tangboot> Without the zfs module, booting the netboot image I (of course) see modprobe: module zfs not found in modules.dep 
2023-08-12 15:55:52 <tangboot> Ideally I'd please like information on how to rebuild the current netboot image and add the zfs kernel module but in my research have not yet found docs for that (maybe https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/scripts/mkimg.standard.sh#L39 ).
2023-08-12 15:55:58 <tangboot> If anyone may please point me to the repo where netboot-3.18.3 gets built that'd be much appreciated (https://gitlab.alpinelinux.org/alpine/infra/alpine-netboot) appears to be the website not the build process.
2023-08-12 15:56:18 <tangboot>  I've also tried pointing iPXE to an extracted alpine-extended-3.18.0-x86_64.iso (vmlinuz-lts & initramfs-lts) but see /sbin/init not found in new root when I do that. Thanks for your time
2023-08-12 16:38:50 <bovis> A long shot, but: I updated Alpine, then 'rails server' (Ruby on Rails) started giving me this error. https://bpa.st/RP7A Rails was working fine before the update. Sqlite and Ruby Alpine packages didn't update, and their gems haven't changed version. Usual routes of uninstalling or updating haven't fixed it. Sandboxing and changing Ruby versions with rbenv didn't fix it.
2023-08-12 16:39:20 <bovis> Is there anything in the past month of Alpine updates that would be connected to this? Or do you have any ideas of where else I can investigate?
2023-08-12 16:43:57 <ikke> bovis: what version of alpine?
2023-08-12 16:46:24 <ikke> Are you using a ruby docker image or something like that? 
2023-08-12 16:46:31 <ikke> /usr/local/lib/site_ruby/3.2.0 means it's not from Alpine Linux
2023-08-12 16:46:51 <bovis> ikke: Edge. My update was on Aug 7.
2023-08-12 16:47:16 <bovis> No docker. Everything is run on my desktop.
2023-08-12 16:48:32 <ikke> No Alpine Linux package would install anything in /usr/local/*
2023-08-12 16:49:14 <bovis> I don't know enough to say, but maybe a gem would use that directory.
2023-08-12 16:49:24 <ikke> It would not
2023-08-12 16:49:59 <ikke> bovis: what does `which ruby` return?
2023-08-12 16:50:27 <bovis> \/usr/bin/ruby
2023-08-12 16:50:34 <ikke> and `gem env`?
2023-08-12 16:51:23 <bovis> https://bpa.st/P43A
2023-08-12 16:51:46 <ikke> That all looks normal
2023-08-12 16:55:09 <bovis> I haven't purposely changed anything to use /usr/local or tried to adjust ruby or gem default install locations.
2023-08-12 16:55:17 <ikke> bovis: "posix_fallocate64: symbol not found"
2023-08-12 16:55:22 <ikke> this is due to musl 1.2.4
2023-08-12 16:55:36 <ikke> but that normally is only a compile time error
2023-08-12 16:55:54 <ikke> but maybe it can crop up with dlopened libraries
2023-08-12 16:59:14 <ikke> it means you need to make sure that gem is compiled against musl 1.2.4
2023-08-12 16:59:15 <bovis> There is no ruby-sqlite package in Alpine. Other packages that install gems don't appear to do much beyond call 'gem install <package>' (https://git.alpinelinux.org/aports/tree/community/ruby-mini_portile2/APKBUILD). Should I make an APKBUILD for the sqlite gem, and can it address the posix_fallocate64 error?
2023-08-12 16:59:41 <ikke> bovis: you might still need to patch the gem to not use any *64 syscalls
2023-08-12 17:03:27 <bovis> ikke: I'll see what I can do. Thank you for the help.
2023-08-12 17:06:56 <ikke> bovis: the gist is that glibc has the *64 syscalls to prepare the 2038 problem. For musl, this is not a problem because the normal syscalls already are 64-bit. It did have *64 shims for compatibility with glibc, but 1.2.4 removed the shims by default
2023-08-12 18:15:42 <quantum`> I'm trying to set up unbound with dnf-over-tls.
2023-08-12 18:16:02 <quantum`> Which is the proper package for TLS certs?
2023-08-12 18:16:54 <quantum`> Could it be the one which provides /etc/ssl/cert.pem ?
2023-08-12 18:17:45 <quantum`> ca-certificates-bundle ?
2023-08-12 18:18:38 <quantum`> Or, ca-certificates ?
2023-08-12 18:18:42 <ikke> quantum`: if it expects a single pem file with all certificates, then yes
2023-08-12 18:20:45 <quantum`> It does, but the former lacks update-ca-certificates, which is in the latter.  
2023-08-12 18:21:14 <quantum`> Does update-ca-certificates only update crt files, or also pem?
2023-08-12 18:21:44 <ikke> it creates a new combined pem file from loose files
2023-08-12 18:21:53 <ikke> if you are not adding your own certs, you don't need it
2023-08-12 18:22:08 <quantum`> Ok, thx.
2023-08-12 20:46:23 <m_ben> is it somehow possible to search the package database for a specific command?
2023-08-12 20:46:48 <m_ben> like which package provides command "foo"?
2023-08-12 20:47:04 <lopid> apk search cmd:foo
2023-08-12 20:47:26 <m_ben> hey a familiar user name :D
2023-08-12 20:47:36 <m_ben> ty 
2023-08-12 20:47:47 <lopid> but is it the familiar person behind the username…
2023-08-12 20:51:06 <m_ben> I'm assuming this will search only in packages that are installed on my system?
2023-08-12 20:51:14 <Ermine> no
2023-08-12 20:53:20 <XFaCiEer> I wish the ed/vi could properly regex
2023-08-12 20:54:50 <m_ben> where is this explained with "cmd:foo"? I just checked apk-search(8)
2023-08-12 21:03:03 <lopid> i originally found it on the wiki
2023-08-12 21:13:08 <lopid> https://docs.alpinelinux.org/user-handbook/0.1a/Working/apk.html#_searching_for_packages
2023-08-12 21:14:09 <m_ben> would be nice if this was explained in the man pages as well
2023-08-12 21:15:45 <lopid> agreed. all the alpine docs are lacking in some respect
2023-08-12 21:53:59 <quinq> I have a local patch for that, should share it…
2023-08-13 00:00:40 <Armote[m]> edge got multiple conflicts. one: abseil-cpp-raw-logging-internal-20230802.0-r0 conflicts: abseil-cpp-raw-logging-internal-20230125.3-r3
2023-08-13 00:01:21 <Armote[m]> all the conflicts are abseil
2023-08-13 00:05:45 <ptrc> Armote[m]: it's kinda obvious that you can't install two versions of the same package simultaneously
2023-08-13 00:06:07 <ptrc> do you have a bit more context on this one?
2023-08-13 00:10:50 <ptrc> ah, got it, it's even more stuff that wasn't rebuilt against new abseil-cpp
2023-08-13 00:44:06 <testburgers> Can anyone here help me understand what this user did to get their alpine linux running smoothly? I am mainly hoping to use this distro on a raspberry pi 4b to run basic videos, chats, or run a coding environment or various automated API web scripts.
2023-08-13 00:44:35 <testburgers> >Alpine After about an hour of coonfiging everything just works, both Wayland and Xorg, Wayland is so stable there's no need in X, which is just slightly slower. starting the system is so easy you can add it to the startup process, but I prefer sudo rc-service sddm start. Yes. SDDM. Just works. RAM usage? 1.13Gb with both Discord and Youtube open, never ran into VRAM issues either. No swap use in hours of uptime, too.
2023-08-13 00:45:06 <ptrc> from experience, it's mostly like that out of the box :p
2023-08-13 01:12:22 <Saijin_Naib[m]> If you aren't using the setup-alpine script with the setup-desktop step, I would highly recommend doing so
2023-08-13 01:19:29 <testburgers> >[18:12] <Saijin_Naib[m]> If you aren't using the setup-alpine script with the setup-desktop step, I would highly recommend doing so 
2023-08-13 01:19:44 <testburgers> i admit I have no idea about it. but the last time I used alpine was a long time ago
2023-08-13 01:19:57 <testburgers> i guess i'll just bite the bullet and hope everything works
2023-08-13 01:20:11 <testburgers> im used to problems when I try to run on my jank hardware though
2023-08-13 01:21:46 <Saijin_Naib[m]> Those scripts, especially setup-desktop, have made achieving a functional desktop environment, like what I'm used to getting with Debian-deriveds, Solus, et al, insanely easy
2023-08-13 01:22:05 <Saijin_Naib[m]> Before having that, I would struggle with setting up user groups, permissions, what services I needed to have, login managers, all that
2023-08-13 01:22:25 <Saijin_Naib[m]> I bet your Pi is more powerful than my craptop that I run Alpine on with XFCE perfectly fine
2023-08-13 01:22:47 <Saijin_Naib[m]> About the only tricky thing I do is force my eMMC SSD to use XFS over LVM and BFQ for the scheduler
2023-08-13 03:56:42 <openlinux21[m]> 
ACTION uploaded an image: (112KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/QuRVhSwoUqbHJjNjgCjSddjl/Screenshot_20230813_111540_com.limbo.emu.main-01.jpeg >

2023-08-13 03:56:57 <openlinux21[m]> my server,GRUB
2023-08-13 10:33:09 <zukunf> greets
2023-08-13 10:33:18 <zukunf> rusticl works on musl
2023-08-13 10:33:29 <zukunf> but what about intel-compute?
2023-08-13 11:04:59 <zukunf> hello?
2023-08-13 11:06:29 <lopid> i don't see an "intel-compute" package
2023-08-13 11:21:08 <zukunf> :(
2023-08-13 11:21:56 <zukunf> https://github.com/intel/compute-runtime
2023-08-13 11:22:42 <zukunf> I wonder the work involved to make a musl port.
2023-08-13 11:43:53 <Apachez> what does that do?
2023-08-13 12:00:39 <zukunf> opencl accel
2023-08-13 12:01:02 <zukunf> darktable, gimp, etc
2023-08-13 12:22:05 <quinq> zukunf, is it not written in a portable manner in the first place?
2023-08-13 12:26:04 <zukunf> ouch!!
2023-08-13 12:28:05 <ikke> zukunf: that was a question from quinq, not a statement
2023-08-13 12:28:24 <zukunf> ah, ok
2023-08-13 12:28:51 <m_ben> how can I modify the keymap files .bmap in /etc/keymap/? or add a new one
2023-08-13 12:28:52 <zukunf> quinq: how can I tell whether is portable or not? as ar as I know it uses LLVM 15 and clang 15
2023-08-13 12:29:10 <ikke> zukunf: try to build it and see if it fails
2023-08-13 12:29:13 <m_ben> they do not seem to be text files
2023-08-13 12:29:24 <zukunf> ikke: ha ha!!
2023-08-13 12:31:16 <zukunf> mm.. I don't know where to look, it's got a bunch of cmake files
2023-08-13 12:33:54 <zukunf> just do make?
2023-08-13 12:34:18 <ikke> no, you first need to run cmake 
2023-08-13 12:34:28 <quinq> cmake is a configuration framework, before building, you need to configure it
2023-08-13 12:34:56 <quinq> Them you build it with whatever chosen backend
2023-08-13 12:35:00 <quinq> s/Them/Then/
2023-08-13 12:35:44 <m_ben> zukunf: cmake will generate a Makefile (or optionally a Ninja file) that you need to run
2023-08-13 12:35:57 <ikke> zukunf: example: https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/community/darktable/APKBUILD#L53
2023-08-13 12:36:24 <ikke> We prefer ninja (samurai) as build system
2023-08-13 12:37:47 <lopid> setup-keymap?
2023-08-13 12:38:56 <ikke> That doesn't allow you to 
2023-08-13 12:39:01 <ikke> modify keymaps
2023-08-13 12:39:07 <ikke> only select which one is active
2023-08-13 12:39:47 <m_ben> setup-keymap will allow to select one of the predefined ones. I don't know how to open the .bmap files and modify them
2023-08-13 12:39:56 <m_ben> I don't even know what format they are
2023-08-13 12:42:07 <zukunf> wow, it requires intel graphics compiler oO
2023-08-13 12:42:15 <zukunf> it's a whole toolchain
2023-08-13 12:42:28 <quinq> :D
2023-08-13 12:42:42 <zukunf> that's gonna be quite hard
2023-08-13 12:43:24 <ikke> m_ben: I'm not sure, but searching shows it may be block map files, but not sure if that's the same or something different
2023-08-13 12:44:25 <zukunf> my guess is that opencl on intel is not a priority perhaps on musl
2023-08-13 12:47:26 <ikke> someone needs to be motivated enough to try to get it built
2023-08-13 13:00:41 <zukunf> ikke: i am motivated!
2023-08-13 13:01:52 <zukunf> dependencies: GmmLib - https://github.com/intel/gmmlib Intel Graphics Compiler - https://github.com/intel/intel-graphics-compiler  Intel(R) Metrics Discovery (MDAPI) - https://github.com/intel/metrics-discovery  Intel(R) Metrics Library for MDAPI - https://github.com/intel/metrics-library 
2023-08-13 13:13:04 <WhyNotHugo> How can I mount all volumes RO in order to clealy execute 'kexec -e'?
2023-08-13 13:15:49 <m_ben> ikke: it seems that .bmap files are loaded with loadkmap. the help page says that it loads a binary keyboard translation table. whatever that means...
2023-08-13 13:18:00 <ikke> I have no idea either
2023-08-13 13:18:56 <WhyNotHugo> I guess my real question is: can I somehow remount all mounts as read-only?
2023-08-13 13:23:20 <ikke> Don't think there is a single / easy command to do that
2023-08-13 13:23:43 <WhyNotHugo> I think that what I want is to actually call reboot with LINUX_REBOOT_CMD_KEXEC
2023-08-13 13:29:54 <ascensionwires> there they say qt5ct or qt6ct good for sway https://wiki.archlinux.org/title/Wayland#Qt
2023-08-13 13:30:40 <ascensionwires> is package on alpine?
2023-08-13 13:30:46 <ikke> what package?
2023-08-13 13:30:54 <ascensionwires> qt6ct
2023-08-13 13:31:02 <ascensionwires> oh it in testing
2023-08-13 14:15:29 <m_ben> ikke: I think I found the answer. the .bmap files are created with "loadkeys -b", provided by the kbd package
2023-08-13 14:15:39 <m_ben> see loadkeys --help
2023-08-13 14:41:03 <Nulo[m]> when upgrading from a minirootfs, alpine-release isn't installed so /etc/os-release remains outdated. when adding the package the new file ends up in /etc/os-release.apk-new....
2023-08-13 14:41:39 <Nulo[m]> the best i can think of is to delete /etc/os-release before manually adding alpine-release, but even then if adding alpine-release fails we end up with no os-release
2023-08-13 15:34:01 <m_ben> "apk add iwd" says "no such package"
2023-08-13 15:34:33 <ikke> m_ben: did you enable the community repo?
2023-08-13 15:35:09 <m_ben> no not yet. thought it would be in the main repo
2023-08-13 15:38:04 <ikke> no, it's in community
2023-08-13 16:12:25 <m_ben> I wished all the builtin commands like ash, find and friends were better documented. TIL that ash supports local variables
2023-08-13 16:13:19 <grawity> WhyNotHugo: not exactly the best way but "echo r > /proc/sysrq-trigger"
2023-08-13 16:13:27 <grawity> WhyNotHugo: er, not 'r' but 
2023-08-13 16:13:33 <grawity> let me check what the command was
2023-08-13 16:13:51 <ikke> m_ben: busybox documentation is a bit lacking, indeed
2023-08-13 16:13:53 <grawity> 'u'
2023-08-13 16:27:02 <Hello71> proper reboot by kexec needs to be handled by the init system, same as regular reboot. it needs to stop all services, SIGTERM all processes, unmount/mount-ro all disks, etc
2023-08-13 16:27:56 <Hello71> e.g. systemctl kexec or openrc-shutdown -k
2023-08-13 16:28:59 <Hello71> but alpine uses busybox init which doesn't support kexec, so you would have to use a restart directive in inittab or maybe openrc shutdown; kexec -e would work
2023-08-13 16:29:18 <Hello71> not sure exactly how openrc kill_all works
2023-08-13 16:31:27 <WhyNotHugo> The sysrq approach "Will attempt to remount all mounted filesystems read-only.". Is there even a way to query if/when this completed successfully?
2023-08-13 16:31:41 <WhyNotHugo> Hello71: Yeah, that was my impression; that init itself needs to handle it.
2023-08-13 16:35:34 <WhyNotHugo> Oh, based on openrc(8), looks like it is the one who does the actual reboot after switching to the "shutdown" runlevel.
2023-08-13 17:46:42 <Hello71> theoretically it could but I don't think it does
2023-08-13 19:06:53 <linux818> hello how can i use initramfs-tools to create a csutom initramfs? i noticed for some reason alpine uses mkinitfs for making initramfs images instead of the initramfs-tools for ubuntu.. seems its missing alot of the functionality that initramfs-tools has like hooks/scripts/etc
2023-08-13 19:07:01 <linux818> https://wiki.alpinelinux.org/wiki/Initramfs_init
2023-08-13 19:07:08 <linux818> https://manpages.ubuntu.com/manpages/xenial/man8/initramfs-tools.8.html
2023-08-13 19:09:54 <linux818> for example initramfs-tools has a function `copy_exec ` that will copy binary and all libraries into the initramfs. without initramfs-tools i would have to do this manually for each binary which is a fking pita
2023-08-13 19:10:29 <zcrayfish> If you´re just adding additional files/modules to the initrd, you could add a new feature (.modules/.files) file to /etc/mkinitfs/features.d, and then add said feature to /etc/mkinitfs/mkinitfs.conf 
2023-08-13 19:11:04 <zcrayfish> looks like globbing is supported which might help
2023-08-13 19:11:17 <minimal> linux818: there are 3 ways to create an initramfs on Alpine, mkinitfs (the official/standard method), booster, and dracut
2023-08-13 19:18:37 <Saijin_Naib[m]> zukunf if you tackle this, I'll be so grateful. Super useful to have for so many compute needs 
2023-08-13 19:25:15 <linux818> zcrayfish: thanks for this.. i see this is only for kernel modules though it wouldnt support adding a binary into the initramfs or maybe i missed something
2023-08-13 19:25:51 <zcrayfish> There are lots of binaries in there, look at the features that end in .files
2023-08-13 19:26:19 <zcrayfish> base.files, for example, put's busybox in the initrd.
2023-08-13 19:33:57 <linux818> zcrayfish: ah thanks. looks like this is what I need. will have a play around with this and report back
2023-08-13 19:35:25 <zcrayfish> cool beans. Glad I could help \o/
2023-08-13 19:38:58 <WhyNotHugo> So what I need to do for kexec is: (1) tell OpenRC to exec shutdown sequence (2) instead of actually calling "shutdown", exec into 'kexec -e'.
2023-08-13 19:39:13 <WhyNotHugo> The question is if openrc actually allows this.
2023-08-13 19:43:02 <zcrayfish> I was thinking about using kexec on some of my boxes... my plan was to just use killprocs/killall5 and mount-ro before doing the kexec. Have not got around to it though
2023-08-13 19:44:09 <zcrayfish> I think I hit a stumbling block of the first box I was going to try it on was running a kernel that isn't configured for kexec. ugh.
2023-08-13 19:45:47 <ikke> Is the reason for using kexec is to have faster reboots (ie, skipping firmware initialization, etc)?
2023-08-13 19:46:29 <zcrayfish> seems to be why most people use it
2023-08-13 19:46:38 <zcrayfish> In my case it's because linode's lassie is SLOW.
2023-08-13 19:46:48 <WhyNotHugo> Yeah, my hardware takes a long time to reboot.
2023-08-13 19:46:49 <ikke> ah yeah, sounds familiar
2023-08-13 19:47:10 <ikke> takes about 30 seconds to a minute before the vps is started again
2023-08-13 19:47:47 <WhyNotHugo> kexec works fine on my workstation too, it's just that I prefer to properly unmount when rebooting.
2023-08-13 19:53:35 <zcrayfish> guess I won't be trying it on my laptop anytime soon:
2023-08-13 19:53:38 <zcrayfish> ~ $ grep CONFIG_KEXEC /boot/config-edge 
2023-08-13 19:53:40 <zcrayfish> # CONFIG_KEXEC is not set
2023-08-13 19:54:00 <zcrayfish> (the -lts kernel doesn't work as good as -edge on my hardware, bummer)
2023-08-13 19:55:38 <WhyNotHugo> I just patched the aport and rebuild.
2023-08-13 19:55:52 <WhyNotHugo> I'll send an MR, I think it would be useful to have it enabled on linux-edge
2023-08-13 19:56:20 <WhyNotHugo> It asked some questions and I just hit enter... I have some reading to do. But it just worked.
2023-08-13 19:56:21 <WhyNotHugo> (tm)
2023-08-13 19:57:06 <m_ben> it seems, busybox's clear command won't erase the scrollback buffer but just scroll up
2023-08-13 19:57:56 <WhyNotHugo> I just use ctrl+l and zsh does some magic to clear the screen.
2023-08-13 19:58:16 <zcrayfish> m_ben: Depending on your terminal . . . . reset? :)
2023-08-13 19:58:23 <m_ben> alias clear='command clear; printf "\033[3J"'
2023-08-13 19:58:35 <m_ben> zcrayfish: yes that's another option :)
2023-08-13 19:59:33 <m_ben> WhyNotHugo: but does it also remove the content from the scrollback buffer, i.e. if you scroll up in the terminal is everything deleted?
2023-08-13 19:59:47 <ikke> afaik it always just scrolls up
2023-08-13 20:00:33 <WhyNotHugo> m_ben: It leaves whatever is on-screen in the scrollback buffer.
2023-08-13 20:00:53 <m_ben> anyway, printf "\033[3J" is the terminal escape sequence that erases the scrollback
2023-08-13 20:00:56 <zcrayfish> I think it depends on the $TERM 
2023-08-13 20:01:05 <WhyNotHugo> E.g.: I can see it by moving the mouse wheel to tap-and-drag.
2023-08-13 20:01:26 <zcrayfish> `clear -T vt100' removes the scrollback on kitty. go figure.
2023-08-13 20:01:50 <minimal> WhyNotHugo: don't you need to pass "kexec_load_disabled=0" on the cmdline?
2023-08-13 20:02:48 <zcrayfish> relevant: https://gitlab.alpinelinux.org/alpine/aports/-/issues/14017
2023-08-13 20:26:41 <WhyNotHugo> No, just patched it and installed kexec-tools
2023-08-13 21:16:14 <m_ben> on my laptop I have gentoo linux installed and htop shows that it uses 180MiB, on my new small minipc alpine linux uses 260MiB :(
2023-08-13 21:16:46 <m_ben> and I haven't even setup anything yet apart from wifi
2023-08-13 21:18:29 <quinq> What are you talking about though?
2023-08-13 21:19:13 <m_ben> I just compared a fresh alpine linux installation with my old gentoo machine, that's all
2023-08-13 21:23:06 <m_ben> I was talking about memory usage
2023-08-13 21:23:49 <m_ben> I'm surprised that alpine linux requires so much more memory
2023-08-13 21:28:17 <quinq> Ah ok
2023-08-13 21:33:49 <zcrayfish> sounds about right to me
2023-08-13 21:34:03 <m_ben> why?
2023-08-13 21:34:07 <m_ben> can you explain?
2023-08-13 21:34:49 <zcrayfish> on 3.18.2 with an extremely basic xfce desktop, alpine requires just under 512mb of RAM to not swap on x86_64, likewise on i686 it requires about 320mb
2023-08-13 21:34:59 <zcrayfish> obviously the requirements are less without a desktop
2023-08-13 21:35:02 <zcrayfish> that's all I know.
2023-08-13 21:35:42 <m_ben> another thing that I noticed: my laptop is from 2012 and the SSD is fully encrypted, yet it boots approx 5-10x faster than on my new minipc with a new AMD processor and no disk encryption. I really need to optimize this further
2023-08-13 21:35:58 <m_ben> zcrayfish: how come it requires so much more memory?
2023-08-13 21:36:18 <zcrayfish> I have not looked into it honestly.
2023-08-13 21:36:51 <zcrayfish> on x86_64, it requires 128mb minimum to even decompress the initrd hehe.
2023-08-13 21:38:35 <m_ben> obviously my gentoo kernel is optimized for my laptop. many kernel options have been disabled. but idk if that accounts for approx 80MiB of RAM difference
2023-08-13 21:39:51 <Hello71> zcrayfish: that's only for default image; sys install will use less. also, it's initramfs, not initrd, so the memory is cleared when switching root
2023-08-13 21:39:59 <Hello71> m_ben: are you definitely excluding cached ram?
2023-08-13 21:41:20 <zcrayfish> good point on all counts Hello71 
2023-08-13 21:41:33 <m_ben> Hello71: I checked Mem used with "free -h"
2023-08-13 22:27:50 <minimal> m_ben: you haven't clearly indicated what exactly you're comparing - Gentoo vs an Alpine run-from-ram or Sys mode install? with desktop or without desktop? etc
2023-08-13 22:28:48 <m_ben> minimal: sys mode install, without desktop
2023-08-13 22:42:59 <minimal> m_ben: so when you see 260MiB ram in use, what exactly is using it? By example, htop on a x86_64 Sys-mode VM shows 27.6M/475M with the likes of rsyslog and Virtualbox agent running
2023-08-14 03:37:12 <Saijin_Naib[m]> Anyone else on edge with linux-lts (6.1.44-0-lts) getting failure to mount XFS volumes on removable with "Filesystem type xfs not configured in kernel"?
2023-08-14 03:46:52 <zcrayfish> Saijin_Naib[m]: I haven't tried it lately. xfs is configured as a module instead of a built-in though.
2023-08-14 03:46:57 <zcrayfish> what happens if you try to modprobe it?
2023-08-14 03:50:15 <Saijin_Naib[m]> Looks like it is loaded already, but mount still fails via udisks2/Thunar Volman
2023-08-14 03:53:23 <Saijin_Naib[m]> My rootfs is XFS so it must be loaded, right?
2023-08-14 03:53:27 <Saijin_Naib[m]> XFS over LVM
2023-08-14 04:12:18 <zcrayfish> Interesting
2023-08-14 04:29:11 <uuidNuniq> Saijin_Naib[m]: have you tried command line mount
2023-08-14 04:29:27 <uuidNuniq> did you check dmesg output?
2023-08-14 05:00:22 <Saijin_Naib[m]> Yeah, commandline mount didn't seem to do it, either
2023-08-14 05:00:24 <Saijin_Naib[m]> I did not check DMESG yet
2023-08-14 05:00:36 <Saijin_Naib[m]> GNOME Disk Utility's mount function did mount it properly
2023-08-14 05:02:04 <uuidNuniq> Saijin_Naib[m]: journalctl -af -p4
2023-08-14 05:02:09 <uuidNuniq> and then mount .... 
2023-08-14 05:02:39 <Saijin_Naib[m]> Mm, let me investigate what that does
2023-08-14 05:06:24 <zcrayfish> journalctl? :o
2023-08-14 05:07:52 <Saijin_Naib[m]> 
ACTION sent a code block: https://matrix.org/_matrix/media/v3/download/matrix.org/xHkOcNXeZMopGhQzTJKdSUBP

2023-08-14 05:08:03 <Saijin_Naib[m]> Looks like dirty FS prevented mount?
2023-08-14 05:08:42 <Saijin_Naib[m]> I do not have journalctl
2023-08-14 05:09:02 <grawity> not all mount-related messages will have 'sda' mentioned in them
2023-08-14 05:09:46 <grawity> would be also useful to see just a dump of *all* recent dmesg messages, not just a grep
2023-08-14 05:10:47 <grawity> a lot of those messages look like "disk was disconnected while mounting" from earlier attempts
2023-08-14 05:11:08 <grawity> not just "dirty FS" in general
2023-08-14 05:11:54 <Saijin_Naib[m]> Ahhh, okay. I guess that fits. It would hang waiting to mount and then throw the error. I'd eject the card and try again, etc
2023-08-14 05:12:07 <Saijin_Naib[m]> ```
2023-08-14 05:13:51 <grawity> yeah, the filesystem is complaining because it gets I/O errors, and most (all?) of the I/O errors are along the lines of "request aborted because the disk is unplugged"
2023-08-14 05:13:55 <grawity> it's a SD card?
2023-08-14 05:14:39 <Saijin_Naib[m]> Yeah, microSDXC to an internal reader
2023-08-14 05:15:38 <Saijin_Naib[m]> What's interesting is that this persisted across reboots until I forced checked it via GNOME Disk Utility 
2023-08-14 05:16:15 <grawity> yeah if sd cards are dead they tend to stay dead across reboots
2023-08-14 05:16:53 <Saijin_Naib[m]> Hmm 🤔 it's back working now
2023-08-14 05:18:02 <Saijin_Naib[m]> https://pastebin.com/AXdMeG5w
2023-08-14 05:46:39 <Saijin_Naib[m]> f3probe seems to have killed it so I guess it was on the way out. Drat. 
2023-08-14 07:44:15 <WhyNotHugo> Building the linux-edge package I'm prompted for a few settings (e.g.: Rust support (RUST)).
2023-08-14 07:44:21 <WhyNotHugo> How can I figure out which CONFIG_ options these match to?
2023-08-14 07:44:35 <WhyNotHugo> I want to update config.x86_64 so that it doesn't prompt any more.
2023-08-14 07:53:18 <WhyNotHugo> Oh, I can `modprobe configs && cat /proc/config | gunzip`
2023-08-14 07:53:26 <WhyNotHugo> And then diff that
2023-08-14 08:19:19 <j_v> WhyNotHugo: when I have configured a custom kernel aport, usually starting from linux-edge, I start by running `abuild fetch verify unpack prepare`, then I cd into src/linux-X.X and run `make ARCH=$(uname -m) menuconfig` to fine tune, and also so I can read the help text that accompanies any CONFIG_ options I want to change or examine. ymmv
2023-08-14 09:41:49 <quinq> Hello
2023-08-14 09:42:11 <quinq> Is there a way to list explicitly installed packages (as opposed to the ones installed as dependencies)?
2023-08-14 09:42:24 <quinq> I couldn't find anything about that in apk-list or apk-info man-pages
2023-08-14 09:45:16 <pj> quinq: cat /etc/apk/world
2023-08-14 09:45:48 <quinq> Oh, that's the actual list? :D
2023-08-14 09:47:32 <quinq> Hummm, but if a package was explicitly installed, and then became an actual dependency to another package, will it be removed from that list or will it stay?
2023-08-14 09:47:46 <quinq> (it seems that is the former behavior, but I might be wrong)
2023-08-14 09:47:53 <pj> if it's not removed manually then not
2023-08-14 09:48:44 <quinq> ok, perfect, tranks :)
2023-08-14 09:50:32 <WhyNotHugo> j_v: I'll make a note of this for future reference.
2023-08-14 09:50:33 <WhyNotHugo> Thanks
2023-08-14 10:18:09 <Milamber75>  /win 1
2023-08-14 10:21:24 <dzilvys> dumb question. If I update grub do I need to update grub.efi?
2023-08-14 10:21:48 <dzilvys> aka, do I need to update the contents under /dev/sda1 -> /boot/efi/EFI
2023-08-14 10:52:03 <grawity> generally yes, the version of grub.efi (grub kernel) needs to match the /boot/grub/*.mod (grub modules)
2023-08-14 13:07:39 <grming> hi to all
2023-08-14 13:08:03 <grming> i have a lichee pi zero and I want to use alpine in this board
2023-08-14 13:08:57 <grming> I follow the tutorial on alpine on how to install alpine in arm boards and I have this output on the ttyS0 of the lichee pi zer
2023-08-14 13:08:58 <grming> https://pastebin.com/tEJZtnZQ
2023-08-14 13:09:15 <grming> this is the alpine tutorial
2023-08-14 13:09:16 <grming> https://wiki.alpinelinux.org/wiki/Alpine_on_ARM#Preparing_installation_media
2023-08-14 13:21:52 <latex> Hey does anyone know how to make my IPv6 server reachable from WAN?
2023-08-14 13:22:10 <latex> When I try pinging it says unknown host on my 4G connection
2023-08-14 13:23:53 <grawity> does your 4G connection have IPv6?
2023-08-14 13:24:42 <latex> Wait
2023-08-14 13:24:45 <latex> It might actually not
2023-08-14 13:24:47 <latex> Weird
2023-08-14 13:24:55 <latex> I thought every 4G provider had IPv6
2023-08-14 13:25:15 <latex> How do I test it now?
2023-08-14 13:25:24 <Ermine> ISPs usually don't care
2023-08-14 13:25:29 <latex> Well shit
2023-08-14 13:25:37 <pj> check your IP
2023-08-14 13:25:38 <ikke> Plenty of websites that offer ping
2023-08-14 13:25:42 <latex> Yeah vodafone has a shit reputation anyways
2023-08-14 13:25:58 <latex> And yeah I don't think I actually have IPv6 on my 4G
2023-08-14 13:28:02 <latex> Now I can't test IPv6 WAN
2023-08-14 13:28:06 <latex> Well that's awkward
2023-08-14 13:31:09 <latex> How do I even host services on IPv6?
2023-08-14 13:31:17 <latex> nc -l [2a02:a212:2c07:9f00::6969] 6969
2023-08-14 13:31:40 <latex> I just tried this command and the port isn't open apparently according to https://port.tools/port-checker-ipv6/
2023-08-14 13:32:00 <latex> Why does it not work?
2023-08-14 13:33:28 <Ermine> Open the port lol
2023-08-14 13:33:42 <Ermine> There's no magic
2023-08-14 13:33:56 <ikke> Is it a home network?
2023-08-14 13:34:13 <ikke> Most consumer routers have a firewall enabled 
2023-08-14 13:34:16 <Ermine> Make sure the service you host supports ipv6 and binds to ipv6 address
2023-08-14 13:34:41 <Habbie> latex, does that nc work from inside your network?
2023-08-14 13:34:59 <Habbie> oh sorry, that's nc -l
2023-08-14 13:35:04 <Habbie> can you connect to it from inside? and from the router?
2023-08-14 13:38:30 <j_v> grming: looks like problem with the initramfs... like it is missing or the initramfs is read only... just guessing, been a while since i messed with a pi and not ever with a lichee
2023-08-14 13:41:35 <j_v> grming: do you still have it in the rescue shell?
2023-08-14 13:42:15 <latex> Habbie: it does not
2023-08-14 13:44:20 <grming> j_v thanks for the reply
2023-08-14 13:45:24 <grming> i just download the tarball and untar in the SD card partition marked as boot
2023-08-14 13:51:08 <j_v> grming: your paste looks like the output of the initramfs dropping you into rescue shell. i am wondering if it is still running with from that point. if so could check if some commands work, because it looks like the busybox links aren't setup... could start troubleshooting from there.
2023-08-14 13:53:24 <latex> Habbie: any idea why it always says connection refused?
2023-08-14 13:53:31 <grming> j_v thanks, can you point me to some references? I just starting with this embedded
2023-08-14 13:55:25 <Habbie> latex, no - does it work from the machine itself?
2023-08-14 13:57:32 <minimal> latex: if you don't have an IPv6 address provided by the 4G connection then of course nothing on the Internet will be able to connect, via IPv6, to any service you run on a port via IPv6...
2023-08-14 13:58:08 <Habbie> the 4G was for testing; the service latex is trying to run would be on cable
2023-08-14 13:58:11 <j_v> well, it booted, so that is a start. like i said been a while since i booted my rasp pi and differences with your lichee pi (a board i am unfamiliar with) could be substantial. but since it booted into the initramfs, maybe there is some evidence of what is wrong (why it dropped into rescue shell). The *no such file or directory* messages make me wonder if the initramfs is
2023-08-14 13:58:12 <Habbie> (i have some context from #openwrt)
2023-08-14 13:58:13 <j_v> somehow readonly
2023-08-14 13:59:47 <minimal> Habbie: ah, ok, I understand now
2023-08-14 14:02:51 <j_v> grming: i do note that the kernel, initramfs and dtb all got loaded and the messages after starting kernel should be from the initramfs. line 99 is good evidence that it IS the alpine initramfs, but that nlplug-findfs wasn'tfound
2023-08-14 14:04:51 <grming> https://wiki.alpinelinux.org/wiki/Create_a_Bootable_Device
2023-08-14 14:05:19 <latex> apparently it doesn't even work from the machine itself Habbie 
2023-08-14 14:05:29 <grming> trubleshoting says: Wrong device name syslinux is wrong
2023-08-14 14:06:09 <grming> but I only had grub.cfg in the boot folder of the tarball
2023-08-14 14:06:42 <minimal> j_v, grmimg: it looks like the initramfs couldn't create the /sbin directory for the busybox links to go into: https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/initramfs-init.in#L341
2023-08-14 14:07:02 <minimal> as the busybox softlinks are not in place then the rest of the init will have problems running the various utils it needs
2023-08-14 14:08:11 <minimal> note that the "busybox mkdir" run just before then does NOT create /sbin
2023-08-14 14:08:13 <j_v> minimal: yep, was just going to ask which download had unpacked on the install media
2023-08-14 14:08:37 <j_v> so i could take a look at it unpacked
2023-08-14 14:10:43 <j_v> grming: if you tell which install file you downloaded, can take a look at it
2023-08-14 14:11:41 <grming> alpine-uboot-3.18.3-armv7
2023-08-14 14:21:12 <latex> I'm so confused
2023-08-14 14:23:35 <j_v> looking at the contents of the initramfs, sbin/nlplug-findfs exists, as well as bin/busybox, so that is not issue, it's a cortex a7, so the arch appears to be right...
2023-08-14 14:24:34 <minimal> latex: what does "netstat -6ln | grep 6969" show?
2023-08-14 14:25:28 <minimal> latex: ah, Busybox netstat doesn't support "-6", so use "-ln" instead
2023-08-14 14:25:39 <latex> netstat: unrecognized option: 6
2023-08-14 14:25:47 <latex> also no output
2023-08-14 14:26:01 <latex> maybe my usage of nc is incorrect somehow
2023-08-14 14:27:34 <minimal> j_v: the problems appears to be that whenever busybox wants to create softlinks in /sbin that /sbin doesn't exist
2023-08-14 14:27:34 <j_v> seems like a readonly or permissions issue
2023-08-14 14:28:13 <minimal> j_v: also I guess the init doesn't find nlplug-findfs as it doesn't find it in its PATH
2023-08-14 14:28:43 <minimal> latex: I already pointed out that "6" wouldn't be recognised and how to deal with that
2023-08-14 14:28:55 <latex> I removed the 6 but it gives no output
2023-08-14 14:28:58 <latex> I also don't see it in ss
2023-08-14 14:29:16 <minimal> latex: so then there is nothing listening on port 6969
2023-08-14 14:29:26 <minimal> so of course nothing can then connect to it
2023-08-14 14:29:34 <j_v> grming: do you have a boot prompt that you can type into?
2023-08-14 14:29:47 <grming> yeah
2023-08-14 14:30:35 <latex> yeah but apparently nc isn't working as I expect it to then
2023-08-14 14:30:48 <latex> But none of my services seem to show up in ss for my IPv6 address
2023-08-14 14:31:32 <latex> Oh okay nc'ing my http service works
2023-08-14 14:31:38 <latex> but idk whether it works from WAN
2023-08-14 14:31:48 <latex> Can someone try establishing HTTP connectivity with [2a02:a212:2c07:9f00::6969]?
2023-08-14 14:33:12 <j_v> grming: what i am thinking is '/bin/busybox find / -exec /bin/busybox ls -l {} +'
2023-08-14 14:33:37 <j_v> to see permissions of initramfs files
2023-08-14 14:33:47 <j_v> and 'echo $PATH'
2023-08-14 14:34:17 <grming> like matrix man
2023-08-14 14:34:23 <grming> doing a lot of stuff in here
2023-08-14 14:34:32 <j_v> if you could paste that, maybe that gives a clue
2023-08-14 14:35:30 <grming> it still generating output
2023-08-14 14:35:35 <grming> should I stope it?
2023-08-14 14:36:13 <j_v> shouldn't take that long if is actually in the initramfs
2023-08-14 14:39:32 <grming> https://pastebin.com/Y9xeG8kd
2023-08-14 14:41:30 <j_v> well, thats wierd. only the /sys directory
2023-08-14 14:43:53 <latex> Anyone know why my nginx HTTP service is not accessible form [2a02:a212:2c07:9f00::6969]?
2023-08-14 14:43:59 <latex> from
2023-08-14 14:44:02 <latex> I can access it locally
2023-08-14 14:44:21 <latex> But apparently on WAN it's not working according to 1 person. (feel free to test)
2023-08-14 14:44:35 <minimal> latex: did you open up port 6969 on your firewall?
2023-08-14 14:44:41 <grming> its still throwing data
2023-08-14 14:44:41 <latex> I don't have a firewall
2023-08-14 14:44:44 <grming> its too long
2023-08-14 14:44:54 <latex> Or does alpine come with a default firewall config for ipv6?
2023-08-14 14:44:58 <minimal> latex: your machine(s) are completely exposed to the Internet?
2023-08-14 14:45:06 <latex> I hope so
2023-08-14 14:45:09 <latex> I don't like firewalls
2023-08-14 14:45:14 <latex> I have firewall PTSD
2023-08-14 14:45:35 <latex> But idk whether there's a firewall on OpenWRT or Alpine by default
2023-08-14 14:45:47 <ikke> Alpine does not enable one by default 
2023-08-14 14:45:53 <minimal> latex: so your Linux machine is directly connected to your Internet connection? there's no router?
2023-08-14 14:45:54 <ikke> Openwrt would
2023-08-14 14:46:00 <latex> There is a router
2023-08-14 14:46:02 <latex> OpenWRT
2023-08-14 14:46:10 <ikke> It has a firewall 
2023-08-14 14:46:13 <latex> But I have kept it at default settings for IPv6
2023-08-14 14:46:19 <latex> so idk whether it's firewalled or not
2023-08-14 14:46:26 <latex> The docs say it's not firewalled but idk for sure
2023-08-14 14:46:32 <minimal> latex: so to repeat my question, did you open up port 6969 on your (router's) firewall?
2023-08-14 14:46:40 <latex> Idk how to do that
2023-08-14 14:46:43 <latex> I'm new to IPv6
2023-08-14 14:46:59 <minimal> latex: the same way you would open up port 6969 for IPv4
2023-08-14 14:47:06 <Habbie> where do the docs say that there is no firewall?
2023-08-14 14:47:13 <Habbie> for example, https://openwrt.org/docs/guide-user/firewall/fw3_configurations/fw3_ipv6_examples
2023-08-14 14:47:17 <latex> Well NAT and firewalls are pretty different but I'll take a look
2023-08-14 14:47:19 <Habbie> very first example "open up a port"
2023-08-14 14:48:03 <latex> Well Habbie I saw in the docs that when configuring IPv6, one should be warned that all devices are directly exposed, so I assumed there was no firewall by default
2023-08-14 14:48:08 <latex> but idk I guess it's not accurate then?
2023-08-14 14:48:21 <Habbie> i don't know; which is why i keep asking you -where- you found specific things
2023-08-14 14:48:39 <grming> https://pastebin.com/X0cDRk2P
2023-08-14 14:48:41 <latex> my web browser is a clutter right now lol
2023-08-14 14:48:46 <latex> So I don't remember
2023-08-14 14:49:18 <latex> But I wonder whether it would be safe to open up everything IPv6
2023-08-14 14:49:29 <latex> Because there are many services that require open ports, e.g. BitTorrent
2023-08-14 14:49:43 <latex> So if I open up the network for everyone, all services work flawlessly
2023-08-14 14:49:58 <latex> But the downside is that malware could possibly exploit devices and immediately phone home with no issues
2023-08-14 14:50:16 <latex> But setting up a firewall makes services fail to work e.g. BitTorrent not working
2023-08-14 14:50:29 <latex> And I need to set up ports for every service. What a hastle
2023-08-14 14:50:33 <latex> I like opening up everything
2023-08-14 14:50:47 <latex> I doubt it really matters that much
2023-08-14 14:50:47 <grming> https://pastebin.com/Bq5FQjcJ
2023-08-14 14:51:03 <latex> Of course on IPv4 all Chinese hackers try to invade your server but on IPv6 there's no way to scan all of it right?
2023-08-14 14:51:06 <latex> all ranges?
2023-08-14 14:51:18 <latex> My logs are already full of CHina
2023-08-14 14:51:25 <grming> https://pastebin.com/RN4LVFkz
2023-08-14 14:51:26 <latex> Chinese SSH attempts
2023-08-14 14:51:26 <minimal> latex: having no firewall in place on your router is extremely insecure
2023-08-14 14:51:35 <latex> Why so?
2023-08-14 14:51:56 <latex> There's no magic port to remotely take over your Linux kernel right?
2023-08-14 14:51:59 <Habbie> because you're gambling that every device in your home is ready to be connected directly to the internet
2023-08-14 14:52:04 <grming> https://pastebin.com/mZLHwEHe
2023-08-14 14:52:29 <latex> Or should I make some zone to manually assign my servers to to be directly exposed?
2023-08-14 14:52:37 <Habbie> that's common, yes
2023-08-14 14:52:47 <Habbie> in v4 times we used to call that a DMZ
2023-08-14 14:52:55 <minimal> you should only expose the specific ports that should be exposed
2023-08-14 14:53:02 <Habbie> ^ but that is best
2023-08-14 14:54:08 <j_v> grming: i should have had skip the /sys and /proc directories, at the least
2023-08-14 14:59:53 <grming> according to the "tutorial" it should be easy
2023-08-14 15:00:05 <grming> untar the tar.gz on the SD card
2023-08-14 15:00:16 <grming> and copy the spl on the card
2023-08-14 15:18:47 <j_v> grming: perhaps someone else will be better suited to help.
2023-08-14 15:19:14 <bovis> ikke: Thanks again for the advice about ruby and sqlite. Compiling the gem locally got me past the posix_fallocate64 error.
2023-08-14 15:19:27 <ikke> bovis: glad to hear
2023-08-14 15:19:52 <ikke> bovis: consider filing an upstream issue
2023-08-14 15:20:25 <bovis> I will
2023-08-14 15:20:35 <grming> just a final question
2023-08-14 15:20:47 <grming> once I untar the file
2023-08-14 15:21:22 <grming> and copy the spl to sd, should I do something else? or like the tutorial said, plug the sd card in the board and boot the alpine
2023-08-14 15:21:25 <grming> ??
2023-08-14 15:22:05 <grming> the tarball 3.8 is the onlyone that has licheepi zero board?
2023-08-14 15:34:46 <j_v> grming: the one you mentioned seems to have the correct dtb file, and it boots. busybox --install -s fails for that /sbin links and /sbin/nlplug-findfs fail to run pointing to possible issue with missing /sbin or permissions of /sbin
2023-08-14 15:37:02 <j_v> grming: unpacked initramfs shows proper permissions on bin/busybox, sbin/nlplug-findfs and even sbin directory
2023-08-14 15:39:28 <grming> i will try that
2023-08-14 15:46:09 <j_v> using nsjail and qemu-arm, i have kind of 'chrooted' into the broke out initramfs. will see if i can figure any thing out from that
2023-08-14 16:00:17 <grming> just a dumb question
2023-08-14 16:00:30 <grming> how do I unpack the initramfs???
2023-08-14 16:01:02 <grming> this should give any error, due to the boot unable to locate the initramfs?
2023-08-14 16:01:31 <grming> i guess the boot just unpacket whle installing 
2023-08-14 16:03:53 <j_v> the error in your first paste are from the initramfs, so it is getting loaded
2023-08-14 16:05:08 <j_v> for reference, to unpack initramfs: `mkdir initramfs ; cd initramfs ; gzip -cd < ../initramfs-lts | cpio -i`
2023-08-14 17:00:43 <grming> well i unpacked the initramfs and got this
2023-08-14 17:00:44 <grming> https://pastebin.com/iaVY5mtv
2023-08-14 17:19:34 <j_v> seems to be a private paste
2023-08-14 17:38:43 <grming> its not booting
2023-08-14 17:40:36 <grming> i already had linux kernel 4.3 runing on another SD card, but its too old and tried to use alpine in this board
2023-08-14 18:52:23 <Zeroine> Hello
2023-08-14 18:53:46 <Zeroine> Does anyone here use systemd on alpine? I wonder if there are any issues
2023-08-14 18:55:09 <Zeroine> well considering alpine deviates from what other main stream distros have, like musl and busybox instead of glibc and gnu coreutils
2023-08-14 18:55:32 <Zeroine> not sure how much if at all that even matters to an init systemd like systemd
2023-08-14 18:56:49 <minimal> Zeroine: Alpine does not support nor package systemd
2023-08-14 18:57:25 <Zeroine> yeah I know it's not in the repos, I'm fine with compiling it if needed
2023-08-14 18:58:05 <minimal> why would you want to run systemd with Alpine?
2023-08-14 18:58:23 <quinq> For the power of modern™ technology
2023-08-14 19:00:05 <minimal> Zeroine: enjoy your patching then...
2023-08-14 19:00:38 <quinq> 20:55:32  Zeroine$ not sure how much if at all that even matters to an init systemd like systemd
2023-08-14 19:00:53 <quinq> It matters in the sense that it relies on glibc implementation specifics
2023-08-14 19:01:03 <quinq> And they're not interested to change that
2023-08-14 19:01:09 <minimal> upstream systemd assumes it is used with glibc and therefore musl use requires various patching
2023-08-14 19:01:10 <Habbie> you'll have a better time trying procd, i bet
2023-08-14 19:01:36 <Zeroine> Because I assume it will be more common to encounter it in jobs. Correct me if I'm wrong but I also think systemd is faster, right? They do more things in parallel or something like that. I'm not sure.
2023-08-14 19:01:49 <Habbie> encountering systemd in jobs, surely
2023-08-14 19:01:54 <Habbie> encountering systemd on alpine in jobs, not so likely
2023-08-14 19:02:42 <Zeroine> yeah
2023-08-14 19:03:04 <Zeroine> quinq: ah okay, thank you
2023-08-14 19:03:29 <Zeroine> guess I'll stick with OpenRC then
2023-08-14 19:05:51 <quinq> I like runit, but openrc works
2023-08-14 19:07:17 <zcrayfish> most people I see comparing systemd's speed with openrc are doing an apples to oranges comparison by having openrc starting services in a serial manner (which is the default).
2023-08-14 19:09:37 <zcrayfish> Folks that are looking for speed could put rc_parallel="YES" into /etc/rc.conf; though it´s unsupported by the maintainers. 
2023-08-14 19:10:27 <Habbie> also, if startup is dominated by that difference, it can probably be improved some other way
2023-08-14 19:10:57 <zcrayfish> Indeed.
2023-08-14 19:13:20 <minimal> it also depends on whether you're fixated on boot speed or not - how often do you boot a machine? (obviously more important for things like microvms)
2023-08-14 19:18:08 <Zeroine> I mean speed isn't that much of a concern, I just asked because systemd is so common, more documentation probably for it and I prefer the unit files they use rather than shell scripts
2023-08-14 19:19:02 <Zeroine> but whatever I don't mind using whatever alpine has
2023-08-14 19:41:50 <m_ben> why does alpine linux still use the old naming scheme for network interfaces wlan0, eth0 etc.? is it because mdev is preferred over udev?
2023-08-14 19:42:29 <ikke> probably because the names are simpler and easier to remember
2023-08-14 19:42:52 <ikke> and you'd use something like macifrename if you want deterministic name
2023-08-14 19:42:55 <ikke> names
2023-08-14 19:48:07 <j_v> Habbie: when you mentioned procd, were you referring to openwrt's procd?
2023-08-14 19:48:14 <Habbie> yes
2023-08-14 19:51:22 <j_v> ah, been running openwrt on my routers for a while. very easy to setup. i still prefer to must of my setup editing the config files, but luci is pretty nice too.
2023-08-14 19:51:48 <j_v> s/to must/to do most/
2023-08-14 19:53:41 <j_v> can't say i know that much about procd, but i perused the source code a few times, and it is amazing small, or seems so to me.
2023-08-14 19:53:53 <Habbie> yeah, especially as it actually manages processes
2023-08-14 19:54:47 <j_v> do you know anyone who has used it on a non-embedded platform?
2023-08-14 19:54:53 <Habbie> i don't
2023-08-14 19:55:08 <Habbie> well, some people run openwrt on bigger devices, but i'm guessing that's not what you meant
2023-08-14 19:57:04 <j_v> no, i guess not. but i have been meaning to setting s6 for a while now, not so much for init, but for supervision... i should a round tuit.
2023-08-14 19:58:27 <j_v> and smaller and more modular inits and system supervision fascinates me
2023-08-14 19:58:58 <j_v> is why i feel right at home with openrc
2023-08-14 20:30:00 <testburgers> does anyone know why alpine is refusing to work in a fresh pi4b install after trying to use xfce?
2023-08-14 20:30:20 <testburgers> I used the default installers and yet it constantly says I have no video drivers.
2023-08-14 20:30:34 <testburgers> mesa-dri-vc4 doesn't seem to be available either
2023-08-14 20:34:41 <kwadroke[m]> perhaps install mesa-dri-gallium https://wiki.alpinelinux.org/wiki/Raspberry_Pi
2023-08-14 20:41:04 <testburgers> I've tried adding that but it doesnt work for startx, I noticed some other instructions but the directions suggest modifying a config file I had to remove from the pi for the install to even boot at all
2023-08-14 20:41:25 <testburgers> the entire flashcard needs to be removed or the pi wont start.
2023-08-14 20:41:41 <testburgers> it will just loop back into installing fresh alpine
2023-08-14 20:42:26 <testburgers> after investigating the media folder, the usb directory is empty, and pis dont have floppy or cdrom.
2023-08-14 21:06:42 <zcrayfish> the three directories are created there regardless.
2023-08-14 21:06:52 <zcrayfish> on all arches and flavours that I know of.
2023-08-14 21:06:54 <zcrayfish> FWIW
2023-08-14 21:12:39 <minimal> m_ben: because that's what it does. Also even if you install eudev on Alpine the (standard) initramfs will still use mdev
2023-08-14 21:15:56 <testburgers> crap
2023-08-14 21:16:19 <testburgers> I got all this working at last, but I just realized barrier isn't available on alpine for raspberry pi, just x86
2023-08-14 21:17:52 <minimal> testburgers? barrier is packages for all archs.....it is however only in testing repo, so only on Alpine Edge
2023-08-14 21:21:17 <testburgers> it is? is there a way i can make my alpine edge? or do I have to restart all over?
2023-08-14 21:21:46 <durrendal> you just need to change you repos in /etc/apk/repos
2023-08-14 21:23:11 <durrendal> http://dl-cdn.alpinelinux.org/alpine/edge/testing <- add this to the end of the repositories file, and change the existing 3.x to edge
2023-08-14 21:23:29 <durrendal> that should be /etc/apk/repositories not repos
2023-08-14 21:39:20 <txnintn> <- brain fade, manual install on RPI4, created 8G volume group, boots fine, if try to add a mirror, it loses a leg.  What am I missing?
2023-08-14 21:39:52 <txnintn> Loses leg on reboot.  
2023-08-14 22:01:49 <minimal> what does "looses a leg" mean?
2023-08-14 22:03:58 <txnintn> one of the lvm mirrored drives fail to load during boot, I guess it term came from "doesnt have a leg to stand on" just a guess
2023-08-14 22:19:02 <txnintn> Could it need fstab updated from /dev/lg0/lv_name to /dev/mapper/vg0/lv_name?
2023-08-14 22:19:53 <dwfreed> both of those are symlinks to the same file
2023-08-14 22:21:56 <minimal> txnintn: is the rootfs on the LVM mirror? If so then this will be handled (i.e. mounted) by the initramfs' init at boot
2023-08-14 22:24:02 <txnintn> I'm using a single / root partition, then I add the mirror but I didn't dd the mirror.
2023-08-14 22:27:24 <txnintn> So if I dd the 8G drive to the mirror it should be good?
2023-08-14 22:30:28 <minimal> how did you add the mirror?
2023-08-14 22:35:22 <txnintn> micro is SD, used sfdisk to copy partition layout, fat32 on partition 1, with boot flags, partition 2 is lvm
2023-08-14 22:39:06 <txnintn> Mirror steps, pvcreate /dev/mmcblk2, lvcreate -L 8G -m1 -n lv_mirror
2023-08-14 22:40:38 <minimal> you have a RPI4 with 2 SDcards in use at same time?
2023-08-14 22:41:07 <txnintn> No, using one usb and one mmcblk
2023-08-14 22:41:38 <minimal> and booting off which? (from a RPI settings perspective)
2023-08-14 22:41:45 <txnintn> Usb
2023-08-14 22:41:59 <txnintn> then mmcblk
2023-08-14 22:48:23 <txnintn> back in 20 mins.  dinner on the table.
2023-08-14 23:08:20 <txnintn> back
2023-08-15 01:04:09 <livy> Hello. Does anyone know how to stop network device name from changing on each boot?
2023-08-15 01:04:38 <livy> I have a bridge 'br0', which has a member interface 'eth2' which is the LAN port on the motherboard.
2023-08-15 01:04:57 <livy> I also have several PCIe network cards installed on the motherboard.
2023-08-15 01:05:37 <livy> Sometimes the LAN port just becomes 'eth1', which is currently not in used, and cause the bridge to be DOWN.
2023-08-15 01:06:15 <livy> In Ubuntu they use naming scheme which reflects the logical position (e.g: enp2s0) to prevent that from happening.
2023-08-15 01:10:18 <somercet1> livy, udev rules can fix a device name
2023-08-15 01:13:01 <livy> So first of all I have to change the default mdev to udev (or eudev)?
2023-08-15 01:48:11 <minimal> livy: there is the macifrename package
2023-08-15 01:49:58 <minimal> for which you can create a file indicating what each interface should be named (based on its MAC address)
2023-08-15 01:56:07 <bovis> livy: I can't be of much help here, but there is this that could point you in the right direction: https://wiki.alpinelinux.org/wiki/Custom_network_interface_names
2023-08-15 03:04:22 <bovis> livy: https://gist.github.com/jirutka/9496300 "Renaming Network Interfaces" This might also help. Looks like you can't rename to anything etc[0-9], but if your bridge can recognize another name you set, this could work.
2023-08-15 04:58:10 <livy> Thanks, I'll take a look at it. I have a virtual network for QEMU virtual machine. Luckily none of my servers has PCIe network cards.
2023-08-15 04:58:29 <livy> They only have 4 internet Ethernet ports, and luckily they have been working correctly.
2023-08-15 04:59:00 <livy> I am having issues on my test PC, but I need to be careful to avoid the same issue on a production server in the future.
2023-08-15 12:22:38 <dzilvys> anyone know why I potentially am getting the following error when trying to mount a squashfs ? mount: mounting /dev/loop0 on tmp failed: Invalid argument
2023-08-15 12:27:06 <ikke> Try to explicitly specify the de type
2023-08-15 12:27:12 <ikke> The fs type
2023-08-15 12:27:21 <ikke> -t squashfs
2023-08-15 12:31:57 <dzilvys> ikke: wonderful it worked
2023-08-15 12:38:00 <grming> anybody knows if there is a IRC channel for u-boot?
2023-08-15 12:40:49 <ovf> i wonder if it's worth having a simple shell script wrapper for busybox mount that does blkid and modprobe, just like util-linux's mount
2023-08-15 13:45:15 <j_v> trying to get urlwatch to the re.sub filter, keeps giving error: ValueError: re.sub needs a pattern
2023-08-15 13:46:10 <j_v> urls.yaml: https://pastebin.com/cs2x797h
2023-08-15 14:08:40 <j_v> starting to think urlwatch not so great
2023-08-15 15:34:58 <handlerug> don't you need to indent lines 7 and 8 more
2023-08-15 15:35:05 <handlerug> just a guess
2023-08-15 15:44:03 <j_v> handlerug: sure enough, that was the problem. thank you!
2023-08-15 16:02:55 <handlerug> yeah that's yaml being yaml :)
2023-08-15 16:05:16 <j_v> only written a little when i was some advanced lxc template stuff, apparently i forgot much i what i learned from that.
2023-08-15 20:11:15 <elly> the 3.18 install I have on my desktop (which I haven't booted in a while...) is hanging at "Loading hardware drivers ..." during boot
2023-08-15 20:11:22 <elly> how (if at all) does one debug this?
2023-08-15 20:12:58 <Habbie> i know nothing, but can you do anything about verbosity?
2023-08-15 20:13:11 <elly> unfortunately removing `quiet` from the grub command line doesn't help, since the trouble is happening after starting OpenRC
2023-08-15 20:13:56 <Habbie> ah
2023-08-15 20:14:30 <elly> time for single-user mode maybe
2023-08-15 20:14:32 <m_ben> does anyone use rEFInd? how did you install it? seems like you need to do it manually
2023-08-15 20:14:59 <Habbie> m_ben, i use it but i did not install it from alpine
2023-08-15 20:15:07 <zcrayfish> elly: rc_interactive="YES" in /etc/rc.conf would be my first choice
2023-08-15 20:15:23 <zcrayfish> rc_logger may be useful too.
2023-08-15 20:15:51 <elly> oho, thanks
2023-08-15 20:15:57 <elly> I'll boot into single and try that
2023-08-15 20:17:10 <elly> that did not help, I think that when it boots into single I am still in the initramfs
2023-08-15 20:20:02 <elly> bless this initramfs though
2023-08-15 20:26:30 <Habbie> i have a terrible response that in no way should be taken seriously: why? are you on a mac?
2023-08-15 20:26:36 <Habbie> (this joke has a limited audience)
2023-08-15 20:26:54 <elly> for future reference, you can supply `noautodetect` on the kernel command-line to disable coldplug device detection
2023-08-15 20:27:03 <elly> which prevented my system from hanging during boot
2023-08-15 20:27:11 <elly> now, exactly which driver was hanging it is still a very good question
2023-08-15 21:42:44 <zcrayfish> 
ACTION drops an eMac G4 on Habbie

2023-08-15 21:42:46 <zcrayfish> =)
2023-08-15 21:46:31 <Habbie> 
ACTION dies

2023-08-15 21:47:22 <zcrayfish> "must be the radiation"
2023-08-15 22:55:09 <zcrayfish> Don't you just love it when you have tainted the kernel for REASONS but need to get it back to normal for troubleshooting? :p
2023-08-15 23:09:40 <zcrayfish> oof. I guess the i915 driver on the lastest revision of Edge has a little trouble with the Iris Xe. \o/
2023-08-15 23:10:32 <Xe> for what it's worth
2023-08-15 23:10:38 <zcrayfish> LOL sorry Xe
2023-08-15 23:10:42 <Xe> this is the first time that GPU has been mentioned here
2023-08-15 23:10:42 <zcrayfish> 
ACTION blames Intel =)

2023-08-15 23:10:45 <Xe> trust me, i know
2023-08-15 23:12:56 <zcrayfish> I'll try to refer to it as "That lame iGPU inside Tiger Lake" ;)
2023-08-15 23:13:20 <zcrayfish> Incidentally it's not bad for an iGPU, it's just that the driver kind of sucks on the current revision of Edge. (at no fault of this distro)
2023-08-15 23:38:03 <zcrayfish> GµC context submission. that was the problem. Not supported on this generation. Curiously had it turned on without issue before all year. lol. 
2023-08-15 23:38:15 <zcrayfish> Glad the HEVC/H.265 micro controller still works though. Will take what I can get
2023-08-16 02:46:29 <Saijin_Naib[m]> I hope to go Xe/ARC soon
2023-08-16 02:46:47 <Saijin_Naib[m]> Dumped all my NVIDIA when I went Alpine, but now I just have my iGPU from my 6700k which is... Terri-bad
2023-08-16 02:47:55 <Xe> i can't wait until they kill that line
2023-08-16 02:49:30 <Saijin_Naib[m]> Which ones?
2023-08-16 03:05:49 <zcrayfish> X e ;p
2023-08-16 03:05:58 <zcrayfish> So that we can stop highlighting his nick every 2 seconds \o/
2023-08-16 03:31:32 <lambdanil> pretty new to Alpine but I'm having a good time so far
2023-08-16 03:31:53 <lambdanil> also I got my Debian chroot working with bwrap so I can run Android Studio
2023-08-16 03:38:30 <Saijin_Naib[m]> hah, awesome! I must admit I've not learned enough about chroot vs distrobox to really make use of either
2023-08-16 03:43:46 <Zeroine> Saijin_Naib[m]: looks like distrobox is just a wrapper for creating containers with podman or docker, I've never used it btw
2023-08-16 03:44:10 <Saijin_Naib[m]> Ah... I thought it was more like a chroot that doesn't need root/doas/sudo
2023-08-16 03:44:17 <Saijin_Naib[m]> See, I have not read enough yet, haha
2023-08-16 03:44:45 <Zeroine> well you can run containers rootless, at least on docker, I think on podman too
2023-08-16 05:47:04 <grawity> Saijin_Naib[m]: Linux "user namespaces" are your chroot that doesn't need root/doas/sudo, some container managers already support those, usually under the name "unprivileged containers"... userns needs to be enabled in the kernel though
2023-08-16 06:26:42 <livy> Hi. I need to use a FTP client in CLI mode. Which package should I download?
2023-08-16 06:31:33 <zcrayfish> inetutils-ftp
2023-08-16 06:32:13 <zcrayfish> (if you need something from main, curl is a great download-only client)
2023-08-16 06:32:35 <zcrayfish> you can upload with it too, I just personally find it more annoying than the traditional client.
2023-08-16 06:32:47 <livy> For downloading, I think wget can do that. But I need to list, put, and push files.
2023-08-16 06:33:00 <zcrayfish> definitely inetutils-ftp from community.
2023-08-16 06:36:16 <zcrayfish> 
ACTION wonders how soon gopher servers will outnumber ftp servers due to everyone dropping support for the latter lately

2023-08-16 07:57:38 <somercet1> livy, I like lftp.
2023-08-16 08:05:07 <quinq> Yeah, lftp is nice
2023-08-16 08:08:52 <ovf> there's also curlftpfs for when you've got a short fuse
2023-08-16 08:34:43 <lopid> long fuse transfer protocol
2023-08-16 14:31:12 <ascensionwires> i installed alpine with these commands: (pc connected access to a wired network)\nsetup-alpine\nanswered "none" when asked for the disk to use, for where to store configs, and for the location of the package cache\napk update\napk add util-linux-misc e2fsprogs\nfdisk /dev/sdb\n(created one "Linux" type partition with boot flag)\nmkfs.ext4 /dev/sdb1\napk del util-linux-misc e2fsprogs\nmount 
2023-08-16 14:31:18 <ascensionwires> /dev/sdb1 /mnt\nvi /etc/apk/repositories\n(uncommented community repo)\napk update\nBOOTLOADER=grub setup-disk -k edge /mnt\nthen i rebooted and\nlogin as root\nsed -i -e 's/v3\\.18/edge/g' /etc/apk/repositories\napk update\napk add --upgrade apk-tools\napk upgrade --available\nsync\nreboot\nand i get this error:\n* Remounting / read only ...\n* failed because we are using /\n* ERROR: 
2023-08-16 14:31:24 <ascensionwires> mount-ro failed to start
2023-08-16 14:31:46 <ascensionwires> sorry bad paste
2023-08-16 14:32:25 <ascensionwires> i installed alpine with these commands: (pc connected access to a wired network)
2023-08-16 14:32:28 <ascensionwires> setup-alpine
2023-08-16 14:32:31 <ascensionwires> answered "none" when asked for the disk to use, for where to store configs, and for the location of the package cache
2023-08-16 14:32:34 <ascensionwires> apk update
2023-08-16 14:32:36 <ascensionwires> apk add util-linux-misc e2fsprogs
2023-08-16 14:32:39 <ascensionwires> fdisk /dev/sdb
2023-08-16 14:32:41 <ascensionwires> (created one "Linux" type partition with boot flag)
2023-08-16 14:32:44 <ascensionwires> mkfs.ext4 /dev/sdb1
2023-08-16 14:32:46 <ascensionwires> apk del util-linux-misc e2fsprogs
2023-08-16 14:32:49 <ascensionwires> mount /dev/sdb1 /mnt
2023-08-16 14:32:51 <ascensionwires> vi /etc/apk/repositories
2023-08-16 14:32:54 <ascensionwires> (uncommented community repo)
2023-08-16 14:32:56 <ascensionwires> apk update
2023-08-16 14:32:59 <ascensionwires> BOOTLOADER=grub setup-disk -k edge /mnt
2023-08-16 14:33:01 <elly> please don't paste a bunch of text here
2023-08-16 14:33:01 <ascensionwires> then i rebooted and
2023-08-16 14:33:04 <ascensionwires> login as root
2023-08-16 14:33:06 <elly> use a pastebin
2023-08-16 14:33:07 <ascensionwires> sed -i -e 's/v3\.18/edge/g' /etc/apk/repositories
2023-08-16 14:33:09 <ascensionwires> apk update
2023-08-16 14:33:11 <ascensionwires> apk add --upgrade apk-tools
2023-08-16 14:33:14 <ascensionwires> apk upgrade --available
2023-08-16 14:33:16 <ascensionwires> sync
2023-08-16 14:33:19 <ascensionwires> reboot
2023-08-16 14:33:22 <ascensionwires> and i get this error:
2023-08-16 14:33:24 <ascensionwires> * Remounting / read only ...
2023-08-16 14:33:27 <ascensionwires> * failed because we are using /
2023-08-16 14:33:29 <ascensionwires> * ERROR: mount-ro failed to start
2023-08-16 14:33:32 <ascensionwires> sorry
2023-08-16 14:34:10 <ascensionwires> ok 
2023-08-16 14:34:14 <ascensionwires> https://pastebin.com/raw/Tr9sEE1s
2023-08-16 14:35:15 <ascensionwires> why this error?
2023-08-16 15:09:40 <ascensionwires> why does this error occur?
2023-08-16 15:14:05 <Ermine> Did it reboot afterwards?
2023-08-16 15:26:13 <ascensionwires> error as i sended shows on rebooting poweroff step
2023-08-16 15:26:32 <ascensionwires> Remounting / read only ...
2023-08-16 15:26:32 <ascensionwires> * failed because we are using /
2023-08-16 15:26:42 <ascensionwires> * ERROR: mount-ro failed to start
2023-08-16 15:27:27 <ascensionwires> as i sended i changed repos to edge then upgraded
2023-08-16 15:27:41 <ascensionwires> mb it related to kernel upgrade
2023-08-16 15:28:03 <ascensionwires> because of it its cant remount /
2023-08-16 15:29:37 <ascensionwires> Ermine , it rebooted and work (i think adequately)
2023-08-16 15:30:55 <ascensionwires> but when rebooting (in poweroff stage of rebooting it showed this error i think becauae upgrade to edge include kernel updare)
2023-08-16 15:32:52 <zcrayfish> hmm
2023-08-16 15:32:57 <zcrayfish> that looks awfully familiar.
2023-08-16 15:34:05 <Ermine> ascensionwires: I have this error too, but apparently it doesn't affect anything, so it can be ignored.
2023-08-16 15:36:02 <zcrayfish> There's a way to fix it, but darn if I can remember
2023-08-16 15:36:07 <zcrayfish> 
ACTION is looking

2023-08-16 15:36:33 <minimal> I thought I'd raised a MR to "fix" this a while ago
2023-08-16 15:39:07 <minimal> though maybe it was the LVM equivalent "error" that I'm thinking of
2023-08-16 15:43:13 <zcrayfish> ascensionwires: I would like you to try something.
2023-08-16 15:43:30 <zcrayfish> edit the file /etc/conf.d/localmount
2023-08-16 15:43:47 <zcrayfish> change the third line to:
2023-08-16 15:43:50 <zcrayfish> no_umounts="/tmp"
2023-08-16 15:44:02 <zcrayfish> See if that fixes your problem.
2023-08-16 15:45:20 <zcrayfish> 
ACTION doesn't see why it would, but seems to remember the fix being something extremely unintuitive.

2023-08-16 15:47:55 <minimal> zcrayfish: wouldn't that be /etc/conf.d/mount-ro?
2023-08-16 15:48:40 <zcrayfish> No, I don't think the mount-ro service is the problem at all, it's just the one complaining.
2023-08-16 15:52:02 <minimal> ah, it is /lib/rc/sh/rc-mount.sh that handles this
2023-08-16 15:55:31 <minimal> a "fuser" issue perhaps? (that seem to sound familiar)
2023-08-16 15:56:30 <ascensionwires> zcrayfish m talking about this because when updating the kernel, grub generate config every time and for some reason it cant detect other oses in this moment
2023-08-16 15:57:42 <ascensionwires> windows on anotger disk and it cant find it when generating grub config during upgrade kernel or some
2023-08-16 15:58:10 <ascensionwires> but i can generate config by myself after rebooting
2023-08-16 15:58:14 <ascensionwires> and it working
2023-08-16 15:58:42 <ascensionwires> for some reason it cant do something with disks when upgrade like thia
2023-08-16 15:58:59 <ascensionwires> mb it not cuz kernel upgrade
2023-08-16 15:59:15 <ascensionwires> but i see this issues only when kernel upgrades
2023-08-16 16:03:30 <zcrayfish> oh bummer. Using grub at all is not something I do, or miss. hopefully someone else can think of something.
2023-08-16 16:05:16 <minimal> Grub does detect other OSes, you need os-prober installed
2023-08-16 16:05:22 <minimal> s/does/does not/
2023-08-16 16:05:34 <ascensionwires> minimal
2023-08-16 16:05:43 <minimal> ?
2023-08-16 16:05:58 <ascensionwires> im about when grub-mkconfig trigger?
2023-08-16 16:06:06 <ascensionwires> when kernel update
2023-08-16 16:06:16 <ascensionwires> *upgrade
2023-08-16 16:06:17 <minimal> I gave you an answer
2023-08-16 16:06:55 <ascensionwires> i about os prober cant see oses when this upgrade
2023-08-16 16:07:15 <minimal> I do not understand what you are saying
2023-08-16 16:07:28 <ascensionwires> also when this upgrade some ERROR: mount-ro failed to start
2023-08-16 16:07:49 <minimal> and you are also talking about 2 unrelated things at the same time
2023-08-16 16:07:50 <ascensionwires> like it wrong if kernel upgrade
2023-08-16 16:08:10 <ascensionwires> this related things
2023-08-16 16:08:36 <ascensionwires> cuz grub mkconfig working after reboot
2023-08-16 16:09:35 <ascensionwires> i can apk add amd-ucode (this also triggers grub-mkconfig) and it see other os 
2023-08-16 16:12:21 <ascensionwires> minimal this things 100% related
2023-08-16 16:13:27 <ascensionwires> it cant see other os when after some upgrade some ERROR: mount-ro failed to start
2023-08-16 16:14:15 <ascensionwires> i think its wrong after kernel upgrade fix me if am wrong
2023-08-16 16:31:32 <ascensionwires> its not critical as upgrade some ERROR: mount-ro failed to start
2023-08-16 16:33:20 <ascensionwires> i think its not critical as zcrayfish mentioned, i can manually call grub-mkconfig after rebpot and it work
2023-08-16 17:02:32 <o101010> Hi everyone, I tried to register on gitlab.alpinelinux.org but I never received confirmation email. I have tried last week and today. Does anyone have information about this issue ?
2023-08-16 17:31:06 <Habbie> o101010, i have no information, but i can tell you the github login button works fine - assuming that's acceptable to you
2023-08-16 17:45:18 <ascensionwires> how add modules? https://wiki.alpinelinux.org/wiki/Kernel_Modesetting
2023-08-16 17:52:36 <zcrayfish> ascensionwires: which GPU?
2023-08-16 17:59:08 <ascensionwires> i cant find for intel on wiki
2023-08-16 18:04:29 <zcrayfish> the -lts and -edge kernel packages both contain the i915 package.
2023-08-16 18:04:57 <zcrayfish> which covers MOST intel GPUs and iGPUs, except for perhaps some of the horrible PowerVR-based ones last time I checked.
2023-08-16 18:06:01 <zcrayfish> Beyond that: https://wiki.alpinelinux.org/wiki/Intel_Video#Kernel_Modesetting_(KMS)
2023-08-16 18:06:28 <ascensionwires> there nothing
2023-08-16 18:07:42 <zcrayfish> By design: Typically there is very little you need to do to get Intel graphics running.
2023-08-16 18:08:57 <ascensionwires> but on amd there intructions with adding fbcon and   amdgpu or radeon to /etc/modules, tgen 
2023-08-16 18:09:16 <ascensionwires> then adding kms to mkinitfs
2023-08-16 18:10:09 <zcrayfish> the AMD article looks like it's slightly out of date, but it's also not relevant to Intel at all.
2023-08-16 18:11:02 <zcrayfish> I think the only thing the Intel article needs is a mention that GµC-equipped models should have the firmware installed as well.
2023-08-16 18:11:09 <zcrayfish> (but even they should work without it)
2023-08-16 18:23:24 <ascensionwires> why kms intel link dont show any about kms
2023-08-16 18:24:14 <zcrayfish> Because it's on by default, at least on PCs.
2023-08-16 18:25:44 <ascensionwires> but why this url on wiki if fragment with thus hashtag actually deleted from intel video page?
2023-08-16 18:25:47 <ascensionwires> nm
2023-08-16 18:26:38 <o101010> @Habbie, I have more the feeling that thes problem come with the email server sender
2023-08-16 18:26:42 <zcrayfish> Because back in 2015 when the article was originally written, the stuff had to be done manually as you are expecting for some reason.
2023-08-16 18:27:17 <zcrayfish> https://wiki.alpinelinux.org/w/index.php?title=Intel_Video&oldid=10914
2023-08-16 18:30:21 <ascensionwires> ok
2023-08-16 18:31:56 <ascensionwires> m behaving badly sorry for my stuffiness pleaae (T.T)
2023-08-16 22:08:56 <Mari[m]> Hey I deleted my boot folder by mistake
2023-08-16 22:09:07 <Mari[m]> How can I fix this and have a bootable system again?
2023-08-16 22:14:38 <XFaCiEer> Boot from USB and chroot, or something.
2023-08-16 22:17:00 <elly> yup - boot a live alpine environment from USB, mount and chroot into your system, then apk fix grub - although you might need to edit the resulting grub config
2023-08-16 22:39:42 <Mari[m]> Ummm it won't let me mount /boot
2023-08-16 22:39:57 <Mari[m]> 
ACTION uploaded an image: (80KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/SNUntKLelyazdacBYklWmcQd/1000023760.jpg >

2023-08-16 22:40:46 <Mari[m]> 
ACTION uploaded an image: (106KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/ARVYOzvsFqCHTYTcULWSsaGn/1000023761.jpg >

2023-08-16 22:40:50 <Mari[m]> Here we go
2023-08-16 22:41:03 <Mari[m]> I can't mount the boot partition on /boot
2023-08-16 22:41:15 <elly> oh yeah? why not?
2023-08-16 22:41:37 <Mari[m]> Idk that's why I asked
2023-08-16 22:42:08 <Mari[m]> It just says mount /dev/nvme0n1p1 on /boot failed invalid argument
2023-08-16 22:46:29 <Mari[m]> Ok I fixed it by formatting the partition but now I can't install grub
2023-08-16 23:12:58 <Mari[m]> Fixed it
2023-08-16 23:13:13 <Mari[m]> I had to chroot and reinstall linux-lts and grub
2023-08-16 23:43:34 <Saijin_Naib[m]> grawity sweet, thank you. Do you know if we do enable that? 
2023-08-17 07:07:56 <wyk72> hi 
2023-08-17 07:08:17 <wyk72> I have a tricky question for Alpine Wizards
2023-08-17 07:09:47 <wyk72> I had to use the (infamous?)  glibc pkg from user sgerrand on github, that installs a "true" glibc binary into Alpine
2023-08-17 07:10:30 <wyk72> can somebody explain to me why it works 100% an Alpine 3.17 and lower but segfaults on Alpine 3.18 and higher ?
2023-08-17 07:11:42 <wyk72> maybe it has something to do with the "DT_RELR" optimization I've read on changelog ?
2023-08-17 07:11:51 <wyk72> on 3.18 changelog
2023-08-17 07:12:51 <wyk72> (I know it's a silly question about an..hack, basically, don't burst into flames pls)
2023-08-17 07:13:52 <wyk72> any glibc binary on 3.18 segafaults with "Code: Unable to access opcode bytes at 0x3566", but works perfectly on 3.17 (and lower)
2023-08-17 08:12:52 <dzilvys> Does anyone know if the following mirror is down? http://ftp.icm.edu.pl/
2023-08-17 08:16:21 <dzilvys>   WARNING: Ignoring http://ftp.icm.edu.pl/pub/Linux/distributions/alpine/v3.17/main: No such file or directory
2023-08-17 08:16:21 <dzilvys> randomly about an hour ago I got the following
2023-08-17 08:16:21 <dzilvys> ERROR: http://ftp.icm.edu.pl/pub/Linux/distributions/alpine/v3.17/main: network error (check Internet connection and firewall)
2023-08-17 08:16:21 <dzilvys> 08:06:32 
2023-08-17 08:16:22 <dzilvys> 08:06:32 
2023-08-17 08:16:23 <dzilvys>   fetch http://ftp.icm.edu.pl/pub/Linux/distributions/alpine/v3.17/community/x86_64/APKINDEX.tar.gz
2023-08-17 08:16:25 <dzilvys> 08:08:41 
2023-08-17 08:16:28 <dzilvys>   ERROR: http://ftp.icm.edu.pl/pub/Linux/distributions/alpine/v3.17/community: network error (check Internet connection and firewall)
2023-08-17 08:16:30 <dzilvys> 08:08:41 
2023-08-17 08:16:32 <dzilvys>   WARNING: Ignoring http://ftp.icm.edu.pl/pub/Linux/distributions/alpine/v3.17/community: No such file or directory
2023-08-17 08:20:24 <rnalrd> https://downforeveryoneorjustme.com/ftp.icm.edu.pl
2023-08-17 08:24:10 <dzilvys> damn alright thank you
2023-08-17 08:24:23 <dzilvys> is therre a way to make sure alpine is always using a valid mirror?
2023-08-17 08:24:39 <aron> I suggest to use dl-cdn.alpinelinux.org
2023-08-17 08:25:17 <aron> i.e. http://dl-cdn.alpinelinux.org/alpine/v3.17/main
2023-08-17 08:25:24 <dzilvys> thank you
2023-08-17 08:28:49 <dzilvys> aron: is there a way to change the mirror via command instead of modifying the config file?
2023-08-17 08:29:15 <aron> dunno, I always edit the file directly
2023-08-17 08:29:20 <aron>  /etc/apk/repositories
2023-08-17 08:29:39 <aron> it's a good ol' text file
2023-08-17 08:30:15 <aron> (you can place comments and empty lines as well, as usual)
2023-08-17 08:30:22 <zcrayfish> setup-apkrepos ???
2023-08-17 08:30:41 <zcrayfish> TBH I always just edit the file. hah
2023-08-17 08:30:52 <aron> lol
2023-08-17 08:32:24 <dzilvys> that might work for me actually
2023-08-17 08:32:25 <dzilvys> tyty
2023-08-17 08:57:31 <dzilvys> Does anyone know where dl-cdl.alpinelinux.org is located?
2023-08-17 08:57:55 <grawity> it's a CDN, so "everywhere"
2023-08-17 08:58:12 <grawity> "wherever Fastly has presence"
2023-08-17 08:58:28 <zcrayfish> DNS round robin + anycast I would assume. 
2023-08-17 08:58:51 <dzilvys> oh so even if im in the UK it will just try to get best speed from the closest fastly server?
2023-08-17 08:58:57 <zcrayfish> yep
2023-08-17 09:00:11 <dzilvys> cool thank you :)
2023-08-17 09:00:31 <grawity> run a mtr or traceroute, or visit https://www.fastly-debug.com/, to get a rough idea of where it takes you
2023-08-17 09:00:47 <grawity> oh they broke fastly-debug
2023-08-17 09:24:53 <quinq> Apparently the closest to me (Spain) is San Francisco, I suppose I should trust fastly to route correctly ;p
2023-08-17 09:26:10 <quinq> Ah no, it's somewhere else, not listed in https://ip-geolocation.fastly.com/
2023-08-17 09:26:43 <quinq> (though I modified the hostname from dl-cdl.alpinelinux.org to dl-cdn.alpinelinux.org, assuming that's the actual target)
2023-08-17 09:27:45 <quinq> Yeah, neither on IPv4 nor IPv6
2023-08-17 10:05:28 <grawity> you can also do `curl -H Fastly-Debug:1 -I https://dl-cdn.alpinelinux.org` to see what's responding to the specific request
2023-08-17 11:08:58 <quinq> grawity, what's useful there, how to use it?
2023-08-17 11:09:21 <quinq> I can see "x-served-by: cache-ams12740-AMS, cache-mad22031-MAD
2023-08-17 11:09:27 <quinq> for example, buuut
2023-08-17 11:47:25 <grming> hi
2023-08-17 16:41:31 <grawity> quinq: 'ams' and 'mad', as in Amsterdam and Madrid
2023-08-17 16:42:07 <grawity> quinq: fastly-debug-path has the user-facing node at the beginning, x-served-by looks like it has that at the end
2023-08-17 17:03:03 <quinq> Thanks grawity :)
2023-08-17 17:03:25 <quinq> I suppose they just don't maintain their ip location list but still advertise it nonetheless
2023-08-17 19:01:16 <latex_> hey does anyone know how to set ipv6 token?
2023-08-17 19:02:06 <Habbie> what's ipv6 token?
2023-08-17 19:02:57 <ikke> https://www.man7.org/linux/man-pages/man8/ip-token.8.html
2023-08-17 19:03:00 <ikke> I suppose
2023-08-17 19:03:10 <Habbie> ah!
2023-08-17 19:03:19 <Habbie> i was just reading back on #openwrt and started to suspect it was something like that
2023-08-17 19:06:12 <latex_> Or should I now quit using /etc/network/interfaces and write my own service?
2023-08-17 19:06:35 <latex_> I love the /etc/network/interfaces syntax
2023-08-17 19:06:42 <latex_> but I need it to be able to do ip token
2023-08-17 19:07:39 <ikke> you can use pre-up / post-up hooks to run arbitrary commands
2023-08-17 22:26:21 <zcrayfish> hm. running setup-apkrepos -c1 on 3.18.2 results in a mixed 3.18/edge /etc/apk/repositories 
2023-08-17 22:26:24 <zcrayfish> bug?
2023-08-18 00:23:01 <elly> that sure sounds like a bug
2023-08-18 00:44:36 <zcrayfish> I'll file an issue in a bit. First food. 
2023-08-18 01:08:20 <zcrayfish> It's known already
2023-08-18 01:08:23 <zcrayfish> https://gitlab.alpinelinux.org/alpine/alpine-conf/-/issues/10554 
2023-08-18 01:22:24 <zcrayfish> the setup-xorg-base script looks better. It needs work too, but it's not as bad...
2023-08-18 01:22:47 <zcrayfish> I mean in regards to enabling the community repo without mixing stable and edge
2023-08-18 05:43:01 <r3dlust> can someone help me with usages of networkmanager?
2023-08-18 05:48:17 <zcrayfish> what's the issue r3dlust ?
2023-08-18 05:48:30 <r3dlust> ok so basically
2023-08-18 05:48:34 <r3dlust> my devices always show up as unmanaged
2023-08-18 05:48:38 <r3dlust> no matter what i do
2023-08-18 05:48:43 <r3dlust> i've tried literally everything
2023-08-18 05:48:49 <r3dlust> deleting all possible packages that could be interfering
2023-08-18 05:48:53 <r3dlust> messing around for hours in configs
2023-08-18 05:49:03 <r3dlust> nothing seems to get networkmanager to manage my devices
2023-08-18 05:50:54 <zcrayfish> mix of wifi and ethernet?
2023-08-18 05:51:43 <r3dlust> i basically have 3 devices i can use at any given time, a ethernet cable, a wifi dongle (thats the one i want to use) and the integrated wifi adapter of the raspberry pi
2023-08-18 05:51:55 <r3dlust> btw i forgot to mention im running a raspberry pi 3b
2023-08-18 05:52:17 <r3dlust> none of the devices are ever managed by networkmanager
2023-08-18 06:03:00 <zcrayfish> what happens if you try to set them as managed in nmcli? (e.g. nmcli device set DEVICE managed yes)  ???
2023-08-18 06:05:25 <r3dlust> absolutely nothing
2023-08-18 06:05:30 <r3dlust> i've tried that multiple times
2023-08-18 06:05:40 <r3dlust> no error, nothing
2023-08-18 06:05:44 <r3dlust> it just, doesnt do anything
2023-08-18 06:07:47 <zcrayfish> apparently damn near all the output is just thrown into the syslog at /var/log/messages
2023-08-18 06:07:50 <zcrayfish> fun
2023-08-18 06:09:04 <r3dlust> no error messages in those logs either. just standard stuff
2023-08-18 07:11:44 <quinq> Hello
2023-08-18 07:12:04 <quinq> Today apk want's to install a new firefox package, firefox-intl, in addition to firefox
2023-08-18 07:12:10 <quinq> Is there a reason for that?
2023-08-18 07:12:40 <quinq> When I then try to see what pulled in that new package, apk tells me there is no other package requiring it
2023-08-18 07:20:06 <ptrc> quinq: see https://gitlab.alpinelinux.org/alpine/aports/-/issues/15163
2023-08-18 07:21:05 <ptrc> tldr if someone wants to willingly break their firefox install to save a whole 20MB of space, they can
2023-08-18 07:21:43 <zcrayfish> Just a meta package to pull in some runtime deps. Pretty boring
2023-08-18 07:23:03 <quinq> Hummm, still unclear why there's a new package
2023-08-18 07:23:25 <quinq> has firefox transitioned to a meta package, and the default one it points to is now firefox-intl?
2023-08-18 07:24:27 <zcrayfish> firefox-intl is a meta package.
2023-08-18 07:24:38 <zcrayfish> you can tell because: https://pkgs.alpinelinux.org/package/edge/community/x86_64/firefox-intl <-- if you click on contents, there's nothing.
2023-08-18 07:26:07 <quinq> When I “try” to del firefox-intl, it tells me it can't because of… webkit2gtk-4.1-dev
2023-08-18 07:26:12 <quinq> Is there a dependency between both? oO
2023-08-18 07:26:42 <ptrc> no, it's just apk being high on crack
2023-08-18 07:27:01 <ptrc> it's kinda known that it doesn't really show stuff properly with install_if rules
2023-08-18 07:27:17 <quinq> ah ok, is there an open ticket for that?
2023-08-18 07:27:23 <quinq> Well, I can lookup myself I suppose
2023-08-18 07:27:47 <ptrc> i.. don't think so?
2023-08-18 07:28:21 <zcrayfish> lol, I just updated mine just to see what nonsense I would get on a simulated del...
2023-08-18 07:28:27 <zcrayfish> World updated, but the following packages are not removed due to: firefox-intl: firefox gtk+2.0-dev gtk+3.0-dev
2023-08-18 07:28:31 <zcrayfish> good stuff.
2023-08-18 07:28:34 <quinq> meh, job is calling, bbl
2023-08-18 07:28:40 <zcrayfish> later
2023-08-18 10:57:43 <ncopa> the firefox-intl is a workaround so you can use firefox without icu-data-full
2023-08-18 10:59:03 <ncopa> basically, you can do `apk add !icu-data-full` or `apk add !firefox-intl` to prevent those from being installed
2023-08-18 12:13:32 <somercet1> ah, no more icu-data-en?
2023-08-18 16:12:12 <dzilvys> 16:01:24 
2023-08-18 16:12:12 <dzilvys> any clue on what could be causing the following when doing apk upgrade?
2023-08-18 16:12:12 <dzilvys>   ERROR: http://dl-cdn.alpinelinux.org/alpine/v3.17/main: temporary error (try again later)
2023-08-18 16:12:12 <dzilvys> 16:01:19 
2023-08-18 16:12:12 <dzilvys>   / # fetch http://dl-cdn.alpinelinux.org/alpine/v3.17/main/x86_64/APKINDEX.tar.gz
2023-08-18 16:12:14 <dzilvys> 16:01:24 
2023-08-18 16:12:15 <dzilvys>   fetch http://dl-cdn.alpinelinux.org/alpine/v3.17/community/x86_64/APKINDEX.tar.gz
2023-08-18 16:12:17 <dzilvys> 16:01:29 
2023-08-18 16:12:19 <dzilvys>   ERROR: http://dl-cdn.alpinelinux.org/alpine/v3.17/community: temporary error (try again later)
2023-08-18 16:12:22 <dzilvys> 16:01:29 
2023-08-18 16:12:26 <dzilvys>   WARNING: Ignoring http://dl-cdn.alpinelinux.org/alpine/v3.17/community: No such file or directory
2023-08-18 16:12:28 <dzilvys> 16:01:29 
2023-08-18 16:12:30 <dzilvys>   v3.17.5-33-g59034473c39 [http://dl-cdn.alpinelinux.org/alpine/v3.17/main
2023-08-18 16:14:34 <ikke> dzilvys: dns or other network issue
2023-08-18 16:15:47 <dzilvys> ah ok, time to speak to our IT team :D
2023-08-18 16:16:06 <quinq> 3.17 is too old! ;)
2023-08-18 16:18:00 <dzilvys> dont
2023-08-18 16:18:13 <dzilvys> i dont wanna get on to that topic again
2023-08-18 16:19:08 <quinq> Sorry, I didn't know this was traumatic to you
2023-08-18 16:19:42 <ikke> 3.17/main is still supported in any case
2023-08-18 16:20:06 <quinq> The wink was an attempt at hinting this was a joke
2023-08-18 16:20:35 <ikke> yes, but you know Poes law
2023-08-18 16:21:12 <quinq> Not yet, but let-me look that up :)
2023-08-18 16:21:18 <dzilvys> no I get it being a joke and dont feel bad about it, I was somewhat jokingly saying it too
2023-08-18 16:21:20 <dzilvys> <3
2023-08-18 16:22:19 <quinq> Then we reached indeed Poes law! :D
2023-08-18 16:22:25 <quinq> (if I understand it correctly)
2023-08-18 16:59:24 <Ermine> quinq: my server is still 3.17
2023-08-18 17:00:55 <quinq> Ermine, noted
2023-08-18 19:38:36 <oleovit> hi. without being too demanding, just a neutral question: chromium 116 has been out for 2 days, including security fixes. unfortunately, no visible sign of the new maintainer yet. vacation or other reasons?
2023-08-18 19:40:19 <dwfreed> every chrom{e,ium} release has security fixes
2023-08-18 19:41:45 <oleovit> right
2023-08-18 19:42:09 <minimal> Chromium probably takes 2+ days to build ;-)
2023-08-18 19:42:45 <oleovit> firefox takes longer minimal (as far as i know)
2023-08-18 19:43:08 <quinq> Ye who enters web browser realm, abandon any sanity
2023-08-18 19:43:12 <dwfreed> they're both awful
2023-08-18 19:43:23 <oleovit> no alternative
2023-08-18 19:44:01 <zcrayfish> https://flathub.org/apps/org.chromium.Chromium
2023-08-18 19:44:11 <zcrayfish> ;)
2023-08-18 19:44:19 <oleovit> from the tv series "devs": they use you, bit they don't like you anymore"
2023-08-18 19:44:47 <oleovit> the problem is flathub here
2023-08-18 19:45:30 <oleovit> don't want to use it anymore (and i was on silverblue before it was cool)#
2023-08-18 19:47:35 <quinq> indeed (don't know what silverblue is)
2023-08-18 19:48:53 <oleovit> fedora, completely-flatpak-everything-ostree-everything
2023-08-18 19:49:06 <dwfreed> https://fedoraproject.org/silverblue/
2023-08-18 19:51:56 <oleovit> yep. opensuse has something similiar. well, these times are over for me. anyway, i hope the new maintainer can maintain weekely chromium-updates (but as far as i know, he or she is related to google resp. shares a history)
2023-08-18 19:55:39 <zcrayfish> 🤔🤔🤔
2023-08-18 19:59:02 <quinq> Thanks :)
2023-08-18 21:12:57 <lambdanil> I use Flatpak for proprietary stuff
2023-08-18 21:13:00 <lambdanil> I guess it works ok
2023-08-18 21:15:12 <Mari[m]> me too
2023-08-18 21:15:28 <zcrayfish> Same. Steam and Chrome being the big ones.
2023-08-18 21:15:40 <Mari[m]> i use flatpak for gaming stuff (steam, dolphin, lutris, etc)
2023-08-18 21:16:07 <Mari[m]> and nix for non gaming stuff (discord, element, skype)
2023-08-18 21:17:08 <lambdanil> I have those as Flatpaks too, though I do have a few packages installed through Guix
2023-08-18 21:36:05 <m_ben> I wonder if I can get the memory usage of alpine as low as gentoo: https://i.imgur.com/B9MRZbq.jpg
2023-08-18 21:36:50 <m_ben> that's my gentoo machine after booting (disk is fully encrypted) before starting the X server
2023-08-18 21:52:32 <zcrayfish> m_ben: should be doable on the 32-bit build
2023-08-18 21:52:38 <zcrayfish> a little harder on the x86_64
2023-08-18 22:00:41 <zcrayfish> m_ben: 19.5 here loading the 32-bit build on x86_64 qemu. lol
2023-08-18 22:02:41 <m_ben> zcrayfish: interesting. how did you get it that low?
2023-08-18 22:02:51 <zcrayfish> By only using the base system.
2023-08-18 22:03:01 <zcrayfish> Note: 96 MB is required to decompress the initrd/initramfs
2023-08-18 22:05:22 <zcrayfish> https://gopher.zcrayfish.soy/I/irc/alpine-standard-3.18.2.i686.qemu.png
2023-08-18 22:05:36 <zcrayfish> qemu started with: qemu-system-x86_64 -m 96 -nic user -boot c -hda alpine.qcow2 -cdrom /home/zachary/Downloads/alpine-standard-3.18.2-x86.iso -accel kvm
2023-08-18 22:06:06 <zcrayfish> This is an installed system (using pretty much all defaults in setup-alpine)
2023-08-18 22:06:19 <zcrayfish> only thing I added was htop, as it is not in base :p
2023-08-18 22:06:55 <minimal> m_ben: Sys-mode Alpine install you mean? Should be fine. Looking at a VM here htop shows 29.6M/475M (that's with stuff like Virtualbox tools, rsyslog, etc running)
2023-08-18 22:07:07 <zcrayfish> Yes
2023-08-18 22:07:12 <m_ben> minimal: yes, sys mode
2023-08-18 22:08:24 <ayushnix> do the alpine kernels sign the ZFS kernel module if I use ZFS?
2023-08-18 22:08:29 <minimal> m_ben: I've run VMs with only 128M ram in Sys mode, think I got down to about 90M before having kernel issues
2023-08-18 22:08:59 <elly> psykose did some experiments a while ago and the threshold before there was insufficient space to decompress the initramfs was around 95-96MB
2023-08-18 22:09:04 <elly> ah someone said that already
2023-08-18 22:09:14 <minimal> elly: yes I was also involved in that conversation
2023-08-18 22:09:15 <elly> post boot usage can be lower
2023-08-18 22:09:17 <elly> great
2023-08-18 22:09:51 <zcrayfish> 96 for i686 and ~128 for x86_64, yes
2023-08-18 22:09:56 <minimal> elly: that was with me having a slimmed down initramfs (only what was needed)
2023-08-18 22:10:30 <zcrayfish> You need about 320 mb to *install* x86_64 or 128 for i686... at least without swapping.
2023-08-18 22:10:45 <zcrayfish> as-of 3.18.2 
2023-08-18 22:11:11 <minimal> zcrayfish: I'm using pre-built disk images
2023-08-18 22:12:11 <zcrayfish> That helps a lot
2023-08-18 22:13:39 <zcrayfish> ayushnix: Yes
2023-08-18 22:14:07 <zcrayfish> I assume you mean the signature as revealed by `modinfo zfs'?
2023-08-18 22:16:40 <ayushnix> zcrayfish: I mean that if I add the module.sig_enforce=1 kernel parameter, will the ZFS kernel module load?
2023-08-18 22:18:32 <zcrayfish> I'm about 99% sure, but if you want 100%, give me a few minutes to spin up a VM. lol
2023-08-18 22:18:56 <ayushnix> sure
2023-08-18 22:23:07 <zcrayfish> yep, works fine
2023-08-18 22:23:31 <zcrayfish> tested on 3.18.2 x86. 
2023-08-18 22:26:12 <ayushnix> nice
2023-08-18 22:26:16 <ayushnix> thanks zcrayfish
2023-08-18 22:26:31 <zcrayfish> no problem
2023-08-18 22:26:45 <ayushnix> I've setup vanilla Arch on my new desktop but now I'm considering switching to alpine on it
2023-08-18 22:28:10 <minimal> m_ben: just spun up a different VM (with Busybox/dropbear rather than rsyslogd, openssh, dhcpcd, etc) and according to htop it is 27.0M/475M
2023-08-18 22:28:34 <minimal> both those VMs are x86_64, haven't checked to see what difference x86 would make
2023-08-18 22:30:00 <minimal> the biggest thing memory-wise is htop itself lol
2023-08-18 22:32:10 <m_ben> :)
2023-08-18 22:32:30 <minimal> followed by ash and then /sbin/init
2023-08-18 22:33:07 <m_ben> with the default kernel image or a custom one?
2023-08-18 22:33:12 <minimal> linux-virt
2023-08-18 22:35:31 <minimal> the only things running are ash, crond, dropbear, init, login, ntpd, syslogd, udhcpc
2023-08-18 22:38:18 <minimal> m_ben: though I could probably slim it down further by reducing loaded kernel modules further (forgot to set for VGA text mode to avoid need for DRM modules and Virtualbox module is loaded)
2023-08-18 22:53:25 <m_ben> are the kernel images identical between "standard", "extended" and "netboot"?
2023-08-18 22:55:05 <minimal> m_ben: there are no standard, extended, netboot kernel images, only linux-lts, linux-edge, and linux-virt
2023-08-18 22:55:32 <minimal> I assume the Alpine standard/extended/netboot downloads all use linux-lts
2023-08-18 23:09:06 <minimal> m_ben: fixed the kernel modules that shouldn't have been loaded and htop now says 23.6M/475M ;-)
2023-08-18 23:10:45 <m_ben> what is the output of lsmod?
2023-08-18 23:10:52 <m_ben> if you have that installed
2023-08-18 23:11:08 <minimal> 24 entries
2023-08-18 23:12:56 <minimal> basically ext4, virtio_net, virtio_scsi, tiny_power_button, sd_mod, vboxguest (for Virtualbox), and their deps..........I did forget to disable ipv6 so that might save a very small ammount
2023-08-18 23:13:51 <minimal> actually hmm, ipv6 is the 2nd largest module...
2023-08-19 01:20:52 <lambdanil> so, I solved my biggest gripe with Alpine
2023-08-19 01:20:57 <lambdanil> no Doom source port in the repos!
2023-08-19 01:21:18 <lambdanil> once I get aports figured out I could submit a PR
2023-08-19 01:48:00 <Saijin_Naib[m]> Ooh, which one are you submitting?
2023-08-19 01:48:33 <Saijin_Naib[m]> We have chocolate, crispy, free,gz, and lz in edge/testing
2023-08-19 01:49:37 <zcrayfish> too bad they left
2023-08-19 01:49:38 <zcrayfish> :/
2023-08-19 01:50:06 <Saijin_Naib[m]> Balls.
2023-08-19 01:50:17 <Saijin_Naib[m]> That's what I get for hiding join/part/update messages, I guess
2023-08-19 01:50:58 <zcrayfish> I love the way WeeChat's join/part hide can be configured: you can set it to hide them for people who have not talked in X number of hours. pretty cool
2023-08-19 01:51:03 <zcrayfish> I usually have it turned off though
2023-08-19 01:51:13 <elly> Saijin_Naib[m]: I hide joins/parts/quits too, you just gotta reply to people by name in your first line & tab complete them
2023-08-19 01:52:09 <Saijin_Naib[m]> Heh, unfortunately with Element, they show up even if disconnected, apparently
2023-08-19 01:52:16 <Saijin_Naib[m]> It must fetch them from history or something
2023-08-19 01:52:42 <zcrayfish> doh. yes I totally ignored the fact that you are using Matrix. Rendering the WeeChat comment irrelevant 
2023-08-19 01:57:34 <Saijin_Naib[m]> Nah, all good. Everyone has their favorite tools, right? Element isn't mine, but Fractal isn't where I need it to be yet and still might miss some features that Element provides so... Electron it is -_-
2023-08-19 02:08:28 <zcrayfish> I'm using the amazingly convoluted setup of Ferdium -> Glowing Bear -> Weechat, typically. it does work.... but it's a great way to waste several GB of RAM. haha!
2023-08-19 02:11:53 <elly> what is Ferdium?
2023-08-19 02:13:38 <zcrayfish> some lame program that basically hosts all of your web chat/social media stuff in these chromium containers
2023-08-19 02:13:49 <zcrayfish> (n.b.: my terminology is probably an oversimplification, but you get the idea)
2023-08-19 02:14:40 <zcrayfish> Ihttps://gopher.zcrayfish.soy/I/irc/LXQt_Wayland_Screenshot.png <-- I have it on the left-half of my screen in this screenshot I took yesterday
2023-08-19 02:19:58 <minimal> now down to 23M (according to htop) once I disabled ipv6 and vboxguest modules. Think that's as small as I'll be able to get
2023-08-19 02:26:39 <Saijin_Naib[m]> zcrayfish: That is quite the setup... thinking about that makes me shudder, and consider just signing up for Beeper
2023-08-19 02:26:52 <Saijin_Naib[m]> minimal: Living up to the name, I see
2023-08-19 02:27:26 <minimal> on occasions, yes ;-)
2023-08-19 02:28:27 <zcrayfish> it's more efficient than what I had before at least: IRC, Discord, and Whatsapp all in separate windows. Good luck with me ever seeing FB or Matrix stuff happening. haha
2023-08-19 04:48:54 <witcher> Mari[m]: i was under the impression that nix only works with systems running systemd. am I wrong?
2023-08-19 04:49:25 <Mari[m]> What no 
2023-08-19 04:49:34 <Mari[m]> I've used nix on guix and alpine
2023-08-19 04:50:09 <witcher> huh, good to know
2023-08-19 14:59:05 <Case_Of> as I can guess, setup-alpine does not have an option globally to not create a swap partition? it is only available to setup-disk?
2023-08-19 15:08:30 <ikke> SWAP_SIZE=0 setup-alpine
2023-08-19 16:23:59 <Case_Of> ikke: thanks
2023-08-19 16:51:20 <ripspin> I just spent a number of hours trying to install alpine, through the setup-alpine to the point of saving to disk, and it ONLY gave me sr0 which failed, I remounted,rw the partition but no effect, right now I give up
2023-08-19 16:52:05 <ikke> sr0 is the cdrom
2023-08-19 16:52:29 <ripspin> yes i know
2023-08-19 16:52:48 <ripspin> that is where the boot came from
2023-08-19 16:53:02 <ikke> setup-alpine does not offer disks that are mounted
2023-08-19 16:55:56 <ripspin> that is clearly the case, not sure how an insall can happen without a mounted partition
2023-08-19 16:56:15 <ikke> it will create partitions and mount it itself
2023-08-19 16:56:45 <ikke> if you want some custom setup, you can create the partitions and then run setup-disk <partition>
2023-08-19 16:57:13 <ripspin> are you saying that it will take the whole HD?
2023-08-19 16:57:22 <ikke> By default it will
2023-08-19 16:58:04 <ripspin> well Im glad Im not going any further with alpine-linux
2023-08-19 16:59:15 <ripspin> what happened to the small part, you know "small fast and secure"
2023-08-19 17:03:53 <ikke> That has very little to do with how it partitions your disk
2023-08-19 17:04:40 <ikke> The default setup is simple, but gives a way to customize
2023-08-19 17:06:33 <ripspin> it is not going to partition my disk, Im glad you like it the way it is..  I bid you farewell
2023-08-19 17:06:53 <ikke> Have a nice day
2023-08-19 17:07:05 <minimal> ripspin: so then you partition the disk they way you want instead, format them, mount the and THEN run setup-disk and tell it the mountpoint for /
2023-08-19 17:07:47 <ikke> I told ripspin that, but apparently that does not suffice
2023-08-19 17:08:38 <minimal> ikke: of course the other relevant point is that setup-disk assumes that Alpine will be the only OS on the disk/SSD, in terms of bootloader config
2023-08-19 17:09:39 <minimal> that might not be an issue for UEFI but would be for BIOS-based booting where it would modify the bootsector
2023-08-19 17:13:03 <ikke> with BOOTLOADER=none, it will not install a bootloader
2023-08-19 17:18:17 <minimal> ikke: right but then that doesn't give a bootable system. My point was that setup-disk is not necessarily designed with multi-OS boot in mind
2023-08-19 17:23:20 <ripspin> I expanded the download ISO put the dirs/files in a partition changed my grub.conf and booted into alpine linux, I even installed bash and a few others, all gone next boot
2023-08-19 17:24:06 <minimal> ripspin: you installed in run-from-ram or Sys mode?
2023-08-19 17:25:21 <ripspin> it didn't say
2023-08-19 17:26:02 <minimal> ripspin: what did you answer to the prompts of setup-alpine?
2023-08-19 17:26:20 <minimal> is this after you manually partitioned the drive?
2023-08-19 17:28:19 <ripspin> yes, my drive has a number of spare partitions
2023-08-19 17:28:51 <minimal> so what command did you run to install Alpine?
2023-08-19 17:29:23 <ripspin> if your previous question was important then "the alpine linux install system"
2023-08-19 17:29:58 <minimal> ?
2023-08-19 17:30:07 <ripspin> should have made a log and then I could send the log to you 
2023-08-19 17:30:27 <minimal> without knowing what exact commands you ran it is hard to diagnose anything
2023-08-19 17:30:55 <ripspin> we agree on that
2023-08-19 18:01:37 <grming> hi to all
2023-08-19 18:01:45 <grming> anybody uses the libgpiod library?
2023-08-19 19:05:31 <WhyNotHugo> grming: It's a dependency for four packages: https://pkgs.alpinelinux.org/package/edge/community/armhf/libgpiod
2023-08-19 19:06:45 <grming> yes
2023-08-19 19:06:56 <grming> the thing is that im using python with gpiod
2023-08-19 19:07:08 <grming> and using gpioset works great
2023-08-19 19:07:35 <grming> but when I wanna read the output pin using the library the in change state to 0
2023-08-19 19:07:47 <grming> and I want to keep the pin to the state
2023-08-19 19:36:49 <M-u-m-p-i-t-z> Hello! I have a question about openfire package. If I start the server "htop" shows me 20 different PIDs all the same Java Server start of openfire. Is It correct? Because it is a extrem loadout and I get timeouts at admin console. If any could help me, please write I will stay same time. thx
2023-08-19 19:39:16 <ikke> M-u-m-p-i-t-z: all in white, or also green processes (threads)?
2023-08-19 19:40:07 <M-u-m-p-i-t-z> all green
2023-08-19 19:40:17 <ikke> those are threads, they belong to one process
2023-08-19 19:40:51 <ikke> press shift+h to hide them
2023-08-19 19:41:16 <M-u-m-p-i-t-z> ok, but this is the reason of laggy admin console?
2023-08-19 19:41:54 <ikke> not on its own, no
2023-08-19 19:42:26 <M-u-m-p-i-t-z> hmm because CPU and RAM is free
2023-08-19 19:43:33 <M-u-m-p-i-t-z> but it is nice to know that the green PID can be hide
2023-08-19 19:46:46 <M-u-m-p-i-t-z> Then only the internal database remains which could cause these lags
2023-08-20 00:14:51 <elagost> would there be a negative consequence to regenerating /etc/machine-id at every boot?
2023-08-20 00:36:14 <uuidNuniq> elagost: what do you mean,
2023-08-20 00:37:07 <uuidNuniq> machine-id is useless on Alpine.
2023-08-20 00:37:17 <uuidNuniq> it is basically a systemd thing. 
2023-08-20 00:38:13 <grawity> not exactly, on non-systemd distros it still serves the same purpose that it did before systemd was invented
2023-08-20 00:38:46 <grawity> though most of its (few) users were systemd-ish back then, i.e. pulseaudio and such
2023-08-20 00:39:49 <grawity> (it was located at /var/lib/dbus in the past)
2023-08-20 00:39:51 <grawity> I don't *think* anything significant would break, if you re-generated it every boot?
2023-08-20 00:40:24 <uuidNuniq> why would one want to regen it on every boot?
2023-08-20 00:41:46 <grawity> ¯\_(ツ)_/¯
2023-08-20 00:42:37 <grawity> somehow it makes people worry more than /etc/ssh/ssh_host_rsa_key does
2023-08-20 00:43:24 <uuidNuniq> are these 2 related?
2023-08-20 00:43:40 <grawity> only in that they're both unique machine identifiers
2023-08-20 00:44:13 <uuidNuniq> host keys are ususally generated the first time openssh is installed
2023-08-20 00:55:16 <zcrayfish> grawity: You should just ditch the RSA keys and use one of the EC keys instead.
2023-08-20 00:56:55 <elagost> ssh isn't enabled on my laptop system. Just concerned about the potential implications of a unique ID.
2023-08-20 00:57:21 <zcrayfish> ah, I got you
2023-08-20 00:57:38 <grawity> zcrayfish: I...don't think that was the point at all
2023-08-20 00:57:51 <zcrayfish> I'm trying to find an article I saw the other day... Basically someone was advocating for all machines to simply have the same ID.
2023-08-20 00:58:06 <zcrayfish> grawity: I know, but screw RSA keys, they're obsolete.
2023-08-20 00:58:25 <zcrayfish> I'm surprised it took over a minute for anyone to eyeroll at me on that one though \o/
2023-08-20 01:00:33 <zcrayfish> Well he left so disregard everything.
2023-08-20 01:00:37 <zcrayfish> :)
2023-08-20 01:03:10 <elagost> zcrayfish: the devuan folks get into it, like they get into everything: https://lists.dyne.org/lurker/message/20190308.124740.2b7329de.en.html
2023-08-20 01:05:12 <zcrayfish> thanks for the URL. 
2023-08-20 01:05:45 <zcrayfish> it was either whonix or tails who had posted something about it
2023-08-20 01:05:47 <zcrayfish> anyway
2023-08-20 01:06:17 <zcrayfish> https://github.com/Kicksecure/dist-base-files/blob/master/etc/machine-id 
2023-08-20 01:06:19 <zcrayfish> Yep
2023-08-20 01:06:24 <zcrayfish> Just steal theirs. Problem solved.
2023-08-20 01:06:27 <zcrayfish> ;)
2023-08-20 01:14:55 <Ermine> It's a dbus thing
2023-08-20 01:47:33 <dalias> host key is only readable by root. i guess you could attempt ssh to localhost to get the pubkey..?
2023-08-20 09:09:21 <abbablack> I'm installing Alpine on a Dell XPS 15.  I've reinstalled it twice using two different install media.  Both times, I run into an issue where it can't properly boot.  It gets to "Loading hardware modules" and then begins complaining about I/O errors for everything (/sbin/getty, etc.).  The drive itself seems fine, as everything is accessible/works when doing chroot/live disk.  I also have nvme 
2023-08-20 09:09:27 <abbablack> configured (it seems that's done automatically via the installer now).  Any pointers?
2023-08-20 09:17:55 <somercet1> hmm, which firmware apks are installed?
2023-08-20 09:23:05 <abbablack> Like the whole list? Is there something specific I should look for? It's a couple pages long
2023-08-20 09:25:38 <somercet1> apk -I list | grep firmware
2023-08-20 09:27:38 <somercet1> with --root mountpoint, pointing to your install partition
2023-08-20 09:28:23 <abbablack> The first error it prints out after Loading modules is related to openrc "veinfo: I/O error" and then it never reaches the prompt (just endlessly prints /sbin/getty IO error). When I chroot from livedisk and run that command I get: ath10k, go7007, i915, intel, mediatek, nvida, other, qcom, rtl_bt
2023-08-20 09:31:52 <somercet1> hmm, dunno, I'm new to openrc, sorry
2023-08-20 09:36:29 <abbablack> no worries, thanks for trying to help anyway
2023-08-21 00:18:45 <fluix> how does one use just plain LUKS (without LVM, encrypted boot, any other stuff) with grub? I modify /etc/default/grub with cryptoroot=UUID=... root=/dev/mapper/root rootfstype=ext4 and I get "error: no such edvice: <annother UUID>"
2023-08-21 00:19:09 <fluix> this error happens before any grub menu or anything and I'm dropped into grub rescue
2023-08-21 00:22:14 <fluix> the UUID I set is the one blkid shows for /dev/nvme0n1p2 and another UUID is that of /dev/mapper/root when I `cryptsetup open /dev/nvme0n1p2 root` on the installation medium
2023-08-21 00:25:29 <minimal> fluix: you said *without* encrypted root, so why would you set cryptoroot=UUID=... ?
2023-08-21 00:26:14 <minimal> ah, sorry, you said not encrypted boot, rather than encrypted root
2023-08-21 00:26:49 <fluix> yeah, just encrypted Root
2023-08-21 00:27:10 <fluix> I'm not sure where grub is even getting the UUID it's failing on, it's not in any config
2023-08-21 00:28:03 <fluix> but it appears multiple times in the generated grub.cfg
2023-08-21 00:28:19 <minimal> did you set "cryptoroot=UUID=..." in /etc/default/grub?
2023-08-21 00:28:38 <fluix> yes and it gets added to the command line in grub.cfg
2023-08-21 00:28:39 <minimal> and also set "cryptdm=<whatever>" there?
2023-08-21 00:29:07 <fluix> what does cryptdm get set to?
2023-08-21 00:29:15 <minimal> the device mapper name
2023-08-21 00:29:32 <minimal> i.e. /dev/mapper/<whatever>
2023-08-21 00:29:59 <fluix> is cryptdevice shorthand for cryptroot and cryptdm?
2023-08-21 00:30:21 <minimal> ?
2023-08-21 00:31:26 <fluix> I've set cryptdm, will try that in a sec
2023-08-21 00:31:55 <fluix> asking because Arch pages say set `cryptdevice=UUID=...:root root=/dev/mapper/root`
2023-08-21 00:32:11 <minimal> also I don't set "root="m just "cryptroot=UUID=... cryptdm=<whatever> rootfstype=ext4"
2023-08-21 00:32:39 <fluix> I believe I tried that, but one sec
2023-08-21 00:32:47 <minimal> Arch pages will be written from the perspective of the initramfs they use (dracut?) so should not be blindly followed
2023-08-21 00:33:28 <fluix> yep, so I'm reading the Alpine ones which is why I used cryptroot but I figured cryptdm was just for lvm
2023-08-21 00:33:46 <fluix> anyways, cryptdm=root and no root= still fails with the same error
2023-08-21 00:34:11 <minimal> what else is in your /etc/default/grub?
2023-08-21 00:35:00 <fluix> will pull it up in a sec, but it's just the defaults plut cryptroot and cryptdm in the command line
2023-08-21 00:35:10 <minimal> did you also enable cryptsetup in features in mkinitfs.conf and rebuild initramfs ?
2023-08-21 00:35:45 <fluix> I believe so; checking
2023-08-21 00:37:31 <minimal> and the UUID valueused for "cryptroot=UUID=..." matches that of /dev/nvme0n1p2 ?
2023-08-21 00:37:50 <fluix> yes that UUID matches
2023-08-21 00:37:59 <fluix> but grub is erroring on the id of /dev/mapper/root
2023-08-21 00:39:28 <fluix> /etc/default/grub contains the defaults with GRUB_CMDLINE_LINUX_DEFAULT="modules=sd-mod,usb-storage,ext4,nvme quiet cryptroot=UUID=... cryptdm=root rootfstype=ext4"
2023-08-21 00:39:57 <minimal> and nvme is present in features list in mkinitfs.conf ?
2023-08-21 00:40:05 <fluix> and /etc/mkinitfs/mkinitfs.conf has features="ata base ide scsi usb virtio ext4 nvme cryptsetup resume"
2023-08-21 00:41:50 <minimal> bios or uefi boot?
2023-08-21 00:42:07 <fluix> I have definitely regenerated both (modification time of /mnt/initramfs-lts is after /etc/mkinitfs/mkinitfs.conf and grub.cfg contains the command line)
2023-08-21 00:42:07 <fluix> uefi
2023-08-21 00:42:31 <minimal> so /dev/nvme0n1p2 is the VFAT ESP partition?
2023-08-21 00:42:41 <minimal> oops p1
2023-08-21 00:42:46 <fluix> p1 is the EFI partition, p2 is ext4
2023-08-21 00:43:00 <fluix> which I've been mounting into /mnt/boot/efi/
2023-08-21 00:43:22 <minimal> LUKSv1 or v2?
2023-08-21 00:43:33 <fluix> whatever the default luksFormat is, which I think is 2 now?
2023-08-21 00:43:48 <minimal> so the generated grub.cfg contains stuff like "insmod luks2" ?
2023-08-21 00:44:11 <minimal> doh, forget that, still thinking of encrypted boot lol
2023-08-21 00:44:16 <fluix> it doesn't mention luks at all, yeah
2023-08-21 00:44:51 <fluix> (not sure if this is relevant, but where does grub get its search UUIDs from? i.e. the UUID it's failing on?)
2023-08-21 00:45:00 <fluix> does /etc/fstab need to have /dev/mapper/root instead of the UUID?
2023-08-21 00:45:36 <minimal> is nvme drive 512/4K or 4k/4k sector sizes?
2023-08-21 00:45:56 <minimal> typically I have an entry for /dev/mapper/<whatever> in /etc/fstab
2023-08-21 00:46:21 <minimal> then again I also have GRUB_DISABLE_LINUX_UUID=true in /etc/default/grub
2023-08-21 00:49:23 <fluix> how do I found that out and how would that be relevant?
2023-08-21 00:50:08 <minimal> the sector sizes? It may be relevant for partitioning layout if drive is 4k/4k
2023-08-21 00:50:40 <fluix> yes
2023-08-21 00:51:40 <minimal> cat /sys/block/nvme0n1/queue/logical_block_size and physical_block_size
2023-08-21 00:53:18 <fluix> 512/512
2023-08-21 00:54:01 <minimal> ok, that's fine
2023-08-21 00:55:11 <minimal> grub should read the UUID you specified in GRUB_CMDLINE_LINUX_DEFAULT's cryptroot=UUID=... to figure out things for the grub.cfg it generates
2023-08-21 00:56:12 <fluix> the only place that UUID exists is in the linux kernel options for the normal and advanced boot menu options
2023-08-21 00:56:48 <fluix> everywhere else as part of `search` commands is the /dev/mapper/root UUID that it fails on
2023-08-21 00:57:40 <minimal> what kernel options are you talking about?
2023-08-21 00:58:07 <fluix> s/kernel options/linux command line/
2023-08-21 00:58:30 <minimal> which linux command line options are you talking about?
2023-08-21 00:58:43 <fluix> i.e. linux /boot/bmlinuz-lts root=UUID=<mapper UUID> ... cryptroot=UUID=<the one in default/grub>
2023-08-21 00:58:53 <fluix> s/bmlinuz/vmlinuz/
2023-08-21 00:59:19 <minimal> this is from "cat /proc/cmdline"? or from some file?
2023-08-21 00:59:23 <fluix> grub.cfg
2023-08-21 01:00:19 <fluix> but grub seems to be failing well before that, because I don't get a grub menu
2023-08-21 01:10:17 <minimal> just building a similar VM locally to see its config
2023-08-21 01:11:43 <fluix> thanks
2023-08-21 01:13:30 <fluix> looks like it's using the UUID from $(grub_get_device_id "${GRUB_DEVICE}")
2023-08-21 01:23:30 <minimal> ok the generated grub.cfg looks like I expected
2023-08-21 01:24:38 <minimal> its contains "root=/dev/mapper/lukspart rootfstype=ext4 cryptroot=UUID=<of the partition> cryptdm=lukspart
2023-08-21 01:25:23 <fluix> so why is mine generating it with root=UUID=<garbage>
2023-08-21 01:26:54 <minimal> this is with /etc/default/grub containing: GRUB_CMDLINE_LINUX_DEFAULT="rootfstype=ext4 cryptroot=UUID=<of the partition> cryptdm=lukspart"
2023-08-21 01:27:42 <minimal> and GRUB_ENABLE_LINUX_LABEL=false and GRUB_DISABLE_LINUX_UUID=true and GRUB_ENABLE_CRYPTODISK=n
2023-08-21 01:29:00 <fluix> still errors the exact same way with the exact same UUID
2023-08-21 01:29:16 <minimal> what options did you use for grub-install?
2023-08-21 01:29:27 <fluix> however setup-disk does it
2023-08-21 01:30:44 <fluix> err that's not true, I ran grub-install separately after the first was failing this way I believe
2023-08-21 01:31:07 <fluix> how should I run it?
2023-08-21 01:31:24 <minimal> well first tell me how you ran it
2023-08-21 01:33:23 <fluix> grub-install --target=x86_64-efi --efi-directory=/boot/efi/ (not exactly sure of that last path)
2023-08-21 01:34:03 <fluix> but yeah, I ran it an extra time because /mnt/boot/efi/EFI/ contains both alpine/grubx64.efi and grub/grubx64.efi
2023-08-21 01:40:33 <minimal> so as you didn't specify the "-no-nvram" option then it (likely) wrote a boot entry to the EFI variables
2023-08-21 01:40:50 <fluix> yes, although I have not been booting it
2023-08-21 01:41:15 <fluix> (but do want to remove it)
2023-08-21 01:42:16 <minimal> you said earlier you'd mounted the ESP partition as /mnt/boot/efi
2023-08-21 01:42:34 <fluix> yes
2023-08-21 01:43:01 <minimal> but you just said when you ran grub-install you told it "--efi-directory=/biit/efi/" not /mnt/boot/efi
2023-08-21 01:43:09 <minimal> s/biit/boot/
2023-08-21 01:43:32 <minimal> why?
2023-08-21 01:43:33 <fluix> that's because I had run it inside of the chroot
2023-08-21 01:43:37 <fluix> I think*
2023-08-21 01:44:07 <minimal> you think? or you did run it inside a chroot?
2023-08-21 01:44:33 <fluix> I don't remember
2023-08-21 01:44:41 <minimal> its important
2023-08-21 01:44:56 <fluix> the path itself is correct whether I ran it inside or outside though (as in, if I ran it outside, I specified the path with /mnt, if I ran it inside I specified it without /mnt)
2023-08-21 01:45:08 <minimal> so the rootfs is actually /mnt ?
2023-08-21 01:45:35 <fluix> if I ran it outside the chroot, yes
2023-08-21 01:45:58 <minimal> and you bind-mounted /sys, /proc/, /dev into the chroot?
2023-08-21 01:46:07 <fluix> every time I've chrooted ,yes
2023-08-21 01:46:25 <minimal> so everything sounds fine
2023-08-21 01:46:44 <minimal> and it works fine for me
2023-08-21 01:46:47 <fluix> yes, there is just an extra /boot/efi/EFI/grub path and the nvram entry
2023-08-21 01:46:56 <minimal> so i'm out of ideas
2023-08-21 01:47:05 <fluix> do you happen to know how to get rid of the nvram entry?
2023-08-21 01:47:19 <minimal> to stop it being created or to remove it?
2023-08-21 01:47:24 <fluix> remove the extra one
2023-08-21 01:47:29 <minimal> use efibootmgr
2023-08-21 01:50:15 <fluix> thanks. and just to check, I don't need to run grub-install at all, right?
2023-08-21 01:50:28 <fluix> (outside of the time setup-alpine does it, that is)
2023-08-21 01:51:02 <minimal> I guess so
2023-08-21 01:51:20 <fluix> okay, thanks for trying. will keep trying things
2023-08-21 01:55:26 <fluix> if you still have it open and want to take a look: what's UUID on the `search` line 2 lines above the `linux ...` command line in grub.cfg
2023-08-21 01:55:43 <fluix> s/what's/which/
2023-08-21 01:57:13 <minimal> it's the uuid of the ext4 boot partition
2023-08-21 01:57:38 <minimal> ah........you don't have a boot partition/fs, just a ESP and rootfs
2023-08-21 01:57:57 <minimal> hang, on, you said you weren't doing encrypted boot, so where's your /boot?
2023-08-21 01:58:39 <fluix> 
ACTION facepalms

2023-08-21 01:58:53 <minimal> that explains things
2023-08-21 01:58:53 <fluix> it's on the luks rootfs
2023-08-21 01:59:08 <minimal> right, without cryptomount enabled lol
2023-08-21 01:59:47 <fluix> I ought to be mounting to /mnt/boot aye
2023-08-21 01:59:52 <minimal> you obviously can't do unencrypted /boot when /boot is part of a encrypted / ;-)
2023-08-21 02:00:12 <minimal> ought to be mounting *what* to /mnt/boot ?
2023-08-21 02:00:18 <minimal> you said you only had 2 partitions
2023-08-21 02:00:29 <fluix> can /boot not live on the efi partition?
2023-08-21 02:00:50 <minimal> you you mount /dev/nvme0p1 as /boot?
2023-08-21 02:01:03 <fluix> not right now, but that would work, right?
2023-08-21 02:01:20 <minimal> but then what about /mnt/boot/efi ?
2023-08-21 02:01:25 <fluix> instead of mounting /dev/nvme0n1p1 to /mnt/boot/efi, I mount it to /mnt/boot and then the boot partition is there
2023-08-21 02:01:32 <fluix> well that's inside of boot
2023-08-21 02:01:43 <fluix> oh I get what you're saying
2023-08-21 02:01:52 <fluix> but it looks like that's fine according to Arch's pages
2023-08-21 02:01:53 <minimal> for a ESP partition the EFI directory has to be at the top-level of the filesystem
2023-08-21 02:02:31 <minimal> so then you'd need to run grub-install specifying /mnt/boot as the efi directory
2023-08-21 02:02:36 <fluix> yes
2023-08-21 02:02:54 <minimal> and you'll end up with EFI in the /boot directory
2023-08-21 02:02:57 <fluix> yes
2023-08-21 02:03:10 <minimal> personally I don't like that and so have a separate /boot partition
2023-08-21 02:03:24 <fluix> fair enough, but that should work, right?
2023-08-21 02:03:51 <minimal> I guess, don't think I've ever tried it
2023-08-21 02:03:58 <fluix> (the way I got here is reading the Arch page but trying to stick closer to the Alpine page and didn't see the difference in mount paths)
2023-08-21 02:04:04 <minimal> as I don't want /boot on a VFAT fs
2023-08-21 02:04:19 <fluix> what's the issue with that?
2023-08-21 02:05:20 <minimal> because I don't want /boot on a VFAT fs
2023-08-21 02:05:37 <minimal> I didn't say there was any issue with it, I said I didn't want that
2023-08-21 02:06:02 <fluix> ah ok
2023-08-21 02:06:26 <fluix> alright, well thanks a lot for the help :)
2023-08-21 10:17:50 <deutschem> upgraded my alpinelinux box (apk update && apk upgrade), after reboot i get the message that I do not have a kernel that supports pae. I have a alix1 board.
2023-08-21 10:19:21 <ikke> What version of Alpine Linux?
2023-08-21 10:25:01 <deutschem> I am not sure, i think 3.17
2023-08-21 10:31:34 <ikke> as far as I know, pae has never been enabled in 3.17. It got briefly enabled in 3.18, but then disabled again, as it breaks support of existing users
2023-08-21 10:37:05 <deutschem> reinstall all with 3.16 ?
2023-08-21 10:42:20 <ikke> What gives that message? (when does it happen)?
2023-08-21 10:45:23 <deutschem> after choosing entry in boot-menu "Linux-lts" shortly come the message "this kernel requires the following feature not present on the CPU:pae" Unable to boot
2023-08-21 10:45:59 <ikke> Ok, that's the opposite of what you mentioned initialy 
2023-08-21 10:46:29 <ikke> deutschem: If you install the kernel from edge, it should boot again
2023-08-21 10:47:06 <ikke> deutschem: https://lists.alpinelinux.org/~alpine/devel/%3C84d45683-3727-8e6e-2f65-04c5f724f0da%40dkcomm.net%3E#%3C20230816181753.4e2a4148@ncopa-desktop.lan%3E
2023-08-21 11:05:55 <deutschem> I do this first time, what are the global steps for installing kernel from edge ?
2023-08-21 11:10:30 <zcrayfish> I didn't think those boards had SSE either. 
2023-08-21 12:16:31 <ascensionwires> how use testing branch?
2023-08-21 12:44:45 <ascensionwires> how use testing branch?
2023-08-21 12:50:20 <ikke> 1. Switch to edge. 2. Add the testing repo to /etc/apk/repositories 
2023-08-21 12:59:38 <ascensionwires> for what is @testing?
2023-08-21 13:54:14 <ikke> ascensionwires: you can use it to make sure you only get packages from that repo that you explicitly specify 
2023-08-21 13:54:50 <ikke> Notably it's not a good solution to be able to install packages from the testing repository when you are on a stable release 
2023-08-21 13:55:41 <minimal> s/not a good solution/not supported/
2023-08-21 14:38:57 <ascensionwires> if i on edge?
2023-08-21 14:56:47 <ikke> On edge there is little value to use it
2023-08-21 17:56:30 <SilentTeaCup> Does anyone know how alpine ensures that the apk-cache is mounted before installing packages during boot in diskless mode? According to my /var/log/dmesg package intall is executed before the device with my apk-cache is mounted ("[    5.393118] Installing packages to root filesystem...", "[   21.384187] EXT4-fs (sda1): mounted filesystem with ordered data mode. Quota mode: none.").
2023-08-21 18:49:25 <ascensionwires> who knows how install ungoogled chromium on alpine?
2023-08-21 18:51:14 <elly> compile it?
2023-08-21 18:51:50 <elly> someone requested it be packaged many years ago: https://gitlab.alpinelinux.org/alpine/aports/-/issues/10801 and nobody did so
2023-08-21 18:51:57 <elly> the reasons given there still seem valid to me
2023-08-21 18:52:01 <boomanaiden154> Chromium requires a pretty extensive patch set to compile/run correctly on Alpine.
2023-08-21 18:52:13 <elly> it sure does
2023-08-21 18:52:50 <ascensionwires> cuz it i asking
2023-08-21 18:53:00 <boomanaiden154> Even the current patch set has some issues. I think there's still an outstanding bug where WebRTC or something will segfault after x minutes.
2023-08-21 18:53:21 <elly> I don't know of such a problem at the moment
2023-08-21 18:53:35 <elly> in the alpine chromium package, I mean
2023-08-21 18:53:55 <ascensionwires> firefox? o
2023-08-21 18:55:02 <boomanaiden154> https://gitlab.alpinelinux.org/alpine/aports/-/issues/13970
2023-08-21 18:55:14 <boomanaiden154> Looks like something might've fixed this at some point.
2023-08-21 18:55:43 <ascensionwires> mb browser as chromium or firefox not good idea on alpine
2023-08-21 18:55:51 <elly> boomanaiden154: looks like it got closed 6 months ago?
2023-08-21 18:55:55 <elly> yeah
2023-08-21 18:56:07 <elly> ascensionwires: uh, that's up to your use case - chromium works basically fine on alpine
2023-08-21 18:56:11 <elly> firefox does too, I bet
2023-08-21 18:57:48 <ascensionwires> have other question, how get clangd on alpine? 
2023-08-21 19:00:19 <SilentTeaCup> It seems fstab entries are loaded after package install during boot in diskless mode. It seems only partitions with a device name (e.g. sda1) directly under /media are mounted correctly. Is this expected behaviour?
2023-08-21 19:00:37 <elly> SilentTeaCup: sounds like it?
2023-08-21 19:02:30 <SilentTeaCup> Hmm that's unfortunate so we can't use UUIDs for apk-cache and lbu mounts.
2023-08-21 19:04:10 <elly> hm
2023-08-21 19:05:48 <SilentTeaCup> Maybe I have to dig into the init scripts but I guess there is a reason these are limited to /media in the setup scripts.
2023-08-21 19:07:36 <elly> it is probably just the simplest implementation someone could think of
2023-08-21 19:07:57 <ayushnix> SilentTeaCup: I'm not sure how it works under the hood but the packages in my apk cache are installed after a reboot without issues.
2023-08-21 19:08:08 <ayushnix> Although I had trouble figuring out how.
2023-08-21 19:08:21 <ayushnix> because there's either no documentation or poor documentation for it
2023-08-21 19:08:28 <ayushnix> what's your issue?
2023-08-21 19:10:35 <XFaCiEer> I'm also lazy.
2023-08-21 19:12:04 <SilentTeaCup> ayushnix I wanted to use apk-cache and lbu on a disk I mounted with fstab. But according to my dmesg log the fstab entries are mounted after the installing packages and loading settings set. Therefore pks are not loaded from the cache correctly during a reboot.
2023-08-21 19:16:44 <ayushnix> SilentTeaCup: what's the value of LBU_MEDIA in /etc/lbu/lbu.conf?
2023-08-21 19:18:13 <SilentTeaCup> ayushnix It was data and in fstab I mounted the parition under /media/data.
2023-08-21 19:19:09 <ayushnix> SilentTeaCup: does /etc/apk/cache point to /media/data/cache?
2023-08-21 19:19:50 <SilentTeaCup> ayushnix> It points to /media/data/alpine/cache.
2023-08-21 19:20:53 <ayushnix> SilentTeaCup: What's the output of /etc/apk/repositories?
2023-08-21 19:21:15 <ayushnix> s/\/etc/cat \/etc/
2023-08-21 19:28:14 <SilentTeaCup> <ayushnix> /media/mmcblk0p1/apks
2023-08-21 19:28:15 <SilentTeaCup> http://alpine.mirror.wearetriple.com/v3.18/community
2023-08-21 19:28:15 <SilentTeaCup> http://alpine.mirror.wearetriple.com/v3.18/main
2023-08-21 19:28:40 <ayushnix> SilentTeaCup: remove the first line
2023-08-21 19:29:13 <ayushnix> and maybe ensure that /etc/apk/cache points to /media/data/cache
2023-08-21 19:29:30 <ayushnix> not /media/data/alpine/cache
2023-08-21 19:29:56 <ayushnix> then, run this as root or with doas 'apk cache -v sync'
2023-08-21 19:29:59 <ayushnix> and reboot
2023-08-21 19:32:07 <SilentTeaCup> ayushnix Why remove the line? If I reboot and the cache isn't found correctly again then the system would be without any pkgs and I think in won't boot.
2023-08-21 19:33:41 <ayushnix> SilentTeaCup: that line points to the apk cache of the initial rootfs image, I think
2023-08-21 19:34:02 <ayushnix> I also had the same issue where my apk cache in /media/sda2/cache wasn't being picked up
2023-08-21 19:34:18 <ayushnix> after I removed that line, apk cache worked as expected
2023-08-21 19:34:32 <ayushnix> but remember to do 'apk cache -v sync' before a reboot
2023-08-21 19:42:01 <SilentTeaCup> ayushnix> Hmm didn't help. data is still not mounted before loading settings nor installing packages. Not even LBU_MEDIA is honored it just loads the file from a different parition (mmcblk0p2).
2023-08-21 20:07:22 <ayushnix> Hmm, no idea about to proceed then
2023-08-22 07:41:13 <ascensionwires> is localization possible on alpine linux?
2023-08-22 08:01:32 <ascensionwires> i mean language support
2023-08-22 08:06:36 <qaqland[m]> it follows linux language support🤔
2023-08-22 08:07:56 <qaqland[m]> it's hard to show different language fonts in tty screen
2023-08-22 08:13:41 <ascensionwires> i generate language locale via changing /etc/locale.gen then run locale-gen
2023-08-22 08:13:49 <ascensionwires> how i can choose locale
2023-08-22 08:14:02 <ascensionwires> ?
2023-08-22 08:25:02 <somercet1> either define LANG in ~/.profile, or in /etc/profile.d/20locale.sh, and/or install musl-locales-lang
2023-08-22 08:27:39 <ascensionwires> there some steps here but i dont understand it https://wiki.alpinelinux.org/wiki/Post_installation#Language_support
2023-08-22 08:28:27 <ascensionwires> what is actually this unicode? need i "yes" ?
2023-08-22 08:30:56 <ascensionwires> there nothing about /etc/locale.gen and locale-gen command on thus wiki page
2023-08-22 08:33:06 <ikke> It does not exist for musl 
2023-08-22 08:37:31 <ascensionwires> yes and how i can do steps to generate locales and choose it?
2023-08-22 08:43:36 <humm> ascensionwires, you don’t generate locales
2023-08-22 08:44:40 <humm> you choose by setting LANG
2023-08-22 08:47:38 <ascensionwires> all locales generated?
2023-08-22 08:47:53 <ascensionwires> *already?
2023-08-22 08:51:26 <ascensionwires> https://wiki.gentoo.org/wiki/Musl_usage_guide#Locales there even LANG env variable not available on gentoo wiki writed
2023-08-22 09:24:21 <humm> that’s wrong
2023-08-22 09:24:55 <humm> ascensionwires, just try
2023-08-22 09:25:21 <humm> run a program that does localization with your favorite LANG
2023-08-22 09:36:44 <ascensionwires> where set LANG system wide?
2023-08-22 09:38:14 <WhyNotHugo> Musl doesn't support locales (e.g.: LC_TIME or LC_NUMERIC).
2023-08-22 09:38:17 <WhyNotHugo> So there's no locale-gen.
2023-08-22 09:44:09 <ascensionwires> mb i need install musl-locales musl-locales?
2023-08-22 09:47:38 <ascensionwires> then apk add lang
2023-08-22 09:47:47 <ascensionwires> and then export LANG
2023-08-22 09:50:00 <ascensionwires> ^*musl-locales musl-locales-lang
2023-08-22 09:51:34 <ascensionwires> only thing where define LANG system wide? /etc/profile.d/20locale.sh.sh?
2023-08-22 09:52:24 <ascensionwires> or just /etc/profile.d/locale.sh
2023-08-22 10:36:57 <ikke> ascensionwires: /etc/profile.d/locale.sh would work, but note that it only affects login shells
2023-08-22 11:03:33 <divVerent_> I have a problem on a laptop... it kinda powered down during apk upgrade. Now the dbus binary is all nul bytes and thus iwd can no longer connect to wifi. I have no ethernet.
2023-08-22 11:03:46 <divVerent_> I'm aware of "apk audit --packages -q | xargs apk fix", but can't do that without internet.
2023-08-22 11:04:14 <divVerent_> Any good idea how I can locate the package files I need and put them on a flash drive so apk fix can actually find them?
2023-08-22 11:05:27 <divVerent_> (argh, wpa_supplicant needs dbus too)
2023-08-22 11:06:54 <ikke> apk cache download <packagelist> will download the packages including dependencies to the cache directory
2023-08-22 11:07:31 <ikke> or apk fetch -R <packages>
2023-08-22 11:07:39 <ikke> which will download it 
2023-08-22 11:07:47 <divVerent_> yeah, but my only Alpine machine has no internet
2023-08-22 11:07:49 <ikke> in the current folder or the directory specified with -o
2023-08-22 11:08:10 <ikke> You can download a static copy of apk-tools and run that anywhere
2023-08-22 11:08:11 <divVerent_> guess I can set up a VM, or maybe use podman, have one machine that has it
2023-08-22 11:08:22 <divVerent_> ah, nice
2023-08-22 11:08:41 <divVerent_> I do have a package list, but first attempt didn't work as versions don't match what's installed so deps problems... with apk fetch -R it should be fine, let's try
2023-08-22 11:09:18 <ovf> you also probably want apk audit --full (otherwise it won't catch things like the dbus binary)
2023-08-22 11:13:51 <divVerent_> yeah, did that
2023-08-22 11:14:15 <divVerent_> now am inside a podman with alpine... I guess that's easiest
2023-08-22 11:21:26 <divVerent_> also, lesson learned: should have done setup-apkcache in advance
2023-08-22 11:27:48 <divVerent_> hm... thanks, but that's not all yet
2023-08-22 11:28:18 <divVerent_> even after doing apk cache download, and then rsyncing the cache dir to the laptop's /var/cache/apk, I still can't run apk fix $(cat packagelist)
2023-08-22 11:28:27 <divVerent_> (with --no-network)
2023-08-22 11:28:40 <divVerent_> still all packages are "APK unavailable, skipped"
2023-08-22 11:36:58 <divVerent_> hm... only now I see dbus's config files are also zero length, and apk fix leaves those alone... I guess I may as well reinstall then, as it was hundreds of packages being upgraded when I ran out of power
2023-08-22 11:37:16 <ikke> if you remove them, apk should restore them
2023-08-22 11:37:22 <divVerent_> but I can't remove dbus
2023-08-22 11:37:26 <divVerent_> the entire world depends on dbus
2023-08-22 11:37:31 <divVerent_> oh, you mean the files only
2023-08-22 11:38:03 <divVerent_> the other issue is, I didn't get apk fix to work yet
2023-08-22 11:38:09 <divVerent_> no idea how to make it actually find the files
2023-08-22 11:38:16 <divVerent_> so I backed up my world file and apk add'd the dbus apk
2023-08-22 11:38:43 <divVerent_> but of course, apk add won't add it again as it's already there
2023-08-22 11:39:24 <ikke> try `apk add path/to/dbus*.apk`
2023-08-22 11:39:30 <divVerent_> doing just that
2023-08-22 11:39:38 <divVerent_> it won't do it anymore as now the version matches
2023-08-22 11:39:48 <divVerent_> after my first attempt restored the binaries (good) but not the config files
2023-08-22 11:39:58 <ikke> and if you provide --available?
2023-08-22 11:40:12 <divVerent_> that flag doesn't exist in apk add
2023-08-22 11:40:44 <ikke> right, only for upgrade
2023-08-22 11:41:39 <divVerent_> I'd think the proper way would be getting apk fix to work without network, just how to make it recognize the cached apkl files
2023-08-22 11:42:19 <ikke> divVerent_: I assume you enabled package caching (setup-apkcache)?
2023-08-22 11:43:36 <divVerent_> yes
2023-08-22 11:43:44 <divVerent_> currently reading strace output
2023-08-22 11:44:47 <divVerent_> it did list all the apk files
2023-08-22 11:45:02 <divVerent_> but says apk unavailable anyway
2023-08-22 11:46:17 <divVerent_> argh, I found one issue - different repository versions
2023-08-22 11:49:40 <divVerent_> also, now deleted all 0 byte files from /etc
2023-08-22 11:49:53 <divVerent_> is there anything else "protected" from upgrading?
2023-08-22 11:51:45 <ikke> no, only /etc/* by defaut
2023-08-22 11:51:49 <ikke> no, only /etc/* by defau;t
2023-08-22 11:53:55 <divVerent_> (also, I've never seen so many mmap calls as in that strace of apk)
2023-08-22 11:54:16 <divVerent_> (looks like it's mmapping every block of the apk index separately)
2023-08-22 11:56:34 <divVerent_> apk fix still won't do it as I have different package versions currently installed than what would be correct
2023-08-22 11:57:10 <ikke> you can specify --upgrade to apk fix, but if these packages are from a different release, then you should not use that
2023-08-22 11:58:19 <divVerent_> that has millions of dep conflicts
2023-08-22 11:58:45 <divVerent_> they are same release (edge), but the laptop had not been powered on for about 6 months, so... kinda expectable
2023-08-22 11:59:03 <divVerent_> well, trying apk upgrade now that I finally have dbus and iwd running again, thanks!
2023-08-22 11:59:26 <divVerent_> all happened because stupid me had the power supply plug in the headphone jack
2023-08-22 14:39:02 <ascensionwires> can i install alpine without wired network?
2023-08-22 14:40:05 <eddsalkield> yes, i'd recommend downloading the extended image
2023-08-22 14:40:12 <eddsalkield> it contains packages needed to set up wifi
2023-08-22 14:40:38 <eddsalkield> if you haven't seen it already, check out https://wiki.alpinelinux.org/wiki/Wi-Fi
2023-08-22 14:43:44 <ascensionwires> i usually do: setup-alpine and answer "none" when asked for the disk to use, for where to store configs, and for the location of the package cache, then i partitioning and formatting, then mount and do setup-disk -k edge /mnt
2023-08-22 14:44:25 <ascensionwires> what i need if no wired connection?
2023-08-22 14:48:19 <ascensionwires> alpine extended is only solution?
2023-08-22 14:49:51 <eddsalkield> i don't think you need any kind of internet connection for the operations you mentioned
2023-08-22 14:50:04 <eddsalkield> you can certainly install alpine in sys mode without any kind of internet connection
2023-08-22 14:50:22 <eddsalkield> because the image contains a package cache
2023-08-22 14:50:29 <ascensionwires> i need connect to wifi during installation?
2023-08-22 14:50:53 <eddsalkield> yeah, sometimes the installer requires downloading a package outside the cache
2023-08-22 14:51:16 <eddsalkield> in that case using the extended image works better, because it'll contain the package you need
2023-08-22 14:51:17 <ascensionwires> can i connect to wifi from standard iso?
2023-08-22 14:51:37 <eddsalkield> i don't think the standard iso contains the wifi packages, no
2023-08-22 14:51:39 <eddsalkield> it's very minimal
2023-08-22 14:51:58 <eddsalkield> you won't get packages you don't need on your installed system if you use extended
2023-08-22 14:52:48 <ascensionwires> on my  wired setup i can install any packages during installation
2023-08-22 14:53:08 <eddsalkield> yes, because you can reach the internet to get those additional packages
2023-08-22 14:57:07 <ascensionwires> ok i cant connect wireless to wifi in standard iso during installation
2023-08-22 14:59:57 <ascensionwires> ?
2023-08-22 14:59:59 <qaqland[m]> apk add says "To upgrade individual packages without modifying world use apk-fix(8)."
2023-08-22 15:01:04 <eddsalkield> ascensionwires: yes, the packages needed to connect to wifi aren't in the standard image
2023-08-22 15:01:14 <eddsalkield> so you need to either:
2023-08-22 15:01:26 <eddsalkield> using the standard image, connect to the internet to download the wifi packages
2023-08-22 15:01:42 <eddsalkield> or use an image where those packages are already present (i.e. extended)
2023-08-22 15:02:21 <ascensionwires> also have a question im now on winfows 10 how i can load in alpine standard liveiso?
2023-08-22 15:02:30 <qaqland[m]> using apk add to upgrade package will modify /etc/apk/world ?
2023-08-22 15:04:07 <eddsalkield> qaqland: apk add does modify /etc/apk/world - it puts an additional package in there
2023-08-22 15:04:40 <eddsalkield> ascensionwires: are you looking to load it onto some removable media for installation?
2023-08-22 15:04:46 <eddsalkield> e.g. a USB stick?
2023-08-22 15:04:55 <eddsalkield> or do you want to run alpine as a VM or within WSL?
2023-08-22 15:07:49 <ascensionwires> usb
2023-08-22 15:08:35 <qaqland[m]> ascensionwires: I use this https://rufus.ie/
2023-08-22 15:09:14 <ascensionwires> i just usually installing alpine to my pc withot uefi, now i want install on laptop with installed windows 10
2023-08-22 15:10:26 <eddsalkield> yeah any kind of iso writer should do, I've also heard of https://github.com/balena-io/etcher
2023-08-22 15:10:31 <eddsalkield> on linux I just use dd
2023-08-22 15:11:41 <ascensionwires> i now have alpine liveiso
2023-08-22 15:11:51 <eddsalkield> nice one
2023-08-22 15:12:02 <ascensionwires> i have standard
2023-08-22 15:12:17 <ascensionwires> this how i install alpine on my pc https://pastebin.com/raw/7b2cgsKW
2023-08-22 15:12:26 <ascensionwires> now i want install on laptop
2023-08-22 15:13:11 <ascensionwires> and i dont know how do this
2023-08-22 15:13:23 <eddsalkield> why didn't you get the extended image?
2023-08-22 15:13:37 <eddsalkield> there's a guide for dualbooting on the wiki: https://wiki.alpinelinux.org/wiki/Dualbooting
2023-08-22 15:14:18 <ascensionwires> just never used uefi on my pc cuz no uefi support
2023-08-22 15:14:35 <ascensionwires> as i understand on laptop uefi support 
2023-08-22 15:15:07 <ascensionwires> but i cant even boot in it
2023-08-22 15:44:52 <elly> you have a laptop without UEFI support?
2023-08-22 16:17:37 <ascensionwires> now i have laptop with uefi support
2023-08-22 16:17:53 <ascensionwires> and windows 10
2023-08-22 16:18:28 <ascensionwires> i even dont know how load in bios in it to run alpine liveiso in uefi
2023-08-22 16:18:46 <ascensionwires> also dont know how install alpine with uefi
2023-08-22 16:19:12 <ascensionwires> i just installed alpibe on my old pc like i sended before
2023-08-22 16:19:30 <ascensionwires> https://pastebin.com/raw/7b2cgsKW
2023-08-22 16:26:43 <ascensionwires> how boot on alpine iso from my laptop?
2023-08-22 16:30:52 <elly> in general all you need to do is dd it onto a usb stick, then select the usb stick while the machine is booting
2023-08-22 16:34:10 <ascensionwires> how select usb disk
2023-08-22 16:34:36 <ascensionwires> i even cant boot in bios
2023-08-22 16:35:58 <zcrayfish> Which brand/marque is your computer?
2023-08-22 17:14:19 <ascensionwires> lenovo
2023-08-22 17:14:47 <ascensionwires> lol i now change boot mode to uefi and loaded from alpibe standard usb
2023-08-22 17:15:02 <ascensionwires> setup-alpibe have wifi support
2023-08-22 17:15:12 <ascensionwires> i connected to wifi
2023-08-22 17:15:28 <ascensionwires> i skipped disk with answering none
2023-08-22 17:15:35 <ascensionwires> what i need now?
2023-08-22 17:15:56 <ikke> create / mount partitions
2023-08-22 17:16:03 <ikke> run setup-disk /path/to/root
2023-08-22 17:16:30 <ikke> But since you are dual booting, you need to be carefull with that, it would try to install a bootloader
2023-08-22 17:16:53 <ascensionwires> i shrink space from windows 10 but cant see it
2023-08-22 17:17:31 <ascensionwires> like 60 gb
2023-08-22 17:18:11 <ascensionwires> cant see it from fdisk
2023-08-22 17:18:19 <ikke> how did you shrink it?
2023-08-22 17:18:25 <ascensionwires> with menu
2023-08-22 17:18:30 <ascensionwires> diskmgmt
2023-08-22 17:22:20 <elly> that might have shrunk the filesystem but not the containing partition
2023-08-22 17:51:50 <ascensionwires> what i need mount to /mnt if i want uefi
2023-08-22 17:52:08 <ascensionwires> windows uefi partition to /mnt/boot ?
2023-08-22 17:52:30 <elly> you do not need to mount anything to use uefi
2023-08-22 17:53:10 <ascensionwires> i mean if i want double boot with windows
2023-08-22 17:53:22 <ascensionwires> or no way?
2023-08-22 17:53:42 <ascensionwires> i even dont know is windows installed with uefi
2023-08-22 17:54:13 <elly> I think one of us is very confused here
2023-08-22 17:54:49 <elly> UEFI is something that happens mostly before boot, your system can either boot with UEFI or legacy bios; you can't have some OSes that boot with legacy BIOS and some that boot with UEFI installed simultaneously
2023-08-22 17:55:01 <elly> Windows 10 almost certainly boots with UEFI on your system
2023-08-22 17:55:15 <zcrayfish> Yes, it is a minimum system requirement of recent windows versions.
2023-08-22 17:55:28 <elly> if you want to, you can boot Alpine from Windows' bootloader (WinBoot), or you can have GRUB (Alpine's default bootloader) load WinBoot, which is called "chainloading"
2023-08-22 17:55:35 <elly> or you can have your laptop's UEFI firmware present you both options
2023-08-22 17:55:54 <elly> in any case, there's no actual need to mount anything after boot - the UEFI part already happened if your system is running
2023-08-22 17:59:49 <ascensionwires> lol i checked no and no uefi partition on windows 10 now
2023-08-22 18:00:09 <ascensionwires> but t480 support uefi
2023-08-22 18:00:53 <ascensionwires> i feel i cant install alpine with uefi and dual booting with windows
2023-08-22 18:01:00 <zcrayfish> ESP is just a FAT16 or FAT32 partition, it's not really special. 
2023-08-22 18:01:40 <elly> my desktop is in a very similar config (UEFI-only, dualbooting Win11 and Alpine) and it worked just fine without any messing around except for convincing grub to chainload winboot
2023-08-22 18:02:09 <zcrayfish> I used to use EasyBCD to convince Windows bootloader to chainload GRUB
2023-08-22 18:02:16 <zcrayfish> these days I just use efibootmgr exclusively.
2023-08-22 18:02:21 <zcrayfish> so there's more than one way to do it.
2023-08-22 18:02:29 <elly> there's like 50 ways unfortunately
2023-08-22 18:02:33 <zcrayfish> yeah
2023-08-22 18:07:00 <elly> I cannot say whether I like x86 or arm booting less :P
2023-08-22 18:08:11 <zcrayfish> I like GRUB the least out of all options, so. =)
2023-08-22 18:37:57 <ascensionwires> elly, what is chainload win boot?
2023-08-22 18:51:34 <elly> I think you should go look that one up on your own
2023-08-22 19:01:35 <XFaCiEer> <zcrayfish> ESP is just a FAT16 or FAT32 partition, it's not really special. 
2023-08-22 19:01:43 <XFaCiEer> But does it support FAT12? That's what we need to know.
2023-08-22 19:02:15 <elly> I hope not
2023-08-22 19:02:56 <XFaCiEer> What are you, anti-DOS?
2023-08-22 19:03:02 <minimal> yes FAT12 is supported by *some* (not sure if all) UEFI implementations, I use a 1M FAT12 ESP for my VMs
2023-08-22 19:04:52 <minimal> as all I have in there is the fallback BOOT/BOOTX64.EFI file
2023-08-22 19:05:36 <XFaCiEer> What's the benefit of fat12? Less file system overhead?
2023-08-22 19:05:42 <minimal> smaller filesystem
2023-08-22 19:05:51 <elly> XFaCiEer: yes I am :P
2023-08-22 19:06:14 <XFaCiEer> @ECHO OFF
2023-08-22 19:06:37 <XFaCiEer> minimal: but what percentage savings are we talking over FAT16?
2023-08-22 19:07:01 <minimal> you cannot have a FAT16 < 8Mb
2023-08-22 19:07:14 <minimal> to be smaller it has to be FAT12
2023-08-22 19:07:31 <XFaCiEer> Understood
2023-08-22 19:07:37 <minimal> that's based on a typical 512 byte sector
2023-08-22 19:07:56 <XFaCiEer> Useful for very tight situations, like some embed use cases
2023-08-22 19:08:11 <minimal> if you're using a device with 4K sectors (like some NVMEs) then a FAT16 cannot be smaller than 256MB
2023-08-22 19:09:25 <minimal> well as I said typically for VMs the only thing I have in the ESP partition is the BOOTX64.EFI file
2023-08-22 19:09:52 <minimal> I could create smaller than 1M ESP partition in that case, I just never got around to it
2023-08-22 19:11:23 <zcrayfish> elly + XFaCiEer: FAT12 required to be supported as part of the UEFI specification. It's a terrible idea. Don't do that. :)
2023-08-22 19:11:45 <elly> surprising
2023-08-22 19:17:27 <minimal> "EFI encompasses the use of FAT32 for a system partition, and FAT12 or FAT16 for removable media."
2023-08-22 19:18:09 <minimal> that implies that FAT16 and FAT32 does not have to be supported for non-removable media (i.e. SATA disks and SSDs, NVME SSDs)
2023-08-22 19:18:26 <minimal> that's from UEFI 2.5 spec
2023-08-22 19:18:50 <zcrayfish> ah, didn't realize it was nuanced like that... but it makes sense!
2023-08-22 19:20:15 <minimal> whereas I was referring to using FAT12 for a disk/ssd with a VM as EDKII's UEFI implementation does support FAT12 in that scenario
2023-08-22 19:27:41 <XFaCiEer> zcrayfish: I will touch the hot stove.
2023-08-22 19:28:29 <zcrayfish> In that case... UDF?
2023-08-22 19:29:39 <dalias> everything about uefi is so despicable
2023-08-22 19:31:50 <XFaCiEer> dalias: EFI shell being DOS doesn't strike your fancy?
2023-08-22 19:33:07 <zcrayfish> 
ACTION thinks UEFI class 0, 1, 2, and 3+ are despicable.

2023-08-22 19:33:49 <elly> bring back Open Firmware
2023-08-22 19:34:50 <zcrayfish> The stuff that late-model MacPPC machines ran?
2023-08-22 19:35:01 <zcrayfish> (e.g. the iMac G3 and the eMac)
2023-08-22 19:35:52 <elly> it had a long history before that
2023-08-22 19:38:10 <XFaCiEer> 
ACTION ponders buying the nearby used eMac

2023-08-22 19:38:51 <zcrayfish> Bad idea if you have good hearing: The flyback transformers in eMacs are the most godawful garbage Apple has ever used.
2023-08-22 19:39:00 <zcrayfish> They're reliably loud.
2023-08-22 19:39:24 <minimal> do they go all the way to 11? ;-)
2023-08-22 19:39:29 <zcrayfish> Indeed!
2023-08-22 19:39:34 <zcrayfish> Nice little antique machines otherwise :)
2023-08-22 19:44:17 <dalias> the whole problem domain of uefi is something that shouldn't exist
2023-08-22 19:45:16 <dalias> the legitimate role of the bios is a bootloaded to pass of execution and a hardware description to your choice of absolute-highest-privilege software as quickly and simply as possible
2023-08-22 19:45:24 <dalias> bootloader*
2023-08-22 19:45:59 <dalias> and this never required making something more elaborate than legacy bios
2023-08-22 19:47:53 <zcrayfish> UEFI level 2 does that job perfectly for me. 
2023-08-22 19:48:36 <zcrayfish> s/2/3 
2023-08-22 19:48:37 <zcrayfish> but yeah
2023-08-22 19:55:54 <minimal> zcrayfish: what are these UEFI levels that you refer to?
2023-08-22 19:57:16 <zcrayfish> the UEFI classes
2023-08-22 19:57:36 <zcrayfish> https://wiki.osdev.org/UEFI#UEFI_class_0-3_and_CSM 
2023-08-22 19:58:58 <zcrayfish> class 3... no legacy cruft? Pretty nice. Downsides: potentially can be turned into level 3+ (i.e secureboot), and the UEFI networking stack I wouldn't miss.
2023-08-22 19:59:57 <minimal> there's no mention of these classes as far as I can see in the UEFI 2.5 spec
2023-08-22 20:01:17 <zcrayfish> Might be vendor specific terminology (intel) https://uefi.org/sites/default/files/resources/Brian_Richardson_Intel_Final.pdf
2023-08-22 21:10:35 <noisp> im building a linux kernel for alpine on macOS, and i would like to use EFI. however, efibootmgr is incompatible
2023-08-22 21:11:35 <noisp> are there any tools that will generate a EFI boot-entry on macOS? I do not mind compiling software
2023-08-22 21:14:18 <minimal> noisp: why do you need to use efibootmgr when compiling a kernel?
2023-08-22 21:16:56 <noisp> minimal: to clarify, i am making a bootable-drive. i included the part about the kernel to show im not using an iso
2023-08-22 21:34:53 <ascensionwires> i got really strange issue
2023-08-22 21:35:07 <ascensionwires> i installed alpine but bo internet
2023-08-22 21:35:19 <ascensionwires> i installed from standard iso
2023-08-22 21:35:40 <ascensionwires> wlan0 was fduring installation
2023-08-22 21:36:01 <ascensionwires> but when i setup-disk, sone errors
2023-08-22 21:36:24 <ascensionwires> I can share installation way throught pastbin
2023-08-22 21:46:22 <ascensionwires> https://pastebin.com/raw/CwyRqcBd
2023-08-22 21:53:52 <minimal> noisp: is this bootable drive a removable (i.e. USB drive)? are you going to be booting Alpine on your Mac?
2023-08-22 22:18:15 <ascensionwires> help
2023-08-22 22:28:10 <ascensionwires> why i have this issue?
2023-08-22 22:35:43 <noisp> The efibootmgr equivalent for macOS is bless, this is included in the OS
2023-08-22 22:36:23 <noisp> Answering a question I asked earlier, and sharing what I found
2023-08-22 22:36:50 <ascensionwires> hwlp please
2023-08-22 22:53:06 <ascensionwires> i have no idea why this wrong - what about i wrote about earlier
2023-08-22 22:56:54 <zcrayfish> ascensionwires: I wasn't paying a lot of attention, but did you use one of the setup scripts to get your internet going to begin with?
2023-08-22 23:02:39 <ptrc> ascensionwires: "some other error" is not very helpful when debugging :/
2023-08-22 23:06:17 <minimal> noisp: I'd asked if you were you going to be booting Alpine on your Mac or not
2023-08-22 23:08:36 <zcrayfish> minimal: he left again
2023-08-22 23:17:06 <minimal> if he's not booting on the Mac then he doesn't need to deal with EFI variables
2023-08-22 23:37:23 <ascensionwires> zcrayfish i sended pastbin link
2023-08-22 23:39:18 <zcrayfish> ascensionwires: you may be able to simply run setup-interfaces again; it seems like networking started running after you had run setup-alpine earlier (setup-interfaces is a part of it)
2023-08-22 23:40:39 <ascensionwires> strange it didnt see wlan0
2023-08-22 23:40:53 <ascensionwires> mb i installed alpine wrong
2023-08-22 23:41:25 <ascensionwires> its very strange
2023-08-22 23:41:58 <ascensionwires> why it work during setup-alpine but not working if setup-disk
2023-08-22 23:42:58 <zcrayfish> I have no idea. your pastebin only shows what you did, but not any of the results (e.g. dmesg output) so I can only guess.
2023-08-22 23:43:12 <ascensionwires> i va
2023-08-22 23:43:30 <ascensionwires> i can do samr steps and say whqt wrong
2023-08-22 23:43:48 <ascensionwires> connection work
2023-08-22 23:44:03 <ascensionwires> it connected from setup-alpine
2023-08-22 23:44:18 <ascensionwires> but when i runned setup-disk all ruined
2023-08-22 23:52:56 <ascensionwires> have no idea what it was
2023-08-22 23:53:36 <ascensionwires> i now added USE_EFI=1 and hope all will work, now not this error
2023-08-22 23:54:56 <zcrayfish> uh....
2023-08-22 23:55:24 <zcrayfish> how does that factor into the networking issue?
2023-08-22 23:55:59 <ascensionwires> now all working
2023-08-22 23:56:06 <zcrayfish> cool
2023-08-22 23:56:06 <ascensionwires> have no idea what it was
2023-08-22 23:56:10 <ascensionwires> mb
2023-08-22 23:56:44 <ascensionwires> some router issue when setupping
2023-08-22 23:56:49 <ascensionwires> have no idea
2023-08-23 00:05:26 <ascensionwires> thanks
2023-08-23 00:05:33 <ascensionwires> gn
2023-08-23 00:18:50 <M-u-m-p-i-t-z> Hello, I will use a rasby with alpine linux and a  XMPP server (openfire) with intern (on the sdcard) database. How can I  save only the database without doing lbu ci command?
2023-08-23 00:53:00 <eddsalkield> ascensionwires: sorry for the outdated info earlier on wifi not being in the standard image. this was fixed since I last heard: https://gitlab.alpinelinux.org/alpine/aports/-/issues/13258
2023-08-23 02:00:09 <linux819> hi, i deleted /var/log/messages and apparently syslogd is not recreating on reboot. killing/starting syslogd does not recreate it either. any ideas. thanks
2023-08-23 02:04:49 <minimal> busybox syslogd? there's nothing in its conf.d/init.d to create any such files
2023-08-23 02:07:44 <minimal> linux819: with BB syslogd I just tested stopping it, renaming messages file, starting it, and message file WAS created
2023-08-23 02:17:41 <linux819> nminimal: thank you for trying this! it seems when run with the debug level argument the file is not created. `/sbin/syslogd -l 1` 
2023-08-23 02:22:14 <minimal> linux819: hmm, you're right, seems like a BB bug
2023-08-23 02:34:22 <divVerent_> ikke: thanks for the help from before - have that laptop recovered now
2023-08-23 02:34:51 <linux819> minimal: that's frustrating. but thanks for your effort in testing! i'll have to find a different way to do things
2023-08-23 02:35:29 <minimal> linux819: why not raise a bug for it?
2023-08-23 02:38:23 <linux819> minimal: indeed, this is the correct way
2023-08-23 05:35:06 <ikke> divVerent_: glad to hear
2023-08-23 09:50:12 <quinq> Opened a new firefox tab, the whole firefox process just quit, without crashing
2023-08-23 09:59:30 <ptrc> quinq: oom maybe?
2023-08-23 09:59:30 <ikke> Did you pay your monthly tab subscription?
2023-08-23 11:28:34 <ascensionwires> yesterday i installer alpine with only wifi connection, and all work
2023-08-23 11:29:05 <ascensionwires> *indtalled
2023-08-23 11:29:10 <ascensionwires> *installed
2023-08-23 11:31:10 <ascensionwires> how do I connect to other wi-fi now?
2023-08-23 11:41:28 <lopid> how did you connect?
2023-08-23 11:43:01 <ascensionwires> its was during installation (in setup-alpine)
2023-08-23 11:43:26 <ascensionwires> dont know how connect to other wifi now?
2023-08-23 12:02:58 <quinq> ptrc, I would have seen it on the dmesg, but no
2023-08-23 12:03:06 <quinq> ikke, sounds like that yeah :D
2023-08-23 12:09:00 <ascensionwires> what i need to connect wifi?
2023-08-23 12:10:15 <ikke> You need to change the wpa_supplicant.conf file in /etc to specify the SSID and psk 
2023-08-23 12:11:53 <ascensionwires> now no actual passphrase in this
2023-08-23 12:12:12 <ascensionwires> i actually dont know how it all setupped wgen setup appibr
2023-08-23 12:12:20 <ascensionwires> *setup-appine
2023-08-23 12:13:43 <ikke> You can try to run setup-interfaces 
2023-08-23 12:15:00 <ascensionwires> and it add wifi, not overwrite?
2023-08-23 12:15:53 <ikke> Yes, it appends to that file 
2023-08-23 12:19:07 <ascensionwires> ok i think setup-interfaces good after it
2023-08-23 13:08:08 <M-u-m-p-i-t-z> Hello, I will use a rasby with alpine linux and a  XMPP server (openfire) with intern (on the sdcard) database. How can I  save only the database without doing lbu ci command?
2023-08-23 13:08:08 <M-u-m-p-i-t-z> Can anyone give me some support?
2023-08-23 15:41:46 <ascensionwires> how connect to portable hotspot (phone)? from alpine
2023-08-23 15:42:46 <ikke> The same as you connect to any wireless network
2023-08-23 15:43:09 <harleen> from the wifi networks
2023-08-23 15:44:33 <harleen> anyone know about the ish on iphone?
2023-08-23 15:44:48 <ikke> I know what it is, but nothing more
2023-08-23 15:45:21 <ascensionwires> ikke for some reason not work
2023-08-23 15:45:39 <ascensionwires> i do setup-interfaces
2023-08-23 15:45:45 <ascensionwires> choose hotsoot
2023-08-23 15:45:50 <harleen> ISH is alpine linux on iphone..id like to know whether it is close to an alipne linux with gui
2023-08-23 15:46:09 <ascensionwires> then dhcp and ping dont work
2023-08-23 15:47:31 <ascensionwires> it connect to wifi
2023-08-23 15:48:44 <ascensionwires> but if i off router and hotdpotting from my phone and setup-interfaces and choosingy phone it not connect to internet for some reason
2023-08-23 15:55:41 <invoked> iwd has worked better than wpa_supplicant for me in some corner cases
2023-08-23 15:55:43 <ascensionwires> have no idea how do it
2023-08-23 15:56:04 <ascensionwires> why wpa_supplicant doesnt work
2023-08-23 15:56:11 <clius[m]> Hi I copied my whole file system as root and so accidentally set the owner of everything to Rolf. Is there any list of folders and their default owners?
2023-08-23 15:56:27 <invoked> ascensionwires: i didn't say it won't work in your case.  it might, i don't know.
2023-08-23 15:56:35 <invoked> captive portals are always an issue though
2023-08-23 15:57:51 <clius[m]> I ment directorys damn deepl translator
2023-08-23 15:58:30 <invoked> ascensionwires: also a problem, potentially... i've seen people complain about personal hotspots when they're running a vpn on the phone that has a kill switch.
2023-08-23 15:58:34 <lopid> copied? so you can rsync over them?
2023-08-23 15:58:44 <ikke> clius[m]: we don't maintain a list, but `apk fix --directory-permissions` might fix it
2023-08-23 15:58:57 <invoked> ascensionwires: if you have a vpn on the hotspot device, try disabling the kill switch.
2023-08-23 15:59:08 <ascensionwires> no
2023-08-23 15:59:11 <ascensionwires> strange
2023-08-23 15:59:43 <ascensionwires> now i tryed load from alpine liveiso 
2023-08-23 15:59:52 <ascensionwires> and running setup-alpine
2023-08-23 16:00:02 <ascensionwires> it connect from setup-alpine
2023-08-23 16:00:21 <ascensionwires> mb i do something wrong
2023-08-23 16:00:36 <ascensionwires> why it not work not from liveiso?
2023-08-23 16:01:41 <ascensionwires> mb i need not setup-intetfaces command
2023-08-23 16:04:07 <elly> you will probably get better help if you do some research and ask more specific questions, rather than "why does x not work?"
2023-08-23 16:04:15 <elly> none of us have the same system setup as you so we'd just be guessing
2023-08-23 16:06:35 <clius[m]> <ikke> "clius: we don't maintain a list,..." <- Thanks for the answer i now only get 8 errors but dont know why http://0x0.st/H9Z5.txt
2023-08-23 16:07:26 <elly> huh, it'd be nice if it said what the errors were :)
2023-08-23 16:08:08 <ascensionwires> https://pasteboard.co/mau4WYu5YY1W.jpg
2023-08-23 16:08:28 <ikke> clius[m]: can you run `apk fix` one more time, see if it mentions the issues?
2023-08-23 16:09:21 <lopid> why *doesn't* it say?
2023-08-23 16:09:36 <ikke> It does say the first time
2023-08-23 16:09:58 <lopid> oh, there was missing info
2023-08-23 16:10:34 <lopid> probably what lead to the -vvv
2023-08-23 16:11:19 <ascensionwires> https://pasteboard.co/wajbxooCK1MX.jpg
2023-08-23 16:11:45 <ascensionwires> first link - is from installation, second link from liveiso
2023-08-23 16:11:51 <ascensionwires> same hotspot
2023-08-23 16:12:19 <ascensionwires> very strange
2023-08-23 16:12:28 <ascensionwires> mb i need run anothet command
2023-08-23 16:12:35 <ascensionwires> not setup-interfaces
2023-08-23 16:12:49 <ascensionwires> or i need do something soecial
2023-08-23 16:13:30 <zcrayfish> after the DHCP lease is obtained, can you do an `nslookup google.com'
2023-08-23 16:14:02 <ascensionwires> ping google not work
2023-08-23 16:14:10 <ascensionwires> strange
2023-08-23 16:14:12 <ikke> do you have an address on the interface?
2023-08-23 16:14:12 <zcrayfish> n.b.: despite it being an absolutely asinine thing to do, some networks filter ICMP pings, hence nslookup
2023-08-23 16:14:25 <ascensionwires> but it work from liveiso
2023-08-23 16:14:29 <ascensionwires> on same laptop
2023-08-23 16:14:34 <ascensionwires> with same hotspot
2023-08-23 16:14:43 <ascensionwires> just from setup-alpine
2023-08-23 16:15:16 <ascensionwires> but from current installation from same laptop it not work
2023-08-23 16:15:26 <ascensionwires> have no idea
2023-08-23 16:15:58 <elly> well, it's either: 1) the interface itself is not up, 2) you aren't associated with the network, 3) you are associated but have no address, 4) you are associated and have an address but no routes, or 5) you have routes but your packets don't actually go anywhere
2023-08-23 16:16:30 <zcrayfish> did he try to determine if the connection was working using any method /other/ than ICMP?
2023-08-23 16:17:41 <ascensionwires> its really strange, mb better way just install alpine with this hotspot, because i cant connect to it not from liveiso
2023-08-23 16:18:21 <elly> you really should stop just reinstalling constantly
2023-08-23 16:18:31 <ikke> ascensionwires: To find out what the issue is, you need to be methodical
2023-08-23 16:18:41 <elly> your actual problem is that you don't understand the system well enough to debug your problems, and instead of learning about it you are just retrying randomly over and over
2023-08-23 16:19:15 <ascensionwires> but setup-alpine do setup-interfaces
2023-08-23 16:19:55 <ascensionwires> why it work only from liveusb setup-alpine?
2023-08-23 16:20:41 <ikke> ascensionwires: to find out, you need to actually investigate
2023-08-23 16:20:44 <elly> the answer is one of the five things I suggested above, and you should figure out which one of them it is by inspecting your system and reasoning about it
2023-08-23 16:20:45 <ikke> we cannot guess why
2023-08-23 16:20:53 <elly> (or it's some sixth thing I didn't think of)
2023-08-23 16:21:26 <ikke> 1. Do you receive an ip. 2. Do you receive a default route. 3. Can you ping the default route
2023-08-23 16:21:34 <clius[m]> <ikke> "clius: can you run `apk fix` one..." <- Dont help http://0x0.st/H9NP.txt
2023-08-23 16:22:21 <ikke> ah ok, it's the --directory-permissions that's returning the errors
2023-08-23 16:22:58 <ascensionwires> ikke you told me ways, but still it very strange why it work from liveiso on same laptop with same hotspot
2023-08-23 16:23:39 <elly> yes, it is strange
2023-08-23 16:23:42 <zcrayfish> I have to assume that the network is working, but ICMPv4 is being filtered.
2023-08-23 16:23:43 <elly> time to figure out why
2023-08-23 16:23:57 <zcrayfish> Since AFAIK no effort at a TCP or UDP connection has been made.
2023-08-23 16:24:03 <elly> zcrayfish: it's the same network regardless though, right? so changing the alpine install's behavior should not affect that
2023-08-23 16:24:13 <zcrayfish> right.
2023-08-23 16:24:15 <ikke> ascensionwires: because you have a working setup and a non-working setup, one way to debug is checking for differences
2023-08-23 16:24:36 <ascensionwires> its same laptop
2023-08-23 16:24:43 <ascensionwires> and same hotspot
2023-08-23 16:25:06 <ascensionwires> i just can connect from setup-alpine in liveiso
2023-08-23 16:25:15 <dwfreed> there's a lot of differences besides those two things
2023-08-23 16:25:29 <ikke> ascensionwires: we cannot work with 'it is working here and not there'
2023-08-23 16:25:35 <ikke> You have to run commands to investigate
2023-08-23 16:25:54 <ikke> ascensionwires: just repeating the same thing does not give us more information
2023-08-23 16:26:10 <ascensionwires> i inswered no now in liveiso and pinged google and it work
2023-08-23 16:26:12 <ikke> ascensionwires: run `ip address show`
2023-08-23 16:26:44 <ikke> We believe you when you say it's working from the live iso
2023-08-23 16:27:00 <ikke> but we are trying to help you to find out why it's not working after installation
2023-08-23 16:27:11 <ikke> But you need to give the information we ask you
2023-08-23 16:27:17 <ikke> otherwise we cannot help you
2023-08-23 16:27:59 <ascensionwires> ok now i poweroff liveiso and back boot in installation
2023-08-23 16:28:38 <ascensionwires> whwt i need to do?
2023-08-23 16:29:38 <ikke> `ip addr show`
2023-08-23 16:29:44 <ascensionwires> ok
2023-08-23 16:29:48 <ikke> is your wireless interface up and does it have an ip address?
2023-08-23 16:30:32 <ascensionwires> i installed this insyltallation with other wifi
2023-08-23 16:30:58 <clius[m]> <ikke> "ah ok, it's the --directory-..." <- Is there any way to see which errors it throws?
2023-08-23 16:31:12 <ikke> clius[m]: I know _a_ way
2023-08-23 16:31:29 <ikke> running it with strace
2023-08-23 16:31:35 <zcrayfish> oh that's gonna be fun
2023-08-23 16:31:39 <dwfreed> I was just going to say 'strace is always an option'
2023-08-23 16:32:40 <ascensionwires> ikke https://pasteboard.co/hlaulyxeg6Ww.jpg
2023-08-23 16:33:04 <ikke> so yes, it's up and has an ip address
2023-08-23 16:33:15 <ascensionwires> but it doesnt hotspot
2023-08-23 16:33:28 <ikke> just be patient
2023-08-23 16:33:35 <ikke> what does `ip route` return?]\
2023-08-23 16:33:41 <ascensionwires> it other wifi i choosed when installed
2023-08-23 16:34:13 <ascensionwires> lol now this other wifi is working
2023-08-23 16:34:46 <ascensionwires> but i need connect not from it but from hotspot
2023-08-23 16:34:55 <ascensionwires> how?
2023-08-23 16:35:53 <ascensionwires> first mb i need dusconnect from it
2023-08-23 16:36:04 <ascensionwires> and then try connect to hotspot
2023-08-23 16:36:24 <ikke> edit /etc/wpa_supplicant/wpa_supplicant.conf
2023-08-23 16:36:31 <ascensionwires> ok
2023-08-23 16:36:43 <ascensionwires> delete from it?
2023-08-23 16:37:23 <ikke> change the ssid and psk 
2023-08-23 16:37:38 <ascensionwires> to hotspot?
2023-08-23 16:37:47 <ikke> to the ssid of your hotspot, yes
2023-08-23 16:38:21 <ascensionwires> and what psk?
2023-08-23 16:38:45 <ikke> The one you have set for your hotspot
2023-08-23 16:39:45 <ascensio1wires> im here, i just reconnected
2023-08-23 16:40:15 <ascensio1wires> what psk?
2023-08-23 16:40:53 <zcrayfish> pre-shared key (i.e. the password)
2023-08-23 16:42:09 <ascensio1wires> mb i need wpa_passphrase 'ExampleWifiSSID' 'ExampleWifiPassword' > /etc/wpa_supplicant/wpa_supplicant.conf ?
2023-08-23 16:42:28 <dwfreed> note that > will wipe out the existing config
2023-08-23 16:42:35 <dwfreed> all of it
2023-08-23 16:42:54 <ascensio1wires> it not bad
2023-08-23 16:43:02 <elly> you can just edit the file normally and add the psk to your config
2023-08-23 16:43:15 <ascensio1wires> what if no password?
2023-08-23 16:44:10 <ascensio1wires> this hotspot without password i setted
2023-08-23 16:44:58 <ikke> try to set it empty psk=""
2023-08-23 16:45:29 <elly> you can set key_mgmt to NONE instead of WPA-PSK for that network, or you can try WPA-PSK with psk="", depending on how the hotspot is configured
2023-08-23 16:46:17 <ascensio1wires> mb good idea to do key_mgmt=NONE first
2023-08-23 16:46:27 <ascensio1wires> ok i do it now
2023-08-23 16:46:36 <ikke> after you modify the file, run `rc-service wpa_supplicant restart`
2023-08-23 16:47:00 <ascensio1wires> ok
2023-08-23 16:47:15 <ascensio1wires> it work
2023-08-23 16:47:33 <ascensio1wires> but i removed other wifi
2023-08-23 16:47:39 <elly> ikke: look at mx. fancy here remembering to use `rc-service` instead of running the init scripts like a peasant :)
2023-08-23 16:47:58 <ikke> heh
2023-08-23 16:49:15 <zcrayfish> 
ACTION hates typing the extra 3 letters and just uses the service symlink. :p

2023-08-23 16:49:34 <ascensio1wires> ikke
2023-08-23 16:49:36 <ascensio1wires> https://pasteboard.co/T3AEsuWi997g.jpg
2023-08-23 16:49:46 <ascensio1wires> this is my hotspot
2023-08-23 16:49:59 <ascensio1wires> and ping google.com work
2023-08-23 16:50:06 <zcrayfish> sweet
2023-08-23 16:50:31 <ascensio1wires> mb i understand issue
2023-08-23 16:50:31 <dwfreed> elly: I read 'mx' and immediately thought mail for some reason and not the gender-unspecified honorific
2023-08-23 16:50:53 <ascensio1wires> this network={}
2023-08-23 16:51:04 <ikke> oh, apparently you can also run wpa_cli reconfigure
2023-08-23 16:51:46 <elly> dwfreed: lol
2023-08-23 16:51:59 <elly> ikke: what does that do?
2023-08-23 16:52:34 <dwfreed> tells wpa_supplicant to reload its config
2023-08-23 16:52:36 <ikke> ^
2023-08-23 16:52:42 <elly> aha
2023-08-23 16:52:45 <elly> I've never used wpa_cli
2023-08-23 16:53:00 <ikke> I haven't hardly used it as well
2023-08-23 16:53:17 <ikke> just saw it in wpa_supplicant.conf(5)
2023-08-23 16:53:37 <ikke> and if you want one more option, you can also send SIGHUP
2023-08-23 16:54:33 <dwfreed> wpa_cli has 10,000 commands that do some weird and esoteric thing, and about 5 commands that do something useful to the average person
2023-08-23 16:54:39 <ascensionwires> this all very strange
2023-08-23 16:54:48 <ikke> dwfreed: my feeling exactly :D
2023-08-23 16:54:49 <ascensionwires> hotspot working
2023-08-23 16:55:08 <elly> I have always just edited wpa_supplicant.conf and restarted it by hand basically
2023-08-23 16:55:19 <ikke> elly: me three
2023-08-23 16:55:29 <ascensionwires> but now i deleted hotspot and setupped wifi back and it dont work
2023-08-23 16:55:42 <ikke> ascensionwires: how did you do that?
2023-08-23 16:55:48 <dwfreed> because setup-interfaces expects wpa_supplicant to be stopped
2023-08-23 16:56:02 <dwfreed> and probably doesn't reload it otherwise
2023-08-23 16:56:13 <ikke> yeah, it just starts it
2023-08-23 16:56:20 <ikke> which is a noop if it's already running
2023-08-23 16:57:27 <dwfreed> yep
2023-08-23 17:00:03 <ascensio1wires> now nothing setup
2023-08-23 17:00:12 <ascensio1wires> no wifi nor hotspot
2023-08-23 17:00:37 <ascensio1wires> i need reboot i think
2023-08-23 17:01:03 <elly> why are you doing this? you had it working a moment ago
2023-08-23 17:01:16 <elly> why did you delete the working config and rerun the setup script?
2023-08-23 17:01:24 <ascensio1wires> yea
2023-08-23 17:02:06 <ascensio1wires> oh no i dont deleted a file
2023-08-23 17:02:54 <ascensio1wires> i just do setup-interfaces and then check my wpa_supplicant.conf
2023-08-23 17:03:12 <invoked> i'm team iwd now, but that 'wpa_cli reconfigure' is a good one to file away for later.
2023-08-23 17:03:35 <ascensio1wires> now i rebooted and all work
2023-08-23 17:03:42 <ascensio1wires> i mean hotspot
2023-08-23 17:04:11 <ascensio1wires> mb running setup-interfaces two times is bad idea
2023-08-23 17:04:31 <elly> yes, it is
2023-08-23 17:04:39 <ascensio1wires> i need see wpa_supplicant.conf example file with two networks
2023-08-23 17:04:48 <elly> just write two network={...} blocks in it
2023-08-23 17:05:02 <elly> you don't need to do anything special
2023-08-23 17:05:09 <ascensio1wires> what i need write?
2023-08-23 17:05:29 <elly> the same stuff you were writing before? look at: man 5 wpa_supplicant.conf
2023-08-23 17:06:31 <ascensio1wires> no man on my system
2023-08-23 17:06:46 <ascensio1wires> https://man.archlinux.org/man/wpa_supplicant.conf.5
2023-08-23 17:06:57 <ascensio1wires> there two network={}
2023-08-23 17:07:37 <elly> indeed
2023-08-23 17:10:21 <zcrayfish> ascensio1wires: you likely need to install mandoc and wpa_supplicant-doc 
2023-08-23 17:10:24 <zcrayfish> to get the manpage
2023-08-23 17:10:27 <ascensio1wires> ok all work i needed reboot
2023-08-23 17:10:37 <ascensio1wires> it all i needed i think
2023-08-23 17:10:42 <elly> yay
2023-08-23 17:11:04 <ascensio1wires> there posdible to run setup-interfaces many times
2023-08-23 17:11:08 <ascensio1wires> all work
2023-08-23 17:12:31 <ascensio1wires> hope everything will be fine in the future
2023-08-23 17:34:13 <elly> if it stops working, you can read the error messages and debug it :)
2023-08-23 18:16:12 <clius[m]> <ikke> "clius: I know _a_ way" <- Thanks for the help apk fix --directory-permition $(apk info) helped for some reasons
2023-08-23 18:36:28 <ikke> clius[m]: ah, good to know
2023-08-23 18:36:40 <ikke> I initially wanted to suggest trying that as well, but I forgot
2023-08-23 18:59:57 <ajrycba> Hello. Does alpine run well on lenovo x13s out of the box?
2023-08-23 19:14:02 <zcrayfish> Hard to say. No alpine users have submitted that model to linux-hardware.org :x
2023-08-23 19:14:07 <zcrayfish> https://linux-hardware.org/?view=computers&vendor=Lenovo&model=ThinkPad+X13s+Gen+%28All%29 
2023-08-23 19:14:18 <zcrayfish> My areas of concern there would be the GPU and touchpad.
2023-08-23 19:47:10 <ajrycba> zcrayfish: thanks
2023-08-23 19:47:33 <zcrayfish> no problem, wish I had a more concrete answer for you!
2023-08-23 19:48:23 <ajrycba> Well, I think it's just a matter of time as ARM based laptops seem to be gaining in popularity 
2023-08-23 19:48:36 <ajrycba> Which I'm  glad about
2023-08-23 19:50:04 <ajrycba> I love the idea of running alpine linux with Sway+vim on a 13 or 14" fanless arm laptop.
2023-08-23 19:50:55 <ajrycba> Minimal, lightweight, quiet, cool.
2023-08-23 19:51:00 <zcrayfish> Yeah that's pretty sweet
2023-08-23 19:51:21 <ajrycba> One day soon :D
2023-08-23 19:59:28 <zcrayfish> I think my next non x86_64-based machine will be an ARM64 as well. maybe a smartphone I can put postmarketOS on.
2023-08-23 20:10:54 <ajrycba> next frontier, linux phone. One can dream.
2023-08-23 20:19:24 <zcrayfish> At least we're there on the kernel side of things.... Most of the rest of android leaves a lot to be desired.
2023-08-23 21:11:37 <M-u-m-p-i-t-z> [114,995:08] 99,99 7,99M-u-m-p-i-t-z:99,99 Hello, I will use a rasby with alpine linux and a  XMPP server (openfire) with intern (on the sdcard) database. How can I  save only the database without doing lbu ci command?
2023-08-23 21:11:37 <M-u-m-p-i-t-z> 14,99[15:08] 99,99 7,99M-u-m-p-i-t-z:99,99 Can anyone give me some support? 
2023-08-23 21:12:17 <M-u-m-p-i-t-z> So I found only lbu include 
2023-08-23 21:13:38 <M-u-m-p-i-t-z> But this save the whole system,any chance only save the database periodic?
2023-08-24 00:32:14 <fluix> probably better to look up your databaes's documentation and follow that
2023-08-24 00:32:31 <fluix> if it's just a file what's preventing you from just copying it?
2023-08-24 00:33:23 <fluix> oh, you mean save as in persist across reboots?
2023-08-24 03:02:05 <Hello71> I don't think that's possible exactly. as I recall lbu stores a tarball, so you can't update a single file. maybe you could try tar -r
2023-08-24 03:03:18 <Hello71> in general if you want to store a lot of files and update each one independently, then you need a filesystem, not a tarball
2023-08-24 06:19:56 <zcrayfish> https://wiki.alpinelinux.org/wiki/Elogind <-- Query: Are there any downsides to having elogind start in the boot runlevel instead of default?
2023-08-24 06:20:46 <zcrayfish> sddm's hibernate/standby/etc buttons are sometimes greyed out... Seemingly because elogind isn't started soon enough.
2023-08-24 06:27:45 <dwfreed> you can make sure that whatever starts sddm depends on elogind
2023-08-24 06:28:45 <zcrayfish> Well that makes it sound like a bug in sddm's openrc service then. :p
2023-08-24 06:31:22 <dwfreed> elogind has 'before display-manager'
2023-08-24 06:32:58 <dwfreed> enable openrc logging and make sure it's starting before sddm?
2023-08-24 06:33:50 <zcrayfish> will do.
2023-08-24 06:45:22 <zcrayfish> user error: had rc_parallel turned on. 
2023-08-24 06:58:32 <dwfreed> yeah, I wouldn't enable that; it's really not maintained
2023-08-24 09:23:11 <WhyNotHugo> busybox shell doesn't handle '-o pipefail'?
2023-08-24 09:23:49 <ikke> It does
2023-08-24 09:24:22 <ikke> Shellcheck in posix mode would complain it doesn't 
2023-08-24 09:25:17 <WhyNotHugo> Oh, I need to include 'set -e' for it to actually do what I want.
2023-08-24 09:25:47 <ikke> I almost always use set -eu -o pipefail
2023-08-24 09:40:10 <M-u-m-p-i-t-z> Hello71: thx than i can give up search and do backup somehow else
2023-08-24 10:28:35 <armin> ikke: i wouldn't just blindly set -e in scripts. catching error codes != 0 manually isn't exactly hard and gives you way more control, and i've seen people shooting theirselves in the foot by using it. 
2023-08-24 10:29:06 <armin> ikke: set -u, hm, yeah, debatable. set -o pipefail yea ok i get that one.
2023-08-24 10:30:15 <ikke> armin: I've seen the opposite as well
2023-08-24 10:30:27 <ikke> commands failing and the script just happily continues
2023-08-24 10:30:42 <armin> ikke: of course it does if you don't check for the error code.
2023-08-24 10:31:03 <ikke> I opt to use set -e instead
2023-08-24 10:31:09 <ikke> which abuild does as well
2023-08-24 10:31:47 <armin> ikke: well sometimes i actually need a script to continue in such a situation, pretty often indeed.
2023-08-24 10:32:19 <armin> ikke: it totally depends. if you know what you're doing, fine, use it, but i disagree with that "just always do this" thing.
2023-08-24 10:32:34 <ikke> That's why I said: 'amost always'
2023-08-24 10:32:42 <armin> fair.
2023-08-24 18:00:36 <linearcannon> running alpine on a ppc64le box and having some difficulty because it's seemingly not using my gpu
2023-08-24 18:01:13 <linearcannon> on other distros i have to blacklist the module for the built-in aspeed video because otherwise that gets used as the default device, and i don't have that plugged in to anything
2023-08-24 18:01:31 <linearcannon> i did that here as well, but it doesn't seem to be doing anything. does alpine build this in rather than as a module??
2023-08-24 18:02:16 <Forza> linearcannon: you can easily check with lsmod 
2023-08-24 18:02:57 <linearcannon> yeah, i don't see it in the module list
2023-08-24 18:08:09 <linearcannon> well that sucks for me
2023-08-24 18:09:29 <linearcannon> maybe this board will let me disable the onboard video
2023-08-24 18:09:35 <linearcannon> i know the more expensive version has a jumper for that
2023-08-24 18:10:19 <ikke> check /boot/config-lts
2023-08-24 18:10:27 <ikke> to see the config for that module
2023-08-24 18:11:04 <linearcannon> CONFIG_DRM_AST=y
2023-08-24 18:11:38 <ikke> That means it's built-in
2023-08-24 18:11:48 <linearcannon> yep
2023-08-24 18:12:33 <linearcannon> ok well, i don't feel like building my own kernel
2023-08-24 18:12:42 <linearcannon> and it looks like there is a jumper to disable the onboard video
2023-08-24 18:12:44 <linearcannon> so i'll just do that
2023-08-24 18:13:03 <linearcannon> i do enough kernel building at work :P
2023-08-24 18:14:14 <ikke> heh
2023-08-24 18:17:46 <Guest510> sooo, got an annoying problem trying to set up daphne. If I run `service daphne start` (either iwth start-stop or supervise) I get `xxx-daemon: fopen '/run/daphne.pid' No such file..` but if I run `/etc/init.d/daphne start` it runs as expected
2023-08-24 18:18:47 <Guest510> running `service -v` doesnt show much extra, the output makes it look like it was successful
2023-08-24 18:18:59 <Guest510> additionally the .pid file DOES exist
2023-08-24 18:39:32 <Guest510> oh, I guess that fopen warning was a red herring, as I've realized sshd, nginx, etc all say it too
2023-08-24 18:39:38 <Guest510> guess daphne is just crashing
2023-08-25 01:46:42 <hanetzer> linearcannon: talos?
2023-08-25 04:33:59 <linearcannon> hanetzer, blackbird
2023-08-25 04:35:12 <hanetzer> nice. gratz.
2023-08-25 04:43:13 <zcrayfish> I guess Leethium61624's insta join/parts ad infinitum give me incentive to configure my client to hide join/parts for people who never talk. heh
2023-08-25 10:54:24 <clius[m]> Hello my encrypted alpine with btrfs no longer boots. When I regenerate with mkinitramfs me the ramfs comes: "/lib/modules/6.4.10-1-edge does not exist or is not a directory" and it does not find the root specified by uuid when I boot.
2023-08-25 10:54:24 <clius[m]> Does anyone have the same problem?
2023-08-25 10:56:51 <ikke> clius[m]: what does uname -r return?
2023-08-25 11:09:39 <clius[m]> For that I first have to boot the system but it dont boot I could chroot into the system and see with apk the kernel version 
2023-08-25 11:10:47 <ikke> Ah, then you need to explicitly provide the kernel version 
2023-08-25 13:53:25 <clius[m]> I have... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/OvxCfmcBYSVYWeOINjBPCBYh>)
2023-08-25 14:11:48 <clius[m]> When I look in /lib/modules i see a different version http://0x0.st/H9LH.txt
2023-08-25 14:12:22 <ikke> I have... <Unusable url>
2023-08-25 14:14:56 <zcrayfish> 0x0.st is being a little funky today. it loaded for me... after forever.
2023-08-25 14:21:18 <ikke> I meant the matrix.org url
2023-08-25 14:21:50 <ikke> Which shows up for IRC users if you sent too long messages or multi-line messages 
2023-08-25 14:22:09 <ikke> And the url will force a download, it's effectively unusable 
2023-08-25 14:32:54 <zcrayfish> content-disposition: attachment; filename*=utf-8''clius%5Bm%5D_2023-08-25_13%3A53%3A25.txt
2023-08-25 14:32:56 <zcrayfish> thanks matrix.
2023-08-25 14:33:04 <zcrayfish> attachment -> inline dammit!
2023-08-25 14:34:01 <idefix> I am trying to install on hardware. It does not have ethernet, only wifi. Setting up wifi requires internet according to the wiki. How can I solve this chicken-egg situation?
2023-08-25 14:35:41 <herkulessi[m]> maybe add the apk files for the drivers to your installmedia?
2023-08-25 14:36:24 <idefix> I had no idea this was possible. Is there a guide for that?
2023-08-25 14:38:35 <Ermine> idefix: easier way: use extended iso
2023-08-25 14:38:53 <herkulessi[m]> You can install the files like this: https://wiki.alpinelinux.org/wiki/Alpine_Package_Keeper#Add_a_local_Package
2023-08-25 14:39:34 <herkulessi[m]> how you get those files I don't fully remember, but I somehow downloaded them from the repos
2023-08-25 14:40:55 <ikke> idefix: even standard should be able to config wifi nowadays 
2023-08-25 14:40:56 <idefix> I already use extended iso, but that doesn't seem to be enough.
2023-08-25 14:41:08 <ikke> What are you missing?
2023-08-25 14:45:12 <WhyNotHugo> When I first installed Alpine, I recall neededing to install linux-firmware, but the firmware directory was mounted RO.
2023-08-25 14:45:13 <idefix> From the wifi wiki page, iwd seems not to exist and from the wpa approach ip link only lists a "loopback" entry.
2023-08-25 14:45:16 <WhyNotHugo> Not sure if that's been fixed since.
2023-08-25 14:45:45 <WhyNotHugo> Also the procedure in the wiki resulting in me seeing only a few wifi networks -- I needed to use iwd to see mine.
2023-08-25 14:46:45 <Ermine> idefix: what wi-fi card do you have?
2023-08-25 14:48:13 <ikke> Earlier someone could connect to wifi from the installer with the standard iso
2023-08-25 14:50:45 <idefix> Idk, never had to check. Surface Pro 3 is the device type (i think)
2023-08-25 14:53:53 <idefix> MS spec site is not helpful, doesn't tell the card...
2023-08-25 14:54:48 <Ermine> Doesn't seem to 
2023-08-25 14:54:57 <Ermine> ... be brand new model
2023-08-25 14:56:10 <zcrayfish> which should make things easier...
2023-08-25 14:57:04 <Ermine> idefix: if you have windows on your device, you can check model in the device manager
2023-08-25 14:58:07 <idefix> I haven't in a long time. I'll try with a linux iso with a fancy desktop which might tell.
2023-08-25 14:58:45 <Ermine> So you had another distro with wi-fi working?
2023-08-25 15:00:27 <idefix> Yes. I unfortunatly wiped that already. today, after not using that hardware in months.
2023-08-25 15:01:32 <idefix> I think it was debian (nonfree), not sure
2023-08-25 15:03:08 <Ermine> from that distro, please send output of lspci -vvv
2023-08-25 15:03:33 <Ermine> through some pastebin site, of course
2023-08-25 15:30:42 <idefix2> http://0x0.st/H9L0.txt @Ermine
2023-08-25 15:32:33 <Ermine> So, your chip is Marvell 88W8897, kernel module mwifiex_pcie
2023-08-25 15:35:52 <Ermine> Can you try to modprobe it under Alpine?
2023-08-25 15:36:53 <clius[m]> <ikke> "I have... <Unusable url>" <- Problem solved I linked /lib/modules/6.4.12-0-edge to /lib/modules/6.4.10-1-edge and had other problems with an other installation
2023-08-25 15:39:53 <idefix> Sorry I'm such a noob with low level stuff. Not sure how to use modprobe, what exactly does it expect?
2023-08-25 15:40:16 <ikke> it expects a module name
2023-08-25 15:40:19 <Ermine> modprobe mwifiex_pcie
2023-08-25 15:41:06 <ikke> one issue could be that the running kernel version is different from the available modules
2023-08-25 15:41:23 <idefix> That outputs nothing
2023-08-25 15:42:01 <Ermine> Use ip link now. Does wlan device appear?
2023-08-25 15:43:12 <idefix> No, nothing changed. I doublechecked the modprobe command, no typo.
2023-08-25 15:43:36 <ikke> does `lsmod` list the module?
2023-08-25 15:43:55 <Ermine> lsmod | grep mwifiex_pcie
2023-08-25 15:44:50 <idefix> At least grep I know :) Yes, listed.
2023-08-25 15:45:28 <Ermine> check dmesg for messages related to this module then
2023-08-25 15:48:32 <idefix> Yes, dmesg shows some things. "Direct firmware load for mrvl/pcie8897_uapsta.bin failed with error -2"
2023-08-25 15:49:42 <idefix> The "Failed to get firmware mrvl/...bin"
2023-08-25 15:49:51 <idefix> *Then
2023-08-25 15:50:41 <Ermine> 
ACTION checks what this code means

2023-08-25 15:51:07 <idefix> Also other stuff, but I guess these two lines are the relevant ones?
2023-08-25 15:51:54 <Ermine> Yes
2023-08-25 15:53:00 <Ermine> Apparently firmware is missing
2023-08-25 15:53:24 <Ermine> try apk add linux-firmware
2023-08-25 15:53:47 <Ermine> (you are on extended iso I guess?)
2023-08-25 15:54:24 <idefix> Yes, I am. Apk now gives looots of errors but keeps going.
2023-08-25 15:54:47 <idefix> 316 errors updating directory permissions
2023-08-25 15:55:29 <Ermine> try stat /lib/firmware/mrvl/pcie8897_uapsta.bin
2023-08-25 15:55:50 <Ermine> Just in case
2023-08-25 15:56:27 <idefix> No such file or directory
2023-08-25 15:57:12 <idefix> All entries in /mrvl start either with sd or usb
2023-08-25 15:58:19 <Ermine> Which Alpine version do you use?
2023-08-25 15:59:43 <idefix> I downloaded alpine-extended-3.18.3-x86_64.iso just today.
2023-08-25 16:00:00 <idefix> Thats what I booted
2023-08-25 16:00:45 <Ermine> And please try apk list -I | grep linux-firmware-mrvl
2023-08-25 16:02:17 <idefix> 1 entry: ...mrvl-20230515-r6 ... (custom) [installed]
2023-08-25 16:03:03 <Ermine> Weird situation
2023-08-25 16:03:39 <idefix> At this point I really gotta thank u for your patience :)
2023-08-25 16:04:53 <Ermine> :)
2023-08-25 16:06:34 <Ermine> Well, internet connection is crappy where I am, so it will be downloaded in a hour 
2023-08-25 16:07:37 <Ermine> an hour*
2023-08-25 18:34:11 <WhyNotHugo> Is there anything like a GUI for viewing zips and stuff?
2023-08-25 18:34:19 <WhyNotHugo> A kind of winzip-for-linux GUI basically?
2023-08-25 18:34:35 <WhyNotHugo> file-roller looks nice, but you can't drag files out of zips, which makes it pretty useless.
2023-08-25 18:36:40 <kwadroke[m]> have you looked at xarchiver?
2023-08-25 18:39:19 <WhyNotHugo> kwadroke[m]:  Dragging files out doesn't work
2023-08-25 18:39:56 <kwadroke[m]> I know you can drag files in. never tried to drag them out
2023-08-25 18:40:33 <kwadroke[m]> and yes, you can drag them out, depending on your file manager
2023-08-25 18:41:14 <kwadroke[m]> but they do create any folders
2023-08-25 18:41:43 <kwadroke[m]> just tried with xarchiver and pcmanfm
2023-08-25 18:46:19 <WhyNotHugo> Oh, apparently file-roller is known to now work (issue 5 years old): https://gitlab.gnome.org/GNOME/file-roller/-/issues/4
2023-08-25 18:46:42 <WhyNotHugo> Wow, and this is the default on a GNOME setup. I can understand why folk coming from windows think that Linux is terrible; dragging a file out of a ZIP is a very common and simple UX.
2023-08-25 18:46:58 <ikke> s/now/not I suppose
2023-08-25 18:47:26 <kwadroke[m]> don't use Gnome so not sure how it handles things. usually use IceWM/JWM or XFCE
2023-08-25 18:48:11 <kwadroke[m]> Gnome UX is... different
2023-08-25 18:48:19 <kwadroke[m]> across the board
2023-08-25 18:49:19 <WhyNotHugo> Using `unzip $FILENAME` is usually fine by me, but firefox auto-opened the zip, and I figured it might be faster to drag the file out into nemo.
2023-08-25 18:49:41 <WhyNotHugo> xarchiver didn't work with nemo. I wonder if nemo has issues here too.
2023-08-25 18:50:24 <ayushnix> lol, that issue tracker is a disaster
2023-08-25 18:50:28 <kwadroke[m]> if firefox auto-opened it, beware it's probably in /tmp somewhere and not in ~/Downloads
2023-08-25 18:50:39 <ayushnix> filled with people pestering devs
2023-08-25 18:50:47 <kwadroke[m]> It's probably Nemo and not file-roller then
2023-08-25 18:51:55 <WhyNotHugo> kwadroke[m]: People on that bug have been reproducing the issue with file roller and all sorts of file managers (including nautilus) for years.
2023-08-25 18:52:05 <WhyNotHugo> It's either file roller or both.
2023-08-25 18:52:11 <WhyNotHugo> omg https://gitlab.gnome.org/AlynxZhou/file-roller/-/commit/80f53ece6714c89f604a80d60a2153e7599060fd
2023-08-25 18:52:46 <kwadroke[m]> pcmanfm and thunar has worked for me for many years now
2023-08-25 20:22:54 <b-over-c> hello all, im attempting build linux-lts-dev-6.1.43 through apk
2023-08-25 20:24:01 <b-over-c> this is a rollback, as we are currently on 6.1.47, so im attempting to revert the history in aports to that version
2023-08-25 20:25:56 <b-over-c> however, it seems that I would have to clone the whole aports repo to do this. _is there any way to clone one specific tree in aports?_
2023-08-25 20:27:16 <b-over-c> or is there a way to get prior APKBUILDs without going through aports? For context, the aports file is way too large for me to download currently
2023-08-25 20:28:17 <quinq> There's also a web interface to the repository
2023-08-25 20:32:38 <b-over-c> quinq thanks for the heads up
2023-08-25 23:36:56 <b-over-c> another question, how does one get build to break world?
2023-08-25 23:37:15 <b-over-c> *abuild to break world
2023-08-25 23:54:43 <lazywalker> any logs or abuild?
2023-08-25 23:55:45 <lazywalker> *of
2023-08-26 00:09:41 <b-over-c> lazywalker: i'll post stdout+stderr
2023-08-26 00:09:56 <b-over-c> >>> linux-lts: Building builder/linux-lts 6.1.43-r0 (using abuild 3.11.1-r0) started Sat, 26 Aug 2023 00:05:13 +0000 >>> linux-lts: Checking sanity of /home/builder/compile-this/APKBUILD... >>> linux-lts: Analyzing dependencies... ERROR: unable to select packages:   gcc-12.2.1_git20220924-r10:     breaks: .makedepends-linux-lts-20230826.000519[gcc>=13.1.1_git20230624]     satisfies: g++-12.2.1_git20220924-r10[gcc=12.2.1_git20220924
2023-08-26 00:10:42 <b-over-c> edepends-linux-lts-20230826.000519:     masked in: cache     satisfies: world[.makedepends-linux-lts=20230826.000519] >>> ERROR: linux-lts: builddeps failed >>> linux-lts: Uninstalling dependencies... ERROR: No such package: .makedepends-linux-lts
2023-08-26 00:12:16 <b-over-c> to summarize, the package im attempting to install won't break necessary functionality - but abuild is blocking it, so I would like to override it
2023-08-26 00:12:49 <b-over-c> however, there doesn't seem to be a "broken-world" option anywhere
2023-08-26 00:23:08 <minimal> b-over-c: you're trying to build linux-lts 6.11.43 which has a build dep on gcc>=13.1.1_git20230624 but you have gcc 12.2.1_git20220924 installed
2023-08-26 00:23:32 <minimal> s/6.11.43/6.1.43/
2023-08-26 00:37:02 <minimal> b-over-c: you're trying to build linux-lts 6.1.43 which has a build dep on gcc>=13.1.1_git20230624 but you have gcc 12.2.1_git20220924 installed
2023-08-26 00:39:42 <b-over-c> minimal: yes, but i am currently on 6.1.43 - which is why I would like to override the program
2023-08-26 00:40:38 <minimal> eh? the APKBUILD declares that gcc >- 13.1.1_git20230624 is *NEEDED* to build it
2023-08-26 00:41:03 <minimal> it obviously needs that version for a reason
2023-08-26 00:41:19 <b-over-c> yeah, im also confused about that
2023-08-26 00:41:27 <minimal> why are you confused?
2023-08-26 00:41:38 <b-over-c> I built it using 12.2.1
2023-08-26 00:41:50 <minimal> you built it how?
2023-08-26 00:42:28 <minimal> and if you built it already then why do you need to build it again?
2023-08-26 00:44:26 <b-over-c> I would like the headers for 6.1.43, and im using abuild. the process seems to get the headers from the kernel once its built
2023-08-26 00:45:29 <minimal> you didn't answer my question, you built it how?
2023-08-26 00:45:41 <b-over-c> its an automated process, and I would like not be to involved in changing it
2023-08-26 00:46:13 <minimal> I am confused as to your problem
2023-08-26 00:47:48 <minimal> why do you not have gcc 13.1.1 installed?
2023-08-26 00:48:06 <minimal> are you on Alpine Edge or 3.18.x?
2023-08-26 00:48:31 <b-over-c> I would like to keep gcc 12.2.1 and install 6.1.43 headers, so I would like to know if it is possible to break world with abuild
2023-08-26 00:48:40 <b-over-c> on 3.18.3
2023-08-26 00:49:14 <b-over-c> if its not possible, then its unfortunate but workable
2023-08-26 00:49:16 <minimal> so why are you using the linux-lts APKBUILD from Edge rather than 3.18.x?
2023-08-26 00:49:42 <b-over-c> 6.1.43 is from 3.18.3
2023-08-26 00:49:52 <b-over-c> current is 6.1.47
2023-08-26 00:51:08 <minimal> the Edge version of linux-lts' APKBUILD declared a makedepends on gcc 13.1.1, the Alpine 3.18 linux-lts' APKBUILD for 6.1.43 did NOT
2023-08-26 00:52:38 <minimal> you're using a APKBUILD from Alpine Edge, not from Alpine 3.18
2023-08-26 00:54:12 <b-over-c> you're correct, thanks for the help
2023-08-26 00:54:30 <b-over-c> I mistook master for stable, when its actually edge
2023-08-26 00:54:57 <b-over-c> this isn't something uncommon, so I should've checked for that
2023-08-26 00:55:24 <minimal> obviously linux-lts on 3.18 cannot makedepends on gcc 13.1.1 as Alpine 3.18 still uses gcc 12.2.1
2023-08-26 00:56:12 <b-over-c> yeah that's what I was saying, I just didn't want to go through the trouble of explaining it
2023-08-26 00:56:37 <b-over-c> you were a massive help, thanks again
2023-08-26 00:56:53 <minimal> eh? you didn't want to go to the trouble of explaining your problem so someone could understand it and help?
2023-08-26 00:59:31 <b-over-c> the gcc difference was enough to point out the flaw, which you did. how I built my kernel doesn't seem relevant
2023-08-26 01:00:19 <minimal> well it seems relevant if you managed to build linux-lts with the "wrong" gcc version than that *required* by the APKBUILD as this didn't make sense
2023-08-26 01:00:37 <b-over-c> but as I said prior, you were a huge help
2023-08-26 05:38:43 <ascensionwires> how organize energy saving in my system?
2023-08-26 05:49:12 <elly> powertop is a good place to start
2023-08-26 05:56:08 <zcrayfish> We don't know what your system is.
2023-08-26 05:56:14 <zcrayfish> laptop?
2023-08-26 05:58:52 <ascensionwires> tlp mb, it possible with defult configurtion just add, rc-update add tlp, rc-service tlp start
2023-08-26 06:00:42 <ascensionwires> zcrayfish : yes, laptop, thinkpad t480 with i5 and intel uhd 620 gpu
2023-08-26 06:06:15 <zcrayfish> n.b.: If you're just using it to change power settings upon AC connect/disconnect, the rc-service is useless. tlp's udev rules should already be handling this for you.
2023-08-26 06:06:33 <zcrayfish> however it shouldn't hurt anything regardless. \o/
2023-08-26 06:59:15 <ascensionwires> i have libudev-zero :(
2023-08-26 07:17:34 <zcrayfish> Took him over an hour to goalpost that time.
2023-08-26 09:32:24 <ascensionwires> powertop not work for me for some reason
2023-08-26 09:33:11 <ascensionwires> i run doas powertop --htpl=power_report.html and it return error
2023-08-26 09:33:35 <ascensionwires> modprobe cpufreq_stats failed
2023-08-26 09:33:51 <ascensionwires> Failed to mount debugfs
2023-08-26 09:36:01 <ikke> `ls -d lib/modules/$(uname -r)`
2023-08-26 09:36:09 <ikke> `ls -d /lib/modules/$(uname -r)`
2023-08-26 09:36:58 <ascensionwires> edge 6.4.12
2023-08-26 09:37:14 <ikke> but does that directory existr?
2023-08-26 09:37:58 <ascensionwires> yeah
2023-08-26 09:57:53 <zcrayfish> Use LTS.
2023-08-26 10:03:49 <ascensionwires> what difference between lts and just linux flavor?
2023-08-26 10:04:18 <ikke> The official alpine kernel is linux-lts (which is mainline linux, lts branch)
2023-08-26 10:04:35 <ikke> There is also an -edge kernel, which tracks the latest mainline version
2023-08-26 10:04:57 <ikke> But there are also configuration differences between lts and edge
2023-08-26 10:05:01 <zcrayfish> config-edge:# CONFIG_DEBUG_FS is not set
2023-08-26 10:05:10 <zcrayfish> FWIW.
2023-08-26 11:42:41 <ascensionwires> after all is it way enable debugfs in edge kernel? 
2023-08-26 11:43:12 <ascensionwires> there on gentoo wiki https://wiki.gentoo.org/wiki/PowerTOP#Kernel
2023-08-26 11:47:11 <ascensionwires> i was thrown out of the chat due to network problems:( sry
2023-08-26 11:51:58 <zcrayfish> It would be substantially easier to use the lts kernel; which has debugfs.
2023-08-26 12:01:39 <ikke> otherwise you would need to recompile the kernel and install it
2023-08-26 12:04:05 <ascensionwires> how disable bluetooth?
2023-08-26 12:05:17 <ikke> rfkill
2023-08-26 12:35:02 <ascensionwires> is it on boot?
2023-08-26 12:35:16 <ascensionwires> as i see it until first reboot
2023-08-26 12:52:17 <ascensionwires> how disable it after reboot
2023-08-26 12:52:19 <ascensionwires> ?
2023-08-26 12:52:49 <Ermine> rfkill
2023-08-26 12:54:01 <ascensionwires> but it work until first reboot
2023-08-26 12:56:26 <Ermine> You can try writing openrc service which will run rfkill on each boot
2023-08-26 13:05:10 <ascensionwires> really
2023-08-26 13:10:05 <ikke> there is a local service
2023-08-26 13:21:39 <ikke> You can put a script in /etc/local.d, enable the local service, and then that script gets executed at boot
2023-08-26 13:22:09 <ikke> another option would be to blacklist the kernel module used by the bluetooth card, but often that's shared with wifi, so you'd lose wifi as wel
2023-08-26 13:35:55 <ascensionwires> i did local.d disable-bluetooth.start, containing only shebang and rfkill block bluetooth, then i runned rc-update add local
2023-08-26 13:36:26 <ikke> did you make the script executable?
2023-08-26 13:37:25 <ascensionwires> really o+x
2023-08-26 13:38:58 <ascensionwires> how i can setup brightness?
2023-08-26 13:40:31 <ikke> apk add brightnessctl
2023-08-26 13:41:28 <ascensionwires> mb i can without?
2023-08-26 13:41:46 <ascensionwires> just changing backlight? or is bad method?
2023-08-26 13:48:07 <elly> you can use any method that works for you I think? it depends what behavior you want
2023-08-26 13:48:36 <Ermine> Of course you can have fun figuring out what to write to /sys
2023-08-26 13:49:06 <armin> ikke: oh don't remind me of having to run a simple command on boot on a systemd based system
2023-08-26 13:50:26 <armin> i mean what about /etc/local to begin with? why did most distributions purposefully remove that?
2023-08-26 13:50:47 <elly> Ermine: that's what I do personally
2023-08-26 13:51:38 <Ermine> elly: you like it I guess?
2023-08-26 13:51:57 <ascensionwires> i want something like default brightness
2023-08-26 13:52:13 <Ermine> What is it?
2023-08-26 13:52:43 <ascensionwires> like i can change /sys/class/backlight/intel_backlight/brightness
2023-08-26 13:52:48 <minimal> default brightness is the default, that's why it is called "default" lol
2023-08-26 13:53:12 <minimal> if you change it then it is no longer default
2023-08-26 13:54:02 <ascensionwires> i mean after rbooting this file i change no more as i writed
2023-08-26 13:54:34 <ascensionwires> it always changed to 549
2023-08-26 13:54:57 <Ermine> Time for another local.d script
2023-08-26 13:55:04 <ascensionwires> ok
2023-08-26 13:56:13 <ascensionwires> like echo <brightness i want> > /sys/class/backlight/intel_backlight/brightness  in script file?
2023-08-26 13:58:46 <elly> Ermine: it works for me :P
2023-08-26 13:59:01 <elly> I have a glorified shell script that manages my brightness
2023-08-26 14:00:29 <Ermine> The problem is that it depends on hardware you have
2023-08-26 14:00:48 <Ermine> Afaik there are different paths and different scales
2023-08-26 14:06:30 <elly> yes, there are
2023-08-26 14:33:48 <ascensionwires> Ermine : bash script possible to check what hardware
2023-08-26 14:34:42 <Ermine> ascensionwires: why? you have only one piece of hardware, just write the script which matches yours
2023-08-26 14:36:50 <ascensionwires> ok
2023-08-26 18:50:19 <ascensionwires> how ude 
2023-08-26 18:50:30 <ascensionwires> how use upower?
2023-08-26 18:50:58 <ascensionwires> there message cannot connect to upowerd
2023-08-26 20:00:09 <somercet1> okay, I noticed earlier: when you run caja FM from wmenu, which has DISPLAY set in the env (as it should), it runs under Xwayland, even though it is GTK3.
2023-08-26 20:00:58 <somercet1> While our internet was down from a storm, and hexchat was off, I saw that deadbeef music player also switched on Xwayland.
2023-08-26 20:01:24 <somercet1> But the DB window itself was not under X11, because it did not respond to xprop.
2023-08-26 20:01:55 <somercet1> So I ran DB from the command line, without DISPLAY,  and it run without triggering Xwayland.
2023-08-26 20:02:03 <somercet1> But now none of the hotkeys work.
2023-08-26 20:02:06 <somercet1> ideas?
2023-08-26 20:05:20 <somercet1> I built and installed caja into /usr/local and now it creates its window in wayland from wmenu.
2023-08-26 22:58:43 <ascensionwires> gn
2023-08-26 23:20:55 <uuidNuniq> asarandi: GM
2023-08-26 23:25:10 <Guest747> I have a mostly default alpine installation that gets its ip address from ethernet and dhcp. I want to permanently change the dhcp provided dns server to something else. How do I do that?
2023-08-26 23:25:44 <uuidNuniq> Guest747: echo 'nameserver x.x.x.x' /etc/resolv.conf.head
2023-08-26 23:26:03 <Guest747> I don't think that's permanent because next time I reboot the dhcp client will overwrite the file.
2023-08-26 23:26:14 <Guest747> Or does that not happen on alpine?
2023-08-26 23:26:24 <uuidNuniq> Guest747: look again
2023-08-26 23:26:37 <Guest747> ".head"?
2023-08-26 23:26:46 <uuidNuniq> echo 'nameserver x.x.x.x' > /etc/resolv.conf.head
2023-08-26 23:26:48 <Guest747> I see. Where is that file documented?
2023-08-26 23:27:33 <Guest747> I'm used to editing /etc/resolvconf.conf but that isn't used by default on Alpine. First time I see the resolv.conf.head file.
2023-08-26 23:27:57 <uuidNuniq> Guest747: that's a common setting. but it don't think it works for you. 
2023-08-26 23:28:10 <uuidNuniq> the default dhcp program is udhcpc
2023-08-26 23:28:33 <Guest747> I figured that out. I didn't figure out how to tell it to use a different dns server.
2023-08-26 23:28:52 <Guest747> Is resolv.conf.head read by udhcpc or by libc when resolving a name?
2023-08-26 23:29:17 <Guest747> Thanks for your help btw. Just curious about where that file comes from.
2023-08-26 23:29:26 <uuidNuniq> Guest747: do this : mkdir -p /etc/udhcpc ; printf '%s' 'dns="8.8.8.8 1.1.1.1"' > /etc/udhcpc/udhcpc.conf
2023-08-26 23:30:25 <uuidNuniq> it is not documented, but it is embedded in /usr/share/udhcpc/default.script
2023-08-26 23:30:34 <Guest747> Should alpine populate this config file by default?
2023-08-26 23:30:49 <Guest747> I'm used to config files in etc existing already after package installation.
2023-08-26 23:32:07 <uuidNuniq> Guest747: dhcp by default uses ifupdown-ng to configure, and then uses udpcpc because it is included in busybox. and then run /usr/share/udhcpc/default.script, and then read /etc/udhcpc/udhcpc.conf
2023-08-26 23:32:37 <uuidNuniq> just do this: do this : mkdir -p /etc/udhcpc ; printf '%s' 'dns="8.8.8.8 1.1.1.1"' > /etc/udhcpc/udhcpc.conf
2023-08-26 23:32:48 <uuidNuniq> and then reboot. 
2023-08-26 23:33:14 <Guest747> I understand your suggestion. Thanks. Still as feedback to whoever maintains this on Alpine, I think this file should be created by default.
2023-08-26 23:33:47 <uuidNuniq> it is not a fallback... it is a replacement for upstream dns from dhcp server.
2023-08-26 23:34:55 <Guest747> I'm not saying it is a fallback. It is a config file that Alpine's default dhcp setup uses. Just like config files for other services in /etc, I would expect this folder and file to exist.
2023-08-26 23:35:06 <uuidNuniq> Guest747: don't expect busybox to document anything... because it has no document other than --help
2023-08-26 23:35:29 <Guest747> That's bad. I don't like using software that has no documentation.
2023-08-26 23:35:50 <uuidNuniq> Guest747: if you installed dhcpcd, then use this file to replace dns : /etc/resolv.conf.head 
2023-08-26 23:36:06 <Guest747> I just have whatever Alpine has by default. 
2023-08-26 23:36:09 <uuidNuniq> because when dhdpcd is installed, then udhcpc will not be used at all. 
2023-08-26 23:37:29 <uuidNuniq> Guest747: using default is good, but udhcpc has a problem, if the cable is disconnected and then connected again, it won't detect.
2023-08-26 23:38:33 <Guest747> Btw, instead of rebooting you can restart the networking service.
2023-08-26 23:38:52 <Guest747> That makes me feel more and more that Alpine shouldn't be using udhcpc.
2023-08-26 23:39:09 <uuidNuniq> anyways: udhcpc : mkdir -p /etc/udhcpc ; printf '%s' 'dns="8.8.8.8 1.1.1.1"' > /etc/udhcpc/udhcpc.conf . dhcpcd :  echo 'nameserver x.x.x.x' > /etc/resolv.conf.head
2023-08-26 23:39:27 <uuidNuniq> so i give 2 options. 
2023-08-26 23:39:27 <Guest747> The udhcpc.conf approach worked.
2023-08-26 23:42:05 <uuidNuniq> Guest747: yes, it works, by digging into /usr/share/udhcpc/default.script
2023-08-26 23:43:18 <uuidNuniq> you have to read the script to understand why it works if that's your pursuing
2023-08-26 23:44:17 <Guest747> It would be nicer if I didn't have to read the script. Like being documented in the Alpine wiki or a man page.
2023-08-26 23:44:44 <uuidNuniq> alpine wiki is poorly maintained.
2023-08-26 23:44:47 <Guest747> I appreciate your help and you've solved the problem. Again, this is feedback for Alpine.
2023-08-26 23:46:44 <uuidNuniq> Guest747: https://wiki.alpinelinux.org/wiki/Udhcpc
2023-08-26 23:47:01 <uuidNuniq> that wiki explains a little about why... 
2023-08-26 23:53:42 <uuidNuniq> Guest747: alpine may be small, but it is not simple :)
2023-08-26 23:54:17 <uuidNuniq> that is to say, sometimes, you got to read the script... mostly shell script.
2023-08-26 23:55:01 <uuidNuniq> if you want to get deeper and deeper
2023-08-27 07:43:32 <quinq> EvgenyK, having connection problems?
2023-08-27 08:25:10 <qaqland[m]> Is there any possible that have the package both in main and community?
2023-08-27 08:25:41 <ikke> Only with a different name
2023-08-27 08:30:10 <qaqland[m]> got it
2023-08-27 09:52:32 <ascensionwires> is there fastfetch in alpine repos?
2023-08-27 09:52:57 <ikke> https://pkgs.alpinelinux.org/contents?file=fastfetch&path=&name=&branch=edge
2023-08-27 09:53:04 <ikke> https://pkgs.alpinelinux.org/packages?name=fastfetch&branch=edge&repo=&arch=&maintainer=
2023-08-27 09:55:35 <ascensionwires> :|
2023-08-27 13:14:44 <qaqland[m]> Some packages are flagged out of date
2023-08-27 13:15:32 <qaqland[m]> Are they manually tagged or automatically
2023-08-27 13:15:54 <ikke> qaqland[m]: can be both
2023-08-27 13:16:46 <qaqland[m]> :)
2023-08-27 13:31:19 <ikke> qaqland[m]: It uses Anitya to automatically notify for updated versions for packages that it knows about
2023-08-27 13:31:26 <ikke> https://release-monitoring.org/
2023-08-27 14:58:43 <ken> Hi. I've been having issues with connecting to external monitors in alpine. They just render the first frame and freeze. Everything is frozen except the cursor on the external display. Is this a know bug? 
2023-08-27 14:59:20 <ikke> ken: Haven't heard of it before. Sounds like a driver / gpu issue
2023-08-27 15:00:51 <ken> I've used arch on this config before and never ran into this issue. I even tried reproducing it on other musl based distros (void musl) but it seems to work fine. 
2023-08-27 15:01:10 <minimal> ken: which graphics card / linux driver?
2023-08-27 15:01:31 <ikke> and which graphical environment
2023-08-27 15:02:28 <ken> minimal: Nvidia RTX 3050 Ti with nouveau
2023-08-27 15:02:34 <ken> ikke: dwm 
2023-08-27 15:02:43 <ken> i tried with wayland as well but same issue 
2023-08-27 16:54:22 <Apachez> where can I find online what the config file for kernel compile looks like for the rpi images (aarch64, armhf and armv7) ?
2023-08-27 16:55:28 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/main/linux-rpi
2023-08-27 16:55:40 <ikke> Note that it's a diff compared to the default config
2023-08-27 16:56:18 <Apachez> and the default config is then located online?
2023-08-27 16:57:06 <Apachez> https://gitlab.alpinelinux.org/alpine/aports/-/tree/master/main/linux-lts ?
2023-08-27 16:57:50 <ikke> The mainline kernel source + https://dev.alpinelinux.org/archive/rpi-patches/rpi-6.1.46-alpine.patch
2023-08-27 16:59:08 <minimal> Apachez: inside the actual APK package is /boot/config-lts which is the full config
2023-08-27 16:59:48 <minimal> so either install the package or extract that file from the APK
2023-08-27 17:03:28 <Apachez> tnx
2023-08-27 17:13:00 <ashpool> is there a command to remove a package from a local repository (removing the APK and the entry in APKINDEX.tar.gz)?
2023-08-27 17:14:19 <ikke> ashpool: are you using abuild?
2023-08-27 17:14:24 <ashpool> yes
2023-08-27 17:14:29 <ikke> abuild cleanpkg
2023-08-27 17:14:34 <ikke> in the package directory
2023-08-27 17:14:40 <ashpool> thanks
2023-08-27 17:14:46 <ikke> or abuild cleanoldpkg if the version is different
2023-08-27 20:05:22 <shrizza> Someone's spamming the apk-tools issues area with torrent links.
2023-08-27 20:05:35 <ikke> shrizza: thanks, taking a look
2023-08-27 20:09:15 <ikke> shrizza: cleaned up
2023-08-27 20:09:19 <ikke> and blocked
2023-08-27 20:09:33 <shrizza> Thanks!
2023-08-28 06:16:06 <zcrayfish> ikke: can I get Gopherbot added to the Bots group on the wiki?
2023-08-28 06:25:57 <ikke> What does gopherbot do?
2023-08-28 06:30:08 <zcrayfish> Right now I only have it setup to audit dead links (after a week of a link being persistently dead it will print a notice to the talk page) and to remove references to unused categories.
2023-08-28 06:30:42 <quinq> Does it provide gopher access to the wiki?
2023-08-28 06:31:29 <zcrayfish> quinq: You can use your user javascript (an actual MediaWiki feature) and write your own ;)
2023-08-28 06:33:55 <quinq> My own what?
2023-08-28 06:34:42 <zcrayfish> lol
2023-08-28 06:35:10 <zcrayfish> You can write your own gopher client in javascript. it would be insane
2023-08-28 06:35:17 <quinq> eeek
2023-08-28 06:35:17 <zcrayfish> I was being silly
2023-08-28 06:36:11 <quinq> But I was talking about the wiki, not the client, there already exist clients, but as you talked about a bot named “gopherbot”, I'm just asking if it's about providing gopher access
2023-08-28 06:39:19 <zcrayfish> No it's just about automating the tedious task of dealing with dead links and occasionally removing unused categories.
2023-08-28 06:40:12 <zcrayfish> I couldn't think of a better name.
2023-08-28 06:40:25 <quinq> oh ok :)
2023-08-28 06:40:57 <quinq> Could be molebot ;)
2023-08-28 06:41:42 <quinq> or garrarufabot
2023-08-28 06:42:07 <zcrayfish> BottyMcBotFace? ;p
2023-08-28 06:42:19 <zcrayfish> BotBot?
2023-08-28 06:49:50 <quinq> :)
2023-08-28 08:38:08 <quinq> Hola
2023-08-28 08:38:35 <quinq> Something strange is happening on *one* of my machines, its dmesg tells me we're friday September 1st
2023-08-28 08:38:45 <quinq> date (local and UTC) is correct though
2023-08-28 08:39:04 <quinq> Any idea what's happening and/or how I could debug that?
2023-08-28 08:41:11 <zcrayfish> Weird. Is your machine inside some kind of DeLorean? :)
2023-08-28 08:41:19 <quinq> :D
2023-08-28 08:41:31 <quinq> It's $work laptop
2023-08-28 08:43:20 <quinq> http://ix.io/4EJY
2023-08-28 08:46:13 <zcrayfish> how does your /proc/uptime compare to the most recent message in dmesg (no flags)?
2023-08-28 08:46:51 <Apachez> is that machine running as vm guest or on bare metal?
2023-08-28 08:47:04 <Apachez> and is ntp setup to sync and save hwclock ?
2023-08-28 08:47:11 <quinq> zcrayfish, it looks correct
2023-08-28 08:47:37 <Apachez> could also be the onboard RTC battery or is getting low on power or such
2023-08-28 08:47:54 <Apachez> check with "sensors" if its within range
2023-08-28 08:48:14 <quinq> So that's a problem of strftime
2023-08-28 08:48:15 <Apachez> "normally" the RTC without proper sync will drift like 2 minutes/week and still be "ok"
2023-08-28 08:48:45 <quinq> Apachez, it's an actual computer
2023-08-28 08:48:56 <Apachez> using rpi who lacks rtc but have that as 3rd party addon those bastards drifts like more than 2 minutes/day
2023-08-28 08:48:57 <quinq> Running chronyd
2023-08-28 08:49:41 <quinq> It doesn't have a CMOS battery sensors :>
2023-08-28 08:55:38 <Apachez> whats output of "chronyc sources" ?
2023-08-28 08:56:22 <Apachez> note that if you use nts thats 2 different ports (TCP4460 for NTS-KE and UDP4123 for NTS)
2023-08-28 08:56:33 <Apachez> where regular ntp is UDP123
2023-08-28 08:57:15 <quinq> I'm using chronyd default conf
2023-08-28 08:57:58 <quinq> http://ix.io/4EK1
2023-08-28 08:58:22 <quinq> RTC is sync to local time
2023-08-28 08:58:42 <quinq> (seems to be the default on void)
2023-08-28 11:08:36 <Ihateunixihateu> can anyone explain why alpine and epiphany dont get along on my raspberry pi? all it says is bwrap cant create symlink at var/run file already exists. no such file can be found when I search for it, nor had any help from chatgpt explained what to do other than install a worse browser. I wish ungoogled chromium worked but according to the forums its not supported in any way and this seems like the only alternative to firefox th
2023-08-28 11:08:41 <Ihateunixihateu> youtube.
2023-08-28 14:36:25 <CcxWrk> zcrayfish: traditionally this kind of bot is called "spider"
2023-08-28 14:37:02 <CcxWrk> wget even has --spider flag for exactly thisd
2023-08-28 17:20:40 <wk> hey guys I've installed Alpine-linux on my 2005 hp pavilion dv1000 and it just rocks 
2023-08-28 17:20:55 <dalias> :)
2023-08-28 17:21:05 <wk> Alpine linux bring life back for my old i686
2023-08-28 17:21:10 <wk> amazing 
2023-08-28 21:11:15 <thresh> howdy.  I'm having a weird isue with Alpine Linux's wget and 403 on github.  Debian's one works fine.  I'm not too sure this is something to do with Alpine and not Github bot detection or whatever, though.
2023-08-28 21:12:34 <thresh> Here's a gist with details: https://gist.github.com/thresheek/b3098a9304c868c80e6926c1bca68630
2023-08-28 21:14:46 <thresh> busybox wget works, too
2023-08-28 21:16:49 <dwfreed> thresh: run wget with '-d' on both, compare the outputs?
2023-08-28 21:18:20 <thresh> I've also tried doing curl with the same set of headers sent, and it was also 200.  I'll compare -d, thanks.
2023-08-28 21:21:34 <quinq>   HTTP/1.1 403 Forbidden
2023-08-28 21:21:38 <quinq> Just got the same
2023-08-28 21:22:31 <pjs> I've got an Epson L3250 printer and in ALpine it only prints blank pages (even the test page from cups).. From Ubuntu, Mac, Windows it prints fine. I've tried a bunch of solutions I've found online and nothing works so far. Anyone ever run across this?
2023-08-28 21:22:33 <quinq> Something to do with the User-agent?
2023-08-28 21:24:07 <thresh> I've tried sending the exact same headers as Alpine wget does, but with curl, and it worked fine
2023-08-28 21:24:51 <quinq> Yeah, works with curl
2023-08-28 21:31:55 <thresh> No significant difference in -d: https://gist.github.com/thresheek/b3098a9304c868c80e6926c1bca68630?permalink_comment_id=4674382#gistcomment-4674382
2023-08-28 21:34:55 <quinq> It seems that by default wget sends an identity accept-encoding header
2023-08-28 21:35:03 <quinq> But even with a deflate/gzip, I get a 403
2023-08-28 21:36:21 <Apachez> some WAF in between triggering on user-agent or missing http headers compared to a regular browser request from the same ip?
2023-08-28 21:37:09 <Apachez> like cloudflare waf triggers if you in a post includes "/usr/local" or similar - workaround is to add a whitespace after the first frontslash like so "/ usr/local"
2023-08-28 21:37:55 <thresh> As I've said, I've tried sending the exact same request with the same headers via curl and it works.
2023-08-28 21:38:19 <thresh> But yeah, maybe it's timing or whatever that triggers their WAF
2023-08-28 21:38:29 <dwfreed> timing or differences in the TLS handshake
2023-08-28 21:38:37 <quinq> Rather that yeah
2023-08-28 21:38:51 <dwfreed> WAFs are fickle beasts
2023-08-28 21:41:18 <dwfreed> relevant: https://github.com/orgs/community/discussions/65227
2023-08-28 21:42:03 <thresh> Woah, your googe-fu is much, much better than mine!
2023-08-28 21:42:08 <thresh> also, google
2023-08-28 21:42:48 <thresh> Yeah, looks like TLS differences are triggering it
2023-08-28 21:42:49 <dwfreed> I didn't even google for it, somebody posted it in another IRC channel I'm in
2023-08-28 21:43:02 <thresh> helps to be social then :) thanks!
2023-08-28 22:02:40 <dalias> wow
2023-08-28 22:02:45 <dalias> such broken
2023-08-28 22:03:27 <zcrayfish> CcxWrk: uh. No.
2023-08-28 22:04:01 <zcrayfish> https://github.com/wikimedia/pywikibot
2023-08-28 22:04:07 <dalias> how do you misconfigure something so bad that you get successful TLS but then http error depending on TLS version/params...
2023-08-28 22:05:55 <fluix> knowing nothing about firewalls, it's probably a bunch of different heuristics
2023-08-28 22:13:02 <thresh> dwfreed, tls decryption happens on the edge, WAF happens somewhere else down the stack..
2023-08-28 22:13:43 <dwfreed> sure, but the TLS decryptor can put various details about said TLS decryption into the request forwarded on
2023-08-28 22:14:44 <thresh> arf, I meant dalias. sorry
2023-08-28 22:16:14 <dalias> yeah. i mean you have to be doing something really intentionally awful...
2023-08-28 22:16:40 <dalias> (which running a WAF to begin with is, but here i mean it would almost have to be intentional that they're making it give an http error for tls parameters they don't like)
2023-08-28 22:16:40 <thresh> that pretty much describes WAFs
2023-08-28 22:17:21 <thresh> and to their defence you don't really want botnet authors to (easily) figure out why exactly you're blocking them
2023-08-28 23:02:32 <Apachez> also a waf in front of github that blocks posts including common *nix paths :D
2023-08-28 23:03:59 <dalias> lmao
2023-08-28 23:52:35 <NepNepdmsalwaysopen[m]> Hm, I'm trying to see why I get a postgres error message at boot saying it failed to start, but it still does. Here's the message:... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/YGLslklxsTzOMlohsJoMhHhY>)
2023-08-29 02:15:56 <uuidNuniq> NepNepdmsalwaysopen[m]: did not start **in time**
2023-08-29 02:16:55 <NepNepdmsalwaysopen[m]> Ohh i see
2023-08-29 02:50:55 <asdon> Hello
2023-08-29 02:50:59 <dalias> hi
2023-08-29 02:51:34 <asdon> Chat seems really slow tonight for some reason
2023-08-29 03:54:17 <troubleshoe> can anyone help get alpine and epiphany to work on my raspberry pi? all it says is bwrap cant create symlink at var/run file already exists. no such file can be found when I search for it, nor had any help from chatgpt explained what to do other than install a worse browser. I wish ungoogled chromium worked but according to the forums its not supported in any way and this seems like the only alternative to firefox that w
2023-08-29 04:12:13 <zcrayfish> troubleshoe: did you install it from apk or flatpak?
2023-08-29 04:26:59 <troubleshoe> apk
2023-08-29 04:32:31 <zcrayfish> are dbus and xdg-desktop-portal running?
2023-08-29 09:31:47 <troubleshoe> >[21:32] <zcrayfish> are dbus and xdg-desktop-portal running?
2023-08-29 09:31:52 <troubleshoe> dbus is
2023-08-29 09:32:07 <troubleshoe> no idea about the other one atm chatgpt didnt ask me to check
2023-08-29 09:32:42 <troubleshoe> the program by the way shows up in internet as "browser" and will load into the task bar and then never show up, and can be force closed
2023-08-29 09:32:56 <troubleshoe> i dont know if thats helpful information
2023-08-29 09:33:22 <troubleshoe> if xdg-desktop-portal is needed for the desktop environment to even exist its probably running but I;'ll have to get home to check
2023-08-29 17:46:54 <cuc> hey, is python is python3 package?
2023-08-29 17:47:17 <ikke> there is no 'python' package
2023-08-29 17:47:19 <ikke> only python3
2023-08-29 17:47:52 <cuc> is in another distros it named just python?
2023-08-29 17:48:09 <ikke> I don't know
2023-08-29 17:50:01 <cuc> is there some way to colorize apk package manager?
2023-08-29 17:56:49 <ikke> No, apk does not support colorized output
2023-08-29 17:57:23 <cuc> is pep668 i need pipx package?
2023-08-29 18:00:30 <ikke> There is a package called pipx
2023-08-29 18:00:33 <ikke> https://pkgs.alpinelinux.org/contents?file=pipx&path=&name=&branch=edge
2023-08-29 18:00:56 <cuc> oh sorry its for pip i need py3-pip
2023-08-29 18:41:58 <cuc> is alpine edge kernel includes rust?
2023-08-29 18:46:25 <Ermine> No
2023-08-29 18:56:53 <ptrc> cuc: do you mean the `pythonispython3` package?
2023-08-29 18:57:15 <ptrc> ( one that sets 'python' to just be a link to python3 )
2023-08-29 18:57:46 <cuc> there no python package, it named python3
2023-08-29 18:57:54 <cuc> have no idea why
2023-08-29 18:58:17 <cuc> but if i do apk add clang it install clang16 for me
2023-08-29 18:58:36 <Habbie> that's because new versions of clang are backwards compatible with older ones
2023-08-29 18:58:42 <ptrc> ah
2023-08-29 18:58:57 <ptrc> python doesn't really get major updates that often :p
2023-08-29 18:59:09 <cuc> as i know in arch python3 package named python
2023-08-29 18:59:20 <cuc> cuz they pissed on py2
2023-08-29 19:19:14 <cuc> is there imagemagick on alpine?
2023-08-29 19:19:20 <cuc> on repos i mean?
2023-08-29 19:19:42 <ikke> yes
2023-08-29 19:19:52 <ikke> use https://pkgs.alpinelinux.org/packages to find packages
2023-08-29 19:20:19 <cuc> i use apk search
2023-08-29 19:20:31 <cuc> and cant see imagemagick
2023-08-29 19:24:17 <cuc> https://pasteboard.co/QGbJA2JmQ3RV.png
2023-08-29 19:28:37 <cuc> https://ibb.co/HDmNjjJ
2023-08-29 19:29:04 <cuc> it so strange
2023-08-29 19:29:44 <cuc> no such package
2023-08-29 19:30:04 <cuc> i on edge repos and comunity have
2023-08-29 19:30:16 <cuc> why?
2023-08-29 19:41:40 <cuc> no package
2023-08-29 19:47:33 <u0_a340> have no idea why no imagemagick
2023-08-29 19:48:07 <ikke> ijhttps://pkgs.alpinelinux.org/packages?name=imagemagick&branch=edge&repo=&arch=&maintainer=
2023-08-29 19:49:01 <kwadroke[m]> also in community
2023-08-29 19:49:06 <cuc> oooh
2023-08-29 19:49:11 <cuc> im stupid
2023-08-29 19:49:19 <cuc> whe i added testing
2023-08-29 19:49:27 <cuc> i delete community
2023-08-29 19:49:30 <cuc> :(
2023-08-29 19:49:38 <cuc> just replaced
2023-08-29 19:55:35 <cuc> is bashrc work in ash?
2023-08-29 19:56:07 <kwadroke[m]> use .profile
2023-08-29 19:56:29 <quinq> Well, use .profile for portable sh
2023-08-29 19:56:45 <quinq> Because no, bashrc will not work in ash
2023-08-29 19:56:50 <cuc> but it not like bashrc
2023-08-29 19:57:00 <cuc> ptofile it on login
2023-08-29 19:57:30 <quinq> (couldn't parse)
2023-08-29 19:57:53 <ikke> cuc: you need to define ENV=$HOME/.ashrc in .profile to have something similar
2023-08-29 19:58:19 <cuc> mb better use bash?
2023-08-29 19:58:27 <cuc> or it bad idea?
2023-08-29 19:58:35 <quinq> What's your objective?
2023-08-29 19:58:35 <ikke> If you want to use it, use it
2023-08-29 19:58:38 <kwadroke[m]> depend on what you're needing to put in there
2023-08-29 19:58:56 <cuc> is ash support colored ps1?
2023-08-29 19:59:01 <ikke> yes
2023-08-29 19:59:04 <quinq> If your terminal does
2023-08-29 19:59:25 <ikke> mv /etc/profile.d/color_prompt.sh.disabled /etc/profile.d/color_prompt.sh
2023-08-29 19:59:46 <zcrayfish> PS1="\e[1;31ma\e[1;33ml\e[1;32mp\e[1;34mi\e[1;36mn\e[1;35me\e[0m\033[30;37;5m😊\e[0m\w " 
2023-08-29 19:59:47 <zcrayfish> ;)
2023-08-29 20:00:35 <cuc> no tput there?
2023-08-29 20:00:58 <quinq> Not sure you want to exec tput on each prompt
2023-08-29 20:01:39 <cuc> but what way to change colors like from 16 pallete?
2023-08-29 20:02:02 <ikke> if you know what terminal you use, and don't switch every other day, it does not provide any benefit
2023-08-29 20:02:27 <cuc> im about pallete displaying while neofetch
2023-08-29 20:02:34 <ikke> You can use it to find the correct escape codes for your terminal, and you could then hardcode them
2023-08-29 20:03:09 <quinq> Are you driving a terminal or a carnival?
2023-08-29 20:03:20 <cuc> carnival
2023-08-29 20:04:04 <quinq> Then yeah, put a few tput calls in subshells in your PS1, and expose it in the Hall of Horrors there ;)
2023-08-29 20:04:11 <cuc> like i just like auto change ps1 colors with changing wallpapers
2023-08-29 20:05:10 <quinq> You can also have a few pre-set PS1s and change them from time to time
2023-08-29 20:06:38 <cuc> as i understand im talking about terminal pallete
2023-08-29 20:06:51 <cuc> cuz neofetch displays that
2023-08-29 20:07:16 <cuc> i need to ps1 promt changing when i change pallete
2023-08-29 20:07:30 <cuc> can i do this without tput?
2023-08-29 20:07:53 <ikke> yes
2023-08-29 20:08:36 <cuc> how?
2023-08-29 20:09:02 <ikke> The palette are the 16 default colors which have fixed escape codes
2023-08-29 20:09:15 <ikke> when you change the pallete, you change the colors associated with each of them
2023-08-29 20:09:24 <ikke> the escape code does not change
2023-08-29 20:09:42 <cuc> how write it?
2023-08-29 20:10:03 <ikke> like how zcrayfish showed
2023-08-29 20:10:48 <zcrayfish> I suggest reading up on ANSI X3.64 (ISO/IEC 6429).
2023-08-29 20:11:02 <zcrayfish> Most terminals since the 80's should either support it, or ignore them and display cleanly. Either way.
2023-08-29 20:11:43 <cuc> how?ok
2023-08-29 20:12:59 <quinq> zcrayfish, do you really have an emoji in your prompt though?
2023-08-29 20:13:36 <ikke> cuc: echo -e '\033[31mred'
2023-08-29 20:14:52 <zcrayfish> lol quinq, I use PS1="\u@\h:\w>" for normal users and PS1="\e[0;31m\u@\h:\w\$\e[m" (or something similar) for daily use.
2023-08-29 20:15:44 <quinq> :>
2023-08-29 20:16:15 <cuc> ikke it printed red but it need reset
2023-08-29 20:16:16 <zcrayfish> the latter red one for root*
2023-08-29 20:16:34 <ikke> cuc: echo -e '\e[31mred\e[0m'
2023-08-29 20:17:54 <cuc> is there some tools online that generate ps1 with this colors not tput
2023-08-29 20:18:10 <pjs> I've got an Epson L3250 printer and in ALpine it only prints blank pages (even the test page from cups).. From Ubuntu, Mac, Windows it prints fine. I've tried a bunch of solutions I've found online and nothing works so far. Anyone ever run across this?
2023-08-29 20:18:26 <ikke> cuc: first google hit: https://ansi.gabebanks.net/
2023-08-29 20:23:20 <cuc> \x1b work in ash but not work in bash?
2023-08-29 20:24:48 <quinq> That looks more like a printf escape
2023-08-29 20:25:04 <quinq> You don't need those fancy escape, just input an escape character directly where it's needed
2023-08-29 20:25:04 <zcrayfish> pjs: No. If that printer supports more than one printer language, I would suggest trying a different one. 
2023-08-29 20:25:57 <zcrayfish> Many Epson printer support both Adobe PS and HP PDL/PCL. FWIW.
2023-08-29 20:58:18 <minimal> the upstream 6.6 kernel is removing the ability to (easily) disable the CPU microcode driver
2023-08-29 20:58:28 <minimal> that doesn't make sense for kernels for VMs
2023-08-29 20:59:09 <minimal> https://lore.kernel.org/lkml/20230812195727.660453052@linutronix.de/
2023-08-29 21:05:54 <ikke> Is https://lore.kernel.org/lkml/169194759403.27769.2355835959167552007.tip-bot2@tip-bot2/ the signal that it has been accepted?
2023-08-29 21:06:38 <zcrayfish> Late loading is being worked on lately too.
2023-08-29 22:18:29 <minimal> ikke: I believe so. I emailed the patch author last week pointing out that it wasn't really a "reasonable configuration" for VMs and today received a reply which was simply a link to https://people.kernel.org/tglx/notes-about-netiquette
2023-08-29 22:19:08 <minimal> so in others he didn't make any comment about my question, I guess he expects me to reply to the kernel mailing list with my question
2023-08-29 23:15:59 <minimal> ikke: with that change I'm assuming microcode can't be enabled/disabled via "make menuconfig" etc (or will appear there but with just no "help" text?) but that CONFIG_MICROCODE will still appear in a .config file and so the Alpine config files for linux-lts can still override it
2023-08-29 23:49:30 <pjs> zcrayfish: as in the driver selection? Not sure where this printer lang would be set
2023-08-30 04:13:02 <invoked> i forget who asked about this, but i helped someone get proton-bridge working on alpine. it builds and runs fine -- bridge --cli works anyway. the only fussy bit is getting it to use 'pass'
2023-08-30 10:05:09 <telmich> For a few days the light utility stopped working on all my laptops and I see light was removed in 71d976f083d3db5db77eee4663dadb54606b5ffd, probably because of the warning on https://github.com/haikarainen/light - do we have a replacement for it or what does one use at the moment for setting the display brigthness?
2023-08-30 10:13:09 <amk> brightnessctl has been working okay for me
2023-08-30 10:14:13 <amk> i had to add a udev rule though, the light package used to provide it: https://paste.sr.ht/~amk/91e9007a50157262f9e752198a6ad71417e78b49
2023-08-30 13:19:16 <zcrayfish> lxqt-backlight is working fine for me: on Xorg and Wayland. 
2023-08-30 18:59:22 <u0_a340> what i need add to fix fcntk64: symbol not found
2023-08-30 18:59:35 <u0_a340> i have gcompat
2023-08-30 18:59:48 <u0_a340> running glibc compilated program
2023-08-30 19:33:00 <ptrc> you might need to write a shim yourself
2023-08-30 19:33:11 <ptrc> compile it as a shared library, then add to DT_NEEDED
2023-08-30 19:33:36 <ptrc> ( also, i assume you mean fcntl64? )
2023-08-30 20:10:33 <u0_a340> i mean dlfcntl
2023-08-30 20:10:46 <u0_a340> i mean fcntl64
2023-08-31 16:02:21 <codebam> I removed wpa_supplicant and installed iwd and set it up. it connects on bootup but udhcpd doesn't run afterwards. how do I fix that?
2023-08-31 16:02:27 <codebam> I can restart networking and it works
2023-08-31 16:03:51 <codebam> also I installed amd-ucode and regenerated my grub config but it didn't load the microcode
2023-08-31 16:06:37 <minimal> codebam: you checked that /boot/grub/grub.cfg mentions the ucode file?
2023-08-31 16:09:23 <zcrayfish> codebam: check `dmesg | grep microcode'
2023-08-31 16:18:36 <codebam> it mentions it in the grub.cfg but `dmesg | grep microcode` doesn't give me a revision https://pastebin.seanbehan.ca/Rtb1ohK5GR.txt
2023-08-31 16:53:43 <invoked> codebam: double check that networking is enabled at boot
2023-08-31 16:54:00 <invoked> you might be restarting it but it was never started in the first place.
2023-08-31 16:54:26 <codebam> how do I check that?
2023-08-31 16:54:57 <codebam> I added it to runlevel default is that right?
2023-08-31 16:55:12 <invoked> rc-update -l
2023-08-31 16:55:28 <invoked> no, it should be at boot
2023-08-31 16:55:47 <codebam> it's in boot already
2023-08-31 16:55:54 <codebam> says its already installed in runlevel boot
2023-08-31 16:56:17 <codebam> should iwd be in default or boot?
2023-08-31 16:56:32 <invoked> boot
2023-08-31 16:56:40 <codebam> okay maybe that was the issue
2023-08-31 16:56:54 <codebam> oh never mind it was in boot
2023-08-31 16:57:18 <invoked> hang on, i'll hop on to my alpine laptop.
2023-08-31 17:00:55 <invoked> i've got networking in boot and default. lol. other than that nothing stands out
2023-08-31 17:01:31 <codebam> should I put it in default too then?
2023-08-31 17:01:56 <invoked> you could try it.
2023-08-31 17:02:04 <invoked> (no warranty)
2023-08-31 17:02:10 <codebam> okay
2023-08-31 17:02:12 <codebam> lol
2023-08-31 17:02:31 <invoked> you don't have anything weird in interfaces, do you?  beyond the usual 'iface wlan0 inet dhcp'?
2023-08-31 17:05:15 <codebam> nope just that and the one for lo
2023-08-31 17:08:09 <invoked> iwd's run script has 'before net' so it should always run before networking does. i'm just thinking that iwd hasn't finished setting up by the time networking runs.
2023-08-31 17:08:18 <invoked> in your case, that is
2023-08-31 17:08:44 <invoked> so maybe you could manually insert a delay somewhere and see if that's the case.
2023-08-31 17:09:03 <invoked> these are all just scripts so you can put that where it makes sense to you
2023-08-31 17:42:46 <oleovit> hi. short question. in order to prevent rc-log messages like "sysctl: error: 'net.ipv6/conf/all/accept_redirects' is an unknown key" , do i have to remove ipv6 module in /etc/modules (ipv6 disabled via sysctl only)...
2023-08-31 17:44:11 <ikke> No, at most that would make it worse
2023-08-31 17:44:25 <p1fk6ohrq9> is it possible that apk del has some leftover files? if yes, how to find them, something like lostfiles on arch?
2023-08-31 17:45:15 <ikke> oleovit: that sysctl key is incorrect, it mixes . and /
2023-08-31 17:45:20 <ikke> it should only use .
2023-08-31 17:46:00 <oleovit> @ikke thx, so nothing i can do i guess. its only cosmetic..
2023-08-31 17:47:46 <oleovit> .. and yes, this was a thought from me yesterday, the / instead . 
2023-08-31 17:48:46 <ikke> oleovit: what service is causing that
2023-08-31 17:49:12 <oleovit> sysctl
2023-08-31 17:49:22 <ikke> oleovit: then check /etc/sysctl.d/*
2023-08-31 17:50:38 <oleovit> that's where i have my stuff (etc/sysctl.d/local.conf)
2023-08-31 17:51:07 <ikke> fix the incorrect key then :)
2023-08-31 17:54:41 <oleovit> 'net.ipv.conf.all.accept_redirects and other stuff that appears in the log, is not defined by me in local.conf. do you mean to add it and  explicitly set it to zero?
2023-08-31 17:55:09 <oleovit> i'm dumb again, like always when i chat
2023-08-31 17:56:58 <oleovit> i meant net.ipv6.conf.all.accept_redirects ... 
2023-08-31 17:57:40 <minimal> oleovit: the standard /etc/sysctl.conf (from package alpine-baselayout) has entries for net.ipv4.conf.all.accept_redirects and net.ipv6.conf.all.accept_redirects
2023-08-31 17:57:57 <minimal> which are correct
2023-08-31 17:59:14 <minimal> now obviously if you have ipv6 disabled then the file that net.ipv6.conf.all.accept_redirects "maps onto" won't exist
2023-08-31 18:00:36 <oleovit> do you mean the /lib/sysctl.d/00-alpine.conf ?
2023-08-31 18:02:25 <minimal> there is that file also
2023-08-31 18:02:46 <minimal> the point is the "error" is due to ipv6 settinngs trying to be set/changes when ipv6 is disabled
2023-08-31 18:10:49 <oleovit> /lib/sysctl.d/00-alpine.conf has all the kernel parameter settings where i get the errors, this file has precedence over my etc/sysctl.d/local.conf (/etc/sysctl.conf 
2023-08-31 18:11:14 <oleovit> .. is empty.. so i get the errors
2023-08-31 18:12:43 <oleovit> what about the / instead . like ikke said? 
2023-08-31 18:13:38 <minimal> do any of those files use "/" instead of "."?
2023-08-31 18:14:31 <oleovit> nope minimal, everything is correct 
2023-08-31 18:15:03 <minimal> well I've already told you you'll see ipv6 related messages as you don't have ipv6 enabled
2023-08-31 18:16:04 <minimal> so either remove any ipv6 settings in any of the sysctl-related config files or else just "put up" with seeing those messages during boot
2023-08-31 18:18:48 <minimal> or modify /etc/init.d/sysctl so it called the sysctl program with the "-e" option to "Don't warn about unknown keys"
2023-08-31 18:20:15 <oleovit> yes, that's clear to me, but doesnt make much sense, because /lib/sysctl.d/00-alpine.conf rct. will be owerwritten. only /etc/sysctl.d/local.conf and /etc/sysctl.conf - not. ah, your last message is a good hint, thx minimal. 
2023-08-31 18:21:42 <minimal> not sure if a MR should be raised for /etc/init.d/sysctl to use "-e" by default - in some situations you may *want* to be told about unknown settings (e.g. typos)
2023-08-31 18:23:45 <oleovit> i think, it's ok with the warning. these err messagesr do not take that much time while boot. chrony takes time lol
2023-08-31 18:24:54 <oleovit> will try the async method from .. "pt..something null"
2023-08-31 18:26:16 <oleovit> so, no worries, no mr from me, but i learned how to do it, thx to ikke and dnscrypt-proxy and other helpers on your gitlab
2023-08-31 18:27:36 <oleovit> .. but maybe, like always.. the wiki,,
2023-08-31 18:28:29 <minimal> oleovit: chrony (depending on config) *purposely* delays boot in order to ensure the clock is correct before continuing
2023-08-31 18:29:03 <minimal> if you stop chrony waiting then it might have some unexpected side effects (if time changes after other processes have started)
2023-08-31 18:29:21 <oleovit> absolutly. i use nts, makestep and a script with chronyc waitsync 
2023-08-31 18:30:02 <oleovit> no problem with waiting. but i can play  with it in a vm.
2023-08-31 18:30:54 <minimal> depending on which hypervisor you are running (and any agent it provides) you may NOT want to run a NTP server
2023-08-31 18:32:17 <oleovit> just to take a look at https://ptrcnull.me/posts/openrc-async-services/ , just playing. on my bare-metal alpine, i will never do that
2023-08-31 18:33:32 <minimal> e.g. Virtualbox's agent if installed will sync time so you likely DON'T want to run a NTP server as well
2023-08-31 18:35:08 <oleovit> it's just playing with a stacked runlevel, that's why i mentioned it
2023-08-31 18:35:57 <minimal> but that seems to be running other services while chrony doesn't yet have time sync...
2023-08-31 18:36:17 <minimal> which defeats the point of chrony waiting for timesync
2023-08-31 18:38:35 <minimal> if you don't want chrony to wait then (re)config chrony not to wait.....but be aware there may be side-effects of doing that
2023-08-31 18:40:06 <oleovit> no, i want chrony to wait, that's why i even wrote a script for my nts/makestep/waitsync config in my postinstall-thing# 
2023-08-31 18:40:39 <oleovit> i dont care about boot-up times
2023-08-31 18:41:03 <minimal> oleovit: but you've changes OTHER SERVICES to not wait on chrony, right? that's my point
2023-08-31 18:42:22 <minimal> is'nt the point of that article to get a login prompt quicker by not having (some) services wait for chrony to finish its initialisation?
2023-08-31 18:44:21 <oleovit> ah, argue in the context of this article. i don't care that much about this.
2023-08-31 18:44:39 <minimal> oleovit: but you've changes OTHER SERVICES to not wait on chrony, right? that's my point
2023-08-31 18:45:29 <oleovit> your english.. what do you mean with you've changes other services"?
2023-08-31 18:45:35 <minimal> s/changes/changed/
2023-08-31 18:45:46 <minimal> what is the purpose of the changes you have made?
2023-08-31 18:45:48 <oleovit> ah, ok
2023-08-31 18:46:36 <p1fk6ohrq9> how to query time from ntpd?
2023-08-31 18:46:55 <minimal> oleovit: ?
2023-08-31 18:47:29 <oleovit> waht so you mean, the chronyc waitsync command?
2023-08-31 18:47:53 <somercet1> about the microcode, I don't see it in dmesg, but it is installed and grub.cfg mentions it. What if my processor has no patches?
2023-08-31 18:48:06 <minimal> oleovit: I mean you said you have made some system config changes (i.e. stacked runlevel) and I am asking what the reason(s) is for making those changes?
2023-08-31 18:48:16 <oleovit> noo
2023-08-31 18:48:32 <minimal> are you trying to have some services run *before* chrony has finished its initial time sync?
2023-08-31 18:48:36 <oleovit> i dont have stacked runlevels, never said that
2023-08-31 18:49:29 <minimal> ok, I misunderstood that point. Anyway, whatever changes related to chrony that you have made, why did you make them?
2023-08-31 18:49:42 <minimal> you seem to be wanting to avoid some services waiting for chrony
2023-08-31 18:50:22 <oleovit> this for example: https://github.com/llightcb/sight/blob/platonic/chrony.conf
2023-08-31 18:50:47 <oleovit> nooo, i dont want to avoid anything
2023-08-31 18:51:29 <oleovit> i said: only in a vm, to play with stacked runlevels, to learn, but not on my bare metal alpine
2023-08-31 18:51:33 <oleovit> #
2023-08-31 18:52:01 <minimal> oleovit: after you earlier said that chrony takes time you then said:
2023-08-31 18:52:03 <minimal> <oleovit> will try the async method from .. "pt..something null"
2023-08-31 18:52:26 <oleovit> yes, in a vm!
2023-08-31 18:52:37 <minimal> so that give the appearance that you were not happy with chrony delaying the boot time and wanted to prevent it slowing boot
2023-08-31 18:52:46 <oleovit> for playing with these runlevel stuff
2023-08-31 18:53:00 <minimal> and that is why I then pointed out that chrony's config delays boot for a reason
2023-08-31 18:53:00 <oleovit> no, you are wrong
2023-08-31 18:53:10 <oleovit> i know
2023-08-31 18:53:40 <oleovit> thats why i have waitsync in combi with makestep
2023-08-31 18:54:13 <minimal> so then why did you said you would try the async method? there must be some reason you considered trying it
2023-08-31 18:54:25 <oleovit> and never change that _on_my:machine_
2023-08-31 18:54:41 <oleovit> yes, its interesting
2023-08-31 18:54:55 <oleovit> i want to try out everything
2023-08-31 18:55:03 <oleovit> _in a vm_
2023-08-31 18:55:08 <minimal> interesting if it risked causing problems due to services starting before time is synced?
2023-08-31 18:55:51 <minimal> do you also try out "rm -Rf /" in VMs as well? ;-)
2023-08-31 18:55:52 <oleovit> jesus, minimal, see my git, i dont care about wait(sync) longer. 
2023-08-31 18:56:25 <oleovit> but i want to play with this stuff in this article, even if useless
2023-08-31 18:58:48 <oleovit> in-a-vm. no risk for me. never with my bare-metal kartoffel. never in my postinstall.
2023-08-31 18:59:56 <oleovit> ok, the sysctl stuff is clear to me now, even while i chat. so thx to you all, minimal, ikke, whole rest, love you all
2023-08-31 19:01:59 <oleovit> well, i overlooked this comment:  "rm -Rf /" and so on. minimal..
2023-08-31 20:06:05 <oleovit> sry, me again, sysctl error, unknown key in rc-log.. it has to do with for example net.ipv6/conf/all/accept_redirects - the / instead of . if i try sysctl net.ipv6.conf.all.accept_redirects=1 for example, no err, even if ipv6 disabled via sysctl condig. why do we have these / while boot?
2023-08-31 20:18:47 <oleovit> nevermind, maybe i simply don't get it. i cant express myself today. good night, it doesnt matter anyhow
2023-08-31 20:58:41 <u0_a340> explain what point of installing dwl via package manager? how configure it after?
2023-08-31 21:02:15 <ikke> If you are fine with the default configuration
2023-08-31 21:04:04 <u0_a340> ok understandable