2021-06-01 02:38:49 <LACampbell> ikke: how do you acutally *use* gcompat? I'm updating the wiki
2021-06-01 02:47:38 <LACampbell> https://wiki.alpinelinux.org/wiki/Running_glibc_programs#gcompat hopefully people will at least know it exists now, but full disclaimer I've never used it
2021-06-01 04:22:02 <chexo4> test
2021-06-01 04:24:31 <chexo4> Hello all! I got an error when running `Xorg -configure` as root, "Number of created screens does not match number of detected devices"
2021-06-01 04:24:38 <iggy> chexo4: re: your question earlier about ryzen 2200g, the biggest thing I'd watch out for is the limited pcie lanes and their usually weird routing
2021-06-01 04:25:02 <chexo4> iggy: makes sense given what limited stuff I've seen with lspci
2021-06-01 04:26:41 <chexo4> On the topic of `Xorg -configure`, I am running an Nvidia GTX 1050 Ti (I know, Nvidia drivers) with xf86-video-nv drivers (I'm assuming here that the nv nvidia drivers are the official/proprietary ones, but I could easily be wrong)
2021-06-01 04:26:51 <chexo4> I have one monitor plugged in
2021-06-01 04:27:43 <chexo4>  I can provide a sample of my Xorg.conf.new but it will take me a minute to upload it to a pastebin service as I am currently on plain TTY with no graphics
2021-06-01 04:28:34 <chexo4> wait why am I doing this, it should autodetect fine, I have a not-weird setup
2021-06-01 04:28:43 <chexo4> going to try if it'll work on it's own
2021-06-01 04:29:31 <ikke> LACampbell: steop 1: you install it, step 2: you run your program
2021-06-01 05:15:11 <LACampbell> ikke: on a headless system at the minute - does it work with GUI apps?
2021-06-01 05:15:52 <ikke> LACampbell: What it does is that it provides the dynamic loaded that glibc programs expect
2021-06-01 05:16:04 <ikke> in that sense, yes, it does work with graphical programs
2021-06-01 05:16:16 <ikke> because that does not matter for what gcompat does
2021-06-01 05:20:21 <LACampbell> cool. I really need to get round to dual booting alpine on this lappy
2021-06-01 05:20:54 <LACampbell> anyway i edited the wiki because I keep hearing about gcompat here, and in HN I think, but people still keep reading the wiki and thinking they have to do a chroot or flatpark or whatever
2021-06-01 05:26:51 <ikke> in some cases, that's still necessary
2021-06-01 05:33:37 <LACampbell> ikke: tell me what they are and I'll add it in
2021-06-01 05:35:14 <ikke> other dependencies, for example
2021-06-01 06:38:46 <whatsyourask> Could someone tell me that I can try? Edge alpine, no wifi on RaspberryPi400, ip link return no wlan0, dmesg | grep brcm return: 
2021-06-01 06:38:52 <whatsyourask> [   11.201600] brcmfmac: brcmf_fw_alloc_request: using brcm/brcmfmac43456-sdio for chip BCM4345/9
2021-06-01 06:38:52 <whatsyourask> [   11.208313] brcmfmac mmc1:0001:1: Direct firmware load for brcm/brcmfmac43456-sdio.bin failed with error -2
2021-06-01 06:38:52 <whatsyourask> [   11.202027] usbcore: registered new interface driver brcmfmac
2021-06-01 06:42:46 <markand> whatsyourask, I think the firmware isn't distributed right now in the upstream firmware repository
2021-06-01 06:43:12 <markand> https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/tree/brcm
2021-06-01 06:43:31 <markand> and alpine follow upstream strictlier
2021-06-01 06:44:25 <whatsyourask> That helps me a lot. 
2021-06-01 06:44:45 <clandmeter> i think we do
2021-06-01 06:44:52 <markand> so not sure where those files are provided though (I can't find on the RPI-Distro github)
2021-06-01 06:45:26 <markand> btw, the pi zero is unable to run armv7 isn't it? you should use armhf?
2021-06-01 06:45:36 <markand> asking because I've just received one after a while
2021-06-01 06:45:42 <clandmeter> yes 
2021-06-01 06:45:53 <clandmeter> its the only armhf device that is kind of active
2021-06-01 06:46:03 <markand> okay, thanks
2021-06-01 06:46:14 <clandmeter> check our release scripts
2021-06-01 06:46:19 <clandmeter> it includes a rpi firmware part
2021-06-01 06:53:55 <markand> clandmeter, and linux-rpi rather than linux-rpi2?
2021-06-01 06:57:32 <clandmeter> markand: i would say yes, but im not 100% sure. its been too long since i set that up...
2021-06-01 06:58:06 <markand> :p
2021-06-01 06:58:17 <markand> according to the rpi armhf it looks like it's linux-rpi 
2021-06-01 06:58:30 <markand> rpi2 seems to be for >= Pi 2 and <= pi3+
2021-06-01 06:58:47 <clandmeter> too many rpi's..
2021-06-01 06:58:51 <markand> definitely
2021-06-01 07:22:14 <markand> fatest boot time ever on the pi 0
2021-06-01 07:22:16 <markand> no I'm kidding
2021-06-01 16:30:41 <Fuentes> hello everyone. I hava a question about Alpine Linux 3.15: Where can i find the virtualbox guest additions kernel modules? The package "virtualbox-guest-modules-virt" only exists on version 3.12
2021-06-01 16:31:16 <Fuentes> *Alpine 3.13 i mean
2021-06-01 16:35:22 <lopid> what 3.15?
2021-06-01 16:35:41 <ikke> They corrected it
2021-06-01 16:35:54 <lopid> oh i thought they corrected the other one
2021-06-01 16:36:08 <lopid> i see then in my 3.13.5
2021-06-01 16:36:10 <lopid> them
2021-06-01 16:36:48 <lopid> virtualbox-guest-additions-* though…
2021-06-01 17:00:16 <Fuentes> If you visit the following link, the package "virtualbox-guest-modules-virt" is not listed: https://pkgs.alpinelinux.org/packages?name=virtualbox-guest-additions-*&branch=v3.13
2021-06-01 17:01:32 <Fuentes> If you review the content of the listed packages, kernel modules are not included
2021-06-01 17:03:23 <y34h> hi
2021-06-01 17:03:46 <Fuentes> hi
2021-06-01 17:03:53 <y34h> i'm getting some error messages while trying to install nfs-utils
2021-06-01 17:04:06 <y34h> they all come down to a "package mentioned in index not found" message
2021-06-01 17:04:15 <y34h> i'm new to the system
2021-06-01 17:04:30 <y34h> is there something I can do or this on the repo's side?
2021-06-01 17:04:58 <lopid> try adding -U to the install command, to sync repo
2021-06-01 17:05:07 <y34h> ok, i'll try
2021-06-01 17:05:09 <y34h> thank you
2021-06-01 17:05:17 <lopid> apk -U add nfs-utils
2021-06-01 17:05:35 <lopid> or sync separately with "apk update"
2021-06-01 17:06:17 <y34h> did that
2021-06-01 17:06:20 <y34h> same thing
2021-06-01 17:06:27 <y34h> three of the packages were not found
2021-06-01 17:06:48 <lopid> what alpine version?
2021-06-01 17:07:24 <y34h> moment
2021-06-01 17:07:46 <y34h> 3.3.15
2021-06-01 17:07:58 <y34h> i'm running it on a raspberry pi 4
2021-06-01 17:08:09 <lopid> was that a typo?
2021-06-01 17:08:56 <Fuentes> any suggestion (some link, google search, etc) to solve my problem, @lopid?
2021-06-01 17:09:10 <y34h> yes
2021-06-01 17:09:16 <y34h> thank for noticing
2021-06-01 17:09:29 <y34h> it's 3.13.5
2021-06-01 17:09:42 <lopid> unfortuantely not, Fuentes. i wish "apk info -L <package>" worked for not installed packages
2021-06-01 17:10:27 <lopid> is that the version used in your /etc/apk/repositories file?
2021-06-01 17:10:39 <y34h> let me check
2021-06-01 17:10:46 <lopid> ok you may
2021-06-01 17:11:46 <y34h> yes: http://linorg.usp.br/AlpineLinux/v3.13/main
2021-06-01 17:12:07 <y34h> the ".5" is not relevant, I suppose
2021-06-01 17:12:20 <lopid> right
2021-06-01 17:13:03 <lopid> fwiw i was just able to install nfs-utils and its deps (x86_64)
2021-06-01 17:13:20 <y34h> probably a raspberry problem
2021-06-01 17:13:26 <lopid> it wouldn't need some from the community repo, would it?
2021-06-01 17:13:34 <y34h> I wasn't able to install neither to rpi1 nor to rpi4
2021-06-01 17:13:49 <y34h> but the erros regarded different packages
2021-06-01 17:13:57 <y34h> something to do with kerberos on rpi1
2021-06-01 17:14:05 <y34h> and 3 other packages on rpi4
2021-06-01 17:14:23 <y34h> I suppose it would not
2021-06-01 17:14:26 <y34h> but let me try
2021-06-01 17:15:18 <y34h> no. same thing.
2021-06-01 17:17:45 <y34h> lopid, just one more thing: do you know what might cause a "permission denied error" on the install of a package, besides root privileges etc?
2021-06-01 17:18:02 <y34h> i'm getting "(4/4) Installing transmission-daemon (3.00-r2)
2021-06-01 17:18:18 <y34h> "ERROR: transmission-daemon-3.00-r2: Permission denied"
2021-06-01 17:19:00 <y34h> just wanted to use the pi as a torrentbox/nfs-server
2021-06-01 17:19:33 <y34h> but I cannot install the nfs or torrent fundamental packages
2021-06-01 17:19:34 <y34h> =/
2021-06-01 17:20:28 <lopid> readonly filesystem, maybe?
2021-06-01 17:20:41 <lopid> though that would be more like "cannot write to…"
2021-06-01 17:21:39 <y34h> thank you
2021-06-01 17:21:41 <y34h> i'll check on that
2021-06-01 17:24:54 <lopid> maybe try "apk -vv add transmission-daemon"
2021-06-01 17:27:06 <y34h> lopid, thank you for all the help
2021-06-01 17:27:13 <y34h> the problem was on the server's side
2021-06-01 17:27:38 <y34h> linorg.usp.br
2021-06-01 17:27:47 <y34h> I changed the repo and all worked well
2021-06-01 17:28:49 <lopid> ah
2021-06-01 17:29:14 <iggy> Fuentes: aren't the kernel modules all upstream now?
2021-06-01 17:36:55 <oliv3r> Hey all, I am doing a scripted bootstrap, and after apk add, modifying some files (/etc/mkinitfs/mkinitfs.conf) I want to force the trigger to run (e.g. re-create the initramfs). Is there a way to force an apk trigger to run, other then creating a dummy package to do touch /usr/share/kernel (in this particular case?)
2021-06-01 17:40:32 <lopid> does "apk fix mkinitfs" help?
2021-06-01 17:54:00 <iggy> Fuentes: according to https://gitlab.alpinelinux.org/alpine/aports/-/commit/fe84974d3e260868d5475b26ba02436b97d42e5f the modules were deleted in favor of the modules in the mainline kernel
2021-06-01 18:02:16 <Fuentes> @iggy daaaamn! I miss that!!
2021-06-01 18:02:25 <Fuentes> many thanks!!
2021-06-01 19:54:18 <thebtm> good afternoon Alpine Linux Channel
2021-06-01 19:54:31 <lopid> busybox's crond doesn't support mailing its job output, does it?
2021-06-01 19:56:49 <lopid> oh, there it is, in MAILTO
2021-06-01 19:56:54 <lopid> must be explicitly set
2021-06-01 20:01:43 <PiRATA> hey thebtm
2021-06-01 20:03:51 <lopid> interestingly, adding MAILTO /etc/crontab/root seems to have no affect
2021-06-01 20:09:28 <lopid> it seems that the only way to change root's crontab is by editing /etc/crontab/root
2021-06-01 21:20:05 <thebtm> is there a way to force lbu add to put in custom init.d scripts since it ignores the directory?
2021-06-01 21:38:47 <thebtm> Nevermind, figured it out, I have explicitly list the full path. can't use relative pathing.
2021-06-01 22:15:11 <wrench> sync
2021-06-01 22:28:22 <Hello71> successfully desynced
2021-06-01 22:28:38 <pinskia> ./gcc_build build test install
2021-06-01 22:28:44 <pinskia> wrong window
2021-06-01 22:51:41 <kernelc> Hi, how to auto install *-doc packages?
2021-06-01 22:52:33 <jn> apk add docs
2021-06-01 22:52:45 <agris> Why can't I apk add this package https://git.alpinelinux.org/aports/tree/community/opensmtpd-filter-rspamd?id=ddc59cbdd69eaefdf836fc2fa5d007514c1e9ca7 in 3.13?
2021-06-01 22:52:48 <jn> then you'll get the -doc packages for everything you install
2021-06-01 22:52:58 <agris> has the autobuild system not ran yet?
2021-06-01 22:53:13 <agris> it moved from testing to stable right?
2021-06-01 22:54:14 <kernelc> jn: thanks! I think it's missing in documentation
2021-06-01 23:11:04 <Hello71> agris: it moved from testing to community
2021-06-01 23:11:13 <Hello71> stable branch is orthogonal
2021-06-01 23:11:24 <Hello71> alpine almost never adds packages to existing releases
2021-06-01 23:11:31 <Hello71> that would defeat the purpose of a stable release
2021-06-01 23:11:35 <agris> o
2021-06-01 23:11:45 <agris> so what version of alpine is that package now on?
2021-06-01 23:11:55 <Hello71> edge release, community repo
2021-06-01 23:11:57 <agris> if not testing and if not stable
2021-06-01 23:11:58 <agris> oh
2021-06-01 23:12:02 <Hello71> testing is not a release
2021-06-01 23:12:30 <agris> would it be problematic if I added edge to my repos.conf file and installed that on stable?
2021-06-01 23:12:53 <agris> also, how does alpine know which release it's on and supposed to upgrade packages from if not repos.conf?
2021-06-01 23:18:34 <Hello71> what is repos.conf
2021-06-01 23:18:44 <Hello71> https://wiki.gentoo.org/wiki//etc/portage/repos.conf
2021-06-01 23:19:32 <Hello71> actually portage doesn't have a concept of releases
2021-06-01 23:20:15 <Hello71> except accept_keywords but i think pms defines only two levels of keywording
2021-06-01 23:20:29 <Hello71> well four if you include ** and -*
2021-06-01 23:21:36 <agris> Hello71: /etc/apk/repositories
2021-06-01 23:34:17 <Hello71> if you add edge to repositories then you'll be running a slightly broken edge
2021-06-01 23:34:27 <Hello71> or very broken depending on how you do it
2021-06-01 23:42:37 <agris> how do you fix from that?
2021-06-02 00:29:44 <chexo4> Test!
2021-06-02 00:29:50 <chexo4> Okay, good.
2021-06-02 00:31:44 <chexo4> Hi, I’m having trouble with detecting an iOS device on my Alpine Linux desktop install (among a few other issues with KDE Plasma in particular, but that’s unrelated). Neither checkra1n (jailbreak tool) not libimobiledevice’s utilities detect my iPhone. I’m going to try a different cable here in a second, but I’m worried it could also be a firmware/driver thing.
2021-06-02 01:23:03 <chexo4> hello
2021-06-02 01:23:43 <chexo4> as usual my connection is intermittent because I am testing things, and I have no bouncer so no logs
2021-06-02 01:23:52 <chexo4> has anything been said since I was here last?
2021-06-02 01:25:16 <chexo4> But yeah, a different cable did not fix the issue with libimobiledevice. I believe the issue is with usbmuxd, it doesn't seem to be working correctly. Is it possible the version on edge-testing is outdated?
2021-06-02 01:28:58 <petrj> is there anyone working on porting `aarch64-none-linux-gnu` and `aarch64-linux-gnu` gcc targets on alpine?
2021-06-02 01:30:39 <petrj> in arch, it is called `aarch64-linux-gnu`, some other distros and arm docs call it `aarch64-none-linux-gnu`, the closest package i have found in aports is `gcc-aarch64-none-elf`
2021-06-02 01:45:19 <chexo4> test
2021-06-02 02:02:31 <pinskia> petrj oh I see you are here too
2021-06-02 02:03:09 <pinskia> the vendor bit in the aarch64-linux-gnu is just empty while in the first case it is none; to GCC they are the same target
2021-06-02 02:04:22 <pinskia> people use the vendor part of the triplet/quadet sometimes for cross compilers so it is easier to figure out who's compiler you are using
2021-06-02 02:04:39 <pinskia> e.g. marvell Octeon SDK uses aarch64-marvell-linux-gnu 
2021-06-02 02:18:16 <Ariadne> petrj: there are no plans for cross-compilers to gnu/linux, no
2021-06-02 02:19:04 <Ariadne> chexo4: i never got usbmuxd to work either
2021-06-02 02:19:57 <Ariadne> i picked up an M1 mac though, so i'll just preserve MacOS somehow on it when asahi matures a little more and boot into MacOS when i need to do stuff to my phone
2021-06-02 02:20:07 <chexo4> ah damn
2021-06-02 02:20:30 <chexo4> Ariadne: what should I do for now, then? would something like a chroot work?
2021-06-02 02:20:44 <Ariadne> seems unlikely
2021-06-02 02:20:51 <chexo4> damn
2021-06-02 02:21:35 <chexo4> one of these days I'll set up some sort of cool VFIO setup and then I will be able to simply pass through the relevant USB controller to a specialized VM
2021-06-02 02:21:42 <Ariadne> right now i am using said mac as daily driver as i am waiting for display assembly to arrive from china for my x86 laptop
2021-06-02 02:22:53 <chexo4> fair enough
2021-06-02 02:23:44 <chexo4> thankfully odysseyn1x exists
2021-06-02 02:25:51 <chexo4> anyways bye, booting that
2021-06-02 02:33:56 <petrj> it's for gnu cross toolchain in ports
2021-06-02 02:35:18 <petrj> required by https://blog.project-insanity.org/2020/04/22/linux-with-wayland-is-now-running-on-iphone-7/
2021-06-02 02:35:30 <petrj> i'm flashing my x for pmos
2021-06-02 13:11:46 <Darksecond> what alternative init system was alpine looking at potentially switching to again?
2021-06-02 13:13:00 <lopid> sarcophagus or soemthing
2021-06-02 13:21:04 <jjuran> s6-rc?
2021-06-02 13:34:51 <petrj> currently most software do not provide openrc counterpart to their systemd service config. it would be nice if s6 brings some kind of higher kinder semantics, or very slim translation layer, so we get same enduser experience as systemd. yes, in theory we can write init script  for all applicable packages, but seeing what i have seen in some cases
2021-06-02 13:34:51 <petrj> it does not happen for years..
2021-06-02 13:36:38 <lopid> i do not want the same experience as systemd, thanks
2021-06-02 13:37:28 <petrj> mm systemd is not the subject of the sentence there you are totally off the tracks dude
2021-06-02 13:37:54 <petrj> i'm speaking of behavior, start/stop service in the context of package
2021-06-02 13:37:59 <lopid> i know ;)
2021-06-02 14:07:56 <danieli> last time i checked, s6 did everything so differently it would've taken quite a bit of effort to become a proficient user
2021-06-02 14:07:58 <danieli> granted that was a long time ago
2021-06-02 14:10:18 <koollman> they still do :)
2021-06-02 14:11:30 <koollman> if you want something less radically different, maybe have a look at 66 from obarun (not sure of the status in alpine). The project helps make s6 feel more 'normal'. Still different (but every init system is quite a bit different from the others anyway)
2021-06-02 14:12:32 <bruoah> After an apk update / apk upgrade sway is now throwing constant Segmentation fault, any ideas what may it be? I tried sway -Vd but it does not show any helpful debug information.
2021-06-02 14:13:47 <Misthios> https://github.com/swaywm/sway/issues/6305
2021-06-02 14:15:57 <bruoah> Thanks a bunch for that one Misthios
2021-06-02 14:16:04 <Misthios> np
2021-06-02 14:18:39 <lopid> is it easy to downgrade?
2021-06-02 14:21:12 <Misthios> its more diffuclt as normal but could be possible if you build wlroots from source/ change the abuild for sway and build that localy
2021-06-02 14:21:54 <Misthios> wlroots is now 0,13 and older sway needs 0,10 (now available as wlroots0.12 i believe)
2021-06-02 14:22:03 <Misthios> 0.12* not 0.10
2021-06-02 14:22:03 <lopid> oh, there's no option to "apk add sway-0.x.y" or something like that?
2021-06-02 14:22:24 <Misthios> that wil just pull the latest wlroots atm
2021-06-02 14:22:46 <lopid> so best make a backup right before updating
2021-06-02 14:23:08 <Misthios> if ur on older sway just dont update atm
2021-06-02 14:24:34 <Misthios> i believe apk also has an option to exclude packages from upgrades?
2021-06-02 14:25:40 <lopid> in general. i updated nextcloud and it broke (expected some other nextcloud package to be installed that i didn't have). did a full system restore after spending an hour trying to fix it
2021-06-02 14:25:57 <ikke> You can pin them 
2021-06-02 14:29:05 <lopid> sorry, not a full system restore. i apk del'd it and added afresh
2021-06-02 16:17:02 <oliv3r> lopid: apk fix linux-rpi does; but let me try apk fix mkinitfs, as that for sure is a cleaner way, but still, what if I wanted to trigger any/all triggers? In case I don't know the changes that require the trigger ...
2021-06-02 16:23:55 <lopid> unknown to this author
2021-06-02 17:04:02 <oliv3r> lopid: would be nice if there was a `apk trigger [mkinitfs]` command :)
2021-06-02 23:41:18 <wart_> Hello! I'm using postmarketOS which is on top of alpine linux.  First time user of both, but long time linux user.  Every once in a while sudo will hang for about a minute or two.
2021-06-02 23:41:37 <wart_> htop doesn't report anything out of the ordinary.  Things resolve after a minute or so.  This is on the PinePhone, but I don't think its hardware.
2021-06-03 01:19:43 <marnold> is there a way to have apk to use custom wget options
2021-06-03 01:59:31 <smcavoy> wart_: could be DNS
2021-06-03 02:00:31 <wart_> smcavoy: i checked /etc/hosts and it seems fine.
2021-06-03 02:04:32 <smcavoy> wart_: sudo does dns lookups so its possible somehwere along the resolution chain there is something slowing things down, causing a wait for a timeout and then continue. putting the host's hostname in /etc/hosts *should* remove that as a possibility 
2021-06-03 02:05:52 <smcavoy> wart_: you could attach `strace` to the sudo PID to see if there is anything obvious it is doing (gethostbyname, or something like that). 
2021-06-03 02:10:15 <wart_> yeah i suspect it is dns too.  its intermittent so i'll see what strace says.
2021-06-03 08:48:00 <iveqy> I'm trying to setup my own apk-repository. How does it work when I update a package with a new version?
2021-06-03 08:48:12 <iveqy> when I do that I've two packages, with different names and different versions
2021-06-03 08:48:33 <iveqy> when trying to install I get a conflict since two packages provides the same things and conflict with eachother
2021-06-03 08:48:51 <iveqy> how can I make apk always choose the latest?
2021-06-03 08:54:00 <lopid> different names?
2021-06-03 08:57:17 <iveqy> lopid: how do you mean?
2021-06-03 08:57:32 <lopid> you said you're updating a package but the name is different?
2021-06-03 08:58:31 <iveqy> I've libgcc1-10.2.0-r0.0.apk and libgcc1-9.3.0-r0.0.apk they both have P:libgcc1 but different V:
2021-06-03 08:59:01 <iveqy> so when I do apk add libgcc1
2021-06-03 08:59:21 <iveqy> apk says they conflict instead of just choosing the one with highest V
2021-06-03 09:04:35 <donoban> iveqy: maybe you have some packages installed that force some ligbcc1 version
2021-06-03 09:04:47 <donoban> could you paste the conflict error?
2021-06-03 09:05:26 <donoban> uhM, did you tried apk update?
2021-06-03 09:07:06 <iveqy> I'm not using any official package repository, so it's impossible for you to recreate the error.
2021-06-03 09:07:29 <iveqy> I'm populating a completely new rootfs, so I've done an apk add --init, apk update, apk add
2021-06-03 09:07:56 <iveqy> I just found that apk add has the -l options, that solved the libgcc1 error, but I get stuck on some other dependency
2021-06-03 09:08:19 <iveqy> if I use a fresh apk repository that only have one version of each package, it all works flawless
2021-06-03 09:09:00 <iveqy> ERROR: unsatisfiable constraints: libsystemd0-1:247.2-r0.0: conflicts: libsystemd0-1:247.4-r0.0
2021-06-03 09:12:21 <iveqy> if I'm trying to run apk version -t "1:247.2-r0.0" "1:247.4-r0.0" I get = as a result, which is wrong...
2021-06-03 09:15:59 <donoban> uhM... libsystemd0?
2021-06-03 09:17:40 <iveqy> donoban: yes?
2021-06-03 09:19:37 <iveqy> Think I found the error, In src/version.c in apk there's tokenizer next_token() that accepts . as a divider but not :, adding : make it work. Unclear if that's a good approach though
2021-06-03 09:24:23 <donoban> I'm not sure if trying to add systemd to alpine is a good idea, did you build that packages?
2021-06-03 09:24:39 <ikke> How do you build the apkindex? abuild by default removes old entries 
2021-06-03 09:24:59 <ikke> donoban: libsystemd is pretty generic
2021-06-03 09:25:29 <donoban> ahm
2021-06-03 09:29:47 <iveqy> donoban: I'm not building alpine, I'm building a special version of poky
2021-06-03 09:30:28 <iveqy> ikke: using apk index, or rather I don't I let artifactory build the index for me with their new apk support. Not sure how they do it since it's closed source.
2021-06-03 09:30:40 <iveqy> ikke: I want to keep old entries
2021-06-03 09:32:23 <iveqy> anyway, using --lastest to apk add seems to solve it. Now I'm on the next error ERROR: unsatisfiable constraints: dbus (virtual): provided by: dbus-1
2021-06-03 09:32:33 <iveqy> for some reason virtual packages is not considered
2021-06-03 14:45:12 <gui1020> I would like to use NextCloud 21 available from the EDGE repository, but this is not advisable for production. Should I wait for a new stable release?
2021-06-03 14:48:45 <lopid> i would
2021-06-03 14:49:32 <lopid> i would, am, also consider not installing it from alpine packages
2021-06-03 14:57:02 <ikke> 3.14 is about to be released
2021-06-03 14:57:40 <lopid> (omg omg omg)
2021-06-03 14:59:07 <ikke> We wait for ppc64le to be finished, after which the first rc will be made 
2021-06-03 15:07:10 <markand> anyone knows why I help E149: Sorry, no help for help.txt in vim?
2021-06-03 15:07:30 <markand> I get*
2021-06-03 15:07:43 <gui1020> Ok thank you guys.
2021-06-03 15:07:49 <ikke> markand: maybe you need to add vim-doc?
2021-06-03 15:07:49 <markand> oh, it's in vim-help rather than vim-doc I guess
2021-06-03 15:07:53 <ikke> ah
2021-06-03 15:07:55 <markand> ikke, I have vim-doc!
2021-06-03 15:07:57 <lopid> do ":help E149"
2021-06-03 15:08:19 <markand> yes installing vim-help did the trick
2021-06-03 15:08:27 <markand> lopid, hard to use :help if the command itself fails ;)
2021-06-03 15:08:43 <lopid> well then simply do ":help :help" O:)
2021-06-03 15:08:59 <markand> you can't use :help if you don't have vim-help
2021-06-03 15:09:02 <markand> as simple as that
2021-06-03 15:45:45 <petrj> in some distros, listing device tree is `find /proc/device-tree/ -type f -exec head {} + > mydevicetree`. is the alpine equivalent just `ls /dev > mydevicetree` or something more clever?
2021-06-03 15:59:06 <Hello71> device tree is totally different from /dev
2021-06-03 16:03:38 <petrj> so you are saying for alpine i should be using ....?
2021-06-03 16:07:34 <lopid> i don't see /proc/device-tree in four distros that i'm on right now
2021-06-03 16:08:28 <petrj> guess 17 votes are wrong then https://unix.stackexchange.com/a/289627
2021-06-03 16:08:43 <petrj> but hey, still looking for the real answer :) 
2021-06-03 16:08:55 <petrj> if anyone knows please chime in
2021-06-03 16:09:35 <lopid> yes, those 17 people ARE wrong if they are not looking on the distros i am
2021-06-03 16:10:06 <ikke> isn't device-tree used for arches like arm?
2021-06-03 16:10:37 <petrj> device-tree is used for all archs during the boot, no?
2021-06-03 16:14:58 <Hello71> no
2021-06-03 16:15:07 <Hello71> acpi is... closer to that
2021-06-03 16:15:30 <ikke> afaik, device-tree is to make up for a lack of enumeration of hw, no?
2021-06-03 16:17:37 <Hello71> more or less
2021-06-03 16:25:28 <iggy> petrj: the only system I have that has /proc/device-tree is my rpi3 (it's not alpine, but /proc is provided by the kernel, so I can't imagine that wouldn't be there under Alpine)
2021-06-03 16:28:51 <petrj> iggy: acpidump gets us close enough to what uboot / LK etc. would pass to the kernel. if we want to see exactly "how kernel sees this device", that usually is a good way. device tree gives a bit flatter view of what acpidump shows. i wanted to find out how or what they refer to when kernel sees the hardware button on my headphones.
2021-06-03 17:02:56 <louisev> after an apk update / apk upgrade, my sound seems to be messed up (using alsa). heres a log: https://paste.debian.net/plainh/86f15696 - any ideas?
2021-06-03 17:03:30 <louisev> before that it said that i'm missing some ucm2 folder in /usr/share/alsa , so i fixed that by doing apk add alsa-ucm-conf , now im getting thrown this, yikes
2021-06-03 17:07:35 <valerius> use alsamixer to see what devices even show up, possibly change the default device
2021-06-03 17:09:39 <louisev> i did set defaults.ctl.card and defaults.pcm.card to 1 in /usr/share/alsa/alsa.conf
2021-06-03 17:09:57 <louisev> so that seems to work... just no sound plays at all
2021-06-03 17:11:27 <valerius> might need to restart any programs that need to play sound
2021-06-03 17:11:37 <valerius> maybe unmute stuff that is muted and you missed
2021-06-03 17:12:11 <louisev> so those errors can be ignored?
2021-06-03 17:12:22 <valerius> I had those errors previously and they went away after a reboot
2021-06-03 17:12:33 <valerius> I also use /etc/asound.conf to set my default card
2021-06-03 17:15:48 <louisev> ill give it a shot. thanks for telling me
2021-06-03 17:17:35 <Hello71> petrj: that's backwards though
2021-06-03 17:33:12 <petrj> Hello71: didn't follow, you are not making any sense.
2021-06-03 18:33:22 <lostdeveloper> Hi All!
2021-06-03 18:36:00 <lostdeveloper> I have been trying to find a method to detect new version releases without having to scrape the Alpine Linux website. Anyone looked at this before?
2021-06-03 18:36:10 <clandmeter> hi
2021-06-03 18:36:19 <clandmeter> i was just reading a similar email 
2021-06-03 18:36:21 <ikke> https://alpinelinux.org/atom.xml
2021-06-03 18:36:24 <clandmeter> what a coincidence 
2021-06-03 18:37:38 <clandmeter> lostdeveloper: our releases are tags not branches
2021-06-03 18:37:49 <clandmeter> if you are the same person as in my mailbox :)
2021-06-03 18:38:11 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/tags
2021-06-03 18:38:29 <clandmeter> and we also provide a kind of hidden yml
2021-06-03 18:38:54 <clandmeter> https://dl-cdn.alpinelinux.org/alpine/v3.13/releases/x86_64/latest-releases.yaml
2021-06-03 18:39:04 <clandmeter> but thats harder to track
2021-06-03 18:40:09 <ikke> This as well: https://alpinelinux.org/releases.json
2021-06-03 18:40:11 <Ariadne> lostdeveloper: https://alpinelinux.org/releases.json
2021-06-03 18:40:15 <ikke> :D
2021-06-03 18:40:37 <clandmeter> yes, i forgot about that one
2021-06-03 18:40:40 <clandmeter> thats your source
2021-06-03 18:41:00 <clandmeter> i actually requested it:D
2021-06-03 18:41:06 <ikke> haha
2021-06-03 18:41:13 <ikke> go code to fetch it: https://gitlab.alpinelinux.org/alpine/go/-/blob/master/pkg/releases/releases.go
2021-06-03 18:49:03 <hodapp> are there any good guidelines for backing up an Alpine system (Ras Pi in this case) while removing extraneous crap from the image that isn't used at bootup?
2021-06-03 18:49:25 <hodapp> I know I can just take a ZIP/tarball, and with Ras Pi it just needs to go onto a FAT partition on a card, but I was curious on clearing out things like cache
2021-06-03 18:49:33 <PureTryOut> Today I realized that even though I have iwd installed and the service running, I've never actually been using it. I only learned that after I realized I still had wpa_supplicanta installed and I uninstalled it, which then caused me to have no networking after rebooting lol
2021-06-03 19:19:20 <lostdeveloper> @clandmeter I most likely was hah - I've been really trying to undertand the architecture of Alpine, what the build process is like and how this team builds the AWS AMIs. Lots to learn!
2021-06-03 20:16:03 <etothepii> Hi there - I'm trying out Alpine on a raspi device that's going to run a single python app, and I had a quick question. When running in diskless mode, do packages automatically update when the device is rebooted, or do I still need to use some tool like unattended upgrades to ensure updates are applied (I'm trying to make the system as
2021-06-03 20:16:03 <etothepii> set-and-forget as possible). 
2021-06-03 20:22:41 <zieq> hello
2021-06-03 20:23:12 <zieq> I'm trying to install a system using setup-disk with an apkovl file*
2021-06-03 20:23:28 <zieq> however, I'm wondering how can I setup mounts in /etc/fstab
2021-06-03 20:24:07 <zieq> If there's a /etc/fstab file in the apkovl, it is renamed to /etc/fstab.old, and I cannot setup additional mounts
2021-06-03 20:24:15 <zieq> am I missing something?
2021-06-03 20:25:28 <zieq> (I am installing with "sys" mode)
2021-06-03 20:35:41 <zieq> I also tried to put a file in /etc/fstab.d, but it does not appear to work either
2021-06-03 20:50:03 <Hello71> hodapp: lbu commit?
2021-06-03 20:50:14 <Ariadne> lostdeveloper: thats a question for #alpine-cloud :)
2021-06-03 21:39:12 <hodapp> Hello71: not sure how that helps with the 'removing extraneous crap' part
2021-06-03 21:40:28 <hodapp> I'm already committing things to the SD card - but was interested in what I can disregard from the SD card itself when taking backups
2021-06-03 22:01:44 <Terror> Hi team, can someone update the dockerhub image? I have failed build due to curl 7.76 :( Thanks!
2021-06-03 22:05:55 <Terror> I mean failed builds due to curl 7.76 vulnerabilities.
2021-06-03 22:06:46 <ngortheone> Guys, how do I install alpine on PineBook Pro? 
2021-06-03 22:09:52 <c705> very carefully
2021-06-03 23:45:02 <Ariadne> Terror: you should use `apk add -U` when building docker images
2021-06-04 01:09:56 <etothepii> If I'm running Alpine without a package cache, are packages supposed to reinstall automatically at boot? I have packages listed in /etc/apk/world, but if I reboot nothing comes back until I run 'apk upddate/apk upgrade'. Is that expected/did I miss a setup step?
2021-06-04 01:11:20 <petrj> you just need `apk fix` after modifying world manually
2021-06-04 01:12:21 <petrj> (i have learned that trick from Ariadne's blog :D)
2021-06-04 01:13:38 <etothepii> Thanks
2021-06-04 01:14:00 <etothepii> Do I need to add that to a init script or something to ensure it gets applied on startup?
2021-06-04 01:20:29 <etothepii> Because I have a new install, and at this point all I've installed is iptables, and even though I've run lbu commit, iptables isn't available after reboot until I run the updater. That's why I'm wondering if there was something I missed during setup.
2021-06-04 01:20:39 <petrj> normally we don't modify world manually. so imo no need for defensive programming.
2021-06-04 01:21:14 <petrj> apk add ...grocery list.. <-- i normally just use that
2021-06-04 01:23:18 <etothepii> but do you have that set to run in a boot script, though? I guess I may not be understanding how diskless works with packages
2021-06-04 01:23:49 <etothepii> (I also should probably clarify - I didn't edit /etc/apk/world manually, it got updated when I ran apk add, I was just saying that its still up to date after a reboot)
2021-06-04 01:26:28 <petrj> if you are running from usb boot, then it will just run everything from ramfs. if you installed only config then i'm not sure. third option is to full install on bare metal which im running, so state is preserved after the boot.
2021-06-04 01:27:07 <petrj> it will just run everything from ramfs* and won't preserve state after the reboot.
2021-06-04 01:28:23 <etothepii> ok, so the packages aren't special/there's no magic that re-downloads them. All the lbu tool is doing is just preserving the configs. 
2021-06-04 01:28:56 <etothepii> that makes sense - so I should then just be able to have the system run apk fix on boot to get the packages back. 
2021-06-04 01:30:36 <petrj> yup, i haven't tried lbu options, but sounds like it should work. https://ariadne.space/2021/04/25/why-apk-tools-is-different-than-other-package-managers/ has some good insights on how stuff actually works under the hood. it's simple and scalable. :) 
2021-06-04 01:37:24 <etothepii> Huh. That's interesting, and makes sense. Thank you - I think I know how to get the system to do what I want now!
2021-06-04 02:26:13 <eletrotupi> hello!
2021-06-04 02:26:28 <eletrotupi> I'm having issues regarding connecting a mini-displayport on my laptop
2021-06-04 02:26:43 <eletrotupi> udevadm monitor doesn't display anything about it, when connecting/disconnecting the adapter
2021-06-04 02:27:11 <eletrotupi> Which is weird, as it works fine on a old mac mini I have laying around
2021-06-04 02:27:28 <Hello71> did you remember to actually start udev
2021-06-04 02:27:28 <eletrotupi> so the adapter works fine
2021-06-04 02:27:47 <eletrotupi> yes, udevd is running
2021-06-04 02:28:00 <eletrotupi> I can see events when disconnecting mouse or keyboard
2021-06-04 02:28:51 <Hello71> alpine on mac mini also?
2021-06-04 02:29:06 <eletrotupi> no, only on the laptop
2021-06-04 02:29:20 <Hello71> i think there may not be events if the monitor is not connected
2021-06-04 02:29:29 <eletrotupi> I should mention that I used to use an different adapter on my laptop before when I was running other distros
2021-06-04 02:29:31 <Hello71> passive adapter and all
2021-06-04 02:29:33 <eletrotupi> it worked fine as well
2021-06-04 02:29:52 <Hello71> otherwise i might be out of ideas
2021-06-04 02:30:06 <eletrotupi> the monitor works fine directly when connecting via hdmi
2021-06-04 02:30:22 <eletrotupi> I wonder if there's a library or something missing
2021-06-04 02:31:44 <Hello71> shouldn't be
2021-06-04 02:32:19 <Hello71> mini dp is kinda funny with its entanglement with thunderbolt
2021-06-04 02:36:14 <eletrotupi> yeah
2021-06-04 03:12:05 <Hello71> did you check lspci and /sys/class/drm
2021-06-04 03:13:19 <Hello71>  I don't know for what exactly
2021-06-04 03:13:19 <Hello71> or maybe archiso?
2021-06-04 10:31:40 <telmich> moin moin
2021-06-04 10:32:01 <telmich> Did the latest firefox upgrade break it for everyone or just me? Rendering/drawing is completely broken for me
2021-06-04 11:03:00 <dekedro> Firefox rendering works for me, although cpu performance is worse
2021-06-04 11:04:23 <telmich> hmm, I cannot even type in it anymore - have to switch tho focus to make any character visible
2021-06-04 12:08:49 <holgersson> telmich: Might be worth looking for an upstream bug report (or filing a new one).
2021-06-04 12:08:49 <holgersson> telmich: I read some similar complains in #ubuntu on libera, but I assume that they use mozilla's (glibc) binary while the alpine package needs to be a musl build.
2021-06-04 12:48:31 <telmich> holgersson: thanks for the pointer - just did that on https://bugzilla.mozilla.org/show_bug.cgi?id=1714564
2021-06-04 12:48:39 <telmich> holgersson: do you have a link to the ubuntu report by any chance?
2021-06-04 12:49:10 <holgersson> telmich: np and no, it was just a complaint inside the IRC channel that I noticed
2021-06-04 12:49:20 <telmich> holgersson: ack, thanks
2021-06-04 12:50:25 <holgersson> maybe it's even completely unrelated, but I just noticed that it was spoken of a few times.
2021-06-04 12:54:19 <nico[m]> What would be the right place to execute `mount --make-shared /` prior to kubectl being executed at boot?
2021-06-04 13:31:00 <CcxWrk> nico[m]: If it can be encoded into fstab I'd do that. Otherwise you'd make new openrc initscript that has { need root; before kubernetes } in depend (or however is the service executing kubectl named)
2021-06-04 14:36:58 <bleb> so my alpine vm died last week and had to be manually rebooted
2021-06-04 14:37:22 <bleb> host says it looks like the vm shut itself down
2021-06-04 14:37:38 <bleb> i had already looked in 
2021-06-04 14:38:38 <bleb> /var/log/messages, saw some unprintable ^@ stuff that could have been added before it shutdown
2021-06-04 14:38:50 <bleb> otherwise no indication of what could have caused it to go down
2021-06-04 14:39:18 <bleb> is there anywhere else I might look to diagnose the issue?
2021-06-04 14:39:48 <bleb> or, some type of logging i might enable to have a better chance of diagnosing it the next time it happens?
2021-06-04 14:41:17 <bleb> some have said the issue sounds like an out of memory problem; could i enable logging that would allow me to catch that?
2021-06-04 14:59:23 <programmerq> any time I've had to diagnose something like that, I end up shipping kernel messages external to the VM. in particular a virtual serial device for kernel logs can be super handy. You can have something on the VM host capture it to a file.
2021-06-04 14:59:48 <programmerq> that should give better info if it's a kernel type failure.
2021-06-04 15:01:48 <programmerq> I reproduced a nasty problem once a long time ago where an AWS image was hard rebooting when a certain action was performed. I spun up a local virtualbox VM and installed the exact same version of every single package that the problem systems had and was able to trigger the same behavior. Kernel messages to serial VM output gave me info about the failure and I was able to find a
2021-06-04 15:01:51 <programmerq> ton of information about the fix (there was a patch release on their specific kernel that was only one or two versions behind)
2021-06-04 15:01:58 <programmerq> that was a long time ago but a ton of fun to diagnose
2021-06-04 17:06:23 <libxkbcommon> Where can I report issues with the website (https://alpinelinux.org) ?
2021-06-04 17:09:14 <ikke> https://gitlab.alpinelinux.org/alpine/infra/alpine-mksite
2021-06-04 17:09:54 <libxkbcommon> thanks
2021-06-04 17:20:45 <libxkbcommon> What about issues with https://pkgs.alpinelinux.org/ ? I tried looking through the other gitlab repositories but nothing stood out.
2021-06-04 17:21:48 <ikke> https://gitlab.alpinelinux.org/alpine/infra/infra is a general infrastructure repo, though, the code is hosted here: >https://gitlab.alpinelinux.org/alpine/infra/aports-turbo
2021-06-04 19:35:33 <lopid> how can one install the previous (or a specific) version of a package?
2021-06-04 19:59:42 <ikke> lopid: we can sadly only keep the latest version
2021-06-04 20:00:11 <lopid> oh :(
2021-06-04 20:01:33 <lopid> oh, i see i have setup-apkcache'd. perhaps one can install it from /var/cache/apk…
2021-06-04 21:43:13 <riverdc> is it possible to add packages from multiple repositories (tagged in /etc/apk/repositories) to the same virtual package?
2021-06-04 21:44:12 <ikke> I would assume so
2021-06-04 21:46:43 <riverdc> just running `apk add -t virt x@a y@b` fails for me
2021-06-04 21:47:21 <riverdc> 'y@b' is not a valid child dependency, format is name([<>~=]version)
2021-06-04 21:59:07 <ikke> Then you have your answer
2021-06-05 13:13:13 <nico[m]> I got problems with two browsers at the moment (firefox and midori) that have rendering issues - I assume that there might be some GTK bug in edge
2021-06-05 14:46:43 <ikke> hmm
2021-06-05 15:07:04 <eletrotupi> hello, after installing mesa-dri-intel, it uninstalled a bunch of things, and somehow it looks like my wifi card stopped working
2021-06-05 15:07:11 <eletrotupi> which is super weird
2021-06-05 15:07:24 <eletrotupi> Is there any place I can audit what was installed and whatnot?
2021-06-05 15:08:14 <eletrotupi> the wifi card doesn't show on rfkill list and neither the interface appears on ip link
2021-06-05 15:08:37 <eletrotupi> I wanted to avoid a full reinstall, unless that's what it takes
2021-06-05 15:09:08 <ikke> No, a full reinstall is not necessay
2021-06-05 15:09:15 <ikke> necessary
2021-06-05 15:12:26 <eletrotupi> history | head
2021-06-05 15:12:30 <eletrotupi> ooh, sorry
2021-06-05 15:12:36 <rob0> if the interface does not exist, the driver is not loaded
2021-06-05 15:13:13 <rob0> "modprobe <your-driver-name>" should change that
2021-06-05 15:13:41 <rob0> but no clue from what you say why that driver would not have already been loaded
2021-06-05 15:15:44 <rob0> I suppose you already checked lspci, and the wireless NIC is listed there?  If not, maybe you have a coincidental hardware failure.
2021-06-05 15:16:31 <rob0> see also if there are errors in dmesg
2021-06-05 15:24:35 <eletrotupi> hm lspci list the driver, but doesn't display the corresponding driver being used
2021-06-05 15:29:16 <eletrotupi> fuck yeah
2021-06-05 15:30:03 <eletrotupi> ok, so I've enabled the corresponding driver (modprobe iwlwifi)
2021-06-05 15:30:13 <eletrotupi> dmesg displayed that the firmware was not found
2021-06-05 15:30:36 <eletrotupi> then I've installed linux-firmware
2021-06-05 15:30:46 <eletrotupi> which probably is a meta package
2021-06-05 15:30:55 <eletrotupi> because it fetched a bunch of other modules
2021-06-05 15:31:07 <eletrotupi> reenabled the module and it's now working
2021-06-05 15:31:36 <eletrotupi> now, by looking at my history I runned apk add mesa-dri-intel 
2021-06-05 15:31:58 <eletrotupi> which were the last command I run before I rebooted into a borked alpine install
2021-06-05 15:32:27 <eletrotupi> I also noticed that mesa-dri-intel shows a notice of DEPRECATED but doesn't say anything more about it
2021-06-05 15:32:43 <rob0> ah, yes, iwlwifi is what I have too, it needs Intel firmware
2021-06-05 15:33:00 <rob0> so indeed all this might have been related somehow
2021-06-05 15:34:09 <ikke> I would not expect just installing mesa-dri-intel to uninstall any firmware
2021-06-05 15:34:23 <ikke> secondly, linux-lts should pull in linux-firmware
2021-06-05 15:34:31 <ikke> do you have some kind of non-standard installation?
2021-06-05 15:39:43 <eletrotupi> no, a pretty basic installation
2021-06-05 15:39:55 <eletrotupi> installed on a thinkpad I had laying around
2021-06-05 15:40:57 <ikke> is linux-lts installed?
2021-06-05 15:41:47 <eletrotupi> yes, 5.10.36 
2021-06-05 15:42:24 <ikke> did you manually install any linux-firmware-* you can check /etc/apk/world
2021-06-05 15:43:05 <eletrotupi> I have linux-firmware and linux-firmware-intel on world
2021-06-05 15:44:24 <ikke> Ok, so when you installed linux-firmware-intel, that caused the other linux-firmware-* (which were automatically installed) to be uninstalled
2021-06-05 15:45:38 <eletrotupi> oh, gotcha
2021-06-05 15:46:09 <ikke> THe idea is that by default you get all firmware, but if you want, you can select specific firmware packages to be installed
2021-06-05 15:46:16 <ikke> but then you need to make sure you get everything you need
2021-06-05 15:46:34 <eletrotupi> you either have everything or you pick your own path?
2021-06-05 15:46:58 <ikke> not sure what you mean with path
2021-06-05 15:47:12 <eletrotupi> oh, by manually picking each firmware, tec
2021-06-05 15:47:13 <eletrotupi> etc
2021-06-05 16:21:36 <martijnbraam> is there anything ready-made to mount iscsi in the initramfs?
2021-06-05 16:54:32 <donoban> martijnbraam: there is a scsi feature
2021-06-05 16:54:55 <martijnbraam> the i in iscsi is very important
2021-06-05 16:55:06 <donoban> ahh sorry
2021-06-05 16:55:31 <martijnbraam> I'm looking into translating the iscsi initfs hook from debian to alpine now
2021-06-05 16:56:35 <donoban> nice
2021-06-05 18:52:48 <ralf1307> hi!
2021-06-05 19:04:43 <thatvillagerguy[m]> Does this finally work?
2021-06-05 19:04:59 <thatvillagerguy[m]> ok, finally
2021-06-05 19:05:24 <thatvillagerguy[m]> sorry for the spam, just had some struggle with irc because I never used it before
2021-06-05 19:05:53 <ikke> np, welcome
2021-06-05 19:06:28 <martijnbraam> the eeeh... configure_ip thing in alpine initfs is slightly broken
2021-06-05 19:06:37 <martijnbraam> it runs before the kernel has found the network interfaces :D
2021-06-05 22:34:38 <Hello71> did you set right modules
2021-06-05 22:35:04 <Hello71> iirc modprobe for pci devices shouldn't return until devices until initialized
2021-06-05 22:35:11 <Hello71> uh, devices are initialized
2021-06-06 02:56:51 <raven523> a4/b move 26
2021-06-06 02:56:52 <raven523> oops
2021-06-06 03:42:54 <LACampbell> Just updated my alpine server. Pretty damn painless, only issues were userspace apps that had changed. this distro is really easy to live with
2021-06-06 09:02:24 <Jenkler> any issues with php7-pecl-imagick ? 
2021-06-06 09:02:52 <Jenkler> Unable to install it 
2021-06-06 09:03:29 <ikke> what arch?
2021-06-06 09:03:39 <ikke> and what release?
2021-06-06 09:03:54 <ikke> I can install it on edge/x86_64
2021-06-06 09:04:19 <Jenkler> https://pastebin.com/9QRrp5b0
2021-06-06 09:04:39 <Jenkler> 3.13 
2021-06-06 09:04:47 <Jenkler> X86_64 
2021-06-06 09:05:03 <Jenkler> broken build log to 
2021-06-06 09:05:17 <Jenkler> https://pkgs.alpinelinux.org/package/v3.13/community/x86_64/php7-pecl-imagick
2021-06-06 09:05:37 <ikke> hmm, yes, I can reproduce
2021-06-06 09:06:55 <Jenkler> edge has later 3.4.4-r8
2021-06-06 09:07:16 <ikke> imagemagick-libs-7.0.11.13-r0 provides:
2021-06-06 09:07:18 <ikke> so:libMagickCore-7.Q16HDRI.so.9=9.0.0
2021-06-06 09:07:20 <ikke> so:libMagickWand-7.Q16HDRI.so.9=9.0.0
2021-06-06 09:07:55 <Jenkler> Maybe backport 3.4.4-r8?
2021-06-06 09:11:05 <ikke> Jenkler: libmagick was upgraded as security fix, but it changed soname
2021-06-06 09:11:15 <ikke> so packages that depend on it, need to be rebuilt
2021-06-06 09:14:26 <Jenkler> ikke, why does edge then work?
2021-06-06 09:14:38 <ikke> because on edge they were already rebuilt
2021-06-06 09:14:52 <ikke> apparently someone forgot to push the rebuilt for 3.13
2021-06-06 09:15:19 <ikke> It has just been pushed, so the rebuilt packages should be available soon
2021-06-06 09:17:24 <Jenkler> but edge has build date imagemagick-libs 2021-05-22 06:42:17
2021-06-06 09:17:41 <Jenkler> 3.13 has 2021-06-03 20:45:35
2021-06-06 09:18:06 <ikke> Jenkler: the security fixes were bacported later
2021-06-06 09:18:17 <ikke> and it's php7-pecl-imagick that needs to be rebuild
2021-06-06 09:19:02 <ikke> Build date: 2020-12-12 13:47:5
2021-06-06 09:19:35 <Jenkler> ahh, that make sense ;)
2021-06-06 09:20:45 <ikke> Usually on stable branches we try to avoid pushing updates with soname changes
2021-06-06 09:20:55 <ikke> but in this case, it was required to fix vulnerabilities
2021-06-06 09:21:09 <Jenkler> why is that? 
2021-06-06 09:21:33 <Ariadne> yeah i forgot to backport the rebuilds
2021-06-06 09:21:43 <Jenkler> does it intruduce a lot of bugs?
2021-06-06 09:22:05 <Jenkler> Or too much to rebuild?
2021-06-06 09:22:08 <ikke> both
2021-06-06 09:22:13 <Ariadne> i had already staged them, and test built them locally
2021-06-06 09:22:15 <ikke> but mostly to avoid breaking changes
2021-06-06 09:22:19 <Ariadne> i just forgot to push
2021-06-06 09:22:28 <Jenkler> ok 
2021-06-06 09:26:27 <Ariadne> anyway, imagemagick had quite a lot of CVEs, we want to make sure alpine users get fixes for those issues
2021-06-06 09:26:50 <Ariadne> unfortunately one of the fixes required bumping soname version :p
2021-06-06 09:42:10 <Jenkler> whats the right channel for asking for updates? Would be nice if someone could update  mongodb-tools to 4.2.14
2021-06-06 09:42:44 <Jenkler> My lokal build of mongodb-4.2.14 workd fine ;)
2021-06-06 09:43:10 <artok> mongodb is non-free, right?
2021-06-06 09:43:20 <Jenkler> yes, but not the tools 
2021-06-06 09:43:57 <Jenkler> https://pkgs.alpinelinux.org/package/edge/community/x86_64/mongodb-tools
2021-06-06 09:48:14 <ikke> Open an issue on https://gitlab.alpinelinux.org/alpine/aports
2021-06-06 09:50:36 <Jenkler> ok
2021-06-06 10:31:00 <Jenkler> ikke, hope this is correct :) https://gitlab.alpinelinux.org/alpine/aports/-/issues/12732
2021-06-06 10:43:33 <donoban> hi, pretty off alpine topic but does someone know if its possibly to use some terminal aware of tmux panes so you can easy copy/paste text?
2021-06-06 10:44:54 <lopid> what the does terminal have to do with it? copypaste how?
2021-06-06 10:45:16 <donoban> Only select text from the current pane
2021-06-06 10:45:44 <donoban> maybe it is not possible
2021-06-06 10:46:01 <donoban> or it's internal tmux mouse support dependant
2021-06-06 10:46:09 <lopid> i don't think a terminal can do that. maybe you can copy from tmux to your x buffer directly
2021-06-06 10:46:09 <ikke> I use tmux vi mode, and have a shortcut that puts the selected text in the clipboard
2021-06-06 10:46:59 <donoban> yeah mi problem is that, if you have e.g. 3 panels, full vertical on left, and two on right
2021-06-06 10:47:17 <donoban> and you do 'ls' on the bottom right, you want to copy that and it will get the text from the left pane too
2021-06-06 10:47:35 <ikke> with tmux vi mode, you can select things just from a single pane
2021-06-06 10:47:44 <donoban> I'm currently doing zoom on that pane for tthat but it's pretty cumberstone
2021-06-06 10:47:53 <donoban> uhM, I'm gonna try it
2021-06-06 10:47:59 <donoban> do I need to restart tmxu?
2021-06-06 10:50:12 <ikke> You can do :source-file I guess
2021-06-06 10:51:03 <donoban> ok ok I'm reading about it
2021-06-06 10:59:18 <donoban> ikke: is this not enough for reach X clipboard? send-keys -X copy-pipe-and-cancel 
2021-06-06 11:00:04 <ikke> (I use copyq): bind-key -T copy-mode-vi y send-keys -X copy-pipe-and-cancel 'copyq add -; copyq select 0'
2021-06-06 11:00:24 <donoban> ok thanks!
2021-06-06 11:00:45 <ikke> When you have text selected and you press y, it would put it in the clipboard
2021-06-06 11:01:13 <donoban> when using just Enter what happens exactly?
2021-06-06 11:01:16 <donoban> it goes to some internal tmux buffer?
2021-06-06 11:01:30 <ikke> yes
2021-06-06 11:01:48 <donoban> do you know how to paste it?
2021-06-06 11:01:50 <ikke> You can paste it with <prefix> ]
2021-06-06 11:01:57 <donoban> ah thanks
2021-06-06 18:12:50 <Forza> Hello. I  have two nics with dhcp, and I want the default route from one of them. How can I prevent dhcp from setting default route on the second interface?
2021-06-06 18:22:33 <Forza> Hm I set the second one to static. 
2021-06-06 18:44:14 <fellowken> Does anyone have experience using 'resume' kernel boot arg in Alpine?
2021-06-06 18:50:20 <chexo4> hello
2021-06-06 18:54:16 <ikke> fellowken: no experience
2021-06-06 19:05:22 <Hello71> i think with initramfs it needs to handle manually
2021-06-06 19:05:36 <ikke> there is resume report there
2021-06-06 19:05:42 <ikke> at least, I've seen it
2021-06-06 19:25:28 <Hello71> hm
2021-06-06 19:56:59 <fellowken> It seems I've found a way to fix resume to accept UUIDs, do we have anyone here to review and/or merge it?
2021-06-06 19:57:53 <fellowken> https://gitlab.alpinelinux.org/alpine/mkinitfs/-/merge_requests/84
2021-06-06 20:21:36 <Hello71> seems reasonable, although i think kernel resume doesn't support UUID (same as root)
2021-06-06 21:02:04 <fellowken> Hello71: AFAIK it doesn't matter, the argument is just being fed to stat and then result is passed to /sys/power/resume, so before invoking stat we can just check if the argument is UUID and replace that with the actual dev node
2021-06-06 21:02:47 <fellowken> UUID for root works fine by the way
2021-06-06 21:15:11 <Hello71> yes because initramfs
2021-06-07 15:57:08 <nangel> Hi all... I'm back. :D
2021-06-07 15:57:30 <ikke> nangel: hey, welcome back!
2021-06-07 16:16:10 <danieli> welcome back
2021-06-07 16:29:18 <lopid> thanks
2021-06-07 17:14:30 <nico[m]> Btw, for those who are interested in the firefox-89 rendering bug, the upstream notes are on https://bugzilla.mozilla.org/show_bug.cgi?id=1714564 now
2021-06-07 17:30:50 <MY-R> I force disabled webrender which look slike is default now and no more flickering/tearing snd slow downs
2021-06-07 17:31:35 <MY-R> webrender on xorg worked nice for me with amd/radeon gpus but not on intel
2021-06-07 17:33:58 <MY-R> nico[m]: did you try: gfx.webrender.force-disabled: true ?
2021-06-07 17:40:57 <PureTryOut> Uh can't say I can reproduce that error...
2021-06-07 17:46:36 <nico[m]> MY-R: not yet, but it is extremely hard to click *anything*
2021-06-07 20:12:21 <nristen> I am trying to follow the wiki page, "Classic_install_or_sys_mode_on_Raspberry_Pi" using the armhf installer on a raspberry pi 3b.  The instruction for "setup-disk -m sys /mnt" is returning the error "ext4 is not supported. Only supported are: vfat" 
2021-06-07 20:13:56 <nristen> the prior instruction describes creating an ext4 fs and then mounting it at the location for the setup-disk target... unless I am missing something?
2021-06-07 20:15:26 <MY-R> nristen: https://gitlab.alpinelinux.org/alpine/aports/-/issues/12353
2021-06-07 20:16:09 <mmlb> hey everyone, my ci just noticed that cfss has been yanked from edge/testing for every arch except for s390. Is that unexpected?
2021-06-07 20:16:16 <mmlb> cfssl*
2021-06-07 20:16:32 <MY-R> nristen: so you can run: FORCE_BOOTFS=1 setup-disk -m sys /mnt
2021-06-07 20:17:36 <nristen> MY-R: ah, thanks for the gitlab reference.  I will try that.  Thank you
2021-06-07 20:18:05 <MY-R> nristen: you are welcome :)
2021-06-07 20:18:10 <ikke> > "Well you cannot boot a RPI if you do not have a VFAT partition (for the bootcode.bin, config.txt, etc files) as well as a root partition "
2021-06-07 20:18:58 <nristen> I have two partitions.  The vfat partition and then the larger ext4 partition.
2021-06-07 20:19:28 <ikke> ok
2021-06-07 20:19:31 <MY-R> ye, on vfat keep all those boot stuff code and kernel stuff and rest / on ext4
2021-06-07 20:19:51 <ikke> ah, now I understand what FORCE_BOOTFS does
2021-06-07 20:19:59 <MY-R> is about root fs
2021-06-08 02:37:14 <mep> the latest powerpc alpine image does not boot properly in a vm after installation
2021-06-08 02:37:42 <mep> installation seems to go fine but openfirmware fails
2021-06-08 02:37:48 <mep> the image itself boots
2021-06-08 02:38:56 <mep> it sees a disk it looks like, but then gives the error "No medium"
2021-06-08 03:06:59 <mep> hmm idk but it doesn't look like the virtual disk is showing up in openfirmware, however the installer could see and access it
2021-06-08 04:12:59 <anjan> Hi, I messed up using apk and now if I run apk fix, I get the following output:
2021-06-08 04:13:02 <anjan> ~ # apk fix
2021-06-08 04:13:05 <anjan> ERROR: Unable to lock database: temporary error (try again later)
2021-06-08 04:13:08 <anjan> ERROR: Failed to open apk database: temporary error (try again later)
2021-06-08 04:13:11 <anjan> Is there any way I can recover apk? 
2021-06-08 04:21:47 <anjan> nvm, I fixed it
2021-06-08 04:21:51 <LACampbell> anjan: judging by other people online who have this issue, it may be a network issue
2021-06-08 04:21:53 <LACampbell> oh, what was it?
2021-06-08 04:22:05 <anjan> I ran abuild -r for the 4th time LACampbell 
2021-06-08 04:22:07 <anjan> lmfao
2021-06-08 04:22:17 <anjan> it worked the 4th time not the 3rd time
2021-06-08 04:22:19 <anjan> 
ACTION shrug

2021-06-08 04:23:12 <LACampbell> oh yeah, lucky 4, should have guessed :|
2021-06-08 04:23:22 <LACampbell> guess it really was temporary and you needed to try again later
2021-06-08 08:40:50 <User14> Good morning, I'm trying to tinker my raspberry pi 4b to have a sys install of Alpine. But setup-disk keeps telling me it does not support ext4. Any suggestions?
2021-06-08 08:42:57 <lopid> which image are you using?
2021-06-08 08:44:43 <User14> I am using: alpine-rpi-3.13.5-aarch64
2021-06-08 08:46:44 <ikke> FORCE_BOOTFS=1 setup-alpine
2021-06-08 08:47:52 <ikke> "FORCE_BOOTFS=1 setup-disk -m sys /mnt:
2021-06-08 08:48:33 <User14> Ok, thanks so much, will try that now
2021-06-08 08:49:45 <User14> That works!
2021-06-08 09:00:13 <User14> Well almost, will do another attempt later. It's tricky, but alpine serves the Pi better being small and fast.
2021-06-08 09:02:07 <User14> Whoever made his excellent guide: https://wiki.alpinelinux.org/wiki/Raspberry_Pi_4_-_Persistent_system_acting_as_a_NAS_and_Time_Machine will need to revisit it to add what ikke just told me :)
2021-06-08 10:04:17 <artok> dl-cdn having problems?
2021-06-08 10:04:24 <danieli> artok: fastly is on fire at the moment
2021-06-08 10:09:32 <clandmeter> yup
2021-06-08 10:09:47 <clandmeter> hope not really on fire
2021-06-08 10:09:51 <danieli> ouch, too soon?
2021-06-08 10:09:57 <danieli> (it was untentional fwiw)
2021-06-08 10:10:53 <clandmeter> yeah, but i still have remembrance of ovh
2021-06-08 10:13:18 <clandmeter> https://status.fastly.com/
2021-06-08 10:29:41 <artok> so many services down, sounds like human error. =)
2021-06-08 10:39:56 <koollman> well, it probably requires some automation to propagate the error efficiently
2021-06-08 10:43:24 <artok> misconfiguration replicated, not first time
2021-06-08 10:48:05 <artok> but at least issue has been found
2021-06-08 10:51:51 <lopid> more comments: https://news.ycombinator.com/item?id=27432408
2021-06-08 10:52:31 <lopid> and it makes the beeb: https://www.bbc.com/news/technology-57399628
2021-06-08 14:13:10 <Hello71> mep: what do you mean by "the latest"
2021-06-08 19:28:19 <voroskoi[m]> `linux-firmware-brcm` contains links to files in `linux-firmware-cypress` so `linux-rpi4` should depend on the latter too. What do you think?
2021-06-08 19:28:30 <voroskoi[m]> https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/commit/brcm?id=0f0aefd733f70beae4c0246edbd2c158d5ce974c
2021-06-08 19:28:55 <voroskoi[m]> I think this upstream commit should be handled this way
2021-06-08 20:39:24 <Hello71> abuild is supposed to automatically trace symlink deps
2021-06-08 20:39:26 <Hello71> is it not working?
2021-06-08 21:07:03 <voroskoi[m]> Upgrading wrecked wifi until I figured this out, so i would say no
2021-06-08 23:27:55 <etothepii> A quick question regarding lbu - is the tool expected to follow symlinks? I tried to save a folder in /var/run (it was /var/run/service_name so my service could write pid files without being root), and I noticed the folder wasn't added to the backup. It did work if I added /run/service_name. 
2021-06-09 00:23:04 <mep> Hello71: I mean alpine-standard-3.13.5-ppc64le.iso
2021-06-09 00:49:07 <baconicsynergy> test123. hello from matrix!
2021-06-09 00:49:20 <baconicsynergy> How are you beautiful people doing today
2021-06-09 00:51:37 <jonathon> 
ACTION feels left out by the question

2021-06-09 00:53:36 <jonathon> :P
2021-06-09 00:59:23 <baconicsynergy> its a trick question - you're all beautiful
2021-06-09 01:07:49 <Hello71> mep: not sure what's wrong, maybe file an issue
2021-06-09 01:08:06 <mep> this seems like a qemu issue
2021-06-09 01:08:10 <jonathon> baconicsynergy: thank you. that helps somewhat... XD
2021-06-09 01:08:11 <mep> more likely a user error
2021-06-09 01:10:02 <mep> ill look into it a bit more later today
2021-06-09 02:32:55 <baconicsynergy> I feel like Alpine Linux deserves more attention from software developers
2021-06-09 02:37:27 <baconicsynergy> I like musl and busybox very very much and I feel Alpine has become more than a "container" OS
2021-06-09 02:38:53 <baconicsynergy> I'm running postmarketOS as my daily phone driver, and I have an Alpine on one of my partitions as a desktop. We already know Alpine is great for servers and containers
2021-06-09 02:39:17 <baconicsynergy> I'm very excited for the future of this project and I wish you all the best
2021-06-09 02:39:22 <baconicsynergy> you have my support!
2021-06-09 02:39:22 <valerius> I run it on my overpowered desktop computer and soon my homemade wireless access point
2021-06-09 02:40:40 <baconicsynergy> homemade wap sounds awesome
2021-06-09 02:43:12 <valerius> yeah, got tired of having to hack consumer level routers
2021-06-09 02:43:22 <valerius> so I bought the parts I need to just build my own
2021-06-09 02:43:54 <valerius> now I will have better control over what vulns are on my network hardware
2021-06-09 02:58:45 <baconicsynergy> im tired of consumer level routers myself. i might look into this...
2021-06-09 03:02:33 <valerius> just get an antenna with good Linux driver support and choose your SBC
2021-06-09 03:15:12 <nangel> Yeah,  I run 2 alpinelinux WAPs in our home, simply because I was tired of my wife telling me she couldn't teach her Zoom class, because our netgear failed agin
2021-06-09 03:16:05 <valerius> nangel, what antenna(s) did you use?
2021-06-09 03:22:05 <nangel> For us, 5db
2021-06-09 03:22:14 <Hello71> alpine was never "container os" in the first place, only ignorant people claimed so (and i guess also charlatans)
2021-06-09 03:34:13 <nangel> valarius: you are probably asking for which radio.  you made me go in my closet, since I keep backups
2021-06-09 03:36:38 <nangel> I'm old school, and we live in the country, so not much wifi overlap.   compex 802.11b/g 200mc mini-pci.  I use them in old (ancient) ALIX boxes
2021-06-09 03:43:42 <nangel> valarius: here is where I go... wherever you live you should be able to pick similar: https://www.mini-box.com/Alix-APU-Systems
2021-06-09 05:15:10 <etothepii> valerius: as a note, the Alix/APU systems don't handle gigabit well, as I understand (if if that matters to you - it was something that came up for us when we built our router)
2021-06-09 08:08:56 <Turodek> what oes it mean it says "breaks: world[youtube-dl=2021.04.26]" when trying to install a different version when it is already uninstalled?
2021-06-09 08:27:09 <smcavoy> Turodek: I think that may be due to the version being specified when it was added `apk add youtube-dl=2021.04.06`.This may fix it `apk add youtube-dl && apk upgrade`
2021-06-09 08:29:27 <ikke> apk add -u youtube-dl
2021-06-09 08:30:39 <Turodek> 04.26 is fairly old and when i tried the latest 06.06-r0 it errored out so, wanted to try the 05.16
2021-06-09 08:42:09 <Turodek> nvm seems to be an overlap between edge main and community 🤦‍♂️
2021-06-09 11:37:58 <donoban> I switched to plasma wayland and I noticed that I can't copy from weechat to other windows :\
2021-06-09 11:42:52 <lopid> switched from…?
2021-06-09 11:43:08 <donoban> from plasmas X11
2021-06-09 11:43:14 <donoban> plasma*
2021-06-09 11:44:00 <lopid> then switch back
2021-06-09 11:44:17 <donoban> I can copy/paste from a konsole window to others window
2021-06-09 11:44:32 <donoban> but not the weechat text, wich is between multiple ssh tunnels and tmux :\
2021-06-09 11:44:43 <lopid> do you need to press shift?
2021-06-09 11:44:49 <donoban> yes
2021-06-09 11:44:55 <lopid> that is normal
2021-06-09 11:46:02 <quadrat> afaik it was fixed in 5.22
2021-06-09 11:46:56 <donoban> 5.22 of what?
2021-06-09 11:47:09 <quadrat> plasma
2021-06-09 11:47:19 <donoban> ah
2021-06-09 11:47:26 <donoban> well, I will investigate it later
2021-06-09 11:47:52 <donoban> at least pipewire is running perfect :D
2021-06-09 12:03:19 <MY-R> donoban: if you dont wanna use mouse in weechat to switch between buffers etc then you can disable mouse support and you will have back your typical copy behavior
2021-06-09 12:03:40 <MY-R> donoban: you can temporary toggle on/off with "Alt + m"
2021-06-09 12:05:15 <MY-R> donoban: otherwise with mouse support enabled pressing "shift" is necessary
2021-06-09 12:10:26 <Jenkler> ncopa, there?
2021-06-09 12:10:47 <Jenkler> I have updated mongodb + tools 
2021-06-09 12:10:49 <Jenkler> https://tpaste.us/g6wv
2021-06-09 12:17:26 <Jenkler> https://gitlab.alpinelinux.org/alpine/aports/-/issues/12732#note_161804
2021-06-09 12:19:09 <clandmeter> Jenkler: the convenient way to do this is to create an MR or send the patch to the aports ML
2021-06-09 12:19:15 <mep> eh
2021-06-09 12:19:21 <clandmeter> this way CI gets a chance to test it
2021-06-09 12:20:01 <mep> i just installed and upgraded alpine in a VM and /usr/bin is set permissions 700
2021-06-09 12:20:05 <mep> is this normal???
2021-06-09 12:20:26 <mep> by upgraded I mean I'm on edge. I used a apk upgrade -a
2021-06-09 12:21:54 <mep> the only really out of the ordinary thing I did was install the testing/unrealircd package
2021-06-09 12:22:20 <mep> which is pretty broken in other aspects... could it have overwritten the directory permissions of /usr/bin?
2021-06-09 12:22:49 <Jenkler> clandmeter, MR=merge request but ML= ?
2021-06-09 12:23:00 <Jenkler> Mailing list 
2021-06-09 12:23:05 <Jenkler> gahh ;)
2021-06-09 12:23:05 <clandmeter> yep
2021-06-09 12:23:19 <clandmeter> we recently have a bridge from ML to GL
2021-06-09 12:23:30 <Jenkler> GL ?
2021-06-09 12:23:35 <ikke> gitlab
2021-06-09 12:23:48 <clandmeter> thanks to ddevault
2021-06-09 12:23:56 <clandmeter> yes sorry
2021-06-09 12:24:02 <clandmeter> im tired of typing :)
2021-06-09 12:24:11 <Entroacceptor> there must be a song in there somewhere. "cd to the CI/CD"
2021-06-09 12:31:15 <Jenkler> ikke, it seams that I do not have any access to fix a merge request
2021-06-09 12:33:27 <Jenkler> Any way to get that perm, or is this only for special members only?
2021-06-09 12:45:04 <donoban> ohh thanks MY-R , I enabled mouse support on first days with weechat but now I really don't use it, it works perfect disabing it
2021-06-09 12:56:18 <ikke> Jenkler: if it's a merge request from someone else, then you could ask that person if you can get permissions on their fork
2021-06-09 12:58:58 <ncopa> jenkler: I think you should keep Marc in contributor comment
2021-06-09 13:01:23 <Jenkler> ncopa, sure. I fix 
2021-06-09 13:10:19 <Jenkler> ikke, what do you mean with "merge request from someone else" ? 
2021-06-09 13:10:57 <Jenkler> https://tpaste.us/kkPz
2021-06-09 13:11:14 <ikke> You said you cannot fix a MR, which MR?
2021-06-09 13:14:09 <ikke> for your own MRs, you should be able to modify them
2021-06-09 13:14:27 <Jenkler> ikke, I cant create one
2021-06-09 13:15:11 <Jenkler> I can create issues in gitlab but not MR 
2021-06-09 13:15:11 <ikke> did you fork aports?
2021-06-09 13:15:30 <ikke> You first need to fork aports, then push your branch to your fork, and create a merge request from there
2021-06-09 13:16:09 <nico[m]> An update on the firefox rendering issue might be related to the mesa libraries on Alpine - https://bugzilla.mozilla.org/show_bug.cgi?id=1714564 ; I'll checkout fedora 34 today to verify whether that's actually the case
2021-06-09 13:16:54 <Jenkler> ikke, ok, so I should not create a patch.
2021-06-09 13:18:09 <Jenkler> ikke, I need to get my own aports fork and modify that.
2021-06-09 13:18:35 <ikke> yes
2021-06-09 13:18:49 <ikke> fork it, clone the repo, create a branch, commit on that branch, push it, create an MR
2021-06-09 13:19:30 <Jenkler> does the branch need a special name?
2021-06-09 13:19:51 <ikke> No, just give it some descriptive name
2021-06-09 13:20:00 <ikke> please don't use master for merge requests :)
2021-06-09 13:20:11 <Jenkler> like jenkler ?
2021-06-09 13:20:41 <ikke> related to what you want to change
2021-06-09 13:20:48 <ikke> upgrade-mongodb-tools for example
2021-06-09 13:20:58 <ikke> it doesn't matter that much, but it helps you keep things organized
2021-06-09 13:21:13 <ikke> especially if you want to create more merge requests later
2021-06-09 13:21:25 <Jenkler> ok, but can i make a MR for both mongodb and mongodb-tools 
2021-06-09 13:21:46 <ikke> as they are related, you can add 2 commits to the same MR
2021-06-09 13:21:46 <Jenkler> * should 
2021-06-09 13:21:48 <ikke> or more
2021-06-09 13:21:49 <samialazar> hello
2021-06-09 13:22:48 <Jenkler> but i need 2 commits, one for each pkg 
2021-06-09 13:23:00 <ikke> yes
2021-06-09 13:23:13 <ikke> https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/COMMITSTYLE.md
2021-06-09 13:25:49 <Jenkler> ikke, thanks
2021-06-09 13:25:56 <samialazar> Iam using openwrt-router as tftp-server for netbooting. Devuan booted sucessfuly but alpine is hanging (kernel-alpine-3.13.5 initramfs-alpine-3.13.5). the server loads both of them then it goes to black screen (monitor has no signal)
2021-06-09 13:30:45 <nristen> change the setup disk command to: FORCE_BOOTFS=1 setup-disk -m sys /mnt
2021-06-09 13:40:17 <Hello71> nristen: wat
2021-06-09 13:40:42 <Hello71> samialazar: what parameters did you set?
2021-06-09 13:40:50 <Hello71> you can't just shove any initramfs you want into pxe
2021-06-09 14:24:25 <samialazar> Hello71: Can you elaborate your question please? I am a newbie to this? Thanks
2021-06-09 14:30:47 <Hello71> what did you do
2021-06-09 14:30:51 <Hello71> explain your process
2021-06-09 14:34:14 <Jenkler> ikke, https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/22179
2021-06-09 14:34:30 <Jenkler> ikke, https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/22178
2021-06-09 14:34:46 <Jenkler> ncopa, I hope this looks fine 
2021-06-09 14:36:44 <Jenkler> Why does it say "Detached merge request pipeline #84153 failed for 3b744c3a" What is the pipeline thing ;)
2021-06-09 14:39:47 <samialazar> Well, the router configuration doesnt have any problem it is clearly doing its job. The problem seems to be from the images. the image's origin can be found here https://code.ungleich.ch/ungleich-public/ungleich-tools/-/blob/9ff028147fdac1741f3dd15edaede76fe9ce907c/alpine-rebuild-initramfs.sh I generated it from here.
2021-06-09 14:39:48 <Hello71> did you open any of the links? or google it?
2021-06-09 14:40:26 <Hello71> samialazar: your premise is flawed
2021-06-09 14:42:25 <Hello71> first response was to Jenkler 
2021-06-09 14:46:14 <Hello71> i wonder how it ever built with --use-system-stemmer
2021-06-09 14:46:34 <Jenkler> Hello71, ?
2021-06-09 14:47:10 <Jenkler> Hello71, What respons, I am confused now ;)
2021-06-09 14:47:14 <Hello71> 14:36 <Jenkler> Why does it say "Detached merge request pipeline #84153 failed for 3b744c3a" What is the pipeline thing ;)
2021-06-09 14:48:26 <Hello71> this question can easily be answered by clicking on the link or googling the error
2021-06-09 14:48:33 <Hello71> http://github.com/search?q=gitlab+pipeline
2021-06-09 14:48:34 <Jenkler> Hello71, it seams that the build is marked failed 
2021-06-09 14:48:35 <Hello71> er
2021-06-09 14:48:39 <Hello71> http://google.com/search?q=gitlab+pipeline
2021-06-09 14:48:50 <Hello71> yes, did you read the error log?
2021-06-09 14:49:17 <Jenkler> but i dont have a pipeline or do I x)
2021-06-09 14:49:31 <Jenkler> Just updated the buildfile 
2021-06-09 14:50:32 <Jenkler> https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/22178 <-- here 
2021-06-09 14:50:51 <quadrat> yeah because the x86-64 build failed
2021-06-09 14:51:36 <quadrat> "Target 'build/opt/mongo/mongo' depends on the availability of a system provided library for 'stemmer', but no suitable library was found during configuration."
2021-06-09 14:53:12 <Jenkler> but i can compile this localy
2021-06-09 14:53:37 <quadrat> probably because you have stemmer installed, the container is a clean state, that's why stemmer is not available
2021-06-09 14:54:57 <Jenkler> I am building this with abuild -r in a clean alpine 3.13 env 
2021-06-09 14:55:09 <quadrat> look at the logs https://gitlab.alpinelinux.org/Jenkler/aports/-/jobs/411248
2021-06-09 14:55:13 <quadrat> i was just pointing it out
2021-06-09 14:57:18 <Jenkler> so the other ARCH works because they are disabled 
2021-06-09 15:04:09 <Jenkler> quadrat, so every pkg that make a MR for is built and checked auto?
2021-06-09 15:04:26 <quadrat> yes
2021-06-09 15:04:39 <quadrat> as long as it's configured in the repo (which it is in aports)
2021-06-09 15:04:53 <Jenkler> quadrat, but it passed before
2021-06-09 15:05:30 <Jenkler> quadrat, when I build it localy its a clean env too 
2021-06-09 15:05:51 <Jenkler> I build it in a chroot 
2021-06-09 15:06:15 <quadrat> idk what's different, just telling you that it's the reason why it failed.
2021-06-09 15:06:41 <quadrat> maybe look what's different in this container image: alpinelinux/alpine-gitlab-ci:latest
2021-06-09 15:06:50 <quadrat> as that's the one it's using to build
2021-06-09 15:09:14 <Jenkler> ok
2021-06-09 15:09:43 <Jenkler> quadrat, about "Fast-forward merge is not possible. Rebase the source branch onto the target branch." 
2021-06-09 15:09:55 <Jenkler> is it becuse I am behind?
2021-06-09 15:11:37 <quadrat> could be but no idea. i am a beginner in sth like this
2021-06-09 15:11:59 <Jenkler> ahh ok 
2021-06-09 15:54:48 <Jenkler> ncopa, is this correct now? https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/22179
2021-06-09 15:55:25 <Jenkler> I says cannot merge. Must i be 100% uptodate to merge?
2021-06-09 15:55:46 <Jenkler> The source branch is 3 commits behind the target branch
2021-06-09 16:17:38 <Jenkler> Now it looks more fine. the one that merges this need too "Rebase the source branch onto the target branch" if it is after master
2021-06-09 17:31:22 <theartist> Hello!
2021-06-09 17:32:34 <theartist> How do i set file limits in alpine?
2021-06-09 17:32:44 <theartist> (persistently)
2021-06-09 17:33:38 <Foucault> file limits meaning? open file descriptors?
2021-06-09 17:33:51 <theartist> yep
2021-06-09 17:34:28 <ikke> for users or services?
2021-06-09 17:34:29 <Foucault> I assume it would he same as with other sysctl values
2021-06-09 17:36:04 <theartist> I have a k3s cluster in running alpine linux VMs, and I was trying to deploy an elastic search instance and got this:
2021-06-09 17:36:09 <theartist> bootstrap check failure [1] of [1]: max file descriptors [4096] for elasticsearch process is too low, increase to at least [65535]
2021-06-09 17:36:19 <ikke> how are the processes started?
2021-06-09 17:36:38 <theartist> ikke: Process is inside k3s (kubernetes) container
2021-06-09 17:37:09 <ikke> so docker?
2021-06-09 17:37:40 <theartist> containerd, but yeah like docker
2021-06-09 17:38:30 <ikke> oh, right, I follow now
2021-06-09 17:42:12 <ikke> https://github.com/rancher/k3os/issues/87
2021-06-09 17:47:01 <ikke> https://github.com/rancher/k3os/pull/343/files
2021-06-09 17:49:54 <theartist> How does this apply to alpine? Adding stuff in sysctl.d/something.conf?
2021-06-09 17:50:51 <ikke> theartist: ulimits are inherrited
2021-06-09 17:51:03 <ikke> and you need privileged access to go beyond the hard limit
2021-06-09 17:51:51 <ikke> so if you run alpine as a container, you need to make sure the hard limit is already increased
2021-06-09 17:51:51 <theartist> ikke: yeah, my question is how do I set the hard limit system-wide and persistenctly
2021-06-09 17:52:08 <ikke> theartist: containers are not virtual machines
2021-06-09 17:52:46 <ikke> so whatever k3s sets, is the hard limit
2021-06-09 17:55:15 <theartist> from where does k3s get the hard limit? Isnt' it from the system? I thought that it is what `ulimit -Hn` reports
2021-06-09 17:55:35 <ikke> theartist: are you running alpine as host or in the containers?
2021-06-09 17:55:45 <theartist> ikke: alpine is the host!
2021-06-09 17:56:40 <theartist> each container has different distro, the containers get the ulimit from the alpine host
2021-06-09 18:00:34 <ikke> ok, then what Foucault set, through sysctl
2021-06-09 18:02:13 <theartist> ikke: well setting file-max through sysctl.d or sysctl.conf does not alter output of ulimit -Hn at all
2021-06-09 18:06:44 <ikke> For services, you can set rc_ulimit="-Hn 65535"
2021-06-09 18:06:51 <ikke> in /etc/conf.d/<servicename>
2021-06-09 18:07:29 <theartist> ah nice, let me try this
2021-06-09 18:22:19 <theartist> well tried to add this in /etc/conf.d/k3s but I suppose that k3s is the controlling process, I somehow need to set this for containerd or even init?
2021-06-09 18:22:53 <ikke> do you manage k3s with openrc?
2021-06-09 18:24:05 <theartist> ikke: yep
2021-06-09 18:24:20 <theartist> but i think k3s talks to containerd or something to spawn containers
2021-06-09 18:24:35 <theartist> I don't see the container processes as children of k3s in htop
2021-06-09 18:31:10 <ikke> the PR I linked to shows that these tools are managing these limits as well
2021-06-09 18:33:28 <theartist> ikke: I do not use this 'k3os' project I use 'k3s': https://github.com/k3s-io/k3s
2021-06-09 18:40:14 <theartist> ikke: isn't there a way to set the ulimit globally and not per service?
2021-06-09 18:40:47 <ikke> /etc/rc.conf
2021-06-09 18:48:12 <theartist> ikke: It finally worked by setting rc_ulimit in /etc/rc.conf!
2021-06-09 18:48:18 <theartist> Thank you really much
2021-06-09 20:27:33 <lopid> alpine's git package doesn't contain contrib, and i don't see it in a separate package…
2021-06-09 20:29:45 <lopid> git-prompt.sh specifically. i see some other stuff in separate packages, but not git-prompt.sh
2021-06-09 20:29:58 <lopid> i guess i must copy it over manually
2021-06-09 20:33:05 <ikke> I suppose we could add subpackages for them
2021-06-09 22:54:24 <thebtm> Good Afternoon, I am wondering if anybody here has used a gitlab-runner in diskless mode on alpine?
2021-06-09 22:54:51 <thebtm> or success with github actions **Cringes** 
2021-06-09 23:04:37 <thebtm> The thought is to run one of the runners on a Raspberry Pi for Alpine to complete some arm comutations or if you know of a online resource that can be a VM Raspberry Pi, that would be awesome too ><
2021-06-09 23:34:27 <Sheila> oracle cloud has a free tier which includes access to Ampere-based aarch64 VMs.
2021-06-09 23:34:45 <Sheila> should be able to tune that to build pi-compatible binaries
2021-06-09 23:36:28 <thebtm> would that build an armv7 and armhf binairies?
2021-06-09 23:38:49 <Sheila> I believe so, yes
2021-06-09 23:41:12 <Hello71> alpine doesn't support multilib but you can use a chroot
2021-06-10 00:40:41 <c705> i just realized I missed 3.13.4 entirely. oops
2021-06-10 04:50:39 <LACampbell> I'm trying to run a binary and I'm getting "not found". it's not a $PATH issue (autocomplete works and I've verified the binary is there)
2021-06-10 04:50:55 <LACampbell> and it's not all binaries, just this one
2021-06-10 04:51:36 <ikke> LACampbell: what does ldd <filename> return?
2021-06-10 04:53:12 <LACampbell> /lib/ld-musl-x86_64.so.1: cannot load /home/lewis/.yarn/bin/http-server: No such file or directory
2021-06-10 04:53:16 <LACampbell> but it's there..
2021-06-10 04:54:24 <LACampbell> eh, it's not that important. I mean I am curious why but I can work around it
2021-06-10 04:57:04 <Sheila> what does `file` say about it
2021-06-10 05:13:44 <LACampbell> actually... how I can tell which path a binary I call is from?
2021-06-10 05:14:06 <ikke> `which <binary>`
2021-06-10 05:14:51 <ikke> one reason this error happens is why you try to run a library built against on glibc on a system using musl
2021-06-10 05:15:22 <LACampbell> ok `which` returns empty. which to me suggests that... ash thinks the binary is in my system, hence why it popups with tab complete, but it's not actually there
2021-06-10 05:16:13 <Sheila> `which` is complicated; if `command -v` doesn't come up with it, then either it's not actually in $PATH or something else is going on.
2021-06-10 05:20:49 <LACampbell> Sheila: ok that shows up
2021-06-10 05:21:32 <LACampbell> hmm, but I've mucked around my path since then so it actually points to a working bin
2021-06-10 05:21:47 <Sheila> so, again, what does `file` say
2021-06-10 05:22:19 <LACampbell> file is fine now.
2021-06-10 05:22:39 <LACampbell> so the issue was... I had some old version of the binary somehow cached, so autocomplete was hitting it
2021-06-10 05:22:43 <LACampbell> but it didn't actually exist
2021-06-10 07:32:53 <markand> I have a raspberry no longer booting because initramfs spawns an emergency shell
2021-06-10 07:33:05 <markand> [   10.839725] Mounting boot media: failed.
2021-06-10 07:33:27 <markand> it's very strange since I can mount the root device correctly, sh -x /init seems to fails at nlplug-finds
2021-06-10 07:46:21 <markand> oh nvm, it looks like root= has been removed somewhere for some reasons
2021-06-10 09:18:17 <voroskoi[m]> libinput do not recognize my rapoo wireless keyboard, am I missing something or these do not work out of the box?
2021-06-10 09:32:53 <PureTryOut> Does anyone know how to get qxl working for virt-based virtual machines? Right now I just get "unsupported configuration: domain configuration does not support video model 'qxl'"
2021-06-10 10:14:33 <donoban> uhM I had same problem and installed all qxl pacakges found by apk search....
2021-06-10 10:15:06 <PureTryOut> all of them? 😂
2021-06-10 10:15:15 <PureTryOut> That seems a bit excessive. But ok, some missing package then
2021-06-10 10:15:25 <donoban> hehe well let me check , it was months ago
2021-06-10 10:15:37 <donoban> lol
2021-06-10 10:15:53 <donoban> gotcha :D
2021-06-10 10:16:33 <donoban> have you tried running it directly with command line?
2021-06-10 10:16:45 <PureTryOut> No, I just run it through virt-manager (GUI)
2021-06-10 10:17:03 <donoban> let me try with qemu-openrc
2021-06-10 10:18:03 <donoban>  * start-stop-daemon: /usr/bin/qemu-system-x86_64 died
2021-06-10 10:18:06 <donoban> ouch
2021-06-10 10:18:25 <donoban> Failed to open module: Error relocating /usr/bin/../lib/qemu/hw-display-qxl.so: qemu_spice_display_switch: symbol not found
2021-06-10 10:18:39 <PureTryOut> Run a apk fix Ig uess
2021-06-10 10:18:46 <PureTryOut> *I guess
2021-06-10 10:19:02 <donoban> uhM, nothing
2021-06-10 10:19:19 <donoban> it seems a linking error
2021-06-10 10:19:27 <donoban> I guess no many people uses qxl here
2021-06-10 10:19:32 <PureTryOut> Well yes, `apk fix <whatever package has that broken file>`
2021-06-10 10:19:50 <donoban> this file qemu/hw-display-qxl.so ?
2021-06-10 10:19:56 <donoban> ah
2021-06-10 10:19:58 <donoban> that package
2021-06-10 10:20:13 <ikke> upstream moved more things to modules
2021-06-10 10:20:24 <donoban> apk fix qemu-hw-display-qxl ?
2021-06-10 10:21:15 <donoban> uhm
2021-06-10 10:21:29 <PureTryOut> Yeah
2021-06-10 10:21:29 <donoban> maybe I'm missing some spice package?
2021-06-10 10:21:41 <donoban> it reintalls the package but the problem persists
2021-06-10 10:22:25 <donoban> probably is some dependency missing
2021-06-10 10:24:35 <donoban> lol
2021-06-10 10:24:46 <donoban> fixed it but didn't notice
2021-06-10 10:24:55 <donoban> because there is another error
2021-06-10 10:24:59 <donoban> Failed to open module: Error relocating /usr/bin/../lib/qemu/ui-spice-core.so: egl_texture_blend: symbol not found
2021-06-10 10:25:25 <PureTryOut> Your system seems screwed up
2021-06-10 10:26:07 <donoban> well, I don't have libvirt, probably I'm missing a lot of packages
2021-06-10 10:30:26 <donoban> I don't know what more to try
2021-06-10 10:43:18 <donoban> ey PureTryOut , are you reading on #alpine-devel?
2021-06-10 10:43:32 <donoban> it finally works
2021-06-10 10:43:48 <donoban> sudo apk add qemu-ui-opengl 
2021-06-10 10:45:20 <PureTryOut> Doesn't fix the qxl issue
2021-06-10 10:45:42 <donoban> ouch
2021-06-10 10:45:49 <donoban> well
2021-06-10 10:47:28 <donoban> try with some of this: https://tpaste.us/X1QD
2021-06-10 10:51:05 <PureTryOut> Nope
2021-06-10 11:07:28 <donoban> ey PureTryOut if you want to add some info about libvirt: https://gitlab.alpinelinux.org/alpine/aports/-/issues/12741
2021-06-10 16:20:20 <c705> is the topic a mistake? "3.13.5"? Should it be .4?
2021-06-10 16:22:46 <trylletre> the latest is .5 though?
2021-06-10 16:24:53 <ikke> https://alpinelinux.org/releases/
2021-06-10 16:26:38 <c705> soo..yes?
2021-06-10 16:27:33 <c705> i'm also seeing an email from ncopa saying "Alpine 3.14 release candidate 1 is out". sooo....
2021-06-10 16:27:49 <trylletre> why would the topic have 3.13.4 when its not the latest release from the above link
2021-06-10 16:28:33 <c705> the above link has edge, then v3.13. I don't even see 3.14 
2021-06-10 16:28:49 <c705> oh 
2021-06-10 16:28:51 <trylletre> because 3.14 isnt out
2021-06-10 16:28:54 <c705> sorry i'm a moron disregard
2021-06-10 16:29:06 <c705> it's a minor release, not a major one
2021-06-10 16:29:31 <trylletre> all good
2021-06-10 16:29:37 <c705> 3.13.5 != 3.15
2021-06-10 16:34:12 <MY-R> trylletre: is almost out! :P https://dl-cdn.alpinelinux.org/alpine/v3.14/releases/
2021-06-10 16:34:44 <trylletre> yeah, ive been checking :) just a bit more
2021-06-10 18:31:20 <thebtm> Good Afternoon channel, I was reading the other day how to convert from diskless to sys mode but now I can't seem to find it. does anybody here know where that would be located?
2021-06-10 18:36:12 <thebtm> Nevermind, found it ^^
2021-06-10 18:36:30 <ncopa> thebtm: its pretty easy. create the partitions and mount them somewhere. for example /dev/sda1 as ext4 for boot and /dev/sda2 for /, then you mount /dev/sda2 /mnt and /dev/sda1 as /mnt/boot
2021-06-10 18:36:36 <ncopa> then you do: setup-disk /mnt
2021-06-10 18:37:05 <ncopa> it can handle lvm partitions, mdadm raid, and lvm on top of mdadm raid
2021-06-10 18:37:42 <ncopa> it will clone the running system to the mounted disks
2021-06-10 18:38:28 <thebtm> I am on a raspberry pi 2, so I still need boot to be on the SD card.
2021-06-10 19:07:08 <MY-R> thebtm: then remember to run: FORCE_BOOTFS=1 setup-disk -m sys /mnt .... if your "/boot" partition will be fat one and / for example ext4
2021-06-10 19:12:51 <thebtm> so that will keep boot on the SD card while the SSD is the root partition?
2021-06-10 19:13:58 <MY-R> thebtm: https://gitlab.alpinelinux.org/alpine/aports/-/issues/12353
2021-06-10 19:14:31 <MY-R> just a bug when using setup-disk on rpi
2021-06-10 19:17:48 <thebtm> Thank you ^^
2021-06-10 19:18:16 <thebtm> though a 240GB SSD is bit of a waste on a Pi 2 but its what I have as a spare drive at the moment ><
2021-06-11 06:29:07 <LACampbell> should the examples in the wiki have 'sudo' in them? I prefer to omit it since it's not in the base system
2021-06-11 07:33:59 <ncopa> +1
2021-06-11 08:57:04 <liske> LACampbell: imho using sudo in docs is wrong - the user should decide howto get privileged (root shell, su, sudo, setcap, suid ...)
2021-06-11 08:59:02 <PureTryOut> Most if the time if a command needs root right, I prefix the line with `#`, and `$` if it doesn't
2021-06-11 09:01:37 <lopid> this can get confusing when you want to add a comment in a shell code block
2021-06-11 09:02:19 <lopid> (and incidentally was the precise reason # was originally chosen to represent root)
2021-06-11 10:35:28 <LACampbell> liske: I'm partial to doas myself
2021-06-11 10:47:41 <baconicsynergy> doas is very cool
2021-06-11 10:48:57 <CcxWrk> Not having to gain privileges from unprivileged process is cooler. 8-)
2021-06-11 10:49:31 <LACampbell> that is the coolest. but when I can't do that I go with doas
2021-06-11 11:20:38 <krystianch> Which package can I install to get Linux man pages? For example pthread_create(3). I can currently access man pages from *-doc packages, I just can't find the specific package I need.
2021-06-11 11:21:30 <LACampbell> oh there's a meta package you need to download
2021-06-11 11:22:20 <ikke> man-pages
2021-06-11 11:23:54 <krystianch> Thanks! I was looking for something-doc, but man-pages was what I needed.
2021-06-11 15:17:27 <bedax> out of interest, does anyone know if alpine's package manager can be used in a source-based way, i.e. compiling everything from source like gentoo instead of download pre-built binaries?
2021-06-11 15:19:24 <ikke> bedax: you can clone aports, use abuild to build your own packages, and install the packages from your own repo
2021-06-11 15:19:31 <ikke> there is no directer way
2021-06-11 15:19:59 <bedax> right, I'm guessing the installer doesn't support that?
2021-06-11 15:20:09 <bedax> or, that can't be done during installation*
2021-06-11 15:22:04 <ikke> apk just knows how to install packages, it has no support for compiling them on the go
2021-06-11 15:22:56 <bedax> ah okay, thanks ikke 
2021-06-11 15:29:46 <Hello71> alpine doesn't really have an installer
2021-06-11 17:04:44 <lopid> don't suppose anyone's gotten tinyfugue to compile against musl?
2021-06-11 17:06:06 <lopid> ooh, it's in edge!
2021-06-11 17:06:19 <lopid> for three years
2021-06-11 17:06:47 <lopid> i spy its patches :)
2021-06-11 17:14:17 <lopid> same error as i was getting before. perhaps there's a working musl-based update in edge that's needed
2021-06-11 17:43:25 <gjabell> if I use a password when creating an lbu apkovl can it still be used when booting into diskless mode?
2021-06-11 17:43:41 <gjabell> ie will it function similar to luks and ask for the password on boot?
2021-06-11 18:41:32 <donoban> yes gjabell it will ask the passphrase for decrypt the overlay
2021-06-11 18:46:31 <gjabell> nice, thanks :)
2021-06-11 19:21:53 <thebtm> Just wondering, what is the difference between Armhf and Armv7? 
2021-06-11 19:23:07 <ikke> armhf is armv6 with hard float
2021-06-11 19:24:13 <ikke> (there were also armv6 softfloat variants)
2021-06-11 22:14:13 <brendonjmb> How does one *fully* upgrade packages on a RPi installation? I have updated /etc/apk/repositories, run `apk upgrade -U`, and rebooted yet `uname` still yields kernel 5.4.72 which is pretty old, I believe. If I try `apk add linux-rpi4` I get module installation errors (known issue, I think?). Is there a proper way to make sure my RPi installation is
2021-06-11 22:14:13 <brendonjmb> up-to-date? Thanks.
2021-06-11 22:19:04 <gjabell> anyone know if there are extra steps required to boot from an encrypted apkovl? I get to the password prompt after boot but it doesn't seem to accept keyboard input
2021-06-11 22:28:52 <fremen2001> hello :) I have a question: is it possible to revert back from edge without reinstalling?
2021-06-11 22:35:41 <quadrat> brendonjmb you would need to save the changes (which is done with the lbu tool)
2021-06-11 22:36:57 <brendonjmb> I forgot to mention that I do run `lbu ci` . Since the RPi setup is diskless, do I just have to extract a new RPi tarball onto my SD card?
2021-06-11 22:37:33 <brendonjmb> Because upgrading packages normally works when I `apk upgrade && lbu ci`
2021-06-11 22:39:21 <quadrat> brendonjmb lbu commit -d?
2021-06-11 22:39:57 <brendonjmb> I didn't use `-d` this time
2021-06-11 22:40:15 <brendonjmb> Not for any particular reason...
2021-06-11 22:46:02 <brendonjmb> I tried it, and it didn't result in a newer kernel version
2021-06-11 22:47:28 <quadrat> sorry, then no idea, back in the day it worked for me, and the wiki still says lbu commit -d should be enough
2021-06-11 22:48:22 <brendonjmb> This article (from 2017) says the RPi diskless install just requires a "reinstall" to upgrade https://dhole.github.io/post/raspberry_pi_alpine_upgrade/
2021-06-11 22:48:29 <brendonjmb> Thanks anyway
2021-06-11 22:50:07 <quadrat> https://wiki.alpinelinux.org/wiki/Raspberry_Pi <- here it says, the 2-3 commands are enough
2021-06-11 22:50:36 <liske> in diskless mode the kernel is *not* handled by apk
2021-06-11 22:51:55 <liske> reinstall manual or using setup-bootable and keeping the apkovl should work
2021-06-11 22:52:17 <brendonjmb> Thanks for the pointer; I'll try it
2021-06-11 23:16:48 <larzen> Can anyone tell me if openjdk 17 is somehow installable on alpine?
2021-06-11 23:30:10 <larzen> I found it here: https://pkgs.alpinelinux.org/package/edge/testing/x86_64/openjdk16
2021-06-11 23:30:22 <larzen> I just don't know how to add it so that I can apk add
2021-06-12 00:26:17 <thebtm> I have successfully upgraded from Alpine 3.5.3 to 3.13.5 in one go with the RPi. There was a few got ya's too since the wget was unable to do the cert checks. 
2021-06-12 00:27:01 <thebtm> I also had to make changes to the steps shared in the Raspberry Pi wiki page for the 3.x to 3.13 upgrade.
2021-06-12 00:28:54 <thebtm> it was based on https://wiki.alpinelinux.org/wiki/Upgrading_Alpine
2021-06-12 00:48:18 <thebtm> What are the steps for adding to the Wiki?
2021-06-12 00:52:02 <zcrayfish> thebtm: create an account, login, and edit or create a new page. (see also: https://wiki.alpinelinux.org/wiki/Help:Editing )
2021-06-12 00:53:34 <thebtm> I am guessing markdown doesn't work on the Wiki
2021-06-12 00:53:51 <zcrayfish> No, the wiki software is MediaWiki, it has its own syntax.
2021-06-12 00:54:18 <zcrayfish> more on that here: https://www.mediawiki.org/wiki/Help:Formatting
2021-06-12 00:57:24 <thebtm> it looks like it could take html, I will get the md converted to html then clean up the html ^^
2021-06-12 01:06:56 <invoked> thebtm: is your rpi diskless install?
2021-06-12 01:10:25 <thebtm> Yes
2021-06-12 01:11:01 <thebtm> I will put my half made MoP up on gitlab in a minute to share for others to review and hopefully add to the wiki.
2021-06-12 01:13:13 <invoked> speaking of gitlab, there were a bunch of vulns with that this week, in case you don't know
2021-06-12 01:17:04 <thebtm> I will get my private gitlab server updated then. thank you for letting me know.
2021-06-12 01:17:28 <thebtm> Here is my rough draft for upgrading Alpine on RPi Diskless - https://gitlab.com/thebtm/alpine-linux-notes/-/blob/main/Raspberry-Pi-Diskless-Upgrade.md
2021-06-12 01:18:52 <thebtm> At this point, only the arm v7 has been tested, the aarch64 is untested but I suspect its really not that much different.
2021-06-12 01:27:03 <thebtm> I took peices of multiple Wiki pages to figure out those steps for the Diskless upgrade from 3.x to 3.13.5
2021-06-12 01:35:48 <brendonjmb> I think I was successfully upgrade my own diskless RPi install, so I will read over this guide and see how it compares to what I did.
2021-06-12 01:43:00 <thebtm> Awesome thank you. if you find any issues, could you open an issue on gitlab? I have to reboot my laptop to windows for the Weekly friday night hangout with my friends (covid times made us improvise) and I only us IRC on linux. 
2021-06-12 01:43:33 <thebtm> We hangout in Destiny 2.
2021-06-12 01:45:17 <thebtm> Have a good evening.
2021-06-12 01:55:56 <invoked> it might be frowned upon, but i only run my rpis sys mode
2021-06-12 02:04:40 <brendonjmb> How does it work? I just went with the recommended diskless to begin with, but given that I had to do some symlinking to get Docker images to persist, I am not sure if it was worth it.
2021-06-12 02:05:36 <brendonjmb> Although I do like being able to mess around and then just restart and go back to the last checkpoint
2021-06-12 02:31:47 <invoked> brendonjmb: works well.  diskless supports the "this is only going to do 1 or 2 things and that will never change" model, whereas sys mode is helpful for heavier installs, development, testing, et al
2021-06-12 02:32:42 <invoked> but the latter case is not something alpine is keen on supporting
2021-06-12 02:33:34 <invoked> eg the 'FORCE_BOOTFS=1 setup-disk' situation
2021-06-12 02:34:45 <invoked> but i'm probably not aware of the arguments/reasons against it.
2021-06-12 06:01:20 <LACampbell> can alpine packages rely on older versions of other packages?
2021-06-12 08:21:57 <kasper81> hi Ariadne, just an fyi, after some hit-and-trial i have found a workaround for usbmuxd issue someone was having with their iPhone in this channel few days ago: https://gitlab.alpinelinux.org/alpine/aports/-/issues/12748. would be nice if a fix is found and we could avoid this workaround. (i personally have no prior experience with udev)
2021-06-12 17:18:32 <wener> hi, using grub, what is the proper way to set rootfs uuid ?
2021-06-12 17:22:09 <lopid> root=UUID=<uuid> iirc
2021-06-12 17:23:03 <lopid> in GRUB_CMDLINE_LINUX_DEFAULT
2021-06-12 17:28:17 <lopid> although https://github.com/torvalds/linux/blob/master/init/do_mounts.c#L255 says it's PARTUUID
2021-06-12 17:32:40 <wener> found GRUB_DEVICE_UUID in /etc/grub.d/10_linux, kind confuesed
2021-06-12 17:34:14 <PureTryOut> Pretty sure it's UUID
2021-06-12 17:36:12 <lopid> i'd go with that
2021-06-12 17:48:56 <wener> uuid in /boot/grub/grub.cfg is generated, set GRUB_DEVICE_UUID in /etc/default/grub generate correct root uuid, thx
2021-06-12 18:10:38 <wener> finally makes the grub+luks works
2021-06-12 19:07:57 <figbert> How do people manage Docker and firewalls with Awall?
2021-06-12 19:09:17 <figbert> Trying to make it so that ports I bind in Docker are only accessible if I authorize them in my firewall
2021-06-12 20:01:59 <quadrat> figbert just disable the docker iptables interaction in daemon.json?
2021-06-12 20:05:43 <Entroacceptor> 
ACTION tells docker to only bind the ports I want

2021-06-12 20:08:02 <quadrat> or that :)
2021-06-12 20:21:04 <endigma> docker has more or less a built in firewall
2021-06-12 20:21:14 <endigma> you can bind on localhost: i think
2021-06-12 20:21:23 <endigma> so it's not available to other network devices
2021-06-12 20:21:38 <endigma> you can do ports: to keep it internal to the docker net
2021-06-12 20:22:13 <endigma> aside from that either disable docker iptables or do your firewalling on your router
2021-06-12 23:38:21 <figbert> quadrat: then I can't access the internet at all from containers (bit of a deal breaker)
2021-06-12 23:39:40 <figbert> I think Entroacceptor and endigma take the most common approach, which is just binding the proper ports – seems like that's what I'll do as well
2021-06-13 01:09:05 <LACampbell> figbert: I just use my cloud providers firewall. though I run alpine straight on the server, not in a container.
2021-06-13 03:13:51 <figbert> LACampbell: good idea, i'll check that out as well
2021-06-13 03:16:27 <figbert> Also planning to run Alpine on the server itself, but I use Docker pretty heavily as well
2021-06-13 09:20:35 <danieli> is awall decent, and is there a good reason to run it? I never delved into that and stuck to managing iptables directly instead
2021-06-13 09:23:02 <clandmeter> the author recently questioned its existence due to nftables. 
2021-06-13 09:23:26 <clandmeter> i am probably not  the right person to ask, just passing the msg :)
2021-06-13 09:29:29 <danieli> clandmeter: fair enough, it was pure curiosity from my side
2021-06-13 13:21:20 <CcxWrk> Which package do I need for C++ debug symbols?
2021-06-13 13:36:53 <Hello71> huh?
2021-06-13 13:38:07 <Hello71> by definition, you need debug symbols for every package you're debugging
2021-06-13 13:38:33 <Hello71> it would defeat the purpose of having separate packages to jam everything into one megapackage
2021-06-13 13:39:36 <trylletre> theres *-dbg packages for.. quite a lot of things if youre looking for one specific one
2021-06-13 13:40:30 <trylletre> musl-dbg, git-dbg etc, they should have the symbols you want
2021-06-13 14:10:30 <CcxWrk> I mean libstdc++ specifically, that is not in musl-dev as musl is libc
2021-06-13 15:02:06 <Entelin> ok i've got an odd issue... I'm running 3.13.5 aarch64 on aws, launched from the alpine website. Things have been working fine for a month or so, but I rebooted it (with reboot from the shell) and it would not come back up. After some digging I was able to get it back up by launching a new instance, then copying /boot from that to mine. Then I ran apk fix linux-virt, it still booted sucessfully. However it would not load 
2021-06-13 15:02:06 <Entelin> existing module see: https://pastebin.ubuntu.com/p/SNHQ2WRD6y/   and then after another reboot it now doesn't boot again...
2021-06-13 15:02:28 <Entelin> any thoughts / suggestions?
2021-06-13 15:03:38 <Entelin> i'm about to try switching to an x86_64 instance
2021-06-13 15:36:11 <Entelin> fired up a brand new x86 instance, updated it, rebooted, fails to boot
2021-06-13 15:36:21 <Entelin> so something is broken
2021-06-13 15:48:42 <Hello71> stl is templates, so it would actually need to be part of the package
2021-06-13 15:48:54 <Hello71> my mistake
2021-06-13 15:49:07 <Hello71> wait, no, i got it double-backwards
2021-06-13 15:49:26 <Hello71> i was right in the first place
2021-06-13 15:49:56 <Hello71> when you do std::vector<int> it makes a new copy of all the std::vector functions specialized for int
2021-06-13 15:58:27 <feuerteufel> Die haben eine Kalibrierung mit drin, und dass kann Tasmota noch nicht hab ich gerade gelesen
2021-06-13 15:58:58 <ikke> Ich glaub nicht das dis has etwas zu machen mit Alpine Linux
2021-06-13 15:59:00 <feuerteufel> Ups, sorry wrong channel
2021-06-13 15:59:03 <ikke> :P
2021-06-13 15:59:10 <thatvillagerguy[m]> xD
2021-06-13 15:59:20 <feuerteufel> sorry again
2021-06-13 16:16:20 <Entelin> anyone using alpine on aws ec2?
2021-06-13 18:38:35 <ZeroBeholder> Hello. Mostly new to alpine Linux... I've followed tutorials in the past, but now is my first foray into doing something myself.
2021-06-13 18:40:44 <ZeroBeholder> I want alpine Linux to be the host OS of my KVM/QEMU VM situation. Should I use standard which has less than half a dozen packages (becuase adding QEMU will bring all the dependencies it needs). Or save my sanity and use the ISO image that blows all the way up to a whopping 150MB or so?
2021-06-13 18:41:09 <ikke> ZeroBeholder: standard image should suffice
2021-06-13 18:41:46 <ikke> extended is mostly for when you need extra packages to be able to install alpine in the first place
2021-06-13 18:42:57 <ZeroBeholder> Cool, thanks. I'll give it a shot. I've set the bar low for myself-- just get it set up so that it has similar functionality to the larger footprint Debian installs I have done. Thanks, I'll give it a try.
2021-06-13 18:43:44 <ZeroBeholder> I do have wired network available. My first time not using it as a Docker container base though.
2021-06-13 21:04:46 <eletrotupi> Hello! Is there a way to make abuild writing it's contents to another place other than ~/packages?
2021-06-13 21:04:53 <eletrotupi> s/writing/write/
2021-06-13 21:04:53 <alpine-meetbot> eletrotupi meant to say: Hello! Is there a way to make abuild write it's contents to another place other than ~/packages?
2021-06-13 21:05:43 <ikke> eletrotupi: check REPODEST in /etc/abuild.conf
2021-06-13 21:05:45 <eletrotupi> I've set the REPODIR on the abuild.conf, but I noticed that it's still writing to ~/packages 
2021-06-13 21:06:13 <ikke> It's REPODEST
2021-06-13 21:07:24 <eletrotupi> Oh, right
2021-06-13 21:07:42 <eletrotupi> Why the fuck I thought it was REPODIR? 
2021-06-13 21:07:49 <eletrotupi> Thanks ikke, always saving my skin 
2021-06-13 21:54:43 <MathGeniusJodie> aha! finally found where to verify my nickserv account
2021-06-13 21:58:26 <ikke> sadly it's necessary
2021-06-13 21:59:32 <MathGeniusJodie> anyway, hi, I'm jodie from the alpine linux discord server
2021-06-13 22:02:37 <holgersson> Hi jodie from discord! Welcome to the IRC world where you don't need browsers to chat :-D
2021-06-13 22:07:33 <MathGeniusJodie> that's right
2021-06-13 22:07:40 <MathGeniusJodie> weechat gang
2021-06-13 22:10:06 <holgersson> This was more meant as an OT rant about discord, but I should not start a flame war :-]
2021-06-13 22:30:13 <MathGeniusJodie> I mean, I only tolerate proprietary software slightly more when it's on the web
2021-06-13 23:22:58 <baconicsynergy> Hello Jodie, I just finished setting up nickserv properly too lmao
2021-06-14 05:36:35 <fission> what's discord?
2021-06-14 07:34:13 <danieli> it's pretty much Slack for zoomers
2021-06-14 07:55:04 <PureTryOut> Instead of Discord have a look at Matrix, stop supporting that shitty proprietary walled garden that is Discord
2021-06-14 07:56:41 <danieli> does it have voip yet?
2021-06-14 07:56:58 <danieli> er, it might be a better topic for #alpine-offtopic now that i think about it
2021-06-14 07:57:11 <clandmeter> yes please do :)
2021-06-14 08:55:50 <Turodek> it has voip (via jitsi), but not voice rooms
2021-06-14 08:56:02 <Turodek> well webrtc and jitsi
2021-06-14 11:37:38 <dngray[m]> matrix is already bridged to #alpine-linux's channels
2021-06-14 11:37:43 <dngray[m]> so that's really nice
2021-06-14 11:39:47 <rapha> hi all
2021-06-14 11:40:06 <rapha> sometimes "apk upgrade" will show something like "2 errors". how do i find out what those are?
2021-06-14 11:40:06 <Newbyte[m]> welcome!
2021-06-14 11:40:16 <ikke> rapha: apk fix
2021-06-14 11:41:01 <rapha> ouch. py3-setuptools trying to overwrite a bunch of stuff installed by python3. how to fix _that_?
2021-06-14 11:41:20 <ikke> what's in your /etc/apk/world?
2021-06-14 11:41:26 <rapha> same with pacman-bash-completion and bash-completion...
2021-06-14 11:41:44 <rapha> ikke: 101 lines of package names.
2021-06-14 11:41:53 <ikke> first: try apk upgrade --available
2021-06-14 11:42:37 <rapha> 2 errors; 1758 MiB in 534 packages
2021-06-14 11:49:13 <rapha> ikke: funny that only bash-completion and python3 are in /etc/apk/world ... the other two are not. so i guess what's in there takes precedence?
2021-06-14 11:49:45 <ikke> packages in world are things that are explicitly installed (ie, with apk add)
2021-06-14 11:49:50 <ikke> The rest are dependencies
2021-06-14 12:17:20 <Xianx> Hey! How would be the correct way to change the file limit displayed by "ulimit -m"? 
2021-06-14 12:17:46 <ikke> For users or services?
2021-06-14 12:18:14 <Xianx> Hmm.. Good question, what would be the difference?
2021-06-14 12:18:35 <ikke> 2 different solutions
2021-06-14 12:18:47 <Xianx> Ah. I would guess it's for a service.
2021-06-14 12:19:29 <ikke> then you can add rc_ulimit="-m <x>" to /etc/conf.d/<service>
2021-06-14 12:19:37 <Xianx> I figured out how to do an openrc script for the service i want. But i could not figure out how to change the ulimit in said service.
2021-06-14 12:19:43 <Xianx> Oh.
2021-06-14 12:19:53 <Xianx> So it just has to match the service name from init.d?
2021-06-14 12:19:58 <ikke> yes
2021-06-14 12:20:08 <Xianx> I'll try it out! Thanks a lot. :D
2021-06-14 12:20:52 <Xianx> I guess it can't go past the hard limit for files on the system. But i guess that one is rather high?
2021-06-14 12:21:05 <ikke> as root you can do that
2021-06-14 12:21:26 <ikke> well, there is the system wide hard limit, but that one is quite high
2021-06-14 12:22:28 <Xianx> Ah, right!
2021-06-14 12:22:38 <Xianx> Do i also need to make that file executable? Or is it just a conf file?
2021-06-14 12:23:39 <Xianx> Ah. Dumb question, none of the files in the dir are executables. :P
2021-06-14 12:24:26 <ikke> yeah, it will be sourced
2021-06-14 12:28:38 <rapha> ikke: okay, so the two that are causing problems are dependencies, got it. doesn't change the fact that they want to overwrite stuff from those that are directly installed. that seems like a grave issue :-/
2021-06-14 12:29:23 <ikke> rapha: py3-setuptools was built-in into python3, but has been separated
2021-06-14 12:31:39 <ikke> we need to tell apk that it's alright for this package to take over the files from python3
2021-06-14 12:32:09 <ikke> rapha: just for my information, what does apk del py3-setuptools return?
2021-06-14 12:41:06 <rapha> ikke: "World updated, but the following packages are not removed due to: glances"
2021-06-14 12:41:19 <rapha> Sorry, "... py3-setuptools: glances"
2021-06-14 12:41:26 <ikke> ok
2021-06-14 12:41:45 <ikke> doesn't help that packages started to have run-time dependencies on setuptools, rather than build-time dependecies
2021-06-14 12:42:00 <ikke> (that command is a quick way to find out what requires the package)
2021-06-14 12:42:55 <rapha> so basically, the glances package would need to get an update?
2021-06-14 12:43:23 <rapha> hmm yeah, removing glances and py3-setuptools, now i'm down to 1 error.
2021-06-14 12:43:43 <ikke> no, we need to fix the py3-setuptools package so that it can claim ownership of the files owned by python3 now
2021-06-14 12:44:52 <rapha> i see
2021-06-14 12:45:42 <rapha> as for the other one, ikke: https://gist.github.com/sixtyfive/f492a5e7b1782baae1ea4ba16a814c8e
2021-06-14 12:47:55 <ikke> similar story
2021-06-14 12:48:38 <rapha> i see
2021-06-14 12:48:50 <rapha> i will just leave all of these uninstalled for now, none are worth the hassle
2021-06-14 12:50:19 <Xianx> ikke: It worked. Just wanted to say thank you! 
2021-06-14 13:03:29 <ikke> Xianx: you're welcome :)
2021-06-14 13:03:52 <ikke> rapha: would you mind creating an issue for each of these on gitlab.alpinelinux.org?
2021-06-14 13:11:01 <rapha> no, not at all, let me see if i already have an account
2021-06-14 13:17:18 <rapha> somehow it was nicer when everyone and their dog were on github instead of having their own gitlab instances ... at least everyone and their dog don't have their own bugzilla instances anymore, which was just torture
2021-06-14 13:22:58 <rapha> ikke: what's the project for these? base-layout?
2021-06-14 13:23:09 <ikke> aports
2021-06-14 13:23:12 <rapha> thx
2021-06-14 13:27:23 <rapha> done
2021-06-14 13:27:36 <ikke> thanks!
2021-06-14 13:27:46 <rapha> thank YOU for an awesome distro :)
2021-06-14 13:57:08 <nico[m]> For kubernetes I have to make /sys and /run shared (`mount --make-shared /sys`) - what would be the best way to persist this in Alpine? I don't think it can be added as a mount flag
2021-06-14 14:56:53 <MathGeniusJodie> hmmm, I might run weechat on a server that runs 24/7 and ssh into it
2021-06-14 14:57:12 <ikke> MathGeniusJodie: that's what I do :)
2021-06-14 14:58:02 <MathGeniusJodie> like that I can put my laptop to sleep and still receive messages
2021-06-14 14:59:16 <Turodek> you could use abduco to keep it running and then attach it to a tmux session via ssh
2021-06-14 14:59:57 <yevaud> or just use mosh.
2021-06-14 15:00:18 <yevaud> SSH over TCP sucks, terminal sync over UDP is flawless. 
2021-06-14 15:00:26 <ikke> I use mosh to tmux
2021-06-14 15:00:30 <programmerq> I run weechat in tmux, and then I connect to it remotely with a tool called glowing bear.
2021-06-14 15:02:00 <Turodek> mosh is basicly ssh
2021-06-14 15:02:35 <MathGeniusJodie> mosh is literally the btw I use mosh meme
2021-06-14 15:02:35 <ikke> it uses ssh to authenticate, but then it uses it's own protocol
2021-06-14 15:02:43 <ikke> over udp
2021-06-14 15:02:56 <yevaud> yeah, mosh is completely unlike SSH 
2021-06-14 15:03:17 <yevaud> it runs a terminal emulator on the remote, and then does sync between the client and the remote states 
2021-06-14 15:14:13 <MathGeniusJodie> I'm quite new to irc
2021-06-14 15:16:38 <ikke> Well, welcome :)
2021-06-14 15:16:58 <MathGeniusJodie> I gotta figure out the optimal setup
2021-06-14 16:09:25 <rapha> MathGeniusJodie: that's what most ppl i know do :)
2021-06-14 20:30:45 <danieli> the default zsh prompt seems to be a bit broken
2021-06-14 20:31:00 <danieli> the prompt is "rendered" as "\h:\w\$ " in the shell
2021-06-14 20:32:09 <caskd> oh yeah i can reproduce that
2021-06-14 20:32:17 <caskd> i never bothered to report it though ^^'
2021-06-14 21:05:56 <danieli> i wonder *why* it happens, it should work
2021-06-14 21:14:06 <caskd> probs a conflict with the ash profile or something
2021-06-14 21:14:23 <caskd> since that's not a zsh problem in itself
2021-06-14 21:29:04 <Sos> hi
2021-06-14 21:37:38 <MathGeniusJodie> hello
2021-06-14 21:43:23 <Sos> i jsut got a risc-v sbc
2021-06-14 21:43:30 <Sos> is there a way to run alpine on it?
2021-06-14 21:45:07 <clandmeter> Sos: yes
2021-06-14 21:45:13 <clandmeter> but you need to build your own kernel
2021-06-14 21:46:07 <Sos> clandmeter, do i need to be on Alpine host to do that?
2021-06-14 21:46:24 <clandmeter> kind of yes
2021-06-14 21:46:32 <clandmeter> not for kernel, but for initramfs
2021-06-14 21:47:14 <clandmeter> which sbc do you have?
2021-06-14 21:47:17 <Sos> nezha
2021-06-14 21:47:26 <Sos> it just came out
2021-06-14 21:47:44 <Sos> they have this wonky debian image but i have zero trust for it
2021-06-14 21:47:58 <Sos> nothing else seems to be booting and all the docs are chinese
2021-06-14 21:48:14 <clandmeter> :)
2021-06-14 21:48:29 <clandmeter> its early also for alpine
2021-06-14 21:48:35 <nangel> imho, libvirt+qemu will be your friend
2021-06-14 21:48:54 <clandmeter> hey nangel 
2021-06-14 21:48:58 <Sos> yeah i know what i was getting into
2021-06-14 21:49:06 <Sos> gonna be fun
2021-06-14 21:49:15 <PureTryOut> We're still building the repositories for it
2021-06-14 21:49:32 <clandmeter> main is  kind of compelte
2021-06-14 21:49:36 <nangel> hey clandmeter!  sorry I've been gone so long
2021-06-14 21:49:45 <clandmeter> yes, shame on you!
2021-06-14 21:49:47 <clandmeter> :)
2021-06-14 21:50:09 <PureTryOut> Oh that Nezha thing seems nice, RPi formfactor, $99 and up to 4K video output
2021-06-14 21:50:21 <clandmeter> yes its a nice sbc
2021-06-14 21:51:18 <clandmeter> Sos: you can run alpine on docker or in a chroot with qemu-user
2021-06-14 21:51:32 <clandmeter> or lxc if you like
2021-06-14 21:51:47 <clandmeter> qemu-user is not that bad
2021-06-14 21:53:11 <Sos> i'm on windows, tho
2021-06-14 21:53:19 <Sos> guess i can just virtualbox
2021-06-14 21:54:36 <clandmeter> WSL :)
2021-06-14 21:56:58 <nangel> :)   for my win10 environment I just paid the $ and got vmware.   But my (slower, less RAM) AlpineLinux box with qemu just "works better" for me.   YMMV
2021-06-14 22:03:58 <Sos> alright i look into it
2021-06-14 22:04:11 <Sos> guess there's no going around building the thing
2021-06-14 22:07:09 <clandmeter> im currently working on a virt kernel, but that wont help you much.
2021-06-14 23:13:25 <MathGeniusJodie> I set up a vpn to run weechat into
2021-06-14 23:14:13 <MathGeniusJodie> I will not quit this channel for the forseeable future
2021-06-14 23:31:23 <LACampbell> MathGeniusJodie: I run weechat on the same machine I run my website. I assume real sys admins think that's not great
2021-06-14 23:32:55 <MathGeniusJodie> eh, linux can run multiple processes 
2021-06-14 23:32:58 <MathGeniusJodie> why not?
2021-06-14 23:35:11 <LACampbell> I don't know. Maybe each should be its own docker instance with kubernetes or something. I have no idea how this works in 2021
2021-06-14 23:36:33 <trylletre> there is nothing wrong with multiple services on one box
2021-06-14 23:37:40 <quadrat> especially with something like an irc client, idk but I am pretty sure it doesn't interact with much on the system
2021-06-14 23:39:06 <MathGeniusJodie> it's funny, I run my irc client on a server, so I'm a client client of a client server
2021-06-14 23:40:15 <trylletre> you could also run a bouncer on the server and use your client locally to connect to that
2021-06-14 23:40:23 <trylletre> but really, no difference, whatever works for ya
2021-06-14 23:40:27 <MathGeniusJodie> I still don't understand what's wrong with the model of multiple users on the same workstation
2021-06-14 23:40:39 <MathGeniusJodie> so don't ask me about containers haha
2021-06-15 00:44:08 <fission> huh. this mosh thing sounds intriguing
2021-06-15 00:58:22 <MathGeniusJodie> it's ssh but slighly snappier
2021-06-15 00:58:39 <MathGeniusJodie> p cool project
2021-06-15 00:59:41 <zcrayfish> ssh is simply unusable over slow (e.g. 2G or 3G) connections without it.
2021-06-15 01:45:07 <LACampbell> fission: I highly recommend it. I'm about a 4 hour flight away from my server (NZ -> AU), and this way I can chat on SSH without noticeable lag. with ssh it's really frustrating
2021-06-15 02:08:49 <fission> LACampbell: cool, good to know
2021-06-15 02:33:24 <endigma> mosh has this issue where you can backspace away your prompt
2021-06-15 02:33:47 <endigma> if they fixed that specific thing along with maybe clientsiding shell completions for fish/zsh it would be perfect
2021-06-15 02:34:05 <endigma> fish currently breaks sometimes due to it's autocompletion showing method
2021-06-15 02:43:52 <zcrayfish> endigma: can't that be fixed with --predict=never ?
2021-06-15 02:44:48 <endigma> perhaps?
2021-06-15 02:45:00 <endigma> I've never had a connection slow enough to reliably reproduce it
2021-06-15 02:48:47 <LACampbell> it's not really the speed of the connection, it's the latency
2021-06-15 07:13:16 <danieli> to mosh it's all just text, it doesn't care about your prompt
2021-06-15 07:13:28 <danieli> and if it did, that would introduce some issues of its own
2021-06-15 07:42:23 <markand> I'm surprised to see that rtc-ds1307 seems to be missing (on linux-rpi4 at least)
2021-06-15 07:42:29 <markand> https://github.com/torvalds/linux/blob/master/drivers/rtc/rtc-ds1307.c
2021-06-15 07:42:33 <markand> seems to be quite old though
2021-06-15 07:43:25 <markand> oh, it's builtin rather than a module
2021-06-15 07:43:52 <markand> then I wonder why I don't get /dev/misc/rtc Oo
2021-06-15 11:43:25 <MY-R> oh sweet, weechat 3.2 will be in Alpine 3.14 :>
2021-06-15 11:44:39 <MY-R> markand: you mean /dev/rtc ?
2021-06-15 11:55:53 <markand> MY-R, don't know but hwclock complains about missing /dev/misc/rtc
2021-06-15 11:56:06 <markand> while we have a ds1307 controller on our raspberry pi hat
2021-06-15 12:06:53 <MY-R> markand: maybe it doesnt load it early enough? did you add correct "dtoverlay=i2c-rtc,<module>
2021-06-15 12:07:11 <markand> aaaah yes maybe, let me check that later :)
2021-06-15 12:07:13 <MY-R> markand: to your "/boot/usercfg.txt"
2021-06-15 12:09:11 <MY-R> in your case it would be "dtoverlay=i2c-rtc,ds1307"
2021-06-15 12:16:17 <markand> thanks
2021-06-15 12:24:56 <MY-R> markand: after reboot check if it is loaded early enough: dmesg | grep rtc
2021-06-15 12:25:23 <MY-R> so should notice something like that: [    0.759054] rtc-ds1307 1-0068: registered as rtc0
2021-06-15 14:20:49 <macc24> does alpine use runit or openrc as default?
2021-06-15 14:22:01 <MY-R> macc24: openrc default
2021-06-15 14:42:03 <ZeroBeholder> Hello, looking for the equivalent of /etc/apt/sources.list in alpine. Just attempting a "semi-automatic" setup script and am somewhat new to using alpine as the host OS. Always used containers that were mostly configured for me previously.
2021-06-15 14:43:03 <nangel> the list of repositories is in /etc/apk/repositories
2021-06-15 14:43:11 <ZeroBeholder> thanks
2021-06-15 14:46:22 <ZeroBeholder> Really excited to see if alpine makes for a better base for KVM/QEMU than Debian. (I love Debian, but best tool for the job... you know? The host OS just needs to run KVM/QEMU.) I suspect less RAM and CPU cycles will be used by alpine.
2021-06-15 14:51:48 <donoban> ey ZeroBeholder , if you want a minimal host for QEMU take a look to qemu-openrc
2021-06-15 14:52:03 <donoban> I was using libvirt and now I'm very haoppy with this
2021-06-15 16:06:56 <quarkyalice> Alpine should work really well, it uses very little RAM on a minimal install
2021-06-15 16:07:10 <quarkyalice> I haven't used Debian as a KVM host in years
2021-06-15 16:07:36 <quarkyalice> I use Gentoo on ARM as an ARM KVM host, although that's not as lightweight as Alpine
2021-06-15 16:44:29 <invoked> the alpine base is much cleaner to work with compared to something like debian
2021-06-15 16:44:59 <invoked> i don't want to have to wonder what's changed in base, and with alpine i don't have to.
2021-06-15 16:45:44 <invoked> (changes to system ssl aside)
2021-06-15 18:19:54 <ZeroBeholder> I will definitely give it a try-- I wasn't super stoked about libvirt since my old configs are... well, old. I'd have to bring them up to current format requirements.
2021-06-15 18:24:51 <wener> use alpine as libvirt host for two years, works great, the guest is also alpine, and the container/pod in guest is still alpine, 😄
2021-06-15 18:48:37 <MY-R> no more fail2ban package in 3.14?
2021-06-15 18:49:22 <ikke> 1ba59cfd185bd51b650b237133672f4e9ae87f06
2021-06-15 18:56:55 <MY-R> ikke: somehow package exist for armhf and mips64 in edge
2021-06-15 18:57:54 <MY-R> aha build date 2021-04-05 so before that commit... cant just skip tests like done for other packages?
2021-06-15 18:58:36 <ikke> It depends
2021-06-15 18:58:49 <ikke> if it's broken, skipping tests does not make it less broken
2021-06-15 19:00:12 <MY-R> dunno how those tests looks, there is too many of them for that package but with fail2ban is obvious that filters/actions not working well at all if using busybox-syslogd
2021-06-15 19:34:59 <lopid> perl 5.34 didn't get in :(
2021-06-15 19:35:23 <ikke> lopid: no, it was too late
2021-06-15 19:37:53 <MY-R> damn fail2ban, there is so many failed tests that no wonder nobody wanna touch it
2021-06-15 19:39:06 <lopid> i touch it
2021-06-15 19:39:13 <ikke> MY-R: right :)
2021-06-15 19:39:45 <MY-R> lopid: touch it like try ti fix it for Alpine :)
2021-06-15 19:41:07 <lopid> i recently installed it on 3.13, it was fine
2021-06-15 19:42:06 <MY-R> I will find something simpler to block access against ssh bots - isnt nice to see 10000 lines weird attemps to touch my sshd server :/
2021-06-15 19:42:45 <MY-R> lopid: I was using it in 3.13, was working fine only with different logger than busybox-syslog one
2021-06-15 19:43:10 <lopid> huh, i am actually using the busybox syslog there. what's wrong?
2021-06-15 19:43:11 <MY-R> lopid: in 3.14 package is not available anymore
2021-06-15 19:43:19 <lopid> oh… :(
2021-06-15 19:43:33 <MY-R> lopid: with default sshd actions/rules? (not talking those provided by alpine)
2021-06-15 19:43:44 <lopid> does that mean if i upgrade it'll get removed?
2021-06-15 19:43:50 <MY-R> lopid: yes
2021-06-15 19:43:55 <lopid> damn!
2021-06-15 19:44:22 <lopid> i need to simulate this upgrade, then
2021-06-15 19:44:42 <MY-R> I was using those official sshd rules which were working with different loggers just fine
2021-06-15 19:44:47 <lopid> with the default rules that came with the alpine package, i just changed the ports
2021-06-15 19:47:05 <MY-R> ye but those Alpine ones are very poor and my logs were full of other type of weird attemps to touch my sshd
2021-06-15 19:47:36 <lopid> i actually didn't notice many differences between them and the ones i was using in slackware
2021-06-15 19:47:46 <lopid> it's a shame update-conf's output isn't coloured
2021-06-15 19:49:24 <MY-R> lopid: update-conf is simple shell script, make a copy of it and add whatever you want there
2021-06-15 20:41:42 <MY-R> oook, thank you Alpine crew! my rpi4 is happy from 3.15 upgrade! :)
2021-06-15 20:42:48 <ikke> hmmm, time flies, didn't know it was december already
2021-06-15 20:43:32 <MY-R> ps is there somewhere something to track removed packages between releases? in my case fail2ban package wasnt removed during upgrade but only -openrc
2021-06-15 20:43:40 <MY-R> 3.14!
2021-06-15 20:44:35 <trylletre> a factorial on a float, brave
2021-06-15 20:44:45 <MY-R> ikke: remember remember, the 15th of December! ;)
2021-06-15 20:44:57 <ikke> :D
2021-06-15 20:45:05 <lopid> not just any float
2021-06-15 20:45:08 <ikke> PI
2021-06-15 20:45:24 <MY-R> rPI
2021-06-15 20:45:27 <ikke> π
2021-06-15 20:45:37 <lopid> mm, pie
2021-06-15 20:45:54 <ikke> []π
2021-06-15 21:38:23 <danieli> is there a package in alpine that creates /etc/skel or is that not used at all?
2021-06-15 21:39:19 <ikke> I guess shadow might use it
2021-06-15 21:39:25 <ikke> but it does not create it
2021-06-15 21:39:27 <ikke> afaik
2021-06-15 21:39:56 <danieli> so it generally isn't used in alpine at all?
2021-06-15 21:40:05 <danieli> if so, i'll have to find a way around that, i need it for a project
2021-06-15 21:41:00 <danieli> it does look like home directories are created without templates of any sort
2021-06-15 21:41:42 <ikke> yeah, bb adduser does not populate home dirs
2021-06-15 21:41:58 <ikke> but shadow useradd should
2021-06-15 21:42:21 <danieli> i sometimes forget if it's adduser or useradd, both will be confusing as heck but i'll live :P
2021-06-15 21:42:22 <danieli> thanks
2021-06-15 21:45:15 <danieli> found something - looks like you need to specify `-k $skel_dir' with bb adduser
2021-06-15 21:45:23 <ikke> aha, ok
2021-06-15 22:13:12 <SamantazFox> danieli: useradd is the original tool from shadow-utils, where adduser is the more friendly tool (that has more option) which was a debian tool, iirc?
2021-06-15 22:13:38 <danieli> SamantazFox: the reason i asked is that one of the tools come with alpine, but that's only the one provided by busybox
2021-06-15 22:13:58 <danieli> and installing shadow has introduced unforeseen issues in the past
2021-06-15 22:15:01 <SamantazFox> danieli: Yeah, that's confusing as hell.
2021-06-15 22:15:22 <SamantazFox> My order is 'useradd', than 'adduser'
2021-06-15 22:15:40 <ikke> adduser is simpler
2021-06-15 22:15:52 <danieli> it is, but it's also meant to be a minimal implementation of it
2021-06-15 22:16:00 <ikke> you don't get a wizard where you need to provide details you mostly don't care about anyway
2021-06-15 22:16:27 <danieli> oh yeah, screw that, who even uses GECOS anymore?
2021-06-15 22:22:50 <nbm> how do I read the error shown by apk upgrade? "1 error; 1205 MiB in xxx packages"
2021-06-15 22:23:15 <danieli> as far as i've understood you don't, you run "apk fix" and hope the magic smoke doesn't escape
2021-06-15 22:23:27 <nbm> ok thank you danieli
2021-06-15 22:24:30 <nbm> it works, just reinstalled libretls all by itself
2021-06-15 22:24:36 <danieli> neat
2021-06-15 22:24:48 <danieli> it might have been some transient issue
2021-06-15 22:25:40 <nbm> i got the error when changed the repos from 3.13 to edge, updated and upgraded and got that 
2021-06-15 22:28:15 <danieli> (tip, you can run apk -vv <command>)
2021-06-16 00:28:51 <twitch6> hi! has anyone installed virtualbox inside alpine linux? how involved is it - do I need to compile from source and patch the kernel?
2021-06-16 00:29:31 <twitch6> would I be better off using the open-vm-tools package?
2021-06-16 07:13:50 <danieli> are there any plans to possibly start over with the/an alpine wiki? certain pages are really old and rotten to the core, written in a highly non-objective way, or has grammar that is so poor it makes the text difficult to read
2021-06-16 07:20:22 <fission> I'd be curious about which ones you suppose to be "rotten to the core"
2021-06-16 07:20:35 <fission> (not disagreeing or agreeing, just curious)
2021-06-16 07:21:10 <danieli> ones that are outdated to the point where the contents are largely not applicable anymore
2021-06-16 07:21:16 <fission> for example...?
2021-06-16 07:23:09 <fcolista> danieli, there's alpine-docs channel where you can share your ideas to improve the wiki. For sure it needs tender care...but there's no many ppl willing to do that over the course of the time
2021-06-16 07:23:33 <fcolista> from time to time someone says "this docs needs to be updated" but actually then nobody does :)
2021-06-16 07:26:22 <danieli> fcolista: yup, i am aware and it wasn't meant as a pure complaint, i was more curious if 'starting over' could be a feasible idea
2021-06-16 07:26:57 <danieli> fission: https://wiki.alpinelinux.org/w/index.php?title=Special:AncientPages&limit=150 click pretty much any page here
2021-06-16 07:27:35 <fission> it's ok, if you don't want to name one
2021-06-16 07:28:05 <danieli> i have to look around to find a good example, and i'm in a meeting at the moment
2021-06-16 07:28:22 <fission> ah but you had time to come here and say they are rotten?
2021-06-16 07:28:22 <fission> I see
2021-06-16 07:30:49 <danieli> this is the one that prompted me to think about it https://wiki.alpinelinux.org/wiki/Setting_up_a_new_user
2021-06-16 07:32:36 <fission> hmm can anyone edit the wiki (after registration) ?
2021-06-16 07:34:14 <danieli> there was a manual process for a while, probably due to vandalism, not sure if that's still in place?
2021-06-16 07:34:37 <fission> I've never actually tried to edit/improve the wiki before
2021-06-16 07:34:59 <fission> I've also never added another user directly in alpine :) but that's besides the point
2021-06-16 07:35:14 <fcolista> danieli, yes I know that you were not complaint :) perhaps sharing some ideas on alpine-docs could be useful. 
2021-06-16 07:35:46 <danieli> fcolista: agreed, it's a better conversation for that channel if there's a separate SIG
2021-06-16 12:36:17 <markand> got some warnings while upgrading to 3.14 like
2021-06-16 12:36:17 <markand> ERROR: boost1.76-dev-1.76.0-r0: trying to overwrite usr/lib/libboost_wserialization.so owned by boost-dev-1.72.0-r6.
2021-06-16 12:36:38 <markand> (was doing apk upgrade -a after a sed s/3.13/3.14 in /etc/apk/repositories)
2021-06-16 13:31:31 <ikke> It needz a replaces=
2021-06-16 14:19:17 <markand> looks like mate has some compositing issues, moving windows make some black glitches at some point (on pi 4)
2021-06-16 14:19:23 <markand> not happening on 3.13 
2021-06-16 14:19:32 <markand> not a big issue though
2021-06-16 15:10:46 <invoked> how many people are running pi4 desktops i wonder? :)
2021-06-16 15:11:10 <hodapp> a fair number perhaps. it's surprisingly usable there
2021-06-16 15:32:40 <invoked> it would probably work well in my case also. i mainly run tui apps in tmux (under sway). i have a threadripper machine at home i only fire up for distcc
2021-06-16 15:40:10 <thebtm> The cost of a Pi4 vs a laptop running MS, I could get a few Pi4s and have multiple desktops intead of just one x86_64 system with the dreaded MS Windows on it.
2021-06-16 15:40:10 <thebtm> Now I admit I have an i7 tower for gamming but thats because Bungie will ban any linux players so I have to use windows :(
2021-06-16 16:03:56 <nbm> good morning, is normal that alpine usb wont boot in uefi mode?
2021-06-16 16:08:23 <invoked> rpis give me mixed feelings though. broadcom sbcs. everything is problematic but i'd feel better about broadcom if they opened up their router firmware, for starters.
2021-06-16 16:09:27 <invoked> also raspbian is not very good + some of the power issues with the pis over the years give me a bad impression
2021-06-16 16:11:52 <invoked> i'll still use them anyway :)
2021-06-16 16:18:34 <nico[m]> nbm: if you just dd it, yes. You'll need to create a uefi/vfat partition. See https://wiki.alpinelinux.org/wiki/Create_UEFI_boot_USB
2021-06-16 16:19:56 <nbm> thank you nico[m] 
2021-06-16 16:20:31 <nbm> i thought those steps were not needed anymore
2021-06-16 16:42:41 <thebtm> invoked: i have my RPi4 8GB running my gitlab server and its better at it then my NAS was :P
2021-06-16 16:52:45 <invoked> thebtm: i don't doubt it, but aspects of the hardware reality are still gross
2021-06-16 16:59:00 <thebtm> no different then modern laptops that are pretty much have everything soldered to the board. 
2021-06-16 16:59:18 <thebtm> or cellphones
2021-06-16 17:00:01 <invoked> i *can* coreboot some laptops
2021-06-16 17:00:12 <invoked> (and have) but sure, it's all pretty gross.
2021-06-16 17:01:38 <thebtm> Fair
2021-06-16 17:03:48 <invoked> where my only choice is firmware blobs from the manufacturer, in 2021, i just don't feel good using any of it
2021-06-16 17:04:56 <invoked> i had some illusion the raspberry pi foundation was working broadcom to a point where maybe we'd have an open device by now
2021-06-16 17:05:31 <invoked> that was several years ago, but i have no such illusions anymore :-/
2021-06-16 17:08:07 <thebtm> I guess I am more System Admin and Software Developer then Embedded Systems so I am enjoying what the Rpi can do. though the Rpi4 8Gb is my first pi. past RPis, I didn
2021-06-16 17:08:21 <thebtm> I didn't feel were there yet.
2021-06-16 17:09:14 <thebtm> Outside of Docker, only really started using Alpine 3 months ago :P
2021-06-16 17:09:23 <jvoisin> congratz
2021-06-16 17:09:45 <invoked> nice
2021-06-16 17:15:46 <trylletre> are there any rpi4-level boards that have open firmware? i havent really looked much
2021-06-16 17:18:45 <asdflkj> the FSF maintains a list IIUC
2021-06-16 17:20:15 <asdflkj> trylletre: here it is: https://www.fsf.org/resources/hw/single-board-computers
2021-06-16 17:25:46 <trylletre> thanks
2021-06-16 17:26:07 <asdflkj> np, yw
2021-06-16 17:30:01 <invoked> the rock64pro is a good choice.  cost is close, the 4gb rock64pro (board only) is roughly the same as a typical rpi4 kit.  so just a tad more expensive
2021-06-16 17:30:22 <invoked> s/rock64pro/rockpro64/
2021-06-16 17:30:22 <alpine-meetbot> invoked meant to say: the rockpro64 is a good choice.  cost is close, the 4gb rock64pro (board only) is roughly the same as a typical rpi4 kit.  so just a tad more expensive
2021-06-16 17:31:22 <invoked> the only problem there is that pine is always out of stock on everything.
2021-06-16 17:31:30 <invoked> so people end up ordering pi4's anyway... :-/
2021-06-16 17:31:47 <trylletre> yeah, that rockpro is using the rk3399, and theres multiple listed for that SoC
2021-06-16 17:31:55 <trylletre> they all seem around the same spec, the 6-core setup and 4gb of memory
2021-06-16 17:32:13 <trylletre> (various vendors not just pine)
2021-06-16 17:32:16 <trylletre> seems like a good choice
2021-06-16 18:24:19 <thebtm> Reading about these small board systems is very interesting. I spent the 10 years before getting into small boards dealing with HPE Rack Mount and Enclosures & blades. 
2021-06-16 18:24:56 <thebtm> and a few mini mainframe units.
2021-06-16 18:57:58 <c705> do we have selinux avaliable in alpine now? I see libselinux avaliable from the repo
2021-06-16 18:58:26 <ikke> There are userspace utilities
2021-06-16 18:58:48 <c705> but is it built ito the default lts kernel?
2021-06-16 18:59:30 <ikke> # CONFIG_SECURITY_SELINUX is not set
2021-06-16 18:59:53 <ikke> CONFIG_SECURITY_APPARMOR=y
2021-06-16 19:03:29 <c705> ok thanks
2021-06-16 20:17:13 <thebtm> Just wondering anybody might know why Alpine 3.13 docker container with --privileged would mount img file but 3.14 fails the mount? giving errors about "Failed to setup loop device"
2021-06-16 20:28:22 <invoked> omni: hello - any eta for tor-0.4.5.9
2021-06-16 20:30:08 <MathGeniusJodie> right now if you compile it yourself
2021-06-16 20:30:12 <MathGeniusJodie> :P
2021-06-16 20:30:27 <invoked> obviously
2021-06-16 20:30:44 <invoked> i wouldn't ping if there wasn't a security issue, tho
2021-06-16 20:36:16 <c705> invoked: if you have a cve, open an issue and it will be addressed
2021-06-16 20:37:07 <ikke> missing here https://security.alpinelinux.org/srcpkg/tor
2021-06-16 20:37:11 <ikke> not a known one yet?
2021-06-16 20:37:34 <invoked> cve is probably lagging, as they do
2021-06-16 20:37:56 <ikke> yup
2021-06-16 20:38:10 <invoked> https://gitweb.torproject.org/tor.git/tree/ChangeLog?h=tor-0.4.5.9
2021-06-16 20:39:09 <c705> there's a cve number right there
2021-06-16 20:39:52 <ikke> nvd does not know about it yet: https://nvd.nist.gov/vuln/detail/CVE-2021-34548
2021-06-16 20:40:46 <invoked> 
ACTION shrugs :)

2021-06-16 20:41:29 <invoked> i really don't know what the whole cve lag issue is about, it's frustrating, but not my wheelhouse
2021-06-16 20:42:06 <c705> you can still put the cve number in an issue and it will be fixed
2021-06-16 20:42:12 <ikke> https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2021-34548
2021-06-16 20:42:14 <c705> it's been assigned, just nothing disclosed yet
2021-06-16 20:42:14 <ikke> yeah
2021-06-16 20:42:22 <ikke> it's reserved on mitre
2021-06-16 20:42:27 <ikke> I guess someone should let them know it's public
2021-06-16 20:42:50 <c705> they know, they've shipped a fix
2021-06-16 20:43:04 <ikke> I mean mitre
2021-06-16 20:43:42 <ikke> when you request a CVE, it remains hidden until it's publically disclosed
2021-06-16 20:45:10 <invoked> upstream disclosure: https://blog.torproject.org/node/2041
2021-06-16 23:19:11 <kedap> Hi
2021-06-16 23:20:17 <MathGeniusJodie> hello
2021-06-16 23:25:43 <kedap> I have a little problem when doing the merge requests in the contributions of alpline linux, apparently pipeline failed in the libressl package dependencies 
2021-06-16 23:25:57 <kedap> I have no experience as it is the first package I try to upload lol
2021-06-16 23:26:57 <kedap> The line of error say: ERROR: libressl3.3-libtls-3.3.3-r0: trying to overwrite usr/lib/libtls.so.20 owned by libretls-3.3.3-r0. It is normal?
2021-06-17 01:00:46 <kedap> I fix the error changing libressl for openssl
2021-06-17 06:03:01 <omni> invoked seem to have left, but I hope for later today for 3.13, 3.14 and edge
2021-06-17 06:07:11 <omni> edge might take longer since I want to upgrade to the new stable release for tor, 0.4.6.5, and builds for s390x and aarch64 fail atm (and I have to run, !22403 if anyone is interested)
2021-06-17 07:50:05 <shyim> Hey, My builds are failing currently as mariadb has been updated. How can I install an older version of the mariadb package? Tried using =10.5.9 but it can't find any package
2021-06-17 07:57:04 <PureTryOut> Outdated packages are removed from the repositories if newer packages appear. You'll have to build it yourself
2021-06-17 07:59:49 <quadrat> how come 3.13 has 10.5.10 while 3.14 has 10.5.9 oO? (talking about mariadb)
2021-06-17 08:00:59 <PureTryOut> That is interesting, I guess the person that bumped the package in 3.13 forgot to do it in edge/3.14
2021-06-17 08:03:17 <shyim> Then I gonna cry :(. The newest mariadb is totally broken https://jira.mariadb.org/browse/MDEV-25672 :(
2021-06-17 09:00:10 <rapha> hi all
2021-06-17 09:01:11 <rapha> i'm trying to use dm-integrity (without encryption) ... the command is `integritysetup format --integrity sha256 /dev/sda` ... unfortunately it fails and then dmesg says "device-mapper: table: 253:2: integrity: unknown target type" ... what am i doing wrong?
2021-06-17 09:15:09 <rapha> hmm, and https://www.google.com/search?q="integritysetup"+"unknown+target+type" has a weirdly short list of resutls. only one legitimate one.
2021-06-17 10:39:52 <rapha> oh, i'm on 3.14.0 with 5.10.43-0-lts
2021-06-17 10:51:00 <rapha> wait, am i in the official channel now, or the community channel? oftc, so should be the official one, right...?
2021-06-17 10:51:15 <yyp> This is the official channel
2021-06-17 10:51:15 <ikke> rapha: this is the correct location
2021-06-17 10:51:53 <danieli> yup, +1, this is the only official channel other than the mailing list
2021-06-17 10:52:19 <ikke> source: https://alpinelinux.org/posts/Switching-to-OFTC.html
2021-06-17 10:55:44 <markand> MY-R, I've tried dtoverlay=i2c-rtc,ds1307 in the config, now I boot I get:
2021-06-17 10:55:51 <markand> [    1.876104] rtc-ds1307: probe of 1-0068 failed with error -121
2021-06-17 11:00:44 <markand> maybe our ds1307 is badly powered, but we have dozen of our cards already running (but on raspbian though)
2021-06-17 11:00:46 <markand> strange
2021-06-17 11:16:29 <Jenkler> Does anyone here have troubles updating from lxc 4.0.6 to 4.0.9. It does not add routes deafult anymore
2021-06-17 11:20:10 <Jenkler> lxc.net.0.ipv4.gateway = 10.0.0.1 does not seam to work 
2021-06-17 11:27:47 <rapha> thanks y'all
2021-06-17 11:28:04 <rapha> but noone has any idea about how to use `integritysetup` with Alpine? :)
2021-06-17 11:29:22 <rapha> (even tried -edge now, same problem)
2021-06-17 11:29:52 <rapha> oh, wait, without specifying the block size now it works, with edge
2021-06-17 11:31:33 <rapha> ah, no, there's a dm_integrity module now
2021-06-17 11:32:31 <rapha> let me try putting that into /etc/modules and then go with 5.11 again
2021-06-17 11:37:33 <rapha> er, 5.10 ... working now after putting dm_integrity into /etc/modules, funny
2021-06-17 11:41:56 <Jenkler> vethd5MpmN@eth0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state LOWERLAYERDOWN qlen 1000
2021-06-17 11:42:44 <Jenkler> ikke, any clue why veth is LOWERLAYERDOWN with lxc 4.0.9
2021-06-17 11:43:02 <Jenkler> no config change from 4.0.6
2021-06-17 11:47:05 <MY-R> markand: install package "i2c-tools" and check if is rly on 68 position with: i2cdetect -y 1
2021-06-17 11:58:38 <markand> no, nothing
2021-06-17 11:59:02 <markand> need to wait my colleague to come back from his holidays to see then :P
2021-06-17 12:05:35 <MY-R> markand: probably that "hat" using different stuff, I noticed I had to use also "enable_uart=1" but you should check specification of that hat before change anything
2021-06-17 12:07:21 <MY-R> my rtc ds3231 module is directly connected with gpio pins in rpi4
2021-06-17 12:18:56 <Jenkler> ncopa, is there any issues with the veth module in 3.14?
2021-06-17 12:19:13 <Jenkler> ncopa, that you know about?
2021-06-17 12:19:44 <Hello71> Shyim: have you considered postgres :p
2021-06-17 12:20:18 <ncopa> Jenkler: not that I am aware of
2021-06-17 12:20:57 <ncopa> Jenkler: only issues I have had with veth modules when I have upgraded kernel and forgotten to reboot
2021-06-17 12:26:02 <Jenkler> ncopa, ok somthing has changed in the config. Do you have a working config with veth and lxc? 
2021-06-17 12:26:21 <Jenkler> *something 
2021-06-17 12:44:57 <markand> I'm installing Alpine in a chroot on a headless machine where the target disk is /dev/sdb (I use minirootfs) to install alpine-base, openssh and such
2021-06-17 12:45:19 <markand> I just wonder how should I install the bootloader (since once I reboot the VPS, the real disk will be /dev/sda rather than /dev/sdb)
2021-06-17 12:45:41 <ikke> Use uuids
2021-06-17 12:46:22 <markand> and just a grub-install /dev/sdb (while being in the alpine chroot) ?
2021-06-17 12:46:33 <markand> (I can create a custom grub.cfg)
2021-06-17 13:00:21 <Jenkler> ncopa, found it "lxc.net.0.flags = up" must be after "lxc.net.0.type = veth" <-- that has changed 
2021-06-17 14:33:04 <tiotags> hello, why does alpine fail to connect to dhcp inside qemu ? it keeps saying udhcp packet with bad magic, ignoring
2021-06-17 14:39:56 <lopid> mine's ok…
2021-06-17 14:41:23 <ikke> Hmm, sounds simular to what clandmeter is facing 
2021-06-17 14:42:00 <clandmeter> Yup
2021-06-17 14:42:17 <clandmeter> tiotags: qemu user?
2021-06-17 14:42:44 <tiotags> ?
2021-06-17 14:43:19 <clandmeter> Which qemu
2021-06-17 14:43:29 <tiotags> you mean version ?
2021-06-17 14:43:35 <clandmeter> No
2021-06-17 14:43:45 <clandmeter> What are you trying?
2021-06-17 14:45:07 <tiotags> to install alpine, I don't know what to say
2021-06-17 14:45:59 <clandmeter> Are you using qemu-system?
2021-06-17 14:46:12 <clandmeter> Or how are you using qemu
2021-06-17 14:46:42 <tiotags> yes
2021-06-17 14:47:00 <tiotags> qemu-system-x86_64 -m 512 -hda hdd.img -cdrom alpine-virt-3.13.5-x86_64.iso
2021-06-17 14:47:27 <clandmeter> Interesting
2021-06-17 14:47:38 <tiotags> that should work, right ?
2021-06-17 14:47:45 <tiotags> I don't use qemu very often
2021-06-17 14:49:01 <tiotags> I've also tried 3.14 virt and standard
2021-06-17 14:50:46 <clandmeter> It should work as qemu will do some magic networking
2021-06-17 14:52:15 <tiotags> I like magic, computers are magic
2021-06-17 14:54:15 <tiotags> is it possible to install openrc inside a docker container ?
2021-06-17 14:55:10 <tiotags> I really want to test something and need openrc
2021-06-17 15:01:02 <clandmeter> tiotags: did you read https://wiki.qemu.org/Documentation/Networking
2021-06-17 15:09:12 <tiotags> don't I just want '-net user' ? I'm reading it and missing it
2021-06-17 15:54:54 <clandmeter> tiotags: which host type are you using?
2021-06-17 15:55:39 <tiotags> host type ?
2021-06-17 15:56:10 <clandmeter> your host is alpine?
2021-06-17 15:56:18 <tiotags> gentoo
2021-06-17 15:56:46 <tiotags> custom kernel
2021-06-17 16:00:14 <clandmeter> qemu-system-x86_64 -m 512 -cdrom https://dl-cdn.alpinelinux.org/alpine/v3.13/releases/x86_64/alpine-virt-3.13.5-x86_64.iso works for me
2021-06-17 16:00:24 <clandmeter> with qemu v5
2021-06-17 16:14:16 <tiotags> it still doesn't work, can it be my kernel ? I'm not very good at configuring kernels
2021-06-17 16:21:58 <MY-R> tiotags: if you dont feel good with kernel configurations then use gentoo-kernel (can reconfigure it whenever want) or if dont want compile and just check if everything is working fine then package: gentoo-kernel-bin
2021-06-17 16:22:47 <MY-R> tiotags: keep in mind that it will use "dractut" to generate initramfs (imho dracat working best and faster from other gentoo options)
2021-06-17 16:23:18 <MY-R> dracut*
2021-06-17 16:53:05 <Jenkler> ncopa, lxc-start seams strange to. Does it work to use -L to log console in background. it worked before but now i dont get any output at all 
2021-06-17 16:58:43 <clandmeter> Jenkler: what is your problem with lcx?
2021-06-17 16:58:43 <clandmeter> lxc
2021-06-17 17:08:14 <Jenkler> clandmeter, I cant get any logs anymore 
2021-06-17 17:08:26 <Jenkler> worked fine in 4.0.6
2021-06-17 17:08:40 <Jenkler> lxc-start -L does not work 
2021-06-17 17:09:52 <Jenkler> if i do lxc-start -F -L then it works but, i do not want it in the forground 
2021-06-17 17:16:19 <clandmeter> Never used that switch
2021-06-17 17:16:40 <clandmeter> It logs console output?
2021-06-17 17:18:24 <Jenkler> Yes
2021-06-17 17:18:45 <Jenkler> -L file/to/log.txt 
2021-06-17 17:19:43 <Jenkler> works when i echo to /dev/console in container but stdout does not log 
2021-06-17 17:51:45 <danieli> what is the current state / health of chromium in alpine?
2021-06-17 18:56:29 <Hello71> tiotags: what version of qemu
2021-06-17 18:56:47 <Hello71> also i guess what version of libslirp
2021-06-17 19:00:33 <tiotags> it seems like 6.0, must of missed the upgrade notice
2021-06-17 19:02:07 <tiotags> whois Hello71 
2021-06-17 19:02:12 <tiotags> oops
2021-06-17 19:02:36 <tiotags> do you have the same problem Hello71 ?
2021-06-17 19:02:40 <tiotags> am I not alone
2021-06-17 19:28:39 <Hello71> libslirp?
2021-06-17 20:48:30 <ncopa> danieli I use chromium daily. Works okish. Tab crashes occasionally 
2021-06-17 20:50:39 <invoked> 10 arb execute vulns recently
2021-06-17 20:51:12 <asdflkj> does `sudo apk add gcompat` make AppImages work identically to other distros?
2021-06-17 20:53:23 <invoked> not magically. no
2021-06-17 20:53:43 <invoked> but then i've not tried, because i don't run gui stuff on alpine
2021-06-17 20:57:23 <asdflkj> I want a distro to run dwm that can suspend to an external (USB) flash drive. I think void-linux would be my ideal distro, but it doesn't support 'diskless' mode AFAIK (I imagine most distros could be made to run diskless, but it would probably take a lot of expertise and tinkering to make most of them do it)
2021-06-17 20:58:31 <qman> I ran alpine on desktop/laptop for a while, but there's a lot of annoyances because most people don't use it for that
2021-06-17 20:58:58 <qman> I switched to artix for desktop use, though I still use alpine for servers, VMs, and docker 
2021-06-17 21:01:29 <qman> chromium worked well enough for me
2021-06-17 21:02:40 <asdflkj> it's seems like there are lots of annoyances for anything that runs in ram/diskless. I've used puppy linux for a long time, and it doesn't have normal multiuser (90% of it runs as root). ATM I'm on tiny core but it seems like most people don't use tc at all, there are pretty few packages, and about 1 not-particularly-expert person active on IRC (besides me)
2021-06-17 21:06:23 <invoked> statically linking desktop stuff isn't generally possible either
2021-06-17 21:06:37 <invoked> GL etc
2021-06-17 21:08:11 <asdflkj> wow, alpine doesn't have sudo by default? if I only use doas, will there be programs that expect sudo and have problems?
2021-06-17 21:08:28 <invoked> you can add sudo
2021-06-17 21:09:52 <asdflkj> I got that, but do I need to? I've been using an OpenBSD server with only doas
2021-06-17 21:10:22 <asdflkj> I guess anything that needs sudo *should* install it as a dep...
2021-06-17 21:10:44 <invoked> no, you don't need to, or you could use doas, but for a minimal distro like alpine it doesn't make sense to have sudo in base.
2021-06-17 21:11:48 <PureTryOut> sudo will indeed be pulled in by applications if it's required, but those are only few and far in between
2021-06-17 21:19:12 <invoked> bit of trivia about sudo: the sudoers format was funded by a payday loans company
2021-06-17 21:24:51 <ikke> That explains a lot
2021-06-17 21:33:42 <MY-R> :D
2021-06-17 21:54:34 <marnold> is there a way to debug apk's http connections
2021-06-17 21:55:39 <marnold> mine is being really slow but i can't figure out why
2021-06-17 21:58:57 <marnold> i suspect it's broken ipv6 at the isp level 
2021-06-17 22:00:24 <ikke> tcpdump
2021-06-17 22:00:44 <ikke> Did you verify the mirror you are using>?
2021-06-17 22:02:00 <MY-R> marnold: try with different mirror like ikke said, even for me main cdn-dl sometimes is weird slow when everything else working as fast possible
2021-06-17 22:04:21 <marnold> thanks
2021-06-17 22:52:24 <MathGeniusJodie> I know I'm a bit late, but I use doas on alpine and haven't had a single issue
2021-06-18 11:36:43 <Jenkler> clandmeter, the lxc -L stuff works diffent than before. But I modified my scripts and now it works fine ;) 
2021-06-18 11:39:01 <clandmeter> Jenkler: good
2021-06-18 11:39:09 <clandmeter> you are talking aobut lxc or lxd?
2021-06-18 11:39:21 <Jenkler> clandmeter, inded. Yes, lxc not lxd 
2021-06-18 17:07:06 <nakedpotato> hello!
2021-06-18 17:07:41 <nakedpotato> is anyone running ZFS?
2021-06-18 17:08:06 <nakedpotato> I was wondering what level you've set your zfs-import and zfs-mount services
2021-06-18 17:14:42 <frojnd> A little help pls. I do sudo serice docker start and then I do docker info but I get: ERROR: Cannot connect to the Docker daemon at unix:///var/run/docker.sock. Is the docker daemon running?
2021-06-18 17:15:22 <artok> frojnd: does your user belong to docker group?
2021-06-18 17:15:46 <frojnd> artok: yes: docker:x:108:falcon
2021-06-18 17:17:38 <frojnd> Can I see some more logs?
2021-06-18 17:17:44 <frojnd> After I restart/start docker service?
2021-06-18 17:18:15 <artok> you can ofcoz check if dockerd is running 
2021-06-18 17:18:35 <ikke> or try to manually start dockerd as root in the foreground it see what it's log output is
2021-06-18 17:19:36 <frojnd> ps aux | grep dockerd nothing returns
2021-06-18 17:20:30 <frojnd> FATA[2021-06-18T19:20:22.341601437+02:00] Your Linux kernel version 3.4.113 is not supported for running docker. Please upgrade your kernel to 3.10.0 or newer.
2021-06-18 17:20:31 <artok> so /var/log/docker.log ?
2021-06-18 17:20:32 <frojnd> :/
2021-06-18 17:20:41 <artok> woaaah
2021-06-18 17:20:52 <artok> there is some ancient kernel running
2021-06-18 17:20:58 <frojnd> And I thought I can run pihole on postmarketos on my moto g 1st gen :/
2021-06-18 17:21:02 <frojnd> Within docker
2021-06-18 17:21:35 <frojnd> Well I guess I must try to manually install pihole on alpinelinux 
2021-06-18 17:22:22 <frojnd> Too bad docker can't be run on old kernels 
2021-06-18 17:23:47 <artok> well it can, 3.10 is old
2021-06-18 17:23:51 <frojnd> Haha
2021-06-18 18:49:48 <nangel> FWIW, if you don't need the web interface of pihole, you can follow this: https://wiki.alpinelinux.org/wiki/Using_Unbound_as_an_Ad-blocker
2021-06-18 19:31:20 <frojnd> Yeah I don't need web interface
2021-06-18 19:58:39 <frojnd> Hm
2021-06-18 19:58:53 <frojnd> How can I test if this unbound is working correctly
2021-06-18 19:58:58 <frojnd> Unbound dns blocking
2021-06-18 20:01:21 <frojnd> Not sure what this means: Dnsmasq configuration
2021-06-18 20:01:32 <frojnd> https://wiki.alpinelinux.org/wiki/Using_Unbound_as_an_Ad-blocker#Dnsmasq_configuration
2021-06-18 20:05:16 <c705> its the configuration for dnsmasq
2021-06-18 20:05:55 <frojnd> Yeah but in this context I don't understand what should I change in /etc/resolv.conf in order for unbound blacklisting to start working
2021-06-18 20:09:40 <trylletre> it works before you reach that section
2021-06-18 20:10:02 <trylletre> that section is so the dhcp sends the unbound ip as the dns for the devices so they reach unbound and not some other dns
2021-06-18 20:10:12 <frojnd> ok
2021-06-18 20:10:22 <frojnd> Ok I forgot I'm doing this on the phone
2021-06-18 20:10:38 <nangel> frojnd: correct
2021-06-18 20:10:59 <frojnd> So phone has unbound setup should have
2021-06-18 20:11:07 <nangel> sorry... I'm in a meeting... trylletre - frojnd is correct
2021-06-18 20:12:14 <frojnd> So now all I have to do in router is set statc DNS to the IP of the phone?
2021-06-18 20:12:22 <frojnd> And I should see no ads on my laptop?
2021-06-18 20:13:13 <nangel> sorry... trylletre explained it correctly... don't type in IRC while in a meeting. 8(
2021-06-18 20:14:09 <nangel> as long as your laptop uses alpine-based unbound as its DNS resolver, you should see no adds.
2021-06-18 20:14:51 <frojnd> Oh I see... I thought I can get replacement for pi-hole
2021-06-18 20:15:05 <nangel> you do.
2021-06-18 20:15:25 <nangel> pihole is just dnsmasq with a very-long hosts file
2021-06-18 20:15:41 <frojnd> Yeah yeah ok so all I have to do now is: https://discourse.pi-hole.net/t/how-do-i-configure-my-devices-to-use-pi-hole-as-their-dns-server/245
2021-06-19 00:22:46 <MathGeniusJodie> lol, new polkit cve 
2021-06-19 01:30:32 <anjan> I am running alpine stable. Do I need to check for new updates (via apk upgrade -Ua) every week or can I just wait for the minor releases?
2021-06-19 01:30:48 <anjan> I want to make sure I get security updates
2021-06-19 01:31:19 <danieli> stable releases are intended to remain stable, e.g. no breaking changes, but security fixes can get backported
2021-06-19 01:33:12 <trylletre> stable doesnt mean there arent (non breaking) package updates, so you should still run that upgrade
2021-06-19 01:34:20 <danieli> yes - i would recommend regularly checking for updates
2021-06-19 01:34:27 <danieli> fwiw there is a security tracker
2021-06-19 04:58:43 <dngray[m]> hey i seem to be getting this error on my server when booting
2021-06-19 04:58:46 <dngray[m]> error: shim_lock protocol not found
2021-06-19 04:59:27 <dngray[m]> i use secureboot and made sure to sign /efi/boot/bootx64.efi and /boot/vmlinuz-lts
2021-06-19 04:59:31 <dngray[m]> normally this is enough
2021-06-19 05:02:21 <dngray[m]> ie
2021-06-19 05:02:22 <dngray[m]> sbsign --key DB.key --cert DB.crt --output /boot/bootx64.efi /boot/bootx64.efi
2021-06-19 05:02:22 <dngray[m]> sbsign --key DB.key --cert DB.crt --output /efi/boot/bootx64.efi /efi/boot/bootx64.efi
2021-06-19 05:03:13 <dngray[m]> rather... sbsign --key DB.key --cert DB.crt --output /boot/vmlinuz-lts /boot/vmlinuz-lts
2021-06-19 08:12:58 <frojnd> Hm..
2021-06-19 08:13:09 <frojnd> I think I setup unbound correctly on the phone
2021-06-19 08:13:56 <frojnd> I've entered custom dns under dhcp (ip of the phone which has alpine linux unbount setup) but when visiting web pages I still see ads
2021-06-19 08:15:30 <frojnd> https://imgur.com/a/b806BuL
2021-06-19 08:15:41 <frojnd> I followed this guide
2021-06-19 08:18:35 <frojnd> Let me try and renew lease
2021-06-19 08:22:37 <frojnd> Oh my still see ads
2021-06-19 08:23:03 <frojnd> Any ideas? I was hoping https://wiki.alpinelinux.org/wiki/Using_Unbound_as_an_Ad-blocker#Dnsmasq_configuration will replace need for pi-hole
2021-06-19 08:29:15 <caskd> does querying the zones you refuse work as expected?
2021-06-19 08:29:25 <caskd> for example when using DiG
2021-06-19 08:32:07 <frojnd> This is nslookup on alpine with unbound: https://termbin.com/w9v8
2021-06-19 08:33:20 <frojnd> And this is my local laptop: https://termbin.com/vqv3
2021-06-19 08:33:21 <caskd> does your DNS server run on your machine or on the 192.168.178.96 machine?
2021-06-19 08:33:29 <frojnd> So on alpine it works but on local machine it does not
2021-06-19 08:34:11 <frojnd> 192.168.178.96 is alpine with unbound 
2021-06-19 08:34:35 <caskd> both of these seem to respond with the same thing
2021-06-19 08:34:47 <frojnd> Hm
2021-06-19 08:35:03 <frojnd> https://wiki.alpinelinux.org/wiki/Using_Unbound_as_an_Ad-blocker#Dnsmasq_configuration
2021-06-19 08:35:11 <frojnd> Last step I didn't configure
2021-06-19 08:35:20 <frojnd> Dnsmasq configuration on 192.168.178.96
2021-06-19 08:36:21 <caskd> did you make unbound listen on the local network or just loopback on the alpine machine?
2021-06-19 08:36:30 <caskd> seems like that one can't even connect to it's own dns server
2021-06-19 08:38:28 <caskd> check if `ss -lu` shows anything listening on port 53
2021-06-19 08:38:35 <frojnd> caskd: this is my /etc/unbound/unbound.conf https://termbin.com/y3p7 all I did was add file with blacklised domains
2021-06-19 08:39:10 <frojnd> caskd ss is under what package? I seem to be missing it
2021-06-19 08:40:04 <caskd> iproute2-ss
2021-06-19 08:41:11 <frojnd> caskd: no 53 port there: https://termbin.com/eyaq
2021-06-19 08:41:43 <caskd> domain is port 53, but only listens on localhost
2021-06-19 08:42:24 <caskd> set `interface: 192.168.178.96` on the machine running unbound
2021-06-19 08:42:42 <caskd> that way it's binding on the local network and is accessible from all machines on it
2021-06-19 08:43:21 <frojnd> Ok. Where do I set `interface: 1921.168.178.961 ?
2021-06-19 08:43:55 <caskd> ideally just after the other examples at the top in the `server:` section
2021-06-19 08:44:09 <frojnd> Oh you mean unbound.conf ?
2021-06-19 08:44:13 <caskd> yes
2021-06-19 08:47:31 <frojnd> Ok now after issuing ss -lu I see 102.168.178.96:domain
2021-06-19 08:47:48 <frojnd> Woops 192.168.178.96:domain
2021-06-19 08:48:20 <frojnd> But I still see ads
2021-06-19 08:49:24 <caskd> try querying a domain in the blacklist.conf with `nslookup DOMAIN_HERE 192.168.178.96`
2021-06-19 08:50:19 <caskd> you should get a `** server can't find DOMAIN_HERE: REFUSED` if it is ready to be used
2021-06-19 08:50:40 <caskd> or 127.0.0.1 if it works differently
2021-06-19 08:50:50 <frojnd> Ok so it seems on local machine if I add ip of the server: https://termbin.com/27j8
2021-06-19 08:50:57 <frojnd> Domain gets refused
2021-06-19 08:51:17 <frojnd> Maybe I have to again renew dhcp lease
2021-06-19 08:51:26 <caskd> that is good, now you only have to set it in dnsmasq on the hosts that you want to use it
2021-06-19 08:51:43 <caskd> (or any other resolv.conf modifier)
2021-06-19 08:52:13 <frojnd> Yeah I thought after adding ip of the server running unbound to the dhcp section on router it would do
2021-06-19 08:52:36 <frojnd> But let me first tyr locally in resolv.conf
2021-06-19 08:52:40 <frojnd> Just for a quick check
2021-06-19 08:54:27 <frojnd> Hm
2021-06-19 08:56:02 <frojnd> This is my dnsmasq.conf on local laptop: https://termbin.com/69l3 and when doing just nslookup linker-network.com I get: REFUSED which seems like it's setup properly: https://termbin.com/86w7 yet I see ads
2021-06-19 08:57:12 <caskd> well, this setup won't fully protect you from ads
2021-06-19 08:57:32 <caskd> some ads need a cosmetic filter rather than a network filter
2021-06-19 08:57:38 <frojnd> Aha
2021-06-19 08:57:53 <frojnd> Can I see on unbound server how many hits it get?
2021-06-19 08:57:56 <caskd> but if you have resources coming from domains blocked, they should stop showing up
2021-06-19 08:57:59 <frojnd> So I can be sure it's working
2021-06-19 08:59:57 <caskd> yes, there's 2 ways
2021-06-19 09:00:09 <caskd> simple one is running `unbound-control stats_noreset`
2021-06-19 09:00:14 <caskd> or running a tcpump
2021-06-19 09:00:18 <caskd> tcpdump*
2021-06-19 09:00:51 <frojnd> Hm
2021-06-19 09:01:05 <frojnd> I see only zeroes except for time
2021-06-19 09:01:42 <frojnd> https://termbin.com/zh0w
2021-06-19 09:02:36 <caskd> well, if you get a refused that's good
2021-06-19 09:02:44 <caskd> not sure why those stats don't update
2021-06-19 09:04:34 <frojnd> http://www.milliondollarhomepage.com/
2021-06-19 09:04:52 <frojnd> This site has the most ads 
2021-06-19 09:05:01 <frojnd> Yet it won't refresh my stats and I see a million pixels there lol
2021-06-19 09:05:26 <frojnd> Oh wait
2021-06-19 09:05:26 <caskd> and your resolv.conf has the dns server set, right?
2021-06-19 09:05:34 <frojnd> Nah I setill see it
2021-06-19 09:05:59 <frojnd> caskd my resolv.conf looks like this: https://termbin.com/e5rc
2021-06-19 09:06:19 <caskd> that looks fine
2021-06-19 09:06:38 <frojnd> Ha
2021-06-19 09:06:56 <caskd> are the domains on the page even blocked in the blacklist.conf? 
2021-06-19 09:07:09 <frojnd> But when running resolvectl status I see this: https://termbin.com/6oc5
2021-06-19 09:07:41 <caskd> uhh
2021-06-19 09:07:41 <frojnd> Like current dns server is still the ip of the router
2021-06-19 09:07:51 <frojnd> and not the one I want it to be which is .96
2021-06-19 09:07:53 <caskd> well, that should not matter...
2021-06-19 09:07:59 <caskd> as long as resolv.conf is fine
2021-06-19 09:09:12 <caskd> could i see your blacklist.conf?
2021-06-19 09:09:43 <frojnd> caskd: yes: https://termbin.com/pb7e
2021-06-19 09:10:45 <caskd> is that the entire file? that seems incomplete
2021-06-19 09:13:26 <frojnd> Yeah not entire it seems tembin died
2021-06-19 09:13:28 <frojnd> or netecat
2021-06-19 09:13:41 <frojnd> Last line is local-zone: "linker-network.com" refuse
2021-06-19 09:14:03 <caskd> and you can't access that domain in your browser anymore, right?
2021-06-19 09:14:35 <frojnd> I can
2021-06-19 09:14:38 <frojnd> I just tired linker-network.com
2021-06-19 09:14:48 <caskd> O.o
2021-06-19 09:14:56 <caskd> i'm confused
2021-06-19 09:15:13 <caskd> so nslookup returns refused when you try `nslookup linker-network.com`, right?
2021-06-19 09:15:15 <frojnd> nslookup says refused yet when going with browser I can access
2021-06-19 09:15:19 <frojnd> caskd: yes
2021-06-19 09:15:40 <caskd> do you have any kind of proxy or DNS over HTTPS/TLS?
2021-06-19 09:16:06 <frojnd> https://termbin.com/ikfo
2021-06-19 09:16:12 <frojnd> No
2021-06-19 09:16:35 <frojnd> I think not
2021-06-19 09:16:49 <frojnd> But let me double check if I setup it long time agon on that computer
2021-06-19 09:17:00 <caskd> probably it is but you're not aware of it
2021-06-19 09:17:03 <caskd> since your system already uses your configured server
2021-06-19 09:17:08 <caskd> but your browser isn't
2021-06-19 09:19:03 <caskd> if you run firefox you can set `network.trr.mode` to 5
2021-06-19 09:19:07 <caskd> if it's not 0
2021-06-19 09:19:27 <frojnd> I see I have bind
2021-06-19 09:20:43 <frojnd> caskd: in firefox I have network.trr.,mode set to 0
2021-06-19 09:21:05 <caskd> and it still loads that page? (even after a restart to clean cache)
2021-06-19 09:21:47 <frojnd> caskd: restart of local computer?
2021-06-19 09:21:56 <caskd> no, just the firefox process
2021-06-19 09:22:04 <frojnd> Yeah, I started firefox after 1 month 
2021-06-19 09:22:08 <frojnd> And first access that page
2021-06-19 09:22:22 <caskd> huh...
2021-06-19 09:22:32 <frojnd> Let me also clean cache from dev tools
2021-06-19 09:23:08 <caskd> Let me see if we missed anything: you have set the resolv.conf on the machine you test from and nslookup 'domain' returns refused, right?
2021-06-19 09:23:08 <frojnd> Can access just fine
2021-06-19 09:23:18 <frojnd> caskd: correct
2021-06-19 09:23:40 <caskd> hmmm.... okay this is weird...
2021-06-19 09:23:45 <caskd> give me one sec
2021-06-19 09:23:47 <frojnd> But I also found out I have bind installed
2021-06-19 09:23:59 <caskd> that should do nothing as long as it's not started
2021-06-19 09:26:46 <caskd> could you try running this and loading that page? doas tcpdump -i any 'port 53' -w /tmp/capture.pcap
2021-06-19 09:26:55 <caskd> (use sudo instead)
2021-06-19 09:30:08 <frojnd> Ok
2021-06-19 09:30:14 <frojnd> Where should I upload pcap
2021-06-19 09:30:25 <caskd> whereever you want 
2021-06-19 09:30:54 <frojnd> I ctrl+c after page was loaded
2021-06-19 09:30:57 <frojnd> but file seems empty
2021-06-19 09:31:10 <frojnd> Almost empty
2021-06-19 09:31:21 <caskd> even if almost empty it should be fine
2021-06-19 09:31:29 <caskd> as long as not entirely empty
2021-06-19 09:34:56 <frojnd> https://ufile.io/k370esks
2021-06-19 09:36:02 <frojnd> Hope it's without ads for you, I just registered with that unknown site
2021-06-19 09:36:20 <caskd> you could use something like 0x0.st alternatively but it's fine
2021-06-19 09:36:39 <caskd> also from what i can see in the log, your machine queries your router instead of the server
2021-06-19 09:36:46 <caskd> or better said the browser does
2021-06-19 09:38:17 <frojnd> Ok this seem to be problem on my site
2021-06-19 09:38:30 <frojnd> I have to force renew dhcp lease on the router
2021-06-19 09:38:59 <frojnd> And/or find the way why my browser tries to use the router dns ip instead of the server
2021-06-19 09:39:09 <frojnd> Thank you very much for extensivelly halping me
2021-06-19 09:39:20 <caskd> no problem, hope you find the solution
2021-06-19 14:53:55 <qsx> hi
2021-06-19 14:54:21 <qsx> i’m trying to boot alpine-standard-3.14.0-ppc64le.iso, but it seems like there is no kernel on it
2021-06-19 14:55:02 <qsx> neither on the hfs+ nor the iso9660 partition
2021-06-19 15:26:21 <Milamber75> What package has the man pages for /etc/network/interfaces and ifup?
2021-06-19 15:31:07 <Hello71> https://pkgs.alpinelinux.org/contents?file=interfaces.*&path=%2Fusr%2Fshare%2Fman%2F*&name=&branch=edge
2021-06-19 15:34:20 <Milamber75> Hello71: Thank you.
2021-06-19 16:05:07 <ikke> qsx: mind opening an issue on gitlab.alpinelinux.org?
2021-06-19 16:16:53 <qsx> ikke: which project?
2021-06-19 16:17:37 <ikke> aports
2021-06-19 16:26:54 <qsx> https://gitlab.alpinelinux.org/alpine/aports/-/issues/12777
2021-06-19 16:29:11 <ikke> qsx: thanks\
2021-06-19 19:12:11 <redavoc> Hi people! I want to setup a virtualized Alpine install in VirtualBox, and need working Virtual Guest Tools/Additions in the VM. The base installation of the OS and Xfce worked fine, but I cannot get the additions to run. I've tried following the guide in the Wiki, but thats kinda confusing: the virtualbox-guest-modules-virt isnt available in Alpine>3.12, so I chose a 3.12 ISO instead of 3.14, but still got issues. 
2021-06-19 19:13:47 <redavoc> Is this even relevant if I use the "Virtual" ISO? I've tried several Alpine versions, packages, DEs today but wasn't able to find the right combination yet. Some maybe has a hint for me? Thanks in advance
2021-06-19 19:15:02 <Sheila> redavoc: aiui, the kernel modules are part of the kernel now.
2021-06-19 19:17:22 <redavoc> Sheila: alright, good to know. So, if I assume correctly, I only need to install the virtualbox-guest-additions pkg (theoretically)?
2021-06-19 19:17:38 <Sheila> on 3.13+, yeah
2021-06-19 19:17:57 <redavoc> cool
2021-06-19 19:39:32 <frojnd> caskd: here?
2021-06-19 19:39:41 <frojnd> It seems now all pages are feused for all the clients lol
2021-06-19 19:40:01 <Hello71> redavoc: although you shouldn't use virtualbox
2021-06-19 19:40:14 <frojnd> Wait no just for some clients
2021-06-19 19:40:20 <caskd> frojnd: yes i am here
2021-06-19 19:41:01 <caskd> do you have any access-control: definitions in your configuration?
2021-06-19 19:41:35 <frojnd> No nothing
2021-06-19 19:42:11 <caskd> the other clients are also on the 192.168.0.0/16 network right?
2021-06-19 19:42:24 <frojnd> Yes
2021-06-19 19:43:00 <caskd> hm
2021-06-19 19:43:10 <caskd> could you list the ones that fail and ones that don't?
2021-06-19 19:45:17 <frojnd> I see now client didn't actually connect to the wifi
2021-06-19 19:45:21 <frojnd> false alarm
2021-06-19 19:45:40 <caskd> oh, those happen way too often
2021-06-19 19:45:53 <frojnd> I must figure out why router won't serve custom dns with dhcp server
2021-06-19 19:46:07 <caskd> i generally just try to prove myself wrong until i can't at which point i'm sure it's not my fault
2021-06-19 19:46:15 <caskd> good troubleshooting tip
2021-06-19 19:46:24 <frojnd> Yeah that's a good one
2021-06-19 20:11:32 <frojnd> caskd: I changed something in unbound.conf that blocks everyting 
2021-06-19 20:11:51 <frojnd> unbound-control stats_noreset actually shows new queries
2021-06-20 02:29:45 <LACampbell> Hello71: why shouldn't someone use virtual box?
2021-06-20 02:36:55 <valerius> qemu/kvm is a better option
2021-06-20 02:40:11 <Hello71> TAINT_CRAP, oracle crap, proprietary locked features...
2021-06-20 02:40:17 <Hello71> and qemu is better on all platforms
2021-06-20 02:43:17 <LACampbell> I see. I have alpine in a hyper-v thing, but I've been to lazy to migrate off WSL
2021-06-20 02:45:20 <Sheila> vbox devs are also notoriously unhelpful when you hit weird scenarios like a time64 guest (this was before 5.6.0 landed).
2021-06-20 03:07:51 <hodapp> ugh, I gave up on VirtualBox after using it 8-9 years, and really never regretted it
2021-06-20 03:08:11 <hodapp> was just so tired of shit breaking with cryptic errors
2021-06-20 03:08:59 <valerius> not sure if anyone is aware or not, but Oracle is evil
2021-06-20 11:07:42 <markpash> Hi, does anyone here know if the alpine pi images will get a dhcp lease on boot before going through the install process?
2021-06-20 11:08:14 <markpash> and if there's an ssh server running?
2021-06-20 11:08:48 <markpash> I have an issue where I can't hook up my pi4 to a screen, but I want to install alpine on it.
2021-06-20 11:17:34 <donoban> markpash: I don't think that any alpine iso sets up network and ssh before setup-alpine
2021-06-20 11:18:35 <donoban> you can create an overlay with the configuration that you need
2021-06-20 11:19:00 <markpash> donoban: can you give a link on how to do that please?
2021-06-20 11:20:23 <markpash> I found something on the wiki: https://wiki.alpinelinux.org/wiki/Raspberry_Pi_-_Headless_Installation
2021-06-20 11:20:24 <donoban> https://wiki.alpinelinux.org/wiki/Alpine_local_backup
2021-06-20 11:20:48 <markpash> donoban: thanks!
2021-06-20 11:21:04 <donoban> but...
2021-06-20 11:21:31 <donoban> you need to create a bootable and writeble device
2021-06-20 11:23:26 <donoban> if do "dd if=.. of=...", the device will be read only and can't save the overlay image in it
2021-06-20 11:24:22 <donoban> you can use 'setup-bootable' https://wiki.alpinelinux.org/wiki/Alpine_setup_scripts#setup-bootable
2021-06-20 11:24:37 <markpash> cool, I'll go do some reading
2021-06-20 11:26:06 <donoban> here are the manual steps: https://wiki.alpinelinux.org/wiki/Create_a_Bootable_Device#Manually_copying_Alpine_files
2021-06-20 11:26:32 <donoban> so you can do it from other distribution
2021-06-20 12:21:25 <alanszlosek> Hi, all! I have alpine 3.13 on a Dell Mini 10v with an Intel 945GME chipset, and am having issues with sway and wayland. I've installed mesa-dri-intel but am only seeing a blank screen and no console output when I start sway. Is this an appropriate place to get help on those? 
2021-06-20 12:21:49 <alanszlosek> (er, sorry, I'm acrually running alpine 3.14)
2021-06-20 12:22:05 <ikke> yes, this is an appropriate place
2021-06-20 12:30:07 <alanszlosek> thank you! i tried again just now and i see it prints the sway version, some envvars, "Preparing Wayland server initialization", "User has no sessions", and then "Couldn't find an active session or a greeter session". the last 2 are in red text, so i assume they're errors.
2021-06-20 12:33:07 <alanszlosek> another discovery: if i run `swaymsg -t get_outputs` I see "Unable to retrieve socket path"
2021-06-20 12:43:12 <Hello71> from *in* sway?
2021-06-20 12:43:16 <Hello71> you can't run it outside sway
2021-06-20 12:43:33 <Hello71> you need to either add your user to video and input group, or install and configure elogind
2021-06-20 12:45:29 <alanszlosek> oh! i was trying to run it from outside .. trying to find the display names so i can set them in the config file. my user already is in the input and video groups
2021-06-20 12:46:02 <alanszlosek> (the reason i wanted output display names is because i thought sway might be defaulting to my VGA connector instead of the built-in screen)
2021-06-20 12:46:15 <alanszlosek> (so i wanted to force it to use laptop display in the config)
2021-06-20 12:51:11 <markand> since openrc logs everything to last console parameter, you can make a beautiful silent boot using command line like : console=/dev/ttyS0,115800,n8 console=/dev/null 
2021-06-20 12:51:32 <markand> so that everything gets on serial console and you'll reach your logging screen/tty without any message on the video output
2021-06-20 12:58:24 <alanszlosek> that's a good hint, but that's not my problem. i'm trying to get sway to show something on the screen. i see the messages i mentioned above, but nothing else. if i switch to another TTY and then back to TTY0 i see a blank screen, as if sway TRIED to init video but failed
2021-06-20 12:59:27 <alanszlosek> (clarification, when i switch back to the first tty where i started sway i see blank scren .... Ctrl+Alt F1)
2021-06-20 13:07:32 <alanszlosek> hmm, i think my brain isn't awake yet ... found a stupid mistake in my startsway.sh script that wasn't redirecting stderr correctly. with that fixed i'm seeing lots more debug and info messages now. so i'll come back later once i've investigated further
2021-06-20 13:08:13 <MathGeniusJodie> progress!
2021-06-20 13:08:18 <Hello71> you did sway 2>&1 >file?
2021-06-20 13:14:16 <alanszlosek> hehe, you nailed it. i had the order wrong ... once i changed it to ">~/sway.log 2>&1" it started logging everything correctly to file. :-) 
2021-06-20 17:14:22 <alanszlosek> i had a chance to look through the logs... this doesn't look good: "00:00:00.208 [ERROR] [wlr] [backend/headless/backend.c:142] Renderer doesn't support XRGB8888". and a bit later: "00:00:01.182 [ERROR] [wlr] [backend/drm/renderer.c:236] Renderer doesn't support format 0x34325258". what does that mean and does that rule out ever being able to use sway on this device?
2021-06-20 17:16:57 <Hello71> better to just post all logs
2021-06-20 17:21:26 <alanszlosek> makes sense, here you go: https://gist.github.com/alanszlosek/c868f13a190d9e3452429d76e4a1e466
2021-06-20 18:00:55 <Hello71> and you say it shows a black screen?
2021-06-20 19:02:13 <markand> hmmm, I've migrated my thinkcentre from Arch to Alpine 3.14, alsamixer/pulseaudio are able to list the HDMI audio output
2021-06-20 19:02:19 <markand> only the analog output is listed
2021-06-20 19:12:17 <markand> oh nvm, it looks like HDMI/analog and speaker is handled by the same card
2021-06-20 19:12:26 <markand> you have to fire up some pacmd to switch to HDMI sink
2021-06-20 19:16:24 <srfsh> Hi folks.  So, I'm following this guide ( https://wiki.alpinelinux.org/wiki/LVM_on_LUKS ) word by word for the EFI installation, but I keep getting this errors: https://0x0.st/-9O4.txt .  The thing is, this not my first try, and I keep getting this no matter what I do.  I verified the UUID of the disk and it matches, so I have no idea why it does this.  Can anyone assits me, please?
2021-06-20 19:17:18 <srfsh> This is 3.13.0 standard amd64, by the way.
2021-06-20 19:47:15 <MY-R> srfsh: you sure your cryptroot=UUID=XXXX-XXXX point to the partition which "blkid" command show it as TYPE="crypto_LUKS"?
2021-06-20 19:51:03 <MY-R> srfsh: is your /boot encrypted or not? in that "howto" BIOS/MBR way looks like unencrypted (dunno why...) where on GPT is encrypted
2021-06-20 19:52:35 <MY-R> srfsh: so if you used BIOS/MBR or have unencrypted /boot then do NOT put this in default/grub "GRUB_ENABLE_CRYPTODISK=y"
2021-06-20 19:52:44 <markand> don't know what happened with bluetooth on Linux those days, my 8bitdo takes an average of 20ms to respond
2021-06-20 19:52:49 <markand> totally unusable…
2021-06-20 20:06:06 <nbm> srfsh: root=UUID=LV_UUID_HERE rd.luks.uuid=LUKS_UUID_HERE
2021-06-20 20:08:34 <nbm> MY-R: rd.luks.uuid will work similar to cryptroot right?
2021-06-20 20:09:45 <MY-R> nbm: wrong distribution, rd.luks.uuid stuff is dracut thing, Alpine using mkinitfs
2021-06-20 20:11:22 <nbm> good to know
2021-06-20 20:11:22 <nbm> oh sorry, thought the same dracut options worked for mkinitfs
2021-06-20 20:12:22 <MY-R> I think he used bios/mbr way and then there is no reason to put in grub those all modules, same as GRUB_ENABLE_CRYPTODISK
2021-06-20 20:13:04 <raven523> could this be an issue with grub not supporting luksv2?
2021-06-20 20:13:32 <nbm> as far as I know, grub 2.06 has luks2 support
2021-06-20 20:13:37 <MY-R> grub support luks2 but /boot still have to be luks1 or just not encrypted
2021-06-20 20:14:24 <MY-R> and Im talking about 2.06 version
2021-06-20 20:14:57 <MY-R> they should explain better what luks2 support means for them in changelog because it isnt so clear like should
2021-06-20 20:19:28 <MY-R> or maybe /boot on luks2 not working with passphrase and support only keys... who knows, that I didnt check
2021-06-20 20:34:38 <proycon> Any idea why keyboard and mouse might not be working in Xorg on Alpine? (no input and freezes after a few seconds)... I used to have this in a VM and then I think it went away when installing xf86-input-keyboard / mouse , but those packages are deprecated and not in 3.14 anymore... and now I have the same issue on bare metal
2021-06-20 20:36:49 <MY-R> proycon: not working at all?
2021-06-20 20:37:14 <proycon> yep, no input at all from either in X11, works fine on console
2021-06-20 20:38:08 <proycon> and nothing in Xorg.0.log
2021-06-20 20:38:23 <proycon> well, nothing indicative of the problem :)
2021-06-20 20:38:34 <MY-R> proycon: if you are not using elogind, then add yourself to group "input" maybe and relogin and be sure you have installed "xf86-input-libinput" package
2021-06-20 20:39:10 <MY-R> proycon: did you run "setup-xorg-base" script?
2021-06-20 20:39:35 <proycon> xf86-input-libinput is install (and -evdev too), and I'm stuck in lightdm so not even past the login stage
2021-06-20 20:40:48 <MY-R> proycon: dbus service is started?
2021-06-20 20:40:59 <proycon> yep
2021-06-20 20:41:41 <MY-R> and nothing in logs?
2021-06-20 20:42:22 <proycon> no, unfortunately not
2021-06-20 20:42:34 <proycon> let me check the main log
2021-06-20 20:43:09 <MY-R> I would try first without any "dm" and just from startx
2021-06-20 20:44:14 <proycon> yeah, I'll give that a try
2021-06-20 20:44:39 <MY-R> proycon: if you done setup-xorg-base then is good to reboot after if you didnt
2021-06-20 20:44:48 <proycon> yep, I did
2021-06-20 20:45:37 <MY-R> so disable lightdm service and try first with startx way
2021-06-20 20:45:49 <proycon> just tried that, same thing... no input
2021-06-20 20:46:11 <MY-R> proycon: and your user is in groups like: audio, video, input?
2021-06-20 20:46:16 <proycon> yep
2021-06-20 20:46:36 <MY-R> what you starting?
2021-06-20 20:46:48 <MY-R> which wm?
2021-06-20 20:46:49 <proycon> openbox now
2021-06-20 20:47:26 <proycon> it really is some lower-level input thing..
2021-06-20 20:48:04 <proycon> I circumvented that last time by just installing xf86-input-keyboard +  mouse probably
2021-06-20 20:48:55 <MY-R> proycon: install "xinput" package and run xinput as a user which you using
2021-06-20 20:49:12 <proycon> oh, now Iu have some output in the console, hold on
2021-06-20 20:50:10 <proycon> hmm, probably my basic xsession output.. 
2021-06-20 20:51:31 <MY-R> try to not use any custom files, add openbox to .xinitrc and then startx
2021-06-20 20:52:26 <proycon> ok, started X session (freezes), but ran xinput over ssh:
2021-06-20 20:52:28 <proycon> $ xinput
2021-06-20 20:52:29 <proycon> ⎡ Virtual core pointer                     id=2 [master pointer  (3)]
2021-06-20 20:52:31 <proycon> ⎜   ↳ Virtual core XTEST pointer               id=4 [slave  pointer  (2)]
2021-06-20 20:52:33 <proycon> ⎣ Virtual core keyboard                    id=3 [master keyboard (2)]
2021-06-20 20:52:35 <proycon>     ↳ Virtual core XTEST keyboard              id=5 [slave  keyboard (3)]
2021-06-20 20:55:05 <MY-R> proycon: no xorg.log or something?
2021-06-20 20:56:04 <proycon> it did write another log somewhere else yeah, lets see
2021-06-20 20:57:00 <MY-R> maybe in ~/.local/share/xorg/Xorg.0.log
2021-06-20 20:57:30 <MY-R> there was some mess with rootless stuff in alpine, not always was working fine... dunno if is fixed
2021-06-20 20:57:38 <proycon> yep: https://download.anaproy.nl/Xorg.0.log
2021-06-20 21:03:25 <MY-R> looks like all load fine and then... terminated, wth :/
2021-06-20 21:03:41 <proycon> yeah, the terminated I did via ssh
2021-06-20 21:03:53 <proycon> because the system itself doesn't take input
2021-06-20 21:03:55 <proycon> it's quite puzzling
2021-06-20 21:05:20 <MY-R> proycon: you have .xinitrc file with "exec openbox" inside, right?
2021-06-20 21:06:22 <MY-R> or openbox-session?
2021-06-20 21:06:24 <proycon> dbus-launch openbox at the end even...
2021-06-20 21:06:53 <proycon> and some other stuff before it for small service, but I can try the most minimal setup to see
2021-06-20 21:06:54 <MY-R> just put "exec openbox-session"
2021-06-20 21:07:04 <MY-R> just with that single line
2021-06-20 21:07:55 <proycon> yep, no difference, as expected
2021-06-20 21:08:45 <MY-R> xorg log ending in same way like previous?
2021-06-20 21:09:00 <proycon> yeah
2021-06-20 21:09:31 <MY-R> probably some package is missing to move to other steps I guess
2021-06-20 21:10:48 <proycon> I'm also reading https://dev.alpinelinux.org/~clandmeter/other/forum.alpinelinux.org/forum/installation/mouse-and-keyboard-do-not-work-under-xorg.html now
2021-06-20 21:10:59 <proycon> already tried to "Disable wfb" option
2021-06-20 21:12:56 <MY-R> you got running udev and dbus
2021-06-20 21:14:02 <MY-R> proycon: have you got installed "xf86-input-evdev" ?
2021-06-20 21:15:28 <proycon> yep
2021-06-20 21:16:07 <proycon> let's try an extra reboot.. probably won't do much but who knows :)
2021-06-20 21:16:46 <MY-R> ye...
2021-06-20 21:17:48 <proycon> ooh!!
2021-06-20 21:17:51 <proycon> it works now!
2021-06-20 21:18:10 <proycon> the only thing I really changed was applying rc-update add udev-postmount default
2021-06-20 21:18:17 <proycon> I had udev but not udev-postmount
2021-06-20 21:18:21 <proycon> I wonder if that did it..
2021-06-20 21:18:39 <proycon> but I did rc-service start it manually after adding it..
2021-06-20 21:19:30 <proycon> thanks for the help in any case :)
2021-06-20 21:20:25 <MY-R> you said that done reboot after, those services should be set up by default
2021-06-20 21:21:03 <proycon> yeah, I did.. odd
2021-06-20 21:21:11 <MY-R> setup-xorg-base didnt run "setup-udev" ? ...
2021-06-20 21:21:33 <proycon> it might have something to do with me having installed a lot manually prior to setup-xorg-base
2021-06-20 21:21:40 <MY-R> installing udev should do the job
2021-06-20 21:22:30 <proycon> I was trying to capture the whole provising in a simple script: https://git.sr.ht/~proycon/alpine-desktop-dotfiles/tree/master/item/install.sh
2021-06-20 21:23:31 <MY-R> ahh, too many stuff will be set up as manualy installed with such a script
2021-06-20 21:24:49 <proycon> I better let it do setup-xorg-base then?
2021-06-20 21:25:26 <proycon> I kinda like the manual way cause then I know exactly what's running
2021-06-20 21:25:53 <MY-R> proycon: no reason to do, just read last line from that script setup-xorg-base
2021-06-20 21:26:15 <MY-R> cat /sbin/setup-xorg-base
2021-06-20 21:26:22 <proycon> was just on it yeah :)
2021-06-20 21:26:37 <proycon> and that's probably what I forgot then yeah
2021-06-20 21:27:10 <MY-R> you said you ran setup-xorg-base so ude should be set up correctly
2021-06-20 21:27:13 <MY-R> udev*
2021-06-20 21:27:32 <proycon> yeah, I did ran it after I found out things were not workig
2021-06-20 21:28:01 <proycon> I see the udev script starts two more services even, udev-trigger and udev-settle
2021-06-20 21:28:27 <MY-R> yep
2021-06-20 21:28:49 <MY-R> so now you will know why didnt have inputs! enjoy!
2021-06-20 21:28:52 <proycon> and I see why it didn't run correctly now: there was a  ! rc-service --quiet udev status  condition in setup-udev
2021-06-20 21:29:00 <proycon> which I already isntalled manually so that part was running :)
2021-06-20 21:29:06 <MY-R> ye
2021-06-20 21:29:23 <proycon> well, mystery solved
2021-06-20 21:29:29 <proycon> I'll adapt my script
2021-06-20 21:31:29 <MY-R> :)
2021-06-20 21:55:11 <alanszlosek> @Hello71 it shows the first ~16 lines of log output then stops updating. when i switch away from the TTY and back to it, the screen is blank
2021-06-20 21:55:37 <Hello71> does it work in other distros
2021-06-20 21:55:38 <Hello71> or windows
2021-06-20 21:56:40 <alanszlosek> the laptop? yes. i previously had debian with a desktop environment and it worked fine
2021-06-20 22:03:33 <MY-R> debian stable or testing?
2021-06-21 01:58:38 <srfsh> MY-R: I'm telling these on top of my head: sdb is the disk I'm trying to install alpine to.  I was following the UEFI guide there, so sdb1 is an EFI partition, and sdb2 is the encrypted luks partition.  I used the UUID of sdb2, so it must be the correct one (I'm gonna check it in a moment to verify nonetheless).  Also, my /boot is encrypted, and I have "GRUB_ENABLE_CRYPTODISK=y" set on my /etc/default/grub.
2021-06-21 01:58:49 <srfsh> raven523: I explicitly used --type luks1 as the guide said.
2021-06-21 01:59:15 <srfsh> Sorry folks I couldn't reply earlier as I was very tried and went to bed.  ^^
2021-06-21 02:25:31 <srfsh> Oh, it seems like I was using 3.13.5.  Sorry for that as well.  ^^
2021-06-21 02:26:53 <srfsh> This is how my setup looks like: https://0x0.st/-94Y.txt
2021-06-21 02:27:25 <srfsh> It contains information about the the disks, mount points and related configurations.
2021-06-21 02:28:39 <srfsh> Just to clearify again, I'm using 3.13.5 standard amd64 with UEFI+GPT+GRUB.
2021-06-21 02:28:58 <srfsh> clarify*
2021-06-21 09:24:30 <ignaloidas> Hi, I was wondering if alpine has anything like debian's configuration packages?
2021-06-21 09:28:28 <lopid> we'd have to know what you meant by that
2021-06-21 09:30:40 <ignaloidas> packages that replace the default configuration of some other packages so that you can manage your configuration as packages
2021-06-21 09:30:52 <ignaloidas> https://wiki.debian.org/ConfigPackages
2021-06-21 09:35:33 <lopid> https://wiki.alpinelinux.org/wiki/Alpine_local_backup ?
2021-06-21 09:42:33 <ignaloidas> you can't stack local backups? I'd have multiple config packages in debian as not all servers need all of the config
2021-06-21 09:44:51 <lopid> i don't know. must they be provided by a distro package?
2021-06-21 09:45:48 <markand> hmmm for some reasons at boot openrc keeps telling me that mount /boot is already mounted or busy (but it's perfectly mounted afterwards)
2021-06-21 09:46:12 <markand> localmount service isn't even enabled
2021-06-21 09:46:25 <markand> (which seems to be the culprit regarding the message format)
2021-06-21 10:02:32 <ignaloidas> lopid, it's easier that way, especially if you already got CI/CD setup for other packages
2021-06-21 12:25:37 <MY-R> srfsh: looks good to me, is it possible that you skip command in chroot like "grub-mkconfig -o /boot/grub/grub.cfg" ?
2021-06-21 12:26:50 <srfsh> MY-R: Nope.
2021-06-21 12:27:18 <MY-R> srfsh: can you send what you got in /boot/grub/grub.cfg?
2021-06-21 12:27:26 <srfsh> Sure thing.
2021-06-21 12:29:23 <MY-R> srfsh: and you didnt forget that command: cryptsetup luksAddKey /dev/sdb2 /mnt/crypto_keyfile.bin
2021-06-21 12:30:13 <srfsh> no, I didn't.
2021-06-21 12:30:35 <srfsh> It is there and it isn't empty.
2021-06-21 12:30:47 <srfsh> I'm pasting it in a moment (grub.cfg).
2021-06-21 12:31:12 <MY-R> ok
2021-06-21 12:31:22 <srfsh> https://0x0.st/-9yv.cfg
2021-06-21 12:39:31 <MY-R> srfsh: dunno, maybe in kernel are missing modules for that non default luks encryption from howto
2021-06-21 12:41:03 <srfsh> I can add more entries to my mkinitfs.conf if you want?
2021-06-21 12:41:28 <MY-R> Im checking what can add at all
2021-06-21 12:42:30 <srfsh> I remember I installed alpine with this setup before, but this time it gives me the forementioned errors above no matter what I do.  :(
2021-06-21 12:42:41 <srfsh> s,setup,guide,
2021-06-21 12:47:31 <minimal> MY-R: I've been looking into Grub 2.06 and LUKS2 /boot for a few days. Seems like its an issue with grub-mkconfig where it does not put the "insmod" and "cryptmount" entries into the created grub.cfg when using LUKS2. I'm currently working on a solution at which point I'll submit a MR to Alpine grub
2021-06-21 12:48:42 <srfsh> minimal: But I use luks1.
2021-06-21 12:49:21 <minimal> srfsh: I was refering to something MY-R said earlier about LUKS2 encrypted /boot
2021-06-21 12:49:42 <srfsh> I see.
2021-06-21 12:49:50 <minimal> LUKS1 should work fine, I have it working here.
2021-06-21 12:50:03 <srfsh> With the guide on the wiki?
2021-06-21 12:50:18 <MY-R> minimal: nice and it will just work if modules will be there?
2021-06-21 12:52:53 <minimal> MY-R: if you look at a grub.cfg for LUKS1 you'll see "insmod" entries to load modules such as "luks", "crypt_sha256" etc as well as "cryptmount=<UUID>". Those are not being added by grub-mkconfig when using LUKS2, so I'm investigating how to fix this in grub-mkconfig
2021-06-21 12:53:35 <minimal> srfsh: I have my own script for creating Alpine disk images and I recently added an option to use LUKS1 which is working. I'm currently trying to get LUKS2 working with it
2021-06-21 12:53:50 <MY-R> minimal: ye that what I noticed and assumed that if grub-mkconfig doesnt detect that then it doesnt support it at all, o rmaybe luks2 boot working with those special keys and not passphrase
2021-06-21 12:54:14 <srfsh> minimal: can you link it so I can try?  :)
2021-06-21 12:57:30 <MY-R> srfsh: I have my own howto with examples but dunno if it wont mess you up, done it time ago but it still working fine
2021-06-21 12:57:56 <MY-R> is just txt file and every command is after #
2021-06-21 12:58:30 <srfsh> I'm not entirely familiar with the commands, but I know what they do, so I can give it a try.  ^^
2021-06-21 12:59:30 <minimal> srfsh: sure, 1 minute for me to push the latest version to my git repo
2021-06-21 13:00:34 <MY-R> srfsh: http://sprunge.us/CaDhvl
2021-06-21 13:03:22 <srfsh> MY-R: If you don't need anything for me to test, can I try to follow these instructions and install it again?
2021-06-21 13:04:08 <MY-R> ye try it
2021-06-21 13:23:08 <srfsh> MY-R: I'm in the "setup-disk -m sys /mnt/" stange.  The only thing different is that I have a separate boot lvm partition.
2021-06-21 13:24:41 <MY-R> srfsh: in Alpine howot have no idea why making boot one at all, and encryption is more generic and support cpu acceleration
2021-06-21 13:26:07 <MY-R> srfsh: srfsh and when removing syslinux it will also remove blkid so there is point to add it back because that busybox blkid version is very poor...
2021-06-21 13:26:13 <srfsh> I would partition my root even more, but let's see where this goes first.  :)
2021-06-21 13:26:21 <MY-R> srfsh: ye
2021-06-21 13:26:33 <minimal> srfsh: Have a look at https://github.com/dermotbradley/create-alpine-disk-image
2021-06-21 13:26:34 <MY-R> after you will do whatever want if will work...
2021-06-21 13:26:56 <srfsh> MY-R: Thank you, I'll check it out if this doesn't do the trick.  :D
2021-06-21 13:27:03 <srfsh> err, wrong person.
2021-06-21 13:27:14 <srfsh> minimal: ^ this one was for you :)
2021-06-21 13:27:18 <minimal> srfsh: try a QEMU disk image using BIOS with encryption - that's what I've been testing recently
2021-06-21 13:27:36 <srfsh> Makes sense actually.
2021-06-21 13:28:30 <minimal> srfsh: I meant when using the script :-) I haven't yet testing UEFI with encryption in the script so I doubt that (yet) works
2021-06-21 13:28:42 <MY-R> I usually test install setups in qemu, ssh in to it and after dump history to file so can do it on bare metal
2021-06-21 13:29:58 <minimal> MY-R: my script is designed to build images for bare metal as well but that's not yet tested as working.....concentrating on finishing up some VM stuff before testing/verifying the physical images
2021-06-21 13:35:42 <MY-R> minimal: yeah, some people with rpi's will be happy too :P
2021-06-21 13:37:21 <minimal> MY-R: I've been using RPI disk images for more than a year but they were created using my previous Ansible playbook. It just takes a little time to copy the functionality into this script and test. Want to get LUKS finished first (and then there's LVM, and LVM-on-LUKS)
2021-06-21 13:38:17 <MY-R> :)
2021-06-21 13:38:48 <minimal> plus I have to start testing the various cloud providers as well.....
2021-06-21 13:39:41 <srfsh> MY-R: Rebooting.  ^^'
2021-06-21 13:39:47 <MY-R> Im not using lately lvm, just creating single luks partition on every disk and if need swap then creating swapfile or not creating at all and using zram, depends form machine
2021-06-21 13:40:32 <srfsh> Same error.  :P
2021-06-21 13:41:11 <minimal> MY-R: LVM is useful on a server if you have separate mounts for things like logs, home dirs, etc as you can increase the size of these individual mountpoints as required
2021-06-21 13:41:11 <MY-R> srfsh: you have uefi, you sure it start from your second disk?
2021-06-21 13:41:34 <srfsh> I mean, I only have one linux os.
2021-06-21 13:41:40 <srfsh> (at least grub)
2021-06-21 13:41:51 <srfsh> So, it is definetelly booting from the right disk.
2021-06-21 13:41:59 <srfsh> As for uefi, it must be.
2021-06-21 13:42:07 <srfsh> Let me really quickly check.
2021-06-21 13:42:07 <minimal> srfsh: what's your error?
2021-06-21 13:42:13 <MY-R> srfsh: you using sdB so what got on sdA ?
2021-06-21 13:42:23 <srfsh> sda is openbsd
2021-06-21 13:42:35 <srfsh> or usb disk, and sdc is openbsd
2021-06-21 13:42:40 <srfsh> you know what i mean
2021-06-21 13:43:03 <srfsh> minimal: I linked the error messages somewhere above, let me find it.
2021-06-21 13:43:10 <srfsh> It is quite long to type.  ^^'
2021-06-21 13:43:14 <MY-R> minimal: his error: https://0x0.st/-9O4.txt
2021-06-21 13:43:36 <srfsh> MY-R: oh, thank you.  ^^
2021-06-21 13:45:06 <srfsh> MY-R: legacy bios is disabled, so it only uses UEFI.
2021-06-21 13:45:18 <minimal> srfsh: LVM-on-LUKS or LUKS-on-LVM? The reference make me think its looking for a LUKS partition inside a LVM one
2021-06-21 13:45:42 <srfsh> The guide on the wiki and the one MY-R linked are both LUKS-on-LVM.
2021-06-21 13:46:28 <MY-R> it is LVM on LUKS
2021-06-21 13:46:35 <MY-R> both
2021-06-21 13:46:36 <srfsh> Oh, sorry.
2021-06-21 13:46:48 <MY-R> :) ye can go crazy with all that
2021-06-21 13:47:10 <srfsh> English is not my mother tongue.  :D
2021-06-21 13:47:44 <MY-R> like 80% of people here, dont worry :P
2021-06-21 13:47:58 <srfsh> hehe
2021-06-21 13:48:04 <minimal> srfsh: never tried LUKS-on-LVM. Not sure if I'd ever want to try that, LVM-on-LUKS makes more sense to me
2021-06-21 13:48:37 <srfsh> I've never seen LUKS-on-LVM.
2021-06-21 13:49:09 <minimal> srfsh: sorry, misread :-)
2021-06-21 13:49:37 <srfsh> yeah, it was my bad.  ^^'
2021-06-21 13:50:25 <minimal> so does the rescue mode have "blkid" or "lsblk" to see the disk layout?
2021-06-21 13:50:41 <srfsh> I tried both; the answer is "no".  :D
2021-06-21 13:50:45 <srfsh> tried both before*
2021-06-21 13:51:20 <srfsh> it has "ls", but I have no idea what are the printed values.
2021-06-21 13:51:32 <srfsh> assuming some not-abstracted layer of disks.
2021-06-21 13:51:41 <minimal> I guess the binaries are specified in /etc/mkinitfs/features.d/crypsetup.files (which would cause them to be copied into initramfs)
2021-06-21 13:51:48 <minimal> s/are/are not/
2021-06-21 13:51:48 <alpine-meetbot> minimal meant to say: I guess the binaries are not specified in /etc/mkinitfs/features.d/crypsetup.files (which would cause them to be copied into initramfs)
2021-06-21 13:52:41 <srfsh> Oh, the bot is not aware of separators other than "/".  :)
2021-06-21 13:53:34 <srfsh> Did you also mean to say "which would cause them *not* to be"?
2021-06-21 13:55:01 <minimal> I guess I was suggesting a way to make them available in rescue mode by adding references to lsblk/blkid in cryptsetup.files and regenerating the initramfs
2021-06-21 13:56:32 <MY-R> srfsh: does it ask you for passphrase first or going directly to grub rescue?
2021-06-21 13:56:42 <srfsh> MY-R: It asks me the password.
2021-06-21 13:57:32 <srfsh> s/asks/does ask/
2021-06-21 13:57:32 <alpine-meetbot> srfsh meant to say: MY-R: It does ask me the password.
2021-06-21 13:59:24 <minimal> srfsh: doesn't that means that the "cryptmount" entry in grub.cfg has the right UUID for the LUKS partition? So then the reference to the root device is the issue...
2021-06-21 14:00:14 <srfsh> I am not sure if I understand the first sentence fo yours.
2021-06-21 14:00:31 <MY-R> srfsh: or try in grub rescue: cryptomount (hd1,gpt2)
2021-06-21 14:00:35 <srfsh> You meant the wrong one?
2021-06-21 14:01:09 <MY-R> hd1 will be second disk right? and gpt2 is second partition so /dev/sdb2
2021-06-21 14:01:11 <minimal> Looking at a grub.cfg here I see the following 2 lines:
2021-06-21 14:01:12 <minimal> cryptomount -u 443d4ca41c7442a8b7b729014134d6ec
2021-06-21 14:01:28 <MY-R> Im always confused by grub disk names/numbers
2021-06-21 14:01:31 <minimal> that tells GRUB the UUID of the LUKS partition
2021-06-21 14:01:35 <minimal> then the next line:
2021-06-21 14:01:46 <minimal> set root='cryptouuid/443d4ca41c7442a8b7b729014134d6ec'
2021-06-21 14:01:47 <MY-R> minimal: ye that is why I want him to check device names and not uuid
2021-06-21 14:02:11 <minimal> that tells GRUB what the root is (so it can find the kernel and initramfs)
2021-06-21 14:03:00 <MY-R> ah ye, cryptmount should be already done and grub looking for a root partition
2021-06-21 14:03:04 <MY-R> blehhh
2021-06-21 14:03:04 <minimal> in my examples that a LUKS UUID as there's no LVM in that setup. In your case it seems its set to "lvmid/PQrFjF-" instead which is a LVM reference that it can't figure out
2021-06-21 14:05:09 <srfsh> MY-R: It tells me: error: no such partition.
2021-06-21 14:05:34 <minimal> so you see the GRUB menu at all after typing in the passphrase?
2021-06-21 14:05:39 <minimal> s/so/do/g
2021-06-21 14:05:39 <alpine-meetbot> minimal meant to say: do you see the GRUB menu at all after typing in the passphrase?
2021-06-21 14:06:18 <srfsh> I don't think so.
2021-06-21 14:09:45 <minimal> do you have a copy of the grub.cfg file?
2021-06-21 14:10:09 <srfsh> The old one (the one you viewed) is gone.
2021-06-21 14:10:53 <srfsh> I can link the new one (the one generated by MY-R's instructions), but it will take a moment.
2021-06-21 14:12:37 <MY-R> minimal: this was after first atempt and same error: https://0x0.st/-9yv.cfg
2021-06-21 14:14:47 <minimal> ok, so "cryptomount -u 94d9e669bd5845d89be711fc8bc7ed6e" is a reference to the UUID of the LUKS partition, if that was wrong the you wouldn't get past the password bit (blkid if available would let you check the UUID)
2021-06-21 14:15:26 <minimal> next line: "set root='lvmid/PQrFjF-floy-tTkt-qTii-Kdmc-j2YB-UJriY0/DHOpoz-Wovy-ptkS-w7af-Ofdg-16mX-7z697o'" is a reference to the root partition, that's the problem you're seeing
2021-06-21 14:16:09 <minimal> so the GRUB LVM module is loaded which is good ("insmod lvm")
2021-06-21 14:16:40 <srfsh> So, it decrypts the luks partition but can't access the lvm partition inside it?
2021-06-21 14:17:26 <minimal> further down you can see the root LVM partition's UUID (look at both lines beginning with "search --no-floppy --fs-uuid" for the number beginning with "45985"")
2021-06-21 14:18:40 <minimal> seems that GRUB doesn't know that the lvmid reference is for part with UUID beginning "45985"
2021-06-21 14:19:56 <minimal> srfsh: hang on, I'm wrong - the above is in the "00_header" included part of the file... I would expect to see sucg stuff there
2021-06-21 14:20:13 <minimal> the Alpine stuff is further down in the 10_linux included section (as usual)
2021-06-21 14:21:45 <minimal> srfsh: so where is the earlier stuff coming from?
2021-06-21 14:22:11 <srfsh> What earlier stuff?  :)
2021-06-21 14:24:35 <minimal> grub-mkconfig takes files in /etc/grub.d/ as templates. One of these files is 00_header and as can be seen from the grub.cfg you provided between the "### BEGIN /etc/grub.d/00_header ###" and "### END" sections there are "cryptomount -u" and "set root" entries as well as in the usual 10_linux section
2021-06-21 14:25:12 <minimal> I'm wondering how these earlier entries came to be in the grub.cfg
2021-06-21 14:26:45 <MY-R> ye, wth those aprts doing at begining
2021-06-21 14:26:49 <srfsh> It is not expected to have anything other than 10_linux section?
2021-06-21 14:26:50 <MY-R> parts*
2021-06-21 14:27:37 <MY-R> it trying do some magic before even get to the grub menu
2021-06-21 14:28:33 <MY-R> srfsh: better send current grub.cfg from that last installation if can
2021-06-21 14:28:37 <minimal> srfsh: there will be stuff in the 00_header section but I don't expect to see anything to do with cryptomount or "set root" there. I'm thinking that (assuming you're following that Wiki guide) that you don't set the GRUB_PRELOAD_MODULES entry in /etc/default/grub as that might prevent these entries appearing
2021-06-21 14:28:44 <srfsh> MY-R: on it.
2021-06-21 14:29:09 <MY-R> minimal: ye, in my "howto" I didnt add those modules so will see how grub generated config without it
2021-06-21 14:29:27 <srfsh> I think I added GRUB_PRELOAD_MODULES.
2021-06-21 14:29:33 <minimal> Grub shouldn't need the GRUB_PRELOAD_MODULES as it should work stuff out itself (MY-R: though it might be a way to get LUKS2 working, its something I have to test yet)
2021-06-21 14:29:42 <MY-R> minimal: exactly
2021-06-21 14:30:00 <minimal> srfsh: that Guide says to add it, I'm not convinced.
2021-06-21 14:30:01 <MY-R> minimal: I added modules manualy when tried with luks2 and it didnt work
2021-06-21 14:30:43 <MY-R> srfsh: so you didnt do exact with my howto
2021-06-21 14:30:52 <minimal> MY-R: though it setting that would workaround the LUKS2 problem you'd have to remember to add "luks2" not "luks" to the list - there are separate modules for v1 and v2
2021-06-21 14:31:01 <srfsh> MY-R: I did.
2021-06-21 14:31:03 <minimal> s/it if /g
2021-06-21 14:31:07 <MY-R> minimal: I was doing with luks2
2021-06-21 14:31:19 <srfsh> I didn't have GRUB_PRELOAD_MODULES in your setup.
2021-06-21 14:31:25 <srfsh> I'll link the files in a moment.
2021-06-21 14:31:29 <MY-R> srfsh: ook good, so waiting for grub.cfg
2021-06-21 14:31:32 <MY-R> :)
2021-06-21 14:32:13 <minimal> I've spent too much time staring at Grub configs and settings in the past week lol
2021-06-21 14:33:38 <srfsh> MY-R: here it is: https://0x0.st/-9ty.cfg
2021-06-21 14:34:45 <srfsh> also my fstab, reflecting the correct UUIDs: https://0x0.st/-9tt.txt
2021-06-21 14:34:49 <MY-R> srfsh: insmod gcry_whirlpool
2021-06-21 14:35:08 <MY-R> you again used that encryption from Alpine howto
2021-06-21 14:35:32 <srfsh> yeah, but does it matter?
2021-06-21 14:35:59 <srfsh> Only the hash is different.
2021-06-21 14:36:08 <MY-R> time ago was issue with modules that didnt load correct encryption modules but ye in your case it cant find correct root or dunno already
2021-06-21 14:36:31 <srfsh> I can setup with sha512 if you want.
2021-06-21 14:37:51 <MY-R> srfsh: nah
2021-06-21 14:37:57 <minimal> srfsh: so the UUID specified in "search --no-floppy --fs-uuid" lines is the UUID of your /boot not your root (as seen from fstab)
2021-06-21 14:38:26 <MY-R> but he shouldnt have any /boot partition if followed my notes
2021-06-21 14:38:41 <srfsh> I said I added one.  :D
2021-06-21 14:38:43 <minimal> when means that also the lvmid/H4iq.... ref is pointing to /boot
2021-06-21 14:39:25 <MY-R> ...
2021-06-21 14:39:26 <MY-R> whatever
2021-06-21 14:39:54 <srfsh> minimal: the one starts with bc120 is vg0/root.
2021-06-21 14:40:31 <minimal> yes but that's not mentioned in grub.cfg, the UUID for /boot is
2021-06-21 14:40:54 <MY-R> I dont see any point to have separated /boot partition if everything is encrypted
2021-06-21 14:41:11 <minimal> MY-R: agreed
2021-06-21 14:42:16 <srfsh> I shouldn't have made a /boot partition for this setup, yeah, but I will for the setup I want.
2021-06-21 14:42:52 <srfsh> minimal: I see what you mean.
2021-06-21 14:43:13 <srfsh> So, if I cange this to use vg0/root, it will work?
2021-06-21 14:43:41 <minimal> so in GRUB_CMDLINE_LINUX_DEFAULT did you put "cryptroot=UUID=bc120b68-2dc5-4b6b-988b-2762dcb273b1"?
2021-06-21 14:44:20 <srfsh> ON, it points to /dev/sdb2.
2021-06-21 14:44:22 <srfsh> No*
2021-06-21 14:44:33 <minimal> sorry, forget that, still thinking in terms of LUKS-only rather than LVM-on-LUKS
2021-06-21 14:45:12 <srfsh> MY-R's guide and the one on the wiki all refer to the luks partition.
2021-06-21 14:45:41 <minimal> srfsh: yeah like I said, was getting confused.
2021-06-21 14:46:27 <minimal> my question earlier about whether you see Grub menu - if so then you could use "e" to edit the settings to try some changes to see if it could then boot
2021-06-21 14:47:27 <srfsh> I see.  Then I think you meant *before* the password (you said *after*).  In that case, I don't see the menu.
2021-06-21 14:51:06 <MY-R> well, /boot is on separated partition where grub.cfg exist with menu, maybe there is some race condition
2021-06-21 14:51:28 <srfsh> I can recrate my lvm partitions to only have swap and root.
2021-06-21 14:51:33 <srfsh> it works for me.
2021-06-21 14:51:36 <MY-R> and dunno how lvm scan interact with all that
2021-06-21 14:52:03 <srfsh> minimal: if you don't have anything for me to try, I'm gonna go ahead and recreate my lvm partitions.
2021-06-21 14:59:55 <minimal> srfsh: I meant after the password - GRUB menu will appear afterwards
2021-06-21 15:00:56 <srfsh> Whether I type the password correctly or not, it doesn't open up the grub menu, just the rescue cli.
2021-06-21 15:19:52 <minimal> I think it needs to be able to read grub.cfg to show the menu
2021-06-21 15:22:37 <srfsh> I just installed the system according to MY-R's guide (except for the hash).
2021-06-21 15:22:49 <srfsh> Now, let's see what happens when I enter my password.  :D
2021-06-21 15:24:14 <srfsh> same error.  :D
2021-06-21 15:25:15 <MY-R> maybe you dont even boot correct grub, or maybe that hash is fucked up
2021-06-21 15:25:41 <srfsh> What do you mean I don't boot correct grub?
2021-06-21 15:26:03 <srfsh> I'll also try with the same hash you have one more time.
2021-06-21 15:26:05 <MY-R> you said you have openbsd on sda
2021-06-21 15:26:14 <srfsh> sdb is a differnt disk.
2021-06-21 15:26:53 <MY-R> some uefi doesnt support multiple esp partitions, some do
2021-06-21 15:27:20 <srfsh> I chose, from my bios menu, to boot from a specific disk.
2021-06-21 15:27:21 <MY-R> or need pick it up from uefi boot menu
2021-06-21 15:27:37 <srfsh> It works without encryption, so I suppose it supports it.
2021-06-21 15:28:13 <MY-R> ye
2021-06-21 15:33:43 <invoked> firejail didn't make it to 3.14 :(
2021-06-21 15:36:54 <ikke> maybe consider bubblejail?
2021-06-21 15:37:24 <caskd> you mean bubblewrap?
2021-06-21 15:38:19 <MY-R> srfsh: what exact commaund you using when creating luks partition?
2021-06-21 15:38:28 <MY-R> srfsh: cryptsetup -v -c serpent-xts-plain64 -s 512 --hash whirlpool --iter-time 5000 --use-random luksFormat /dev/sdb2
2021-06-21 15:38:30 <MY-R> ?
2021-06-21 15:38:37 <MY-R> exact one
2021-06-21 15:40:35 <ikke> caskd: https://github.com/igo95862/bubblejail
2021-06-21 15:42:19 <caskd> oh
2021-06-21 15:48:58 <srfsh> MY-R: everything is same as yours except for the hash.
2021-06-21 15:49:55 <srfsh> The exact command (on top of my head): cryptsetup -v -s 512 -h whirlpool -i 4000 --type luks1 luksFormat /dev/sdb2 pv0
2021-06-21 15:54:32 <invoked> firejail does quite a bit that bwrap does not
2021-06-21 15:55:00 <ikke> not bwrap
2021-06-21 15:55:06 <ikke> bubblejail
2021-06-21 15:55:20 <ikke> invoked: and it does not matter what it does, if it does so insecurely
2021-06-21 15:55:25 <jvoisin> isn't firejail a complete shitshow security-wise?
2021-06-21 15:55:31 <Sheila> the most important thing firejail does that bwrap does not is introduce security holes. :)
2021-06-21 15:55:44 <ikke> :D
2021-06-21 15:55:54 <invoked> i don't have an opinion about it that way
2021-06-21 15:56:28 <invoked> everything has holes
2021-06-21 15:56:39 <invoked> if you think otherwise, you're extremely naive
2021-06-21 15:56:57 <Sheila> so you're saying we shouldn't bother to address them when we encounter them?
2021-06-21 15:57:30 <ikke> "One of the biggest issues with Firejail is that you can accidentally run unsandboxed applications and not notice."
2021-06-21 15:57:56 <invoked> Sheila: that's putting words in my mouth, i'm not saying that (obviously)
2021-06-21 15:58:24 <Sheila> I mean, it was a very logical extrapolation from your statement.
2021-06-21 15:58:36 <invoked> i dunno what planet you're on, but no.
2021-06-21 15:59:02 <ikke> No ad-hominems, please
2021-06-21 15:59:06 <Sheila> the one where it actually matters if something is insecure relative to something else.
2021-06-21 16:00:51 <ikke> A 'security' sandbox that fails to secure is broken
2021-06-21 16:01:58 <invoked> i don't have an opinion either way. i don't use it myself. someone else does, and they installed 3.14 and firejail's not there anymore.
2021-06-21 16:02:06 <invoked> so i noted it here
2021-06-21 16:02:18 <invoked> but i'm taking notes on everyone's opinions because i want to make sure you feel valued
2021-06-21 16:02:18 <ikke> And I gave an alternative
2021-06-21 16:04:35 <MY-R> srfsh: what that pv0 at the end of command doing?
2021-06-21 16:05:14 <ikke> invoked: we're sorry for the inconvenience it causes for that person
2021-06-21 16:05:16 <srfsh> MY-R: Opps, sorry for that, it was for luksOpen, wasn't it?  :D
2021-06-21 16:05:52 <MY-R> srfsh: I rly wanna help you but Im not even abble to reproduce same steps as your because you changing something all the time
2021-06-21 16:06:15 <rob0> 
ACTION feels valued, thanks

2021-06-21 16:07:32 <srfsh> MY-R: the hash and boot was the only thing I changed, and later I deleted the boot part as well.
2021-06-21 16:08:20 <srfsh> I'm gonna do it with the same hash now.
2021-06-21 16:08:23 <MY-R> srfsh: no, you didnt set up cipher too
2021-06-21 16:08:49 <srfsh> MY-R: It defaults to aes-xts-plain64.
2021-06-21 16:09:13 <srfsh> From cryptsetup 2.1.0+, aes-xts-plain64 is the default value of the cipher.
2021-06-21 16:09:41 <srfsh> --use-random and --verify-passhrase are also the defaults.
2021-06-21 16:12:25 <MY-R> srfsh: so "cryptsetup -v -s 512 -h whirlpool -i 4000 --type luks1 luksFormat /dev/sdb2" is what you all time using, right?
2021-06-21 16:12:43 <srfsh> Exactly.
2021-06-21 16:17:01 <invoked> ikke: /shrug i was told they searched for notes about firejail first. but i see that aports #12643 answered that already, so that's on me for assuming people do what they say.
2021-06-21 16:18:18 <ikke> #12635
2021-06-21 16:21:33 <srfsh> MY-R: Currently created a new LUKS volume with the exact cryptsetup command you used.
2021-06-21 16:23:25 <srfsh> Running "setup-disk -m sys /mnt/" now.
2021-06-21 16:37:14 <srfsh> Here's the new setup's files: /etc/fstab: https://0x0.st/-9v_.txt , /etc/default/grub: https://0x0.st/-9vL.txt , /boot/grub/grub.cfg: https://0x0.st/-9v9.cfg
2021-06-21 16:37:49 <srfsh> rebooting
2021-06-21 16:39:34 <srfsh> the same error.  :'(
2021-06-21 16:40:12 <srfsh> MY-R: This time I have the same hash.
2021-06-21 16:40:46 <MY-R> srfsh: gimme a sec
2021-06-21 16:47:07 <minimal> srfsh: why "cryptkey=yes" in /etc/default/grub, why not just "cryptkey"?
2021-06-21 16:47:08 <MY-R> srfsh: are you here? I
2021-06-21 16:47:18 <MY-R> minimal: 3.13.5 is fucking broken
2021-06-21 16:47:59 <MY-R> 3.14 working fine, already reproduced it with same steps, on 3.13.5 cant mount root, ane on 3.14 all working fine
2021-06-21 16:48:20 <MY-R> srfsh: please use 3.14 iso
2021-06-21 16:48:37 <srfsh> MY-R: Okay.
2021-06-21 16:48:45 <srfsh> I'm gonna download it.
2021-06-21 16:48:53 <MY-R> srfsh: and one sec I will give you my whole history just in case
2021-06-21 16:49:08 <srfsh> on 3.13.5 or 3.14?
2021-06-21 16:49:14 <MY-R> on 3.14
2021-06-21 16:49:18 <MY-R> which working
2021-06-21 16:49:28 <srfsh> So, you reproduced it on 3.13.5 and it wasn't working?
2021-06-21 16:49:34 <MY-R> wasted so much time on that 
2021-06-21 16:49:40 <srfsh> Me too.  :)
2021-06-21 16:49:52 <MY-R> ye on 3.13.5 same steps giving root mount error with kernel panic in rescue mode when type exit
2021-06-21 16:50:02 <minimal> srfsh: also why "cryptdm=pv0"? is "dm0" what you specified on the luksOpen command?
2021-06-21 16:50:11 <minimal> s/dm0/pv0/g
2021-06-21 16:50:11 <alpine-meetbot> minimal meant to say: srfsh: also why "cryptdm=pv0"? is "pv0" what you specified on the luksOpen command?
2021-06-21 16:50:21 <srfsh> minimal: everything is according to MY-R instructions, ask him/her.
2021-06-21 16:50:24 <MY-R> minimal: yes
2021-06-21 16:50:36 <MY-R> one sec, will paste all commands
2021-06-21 16:51:49 <srfsh> MY-R, minimal: you've been very helpful; thanks a lot!
2021-06-21 16:52:44 <minimal> srfsh: well I'm already working on this in my script anyway :-)
2021-06-21 16:53:13 <MY-R> minimal srfsh http://sprunge.us/oJz4Gp
2021-06-21 16:54:52 <minimal> MY-R: you may want to use rng-tools instead of haveged these days
2021-06-21 16:55:41 <MY-R> minimal: it is old howto, Im already using everyhere rng :)
2021-06-21 16:55:51 <MY-R> srfsh: grub config: http://sprunge.us/uXitQ7
2021-06-21 16:55:53 <srfsh> The disk didn't have any creds stored plaintext, so I don't really worry about that.
2021-06-21 16:56:08 <srfsh> in plaintext*
2021-06-21 16:57:01 <MY-R> srfsh: you can change GRUB_DISABLE_OS_PROBER=false since grub changed logic of that... and add GRUB_DISTRIBUTOR="Alpine"
2021-06-21 16:57:38 <minimal> MY-R: rng-tool now has support for jitter-rng (as well as any HW RNG) so it can do the same job as haveged
2021-06-21 16:57:38 <srfsh> I see.
2021-06-21 16:58:57 <minimal> MY-R: why do you "chmod 000" the keyfile? doesn't that mean even root can't read it?
2021-06-21 16:59:11 <MY-R> minimal: root will read it without problem :)
2021-06-21 17:03:42 <minimal> MY-R: in general your command list looks fine. Wonder what's up with 3.13.5 then
2021-06-21 17:06:08 <MY-R> minimal: yeah, used same commands in 3.13.5 and couldnt find root, typed 'exit' and saw kernel panic, so it decrypt luks partition but couldnt load initramfs
2021-06-21 17:07:37 <MY-R> or was loading it and crashed, dunno, Im already tired of digging in it
2021-06-21 17:07:48 <minimal> MY-R I see to remember some people were seeing nl-findfs issues (in initramfs' /init) with things like ZFS (and I think LVM too) where it wasn't working right, could be another similar issue.
2021-06-21 17:07:55 <MY-R> and hope srfsh will enjoy with fully encrypted Aline and can move on!
2021-06-21 17:08:06 <srfsh> Aline lol
2021-06-21 17:08:11 <srfsh> MY-R: Thanks a lot!
2021-06-21 17:08:16 <MY-R> I think that was the problem for ikke too maybe
2021-06-21 17:08:37 <MY-R> and I was doing it in qemu with uefi
2021-06-21 17:08:39 <minimal> well in another couple of days I should get the LVM and LVM-on-LUKS stuff into my script and then it'll be easy to build various combinations of Alpine version, BIOS/UEFI, etc
2021-06-21 17:08:50 <srfsh> I've been trying to set this up for about a month (regularly rage quitting).  :D
2021-06-21 17:09:08 <MY-R> minimal: yeah and then can debug it but for sure it is broken in 3.13.5
2021-06-21 17:09:10 <ikke> #12325
2021-06-21 17:10:03 <MY-R> ikke: yeah so I can confirm that can reproduce it qemu in efi mode with lvm
2021-06-21 17:10:05 <minimal> thanks ikke, I was working from memory
2021-06-21 17:10:05 <MY-R> + encryption
2021-06-21 17:11:56 <MY-R> good that working fine in 3.14 but what changed
2021-06-21 17:13:24 <minimal> 3.14 is3.14 is GRUB 2.06. mkinitfs is same version in both 3.13 and 3.14 so unlikely to be it
2021-06-21 17:13:52 <minimal> s/3.14 is3/3/g
2021-06-21 17:13:52 <alpine-meetbot> minimal meant to say: 3.14 is GRUB 2.06. mkinitfs is same version in both 3.13 and 3.14 so unlikely to be it
2021-06-21 17:16:25 <MY-R> ah ye minimal osclock working on my rpi4 after when installed it on disk :) so in diskless mode hwclock ws enforce to start by something 
2021-06-21 17:18:22 <MY-R> minimal: dont have to be cryptkey=yes? cryptkey enough? good to know :)
2021-06-21 17:18:53 <minimal> MY-R: don't use diskless mode with RPI myself. I guess you could do a "rc-update del hwclock", "rc-update add osclock boot" and then use "lbu" to save the setting
2021-06-21 17:19:42 <minimal> MY-R: I believe "cryptkey" is fine. If you want to use a non-default filename for the keyfile then use "cryptkey=/whatever-file"
2021-06-21 17:19:46 <MY-R> minimal: nah, it didnt work, it always started hwclock :) checked 3 times and of course lbu ci after
2021-06-21 17:20:41 <minimal> MY-R: ah, I'm vaguely recalling that something might be enabling hwclock in diskless mode during each boot
2021-06-21 17:20:50 <minimal> like I said, haven't really used diskless mode
2021-06-21 17:21:10 <MY-R> :)
2021-06-21 17:21:57 <minimal> MY-R: I'll let you know once I've got the LVM-on-LUKS stuff and working LUKS2 in my script
2021-06-21 17:24:07 <MY-R> minimal: thanks!
2021-06-21 17:24:59 <MY-R> but I still think that luks2 on /boot working only with some special keys and not passphrase, at least that how understood some comments, dunno, hope Im wrong
2021-06-21 17:26:27 <minimal> MY-R: will find out soon enough. I need to do some other testing to hopefully push MRs for remote LUKS unlocking first.
2021-06-21 17:27:07 <MY-R> ok :)
2021-06-21 17:29:41 <minimal> later...
2021-06-21 17:33:31 <MY-R> minimal: ahh, have to be "yes" I guess: https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/initramfs-init.in#L483
2021-06-21 17:33:33 <MY-R> ah, quit
2021-06-21 17:34:24 <srfsh> YES
2021-06-21 17:34:26 <MY-R> srfsh: so better use cryptkey=yes in default/grub to be sure that will work
2021-06-21 17:34:40 <MY-R> srfsh: all working!?
2021-06-21 17:34:43 <srfsh> YES
2021-06-21 17:34:44 <srfsh> YES
2021-06-21 17:34:45 <srfsh> YES
2021-06-21 17:34:46 <srfsh> :D
2021-06-21 17:35:25 <MY-R> good to hear :)
2021-06-21 17:35:31 <srfsh> :)
2021-06-21 17:35:44 <MY-R> damn bug :\
2021-06-21 17:36:49 <insep> ping
2021-06-21 17:36:55 <ikke> yes
2021-06-21 17:36:55 <srfsh> I should've checked the latest version... but I started to try this about a month ago when 1.14 was't available.  :P
2021-06-21 17:37:05 <ikke> insep: you are here nopw
2021-06-21 17:37:21 <Habbie> insep, i see you!
2021-06-21 17:37:47 <ikke> MY-R: I missed your message regarding reproducing it
2021-06-21 17:38:22 <ikke> MY-R: could you add those details to the issue? ncopa tried to reproduce it
2021-06-21 17:38:25 <MY-R> ikke: is enough to do LVM on LUKS from Alpine how to and do it in qemu with efi enabled
2021-06-21 17:38:52 <MY-R> so efi + lvm are two common things
2021-06-21 17:39:07 <ikke> The system that was affected for us uses grub + legacy boot
2021-06-21 17:39:39 <ikke> And another system with eufi was not affected
2021-06-21 17:40:11 <MY-R> ikke: will try with normal bios and same steps, but not today, tomorrow becouse Im already tired of that when was trying to help :P
2021-06-21 17:40:24 <ikke> appreciated
2021-06-21 17:40:30 <MY-R> so maybe it is grub + lvm
2021-06-21 17:40:39 <ikke> possibly
2021-06-21 17:46:20 <Habbie> writing something for insep to test their connection
2021-06-21 17:46:28 <Habbie> lorem ipsum dolor sit alpine
2021-06-21 17:46:34 <insep> finally
2021-06-21 17:46:42 <Habbie> :)
2021-06-21 17:47:12 <insep> so now
2021-06-21 17:47:26 <insep> how different are uboot images from http://dl-cdn.alpinelinux.org/alpine/latest-stable/releases/armv7/ and "disk" installation type?
2021-06-21 19:02:12 <markand> I have a very strange framebuffer display 
2021-06-21 19:02:28 <markand> I have my thinkpad that is docked with an external screen (24" the screen, 4K resolution) 
2021-06-21 19:02:46 <markand> but on the tty, there is very large black border at the bottom and right edges
2021-06-21 19:02:55 <markand> so I lose somewhat like 1/3 of the screen space
2021-06-21 19:36:02 <MY-R> ikke: nope ikke, I cant reproduce it anymore, repeat exact same steps and now 3.13.5 install fine
2021-06-21 19:37:25 <ikke> MY-R: ugh, so anoying
2021-06-21 19:37:42 <MY-R> ikke: but noticed already years ago that sometimes when using two different parition tool then alpine randomly was losing partitions, they not always were visible in /dev/ and it happened on many different machines and in qemu too
2021-06-21 19:38:37 <MY-R> like something didnt want sync, so maybe that is the issue, some last pieces of data during partitioning or boot loader setup is wrong
2021-06-21 19:40:12 <MY-R> remember I was using "parted" to create 2 partitions and when ran after "cfdisk" to check it if all fine and quit without save anything then somehow one of partition was gone from /dev/
2021-06-21 19:41:30 <MY-R> but reproduce it is rly hard, kind of random
2021-06-21 19:41:54 <MY-R> that is why I never use Alpine to create/manage partitions
2021-06-21 19:42:19 <MY-R> I wanted create issue on gitlab but I have no idea from what start
2021-06-21 22:34:00 <asdflkj> what are the 'common used packages' included in Extended on https://alpinelinux.org/downloads/ ? (also, would 'commonly used' be more grammatical?)
2021-06-21 22:44:25 <MY-R> asdflkj: https://gitlab.alpinelinux.org/alpine/aports/-/blob/3.14-stable/scripts/mkimg.standard.sh#L36
2021-06-21 23:54:42 <nakedpotato> \q
2021-06-21 23:54:46 <nakedpotato> oops
2021-06-22 00:31:00 <whatsyourask> Hi all, let's see if someone can help me a little bit, how can I add some files to /lib/firmware/brcm? The file system is read-only, I am using a raspberry pi with a diskless installation. Is there a folder in /usr/* or /root/* where additional modules are loaded?
2021-06-22 02:04:22 <Hello71> firmware isn't on the modloop, is it?
2021-06-22 02:43:21 <mdekstrand> I have Alpine x86_64 installed on a machine that usually runs headless. It gets into situations where it will no longer boot headless; this reliably happens after a kernel upgrade, and most recently happened after a power failure.
2021-06-22 02:43:43 <mdekstrand> If I boot the machine with a monitor attached (doesn't even need a keyboard!), it will successfully boot up without any intervention.
2021-06-22 02:44:13 <mdekstrand> After a clean boot, it's usually good for further reboots or power-downs, until a kernel upgrade or something unexpected.
2021-06-22 02:44:28 <mdekstrand> (I don't think specific hardware matters, but this is on an Atom-based QNAP NAS)
2021-06-22 02:45:59 <mdekstrand> (root FS is ext4 on mdraid; two additional FS's are xfs on mdraid. mkinitfs.conf has features="ata base ide scsi usb virtio ext4 xfs raid")
2021-06-22 02:46:36 <invoked> are you running X?
2021-06-22 02:46:49 <mdekstrand> no
2021-06-22 02:47:58 <Hello71> sounds suspiciously like some crap bios nonsense
2021-06-22 02:48:38 <mdekstrand> That wouldn't surprise me, but I also haven't had this problem with either NixOS or Debian on the machine - only Alpine has exhibited this.
2021-06-22 02:51:03 <Hello71> what bootloader
2021-06-22 02:51:51 <invoked> i think the main difference there would be busybox
2021-06-22 02:52:07 <Hello71> wat
2021-06-22 02:52:41 <invoked> i'm not saying it's busybox, but that's the obvious diff between alpine
2021-06-22 02:53:06 <invoked> but yeah, bios
2021-06-22 02:53:44 <invoked> there's info missing about the config (uefi?)
2021-06-22 02:53:55 <mdekstrand> I'm using grub
2021-06-22 02:54:03 <mdekstrand> not in EFI mode
2021-06-22 02:54:24 <mdekstrand> (I don't think it's in EFI mode, anyway. Double-checking.)
2021-06-22 02:55:13 <mdekstrand> yep, definitely not EFI.
2021-06-22 02:56:08 <invoked> my second guess after i asked about X was power mgmt
2021-06-22 03:00:35 <mdekstrand> I was just able to force it into this state with a hard power-off (with current Alpine 3.14 install + kernel).
2021-06-22 03:00:50 <mdekstrand> In the hung state, when I plugged the monitor in while the machine was already powered on, the monitor complained of no signal.
2021-06-22 03:01:02 <mdekstrand> and the power button instantly powered it off.
2021-06-22 03:01:23 <mdekstrand> I've now plugged the monitor in, powered it back on, and it's up and running and responding to network :shrug:
2021-06-22 03:01:41 <Hello71> why do you think it's related to the monitor
2021-06-22 03:02:16 <mdekstrand> because with a monitor plugged in, it successfully boots; without the monitor plugged in, it does not. that's the only difference between the booting and non-booting configurations.
2021-06-22 03:03:02 <mdekstrand> I discovered this because it was in a non-booting state, so I dug out the monitor & keyboard to fix it from the console, and it just booted without needing any repair.
2021-06-22 03:03:16 <invoked> if you force power down by the power switch (if you run that test again), switch off the whole unit at the power supply if you can (or unplug), wait a few, then bring it back up without a monitor
2021-06-22 03:03:27 <mdekstrand> after this happened 2-3 times, I tried just monitor and no keyboard to try to isolate which peripheral's presence was allowing it to boot.
2021-06-22 03:04:09 <mdekstrand> @invoked powered off, waiting now.
2021-06-22 03:04:44 <invoked> did you remove it from the power source?
2021-06-22 03:04:49 <invoked> (humor me)
2021-06-22 03:05:04 <invoked> electronic switches don't actually turn things off
2021-06-22 03:05:09 <mdekstrand> yep, I did.
2021-06-22 03:05:12 <invoked> ok
2021-06-22 03:08:37 <mdekstrand> plugged back in and it's powering up now.
2021-06-22 03:13:22 <mdekstrand> @invoked it's back up but no ping, and router logs show it hasn't asked for an IP.
2021-06-22 03:13:35 <mdekstrand> and a successful boot responds to ping & ssh by now
2021-06-22 03:14:59 <invoked> ok. just trying to help you narrow it down, as in whether it's an acpi issue with alpine on a hard reboot
2021-06-22 03:15:07 <mdekstrand> np, I appreciate it :)
2021-06-22 03:15:43 <mdekstrand> on a lark i'm going to try adding nomodeset to my kernel params
2021-06-22 03:18:01 <invoked> good idea. also maybe fiddle/inspect what acpi is doing
2021-06-22 03:22:24 <mdekstrand> nomodeset did not fix it. (also, with nomodeset and a monitor to make sure boot still worked, it still displayed the console in a high-resolution mode.)
2021-06-22 03:22:46 <mdekstrand> what would i need to do to inspect acpi?
2021-06-22 03:22:52 <mdekstrand> especially in the no-boot scenario?
2021-06-22 03:25:49 <mdekstrand> I guess I don't have evidence that it's actually getting to Linux - it's possible that Grub is failing the graphics setup somehow.
2021-06-22 03:31:37 <invoked> sorry, i have to crash early because of 4am/tomorrow reasons. re: acpi, acpi=off kernel param, see if you have any logs from your failed boots (particularly apic)
2021-06-22 03:31:49 <invoked> at least i think it's 'apic=off'
2021-06-22 03:31:55 <invoked> er, 'acpi=off'
2021-06-22 03:32:05 <mdekstrand> no problem, I also have to log off pretty soon.
2021-06-22 03:32:11 <mdekstrand> Thank you for your help :)
2021-06-22 03:32:21 <invoked> no problem, good luck!
2021-06-22 03:32:50 <invoked> note: try adding apic=off manually first before changing grub.conf
2021-06-22 03:33:09 <invoked> argh. acpi!
2021-06-22 03:33:25 <invoked> anyway, your system may not boot with acpi=off at all
2021-06-22 03:33:44 <invoked> so better to test it first, but... that's what i would do.
2021-06-22 03:34:06 <invoked> acpi/apic dyslexia
2021-06-22 03:40:02 <mdekstrand> acpi=off results in a normal boot with the monitor on, still fails with the monitor unplugged. will go peek at logs. will also try either turning off all of Grub's video stuff, or booting with extlinux, to see if it's a pre-kernel problem.
2021-06-22 03:40:27 <whatsyourask> Hello71, ty for the tip
2021-06-22 03:56:18 <mdekstrand> And we have a solution. It was a Grub problem, not a kernel problem. Configuring GRUB_TERMINAL=vga_text and GRUB_GFXPAYLOAD_LINUX=text in /etc/default/grub cleared it up.
2021-06-22 03:56:27 <mdekstrand> thank you again @invoked
2021-06-22 11:42:51 <user8711> hey, having bit of a problem with dbus, it refuses to start because of 'error relocating symbol ...', anyone had the same issue? I've tried to update dbus and musl already, im on edge, couldnt figure out how else to ask ddg
2021-06-22 11:46:17 <danieli> sounds like you may need to upgrade something else that dbus depends on
2021-06-22 11:55:12 <user8711> danieli: it depends on busybox dbus-libs musl and expat, all are up to date, though edge dbus has been flagged in pkgs.alpinelinux.org, don't know why however
2021-06-22 11:57:16 <danieli> what's in your /etc/apk/repositories?
2021-06-22 11:57:23 <danieli> you can use tpaste.us
2021-06-22 12:42:23 <user8711> danieli: https://paste.ubuntu.com/p/KWQs5YcbcJ/ 
2021-06-22 12:42:57 <danieli> that *should* be fine
2021-06-22 12:43:05 <danieli> what does `apk -v fix' show?
2021-06-22 12:44:07 <user8711> on the dbus package or by itself? OK: 333 packages ...
2021-06-22 12:44:42 <user8711> that is the output of the latter
2021-06-22 12:53:43 <Hello71> what exactly is it error relocating symbol
2021-06-22 12:56:36 <user8711> https://paste.ubuntu.com/p/Z7xSVJQtvm/ Hello71:
2021-06-22 14:24:23 <MY-R> anyone using Alpine with xorg could grep Xorg log file and check what return: grep -i silken Xorg.0.log
2021-06-22 14:26:31 <markand> I feel like glade is completely broken under Alpine
2021-06-22 14:26:39 <markand> I can't even add a container into a GtkWindow
2021-06-22 14:26:54 <markand> MY-R, [   102.646] (==) modeset(0): Silken mouse disabled
2021-06-22 14:27:34 <markand> (glade:4125): GladeUI-WARNING **: 16:25:54.858: Failed to load libgladegtk.so: Error loading shared library libgladegtk.so: No such file or directory
2021-06-22 14:27:39 <markand> hmmm that may explain some things
2021-06-22 14:28:27 <markand> okay, so glade actually *needs* glade-dev as runtime dependencies
2021-06-22 14:42:29 <user8711> MY-R: https://paste.ubuntu.com/p/4HzN6X4Xtw/ Lenovo X200, plain X
2021-06-22 14:43:48 <user8711> lol, forgot you just wanted a line. [ 20425.621] (==) intel(0): Silken mouse disabled
2021-06-22 14:58:31 <MY-R> markand user8711 thank you!
2021-06-22 14:59:26 <MY-R> ok so wont install Alpine for now, meson still not building xorg with silken mouse enabled and last thing I want is laggy mouse cursor in wm
2021-06-22 20:33:47 <ngortheone> Hey folks. Did anyone had tried (and succeeded) running Zoom on Alpine?
2021-06-22 20:35:17 <ikke> I would imagine you need a chroot or something like flatpak
2021-06-22 20:38:39 <PureTryOut> Flatpak is the easiest way to go and works well, I've used it for Zoom once
2021-06-22 20:39:33 <ngortheone> Thanks! I will try that
2021-06-22 23:10:22 <asdflkj> https://docs.alpinelinux.org/user-handbook/0.1a/Installing/setup_alpine.html#_full_setup_alpine seems to say I need to use an answer file and/or environment variables to customize setup-alpine's behaviour (because "You have a single (currently) physical disk and you want to install to it: I have both an internal HDD that I don't want to ever be mounted or touched at all, and a
2021-06-22 23:10:23 <asdflkj> USB flash drive that I want to Alpine to boot from (which it has), later save customizations to, and eventually be able to suspend to without using it as swap while running. do I (and if so, which variables), or can I just use diskless mode for now?
2021-06-22 23:11:04 <asdflkj> s/you want to install to it/&" doesn't apply to me/
2021-06-22 23:11:04 <alpine-meetbot> asdflkj meant to say: https://docs.alpinelinux.org/user-handbook/0.1a/Installing/setup_alpine.html#_full_setup_alpine seems to say I need to use an answer file and/or environment variables to customize setup-alpine's behaviour (because "You have a single (currently) physical disk and &" doesn't apply to me: I have both an internal HDD that I don't want to ever be mounted or touched at all, and a
2021-06-23 00:05:07 <asdflkj> for now I just want to set alpine up without touching the HDD, setup-alpine with disk mode "none" is enough to do that, right?
2021-06-23 00:20:05 <asdflkj> 
ACTION seems to have found #alpine-linux's naptime, or found himself on some /ignore lists :/

2021-06-23 00:28:15 <trylletre> pointing it to the USB only should work? in the disk/lbu opts
2021-06-23 00:31:31 <trylletre> -m sys /dev/yourusb or -m data /dev/yourusb shouldnt touch your hdd
2021-06-23 00:32:13 <trylletre> in the answer file, or you can run setup-disk and the others yourself, from https://wiki.alpinelinux.org/wiki/Alpine_setup_scripts#setup-disk
2021-06-23 00:35:16 <asdflkj> thanks
2021-06-23 00:35:40 <asdflkj> my only worry with those is that swapping will kill my flash drive
2021-06-23 00:36:41 <trylletre> -s 0 in diskopts disables swap
2021-06-23 00:37:19 <asdflkj> oh, read that and alt-tabbed back just as you said it
2021-06-23 00:37:27 <trylletre> :)
2021-06-23 00:54:26 <asdflkj> thanks again, running setup-alpine now!
2021-06-23 01:00:58 <asdflkj> doing timezone now (I'm in America/New_York), https://docs.alpinelinux.org/user-handbook/0.1a/Installing/manual.html#_timezone says "it is recommended that you perform this step entirely by hand", but I can do that after setup-alpine exits normally, I don't have to ^C now to do it, right?
2021-06-23 01:08:35 <trylletre> yep
2021-06-23 01:10:49 <asdflkj> thank goodness. does it make a difference if I select the right zone or UTC in setup-alpine?
2021-06-23 01:12:32 <trylletre> i dont know, but you can change it later regardless
2021-06-23 01:25:07 <etothepii> Hello - I'm trying to get zfs running with Alpine and my Raspberry Pi 4. I'm not looking to run a zfs root, just have an external drive. However, when I installed the zfs-rpi4 package, it pulled in linux-firmware-brcm, which fails to install with lots of 'failed to rename lib/firmware/brcm/.apk.[long string]
2021-06-23 01:25:07 <etothepii>  to lib/firmware/brcm/bcm4329-fullmac-4.bin' type errors. Does the raspi have other packages I need to install prior to attempting to install the zfs modules?
2021-06-23 01:34:18 <asdflkj> 
ACTION misunderstood the docs on DISKOPTS thinking the partition was part of -m, setup-alpine exited after printing busybox readlink's usage and (IIRC) something about -s not being a valid partition (scrolled off the screen when running less custom_answerfile), time to retry with DISKOPTS="-s 0 -m data /dev/sdb" instead of "-m data /dev/sdb -s 0" :P

2021-06-23 02:05:29 <etothepii> OK - I think it looks like the issue with installing the zfs module on the raspi might be related to it being a diskless install? Is there any documentation covering installing modules on diskless? I've seen a few reddit posts with people asking how to get wireguard working on diskless installs, but nothing seems conclusive.
2021-06-23 04:02:56 <anjan> is there a security advisory email for alpine's security issues?
2021-06-23 04:03:22 <anjan> is there a security advisory anounce email for alpine's security issues?
2021-06-23 04:03:26 <anjan> sorry if repeat, I didnt know if my first message sent
2021-06-23 04:26:15 <baconicsynergy> There should be one, lemme see
2021-06-23 04:28:11 <baconicsynergy> I actually cannot find it. Hmm
2021-06-23 04:30:45 <insep> https://lists.alpinelinux.org/~alpine/security-announcements
2021-06-23 04:38:12 <trylletre> didnt know alpine ran the sourcehut mailing list interface
2021-06-23 08:29:47 <gopher22> Hello, I'm trying to follow this guid to enable php-fpm under apache2: https://wiki.alpinelinux.org/wiki/Apache_with_php-fpm. There is a mention there that "This step is not needed any more". Has anyone have an ideea to which step does it refer? The one above the text, the one below, or the all the rest below?
2021-06-23 08:35:57 <lopid> well, is httpd-mpm.conf loaded by default?
2021-06-23 09:02:24 <gopher22> lopid nope, doesn't look so there is no such conf under ./conf.d/
2021-06-23 09:04:11 <lopid> the steps seem straighforward enough to figure out from your existing setup
2021-06-23 09:04:33 <lopid> though steps that aren't needed ought to be removed, not left in for confusion
2021-06-23 09:07:05 <gopher22> agreed
2021-06-23 09:08:25 <gopher22> ok, so there is no ./original/extra/ dir under /etc/apache2, so this step should be omitted I suppose
2021-06-23 11:15:18 <dekedro> Hi, noticed that adwaita gtk2 theme was removed from aports (https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/21680), even though there are applications which could benefit from it. Is there some active effort to drop gtk2?
2021-06-23 15:44:58 <asdflkj> so, I ran setup-alpine with DISKOPTS="-s 0 -m data /dev/sdb" in the answerfile, but it exited with an error because /dev/sdb (the USB flash drive I booted from) is still mounted (and umount /dev/sdb1 says "can't umount /media/usb: Resource busy"). is that expected?
2021-06-23 15:46:07 <asdflkj> (I thought that the standard iso booted in diskless mode. maybe setup-alpine mounted it during its previous aborted run, when I had put -s 0 after the partition instead of before it?)
2021-06-23 16:58:29 <kevint> dekedro if I remember right, there is an active effort to drop gtk2
2021-06-23 17:16:15 <dekedro> Got it, thanks. Is there some timeline for that?
2021-06-23 17:58:25 <kevint> Hmm now that I am not sure about
2021-06-23 17:59:10 <asdflkj> 
ACTION just went fully diskless with no config backup or cache for now

2021-06-23 17:59:22 <user8712> hey, I was here a day ago about dbus issues (not finding symbols) but lost connection if anyone found a solution/an idea, you'll have to send it again, sorry :]
2021-06-23 18:00:05 <asdflkj> is setting up my main user with bash (not ash) as the login shell likely to cause any issues?
2021-06-23 18:01:31 <user8712> not if you have scripts in .ash_* asdflkj
2021-06-23 18:02:13 <asdflkj> user8712: I've been connected continuously and I don't think you missed anything about dbus
2021-06-23 18:03:06 <user8712> thanks :), about the scripts, I mean if you have your personal scripts set up to run after logon, you'll need to rename the files accordingly for bash :]
2021-06-23 18:04:49 <asdflkj> np, and thank you :D. so alpine itself and its packages don't rely on any ash-specific behaviour, at least without setting the right shebang?
2021-06-23 18:07:17 <user8712> pretty sure all of their scripts are POSIX compliant so they should work with *sh
2021-06-23 18:08:03 <ikke> We do use some functionality outside of POSIX
2021-06-23 18:08:25 <ikke> but you do not need to use ash as shell
2021-06-23 18:09:02 <asdflkj> I imagine it's probably mostly fine, but if no one has tried it before I think I'll adduser -s /bin/ash and just set bash to open by default in tmux/st/dwm/sxhkd, just to be safe
2021-06-23 18:09:48 <ikke> Changing the shell of your user should not affect anything else
2021-06-23 18:10:13 <ikke> I use zsh as shell, others use fish
2021-06-23 18:10:15 <ikke> or bash
2021-06-23 18:10:40 <ikke> system scripts all use a shebang line to provide the interpreter (and I believe it defaults to /bin/sh anyway)
2021-06-23 18:11:50 <asdflkj> great, I'll use bash and get my perfect readline bindings and process substitution
2021-06-23 19:08:14 <asdflkj> I don't speak any language besides US English, but I do use Unicode often. do I need to do all the language support stuff in https://wiki.alpinelinux.org/wiki/Installation#Post-Install ?
2021-06-23 19:08:38 <ikke> no
2021-06-23 19:18:38 <MY-R_> well, unicode in /etc/rc.conf isn't bad idea
2021-06-23 19:30:41 <Hello71> dekedro: why fpm for apache?
2021-06-23 19:43:31 <invoked> to make things a little easier between dists/machines i'll shebang /usr/bin/env.  at least where it concerns linux.
2021-06-23 19:46:32 <yyp> /bin/sh or /usr/bin/<shell> is more portable :)
2021-06-23 20:03:42 <invoked> well in linux anyway, env is in both busybox and coreutils (with some major differences, but none that matter to me)
2021-06-23 20:12:59 <asdflkj> 
ACTION would like to interject for a moment, this is an example of why calling GNU "linux" is stupid, android runs linux too, and it uses toybox, whereas invoked probably meant to refer to GNU coreutils

2021-06-23 20:25:20 <invoked> i was speaking of the environment generally, and i'm not a fan of gnu prefixing
2021-06-23 20:26:21 <invoked> ain't nobody got time for that
2021-06-23 20:27:38 <asdflkj> IMO, "GNU" is a better way to refer to the environment of Not Unix(TM)s than "linux", and, as only 3 characters, easier and faster to type
2021-06-23 20:33:22 <MathGeniusJodie> gnu linux makes sense for gnu linux
2021-06-23 20:33:39 <MathGeniusJodie> linux just isn't always gnu
2021-06-23 20:34:02 <asdflkj> gnu isn't always linux, either
2021-06-23 20:34:55 <asdflkj> in the particular case of busybox, it was created for GNU (for Debian, while the FSF sponsored it and it had better support for more kernels IIUC), and specifically imitates the GNU coreutils (all options it has that aren't POSIX are GNU-style (as opposed to *BSD-style) IME)
2021-06-23 20:35:20 <invoked> great. tell me again why i have to type gnu every time
2021-06-23 20:36:20 <asdflkj> it's shorter than linux, tell me why you want to type like 66% more characters
2021-06-23 20:37:12 <invoked> it's in the man^H^H^Hinfo page so i think it's easier to just refer to the subject rather than the org (in this case, coreutils)
2021-06-23 20:37:33 <invoked> but if you want to type gnu coreutils (et al, ad nauseum), knock yourself out.
2021-06-23 20:41:08 <asdflkj> typing coreutils instead of gnu coreutils makes perfect sense to me, I was only nitpicking the "in linux generally"
2021-06-23 20:42:48 <invoked> linux generally consists of a lot of contributors who have nothing to do with gnu.
2021-06-23 20:43:17 <invoked> for gnu to put its stamp on everything in the linux ecosystem is hubris
2021-06-23 20:43:56 <invoked> sorry but i'm not here to join cults
2021-06-23 20:44:01 <asdflkj> but the opposite is true IMO
2021-06-23 20:44:25 <asdflkj> gnu consists of lots of contributors who have little to do with linux
2021-06-23 20:46:05 <asdflkj> and linux putting its stamp on everything in the gnu ecosystem is hubris, especially because it often covers gnu's, whereas gnu usually just puts its stamp next to linux
2021-06-23 20:58:32 <asdflkj> 
ACTION could advocate for the great gnu cult all day, maybe in #alpine-offtopic :P ... anyway ...

2021-06-23 20:59:57 <asdflkj> I forgot to add myself to wheel until now, it's not possible to make my currently opened shells reflect that, only login new ones that do, right?
2021-06-23 21:11:02 <lopid> right
2021-06-23 21:11:31 <asdflkj> thanks
2021-06-23 21:17:33 <MathGeniusJodie> I specifically avoid gnu software
2021-06-23 21:17:53 <MathGeniusJodie> now, alpine uses gcc as it's default compiler
2021-06-23 21:18:02 <MathGeniusJodie> making it now totally gnu free
2021-06-23 21:18:10 <MathGeniusJodie> but eh, alpine is convenient 
2021-06-23 21:18:23 <MathGeniusJodie> *not
2021-06-23 21:20:13 <invoked> is any linux-based os bootstrapped with only clang/llvm yet (no gcc)? i don't even know.
2021-06-23 21:20:35 <invoked> i'm sure there are toys out there, but i mean, actual releases.
2021-06-23 21:21:04 <asdflkj> isn't android?
2021-06-23 21:21:37 <invoked> maybe? i dunno.
2021-06-23 21:22:36 <asdflkj> this seems pretty #alpine-offtopic to me, but I specifically seek out gnu software
2021-06-23 21:23:16 <asdflkj> I'm typing this on OpenBSD, and using it has only deepened my appreciation for gnu (and dislike for linux)
2021-06-23 21:23:44 <invoked> openbsd has deepened your appreciation for gnu? wild.
2021-06-23 21:28:57 <asdflkj> yes, I've only use bsd on this shell account and a vps so far, but I've missed GNU bash (particularly process substitution) and all readline programs' consistency and editing power. better M-., better C-r, better C-u. OTOH the great manpages have helped me realize how bad linux docs are (particularly linux man-pages is pretty bad, gnu info is pretty decent except the viewer's
2021-06-23 21:28:59 <asdflkj> foreign to a vi user)
2021-06-23 21:34:00 <asdflkj> I've also used busybox-based distros basically the whole time I've used linux, with varying amounts of GNU utils replacing them, which has grown a resentment in me for some of busybox's annoyances, like ash's lack of process substitution and yanking, and busybox less and vi's way of polluting scrollback (although less is not GNU, and my favorite vi is vis)
2021-06-23 21:37:34 <etothepii> Is there any guidance for what can be a full page in the wiki? I'd like to take a stab at consolidating some of the 'setup SD card for Raspi' materials into its own page, since those appear across different articles, and the instructions sometimes conflict (some pages have duplicate data, and others have conflicts - for example the sys install page
2021-06-23 21:37:34 <etothepii> suggests fat16 is required for the boot partition, others say fat32).
2021-06-23 21:37:55 <invoked> asdflkj:  every os has opinions. it seems to me that if you are hardcore gnu, guix or hurd or something might be more suitable to your endgame than alpine.
2021-06-23 21:40:28 <Habbie> etothepii, i have no guidance but i struggled making sense of 'alpine on raspi' so you have my gratitude already for wanting to improve that documentation
2021-06-23 21:40:39 <asdflkj> invoked: yes, guix would probably be better, I'm mainly on alpine just because I need to keep this box's HDD unmounted
2021-06-23 21:43:06 <asdflkj> plus I want to get a pinephone (I'm a broke teenager and physically can't afford the better Librem 5), and use sxmo, based on alpine (/postmarketOS)
2021-06-23 21:44:14 <etothepii> Habbie: Happy to help - I've been running into some odd corner cases myself so I'm hoping I can at least make it easier for the next person.
2021-06-23 21:44:52 <Habbie> etothepii, i did not get to the point of running into corner cases - i read a few things, asked a few things in here, got hints that i couldn't match with the docs, and at the time i was looking for something easy, so i bailed out :)
2021-06-23 21:46:25 <etothepii> Habbie: If you're not installing kernel modules/are ok with a diskless install, the current 'install on pi' seems mostly correct (at least as of the most recent build). Do you recall what you stumbled on?
2021-06-23 21:47:33 <Habbie> etothepii, it's been a while, but as i recall it, i could 'run from ramdisk' or 'do a 15 step bootstrap thing'
2021-06-23 21:47:44 <Habbie> etothepii, i could totally be exaggerating here
2021-06-23 21:48:19 <Habbie> etothepii, i should also point out i've never installed alpine on a PC either so i might be comparing apples to tomatoes
2021-06-23 21:48:42 <Habbie> etothepii, that all said, i'd be happy to review your work if you get around to it
2021-06-23 21:49:49 <etothepii> Habbie: Those are actually two of the things I want to improve: both what the diskless install means for pi users, since I think most pi users will want it, and I want to take a stab at cleaning up the traditional persistent install on pi section. 
2021-06-23 21:50:04 <Habbie> cool
2021-06-23 21:51:35 <etothepii> Habbie: If you're willing to review, that would be fantastic. I can ping you once I get around to updating the pages. I'm probably going to see if I can consolidate the sd card formatting first (assuming there's no 'new pages must be X'), just because it clogs up a lot of the individual install pages on its own, and then tackle cleaning up some of
2021-06-23 21:51:35 <etothepii> the other docs. 
2021-06-23 21:55:28 <Habbie> etothepii, yes, a ping would be perfect. I have no idea about policy - but if I have to guess, making entirely new pages so they can be reviewed is probably something you would get away with?
2021-06-23 21:59:43 <asdflkj> why does bash[-complete] not tab complete tmux subcommands?
2021-06-23 22:00:08 <asdflkj> I ran apk fix on both bash-completion and tmux
2021-06-23 22:00:49 <asdflkj> ...and now bash as well
2021-06-23 22:16:50 <asdflkj> found the completion file, apparently it's kept in a separate repo, and thus omitted from the alpine package
2021-06-23 22:28:15 <asdflkj> can apk find the real, deps-included installed size of a package? (eg. build-base is not really 4096 B, more like 1-200 MB)
2021-06-23 22:58:40 <asdflkj> dumb question, startx must run as root, right? (wish the docs mentioned that / xenocara was packaged)
2021-06-23 22:59:43 <c705> startx should never be run as root
2021-06-23 23:01:08 <asdflkj> truly dumb question then :P
2021-06-23 23:03:16 <asdflkj> do I need to install some gpu-specific packages to fix https://termbin.com/n05r ?
2021-06-23 23:05:40 <asdflkj> the docs seemed to imply setup-xorg-base would get at least basic unaccelarated framebuffer support
2021-06-23 23:08:27 <c705> [ 87370.840] (EE) Failed to load module "fbdev" (module does not exist, 0)
2021-06-23 23:09:00 <c705> you need the xf86 video package for your card(s)
2021-06-23 23:10:07 <asdflkj> I thought setup-xorg-base would get that, but I'll go look up my apu
2021-06-23 23:10:32 <c705> what de are you using?
2021-06-23 23:11:37 <asdflkj> some AMD APU, Athlon IIRC
2021-06-23 23:11:56 <asdflkj> I'm trying to remember where the file is in /sys
2021-06-23 23:13:34 <c705> no display environment
2021-06-23 23:13:40 <c705> xfce, kde, wayland?
2021-06-23 23:13:53 <asdflkj> no, I just want dwm
2021-06-23 23:14:42 <c705> you can start it with dwm, not startx
2021-06-23 23:14:57 <c705> if you want to use startx, you need to modify your .xinitrc
2021-06-23 23:15:00 <c705> https://wiki.archlinux.org/title/Dwm#Starting
2021-06-23 23:15:16 <asdflkj> I did `echo 'exec dwm' > .xinitrc`
2021-06-23 23:15:23 <c705> why?
2021-06-23 23:15:47 <trylletre> that should work by itself anyway
2021-06-23 23:16:11 <c705> well, no, you just ran over your entire .xinitrc if you used > and not >>
2021-06-23 23:16:19 <asdflkj> it didn't exist before
2021-06-23 23:16:31 <trylletre> a .xinitrc with nothing but 'exec mywm' works fine though
2021-06-23 23:17:32 <asdflkj> that's what I thought
2021-06-23 23:17:39 <c705> sure. what happens if you try dwm instead of startx
2021-06-23 23:18:02 <asdflkj> "dwm: cannot open display"
2021-06-23 23:20:33 <asdflkj> I think I just need to find drivers setup-xorg-base somehow failed to get, maybe because APUs are so unusual
2021-06-23 23:20:42 <trylletre> you want uhh.. something like xf86-video-ati, depending on what you specifically have
2021-06-23 23:20:48 <trylletre> i think the older apus use that
2021-06-23 23:20:51 <c705> yeah that ^
2021-06-23 23:20:56 <trylletre> https://wiki.alpinelinux.org/wiki/Radeon_Video
2021-06-23 23:21:06 <trylletre> you also do need to run the script
2021-06-23 23:21:12 <trylletre> as for fbdev i think its -video-fbdev? been a while
2021-06-23 23:21:16 <c705> or xf86-video-amdgpu or xf86-video-raedon or whatever is most appropriate
2021-06-23 23:21:53 <trylletre> i dont think apk magically fetches the drivers you need with the xorg-base package, you do have to manually figure that out
2021-06-23 23:22:41 <trylletre> based on the xorg log, it does try to load them at least, so it should work as soon as you have something valid installed
2021-06-23 23:22:53 <c705> i don't think the script is nessecary anymore. all I had to do was install xf86..amdgpu + sway + some other crap and it takes the modesetting fine
2021-06-23 23:23:01 <asdflkj> trylletre: yeah, I just thought it would have a good fallback
2021-06-23 23:23:15 <c705> no, X doesn't fall back to anything it just craps itself
2021-06-23 23:23:51 <trylletre> it does have a fallback, it tried everything and found none of the drivers because none are there
2021-06-23 23:23:58 <trylletre> the log tried to load like all of them
2021-06-23 23:27:24 <asdflkj> basically I just need to figure out how new "newer" devices (as https://wiki.alpinelinux.org/wiki/Radeon_Video says) are, or install both ati and radeon drivers
2021-06-23 23:32:21 <asdflkj> oh, I wanted /proc/cpuinfo, not anything in /sys
2021-06-23 23:37:35 <trylletre> athlon is definetely old
2021-06-23 23:39:00 <trylletre> amdgpu is roughly since.. GCN cards? late HD77+, r7/r9/rx etc
2021-06-23 23:39:23 <trylletre> i dont think the older gpus are in it, but regardless it shouldnt take long to install one, check, uninstall install other, check, etc
2021-06-23 23:40:25 <asdflkj> it's an "Athlon A4-5300B APU with Radeon(tm) HD Graphics"
2021-06-23 23:41:37 <asdflkj> annoyingly linux-firmware-amdgpu's license is "custom:multiple" with no indication of if it's libre or not
2021-06-23 23:43:47 <c705> well, the firmware on your card isn't "libre" so might as well toss it if that's not kosher with you
2021-06-23 23:45:15 <asdflkj> woah: https://termbin.com/2yn4
2021-06-23 23:46:19 <trylletre> almost no gpus have open firmware.. certainly not from amd/nvidia etc
2021-06-23 23:46:48 <asdflkj> I didn't buy this PC (just a family one) I'll buy a librebooted thinkpad when I'm not broke
2021-06-23 23:47:04 <asdflkj> for now I might not mind using X unaccelerated
2021-06-23 23:47:07 <trylletre> no issue with that
2021-06-23 23:47:23 <trylletre> but also, theres no harm really in using 'non libre firmware' just to use what you have
2021-06-23 23:47:46 <trylletre> pick something that aligns with what you want when you buy it sure, but if you have some older hardware... might as well put it to use vs scrapping it
2021-06-24 00:02:34 <asdflkj> not planning on scrapping the whole pc for lack of libreboot, but I would prefer to disable the gpu to avoid loading it with nonfree firmware, which feels nasty and like a security issue (although probably far from the biggest one, particularly with the ID-10-T PEBKAC (: )
2021-06-24 00:27:09 <Hello71> why would you install xf86 drivers with sway
2021-06-24 00:27:49 <Hello71> [ 87370.840] (EE) open /dev/dri/card0: Permission denied
2021-06-24 00:28:02 <Hello71> you need to add your user to video group or set up elogind
2021-06-24 00:28:09 <Hello71> maybe setup-xorg-base should do that actually
2021-06-24 00:29:25 <asdflkj> I don't want sway, I want dwm
2021-06-24 00:30:28 <asdflkj> do I still need to add my user to video? (that sounds familiar, but I don't remember where in the docs it was buried)
2021-06-24 00:40:11 <asdflkj> I did doas adduser asdfklj video, logged out and in, reran startx, basically same error, this log: https://termbin.com/wq1q
2021-06-24 00:41:54 <asdflkj> s/error, this log:/errors in the console, but I think different in the log:/
2021-06-24 00:41:54 <alpine-meetbot> asdflkj meant to say: I did doas adduser asdfklj video, logged out and in, reran startx, basically same errors in the console, but I think different in the log: https://termbin.com/wq1q
2021-06-24 00:44:40 <LEdoian> Is there something wrong with the build service? According to https://build.alpinelinux.org the host build-edge-aarch64 has been building java-jna since Sunday.
2021-06-24 00:45:43 <c705> asdflkj: are you using dwm or startx?
2021-06-24 00:46:04 <asdflkj> c705: startx, with nothing but 'exec dwm' in .xinitrc
2021-06-24 00:46:15 <asdflkj> I don't think dwm starts x itself
2021-06-24 00:46:26 <c705> it must
2021-06-24 00:46:46 <asdflkj> starting it via startx is the dev's intended method AFAICT
2021-06-24 00:46:51 <c705> either way, i think you're still failing to load the video modules
2021-06-24 00:47:05 <Hello71> iirc this is related to udev somehow
2021-06-24 00:47:24 <Hello71> did you reboot after setup-xorg-base
2021-06-24 00:47:37 <Hello71> c705: no and no
2021-06-24 00:47:47 <asdflkj> no, I don't have writable storage
2021-06-24 00:48:29 <c705> does dwm replace X?
2021-06-24 00:48:41 <Hello71> i think not that many people use xorg on diskless
2021-06-24 00:49:00 <Hello71> try apk add xf86-video-ati
2021-06-24 00:49:05 <c705> this person is trying to
2021-06-24 00:49:20 <c705> and for ther record, i'm running sway on diskless, not that matters
2021-06-24 00:51:47 <asdflkj> same-looking console error, here's the log: http://termbin.com/4a8r
2021-06-24 00:52:10 <asdflkj> I think the problem is KMS is not set up right
2021-06-24 00:52:17 <Hello71> no
2021-06-24 00:52:26 <Hello71> [ 94999.098] (EE) Caught signal 6 (Aborted). Server aborting
2021-06-24 00:53:57 <asdflkj> [ 94999.007] (II) [KMS] Kernel modesetting enabled.
2021-06-24 00:53:58 <asdflkj> [ 94999.007] (WW) Falling back to old probe method for modesetting
2021-06-24 00:54:26 <Hello71> yeah that's because installed ati
2021-06-24 00:55:02 <asdflkj> oh, I should modprobe that I guess
2021-06-24 00:55:12 <Hello71> there's no ati module
2021-06-24 00:55:23 <Hello71> it's radeon but loaded automatically
2021-06-24 00:55:28 <c705> it's either amdgpu or raedon
2021-06-24 00:55:28 <asdflkj> modprobe amdgpu*
2021-06-24 00:57:26 <asdflkj> as dhttps://wiki.alpinelinux.org/wiki/Radeon_Video#Kernel_Modesetting_.28KMS.29
2021-06-24 00:57:39 <asdflkj> s/as d//
2021-06-24 00:57:39 <alpine-meetbot> asdflkj meant to say: https://wiki.alpinelinux.org/wiki/Radeon_Video#Kernel_Modesetting_.28KMS.29
2021-06-24 00:59:23 <Hello71> your gpu is NI, amdgpu doesn't support terascale
2021-06-24 00:59:41 <Hello71> get a core dump i guess
2021-06-24 00:59:47 <Hello71> ulimit -c unlimited; startx
2021-06-24 01:04:42 <asdflkj> https://termbin.com/gdxi
2021-06-24 01:07:46 <Hello71> i mean that's just the same log again?
2021-06-24 01:07:56 <Hello71> oh, that's what you thought i meant
2021-06-24 01:08:09 <Hello71> like, gdb core.*, bt, etc
2021-06-24 01:08:42 <asdflkj> I don't know how to use gdb, I have uploaded the core file to 0x0.st if you want it
2021-06-24 01:10:12 <asdflkj> it shouldn't have sensitive stuff in it, should it?
2021-06-24 01:33:06 <asdflkj> gosh darn alpine xorg. at this rate maybe I might as well get openbsd or hyperbola for xenocara and install them to my flash drive; they may write it to death, but I should be able buy a new drive in almost less time than this is wasting :(
2021-06-24 01:33:25 <asdflkj> well, the morning is far wiser than the night, hopefully see you then o/
2021-06-24 01:49:32 <AppAraat[m]> 
ACTION uploaded an image: (5KiB) < https://matrix.org/_matrix/media/r0/download/matrix.org/gNdVisJBcLVSPNpJsljwasuC/clipboard.png >

2021-06-24 01:49:41 <AppAraat[m]> Why is the installer dividing by zero again?
2021-06-24 02:01:51 <zcrayfish> I'd guess that the random selection does not work when the mirrors.txt file cannot be downloaded.
2021-06-24 02:03:33 <zcrayfish> line 30 in /sbin/setup-apkrepos 
2021-06-24 02:08:34 <AppAraat[m]> `local random_mirror_index=$(( $RANDOM % $mirror_count ))` - ye
2021-06-24 02:08:56 <AppAraat[m]> Though I assumed that `local` was a bashism 🤔
2021-06-24 02:14:00 <AppAraat[m]> yeah I don't see it at https://pubs.opengroup.org/onlinepubs/9699919799/utilities/V3_chap02.html but it is in the /bin/dash man page, so I guess they must have added it in. If that's the case then I'm sure busybox sh also supports that.
2021-06-24 02:17:28 <Hello71> i'm pretty sure busybox ash predates dash
2021-06-24 02:18:11 <zcrayfish> busybox ash supports /some/ bashisms.
2021-06-24 02:19:35 <c705> very few in my experience
2021-06-24 02:19:48 <zcrayfish> Just enough to make it usable xD
2021-06-24 02:28:19 <xordspar0> Busybox ash is a fork of dash. Also, "local" in ash variants predates bash, or at least evolved around the same time.
2021-06-24 02:28:50 <AppAraat[m]> TIL!
2021-06-24 02:33:00 <xordspar0> Both ash and bash started development in 1989. Ash had local in its first release. Bash had it at the time of its 2.0 release. I can't find changelogs from before that...
2021-06-24 02:33:43 <xordspar0> This page is fun. I love digital history stuff like this. I'm glad that someone did the work to preserve this. https://www.in-ulm.de/~mascheck/various/ash/
2021-06-24 02:36:20 <Hello71> hm, that makes sense
2021-06-24 02:36:24 <AppAraat[m]> Interesting, I thought ash was aiming for strict POSIX compatibility but apparently not.
2021-06-24 02:37:08 <Hello71> also worth noting that lots of "bashisms" are really ksh/zsh features imported to bash
2021-06-24 02:39:58 <AppAraat[m]> Also: I figured out why I didn't have network (in fact, I didn't even have the eth0 interface!) I booted a VM using `alpine-virt-3.14.0-x86_64.iso` and in the `virt-manager` I specified the NIC device model to be `rtl8139` but after I set it to `virtio`, eth0 appeared upon boot.
2021-06-24 03:11:19 <xordspar0> Oh, also, ash predates POSIX shell. The shell was specified in POSIX in 1992, while ash started in 1989. So it wouldn't be fair to say that ash aims for strict POSIX compliance. :) It's a conservative shell, but it does have some features that bourne shell didn't have.
2021-06-24 03:17:34 <AppAraat[m]> wut
2021-06-24 03:17:53 <AppAraat[m]> now that's what I call a plot twist
2021-06-24 07:59:09 <gopher22> Hello fellow alpinists!
2021-06-24 08:01:12 <gopher22> Has anyone tried the package shellinabox? I got it running, but it's not clear for me how to pass the options for the shellinaboxd daemon. I have opened /etc/init.d/shellinaboxd. There is avalue, command_args="${SIAB_OPTS}". How can I pass the concrete settings for the daemon using this variable? Should I hard-code the daemon args into this rc file
2021-06-24 08:01:12 <gopher22> directly?
2021-06-24 08:01:46 <yyp> gopher22: they are in /etc/conf.d/shellinabox
2021-06-24 08:02:02 <yyp> *shellinaboxd
2021-06-24 08:03:41 <gopher22> oooh...aah... *akward surpsied noises intensify*. thank you! I always forget about the presence of /etc/conf.d
2021-06-24 08:04:20 <gopher22> first reaction for me is /etc/nameofyourapporservice. if that's non-existent then on to the rc.d hackeries :)
2021-06-24 08:07:59 <ikke> /etc/conf.d is for service related settings 
2021-06-24 08:08:23 <gopher22> got it
2021-06-24 08:57:07 <yyp> How/where the Rasbperry Pi image (the one on alpinelinux.org/downloads) is built?
2021-06-24 08:57:44 <ikke> On our arm builders 
2021-06-24 08:59:08 <ikke> In the aports repository there is a scripts directory 
2021-06-24 09:14:10 <user8712> hey, I was here about dbus problems yesterday, `dbus-dev` solved it because it provides libdbus-1.so. Odd that I need a dev package for dbus to work. I initially followd the pipewire installation tutorial which didn't mention dbus-dev, just throwing this out here. :)
2021-06-24 09:35:21 <user8712> but I do have another issue that I have encountered on another distro a while ago (arch), the startx command does not run i3, I have .xinitrc and .xserverrc in my home directory and they are excecutable but they don't seem to run. I can initiate the 'stack' with startx, then DISPLAY=:0 i3 but it won't automagically work right now. Also noting, it's a rootless Xorg setup.
2021-06-24 09:36:36 <trylletre> you do have 'exec i3' in startx and not some other window manager right
2021-06-24 09:37:02 <user8712> I've tried with and without exec. I only have i3, yes.
2021-06-24 10:27:40 <ikke> user8712: You should certainly not need -dev packages to run things
2021-06-24 11:10:28 <zcrayfish> must be a bug, that package definitely contains the shared library that he mentioned.
2021-06-24 13:24:15 <LEdoian> As per https://build.alpinelinux.org/ the host build-edge-aarch64 seems to be stuck at building java-jna since Tuesday.
2021-06-24 14:22:52 <fluix> it seems to have failed but didn't get detected
2021-06-24 15:03:33 <q66> is it possible to check if a package is available in repositories, disregarding its installed status? (i.e. if it's installed but not available in repos, it should be disregarded)
2021-06-24 15:06:42 <lopid> looks like `apk info -e package` will return success if package is installed, failure otherwise
2021-06-24 15:07:10 <lopid> man apk-info
2021-06-24 15:07:39 <lopid> also failure for an unknown package
2021-06-24 15:07:55 <lopid> hopefully this helps you
2021-06-24 15:09:57 <lopid> you might also try simulating an install
2021-06-24 15:14:29 <q66> that wasn't the question though
2021-06-24 15:14:42 <q66> i want to check explicitly whether the package is in repos
2021-06-24 15:14:46 <q66> regardless of its installed status
2021-06-24 15:15:17 <q66> i.e. installed but not in repos => false, not installed and not in repos => false, installed and in repos => true, not installed and in repos => true
2021-06-24 15:16:11 <q66> the problem is that apk info even without -e will not fail if the package is not in repos but also is installed
2021-06-24 15:16:23 <q66> so if it's installed, there doesn't seem to be a way to check if it's in repos or not
2021-06-24 15:17:10 <lopid> apk search -e package ? however, it always seems to return success, but you could use its output
2021-06-24 15:17:36 <q66> well, one could do some hack like, init a db in an empty directory, then point --root to it, and --repositories-file to the actual repos
2021-06-24 15:17:41 <q66> but that's not great
2021-06-24 15:19:40 <MY-R> q66: check: apk -h list
2021-06-24 15:20:37 <q66> yeah i guess apk list's output could be used
2021-06-24 15:21:31 <MY-R> still not great but should fit with some tweaks
2021-06-24 15:21:38 <yyp> Hmm, apk list -a will show locally installed packages that are not in the repos
2021-06-24 15:24:28 <q66> well, apk list appears to print both what's installed and what's available
2021-06-24 15:24:30 <q66> which is... fine
2021-06-24 15:25:51 <q66> but actually lopid was closer
2021-06-24 15:25:58 <q66> apk search -e -o -q <foo>
2021-06-24 15:26:01 <q66> does exactly what i want
2021-06-24 15:26:20 <q66> it will print nothing if not in repo
2021-06-24 15:26:23 <q66> and the package name if in repo
2021-06-24 15:26:44 <q66> -e will do an exact match (i.e. not a substring)
2021-06-24 15:27:13 <yyp> Hmm, apk search is not documented but it works (neither man page nor --help)
2021-06-24 15:27:15 <q66> i guess i don't even need -o
2021-06-24 15:27:16 <lopid> where is -o documented? `apk search` is the only form that doesn't have a man page, i think
2021-06-24 15:27:17 <q66> just -e -q
2021-06-24 15:27:31 <q66> -o will print origin (in case of subpackage)
2021-06-24 15:27:34 <q66> which i don't need
2021-06-24 15:27:39 <q66> or rather, don't care for
2021-06-24 15:27:44 <q66> i just want it to check existence
2021-06-24 15:27:46 <q66> so -e -q
2021-06-24 15:27:51 <MY-R> lopid: in sources :D
2021-06-24 15:27:57 <lopid> indeed
2021-06-24 15:27:58 <q66> apk search is not documented in 2.x, but it's documented in 3.x
2021-06-24 15:28:05 <q66> i figured it's not available in 2.x at all, but turns out it is
2021-06-24 15:28:08 <q66> which simplifies things for me
2021-06-24 15:28:37 <q66> 
ACTION is building a distribution on apk and needs this in the build system (not using abuild)

2021-06-24 15:29:11 <lopid> it isn't here, in 3.13.5…
2021-06-24 15:29:24 <q66> https://github.com/alpinelinux/apk-tools/blob/master/doc/apk-search.8.scd
2021-06-24 15:29:36 <lopid> huh
2021-06-24 15:29:37 <q66> documented 14 days ago
2021-06-24 15:29:51 <lopid> doh
2021-06-24 15:30:14 <q66> anyway, thanks
2021-06-24 15:42:23 <q66> ok, works perfectly
2021-06-24 15:42:57 <lopid> great
2021-06-24 15:45:29 <lopid> oh, i see now. apk-tool version 2 vs 3, not alpine version O:)
2021-06-24 15:45:40 <q66> yes
2021-06-24 15:45:57 <q66> btw, while at it, any way to get the origin repository of a package?
2021-06-24 15:46:18 <lopid> apk info -a
2021-06-24 15:46:41 <lopid> oh, apparently not
2021-06-24 15:46:42 <q66> yeah it's not there as far as i can tell
2021-06-24 15:46:57 <q66> since it's not metadata
2021-06-24 15:47:03 <q66> but apk should know of it anyway
2021-06-24 15:47:36 <q66> this one is not super important but would be nice for informational reasons
2021-06-24 15:48:45 <q66> i guess one could just loop over lines of the repositories file and then check apk search for each, but doesn't sound great
2021-06-24 15:49:10 <lopid> https://pkgs.alpinelinux.org/packages :/
2021-06-24 16:00:22 <q66> ?
2021-06-24 16:00:55 <lopid> shows you which repo packages are in
2021-06-24 16:04:07 <q66> no it doesn't
2021-06-24 16:04:12 <q66> since alpine packages are not my packages
2021-06-24 16:04:18 <lopid> oh
2021-06-24 16:04:27 <Hello71> apk policy
2021-06-24 16:05:30 <Hello71> pkgs.alpinelinux.org is lua based, i think apk-tools has a lua api
2021-06-24 16:12:09 <q66> well, this is irrelevant to me anyway
2021-06-24 16:12:30 <q66> it's also not very important in the end, guess i just won't be printing the repo path for found dependencies during build
2021-06-24 20:26:12 <user8712> hey, does rootless xorg currently function on alpine? I thought I was running it rootless before but I checked and it's not the case. This in return causes my startx not to initialize my wm.
2021-06-24 20:26:45 <user8712> im on 3.13.5 fwiw
2021-06-24 20:27:38 <ikke> should be enabled: add17bdb3ef03ac44c05b51fa0ed0b99b4c336a7
2021-06-24 20:27:52 <ikke> it has temporarily reverted, but later enabled again
2021-06-24 20:28:45 <ikke> It mentions Xorg.wrap
2021-06-24 20:29:05 <user8712> well I stripped the suid bit from my Xorg bin and now it refuses to start because I don't have tty permissions. 
2021-06-24 20:30:58 <user8712> I added myself to the tty group but it still complains about not having rights for tty7, (no suid, no tty group) (EE) parse_vt_settings: Cannot open /dev/tty0 (Permission denied), (no suid, tty group) (EE) xf86OpenConsole: Cannot open virtual console 7 (Permission denied)
2021-06-24 20:42:35 <user8712> okay, I put the suid back and ran startx from the device console (now the X is running rootless) but my wm is still not being started despite the `ps` showing it's 'using' my local .xinitrc
2021-06-24 20:46:59 <invoked> that reminds me: is xlsclients(1) somewhere not obvious? asking for a friend.
2021-06-24 20:55:36 <user8712> dumbest fix, I made an empty .xinitrc instead of copying from /etc/X11/xinit/ and it started my wm. oh the time lost on this :)
2021-06-24 22:54:20 <AppAraat[m]> "To get the latest list of available packages, use the update command. The command downloads the `APKINDEX.tar.gz` from each repository and stores it in the local cache, typically `/var/cache/apk/`, `/var/lib/apk/` or `/etc/apk/cache/`. " - Why 3 different dirs?
2021-06-24 22:54:40 <AppAraat[m]> (From https://wiki.alpinelinux.org/wiki/Alpine_Linux_package_management)
2021-06-24 23:21:02 <AppAraat[m]> Ooh `apk info --who-owns <package>` is super useful. I've missed this feature quite a lot when working with `apt`.
2021-06-24 23:28:32 <Hello71> dpkg -S
2021-06-24 23:29:06 <Hello71> although i would assume that apk and pacman are faster than dpkg and rpm
2021-06-24 23:47:59 <AppAraat[m]> Oh hmm, so I guess that's not the feature I was missing. I think I got it confused with searching for an uninstalled package according to a filename that the package contains. I believe I had to use a 3rd party tool for that.
2021-06-24 23:56:38 <asdflkj> IIUC it is a separate tool from apt, but it's installed by default on Ubuntu and runs whenever you trigger 'command not found'
2021-06-25 00:25:05 <AppAraat[m]> Interesting, I should look into that.
2021-06-25 00:59:27 <Hello71> command-not-found is different from apt-file
2021-06-25 01:00:32 <Hello71> i don't know of any package managers that fetch file lists by default, and only know of pacman that even has it integrated by default
2021-06-25 01:00:52 <Hello71> and even pacman's isn't all that good
2021-06-25 01:07:17 <quadrat> rpm does it
2021-06-25 01:18:07 <Sheila> there's an apk impl, iirc.
2021-06-25 01:25:46 <Hello71> huh. i mean technically yum/dnf, but still
2021-06-25 11:29:03 <rapha> hi all!
2021-06-25 11:29:12 <rapha> how can i wget the package listed at https://pkgs.alpinelinux.org/package/edge/main/x86/libarchive-tools ?
2021-06-25 11:47:27 <lopid> get it from the git repo
2021-06-25 12:40:47 <rapha> th
2021-06-25 12:40:48 <rapha> x
2021-06-25 12:43:36 <rapha> lopid: nope, that doesn't help ... the git only contains an APKBUILD file, no binary
2021-06-25 12:45:43 <rapha> also installed it on another box and tried to `find` it there in the file system ... to no avail :(
2021-06-25 12:47:34 <lopid> /usr/bin/bsd{cat,cpio,tar}?
2021-06-25 12:50:49 <rapha> ah, right, i could just copy the binary i need from somewhere ... no need for the .apk
2021-06-25 12:51:04 <rapha> (was looking for the .apk)
2021-06-25 12:51:22 <rapha> hmm ... this is going to be a fun ssh tunnel over 5 hops or so, but oh well
2021-06-25 12:53:31 <rapha> worked, and the binaries run fine :)
2021-06-25 12:53:33 <lopid> how about apk fetching it?
2021-06-25 12:53:38 <lopid> ok
2021-06-25 12:53:44 <rapha> aha, a new command!
2021-06-25 12:53:47 <rapha> also need unrar
2021-06-25 12:53:51 <rapha> so will try the apk fetch thing
2021-06-25 12:54:11 <lopid> with -R to get the deps
2021-06-25 12:54:23 <rapha> oh nice
2021-06-25 12:54:32 <lopid> although, they may be different on a different installation
2021-06-25 12:54:45 <rapha> btw, is there a trick to apk upgrade from 3.12 to 3.13 if you're blocked by the libffi versions dependency issue?
2021-06-25 12:54:56 <lopid> i didn't know about that
2021-06-25 12:55:07 <rapha> (the "source" box is still on 3.12, the "target" box already on 3.14)
2021-06-25 12:55:14 <lopid> i might remember reading being able to hold specific packages
2021-06-25 12:55:19 <rapha> lopid: https://gitlab.alpinelinux.org/alpine/aports/-/issues/11605 ... at least i _think_ i'm hitting that
2021-06-25 12:55:53 <rapha> ah, now, was because of other stuff back then
2021-06-25 12:56:46 <rapha> bit reluctanct to delete qemu with a bunch of VMs running ... although it should be fine, really
2021-06-25 12:57:36 <rapha> samba and nfs too ... basically all my services :(
2021-06-25 12:57:54 <rapha> do the "apk del" scripts tend to stop things?
2021-06-25 12:58:17 <lopid> can't say i've noticed
2021-06-25 12:58:22 <lopid> (so i won't)
2021-06-25 12:58:31 <lopid> upgrading doesn't restart them
2021-06-25 12:59:11 <rapha> hmm
2021-06-25 12:59:12 <lopid> maybe they have a deletion script that you can see
2021-06-25 12:59:23 <rapha> sad that there's no "apk del --force-broken-dependencies"
2021-06-25 13:01:42 <rapha> devserver != productionserver ... will just try
2021-06-25 13:22:16 <rapha> hmm. i had a "libguestfs" installed from testing, that has now magically disappeared from pkgs.alpinelinux.org (google still knows the URLs, but they've become 404s) ... what was that and what was it good for?
2021-06-25 13:44:20 <insep> 3665d7c1
2021-06-25 13:44:31 <insep> 3665d7c1
2021-06-25 13:44:40 <insep> wait how do i use that boot
2021-06-25 13:44:43 <insep> *bot
2021-06-25 13:44:50 <insep> https://gitlab.alpinelinux.org/alpine/aports/-/commit/3665d7c1c7bb325fd836657b2804274b23732acc
2021-06-25 13:44:50 <ikke> You need to paste the full hash 
2021-06-25 13:45:01 <insep> 3665d7c1c7bb325fd836657b2804274b23732acc
2021-06-25 13:45:10 <insep> i see
2021-06-25 14:41:22 <rapha> uh shit
2021-06-25 14:41:34 <rapha> bootign with init=/bin/bash and the keyboard won't work
2021-06-25 14:41:42 <rapha> did something recently change in the initrd?
2021-06-25 14:41:51 <rapha> (-lts)?
2021-06-25 14:45:06 <rapha> and i'm on the out-of-band-controller's keyboard; not even USB
2021-06-25 14:46:05 <rapha> help, anyone?
2021-06-25 14:50:21 <rapha> booting with "single" => same problem
2021-06-25 14:50:32 <rapha> it seems the initrd doesn't have keyboard drivers anymore?!
2021-06-25 14:52:13 <rapha> physical usb keyboard also doesn't work
2021-06-25 14:52:15 <rapha> whaaa
2021-06-25 14:53:40 <rob0> hmm, I encountered things like that when using USB
2021-06-25 14:54:27 <rob0> but if you have a PS/2 connector that should work for most sane kernels, I would think
2021-06-25 14:55:32 <rapha> nope, no ps2 connector
2021-06-25 14:56:00 <rapha> rob0: also, working on an out-of-band controller (java applet thingy) ... so this may technically not be usb at all
2021-06-25 14:56:13 <rob0> ah
2021-06-25 14:56:19 <rapha> and it'd be sad if alpine's -lts kernel wasn't "sane" :-(
2021-06-25 14:56:43 <invoked> what host device is this
2021-06-25 14:57:25 <rapha> invoked: not sure what you mean. the machine is a Dell PowerEdge T320
2021-06-25 14:57:37 <invoked> that's what i meant :)
2021-06-25 14:57:41 <rapha> ok :)
2021-06-25 14:57:52 <rapha> been using single mode multiple times before on it
2021-06-25 14:58:20 <invoked> sbcs have been funky with usb keyboards, hence my question
2021-06-25 14:58:41 <rapha> makes sense
2021-06-25 14:59:04 <rapha> will try with modules=usbcore,usbhid acpi=off nomodeset single just for the heck of it
2021-06-25 14:59:15 <rapha> hmm, no dice
2021-06-25 14:59:27 <rapha> even the numpad LED on the physical keyboard doesn't toggle
2021-06-25 14:59:57 <invoked> are you sure usb is working
2021-06-25 15:01:09 <rapha> physically? yes
2021-06-25 15:45:40 <rapha> invoked: perhaps i should mention explicitly that the USB keyboard is working in GRUB.
2021-06-25 15:53:12 <Hello71> i think mkinitfs hasn't included usb keyboard drivers for a while
2021-06-25 15:53:50 <Hello71> iirc there is an issue open for that somewhere
2021-06-25 16:09:41 <rapha> then it should be a matter of booting with a kernel command line including modules=usbhid Hello71, but that doesn't help unfortunately
2021-06-25 16:09:58 <rapha> and fwiw i just booted the latest alpine .iso and the keyboard is working just fine
2021-06-25 16:10:42 <rapha> unfortunately that includes no dhcp client so i'm dead in the water again because it also doesn't include LVM
2021-06-25 16:10:44 <rapha> jesus
2021-06-25 16:11:49 <rapha> ah, nm ... starts with a "u", not a "d"
2021-06-25 16:11:51 <rapha> my bad
2021-06-25 16:15:53 <rapha> then after lvm is installed and the scan is successfully run, no /dev/<volgroup>/ directory ...wth
2021-06-25 16:17:46 <rapha> and --mknodes doesn't actually make any nodes ... doesn't get much stranger
2021-06-25 16:19:33 <rapha> nm, forgot -a y ... little panicked here because of the whole t hing
2021-06-25 16:28:33 <Hello71> no, the initramfs doesn't contain the usb keyboard modules at all
2021-06-25 16:28:38 <Hello71> it's not that it doesn't load them
2021-06-25 16:42:37 <rapha> anyhow, something is happening when not booting in single mode which makes the keyboard work
2021-06-25 16:43:14 <rapha> also the "fix" to #220747 in /etc/init.d/nfs isn't working properly and just causes killall usage instructions to be displayed
2021-06-25 16:49:19 <programmerq> are there any other distros that use the apk package manager? it'd be neat having apk+glibc+systemd
2021-06-25 16:49:35 <programmerq> (just academic curiosity-- I'm not trying to solve a specific problem)
2021-06-25 17:14:10 <ikke> adelie
2021-06-25 17:15:04 <q66> i'm working on one
2021-06-25 17:15:27 <q66> (no glibc/systemd though)
2021-06-25 17:32:35 <rapha> apk, no glibc, no systemd ... i think i've heard of such a distro ... was named alpine or something like that
2021-06-25 17:33:18 <PureTryOut> programmerq: there are multiple distros, or at least one other, using apk, but no distro using apk in combination with glibc or systemd
2021-06-25 17:33:27 <PureTryOut> There is no reason apk couldn't handle such a distro though
2021-06-25 17:33:39 <q66> rapha: not quite
2021-06-25 17:33:48 <q66> i also have no gcc or other gnu components
2021-06-25 17:34:10 <q66> and abuild is not used
2021-06-25 17:35:08 <q66> i originally intended to use a different package manager, but ended up settling on apk-tools after it proved decent for the purpose
2021-06-25 17:36:20 <ikke> There is/was one individual who was working on a glibc port of alpine
2021-06-25 17:37:20 <rapha> q66: wait, no gnu components at all? that can't be easy...
2021-06-25 17:37:30 <ikke> gcc
2021-06-25 17:37:33 <q66> well, right now there's gnu make for a few components
2021-06-25 17:37:44 <q66> i have plans to deal with that situation
2021-06-25 17:38:01 <q66> and there's just a few coreutils left, but that's a short term thing
2021-06-25 17:38:04 <rapha> short of writing a gnu make replacement, what could you do?
2021-06-25 17:38:11 <q66> writing a gnu make replacement
2021-06-25 17:38:17 <rapha> and this sounds political also ...
2021-06-25 17:38:30 <q66> (and a lot of things can use bmake alright)
2021-06-25 17:38:41 <q66> (mostly stuff that has handwritten makefiles can't, and there's not that much of it)
2021-06-25 17:38:44 <ikke> there is a make fork
2021-06-25 17:38:50 <rapha> ah, thought so. "no replacement for more displacement than more displacement" as the old adage goes
2021-06-25 17:39:13 <rapha> other than*
2021-06-25 17:39:41 <q66> also there's nothing wrong with being political :P
2021-06-25 17:39:47 <q66> but it's also an experimental project
2021-06-25 17:39:58 <rapha> i didn't mean to say there was
2021-06-25 17:40:40 <rapha> and there's something to be said for people who proclaim freedom and right thereafter proclaim The One True Way, so...
2021-06-25 17:41:25 <q66> well, i don't proclaim anything, i'm just working on a project, people can take that as they will :P
2021-06-25 17:43:02 <q66> either way it's an experimental project to seek 1) an alternative userland 2) a fresh package build system 3) an LLVM-based whole linux system build
2021-06-25 17:43:16 <q66> i wouldn't say any of these things are not worth pursuing or not interesting
2021-06-25 17:43:31 <rapha> not at all
2021-06-25 17:43:47 <rapha> wondering about (2) though ... what is your motivation there?
2021-06-25 17:44:37 <q66> i think shell is a bad language to write a package build system in, since it's slow, not introspectable, and soon becomes complex and difficutl to maintain
2021-06-25 17:44:54 <q66> so i'm writing one that does not utilize shell
2021-06-25 17:46:03 <q66> it cuts down the overhead considerably, e.g. musl build with my old shell based system would take 21 secs while now it takes 9 secs (the rest was just the overhead of the build system which has been reduced to nothing)
2021-06-25 17:46:45 <q66> i also want to be able to do things like, have a source package collection with many thousand templates, and be able to process it via API (let's say, generate some statistics out of it) and do it all at once with no preprocessing or caching and have it take seconds instead of hours
2021-06-25 17:49:11 <q66> ./xbps-src -c ppc64le-musl -j16 -t pkg musl  87.53s user 25.95s system 527% cpu 21.523 total
2021-06-25 17:49:11 <q66> python3 cbuild.py -j16 pkg musl  74.52s user 17.96s system 1040% cpu 8.886 total
2021-06-25 17:50:44 <q66> of course, shell overhead can be cut down too if you take some care, and there can be dramatic differences, but in the end it'll always be slow in general
2021-06-25 17:54:00 <rapha> hmm okay that's a compelling argument :)
2021-06-25 17:54:13 <rapha> why not something like rust then, though?
2021-06-25 17:55:06 <q66> because an interpreted language is 1) portable and robust 2) quicker to get going
2021-06-25 17:55:14 <ikke> exactly
2021-06-25 17:55:20 <ikke> you need to thing about the bootstrapping scenario
2021-06-25 17:55:40 <ikke> the more dependencies, the longer it takes before you can actual leverage the build system
2021-06-25 17:55:46 <q66> yes
2021-06-25 17:56:01 <q66> python works on mostly every platform imaginable, and a good enough setup can be built with pretty much no dependencies
2021-06-25 17:56:59 <ikke> https://tpaste.us/jnEm
2021-06-25 17:57:07 <q66> a rust thing would've grown into a monstrosity with 400 statically linked cargo dependencies
2021-06-25 17:57:24 <q66> ikke: for python? you can cut that down a lot
2021-06-25 17:58:29 <q66> mostly need pkg-config + libffi, openssl, bzip2/zlib/liblzma, plus a compiler and that's all it can build with
2021-06-25 17:58:34 <rapha> seems i need to stop thinking of rust as a viable C replacement
2021-06-25 17:59:02 <q66> it would've been if cargo didn't suck
2021-06-25 17:59:17 <q66> but it does, and it's the main reason i probably won't ever use it for anything
2021-06-25 17:59:24 <q66> the language is pretty solid
2021-06-25 17:59:36 <q66> but even if it was a viable C replacement it still wouldn't have been a good fit for this
2021-06-25 17:59:52 <ikke> Portability is also an issue
2021-06-25 17:59:58 <q66> yes, definitely
2021-06-25 17:59:59 <ikke> we still do not have rust for all of our arches
2021-06-25 18:00:13 <q66> we have it for all in void, but it's pretty clear that on some it's definitely tier 3
2021-06-25 18:01:02 <q66> by all i mean x86_64/armv67/aarch64/ppc64le/ppc64/ppc and i think it worked on mips at one point too
2021-06-25 18:01:10 <ikke> But that's with glibc?
2021-06-25 18:01:14 <q66> glibc and musl
2021-06-25 18:01:19 <ikke> Hmm, ok
2021-06-25 18:01:25 <ikke> s390x is missing, right
2021-06-25 18:01:28 <ikke> for us as well
2021-06-25 18:01:35 <q66> yeah void does not have s390x
2021-06-25 18:01:45 <q66> i think rust does work on s390x
2021-06-25 18:02:32 <q66> 
ACTION will attempt to port it to 32-bit little endian powerpc this weekend

2021-06-25 18:02:45 <q66> since it's blocking our ppcle port in a few places
2021-06-25 18:03:17 <q66> and i guess people interested in m68k etc are going to have a problem as well
2021-06-25 18:04:30 <rapha> hmm
2021-06-25 18:04:48 <ikke> We need to port it to rust
2021-06-25 18:04:59 <ikke> riscv64*
2021-06-25 18:05:27 <rapha> question about OpenRC ... if there's an 'after foo' in bar's depend(), why would bar ever run before foo? (i even rc-update del'd and rc-update add'd both of them)
2021-06-25 18:07:28 <q66> anyway, my distro thingy can bootstrap itself from source on either a void-musl system or alpine system or any other musl system, and you basically need... a python installation, scanelf (from pax-tools), apk-utils, openssl (for signing), git (not entirely mandatory), bubblewrap, tee, + for host dependencies for bootstrapping: clang (enabled with libc++ and compiler-rt and lld), cmake+meson, pkg-config, make+bmake, ninja, byacc, flex, perl and m4
2021-06-25 18:07:32 <q66> ideally i'd like a smaller bootstrap set
2021-06-25 18:07:38 <q66> but i think can't do anything much more about that :/
2021-06-25 18:08:07 <q66> (once bootstrapped, you only need the former set)
2021-06-25 18:08:33 <q66> i think cmake is the ugliest dependency but since llvm itself builds with that, nothing i can do
2021-06-25 18:10:06 <Hello71> what do you need scanelf for that llvm-readelf can't do?
2021-06-25 18:10:44 <q66> well scanelf is more convenient, i could probably write a shlib scanner about llvm-readelf fine
2021-06-25 18:10:46 <rapha> q66: sounds like you already put enough work in to try and build a community now.
2021-06-25 18:10:47 <q66> maybe as a fallback
2021-06-25 18:11:08 <q66> since scanelf is going to be much faster (it has a recursive scan and pretty much the exact output i need all at once)
2021-06-25 18:11:28 <Sheila> have you looked at using age/rage for signing instead?
2021-06-25 18:11:46 <q66> with llvm-readelf i'd have to recursively glob the destdir, filter elf files, then readelf and parse the output for SONAME and NEEDED
2021-06-25 18:12:10 <q66> scanelf also gives me info about textrels without any more work which is nice
2021-06-25 18:12:20 <Hello71> hm, i thought llvm-readelf was vaguely binutils compatible
2021-06-25 18:12:36 <Hello71> like lld
2021-06-25 18:13:02 <q66> Sheila: no
2021-06-25 18:13:14 <q66> age is written in go so it's outright a nope
2021-06-25 18:13:33 <Sheila> yeah. I keep hoping someone does a C/C++ version.
2021-06-25 18:13:40 <q66> and rage is rust which is only a little bit better when it comes to portability/bootstrapping
2021-06-25 18:13:54 <Sheila> (won't be me because I'm not comfortable writing crypto code.)
2021-06-25 18:14:46 <q66> yeah can't say i feel like it either :P
2021-06-25 18:14:54 <q66> i'll leave it to someone actually good at crypto
2021-06-25 18:17:39 <q66> Hello71: well tbh pax-tools is not much of a dependency considering all you need for it is a C compiler
2021-06-25 18:17:44 <Hello71> mmhmm
2021-06-25 18:17:48 <q66> and the whole suite is like 400kB installed size
2021-06-25 18:17:53 <q66> and you need one binary out of it
2021-06-25 18:18:26 <q66> i actually previously had a shlib scanner written around objdump
2021-06-25 18:18:28 <q66> but i dropped it
2021-06-25 18:18:42 <q66> it was slower (though it was fast enough) and elftoolchain (which i use in place of binutils) does not have it
2021-06-25 18:18:53 <Hello71> llvm-objdump?
2021-06-25 18:18:58 <q66> sure
2021-06-25 18:19:09 <q66> you can use that
2021-06-25 18:19:13 <q66> though it was still subpar
2021-06-25 18:19:26 <q66> though i could've improved it
2021-06-25 18:19:42 <q66> but then, not much point
2021-06-25 18:20:01 <q66> sometimes i just forget llvm pretty much provides half of binutils by itself :)
2021-06-25 18:20:19 <q66> (it has no libelf though :P)
2021-06-25 18:20:22 <Hello71> hm, why are there so many binutils (re)implementations? libelf (obsolete), elfutils, elftoolchain, llvm...
2021-06-25 18:20:37 <q66> dunno
2021-06-25 18:20:47 <Hello71> rapha: i vaguely recall that they need to be in the same runlevel
2021-06-25 18:20:53 <q66> elftoolchain is more like a binutils *and* elfutils together
2021-06-25 18:21:06 <q66> it provides an actual libelf implementation
2021-06-25 18:21:22 <Hello71> what's wrong with elfutils?
2021-06-25 18:21:34 <q66> elfutils provides some binutils tools by itself though, and i have no idea why :P
2021-06-25 18:21:46 <q66> hm, nothing much i guess
2021-06-25 18:21:52 <q66> besides gpl3
2021-06-25 18:22:08 <q66> while elftoolchain is a bsd licensed impl of much the same thing
2021-06-25 18:22:33 <Hello71> mmhmm
2021-06-25 18:25:52 <q66> iirc elfutils had issues with llvm
2021-06-25 18:31:40 <rapha> Hello71: made sure both are, and nfs still starts before libvirtd despite having been explicitly told to start "after libvirtd" in depends()
2021-06-25 18:41:36 <Hello71> nfs sounds like remote-fs
2021-06-25 18:41:51 <Hello71> circular dependency
2021-06-25 18:49:43 <rapha> Hello71: it was a dependecy of libvirtd's which had an "after nfs", which had to be changed to a "before nfs"
2021-06-25 18:50:06 <rapha> and yeah, this will be bad in any environent where it's not made sure to not be bad
2021-06-25 18:50:12 <ikke> rapha: note that adding / removing services from a runlevel does not help anythign in this regard
2021-06-25 18:52:20 <rapha> ikke: yeah i just wasn't sure. thanks for confirming!
2021-06-26 15:50:40 <markpash> Hi, I've got a failed aports pipeline job for armhf for a new aport. I get the feeling that it has something to do with IPv6. Can someone also take a quick peek and maybe point me in the right direction? https://gitlab.alpinelinux.org/MarkPash/aports/-/jobs/424222
2021-06-26 16:01:23 <ikke> markpash: looking
2021-06-26 16:01:29 <markpash> thanks
2021-06-26 16:04:29 <ikke> markpash: I see we forgot to enable ipv6 in docker for the armhf ci runner
2021-06-26 16:04:31 <ikke> will enable that in a bit
2021-06-26 16:04:54 <markpash> ah cool, thanks for looking into it
2021-06-26 16:46:50 <ikke> markpash: restarted docker now
2021-06-26 16:48:10 <markpash> @ikke: great I'll kick off the pipeline again in a few mins and let you know if anything is wrong. Thanks again.
2021-06-26 16:48:42 <Shyim> I am trying to build the mariadb by own. I use the gitlab docker image but it fails always with Check for working C compiler: /usr/bin/gcc - broken. Idk how its running in the ci. its the same docker image :D 
2021-06-26 16:50:46 <insep> well, gcc was updated today, maybe that's the reason
2021-06-26 16:52:04 <superduck> I'm having issues with abuild - >>> ERROR: *: Could not find ./APKBUILD (PWD=/home/$USER/src/aports/main/tree/src/src/tree-1.8.0)
2021-06-26 16:52:30 <ikke> You need to run it in /home/$USER/src/aports/main/tree
2021-06-26 16:52:37 <superduck> Why is apkbuild being looked for inside the fakeroot, and why isn't abuild copying it there if it's looking for it?
2021-06-26 16:53:22 <cyberia> hello
2021-06-26 16:53:36 <cyberia> are there any guarantees around package stability in release branches?
2021-06-26 16:53:49 <ikke> cyberia: what kind of guarantees?
2021-06-26 16:54:12 <cyberia> well like, iirc, debian stable says packages only get security patches; no feature upgrades or breaking changes
2021-06-26 16:54:12 <Shyim> @insep I will try an older docker image. thanks for the tip :)
2021-06-26 16:54:38 <ikke> We have a similar policy
2021-06-26 16:55:05 <cyberia> ok thanks
2021-06-26 16:55:09 <cyberia> is it documented anywhere?
2021-06-26 17:06:54 <ikke> superduck: I can build tree with rootbld without issue
2021-06-26 17:25:33 <tiotags> is it possible to delete a commit from a merge request ? or edit it's commit message
2021-06-26 17:26:36 <lopid> i don't know about merge requests but you can manage unpushed commits by doing an interactive rebase
2021-06-26 17:27:47 <tiotags> It's already pushed
2021-06-26 17:28:53 <ikke> You can fix it locally, and then force push it
2021-06-26 17:30:05 <tiotags> so reset --soft and then push --force ?
2021-06-26 17:31:15 <tiotags> I forgot to add the package name to the commit message
2021-06-26 17:54:42 <begs> You can just do git rebase -i and reword commit message
2021-06-26 18:11:40 <superduck>  @ikke: how does abuild-rootbld work, it's an empty package for me
2021-06-26 18:12:13 <ikke> superduck: it's a meta package
2021-06-26 18:12:22 <ikke> It's just pulls in the dependencies needed
2021-06-26 18:12:27 <superduck> bubblewrap
2021-06-26 18:12:32 <ikke> right
2021-06-26 18:12:34 <superduck> I reran abuild and no change
2021-06-26 18:12:55 <superduck> >>> tree: Entering fakeroot...
2021-06-26 18:12:57 <superduck> >>> ERROR: *: Could not find ./APKBUILDY
2021-06-26 18:13:15 <superduck> I'm using fakeroot-tcp because wsl1 doesn't work with fakeroot.
2021-06-26 18:24:30 <Hello71> wsl1 doesn't work properly in general
2021-06-26 19:45:14 <tiotags> did I understand this correctly, if a merge request gets accepted all it's commits end up in the commit log of the target ?
2021-06-26 19:46:18 <ikke> Yes
2021-06-26 19:46:26 <ikke> We do not squash by default
2021-06-26 19:46:32 <ikke> but rebase commits
2021-06-26 19:47:20 <tiotags> I really want to get rid of a commit but can't seem to figure it out
2021-06-26 19:47:31 <ikke> interactive rebase, drop the commit
2021-06-26 19:47:48 <ikke> (or squash it if you want to keep the changes)
2021-06-26 19:47:49 <tiotags> do you have a link that explains it, I don't really understand
2021-06-26 19:48:06 <tiotags> I do want to keep the changes ?
2021-06-26 19:48:14 <tiotags> how do I squash ?
2021-06-26 19:48:35 <ikke> git rebase -i master # assuming you based your branch off master
2021-06-26 19:49:08 <ikke> Then you get a list of commits
2021-06-26 20:13:16 <ikke> tiotags: Ah, I see you have commited directly on your local master
2021-06-26 20:15:58 <tiotags> should I have done it differently ?
2021-06-26 20:16:45 <ikke> typically you create a dedicated branch for each merge request
2021-06-26 20:17:23 <cyberia> you could do `git rebase -i origin/master`
2021-06-26 20:17:27 <cyberia> or rename your local branch
2021-06-26 20:23:01 <tiotags> I don't really understand git enough to understand what you're trying to tell me, I have to look it up tomorrow
2021-06-26 20:23:58 <tiotags> thank you, I hope I can figure this out
2021-06-26 20:52:38 <kevint> tiotags this may be helpful, specifically the "Squashing Commits" section: https://git-scm.com/book/en/v2/Git-Tools-Rewriting-History
2021-06-26 20:57:09 <tiotags> thank you it does look like what I'm looking for
2021-06-26 21:38:46 <MathGeniusJodie> has anyone gotten libudev-zero to work with libinput and xorg?
2021-06-27 00:42:14 <whatsyourask> Can someone tell me what tool I can use to create a new "modloop-rpi4" file with the drivers I need to add?
2021-06-27 00:42:48 <whatsyourask> squashfs-tools maybe?
2021-06-27 00:46:54 <Hello71> update-kernel
2021-06-27 00:48:36 <whatsyourask> Hello71, Thx!
2021-06-27 01:10:48 <asdflkj> good news: I got xorg and dwm to start! (I think by installing xf86-video-vesa). bad news: it seemed instantly frozen, wouldn't accept any input from a usb mouse or a usb or even ps/2 keyboard, and soon the screen turned off (maybe because I tried to switch virtual terminals without a kernel module needed for kms? if so it took a long time), and still hasn't turned on :(
2021-06-27 01:21:18 <asdflkj> does the user xorg runs as need to be added to a bunch of groups (if so, which / is it documented anywhere)?
2021-06-27 02:07:08 <asdflkj> hmm, reading #suckless backlog it sounds like musl's locale weirdness may have frozen everything bc I tried to open st immediately after dwm opened
2021-06-27 04:11:09 <etothepii> Habbie - I've started updating the Raspi articles - when you get a chance, can you look at the main one, and see if you find that info on diskless vs sys more understandable? (https://wiki.alpinelinux.org/wiki/Raspberry_Pi)
2021-06-27 04:11:59 <etothepii> I still need to consolidate the duplication of SD card partitioning and sys-mode install info, but I'll tackle that later, as that involves moving info around different pages.
2021-06-27 06:01:58 <russkel> hi, is it common to compile in tree kernel modules with DKMS?
2021-06-27 06:08:48 <raven523> not really, unless you want to use a module that used to be out of tree with an older kernel
2021-06-27 06:16:23 <russkel> I just found it a pain to rebuild only a single driver from the tree, and I'll have to do the same thing next kernel update etc
2021-06-27 06:16:32 <russkel> s/rebuild/build/
2021-06-27 06:16:33 <alpine-meetbot> russkel meant to say: I just found it a pain to build only a single driver from the tree, and I'll have to do the same thing next kernel update etc
2021-06-27 06:18:51 <russkel> which lead me to believe I am doing something wrong here
2021-06-27 15:58:49 <nescafe> Hello, got a segfault on net-snmp which I could only reproduce on Alpine docker image: 1) docker run --rm -it alpine:3.13 2) apk install sudo net-snmp 3) sudo snmpd -f 4) ctrl-c (Segmentation fault (core dumped)) 5) Again: sudo snmpd -f 6) Segmentation fault (core dumped)
2021-06-27 15:59:51 <nescafe> Only after commenting out "iquerySecName   internalUser" and "rouser          internalUser" in /etc/snmp/snmpd.conf, the service starts and runs.
2021-06-27 22:11:52 <joe7> I am trying to strace a process and this is what strace shows. Nothing happens even though the process is still running.. http://okturing.com/src/11475/body
2021-06-27 22:12:07 <joe7> any suggestions on how to see the ptrace of what the process is doing?
2021-06-27 22:24:26 <joe7> what is the alpine equivalent of ps -efL ? to show all the threads of a process.
2021-06-27 22:24:43 <Hello71> strace -f
2021-06-27 22:24:48 <Hello71> ps --help
2021-06-27 22:26:07 <joe7> thanks.
2021-06-27 23:19:00 <whatsyourask> Hello!, When I use update-kernel --modloopfw FILE, it works but does not add the firmware to the generated modloop-rpi4 file. Could someone tell me where I have to put the firmware file?
2021-06-27 23:36:25 <Hello71> https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/update-kernel.in#L296
2021-06-27 23:40:54 <whatsyourask> I am looking at exactly that file but I am not able to understand very well. The paths where it looks for firmware are read-only. 
2021-06-27 23:48:27 <Hello71> seems like it copies the files from /lib/firmware
2021-06-27 23:49:19 <Hello71> what files are you trying to add anyways
2021-06-28 00:12:01 <MY-R> isn't enough to copy firmware file to '/lib/firmware/' then do 'lbu include /lib/firmware/blabla.bin' and 'lbu commit' ?
2021-06-28 00:14:48 <whatsyourask> dont work
2021-06-28 00:24:16 <Hello71> i think you need to reprobe
2021-06-28 00:33:32 <whatsyourask>  /lib/firmaware and /.modloop/modules/firmware/brcm are read-only
2021-06-28 00:35:35 <whatsyourask> https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/update-kernel.in#L328
2021-06-28 00:37:02 <whatsyourask> If I can identify what arguments the script passes to the command I might be able to do it myself. I create a duplicate /lib/firmware folder where I have write permissions and add the firmware I need.
2021-06-28 00:57:07 <whatsyourask> Hello71, MY-R,  Thank you very much for your help, I will keep trying.
2021-06-28 01:03:34 <MY-R> whatsyourask: hmm, could you try put your firmware in /boot/firmware ?
2021-06-28 01:06:25 <MY-R> whitenite: what that command return to you on your rpi? awk '$2 !~ /^\/(sys|proc|dev|run)/ && $2 != "/" {print $2}' /proc/mounts
2021-06-28 01:06:28 <MY-R> sry
2021-06-28 01:06:34 <MY-R> ^ whatsyourask 
2021-06-28 01:21:30 <MY-R> ah whatever, that wont work
2021-06-28 07:52:54 <markand> /etc/os-release gets never updated?
2021-06-28 07:53:14 <markand> it still telling 3.11 on my box (while I've upgraded to 3.12, 3.13 and 3.14 in many years)
2021-06-28 07:55:01 <yyp> If you just upgrade the system (ie apk upgrade), it won't get updated. It's only included in the ISO
2021-06-28 07:55:16 <markand> okay
2021-06-28 07:55:40 <clandmeter> markand: should get updated
2021-06-28 07:55:46 <clandmeter> what does this say? 
2021-06-28 07:55:48 <clandmeter> apk info -W /etc/os-release
2021-06-28 07:56:14 <markand> could not find owner package
2021-06-28 07:56:34 <clandmeter> is it a regular disk install?
2021-06-28 07:57:06 <markand> nope :p
2021-06-28 07:57:21 <markand> it's on a VPS, so you're quite stuck to use a chroot install
2021-06-28 07:57:55 <clandmeter> and you did t hat how>
2021-06-28 07:58:02 <clandmeter> i think you are missing alpine-base
2021-06-28 07:58:07 <markand> oh so it looks like I only have alpine-baselayout
2021-06-28 07:58:07 <clandmeter> can you apk add alpine-base?
2021-06-28 07:58:12 <markand> yes, alpine-base isn't there
2021-06-28 07:58:37 <markand> done, thanks!
2021-06-28 07:58:38 <markand> !next
2021-06-28 07:59:17 <markand> clandmeter, it was somewhat like 4 years ago, unable to remember now :P
2021-06-28 07:59:35 <markand> but I think something like tar -xvzf alpine-minirootfs* + some chroot apk add 
2021-06-28 07:59:55 <markand> given the only way to install is to boot the VPS into rescue mode and have ssh access into it
2021-06-28 08:23:16 <nescafe> Hmm, glad my provider offers netboot.xyz in rescue mode.. installation was a breeze
2021-06-28 08:25:42 <markand> I like to install alpine on all my stuff using minirootfs, it makes the system so minimal
2021-06-28 08:26:29 <markand> with your boot manager of choice :)
2021-06-28 08:26:46 <markand> efibootmgr on x86_64, raspberrypi-bootloader on Pi's.
2021-06-28 09:28:59 <jonah1024> Hey, has there been done any MIPS port? I've seems some posts from 2017 that there was some interest. Any progress on that?
2021-06-28 09:30:50 <ikke> You mean 32-bits?
2021-06-28 09:32:40 <Grabunhold> ahoi! i'm running into a little problem running xorg on my latest alpine installation (3.14.0). i'm using /etc/inittab to spawn a session for a user via "login -f username", then have "exec startx" in that users ~/.profile.
2021-06-28 09:32:40 <Grabunhold> it all works well except when Xorg starts, in the newer installation it now drops out with "(EE) Server terminated with error (1)". older installations work fine. the user is a member of the video and input groups. what am i missing?
2021-06-28 09:32:40 <Grabunhold> my .xinitrc then contains a single line to start xterm (eventually i start chromium, but let's keep it at xterm to make it simple).
2021-06-28 09:38:09 <jonah1024> @ikke: yes. 32bit mipsel
2021-06-28 09:39:26 <ikke> Since wave technology pivoted, that effort was discontinued 
2021-06-28 09:41:41 <jonah1024> @ikke: bummer. What about chroot?
2021-06-28 09:42:46 <ikke> We don't have a port, so no minirootfs either 
2021-06-28 09:43:06 <ikke> We only have 64bits 
2021-06-28 09:50:23 <clandmeter> jonah1024: you could try to boostrap it yourself, not sure what the status is in aports.
2021-06-28 12:25:29 <jonah1024> @ikke: @clandmeter  Thanks. Will probably go a different way. I'm currently running OpenWrt but some tools are not available there. Will try with Debian chroot I think
2021-06-28 12:53:43 <AppAraat[m]> Hi, Pi Zero W(H) is armhf arch, correct?
2021-06-28 12:54:19 <ikke> Yes, should be 
2021-06-28 12:55:45 <AppAraat[m]> cool, thx!
2021-06-28 13:26:04 <kleo> How do I pass an environment variable to /alpine/enter-chroot script.sh? I'm trying to do it with ${{ env.VARIABLE }} on Github Actions
2021-06-28 13:57:28 <kleo> curl https://raw.githubusercontent.com/alpinelinux/alpine-chroot-install/master/alpine-chroot-install | sh -s -- -k VARIABLE
2021-06-28 15:06:33 <Grabunhold> is there a way to pass a switch to force overwriting filesystems when using an answerfile for setup-alpine?
2021-06-28 17:49:48 <whatsyourask> Hello71, I think I found the error. Somewhere in the update-kernel script there is a blank space in the variable "$_xfw". That causes that it doesn't copy any fimware that you pass as a parameter.
2021-06-28 17:50:40 <whatsyourask> """install: can't stat '/tmp/update-kernel.nJlNHm/root/lib/firmware/ brcm': No such file or directory""
2021-06-28 17:51:18 <whatsyourask> There is a space after /firmware/
2021-06-28 17:56:30 <Hello71> it should be for _xfw in $modloopfw
2021-06-28 17:56:35 <Hello71> apparently it's always been broken
2021-06-28 17:58:00 <ikke> We should run shellcheck on all or shell scripts
2021-06-28 17:59:23 <whatsyourask> https://github.com/alpinelinux/alpine-conf/commit/4f960a81e65c7ee1e37b5a4029e2aa47e63e654f
2021-06-28 18:00:05 <Hello71> yeah but it was already broken
2021-06-28 18:00:05 <whatsyourask> That part is what I think is wrong. It was changed a few days ago.
2021-06-28 18:00:58 <whatsyourask> I managed to copy my firmware to the temp folder while the script was working. And it worked.
2021-06-28 18:01:47 <whatsyourask> I am not knowledgeable enough to figure out what exactly is wrong with the script.
2021-06-28 21:15:02 <AppAraat[m]> On https://wiki.alpinelinux.org/wiki/Setting_up_a_new_user#Creating_a_new_user there's `doasedit` mentioned, but I can't find it anywhere, even in `doas-doc` package. Is the wiki incorrect in this case?
2021-06-28 21:16:30 <ikke> seems so, maybe the user had an alias or something
2021-06-28 21:16:50 <ikke> No mention of it here: https://github.com/Duncaen/OpenDoas/
2021-06-28 21:17:10 <ikke> https://github.com/Duncaen/OpenDoas/issues/70
2021-06-28 21:21:07 <AppAraat[m]> ikke: would you mind if I update that wiki page to reflect this? If so, can I link the script mentioned in that GH issue?
2021-06-28 21:21:54 <ikke> Feel free to update it
2021-06-28 21:42:13 <AppAraat[m]> done :)
2021-06-28 21:49:04 <AppAraat[m]> oops, was too quick. Getting permission denied errors when using that script (even as root lol)
2021-06-28 21:58:52 <AppAraat[m]> spoke too soon once again, forgot to chmod +x it :p
2021-06-28 22:12:58 <etothepii> Another quick wiki question for fellow Alpine wiki users - how long after proposing a section merge should I wait for feedback before doing the merge? Is there a standard "wait a week" or something like that? I didn't see anything in the wiki guide.
2021-06-28 23:05:03 <AppAraat[m]> Where can I find the tput command? Can't seem to find it anywhere in the descriptions.
2021-06-28 23:07:05 <trylletre> ncurses?
2021-06-28 23:10:16 <AppAraat[m]> yep, that worked. Thanks!
2021-06-28 23:42:42 <Hello71> https://pkgs.alpinelinux.org/contents?file=tput&path=&name=&branch=edge
2021-06-28 23:58:49 <AppAraat[m]> that is one very handy page, thanks!
2021-06-29 01:25:17 <clandmeter> AppAraat[m]: another way is apk search cmd:tput
2021-06-29 01:26:59 <AppAraat[m]> oh that's pretty cool, thanks a bunch!
2021-06-29 01:30:40 <c705> is apparmor only avaliable in testing?
2021-06-29 02:28:32 <LACampbell> clandmeter: how do you find about that? `man apk-search` doesn't exist, `man apk` doesn't mention search anywhere
2021-06-29 02:28:59 <clandmeter> LACampbell: find what?
2021-06-29 02:29:03 <clandmeter> cmd:xxx?
2021-06-29 02:29:07 <LACampbell> yeah
2021-06-29 02:29:12 <clandmeter> its a feature
2021-06-29 02:29:14 <clandmeter> of apk
2021-06-29 02:29:43 <clandmeter> anything in $PATH will get cmd:xxx added in apkindex as provides
2021-06-29 02:29:45 <LACampbell> ok I get that. But I'm trying to be more self-sufficient and RTFM
2021-06-29 02:29:56 <LACampbell> but which FM do I R?
2021-06-29 02:30:03 <clandmeter> RTFS
2021-06-29 02:30:08 <LACampbell> S?
2021-06-29 02:30:16 <clandmeter> source
2021-06-29 02:30:31 <clandmeter> im not sure its documented
2021-06-29 02:31:04 <Hello71> i mean this isn't even an apk feature
2021-06-29 02:31:08 <Hello71> it's abuild
2021-06-29 02:31:30 <clandmeter> correct
2021-06-29 02:31:38 <clandmeter> you can put anything in provides
2021-06-29 02:31:55 <clandmeter> apk just searches it
2021-06-29 02:32:12 <LACampbell> so you guys find this stuff by reading the source code?
2021-06-29 02:32:30 <clandmeter> or by hanging out for the last decade or so
2021-06-29 02:33:05 <clandmeter> its a know feature of alpine to not have proper documentation
2021-06-29 02:40:59 <riverdc> I'd like to install a package from edge/testing, which has dependencies in edge/community (I don't think they're in 3.13). I have edge/testing tagged as @testing in my /etc/apk/repositories, so normally I just do apk install p@testing. But I can't do that in this case becuase I need deps from edge/community. any suggestions?
2021-06-29 02:41:18 <riverdc> (the package is blender)
2021-06-29 02:43:26 <clandmeter> pin the missing packages?
2021-06-29 02:43:38 <Hello71> or just don't mix releases
2021-06-29 02:46:12 <artok> or use blender on flatpak ?
2021-06-29 02:46:56 <riverdc> clandmeter: yeah I guess that's the best option, barring just upgrading everything to edge
2021-06-29 02:47:02 <riverdc> maybe I should do that
2021-06-29 02:47:28 <artok> blender is picky anyways of dependencies
2021-06-29 02:49:49 <artok> I would grab official binary from blender.org and run it under glibc chroot
2021-06-29 15:14:30 <shrizza> Whew.
2021-06-29 15:17:28 <shrizza> Question: Has anyone successfully configured an anycast IP on a loopback in Alpine?
2021-06-29 15:17:42 <Thermi> What's your criteria for that?
2021-06-29 15:19:21 <shrizza> Well, I've added something like the following to /etc/network/interfaces: https://pastebin.com/kQkZkxqM
2021-06-29 15:20:36 <shrizza> Restarted networking but no lo:0 interface actually appears.
2021-06-29 15:22:01 <shrizza> Now if I do this but with inet6 instead of inet, and configure an IPv6 address similarly, it does work (though the IPv6 address appears below the plain "lo" loopback interface, which is fine).
2021-06-29 15:23:36 <shrizza> I guess I should say floating IP instead of anycast IP.
2021-06-29 15:24:11 <shrizza> I plan to use it as an anycast IP in my application, but yeah I think this technically acts as a floating IP in the context of the server.
2021-06-29 15:25:35 <mercenary> by hand, just 'ip address add 172.16.13.13/32 dev lo' works fine. Note that you need the forwarding sysctl enabled.
2021-06-29 15:28:55 <Hello71> why are you adding it to lo
2021-06-29 15:29:08 <Hello71> sub-interfaces are obsolete on linux
2021-06-29 15:30:33 <shrizza> Well, I didn't want to associate it with any physical interface.
2021-06-29 15:30:58 <shrizza> And I thought it would be dangerous to add to lo proper.
2021-06-29 15:31:09 <mercenary> having it on a loopback is perfectly valid, but lo:0 is rather old-style yes. ifupdown-ng should handle the lo case though
2021-06-29 15:32:34 <mercenary> There is also the 'dummy' interface type, create a dummy0 and stick the address on that
2021-06-29 15:35:42 <mercenary> Only reason not to have it on lo would be firewall rulesets that assume any traffic from/to lo is allowed
2021-06-29 15:45:54 <shrizza> Hm, maybe I'll just create a bridge with no associated ports.
2021-06-29 15:46:05 <shrizza> And add the floating IP to that.
2021-06-29 15:50:20 <shrizza> That seems to have worked.
2021-06-29 15:53:04 <mercenary> Works too yes
2021-06-29 15:53:13 <shrizza> And it's actually better because I can avoid the IPv6 behavior above where it mysteriously associates it not with a sub-interface but with the lo interface itself.
2021-06-29 15:54:06 <Hello71> i think linux doesn't support ipv6 subinterfaces
2021-06-29 16:45:49 <Thermi> shrizza, mercenary None of the things you wrote are true. Both of you. "subinterfaces" (you probably mean aliases" are deprecated (for like 20 years now). Forwarding of packets is only relevant of the destination is on another host. So it has no relevance for packets where the destination is just an IP address on an interface that is not the one on which the packet is received on. It also means that such packets are treated like any other 
2021-06-29 16:45:49 <Thermi> incoming packet, not as a forwarded packet. Packets will go through, for example, *filter INPUT and *filter OUTPUT, not *filter FORWARD. Also, adding IP addresses to loopback is fine. But so is adding a dummy interface. The difference is only in how you manage the IP addresses then. E.g. you could seperate anycast IPs from "normal" unicast IP addresses by assigning the first always to a specific dummy interface, and then get rid of all of 
2021-06-29 16:45:50 <Thermi> them in one swoop by deleting it. Also, just adding IP addresses to interfaces do not create subinterfaces.
2021-06-29 16:59:18 <mercenary> Thermi: you are right, forgot that linux is still weak host model, so it does not count it as forwarding
2021-06-29 16:59:59 <mercenary> the rest still stands
2021-06-29 18:46:39 <AppAraat[m]> Hi, so I've been trying to get a Pi Zero W jukebox thing going and I'm wondering what installation mode I should go with. The idea is that I load the OS and the music onto the SDcard, then OS boots and then starts wifi in AP mode and starts the mpd server. Then an Android phone running an mpd client can connect to the wifi network, and then to the mpd server running on the Pi.
2021-06-29 18:46:46 <AppAraat[m]> Since it's an SDcard, my primary concern is to minimize writes to it. But... Since it's running a wifi network, I'd also like to update the kernel (and drivers) from time to time.
2021-06-29 18:46:58 <AppAraat[m]> Since diskless and data modes are meant to be used in conjunction with lbu, this means that I can save stuff like configs (and the updated mpd db before rebooting), but updating the kernel looks kind of like a pain. So perhaps it's better to use the sys mode but mount /var/log/ as tmpfs? What do you think?
2021-06-29 18:48:22 <ikke> AppAraat[m]: afaik, there is an update-kernel command to update the kernel on the image
2021-06-29 18:49:05 <AppAraat[m]> but IIRC according to the wiki it requires 8 GB free RAM, so I'm guessing it just compiles the new kernel?
2021-06-29 18:50:39 <AppAraat[m]> if that's the case, then the Pi Zero will just... obliterate itself into an endless void.
2021-06-29 18:52:10 <Thermi> No, it probably will just install from the APK
2021-06-29 18:52:31 <ikke> It says it has to do with the modloop, so I assume it generates a new modloop
2021-06-29 18:54:06 <Thermi> AppAraat[m], It just makes a new modloop.
2021-06-29 18:57:27 <AppAraat[m]> for that I have to first find out what on earth the modloop is :p
2021-06-29 18:57:51 <AppAraat[m]> Apparently it's some kind of a SquashFS thing containing the entire kernel contents? https://superuser.com/a/1316205
2021-06-29 19:51:13 <MY-R> AppAraat[m]: or you can wait for next release, download alpine-rpi-*.tar.gz remove everything and leave only /cache .apkovl.tar.gz usercfg.txt and unpack to mmcblk0p1, reboot, done
2021-06-29 19:52:31 <ikke> AppAraat[m]: it contains the kernel modules
2021-06-29 19:52:40 <ikke> AppAraat[m]: without the modloop, all the modules would be stored in the tmpfs
2021-06-29 19:52:49 <ikke> requiring a lot of ram in runtime
2021-06-29 19:53:33 <ikke> to aleviate that, the modules are placed in the read-only squashfs file, which is mounted via a loop mount
2021-06-29 19:54:19 <etothepii> ikke - Is it actually possible to install/edit modules in disk less? As far as I could tell from the mailing list, that's not really supported? 
2021-06-29 19:54:35 <ikke> etothepii: nod with the modloop in place
2021-06-29 19:54:40 <ikke> as it's read-only
2021-06-29 19:55:59 <etothepii> ikke - wait, is that "nod, it's not supported, as it's read only" or "not
2021-06-29 19:56:08 <etothepii> "nod it's possible" 
2021-06-29 19:56:20 <ikke> sorry, not*
2021-06-29 19:57:18 <etothepii> OK, cool, that's what I suspected. I recently updated the raspi wiki page to say if you needed extra modules you needed a sys installation. Wanted to make sure there wasn't an error there. 
2021-06-29 20:07:22 <etothepii> MY-R: just to check, if an update is done by replacing the root with a new release, but the apkovls are left intact, the user would still need to update /etc/apk/repositories, correct? Since wouldn't the overlay revert them to the old release?
2021-06-29 20:08:17 <ikke> correct
2021-06-29 20:08:36 <ikke> updating the image only brings you the new kernel
2021-06-29 20:11:00 <etothepii> So the update procedure would be: 1) clear the card, except for HOSTNAME.apkovl, 2) copy the new release, 3) update repos?
2021-06-29 20:12:20 <Hello71> that's one way to do it
2021-06-29 20:12:50 <Hello71> the safer method is probably to get new media, copy the new release, then copy the ovl
2021-06-29 20:13:12 <Hello71> i guess you could put the backup in a subdirectory too
2021-06-29 20:14:10 <etothepii> Ah, I was assuming I would be making these changes in a second system, but yeah I could see how that could be messy on the running media. 
2021-06-29 20:15:06 <etothepii> If the kernels aren't updatable, does that mean diskless can be vulnerable to security issues if there's a kernel bug until an upgrade? Or is there a way for apk to patch them without requiring a reinstall?
2021-06-29 20:18:05 <etothepii> I'm thinking for people that just set things to auto-upgrade that could be worth calling out - that kernel security patches require intervention in a way others might not.
2021-06-29 20:36:23 <AppAraat[m]> ikke: so (a part of?) the kernel in the new image gets extracted into a read-only squashfs which itself resides on the SDcard in this case?
2021-06-29 20:36:49 <ikke> AppAraat[m]: the modloop is pregenerated when the image is built
2021-06-29 20:37:13 <AppAraat[m]> MY-R: Why wait for the next release? Does it have features that make what you described possible?
2021-06-29 20:37:17 <MY-R> is enough to backup whole sdcard with rsync or something, if something go bad then can take out sdcard and copy back to it
2021-06-29 20:38:23 <MY-R> etothepii: dont have to change version in repository when upgrading from for example 3.13.1 to 3.13.2 etc
2021-06-29 20:38:27 <AppAraat[m]> ikke: So update-kernel builds a new image?
2021-06-29 20:39:14 <AppAraat[m]> I also can't seem to get update-kernel's help page, as on the sys mode install I get:
2021-06-29 20:39:30 <AppAraat[m]> `/sbin/update-kernel: exec: line 20: fakeroot: not found`
2021-06-29 20:39:33 <ikke> AppAraat[m]: an educated guess: it updates the kernel on the boot drive, and regenerates the modloop
2021-06-29 20:40:32 <AppAraat[m]> hmm
2021-06-29 20:40:49 <ikke> I guess it installs the kernel in a chroot or something like that, and uses that
2021-06-29 20:44:17 <MY-R> etothepii: with upgrade like from 3.13 to 3.14 you changing repository, doing -a upgrade and have to remember to run "update-conf -a" because there will be major changes in packages, then stop modloop service, mount media writable and replace the root
2021-06-29 20:45:43 <AppAraat[m]> hmm, this is kind of unexplored territory for me as of now and it's kind of difficult to wrap my head around it so far, so I think I'm going to go with a sys install for now and put /var/log/ in a tmpfs. I'll postpone experimentation for later iterations since time is kind of tight right now.
2021-06-29 20:52:30 <MY-R> AppAraat[m]: I can give you my notes so you can decide if you can deal with it :P
2021-06-29 20:55:43 <etothepii> be auto-updated, and there could be a security issue")
2021-06-29 20:55:43 <etothepii> MY-R - that makes sense. I'm mostly thinking on how the docs can be improved for upgrades, so the process is a little less daunting, and to make sure people who are doing set-and-forget installs have info if there are landmines (like it might be worth saying "you need to check if there's a new minor release version every so often, since you won't
2021-06-29 20:56:01 <AppAraat[m]> MY-R: That would be greatly appreciated (especially for later iterations)
2021-06-29 20:57:42 <MY-R> AppAraat[m]: https://tpaste.us/bVMW
2021-06-29 20:58:14 <AppAraat[m]> thanks a lot!
2021-06-29 20:58:29 <MY-R> was using this way on my diskless rpi4 for more than half year and never got problem with upgrades
2021-06-29 21:00:08 <AppAraat[m]> ah ok so it's diskless mode instead of data mode?
2021-06-29 21:00:09 <MY-R> and of course before start doing backup
2021-06-29 21:00:24 <MY-R> AppAraat[m]: ye
2021-06-29 21:00:57 <AppAraat[m]> cool
2021-06-29 21:00:58 <MY-R> what is data mode? with extra partition?
2021-06-29 21:01:19 <MY-R> Im sometimes confused by naming of those modes :P
2021-06-29 21:03:17 <AppAraat[m]> According to https://wiki.alpinelinux.org/wiki/Installation#Data_Disk_Mode at least - "This mode also runs from system RAM, thus it enjoys the same accelerated operation speed as "diskless" mode. However, swap storage and the entire /var directory tree get mounted from a persistent storage device (two newly created partitions)."
2021-06-29 21:03:26 <AppAraat[m]> but I haven't tried it yet
2021-06-29 21:05:03 <MY-R> in your case you dont need persistant /var but just extra partition for own stuff and all rest will be in ram so your sdcard will have longer life
2021-06-29 21:05:57 <etothepii> MY-R: would setup-bootable also work here? There's currently a procedure listed for updating the diskless system that looks similar to the steps you provided (https://wiki.alpinelinux.org/wiki/Upgrading_Alpine#Upgrading_Alpine_Linux_on_other_removable_media_.28such_as_CF.2FUSB.29)
2021-06-29 21:06:46 <ikke> yes
2021-06-29 21:06:53 <AppAraat[m]> normally under sys mode apk update && apk upgrade will also upgrade the kernel (if there are kernel updates) right?
2021-06-29 21:07:20 <ikke> yes
2021-06-29 21:07:32 <ikke> but with diskless, there is no kernel installed with apk
2021-06-29 21:08:18 <AppAraat[m]> cool, so my understanding is ok so far.
2021-06-29 21:09:27 <MY-R> it is all about kernel upgrade (mostly modloop and firmware stuff), normaly all rest packages getting upgrade without that magic
2021-06-29 21:09:47 <ikke> ahuh
2021-06-29 21:10:54 <MY-R> etothepii: probably it could work ye, but I didnt bother to check it, I wanted at some point but then bought usb sata adapter :)
2021-06-29 21:53:13 <etothepii> Oh, MY-R, do you mind then if I put your instructions in the wiki as an alternate to using the setup-bootable?
2021-06-29 21:53:18 <etothepii> just so it's preserved?
2021-06-29 21:56:38 <MY-R> etothepii: do whatever want with it but be sure that it working! :)
2021-06-29 21:58:03 <MY-R> because in Alpine wiki there is many random notes which not always working because outdated, complicated without reason or wrotten in some random way
2021-06-29 21:58:27 <etothepii> Have you run those commands recently? On a recent build? I can do a quick test later if not, but I was thinking of presenting it as an alternate, esp for people who want to update without the script, on a second machine.
2021-06-29 21:59:14 <MY-R> etothepii: ye, upgraded diskless rpi4 from last 3.13.x to 3.14 so still working :)
2021-06-29 22:00:07 <etothepii> OK - cool, so I think I can safely add it just as an alternate upgrade path in the Upgrading_Alpine article
2021-06-29 22:01:00 <MY-R> etothepii: yep, at least people will see how that process looks
2021-06-29 22:01:14 <etothepii> "in Alpine wiki there is many random notes which not always working because outdated, complicated without reason" - agreed, a lot of the raspi data in particular is scattered
2021-06-29 22:01:24 <MY-R> ye
2021-06-29 22:02:22 <MY-R> and noticed people on irc who got updated rpi's but not kernel because they didnt even know how it working in diskless mode
2021-06-29 22:13:47 <MY-R> etothepii: keep in mind that mmcblk0p1/config.txt will be overwriten by file from .tar.gz so can add info about and copy somewhere before unpack 
2021-06-29 22:15:06 <MY-R> that is why better put stuff in usercfg.txt BUT not all settings like gpu_mem= will work from there
2021-06-29 22:17:20 <AppAraat[m]> any advice on chrooting into alpine sdcards to update the sys install there?
2021-06-29 22:17:39 <etothepii> MY-R: Ok, here's the updated text. https://wiki.alpinelinux.org/wiki/Upgrading_Alpine#Updating_a_USB.2FCF_diskless_install_without_setup-bootable
2021-06-29 22:17:43 <AppAraat[m]> this guide for example assumes a raspbian image (and Ubuntu host IIRC) - https://gist.github.com/htruong/7df502fb60268eeee5bca21ef3e436eb
2021-06-29 22:17:46 <etothepii> Does that seem correct to you?
2021-06-29 22:19:06 <etothepii> AppAraat[m]: Why would you want to chroot into it to upgrade if it's a sys install?
2021-06-29 22:19:50 <AppAraat[m]> because otherwise I have to start the pi, and change it from an AP to a wifi client, connect to a network and then download the updates.
2021-06-29 22:19:57 <MY-R> etothepii: please remove "provided by user 'MY-R'" nobody care about such a infos and especially me :)
2021-06-29 22:21:07 <etothepii> MY-R: sure thing - that's an easy enough fix. any other concerns? 
2021-06-29 22:21:11 <AppAraat[m]> etothepii: vs. insert sdcard into sdcard slot of the laptop, chroot into it and update it that way.
2021-06-29 22:21:28 <MY-R> etothepii: usercfg.txt wont be overwriten because not exist in .tar.gz but config.txt exist and that should backup just in case
2021-06-29 22:22:11 <etothepii> ok, so leave usercfg.txt, and config.txt
2021-06-29 22:22:50 <MY-R> etothepii: lbu ci -d tbh isnt necessary when cleaning up cache so can remove
2021-06-29 22:23:05 <MY-R> is just a habbit to lbu ci all time in diskless mode :P
2021-06-29 22:23:29 <shrizza> Thermi: Understood on the point about subinterfaces being deprecated, but I was never able to add a floating IP to lo only using declarations in /etc/network/interfaces. Also AFAICT, my conclusion of using bridges for this function should be OK, unless I've missed something. Linux bridging mechanism should be mature and is intuitive to me, versus this dummy interface that I've never heard
2021-06-29 22:23:31 <shrizza> of and seems to require a kernel module being enabled.
2021-06-29 22:25:07 <etothepii> shrizza: Would it work for you to bridge over a dummy interface? There's a tutorial on that for LXC containers that might meet your needs?  (https://wiki.alpinelinux.org/wiki/LXC#Creating_a_LXC_container_without_modifying_your_network_interfaces)
2021-06-29 22:26:19 <shrizza> I'm saying bridges without dummy interfaces works.
2021-06-29 22:26:58 <shrizza> Just associate the floating IP to a bridge with no bridge-ports.
2021-06-29 22:27:01 <MY-R> etothepii: you forgot about second link to image, the one which you left is to .sha256 hash :)
2021-06-29 22:27:43 <MY-R> etothepii: this you forgot: wget https://dl-cdn.alpinelinux.org/alpine/v3.13/releases/aarch64/alpine-rpi-3.13.0-aarch64.tar.gz
2021-06-29 22:28:08 <etothepii> Oh dear. Yes I did. Updating (and merging in your other chances)
2021-06-29 22:28:11 <etothepii> *changes
2021-06-29 22:28:58 <etothepii> OK, credit removed, file names changed, missing line added, hopefully I've gotten that right. 
2021-06-29 22:29:22 <AppAraat[m]> how long does it usually take for a package to go from edge/testing to edge/community?
2021-06-29 22:32:10 <MY-R> etothepii: you can delete "&& sha256sum -c *.sha256" at the end of wget line since you already put that linde under :)
2021-06-29 22:33:23 <MY-R> etothepii: you can change description "Restore the original usercfg.txt:" to config.txt
2021-06-29 22:34:19 <etothepii> MY-R: Got em - both should be fixed now!
2021-06-29 22:36:46 <MY-R> etothepii: sha256sum -c *.sha256 put in new line, people will forget about it when copy/paste url
2021-06-29 22:37:13 <MY-R> and it is important to check that .tar.gz isnt corrupted
2021-06-29 22:39:14 <MY-R> but ye, good job, will be useful, especially when somebody will come here and ask how to upgrade kernel :)
2021-06-29 22:40:50 <etothepii> MY-R: updated. And thank you for providing the steps!
2021-06-29 22:41:53 <MY-R> :)
2021-06-29 23:32:28 <Thermi> shrizza, network/interfaces is so old, it only ever supported the extremely deprecated way of ifconfig
2021-06-29 23:32:59 <Thermi> I do not know of any proper rework of the file and syntax to support any network features added after its conception
2021-06-29 23:34:28 <shrizza> I just need a floating IP on a non-physical interface. Bridge config in /etc/network/interfaces accomplishes this.
2021-06-29 23:34:46 <AppAraat[m]> Are there any files/dirs other than /var/log/ that do frequent writes which aren't the result of (new) package installs?
2021-06-30 01:05:29 <etothepii> AppAraat[m] - /var/log and /tmp potentially?
2021-06-30 01:06:17 <etothepii> Arch uses a tempfs on /var/run, /var/lock, and /tmp by default, I believe. 
2021-06-30 05:45:54 <AppAraat[m]> etothepii: huh, I thought that /tmp was always mounted as tmpfs.
2021-06-30 07:23:22 <kmmndr> Hi all :-)
2021-06-30 08:10:17 <AppAraat[m]> Q: Which tool would you recommend for monitoring which files/dirs have been written to lately? According to https://unix.stackexchange.com/a/356402 `fatrace` should be the tool to use, but I can't find it in packages. Another one seems to be iosnoop, but it only gives me aggregate info on filesystems. The 3rd option is `inotifywait` but considering its default limitation, would it be a good tool for this purpose?
2021-06-30 08:28:21 <lopid> inotifywait, inotifywatch, etc. maybe one of those
2021-06-30 08:54:17 <AppAraat[m]> ok, inotifywait it is then (with `echo "131072" > /proc/sys/fs/inotify/max_user_watches` just in case)
2021-06-30 08:55:07 <lopid> some time ago i wrote a simple shell loop that used to it detect if a particular filetype changed, and perform an action
2021-06-30 11:13:23 <arlecchino> Hey, can anybody explain how it is done/decided to push a newer package version from edge to a stable version? There is an up-to-date package in edge which has closed a security problem with CVE, but the package in 3.1[234] doesn't get updated.
2021-06-30 11:13:54 <ikke> arlecchino: secfixes should be bacported
2021-06-30 11:14:00 <ikke> which package?
2021-06-30 11:14:04 <arlecchino> dovecot
2021-06-30 11:14:57 <ikke> Which CVE?
2021-06-30 11:16:34 <arlecchino> I think all from June 21st at https://www.dovecot.org/security
2021-06-30 11:17:02 <ikke> https://security.alpinelinux.org/srcpkg/dovecot
2021-06-30 11:19:12 <ikke> This one does not appear to have CPE match rules
2021-06-30 11:19:14 <ikke> https://security.alpinelinux.org/vuln/CVE-2021-29157
2021-06-30 11:19:41 <ikke> nvd has apparently not analyzed it yet
2021-06-30 11:22:17 <arlecchino> Let's look at CVE-2021-33515. It is marked as fixed in edge-main with 2.3.15-r0. But the stable releases doesn't have a fix for it. Upstream it got fixed with 2.3.14.1.
2021-06-30 11:22:36 <ikke> Yeah, for now, we are using the CPE match rules to find out vulnerable versions
2021-06-30 11:22:49 <ikke> in this case, they are not available, but the package in edge marks it as fixed
2021-06-30 11:22:56 <ikke> so the only datapoint we have is that
2021-06-30 11:23:06 <ikke> (for the securitry tracked, I mean)
2021-06-30 11:31:49 <arlecchino> @ikke where do I find these CPE match rules? do you know?
2021-06-30 11:33:20 <arlecchino> Is this the stuff like this: "cpe23Uri" : "cpe:2.3:a:ibm:db2:11.1:*:*:*:*:*:*:*"
2021-06-30 11:33:27 <ikke> Yes
2021-06-30 11:34:21 <ikke> https://nvd.nist.gov/vuln/detail/CVE-2021-29157
2021-06-30 11:38:10 <arlecchino> mmh. I've downloaded the nvdcve-json and there are many CVEs without cpe. Do they get one in the future? Or will have stable alpine the old packages forever?
2021-06-30 11:42:15 <ikke> They get updated, but note that it's not a requirement for security fixes to get backported 
2021-06-30 11:43:56 <ikke> It's just one way we can find out if we miss something 
2021-06-30 11:44:23 <ikke> But dovecot will be updated soon 
2021-06-30 11:45:35 <arlecchino> ok. I appreciated your information. thanks a lot
2021-06-30 19:44:02 <AppAraat[m]> So the full command I'm using is: `doas inotifywait --monitor --format '%T ::: %w%f ::: %e' --timefmt '%F__%T%z__%a' --event modify --recursive / --exclude /dev`
2021-06-30 19:44:55 <AppAraat[m]> I had to figure out how I could get the full path in there since the description of `%w` is pretty confusing: "This will be replaced with the name of the Watched file on which an event occurred."
2021-06-30 19:45:01 <AppAraat[m]> Is it even accurate?
2021-06-30 19:53:31 <yyp> Is it possible to completely disable root check in abuild globally to not pass it -F every time?
2021-06-30 20:10:02 <yyp> nevermind, alias abuild='abuild -F' actually works in .profile when using chroot
2021-06-30 20:12:00 <Sheila> I just create a user account and use `--userspec`, tbh
2021-06-30 20:12:18 <Sheila> almost never need root in a chroot once configured properly anyway, so…
2021-06-30 20:20:14 <AppAraat[m]> `2021-06-30__21:57:37+0200__Wed ::: /etc/resolv.conf.26292 ::: MODIFY` - for some reason resolv.conf keeps changing but it's always the same value.
2021-06-30 20:21:46 <Sheila> are you using dhcp
2021-06-30 20:35:44 <MY-R> AppAraat[m]: probably every time when your dhcp client doing renew it grab info about dns server from dhcp
2021-06-30 20:38:02 <MY-R> depends what dhcp client you using you can tell to stop touching resolv.conf
2021-06-30 20:59:33 <rapha> what would the appropriate place be in alpine to have "for x in a b c d; do integritysetup open --integrity sha256 /dev/sd$x sd$x; done" run before "mount -a" is run during boot?
2021-06-30 21:00:59 <ikke> I guess you could add a service with before localmount
2021-06-30 21:03:32 <rapha> then a custom service it'll be, thank you ikke 
2021-06-30 21:14:01 <AppAraat[m]> Sheila: Yes.
2021-06-30 21:14:33 <AppAraat[m]> I'll try to setup a longer lease in that case.
2021-06-30 21:28:44 <MY-R> AppAraat[m]: keep in mind that renew is done in half time of lease
2021-06-30 21:30:21 <AppAraat[m]> oh, hmm. Well I'm going to be running the machine as a default gw / wifi AP anyway, so I guess I can fudge around more with the lease / renew times.
2021-06-30 21:37:21 <MY-R> AppAraat[m]: if that machine on which trying prevent frequently write access to resolv.conf is connected to your ISP then not always is possible to enforce different lease time
2021-06-30 21:39:08 <AppAraat[m]> yeah true, but in my case it's going to be a Pi ZW without any Internet connection. Just an AP you connect to (with a phone) and then the phone acts as sort of a remote control for playing music.
2021-06-30 21:39:57 <MY-R> oook :)
2021-06-30 21:41:52 <AppAraat[m]> yeah it's kind of dumb, especially considering I'll have to use 2 phones: 1 for normal phone use unrelated to my project and 1 for remote control usage lol
2021-06-30 21:42:25 <AppAraat[m]> that's because on Android you can't connect to the Internet with mobile data and be connected to another wifi network at the same time... ugh.
2021-06-30 21:42:49 <AppAraat[m]> (unless you're rooted, that is)
2021-06-30 21:44:52 <MY-R> cant? now many phones got that weird hybrid network...
2021-06-30 21:46:40 <AppAraat[m]> AFAIK yes, you'll have to mess around with Android host firewall settings.
2021-06-30 21:46:56 <AppAraat[m]> and I bet those hybrid networks are hardware dependent.
2021-06-30 21:47:38 <MY-R> saw it in many samsungs and honor/huaweis but not sure how it work
2021-06-30 21:48:38 <AppAraat[m]> I'll have to dig into that then. As of now if I connect to wifi then my mobile data just cuts out.
2021-06-30 21:52:09 <MY-R> it is called like speed/download booster and it trying use best network for downloading data but probably wont work like you would like to :)
2021-06-30 21:53:19 <MY-R> but would be cool to have setting in which can set up which app got priority to which network 
2021-06-30 21:53:51 <AppAraat[m]> +1
2021-06-30 21:55:36 <MY-R> like it is with dual sim slots, can pickup which is default or got question which using for what, but ye off topic :)
2021-06-30 21:56:45 <AppAraat[m]> heh
2021-06-30 21:58:45 <MY-R> AppAraat[m]: https://jrklein.com/2019/04/06/using-mobile-internet-data-while-connected-to-a-wifi-network-ios-android/
2021-06-30 21:59:24 <MY-R> you could try that, setting up static IP and remove gateway/router so it should mobile data, or try with "use always mobile data"
2021-06-30 22:01:25 <AppAraat[m]> hmm, I'll check it out thanks!