💾 Archived View for gemini.ctrl-c.club › ~phoebos › logs › kisslinux-2022-02-08.txt captured on 2024-05-26 at 16:10:27.

View Raw

More Information

⬅️ Previous capture (2022-03-01)

-=-=-=-=-=-=-

[2022-02-08T01:37:11Z] <illiliti> libinput quirks fixed!
[2022-02-08T01:37:57Z] <illiliti> i have fully working wayland session without libudev[-zero]
[2022-02-08T01:38:29Z] <noocsharp> nice
[2022-02-08T01:38:38Z] <noocsharp> what was forcing you to use libudev-zero?
[2022-02-08T01:39:25Z] <illiliti> wlroots, sway, libinput
[2022-02-08T01:40:24Z] <illiliti> i also intend to upstream patches if possible
[2022-02-08T01:41:19Z] <noocsharp> what device manager do you use?
[2022-02-08T01:41:27Z] <illiliti> mdev
[2022-02-08T01:42:45Z] <illiliti> i'll push patches shortly
[2022-02-08T01:42:53Z] <illiliti> so you can try it...
[2022-02-08T01:48:49Z] <illiliti> https://github.com/illiliti/libdemi
[2022-02-08T01:48:53Z] <illiliti> libinput https://termbin.com/hmcp
[2022-02-08T01:48:57Z] <illiliti> wlroots https://termbin.com/s2yk
[2022-02-08T01:49:23Z] <illiliti> sway https://termbin.com/dvvk
[2022-02-08T01:53:06Z] <rohan> this works with hotplug?
[2022-02-08T01:53:27Z] <illiliti> yep
[2022-02-08T01:53:39Z] <illiliti> i plan to add freebsd support to libdemi, so freebsd users can use wayland without hacks like libudev-devd
[2022-02-08T01:54:32Z] <illiliti> netbsd support is already implemented, but netbsd doesn't evdev...evdev is mandatory for libinput
[2022-02-08T01:58:25Z] <illiliti> for those who are out of the loop, the whole work was done to fix this: https://gitlab.freedesktop.org/wlroots/wlroots/-/issues/2257
[2022-02-08T02:18:42Z] <dilyn> illiliti: do you know to what extent this works with non-sway compositors, or are you looking for testers?
[2022-02-08T02:18:48Z] <dilyn> or is it sway specific
[2022-02-08T02:19:33Z] <dilyn> (testers in the sense of, willing to extend support beyond libinput on compositors)
[2022-02-08T02:21:57Z] <illiliti> this doesn't work with non-wlroots based compositors
[2022-02-08T02:22:12Z] <illiliti> because almost all of them depend on libudev
[2022-02-08T02:34:39Z] <dilyn> yes I meant all the wlroots compositors (for instance I'd love to attempt this on wayfire, with no idea on potential success)
[2022-02-08T02:39:36Z] <illiliti> should work if udev isn't directly used in code
[2022-02-08T02:42:55Z] <illiliti> i suggest you to wait until i open MR to wlroots. i'll also document build intructions there
[2022-02-08T02:49:35Z] <dilyn> gotcha gotcha. Lemme know. Is it wlroots 0.15 specific?
[2022-02-08T02:50:06Z] <dilyn> also am I right in thinking that this almost entirely solves the problem of wayland on BSDs??
[2022-02-08T02:51:45Z] <illiliti> you need wlroots master
[2022-02-08T02:52:02Z] <illiliti> yes. it solves the major part of the problem
[2022-02-08T02:52:15Z] <illiliti> the other problem is evdev
[2022-02-08T02:53:18Z] <illiliti> someone(not me plzz) needs to rewrite libinput to support other backends such as wscons
[2022-02-08T03:02:46Z] <dilyn> jesus ew
[2022-02-08T03:34:10Z] <illiliti> up-to-date patches will be there: https://github.com/illiliti/libinput and https://github.com/illiliti/wlroots
[2022-02-08T03:34:16Z] <illiliti> branch libdemi
[2022-02-08T04:10:57Z] <testuser[m]> Hi
[2022-02-08T04:15:37Z] <illiliti> hi
[2022-02-08T04:15:42Z] <illiliti> do you have latest chromium bin for musl?
[2022-02-08T04:16:32Z] <testuser[m]> I'll try to build it
[2022-02-08T04:16:44Z] <testuser[m]> Today
[2022-02-08T04:16:56Z] <testuser[m]> I have to do v98 update 
[2022-02-08T04:18:18Z] <illiliti> many thanks
[2022-02-08T04:34:01Z] <rohan> any place to get openjdk bin for kiss?
[2022-02-08T06:43:58Z] <illiliti> rohan: you can try to extract bin from alpine and convert to kiss package
[2022-02-08T09:23:50Z] <testuser[m]> dilyn is pipewire 0.3.45 working for you?
[2022-02-08T09:24:04Z] <testuser[m]> it just says "no node found for 35, stream error" when i try to use my headphone or mic
[2022-02-08T09:24:09Z] <testuser[m]> 0.3.43 works
[2022-02-08T09:28:12Z] <testuser[m]> 0.3.44 also works
[2022-02-08T09:28:15Z] <testuser[m]> weird
[2022-02-08T09:46:49Z] <testuser[m]> looks like its not detecting devices at all
[2022-02-08T09:47:17Z] <testuser[m]> changed log level to 4 and compared it's reaction with old ver, it just sits there doing nothing
[2022-02-08T09:47:39Z] <illiliti> git bisect your best friend
[2022-02-08T09:48:44Z] <testuser[m]> yeah
[2022-02-08T09:48:45Z] <testuser[m]> doing
[2022-02-08T09:49:41Z] <illiliti> no wonder if this is udev related bug
[2022-02-08T09:50:37Z] <illiliti> they constantly break it by putting more udev nonsense
[2022-02-08T09:51:50Z] <testuser[m]> Yay found the sussy commit on first look
[2022-02-08T09:51:55Z] <testuser[m]> https://github.com/PipeWire/pipewire/commit/b369401c8ebe46987dba560be0fa978806fff4d4
[2022-02-08T09:52:24Z] <testuser[m]> works after reverting it
[2022-02-08T09:57:21Z] <testuser[m]> btw why are they using `unsigned int var:1;` ? are they allergic to bool or something lol
[2022-02-08T10:03:50Z] <illiliti> this is redhat
[2022-02-08T10:03:56Z] <illiliti> they are so desperate to needlessly obfuscate code
[2022-02-08T10:15:28Z] <illiliti> btw, 97a5fe80c1d0fb0d8173a61a87051053cff7a5c2 should fix the bug, no?
[2022-02-08T10:16:23Z] <testuser[m]> oh looks like it didnt make it into the release
[2022-02-08T10:16:23Z] <testuser[m]> bruh
[2022-02-08T10:19:21Z] <testuser[m]> yeah works
[2022-02-08T10:20:17Z] <illiliti> nice
[2022-02-08T10:23:42Z] <testuser[m]> [34466/48546]
[2022-02-08T10:23:46Z] <testuser[m]> soon
[2022-02-08T10:24:48Z] <illiliti> jesus
[2022-02-08T10:27:12Z] <illiliti> have you tried mold?
[2022-02-08T10:27:50Z] <illiliti> it should speed up build a little
[2022-02-08T10:30:41Z] <testuser[m]> Isn't linking done at the final step only? All the targets are just invoking CXX to build the object files
[2022-02-08T10:31:44Z] <illiliti> it is
[2022-02-08T10:33:41Z] <illiliti> the last time i tried to link chromium my cacl blown up
[2022-02-08T10:38:01Z] <illiliti> even 100G swap didn't help lol
[2022-02-08T10:38:58Z] <testuser[m]> Cacl?
[2022-02-08T10:39:06Z] <illiliti> calc
[2022-02-08T10:39:12Z] <illiliti> calculator
[2022-02-08T10:39:20Z] <testuser[m]> Bruh
[2022-02-08T11:09:55Z] <testuser[m]> What's the correct way to control pipewire volume? I've been using alsamixer but recently it stopped giving me the controls for my mic
[2022-02-08T12:21:29Z] <testuser[m]> https://github.com/kiss-community/bin/commit/b70532ee79e9949e3e4223303c8950b9f29ebf90
[2022-02-08T12:21:33Z] <testuser[m]> illiliti: tell me if it works
[2022-02-08T12:34:06Z] <illiliti> segfault at startup
[2022-02-08T12:34:26Z] <testuser[m]> yea its giving me that too but i thought it was cuz im in chroot
[2022-02-08T12:34:33Z] <testuser[m]> weird, cuz nothing has changed
[2022-02-08T12:34:41Z] <testuser[m]> in patches
[2022-02-08T12:35:59Z] <testuser[m]> I did built it with vaapi though but that shouldn't be related
[2022-02-08T12:38:42Z] <testuser[m]> https://github.com/git-bruh/repo-community/commit/4e53c91b46668b8cd8655cc2a9da33bcc3ebcdd4
[2022-02-08T12:38:47Z] <testuser[m]> Piece of shit
[2022-02-08T12:45:21Z] <testuser[m]> Btw by segfault you meant SIGTRAP?
[2022-02-08T12:45:52Z] <illiliti> yes
[2022-02-08T12:46:27Z] <illiliti> log is different on each startup
[2022-02-08T12:46:43Z] <illiliti> probably related to multithreading
[2022-02-08T12:46:53Z] <testuser[m]> what does it say?
[2022-02-08T12:47:44Z] <illiliti> Zygote could not fork: process_type gpu-process numfds 4 child_pid -1
[2022-02-08T12:47:59Z] <testuser[m]> what if u do --disable-gpu ?
[2022-02-08T12:48:49Z] <illiliti> same
[2022-02-08T12:48:55Z] <illiliti> broker_posix.cc(43)] Invalid node channel message
[2022-02-08T12:49:19Z] <illiliti> gpu_init.cc(454)] Passthrough is not supported, GL is egl, ANGLE is << not sure if this is fatal error
[2022-02-08T12:49:29Z] <testuser[m]> it is
[2022-02-08T12:49:56Z] <testuser[m]> i mean
[2022-02-08T12:49:59Z] <testuser[m]> did older versions say that too? if u remember
[2022-02-08T12:51:07Z] <illiliti> yeah, they did
[2022-02-08T12:51:17Z] <testuser[m]> hmm
[2022-02-08T12:51:24Z] <illiliti> https://termbin.com/8ezl
[2022-02-08T12:51:24Z] <illiliti>  chromium --log-level=0 --enable-logging=stderr
[2022-02-08T12:52:48Z] <testuser[m]> nothing seems wrong in ur log
[2022-02-08T12:53:04Z] <illiliti> --no-sandbox not working too
[2022-02-08T12:53:20Z] <testuser[m]> Failed to get drm magic
[2022-02-08T12:53:23Z] <testuser[m]> hmm
[2022-02-08T12:53:57Z] <testuser[m]> btw any reason for not using ff + webrender?
[2022-02-08T12:55:26Z] <testuser[m]> `#0  0x000055555e1fcb71 in gfx::PlatformFont::CreateDefault() ()` what i get with gdb on unstripped binary
[2022-02-08T12:55:38Z] <illiliti> firefox is slow
[2022-02-08T12:55:44Z] <illiliti> even with webrender
[2022-02-08T12:56:40Z] <testuser[m]> i dont think its even possible to build chromium with debug symbols on a normal system, not sure how to proceed
[2022-02-08T12:56:53Z] <testuser[m]> i tried it with webkitgtk once it exploded
[2022-02-08T12:58:21Z] <illiliti> SIGTRAP << i think something wrong with seccomp
[2022-02-08T12:58:35Z] <testuser[m]> no it traps for any error
[2022-02-08T12:58:41Z] <illiliti> btw, does glibc work?
[2022-02-08T12:59:14Z] <testuser[m]> not built for it, switched to ff webrender since nvidia hw accel works fine on the wlroots fork
[2022-02-08T13:01:34Z] <illiliti> pwd
[2022-02-08T13:01:37Z] <illiliti> oops
[2022-02-08T13:02:46Z] <illiliti> fresh profile doesn't work either
[2022-02-08T13:05:54Z] <testuser[m]> https://github.com/archlinux/svntogit-packages/blob/fe85b648f60af1e25d16965ab974cc93ce496d30/trunk/PKGBUILD#L158
[2022-02-08T13:05:57Z] <testuser[m]> what does this mean
[2022-02-08T13:06:06Z] <testuser[m]> how can symbol level of 0 show any backtrace
[2022-02-08T13:12:59Z] <testuser[m]> i think this is the relevant CreateDefault() function calls this constructor https://termbin.com/ruvk
[2022-02-08T13:13:35Z] <testuser[m]> god
[2022-02-08T13:13:36Z] <testuser[m]> that was it
[2022-02-08T13:13:40Z] <testuser[m]> WTF why didnt it say anything?
[2022-02-08T13:13:43Z] <testuser[m]> i didnt have a font installed
[2022-02-08T13:14:16Z] <testuser[m]> bruh bruh bruh
[2022-02-08T13:22:40Z] <illiliti> i have no logs with --single-process 
[2022-02-08T13:22:46Z] <illiliti> only sigtrap
[2022-02-08T13:22:59Z] <testuser[m]> did u install a font?
[2022-02-08T13:23:52Z] <illiliti> i did
[2022-02-08T13:25:32Z] <testuser[m]> http://0x0.st/oX1Q.png
[2022-02-08T13:25:38Z] <testuser[m]> #worksforme
[2022-02-08T13:25:49Z] <testuser[m]> what about your original command ? the --log-level one
[2022-02-08T13:27:10Z] <testuser[m]> `chromium --disable-gpu --log-level=1 --enable-logging=stderr` this works for me and shows logs too
[2022-02-08T13:27:15Z] <testuser[m]> disable gpu since im in chroot
[2022-02-08T13:30:21Z] <illiliti> do you have libva installed?
[2022-02-08T13:30:39Z] <testuser[m]> Yes
[2022-02-08T14:01:19Z] <illiliti> i have a strong feeling that this crash caused by extra optimizations
[2022-02-08T14:03:02Z] <testuser[m]> What cpu?
[2022-02-08T14:03:41Z] <illiliti> Intel(R) Core(TM) i5-7300HQ CPU @ 2.50GHz
[2022-02-08T14:03:48Z] <illiliti> you were able to run chromium when you recompiled it with debugging symbols, right?
[2022-02-08T14:04:18Z] <testuser[m]> I didn't recompile anything, just inspected the 1 line backtrace from the old, unstripped binary leftover in KISS_TMPDIR
[2022-02-08T14:04:21Z] <testuser[m]> And installed a font
[2022-02-08T14:04:44Z] <testuser[m]> I can send u that binary and u can check what backtrace it gives
[2022-02-08T14:08:11Z] <illiliti> do it
[2022-02-08T14:14:23Z] <testuser[m]> illiliti: curl -LO http://45.76.184.72:6666/chromium.tar.xz
[2022-02-08T14:14:55Z] <testuser[m]> sha256 5a24126bea842721f732c5f26bd9ec97b23920f63441676d54618dfa6b26c34e
[2022-02-08T14:15:32Z] <testuser[m]> u dont need to install it btw, just untar and run ./chromium from that dir
[2022-02-08T14:16:11Z] <illiliti> ok
[2022-02-08T14:18:01Z] <testuser[m]> downloaded? Asking cuz i need to close the process on server
[2022-02-08T14:18:23Z] <illiliti> 300k/s
[2022-02-08T14:18:27Z] <illiliti> xd
[2022-02-08T14:18:38Z] <testuser[m]> bruh
[2022-02-08T14:20:13Z] <illiliti> done
[2022-02-08T14:20:32Z] <testuser[m]> good
[2022-02-08T14:23:11Z] <illiliti> fucking hell
[2022-02-08T14:23:26Z] <testuser[m]> empty?
[2022-02-08T14:23:43Z] <illiliti> device::UdevLoader::Get()
[2022-02-08T14:23:46Z] <illiliti> i'm gonna kill it with fire
[2022-02-08T14:23:50Z] <testuser[m]> bruh
[2022-02-08T14:24:07Z] <illiliti> it works now
[2022-02-08T14:24:10Z] <illiliti> thanks
[2022-02-08T14:24:26Z] <testuser[m]> what was the issue?
[2022-02-08T14:24:31Z] <testuser[m]> mdev not running or something?
[2022-02-08T14:24:53Z] <illiliti> i forgot that chromium requires libudev
[2022-02-08T14:25:02Z] <testuser[m]> oh
[2022-02-08T14:25:12Z] <testuser[m]> ok this is weird
[2022-02-08T14:25:18Z] <testuser[m]> earlier it used to print out proper messages and trap
[2022-02-08T14:25:23Z] <testuser[m]> now its just trapping
[2022-02-08T14:25:30Z] <illiliti> this is very confusing
[2022-02-08T14:44:22Z] <illiliti> testuser[m]: https://source.chromium.org/chromium/chromium/src/+/main:device/udev_linux/BUILD.gn;l=7
[2022-02-08T14:45:32Z] <illiliti> use_udev=false
[2022-02-08T14:45:36Z] <illiliti> does it work?
[2022-02-08T14:46:43Z] <testuser[m]> I don't remember i think there were some issues
[2022-02-08T14:46:46Z] <testuser[m]> I'll try it tomorrow
[2022-02-08T14:48:13Z] <testuser[m]> https://github.com/kiss-community/community/issues/53#issuecomment-766639062
[2022-02-08T14:48:33Z] <testuser[m]> Damn its been 1 year
[2022-02-08T14:51:06Z] <illiliti> why this option still exist if doesn't work lol....
[2022-02-08T14:55:22Z] <illiliti> https://source.chromium.org/chromium/chromium/src/+/main:device/udev_linux/udev_loader.cc;l=22-39?q=UdevLoader&ss=chromium%2Fchromium%2Fsrc
[2022-02-08T14:56:39Z] <illiliti> i think this function should be nulled
[2022-02-08T14:57:45Z] <testuser[m]> return nullptr is after the CHECK(false) so im pretty sure stuff would start exploding if g_udev_loader was null
[2022-02-08T14:57:54Z] <testuser[m]> But I'll try
[2022-02-08T15:00:21Z] <testuser[m]> Also i don't get why they have this CHECK() bullshit just crash and print the file and line names...
[2022-02-08T15:02:20Z] <illiliti> i don't even know what it does.
[2022-02-08T15:02:38Z] <illiliti> maybe it triggers sigtrap, who knows...
[2022-02-08T15:03:44Z] <testuser[m]> Yeah that's what its for
[2022-02-08T16:23:36Z] <testuser[m]> Anyone got a golang binary
[2022-02-08T17:00:57Z] <phoebos> testuser[m]: https://tmp.bvnf.space/go@1.17.6-1.tar.gz
[2022-02-08T17:02:09Z] <midfavila-laptop> sup nerds
[2022-02-08T17:05:47Z] <testuser[m]> phoebos: Thanks but i ended up building on my server
[2022-02-08T17:05:50Z] <testuser[m]> Had to make swap
[2022-02-08T17:05:56Z] <testuser[m]> midfavila-laptop: Hi
[2022-02-08T17:11:36Z] <midfavila-laptop> https://laptop.pics/panasonic-toughbook-cf-m34/
[2022-02-08T17:11:43Z] <midfavila-laptop> check this guy out
[2022-02-08T17:11:46Z] <midfavila-laptop> it's so *small*
[2022-02-08T17:12:16Z] <midfavila-laptop> 9" ruggedized laptop
[2022-02-08T17:13:05Z] <phoebos> hi mid
[2022-02-08T17:13:14Z] <testuser[m]> Rich
[2022-02-08T17:13:22Z] <phoebos> your blog page has unclosed <p> tags
[2022-02-08T17:13:53Z] <midfavila-laptop> yeah probably
[2022-02-08T17:14:00Z] <midfavila-laptop> last time I was on SDF I tried using se to edit html
[2022-02-08T17:14:03Z] <midfavila-laptop> big mistake
[2022-02-08T17:14:09Z] <phoebos> big mistake
[2022-02-08T17:14:33Z] <midfavila-laptop> forked it the other day, after a few changes I think se'll be very serviceable
[2022-02-08T17:14:38Z] <midfavila-laptop> link to page btw?
[2022-02-08T17:15:21Z] <testuser[m]>  make site generator vro
[2022-02-08T17:15:26Z] <midfavila-laptop> soon:tm:
[2022-02-08T17:15:42Z] <midfavila-laptop> once i get around to learning m4 and rc properly I'm going to write a set of templates 
[2022-02-08T17:16:00Z] <testuser[m]> but vro sed
[2022-02-08T17:16:27Z] <midfavila-laptop> define "vro"
[2022-02-08T17:16:41Z] <testuser[m]>  bro
[2022-02-08T17:16:46Z] <midfavila-laptop> ah
[2022-02-08T17:16:47Z] <midfavila-laptop> i mean
[2022-02-08T17:16:51Z] <midfavila-laptop> sed is okay i guess
[2022-02-08T17:17:00Z] <midfavila-laptop> but I wouldn't want to use it for page generation
[2022-02-08T17:17:07Z] <midfavila-laptop> macro processor and shell is plenty for that
[2022-02-08T17:17:23Z] <midfavila-laptop> s/shell/rc/
[2022-02-08T17:17:32Z] <midfavila-laptop> if it wercs for uriel it wercs for me
[2022-02-08T17:19:18Z] <phoebos> midfavila-laptop: /pages/bloglist.html
[2022-02-08T17:19:33Z] <midfavila-laptop> danke
[2022-02-08T17:26:32Z] <midfavila-laptop> cool that's fixed
[2022-02-08T17:26:36Z] <midfavila-laptop> there were a few other things
[2022-02-08T17:26:50Z] <midfavila-laptop> forgot to declare elements in a list as list elements
[2022-02-08T17:30:25Z] <midfavila-laptop> looks like there are some broken image embeds too... I really should write a page generator
[2022-02-08T17:39:15Z] <illiliti> https://gitlab.freedesktop.org/wlroots/wlroots/-/merge_requests/3451
[2022-02-08T17:39:32Z] <midfavila-laptop> okay yeah I'm actually going to go write a generator right now
[2022-02-08T17:39:40Z] <midfavila-laptop> fixing this by hand isn't something I want to do today
[2022-02-08T17:39:41Z] <midfavila-laptop> or ever
[2022-02-08T17:42:27Z] <midfavila-laptop> huh. turns out m4 might actually be perfect for this
[2022-02-08T17:42:56Z] <midfavila-laptop> never really used it before so I was worried it would be tricky to impart structure on pages, but it apparently accepts parameterized macros