ChanServ changed the topic of #kisslinux to: Unnofficial KISS Linux community channel | https://kisscommunity.org | post logs or else | "An idiot admires complexity, a genius admires simplicity." -- Terry A. Davis
GalaxyNova has quit [Quit: Whoooooshh]
GalaxyNova has joined #kisslinux
davidgarland has joined #kisslinux
Andrei[m]1 has joined #kisslinux
barpthewire_ has joined #kisslinux
<GalaxyNova> hi barpthewire_
barpthewire has quit [Ping timeout: 252 seconds]
<GalaxyNova> ...
<testuser[m]> Hi
<testuser[m]> dilyn: no im just doing meson --prefix=/usr . output
davidgarland has quit [Quit: davidgarland]
dbz has joined #kisslinux
<GalaxyNova> is midori as lightweight as people make it out to be?
<testuser[m]> Is there a way to make pipewire fork() after setting itself up ? I'm currently running it via pipewire & but i need to poll for it's socket file to appear
<noocsharp> GalaxyNova: midori uses the chromium browser engine
<noocsharp> which i just learned by reading the wikipedia page
<GalaxyNova> oh
<GalaxyNova> i though it used webkit
<GalaxyNova> eitherway now I'm trying out badwolf
<testuser[m]> Webkit is about as fat as firefox so might as well use the usable one
<GalaxyNova> I'd like to disagree
<GalaxyNova> you don't need rust to compile webkit
<GalaxyNova> and generally I've found webkit to compile a lot faster than both firefox and chromium
<testuser[m]> Oh
<testuser[m]> In my exp both took around the same
<GalaxyNova> the rust compilation slows everything down
<GalaxyNova> it might also depend on what make jobs you have running
<GalaxyNova> i use -j16
<aws> browser war
<aws> tbh i can't stop finding myself browser hopping
<GalaxyNova> same
<GalaxyNova> they all seem to suck in some way or another
<GalaxyNova> I wish we could have firefox from ~2010 back :(
<GalaxyNova> palemoon doesn't work on wayland
<GalaxyNova> or so i think
<testuser[m]> In a few years when it's rebased
<testuser[m]> It might
<GalaxyNova> if it did I'd switch to it instantly
<GalaxyNova> firefox just seems to be getting worse and worse
<GalaxyNova> is there really any reason to use firefox over chromium?
<GalaxyNova> i think it builds slightly faster and that's it
<aws> firefox has been getting faster and snappier for me, and i personally like the new ui
<aws> but it doesn't has support for PWAs
<testuser[m]> ff input bar seems laggy
<testuser[m]> like whn you press enter there is a delay
<testuser[m]> no that was due to something else
<GalaxyNova> oh god
<GalaxyNova> aws: That's a good thing
<aws> not supporting PWAs ?
<GalaxyNova> yes
<aws> If webapps are here to stay, i'd rather use PWAs than electron
<GalaxyNova> I'd rather not use webapps
<GalaxyNova> btw can any other wyverkiss users confirm ruby not building correctly?
<testuser[m]> whats the error
<GalaxyNova> there's no error
<GalaxyNova> it just hangs
<GalaxyNova> on "config.status: creating ruby-runner.h"
<GalaxyNova> it's been stuck like that for 15 minutes
<GalaxyNova> I've tried building it in a regular kiss chroot and it builds
<testuser[m]> probably something with coreutils
<GalaxyNova> it is using 100% cpu though
<GalaxyNova> so
<GalaxyNova> should i leave it overnight or smt
<GalaxyNova> it's using the cpu so it must be doing... something
<aws> if only that was the case
<GalaxyNova> lol
<GalaxyNova> oh well...
<GalaxyNova> it seems to be working fine in a wyverkiss chroot
<GalaxyNova> something must be wrong with my system
<GalaxyNova> time to rebuild...
<testuser[m]> Alright finally got this working
<aws> are you using sway with the proprietary nvidia drivers
<testuser[m]> yeah
<aws> how well does it work
<testuser[m]> Very
<GalaxyNova> I'm guessing you're using it in gkiss?
<testuser[m]> other than no hw accel
<testuser[m]> for videos
<aws> neat
<GalaxyNova> is there any way to see which non-existant directories kiss is complaining about?
<GalaxyNova> when `kiss i pkg`
<GalaxyNova> KISS_DEBUG is not showing anything
<testuser[m]> KISS_DEBUG is for persistent package build dirs
<testuser[m]> do sh -x /usr/bin/kiss
GalaxyNova has quit [Quit: GalaxyNova]
GalaxyNova has joined #kisslinux
<GalaxyNova> testuser[m]: Confirmed the issue in a wyverkiss chroot, it might also happen in a normal kiss chroot
<GalaxyNova> it works fine in the original chroot but after you update the system it doesn't work
<testuser[m]> checking
<testuser[m]> its working fine GalaxyNova https://termbin.com/n3il
<testuser[m]> hmm but its stuck on version.c
<testuser[m]> not the same line you said
<GalaxyNova> oh yeah
<GalaxyNova> it's always near
<GalaxyNova> the `./revision.h unchanged`
<GalaxyNova> either way it's a problem
<GalaxyNova> testuser[m]
<GalaxyNova> micro_0
<GalaxyNova> micro_O
<GalaxyNova> i never know if it's a 0 or an O
<cem> regarding daemonization of pipewire, why don't you install it as a user service?
<testuser[m]> cuz im not running it as intended, just running in a sandbox along with dbus stuff so i can do screen sharing
<testuser[m]> nothing on host
<cem> Ah, I see
<testuser[m]> i just did a while [ ! -S $XDG_RUNTIME_DIR/pipewire-0 ] after forking it
<testuser[m]> forking meaning &
<cem> I mean that's basically how systemd socket units work :P
<GalaxyNova> I should try to see if the ruby problem happens in a an updated regular kiss chroot too
<cem> low level virtual mess
<testuser[m]> it does
<testuser[m]> GalaxyNova:
<cem> What version of ruby?
<testuser[m]> 3.0.2
<GalaxyNova> 3.0.2
<cem> I'll try rebuilding it, I have it installed
<GalaxyNova> it doesn't even error which is annoying
<GalaxyNova> just gets stuck after a while
<testuser[m]> maybe another busybox regression
<testuser[m]> but its a compiler task
<testuser[m]> so thats weird
<GalaxyNova> 3.0.0 and 3.0.1 don't build either
<GalaxyNova> so it has to be a problem with the toolchain
<GalaxyNova> Anyway, I gtg
<GalaxyNova> if you find the solution let me know
GalaxyNova has quit [Quit: nyaa~]
<cem> builds just fine for me on carbs
<cem> It's probably a toolchain issue
<testuser[m]> its a busybox issue confirmed
<testuser[m]> 1.33 works
<cem> Huh
<testuser[m]> the version.c output is misleading
<testuser[m]> probably cuz of multiple threads
<cem> Or maybe version.c is generated by the build system
<testuser[m]> why would is say "Compiling version.c" then
<testuser[m]> i did make V=1 -j1 it gives blank output, without -j1 it tells its generating some header
<cem> Although it's best to run a single thread build
<testuser[m]> (when stuck)
<cem> Can you send the manifest of busybox
<testuser[m]> https://termbin.com/ceo0 old
<testuser[m]> https://termbin.com/5dlv new
<testuser[m]> they had a regression in awk aswell
raph_ael has quit [Quit: WeeChat 3.1]
raph_ael has joined #kisslinux
<cem> Okay, I have lots of alternatives https://termbin.com/xt6w
<cem> I'm using nawk and gnu sort, I think those might be relevant
<cem> I'll try to rebuild using busybox utils
<testuser[m]> making mjit_config.h
<testuser[m]> it gets stuck when calling make_arch_flags from the sourced sh file
<testuser[m]> parse*
<testuser[m]> `sh` is the problematic utility
<testuser[m]> bruh
<cem> Ugh, I've been looking at that script for over ten minutes now
<cem> no idea why it can be stuck at all
<cem> maybe disable jit if you don't need it
<testuser[m]> i'd still like to fix the issue tho
<testuser[m]> what its supposed to do
<testuser[m]> (with the #define stuff in quotes)
<testuser[m]> bruh it works just fine if i do it manually
<testuser[m]> oh i forgor
<testuser[m]> im on the shell from 1.33.0
<testuser[m]> nvm
<testuser[m]> oh its a syntax error
<testuser[m]> why doesnt make quit
<testuser[m]> oh, after setting +x termbin.com/fbnp
<cem> but i'm also using busybox shell
<cem> even if they were not found, why would it be stuck anyways?
<cem> did you source tool/mjit_archflag.sh, because that's what the Makefile does
<testuser[m]> cem: I did set +x in the makefile so yeah
<cem> ah, I thought you ran those commands manually for a second
<cem> You should do the other way around and 'set -x'
<testuser[m]> Isn't -x for turning off debugging
<testuser[m]> So it would be the defult
<cem> no, +x turns it off
<testuser[m]> Oh
<cem> idk the idea behind it
<cem> same with -v and +v
<testuser[m]> same thing but just shows the commands it tried to run
<cem> huh
<testuser[m]> why doesnt it show the . command
<testuser[m]> the one for sourcing
<testuser[m]> oh i forgor the semicolon
<testuser[m]> so it ignored the rest of the line
<testuser[m]> lol
<cem> lmao
<testuser[m]> yeah so it just gets stuck after arch_flag=
<testuser[m]> wtf
<testuser[m]> Ok so the script assigns those 2 vars after that the next operation done by the makefile is stick, not the variable assignment
<testuser[m]> test "$(Q)" = @ || set -x;
<testuser[m]> no
<cem> It's parse_arch_flags
<testuser[m]> yeah i misred
<testuser[m]> read
<cem> both vars are empty for me
<testuser[m]> yeah
<testuser[m]> running manually works fine
<testuser[m]> but it doesnt get that far cuz if i put random echoes in parse_arch_flags it does nothing
<cem> Add a ':; return 0' to the top of parse_arch_flags function
<testuser[m]> same thing, nothing in debug log thingy either
<cem> Try specifying a different shell on the top of the Makefile
<cem> like SHELL = /bin/ksh or something
<testuser[m]> yeah that would fix it cuz the i confirmed that the issue is with the new busybox sh
<testuser[m]> oh
<testuser[m]> The output would still be helpfup
<testuser[m]> it works fine in a standalone makefile
<testuser[m]> ohhhhhhhh
<testuser[m]> the redirection screws it up
<cem> it doesn't work when you redirect it?
<testuser[m]> yeah if i remove the redirect it works fine
<cem> try redirecting with tee?
<testuser[m]> works
<cem> what a weird fucking bug
<testuser[m]> whenever a source is involved it fucks up, just putting an echo or something works fine aswell
<testuser[m]> letme try with bmake aswell
<cem> I don't understand why I can't reproduce it, though
<testuser[m]> what version
<testuser[m]> alright its definitely a busybox issue, same with bmake
<testuser[m]> there were like 10 commits to ash between 1.33 and 1.34
<testuser[m]> smallest repro example `{ . /usr/bin/kiss; } > /tmp/some`
<cem> Ah, I didn't know busybox 1.34 was released
<cem> I'm at 1.33.1
<cem> I thought you were mentioning 1.33.0 when you said 1.33
<testuser[m]> i mentioned 1.33.0 as working and 1.34.0 borked
<cem> I thought you meant 1.33.0 as working and 1.33.1 borked
<cem> lol
<cem> I'm building 1.34.0 now
<testuser[m]> it gets stuck after a call to close(), time to put in some printf's
<cem> yup, can reproduce now
<testuser[m]> almost there
<testuser[m]> soon
riteo has joined #kisslinux
<riteo> hiiiiiii!
<testuser[m]> hiiiiiiiiiiiiiiiiiiiiiii
mobinmob has joined #kisslinux
<testuser[m]> Lol it was fixed already
<testuser[m]> I even looked at this commit earlier
<riteo> talking about busybox, does anybody know how to allow `poweroff`, `reboot` and `halt` to work as a normal user without `doas`? I saw that Dylan did some interesting things with the `busybox` package, but I'm not sure if I should fork it to accomplish that
<testuser[m]> riteo: suid ?
<testuser[m]> better to make it specific to a group
<riteo> idk if I messed up something, but suidding /bin/{poweroff,reboot,halt} suidded the whole binary
<testuser[m]> Check the ssu package
<testuser[m]> Oh
<testuser[m]> Yeah sorry
<testuser[m]> It's all symlinks
<testuser[m]> busyvox and busybox-suid are 2 bins
<testuser[m]> So you can mess with it and make poweroff in busybox-suid
<riteo> that's what I feared
<testuser[m]> Better to suid the kpow utility instead
<riteo> kpow?
<testuser[m]> Type it
<riteo> eh, I'm not on my laptop, gimme a sec
<testuser[m]> its for reboot and shutdown, provised by baseinit
<testuser[m]> So it won't mess with busy box
<riteo> oh, I see
<riteo> so busybox uses that?
<testuser[m]> Wdym
<testuser[m]> Poweroff halt are provided by busybox, kpow is a 10 line C file written by dylan
<testuser[m]> Busybox one is independent of kpow
<riteo> oh I see
<riteo> so I should use that for shutting down my laptop?
<testuser[m]> kpow doesn't depend on a specific init either
<testuser[m]> Yeah
<riteo> I wonder why dylan left them in
<riteo> he could've just made them shell scripts with kpow{ p, r} in them
<testuser[m]> Wdym
<testuser[m]> That would call back into busybox
<riteo> wait, so busybox is independent of kpow but not the other way around?
<riteo> wait, I gotta see this program, gimme a sec
<testuser[m]> both are independent
<riteo> I'm confused
<testuser[m]> kpow is an init-agnostic binary to shut down
<testuser[m]> busybox poweroff is specific to busybox runit
<testuser[m]> busybox doesnt depend on kpow
<testuser[m]> kpow doesnt depend on busybox
<testuser[m]> kpow {r, p} == busybox {reboot, poweroff} functionally
<riteo> ok, there's no trouble on my side in this part
<testuser[m]> so if you setuid on the kpow binary it wont mess with busybox
<riteo> I see
<riteo> one last thing though
<riteo> kpow doesn't kill each process gracefully, does it?
<testuser[m]> it does what `man 3 reboot` describes
<testuser[m]> oh its man 2 not man 3
<testuser[m]> it sync()'s the disks though
<testuser[m]> the kpow program
<testuser[m]> busybox halt does the same thing
<testuser[m]> just that you got some more arguments to configure the behaviour
<riteo> mh, that would be a nice solution, but it completely skips whatever init system I have
<testuser[m]> why does it matter though
<testuser[m]> programs wont explode if incorrectly shut down, most of them dont even keep a persistent state
<testuser[m]> eg a database
<testuser[m]> even if they do , disks are sync()'d
<riteo> that doesn't sound either very elegant nor intended in a way. The comments on top of kpow.c talk about the end of the init system shutdown, not the start
<riteo> idk, thanks for letting me know about this program, but this feels wrong. rc.d supports hooks for shutdown too and this wouldn't run them if I were to add any
<testuser[m]> Oh
<testuser[m]> Yeah
<testuser[m]> Just use ssu :p
<testuser[m]> Password not needed for anything
<riteo> what's ssu
<testuser[m]> suid privilege escalation utility
<riteo> I wonder why dylan couldn't just have enabled the suid config file for busybox and bundled one there
<riteo> ig for safety? But then, to change that file you undeniably gotta be root (iirc it even has to be 600 with root as an owner)
<testuser[m]> You could also write a 2 line C program that just execve's the busybox poweroff symlink lol
<riteo> eh, true
<riteo> I might resort to that if I still don't get an answer to my problems
<riteo> I might as well just alias `poweroff` to `doas poweroff`, afterall I already set it to not ask a password
<riteo> I'll probably send him an email asking why he done that, toghether with another weird thing regarding unicode
<riteo> for the courious, he set the config to disable locale support (understandable for this distro's goals) but he also made unicode work only when LANG ends with utf8
<testuser[m]> Unicode support in what
<riteo> I messed up my terminal countless times because I would accidentally type è without noticing it
<riteo> testuser[m]: all busybox programs
<testuser[m]> Oh
<riteo> ik, that's weird as hell
<riteo> for now I added a script to /etc/profile.d
soliwilos has quit [Remote host closed the connection]
soliwilos has joined #kisslinux
mobinmob has quit [Quit: Connection closed for inactivity]
dbz_ has joined #kisslinux
dbz has quit [Read error: Connection reset by peer]
dbz_ has quit [Remote host closed the connection]
dbz_ has joined #kisslinux
jslick has quit [Ping timeout: 252 seconds]
jslick has joined #kisslinux
davidgarland has joined #kisslinux
davidgarland has quit [Quit: zzz]
Uks2 has quit [Quit: Byee]
Uks2 has joined #kisslinux
jess has quit []
mobinmob has joined #kisslinux
dilyn has joined #kisslinux
<dilyn> testuser[m]: chromium 93.0.4577.63 seems to no longer suffer that weird black box problem
dilyn has quit [Client Quit]
dilyn has joined #kisslinux
<riteo> well that was quick
<dilyn> it DOES however randomly crash tabs, I guess...
<riteo> oh you're back
<riteo> lol
<testuser[m]> dilyn: its fixed
<testuser[m]> chromium-InkDropHost-crash.patch
<testuser[m]> also .82 just dropped so you might as well rebuild it completely
<testuser[m]> :p
<dilyn> yeah well
<dilyn> i accept the zero days
<dilyn> also got it to build with pgo so that's fuuuuuunnnn
<dilyn> though when i generated the PGO data... the browser did NOT work. at all. everything crashed constantly. I hope that's a feature(tm)
<testuser[m]> arent you supposed to use the data given by poogle
<dilyn> they certainly offer pre-generated profiles updated four times a day
<dilyn> but this is called UNGNUGLED CHROMIUM sir
<testuser[m]> you can un-ungoogle it a bit
<testuser[m]> 345 line
<dilyn> lol
davidgarland has joined #kisslinux
Woodi has quit [Quit: ab -> c]
Uks2 has quit [Quit: Byee]
Uks2 has joined #kisslinux
mahmutov has joined #kisslinux
mobinmob has quit [Quit: Connection closed for inactivity]
mobinmob has joined #kisslinux
michalsieron has joined #kisslinux
michalsieron has quit [Ping timeout: 260 seconds]
<noocsharp> jesus christ why did i use epik
dbz_ has quit [Read error: Connection reset by peer]
felinae has quit [Killed (NickServ (GHOST command used by felinae-!~felinae@c-71-227-156-11.hsd1.wa.comcast.net))]
felinae has joined #kisslinux
<riteo> I'll go for now, bye everyone!
riteo has quit [Quit: epic busybox moment]
mobinmob has quit [Quit: Connection closed for inactivity]
davidgarland has quit [Quit: davidgarland]
mahmutov has quit [Ping timeout: 245 seconds]
felinae has quit [Killed (NickServ (GHOST command used by felinae-!~felinae@c-71-227-156-11.hsd1.wa.comcast.net))]
felinae1 has joined #kisslinux
Uks2 has quit [Quit: Byee]
Uks2 has joined #kisslinux
GalaxyNova has joined #kisslinux
<GalaxyNova> heyooo
<GalaxyNova> testuser[m]: did you figure out the reason why ruby wasn't building?
Uks2 has quit [Quit: Byee]
Uks2 has joined #kisslinux
Uks2 has quit [Ping timeout: 252 seconds]