<msk>
did any of you use xcape before switching to wayland? I can't find a good replacement
<dilyn>
you can usually just use the compositor's config itself
<msk>
dwl only has single key --> command configuration options
GalaxyNova has quit [Quit: Whoooooshh]
<msk>
I don't like how compositors have so many roles to fill
micr0 has joined #kisslinux
micr0 has quit [Changing host]
micr0 has joined #kisslinux
Uks3 has joined #kisslinux
<micr0>
hey, anyone remember that hooks package that makes it easy to add and remove individual hooks?
<micr0>
I am looking to package individual kiss hooks, like kiss-hooks-remove-man-pages and kiss-hooks-build-time so people can just install or uninstall them
<micr0>
but making a "proxy" script seems kinda weird
Uks2 has quit [Ping timeout: 276 seconds]
GalaxyNova has joined #kisslinux
<illiliti>
you mean system hooks?
<illiliti>
this feature isn't implemented in kiss
<GalaxyNova>
issue with ccls is that it doesn't support gotodefinition
<GalaxyNova>
which is unfortunate
<GalaxyNova>
but clangd is impossible to package :/
<humaid_02>
Package 'gl', required by epoxy, not found.
<micr0>
ok, well I just made some packages to easily install hooks for people
<micr0>
`kiss build kiss-hooks-build-duration` for example
<micr0>
does two things: installs kiss-hooks, which sets KISS_HOOK in /etc/profile.d/kiss-hooks.sh, and a wrapper script in /etc/kiss/hooks/kiss-hooks
<micr0>
the wrapper script just looks for everything in /etc/kiss/hooks/*.sh and sources them in order, with the appropriate environment variables
<micr0>
this makes it easy to add or remove hooks
<micr0>
fair warning, this is in my personal repo as hooks *could* be dangerous. if enough people are into it, i'll make a PR to kiss-community/community or maybe make a new hooks repo under kiss-community
<micr0>
also happy for people to contribute hooks
<humaid_02>
Nice. This should be upstream..maybe.
<micr0>
I basically just copy-pasted section 6.0 of `kiss help package-manager`, as I anticipate some people might want `kiss install kiss-hooks-remove-man-pages` once #263 or whatever the issue number is closed
<micr0>
humaid_02 which upstream? I think dylan has enough on his plate, and its probably worth having a couple of weeks of people trying it out, coming up with competing versions, etc
<micr0>
thanks illiliti
<micr0>
for example, I was thinking of creating a KISS_HOOKS or KISS_HOOKS_PATH variable, that works like KISS_PATH, though not sure if thats simpler or more complex
msk has quit [Quit: leaving]
<illiliti>
KISS_HOOK is really messy thing
<illiliti>
i dislike it
<humaid_02>
micr0: I think you are right.
<micr0>
illiliti its a *little* messy but to be fair, I was able to package up independent hooks using only 5 lines of supporting shell script
<micr0>
in like <30 minutes, porting 3 hooks to it
kyxor has quit [Read error: Connection reset by peer]
<humaid_02>
Just wanted to report, did kiss-rest. and kiss removed baseinit... resulted in read-only file system.
<micr0>
hmm does anyone have a resource for converting bytes into grams?
akira0172 has joined #kisslinux
akira0172 has quit [Client Quit]
<dilyn>
baseinit is not included with the base system :V
<akira01>
hey guys
<humaid_02>
dylin: really? Why not?
<akira01>
anyone have problem with kiss logs?
<humaid_02>
Hi akira.
<humaid_02>
by the way dylin.. what is this gl package required by epoxy?
GalaxyNova has joined #kisslinux
claudia02 has joined #kisslinux
<claudia02>
humaid_02: you are transitioning from xorg to wayland rn?
<micr0>
i made a dumb function that tells you the weight of your kiss install in picograms
<claudia02>
this is prob because existing packages still link to mesa with libgl or something.
<claudia02>
The kiss-rest you have just done is the right thing.
<claudia02>
Make sure to delete your $HOME/.cache/kiss/bin
kyxor has joined #kisslinux
<dilyn>
baseinit is not included because... idk why. It just hasn't been a part of the tarballs since version 1 and it's stayed that way
<claudia02>
This error results from detecting stuff and setting automatically the wriong configure args
<dilyn>
even though it's the only init in repo/ ... so...
<dilyn>
as for GL, it's like what claudia said. you probably have something off with a package. we'd have to know what you've done
<GalaxyNova>
guess it makes the tarball smaller?
<GalaxyNova>
by a few kilobytes
<dilyn>
make sure you have the right remote, remove any cached packages, etc. because you've already done a reset, you just have to do kiss b sway and you're basically done
<micr0>
illiliti what do you mean by packages cannot use KISS_HOOKS_PATH reliably? if you want an example of how i currently have independent packages appending to KISS_PATH, I can share that
<illiliti>
KISS_HOOKS_PATH can be anything
<illiliti>
it's user controled variable
<claudia02>
I see kiss-reset resets towards a working chroot. There is no baseinit needed.
<micr0>
sure, which is why a well-behaving package really only has one safe option - to append
<micr0>
i manage my package repositories with `kiss`
<micr0>
in addition to cloning the correct repo, each package adds a file in /etc/profile.d/ with a single line: export KISS_PATH="/path/to/my/repo:$KISS_PATH"
<micr0>
so the user is free to add, remove, overwrite the variable
<noocsharp>
how meta
<micr0>
i mean honestly its a testament to the simplicity of the package management of kiss, that its so easy to add features like this without them needing to be hacked or forked or whatever
<illiliti>
micr0: how do you maintain order of KISS_PATH?
<illiliti>
just curious
<micr0>
illiliti thats a very good question
<micr0>
the same way i maintain order with when I managed it manually
<micr0>
that is, i just appended stuff
<micr0>
and when things conflicted or broke, i tried to not have overlapping packages
<micr0>
so i would either fork or stop using said repository
<micr0>
if i really needed two repositories to have a different order than the paths came out to be, I would just prepend that locally (i.e. in .profile)
<micr0>
I do think its a limitation that i cannot specify kiss/extra/firefox vs jedahan/firefox, but ... surprisingly that limitation hasnt been too painful
<noocsharp>
couldn't you also do the typical thing where you start your file in profile.d with a number?
<noocsharp>
that way ordering could be in your repo repo
<micr0>
noocsharp sure, but then you'd have one of two issues: centralized ordering that may conflict with what a user wants
<micr0>
dynamic ordering that would depend on install time O_O
<micr0>
though honestly, kiss is full of dynamic ordering, so maybe thats not an issue
<noocsharp>
true, i was assuming the user controls the repo repo anyway so they have control anyway
<micr0>
(or rather, dynamic install-time stuff)
<noocsharp>
wow, i should read when i type
<micr0>
oh, i mean yeah if each person had their own repo-repo then yeah that works fine
<micr0>
my repo-repo is meant to be used by other folks, if helpful
<noocsharp>
oh
<micr0>
if we wanted to get REAL meta about it...i had an experiment where i added repositories to the dependencies of packages
<micr0>
so like, if a package in my repo, depended on a package in your repo, it would have repo-noocsharp in the depends file
<micr0>
that way I could just `kiss build mycoolpackage` and it would also clone and add your repo, etc.
zenomat has quit [Ping timeout: 276 seconds]
<noocsharp>
i mean at some point we'll need a repo of repo repos
<micr0>
it felt a little too experimental to share
<micr0>
noocsharp i think thats whats called an "awesome list" which i kinda abhor
<noocsharp>
cool idea nonetheless
<micr0>
thanks. it works for me. i mostly dont think about it
zenomat has joined #kisslinux
zenomat has joined #kisslinux
zenomat has quit [Changing host]
<akira01>
man wtf firefox wayland had?
<akira01>
i did the same test as i did in firefox xorg
<akira01>
i only get criticals in browseraudit.com
<GalaxyNova>
akira01: What is browseraudit?
<akira01>
website to test your browser security
<kiedtl>
micr0: there was this thing called "fileweight" (search for it in lobste.rs or HN's submissions, I don't have the link rn), I think it's similar to what you're tyring to do
humaid_02 has quit [Remote host closed the connection]
humaid_02 has joined #kisslinux
<micr0>
btw illiliti I can add /usr/share/kiss/hooks to the kiss-hook wrapper
<micr0>
i mean what you are describing is just setting a default for KISS_HOOKS_PATH="/etc/kiss/hooks:/usr/share/kiss/hooks"
<micr0>
I'll spend an hour seeing if i can make it work that way
<micr0>
kiedtl perfect thanks
clemens3 has quit [Ping timeout: 272 seconds]
<akira01>
anyone uses firefox?
<humaid_02>
Claudia dylin, thank you.. I'll do it.
clemens3 has joined #kisslinux
<konimex>
akira01: currently building 90
<akira01>
so when you build
<akira01>
can confirm if you about:config had this option "media.ffmpeg.dmabuf-textures.enabled" ?
<konimex>
in my current ff89, no media.ffmpeg.dmabuf-textures.enabled, but I have media.ffmpeg.dmabuf-textures.disabled (currently false)
<akira01>
how can i truly know if firefox is using hardware acceleration?
<GalaxyNova>
> open htop
<GalaxyNova>
> resize firefox window like crazy
<GalaxyNova>
watch the cpu output
<micr0>
illiliti I updated my kiss-hooks experiment to look by default in /etc/kiss/hooks and /usr/share/kiss/hooks. these can be changed to any paths you want by updating KISS_HOOKS_PATH, much like KISS_PATH.
<illiliti>
nice
akira01 has quit [Ping timeout: 246 seconds]
akira01 has joined #kisslinux
humaid_02 has quit [Remote host closed the connection]
<akira01>
GalaxyNova: when i resize firefox like a crazy
<akira01>
The cpu is getting to 20%
<akira01>
This is good or not?
<GalaxyNova>
not a good sign xD
<akira01>
Gosh lol
<akira01>
what can i do? i follow the steps in wiki
<micr0>
ok, last improvement: kiss-hooks will now source hooks named after their type, git-style. so you dont need to match against the type if you dont want to. this is in addition to running anything with a .sh extension, like current KISS_HOOK style
necromansy has joined #kisslinux
* necromansy
waves
* micr0
o/
<necromansy>
whats up
<micr0>
illiliti the kiss-hook package has turned into a truly bloated 17 lines of shell
<micr0>
how will i live with the added 1.043448e-11 grams
claudia02 has quit [Ping timeout: 246 seconds]
<micr0>
necromansy hacking on hooks so people can just `kiss install kiss-hooks-remove-man-pages` or `kiss remove kiss-hooks-build-duration`, etc.
<necromansy>
im guessing the man pages hook is for internal consistency? but noice, i really should make more use of hooks
<akira01>
anyone can use volume keys and amixer?
<micr0>
necromansy both hooks are just ones I copied from `kiss help package-manager` section 6
<akira01>
amixer not work with volume keys in sway-tiny and i dont know why
<necromansy>
micr0 ah fair nuff, and yeah that that issue was what i was referring to
<acheam>
dilyn: your no atomic patch doesnt apply with busybox or openbsd patch btw
<micr0>
as for motivation behind the changes, my interpretation of the general direction of kiss dylan is converging on i could provide but might be a bit far removed from what you were asking
<micr0>
necromansy I'd say try out my kiss-hooks packages, but I wrote them this evening over about an hour so itll probably light your computer on fire
<necromansy>
yeah i'll geeze at them *carefully*
<dilyn>
my patch for what now?
<acheam>
removing libatomic from chromium
<dilyn>
well I'm sorry your terrible patch implementations don't respect my hard work >=|
<dilyn>
I would wager that this patch is just a sed anyways...
<acheam>
well its on two files
GalaxyNova has quit [Read error: Connection reset by peer]
<acheam>
which makes it less convinient for sed
<acheam>
but yeah its literally just removing 1 word
<dilyn>
i was just too lazy to check at the time if that single word appeared multiple times
<dilyn>
pax does an interesting thing when extracting chromium...
<testuser[m]>
I didnt have such issues even on nvidia gpu lol
<testuser[m]>
gdb /usr/bin/firefox
<testuser[m]>
run
<testuser[m]>
backtrace
<GalaxyNova>
firefox reset itself
<GalaxyNova>
aaaannd nope
<GalaxyNova>
keeps segfaulting
<GalaxyNova>
I'm going to just rebuilt it
<testuser[m]>
That won't do much
<testuser[m]>
Check where the segfault is
<GalaxyNova>
maybe firefox 90 fixes something
<GalaxyNova>
ok
<GalaxyNova>
built firefox 90
<GalaxyNova>
maybe i have to rebuild ffmpeg?
<testuser[m]>
You need to rebuild the packages that depend on a library in case of ABI breakage, not the library itself
<testuser[m]>
ffmpeg hasnt been updated after ff update anyways
<GalaxyNova>
testuser[m]: How did you get video accelaration working in firefox?
<testuser[m]>
i didnt
<testuser[m]>
Did you check the post install ? I think dylan put sone instructions there
<testuser[m]>
My gpu will use vdpau anyway, i don't think that works kn wayland
<GalaxyNova>
following the instructions lead to the segfault
<GalaxyNova>
> the about:config changes
schillingklaus has joined #kisslinux
<GalaxyNova>
hi schillingklaus
<GalaxyNova>
switch to wayland so you can suffer like me
<testuser[m]>
Mind doing the gdb thing i said
<GalaxyNova>
sure
consolers has joined #kisslinux
<consolers>
i'm still getting the same firefox build error that i saw reported here 2 months ago
<consolers>
firefox/build.amd64/dist/include/nsTHashtable.h:317:27: error: no matching function for call to 'nsTHashtable<detail::VoidPtrHashKey>::WithEntryHandle(const void*&, const fallible_t&, nsTHashtable<detail::VoidPtrHashKey>::PutEntry(nsTHashtable<detail::VoidPtrHashKey>::KeyType, const fallible_t&)::<lambda(auto:9)>)'
<travankor>
might be interesting if you want a better maintained set of bsd style utils
<acheam>
bsdutils has a wacky build system, and annoying dependencies
* travankor
didn't use it
<acheam>
(I packaged most of them already though)
<travankor>
are you using that as a busybox replacement?
akira01 has joined #kisslinux
<akira01>
hi guys
<akira01>
any progress with vaapi in firefox?
<acheam>
travankor: no
<acheam>
but im looking at my options
<acheam>
most likely right now is obsd baseutils + procps-ng + util-linux
<soliwilos>
I've noticed some other musl distro's changing stack-size with ldflags for lld and clang, and kiss packages are not, what is the difference?
<travankor>
acheam: wc output is formatted a bit differently, and grep -P being missing might break some things
<travankor>
that's off the top of my head, there might be other issues too
<testuser[m]>
soliwilos example ? Most packages don't need it
<testuser[m]>
Stack size issues will be runtime problems so its possible that the maintainers just didnt encounter them
<testuser[m]>
And it's only an issue on threaded programs
<travankor>
soliwilos: testuser[m]: this is a good article on the topic:
<noocsharp>
kiss has changed so much in the past week
<noocsharp>
maybe it's time to finally switch my repos
<kqz>
exciting times
mahmutov has joined #kisslinux
<dilyn>
gross
<dilyn>
I hate it
<noocsharp>
the version variables in the build file?
<konimex>
will the version variable thing even be possible in the C version of the package manager?
<testuser[m]>
yeah it seems unnecessary
<ang>
just runnin a hook, instead of sourcing it, is much better
<ang>
I can finally write my kiss hooks in nodejs
<dilyn>
#define thisCommit SuxBalls
<acheam>
idk I like it
<kqz>
will have to play around with it but that seems like a nice change, seems like a nice qol addition for updating packages
<acheam>
its safer, and better matches the behavior of the build file
<kqz>
less maintenance burden ayaya
<ang>
dylan mentioned making queue numbers available to hooks but I don't see it
<dilyn>
it feels like basically the same amount of burden
<testuser[m]>
what should be the norm of community repos with regard to kiss features ? can they be adopted as soon as a new release is out ?
<dilyn>
yes
<kqz>
i mean you just save a step, i'll take it
<dilyn>
what step is being saved though...
<kqz>
updating a sources file isn't a mundane task at all, but when doin things in bulk i'll take those savings
<dilyn>
idgi
<noocsharp>
konimex: i think the c program can just add variables to the environment of the build script
<dilyn>
I'm not understanding what anyone is going to do with this that will reduce hassle. it only seems like more of a hassle
humaid_02 has joined #kisslinux
<kqz>
what im seeing here is that now i can just set a sources file once with the necessary variables, then whenever i have to update a package, all i have to do is update the version file and kiss c and im off to the races, as opposed to updating version + sources
<testuser[m]>
not for everything tho, like libexif pkg version 0.6.22 but source libexif-0_6_22-release/libexif-0.6.22.tar.xz
<kqz>
ah yeah it'll fall through for stuff like that
<acheam>
this is how APKBUILDs and PKGBUILDs work
<dilyn>
I'm still very confused
<konimex>
dilyn: I believe the proper term is "outsourcing the hassle", instead of making the user responsible, the package manager will take some of the burden instead
<acheam>
it seems to work well for those MUCH bigger distributions
<dilyn>
the problem with what PKGBUILDs do with versions and sources is that they're cludgy and useless
<acheam>
bruh
<dilyn>
?!
<dilyn>
it's just not good
<testuser[m]>
tbh if this scheme can work for packages, then a simple sed 's/0.1.1/0.1.2/g' version source -i will work just as well
<dilyn>
precisely
<dilyn>
and for packages where a sed doesn't work, neither does this
<dilyn>
this doesn't help rust packages if they're packaged according to guidestones
<claudia>
I might remember a much earlier discussion the an advantage of the kiss pkg format is, thats is clean and easy readable. Not like apk/pkgbuilds.
<kqz>
yeah it's not the most readible for sure
<cem>
This is an example m4 file for sqlite
<dilyn>
why is m4 so strange
<kqz>
is there an example sources file with this change?
<acheam>
oh so just using some other pgp implementation
<acheam>
unfortunately all the good ones out there right now are rust
<akira01>
guys the git.k1sslinux.org is missing bemenu
<acheam>
dilyn: have you done it?
<dilyn>
akira01: not anymore it isn't
<testuser[m]>
Most rust stuff has bins so you don't need rust
<acheam>
testuser[m]: I like to avoid precompiled bins
<acheam>
especially for something security related
<acheam>
haskell (pandoc, shellcheck) is my exception to this
<akira01>
dilyn: we will use only github?
<dilyn>
acheam: i've done a little (just enough to copy a password)
<acheam>
using what gpg altenative?
<acheam>
thats all I need it for
<dilyn>
akira01: sorry that was a confusing thing to say. bemenu should be on git.k1sslinux.org now
<dilyn>
BUT! I don't know how long I will keep git.k1sslinux.org up for
<acheam>
4eva
<dilyn>
i'll let y'all know when it's going away tho
<dilyn>
lolno
<dilyn>
it would migrate tho :P
<dilyn>
I'm mostly playing with rnp, though something that doesn't require cmake+botan+json-c would be nice. so signify, maybe
<testuser[m]>
rnp?
<testuser[m]>
oh
<acheam>
I thoughtwas only for signing, not for encryption dilyn
<dilyn>
you mean for passwords?
<acheam>
yes
<acheam>
for pash
<dilyn>
I don't find myself making new passwords all that often lol
<dilyn>
I think I've made six in the last two years or something...
<acheam>
yeah but for decrypting it
<dilyn>
and that's a lot
<dilyn>
oh well
<dilyn>
idk mang
<dilyn>
i haven't decided on what my use case will look like :P maybe I'll go back to sane passwords
<kqz>
https://github.com/stealth/opmsg this has always piqued my interest but i've never implemented anything with it, will probably be my next project though while waiting to start new job
<dilyn>
afaik you can't alter gpg with git hooks, just git config (gpg.command)
mahmutov has quit [Ping timeout: 245 seconds]
<acheam>
okay rebuilding with fixed installation steps and vaapi
<acheam>
its in the ccache so hopefullt should go quick
<testuser[m]>
its gonna be updated soon
<testuser[m]>
oh
<acheam>
eh
<testuser[m]>
yeah it'll be done in 10 min
<acheam>
oh
<acheam>
ill wait then lol
<testuser[m]>
i mean the ccache build
<dilyn>
lol
<acheam>
oh
<acheam>
extracting sources takes looong time
<testuser[m]>
maybe we should release a Kromium tarball with all useless 3rd party stuff removed
<acheam>
like what?
<acheam>
also it pains me that yalls chromium builds have the properitary stuff enabled
<dilyn>
i want to watch twitch chess streams acheam
<dilyn>
fight me
<testuser[m]>
videos dont work without it
<acheam>
idc
<acheam>
mpv exists
<testuser[m]>
<acheam "like what?"> pretty much everything in third_party (6gb) is useless
<acheam>
i can play video just fine
<testuser[m]>
also it has chromeos
<dilyn>
mpv is gpl...
<dilyn>
also why would I ever want to do that
<cem>
mpv is really good
<acheam>
oh wait
<dilyn>
well it isn't good for my meme machine cem
<acheam>
i had proprietary codecs enabled
schillingklaus has quit [Ping timeout: 255 seconds]
<dilyn>
it was only a matter of time until acheam realized
<dilyn>
he fucked up
<acheam>
lol
<testuser[m]>
dilyn qt5 build fails with ERROR: The OpenGL functionality tests failed!, it needs libglvnd ? https://termbin.com/yjqe
<acheam>
i redid the build script like 3 times it must have gotten lost in the shuffle
<dilyn>
definitely doesn't
<cem>
use ffmpeg's player, I guess?
<dilyn>
ffmpeg is also gpl if you enable anything useful
<cem>
well, you need to enable useful things in order to stream
<dilyn>
lol
<dilyn>
chromium bundles ffmpeg et al anyways so there's honestly no reason I shouldn't have that whole stack
<dilyn>
except to say that i only have 80 packages installed neener neener neener
<acheam>
I use system ffmpeg
<acheam>
do you never need ffmpeg outside of your browser?
<cem>
One day dilyn will break and use GPL software
<acheam>
nobody tell him about Linux
<dilyn>
i use GPL software on every install except *this one*
<cem>
nobody tell him about GNU make
<dilyn>
:(
<acheam>
every install but this one is 50%?
<dilyn>
ten gpl packages, 4 of which are make-time only, one of which is the kernel, another is util-linux, two for sound, one for git, one for signing commits, libelf because amdgpu fucking SUCKS
<dilyn>
technically i could say perl is licensed differently I suppose
<ang>
guys, never mention bsd infront of dilyn
<dilyn>
:P
<akira01>
how can i change border in sway?
<dilyn>
BSD uses GCC and is FINE WITH THAT ANG
<akira01>
just put in config file?
<acheam>
dilyn: no?
<acheam>
I thought fbsd was gcc free?
<ang>
just uninstall gcc then, no need to compile from source :p
<acheam>
and anyways none of them use gcc 4+, beacuse gpl3
<kqz>
yes akira01, you use the config file to configure sway things ;p
<dilyn>
one of some such bsds has gcc somewhere
<dilyn>
if I wanted to use BSD I'd be using bsd
<acheam>
meanwhile im over here stealing as many obsd programs as i can
micro_O has quit [Ping timeout: 252 seconds]
<dilyn>
cem stop fekkin stalking me
<cem>
dilyn: nö
<dilyn>
|_|
<cem>
dilyn: I'm right outside open the door
<dilyn>
now THAT would be incredible
<cem>
lmao
schillingklaus has joined #kisslinux
<cem>
I'm porting some more openbsd tools to possibly replace busybox
<schillingklaus>
does landley's toybox not do the job?
<cem>
I'm more so aiming for sbase + obsd tools
<acheam>
cem: thank you for your noble work
<acheam>
have you seen baseutils?
<acheam>
github/ibara
<acheam>
he has good stuff
<schillingklaus>
do obsd tools include the openbsd korn shell?
<cem>
I haven't ported ksh yet, but I will
<acheam>
there is ibara/oksh
<acheam>
its what im using at the moment
<schillingklaus>
the fork mksh is more easily portable
<acheam>
but its less close to upstream ksh
<schillingklaus>
from the makers of pax mirabilis
micro_O has joined #kisslinux
<acheam>
mirbsd does good stuff
<cem>
acheam: yeah, I know ibara's ports, though they are quite outdated
<cem>
I initially created the repository ~a year ago to port to bearssl
<cem>
then I found out about michael's patches lol
<schillingklaus>
would that divorce openbsd userland from libressl?
<cem>
Nah, it still works for both
<acheam>
cem: they're not that bad
<acheam>
oksh is kept up to date
<acheam>
baseutils is two years old
<acheam>
as long as they're like 6.6 or newer thats good enough for me
<cem>
yeah, that's true
<schillingklaus>
there was recently aalbus, which ported netbsd userland to musl linux; but there has been no life sign from them since half a year
<acheam>
kyxor: that issue that your reported of the shell exiting before the terminal or something seems to happen with oksh as well
<acheam>
so not a busybox problem
<kyxor>
doesn't happen with bash for some reason though
<kyxor>
I have a feeling that kernel is bugged
<kyxor>
cause on other pc the only difference I find is its running kernel 5.10, and the bugged runs 5.13, all xorg libs are same version, same busybox version, etc
<kyxor>
or its a complier bug
<kyxor>
because cpus are different so different optimizations are in the software
mrlix has quit [Ping timeout: 246 seconds]
<kyxor>
or some other undefined behavior code that is optimized differently. I haven't actually debug it yet, but these are the options
<kyxor>
Thing is I don't remeber getting this bug before I updated to 5.12+, but maybe I just wasn't lucky to get it
lolololol has joined #kisslinux
lolololol is now known as sad_plan
<sad_plan>
hi
<sad_plan>
rust seems to fail for me. EOFError: compressed file ended before the end-of-stream marker was reached.
<acheam>
hmm im torn between the obsd baseutils and sbase
<acheam>
also need a good sed implementation
<sad_plan>
sbase is nothing new, why not give obsd a try instead?
<acheam>
thats fair
<acheam>
and obsd is... well, obsd
<acheam>
beutiful
<sad_plan>
another reason to choose obsd baseutils
<acheam>
the only problem is that its sed doesnt work with kiss
Uks2 has joined #kisslinux
<kyxor>
or this is that prominent fork() race condition bug
<sad_plan>
fork it? or avoid it, and just use a separate sed
mrlix has joined #kisslinux
<kyxor>
cause its likely to reproduce if closing and opening terminals fast
helby has joined #kisslinux
<acheam>
sad_plan: yeah I just need to find a good sed to use
<mrlix>
zsh is the best i have used
<acheam>
?
<acheam>
sed not shed
<acheam>
s/shed/shell/g
<acheam>
my brain is bad lol
<mrlix>
oh i was replying to some old stuff
<acheam>
ah
<mrlix>
but zsh doesn't have that bug where text glitches once it wraps around
Uks3 has quit [Ping timeout: 265 seconds]
<mrlix>
it just wraps around nicely
<sad_plan>
there is this sed/sed-i in community though. atleast there was. if it hasnt been dropped aswell
<acheam>
sed-i is just a wrapper around another sed
<sad_plan>
ah
<acheam>
and -i isnt the problem
<sad_plan>
I see
<acheam>
it just hangs on certain expressions
<acheam>
idk why
<sad_plan>
hm
<akira01>
anyone knows a way to automate the use of wbg?
<kyxor>
archeam: wait, st launches shell as a sub process. that's right. there was a similar bug some time ago with st launching child processes using -e flag, I reported that to suckless and they thought I was crazy, but then that bug went away with some kernel update or whatever
<kyxor>
so I think this bug has the same nature, as the other one, but only for terminating of child processes
<kqz>
akira01: wym u can't set it in sway just add "exec wbg path/to/file.png" to your sway config
<acheam>
ack ubase sed doesnt have -i
<acheam>
that leaves me with either fixing 75 sed -i usages, or finding a new sed
<kyxor>
hiltjo, didn't recognize it as a bug, I can't find that email but it may be in the mailing list archives, I wonder if I can still reproduce it though
<noocsharp>
there's that sed wrapper, right?
<acheam>
yes
<acheam>
but id rather not overcomplicate things by using it
<acheam>
so instead i'll just use a more bloated, non posix conforming sed
<acheam>
i guess i'll just use busybox sed for now
mrlix87 has joined #kisslinux
<sad_plan>
did anyone ask about the DRMSetMaster for sway-tiny?
mrlix has quit [Ping timeout: 246 seconds]
<riteo>
gtg for now, cya later!
riteo has quit [Quit: epic short session moment]
<acheam>
oh crud I also need a device manager
<sad_plan>
smdev!
<sad_plan>
or just use busybox's mdev for now
<acheam>
well i'll have to replace it at some point
<acheam>
might as well be now
<acheam>
but I think those are the last two busybox things
<kyxor>
archeam: oh yeaha that race condition bug is still present. The reason I can't experience it anymore is because hardware gotten faster, so no race condition, but calling usleep(99999) triggers it
<acheam>
sed and mdev
<acheam>
still no r
<acheam>
lol
<acheam>
hmm thats annoying
<sad_plan>
thats my thought aswell. if busybox goes anyway, theres no reason to not switch out parts of it now anyway. one step closer
<kyxor>
it's messed up, and there is no way to fix the size
<akira01>
kqz: i mean if i login in sway the wallpaper will just be set and no need to do a command or somenthing
<kqz>
yes, that is what that command will do
Uks2 has quit [Quit: Byee]
Uks2 has joined #kisslinux
<kqz>
are you trying to set the wallpaper via telekinesis? what other way would you run it?
<akira01>
first
<akira01>
that you mentioned
<akira01>
is to put in my sway/config?
<akira01>
because if is
<kqz>
yes, typically this is ~/.confg/sway/config
<akira01>
congrats dont work
<acheam>
actually I think I will go for the full suckless setup
<kqz>
did you restart sway after putting it in? exec will only run upon sway starting, not reloading
<acheam>
because baseutils isnt very complete
<acheam>
(to be fair, dylan did warn me of that)
<akira01>
yeah i restart
<acheam>
nbase is more complete but also has a wacky build system
<kqz>
does it work if you run the same command from a terminal in sway?
<akira01>
yeah
<akira01>
and if i kill
<akira01>
the bg disapear
<akira01>
if i put in .profile
<akira01>
dont work too
<kqz>
yeah it's a daemon so it has to stay running, try creating a script that has "wpg path/to/file.png" and then exec script.sh in ~/.config/sway/config and see if that works
<kqz>
though arguments should work fine in the sway config file
<akira01>
i will try
<akira01>
already create the script
<akira01>
i used with bemenu but i dont run lol
<akira01>
just in term
<kqz>
also could be the environment, if $PATH isn't set before sway starts it may not find it
<akira01>
other stranger thing
<akira01>
amixer not work with sway
<akira01>
Xf86Volume works with other things but cant work with amixer
<omanom>
can you run amixer in terminal?
<acheam>
the wallpaper setting utility is a daemon???
<acheam>
dilyn: how possible is it to not use perl in the chromium build
<acheam>
I assume that you'd have tried already
<dilyn>
uh iirc it's...
<dilyn>
very not possible
<acheam>
whats it used for?
<acheam>
I might just create a chromium megapackage
<acheam>
with chromium, python2, nss, etc all in one package
<acheam>
ok, after a day of using it, I understand why dilyn likes chromium so much
akira01 has quit [Ping timeout: 245 seconds]
<acheam>
and thats coming from someone who wrote his own browser (sort-of)
<dilyn>
yeah, it's a *good* browser
<dilyn>
I assume perl is used pretty extensively
micr0 has joined #kisslinux
micr0 has joined #kisslinux
micr0 has quit [Changing host]
<acheam>
can someone help me understand linux-libre?
<acheam>
like if I build my very slimmed down linux kernel
<acheam>
with none of the device specific stuff enabled
<acheam>
will I still have binary blobs?
<acheam>
like, I don't get the difference between binary blobs, and linux-firmware
<noocsharp>
linux-firmware contains binary blobs
<acheam>
okay I understand that much
<acheam>
but if I don't use linux-firmware, amd I still using binary blobs?
dilyn has quit [Quit: Connection closed]
<noocsharp>
not that i'm aware
<noocsharp>
i suppose it depends on hardware
<acheam>
then what does linux-libre do?
<noocsharp>
i guess gets rid of things that require binary blobs in linux-firmware
<noocsharp>
maybe reimplements them as free software
<noocsharp>
according to gnu, there are things in the kernel source code that are not free, but i don't know what they are
<acheam>
hmmm
<acheam>
I would bet that they're not in my kernel
<acheam>
just because I dont enable any of the device specific sutff
<micr0>
konimex kudos on making gnu-less kiss easier
<micr0>
(or at the very least, gcc-less)
dilyn has joined #kisslinux
<dilyn>
what the hell is happening to my computer http://ix.io/3t2d <
<micr0>
> too many errors emitted, stopping now :bead_of_sweat:
kyxor has quit [Ping timeout: 255 seconds]
claudia has quit [Ping timeout: 258 seconds]
akira01 has joined #kisslinux
kyxor has joined #kisslinux
<micr0>
does anyone have busybox acpid working? if I ssu sv restart acpid, it times out, and goes down, but if i manually `ssu acpid -f` and `tail -f /var/log/acpid.log`, everything works fine
Uks3 has quit [Ping timeout: 272 seconds]
Uks2 has joined #kisslinux
<micr0>
also, with latest sway, the post-install says seatd is needed, but i dont see seatd in kisslinux/wayland or kisslinux/extra
<noocsharp>
kisslinux/wayland/libseat
<micr0>
thanks noocsharp
akira01 has quit [Ping timeout: 265 seconds]
kyxor has quit [Quit: kyxor]
<acheam>
micr0: i'd reccomend using acpid2 from the repo
<acheam>
much nicer config language, works better
akira01 has joined #kisslinux
<micr0>
acheam which repo? didn't show up with kiss-find
<acheam>
uhh check git log?
<acheam>
might have been removed
<acheam>
but it was in the main repo i think
<micr0>
I just migrated my suspend-on-close script to busybox acpid xD
<acheam>
from what?
<micr0>
acpid2
<acheam>
haha
<micr0>
since `kiss update` showed that acpid2 was orphaned
<acheam>
busybox acpid is painful
<acheam>
oh i see
<acheam>
so dylan must have dropped it
<acheam>
not much I can do to help you then
<micr0>
busybox acpid was just a 1 line file to get suspend working, it actually was simpler than acpid2, but way less readable
<acheam>
yeah acpid2 is easier to understand, albeit more compex