asimovc: we're not going out looking for bugs in all the software. usually it's the responsibility of the software upstream to make a new release to fix a bug, and we'll package that. otherwise, it takes someone like you to tell us there's a bug needing a patch, otherwise we won't notice
<phoebos>
midfavila: dilyn stopped BDFLing when dylan returned, and didn't then start again. so we've had a few lead maintainers since then
<phoebos>
what was the problem with curl?
...
asimovc
joined
<asimovc>
hi guys
<asimovc>
can someone help me to use xdg portal?
asimovc
quit (Read error: Connection reset by peer)
<midfavila>
hrmmmmmmm
<midfavila>
might finally have a lead on making a real disk image now...
<midfavila>
crazy idea but
<midfavila>
to test things i might try to graft alpine onto the uconsole
<midfavila>
honestly, working on a KISS port to the uconsole feels like programming with athena for the first time all over again...
<midfavila>
not enough docs and what docs there are are scattered all over the place, out of date, and poorly written
<midfavila>
neat project to maybe bring up on a CV tho
* midfavila
thoinks
phinxy
quit (Ping timeout: 244 seconds)
zlg
joined
<midfavila>
man i have no idea why arch's PKGBUILD scripts are so insane
<midfavila>
or, at least, this one guy's for the r01 kernel is
<midfavila>
not like, insane because it's a complex job, just insane that it's written so verbosely. three layers of internal functions just to compose a command that boils down to make -jX CROSS_COMPILE=target ARCH=arch
<midfavila>
theres an entire function that does nothing but prepend its first arg to PATH. like
* midfavila
squints
<midfavila>
ngl tho if i can smash upstream and this guy's notes together and get a process that works, i might be able to refine this into a reliable way to create community images for the r01...
<midfavila>
as is, there's only the factory image, which is an ancient ubuntu 21.04
ivyfri12_
quit (Ping timeout: 260 seconds)
...
<midfavila>
okay there, i think i have the kernel done...
<midfavila>
fwah
<midfavila>
i'm giving up for now
<midfavila>
i don't have enough experience and the state of the software and docs is too chaotic for me to figure out right now
<midfavila>
maybe once it runs a mainline kernel and uboot
* midfavila
sighs
<midfavila>
it's frustrating that people only put effort into the cm4 module
<midfavila>
like yeah i get it being able to carry a cm4 in your pocket is awesome
<midfavila>
but how many debians do you need to port to it before you say 'Maybe it's time to do something else'?
<midfavila>
the worst part is that so many people i've seen with a uconsole fall into either the bing-bing-wahoo tech consoomer camp, or the what's-the-usecase camp. like just go away >:C
<midfavila>
WAIT I JUST GOT SOMETHING
<midfavila>
it didn't *finish* the boot process - i don't *THINK*
<midfavila>
but it *started* and finishing initialising the display and keyboard-
<midfavila>
okay enough blogposting
ivyfri12
joined
...
ivyfri12
quit (Ping timeout: 276 seconds)
fultilt
quit (Quit: Leaving)
ivyfri12
joined
...
ivyfri12
quit (Ping timeout: 252 seconds)
phinxy
joined
<phinxy>
When starting sway with `seatd-launch sway` the message output from both programs gets mixed up together and I can not read the full sway error messages.
<phinxy>
setting seatd loglevel to silent fixed it. EGL fails to initialize. Hmm i915 kernel module is builtin to kernel. mesa has crocus gallium driver, everything looks fine in mesa config output. chipset: HD4000 (Ivy Bridge).
<phinxy>
I noticed fully PREEMPT kernel wasnt possible with i915 module, so I chose dynamic "low latency desktop" preempt kernel.