Spent some time with a friend today, helped them
draft a plan for getting mental health issues
addressed since their mom isn't doing anything to
help. Hopefully it works out.

Still dealing with GPU-related crashes. It's getting
really annoying now. Still trying to find a fix of
some sort. It's been 3 days in a row this time when
normally it's once or twice a week. Apparently this
is an AMD GPU thing that just happens and is still
trying to be pinpointed. Nothing has worked yet.
Kernel downgrades, driver changes, firmware downgrades,
changing kernel parameters, you name it. Just stuck with
our screen freezing and going to black after a while.
It's frustrating. May need to get into using qutebrowser
as our main web browser because Firefox sets it off more
often than most things (though it's not Firefox-specific).
Been using qutebrowser a bit and it's alright, just miss
bookmarks. The Vim bindings are nice though.
Ugh, just tired of this issue. It's been months with no
fixes we could find despite a lot of research and
experimentation. It's definitely a kernel issue from
what we've found, which means we're waiting for a fix.
But it's near-impossible to reproduce intentionally.

Sway is set up the way we like it now at least. Looks
good.

Update: Trying yet another fix that has worked for some
people. Turning off GPU DPM, will see if that does
anything. Some people had success with it and one guy
thinks it's a power management problem, so we'll see if
this does anything. Going to use normally and see if a
crash happens since that's really the only way to test it.
It's weirdly sporadic.

Hey, another plus side to this. We're getting good at
finding obscure forum threads and messing with kernel
parameters. Maybe we'll learn something.
Also interesting: the GPU model itself is definitely
relevant. Most people with this same issue (us included)
have an AMD Radeon Vega and/or Ryzen CPU. Have yet to see
an NVIDIA user have this issue.

Update: VICTORY!!! It still froze for a second, but IT
RECOVERED THIS TIME! The active window was visually
funky but the others are fine, and closing it solved
the issue there. But the key is that we're still using
the computer without a hard reboot! I'll take it, dangit!
That confirms it's a power management issue, too. It was
a few lines added to kernel configs. For the record,
yes, the solution was in the Arch wiki. :')
The lines in question for anyone else having this issue
were:
amdgpu.vm_update_mode=3
radeon.dpm=0
radeon.cik_support=0
radeon.si_support=0
amdgpu.cik_support=1
amdgpu.si_support=1

That's what we added this time and we're still typing
right after a crash happened! I don't mind visual
weirdness as long as the computer is usable; at least
we can troubleshoot that without a restart.

But holy shit guys, this is exciting. We've been dealing
with this interupting everything for months. It forced us
to do hard restarts at least weekly because it would just
crash to black with an unresponsive caps lock key and no
tty access, all without any warning. It was annoying as hell
and we've finally made progress after trying fix after fix
from obscure forum threads. It's about time!

Time to go add this to the computer documentation notebook
in case we ever need to do this again.