11:29 <n3ph> I have news regarding my difficulties starting up steam. After chaos_princess was able to reproduce the threading bug in FEX and mentioned race conditions due to timing differences between m1 and m2, I got unhinged. Especially, because from the logs I saw steamwebhelper starting right after the dialog popping up complaining about it being unresponsive. So I tried a couple of things

11:29 <n3ph> back and forth again and it turns out, while the system is on heavy load, I can reliably start steam. 🤯

11:32 <n3ph> I am currently emerging firefox on notmpfs to do so. I am still trying to figure out if this is due to heavy IOload or concurrent processes handled by the scheduler. \

11:48 <n3ph> It seems to be IO related. I can repro the startup fix using something like `while true; do dd if=/dev/urandom of=test bs=1M count=1024; done`. And it's no difference if I write to dick or a tmpfs.

11:48 <n3ph> *disk - LOL

11:51 <n3ph> This is the case for reduced compute load by using /dev/zero instead of /dev/urandom as well - on both, tmpfs and disk.

11:55 <n3ph> I probably need to bisect muvm, as all of this seems to be related to 0.1.4. But for now RL demands attention.