00:21 brotofthemandel: defining performance simulated hardware inside itself for itself can be performed on ASIC too when some OS runs on it already. For an example 486 was in-order pipeline still widely used or deployed in military equipment from earlier releases of these. They do not renew the tech inside, they renew only the sw parts. So for instance, I have access to MIL -XXX specification of the HP
00:21 brotofthemandel: tablet, produced with similarly weak by default sw's atom CPU. That would just need to run proper sw, they do not renew it's internal hardware or refurbish those etc. However we'll skip over those parts for now, i have access to both atom and low-end ARM/NVIDIA hw, skip cause the hypothetical software improvements i have not tested yet to be one to one matching in outputs, i know at
00:21 brotofthemandel: minimal FS can be done to improve small-int compression, i am testing quite soon however - my hypothesis that quantum computer lives inside every hardware. Personally i am not demanding ultra performance for my daily usage at all, nor nothing about security but yes safety is crucial. And that is none the less Linux which does such stuff. No security, no performance, but safety first
00:21 brotofthemandel: from the firmwares, so it uses a lot of energy cause of human safety, and designs vulnerabilities to be able to put criminals behind the bars when there is a need. Shortly this is how i understand it's operation.
00:22 brotofthemandel: And i am entirely fine with Linux, but such systems can not be defense ones in say like full scale war.
00:23 orowith2os[d]: mods, send this spammer to the shadow realm
00:23 orowith2os[d]: :wires:
00:34 redsheep[d]: `my hypothesis that quantum computer lives inside every hardware`
00:34 redsheep[d]: Might be the funniest thing I have ever read from these
00:56 grudiosspec: orowith2os[d]: it's very clear that you are in zoopark for the rest of your life soon, and i have to do nothing to get you there, all you did was to yourself like it happened for the others doing same things as you. redsheep[d] i am pretty certain that all except IO write outs would work inside quantum model of modern computation, where the intrinsics and hashes are composed with low
00:56 grudiosspec: uncertainty/entropy but hundred percent probability. But all the testing to there to confirm that happens this year for me and there is a lot of work to perform, but as you see i am doing this under Linux and i have a huge tail of criminals at me, so it's somewhat unfortunate if not or so that everyone gets the results of my labs.
00:56 HdkR: 🎉
00:56 orowith2os[d]: *insert spinning rainbow cockroach here*
02:28 tramadolneeded: those are incredible ways that i succeeded preliminary under. I had to practice a lot for this, cause you were not exactly encouraging and soul sweet to me, you did everything for me to suffer under immaturity and fail in testing, but i succeeded in a long run made a lot of mistakes, than like settled in. So the data has similar algorithm but in the first phase where the bounds are
02:28 tramadolneeded: determined in that phase it seems like it's almost in reverse to the transition values in compute logic, so the numbers and their arithmetic that sicence has offered do work under all scenarios. I had already arrived. It's multiphase system, it puts the numbers into hash after calculating their bounds, what we know is the indexes and count of those, to start pulling some indexes and
02:28 tramadolneeded: those we do not pull, we mark as primes/even or odd, and as we know the format and count, then with making a general magic value a global hash that is common for any memory locations content transitions the double packed remainder out, and hence the insane compression rates become the reality on any hardware. So old hw is techinically also capable one to play photorealistic video games
02:28 tramadolneeded: or 3d games. I am not very poor my own anymore, i handled my expenses, and i am no longer even in debts soon. And i am also entirely functional though i see things differently.
02:29 HdkR: 🎉
02:55 rabbitcarleone: Ryan Houdek is one of those rare debiliated people who achieved nothing in life, but harassing women who despise such, and calling me a toxic monster after failing to kill me lol, so he hacked his hair off without any real results and is googling for polonium hair transplants and blaming others for the bad luck. The project is called a tegra/nvidia hair transplant with polonium
02:55 rabbitcarleone: nutrition. :D
02:56 HdkR: 🎉
03:03 HdkR: Speedrunning some classic misdirection. I can handle hate thrown my way, better for me to consume it than random users.
04:06 redsheep[d]: https://xkcd.com/965/
04:08 soreau: HdkR: wow, you're popular ;)
04:18 HdkR: Very magical
07:42 mangodev[d]: https://cdn.discordapp.com/attachments/1034184951790305330/1358346047151280218/image.png?ex=67f381d6&is=67f23056&hm=7d41c1d8289d9158697758a8b608192f812a5d69f340407b4b04fcc0677257d8&
07:42 mangodev[d]: i love how nvidia card detection kind of screws over nvk in some ways
07:42 mangodev[d]: here's sodium (a minecraft rendering optimization mod) disabling certain optimizations because it's assuming i'm using the nvidia drivers because a nvidia card was detected :P
07:42 mangodev[d]: (iirc gl threaded optimization works fine through nvk/mesa, correct me if i'm wrong)
07:47 tiredchiku[d]: mangodev[d]: different env vars for gl thread opts on nvprop compared to mesa (if any exist)
07:47 tiredchiku[d]: so it probably does nothing on nvk :p
08:00 mangodev[d]: tiredchiku[d]: didn't you bring up that this MR would perhaps help with the lag that comes with screenshotting tools?
08:10 tiredchiku[d]: I did not
08:10 tiredchiku[d]: but it doesn't hurt to try
09:34 kayliemoony[d]: mangodev[d]: you could almost definitely file an issue on <https://github.com/CaffeineMC/sodium> about this!
09:34 kayliemoony[d]: they care a lot about driver and card compat
09:38 kayliemoony[d]: as long as nvk identifies itself somewhere they can likely fix it
09:40 kayliemoony[d]: of note:
09:40 kayliemoony[d]: - NVIDIA_THREADED_OPTIMIZATIONS_BROKEN only has any effect on windows it seems
09:40 kayliemoony[d]: - NO_ERROR_CONTEXT_UNSUPPORTED is to do with wayland, not NVIDIA
09:40 kayliemoony[d]: so atm that flag combination doesn't actually do anything nvk related
09:44 marysaka[d]: kayliemoony[d]: For NVIDIA_THREADED_OPTIMIZATIONS_BROKEN, no the code actually check for drivers that cannot reliably disable thread optimization and they set `__GL_THREADED_OPTIMIZATIONS` to 0 on linux
09:44 kayliemoony[d]: okay, must've missed it
09:45 kayliemoony[d]: (i did a quick code dive for both, so, oops)
09:45 marysaka[d]: yeah I got lost a bit there, the code that check for != LINUX has to do with a popup to upgrade drivers ect
12:32 asdqueerfromeu[d]: mangodev[d]: They could check the OpenGL renderer string for false-positives I guess
15:27 calico: I finally found out why auto reclocking didn't work on my GTX 1650 ... that's cause I'm still on kernel 6.6.
15:27 calico: kek
15:30 orowith2os[d]: "your gpu doesn't reclock because you're missing GSP. My GPU doesn't reclock because I don't have GSP. We are not the same"
15:53 calico: ?
15:59 pavlo_kozlenko[d]: orowith2os[d]: haha
15:59 pavlo_kozlenko[d]: It's both funny and sad
16:06 JB0xc390cc: Hell,
16:06 JB0xc390cc: oh wawnted just a line break
16:21 JB0xc390cc: So, Hello! I have a GT710 video card and currently can't get it working with nouveau+xorg. Maybe like 3-4 years ago i did it by compiling the xf86 from source, because the xf86 package from arch repositories didn't work(No screens found in xorg log) (on the non default branch?, or with other config options? I don't remeber). Also is this video card supported by nouveau? I think so because the GK208 chipset is listed on the webpage (however exactly
16:21 JB0xc390cc: GT 710 not). If this video card supported by nouveau what should be the way to get it working.
16:23 DodoGTA: JB0xc390cc: Have you tried running Xorg without the xf86-video-nouveau package installed?
16:24 magic_rb[d]: That one shouldnt be used afaik, everything goes through modesetting these days
16:24 pavlo_kozlenko[d]: use xf86-video-modesettings
16:54 JB0xc390cc: oh it did work thx
21:26 airlied[d]: dang it 570 appears to have moved the push buf location inside their vk command buffer
22:53 airlied[d]: hmm got a shader dump, LDCU.TEXUNPACK hmm
22:55 airlied[d]: bit 80
23:30 airlied[d]: should be fun to route through things