In some sense, I have adopted the onepiece@lemmy.world community. not as a mod, but I have been posting stuff from reddit and yt theories, and have seen some more upvotes than usual. hope the community grows more.
thanks for checking it out.
I have also heard that icc profiles should be independent.
I would request you to try something. use some profile generated in windows, in both windows and linux. now take some sample file (ideally non hdr) and open it in a some program available in both windows and linux, so for example, mpv. even though mpv has some differences in it's default config on these 2 systems, but try to use software encoding to prevent those differences. and ensure the output is fullscreened (ideally by a mpv cli option). now ideally same colors should be produced. if they are, then you have your problem solved, if not, well then we know same profile does not work.
Still thanks for trying. linux is currently weak in these cases, because much of these has been driven by artists who do color accurate work, who have traditionally only worked on ms and now apple stuff. so there has been very little work in fixing this stuff. also since many of these color production stuff is licensed, it requires manufacturer to test in linux while certification, which no one does. commercial interest is just not there.
maybe they can atleast serve as fine staring point. i have never edited them before, and not even sure what format are they in (i am guessing they should be something editable by hand, but just guessing)
I am sorry, but you might be correct here that linux icc profile support is not great. my generic advice would be to check out arch wiki (but seeing from your answer, you have already tries some of the things that wiki lists, you have probably already read it).
if i am not wrong, many original display manufacturers do provide icc profiles for panels, and you could find them in their spec somewhere (i think i have done this once for a older display)
not all vaccines. some vaccines (most olden ones) were just differnt similar or weaker or less potent strains, which while still infected you, you get some immunity because your body made "ammunition" against the actual threat.
then there are vaccines like mrna vaccines, where we find the mrna sequence that enocdes for specific protiens, for example, some surface protien. this protien is a part of virus, but alone the protien is pretty much useless. your body still sees that their is a foreign protein, and builds ammunition against this protien. when the actual protien with the specific virus comes, our body sees the same virus, and uses the same ammunition.
or comms as well
part 7 is steel ball run? i have to start reading jojo someday
they are indian probably (someone else said the same), and i remember these shows airing in 00s. 1,3 being indian productions, and 2 and 4 as dubs (not sure about 2, but it is likely spread, 4 was a very famous dub here)
you have 2 major package managers (ootb) on bazzite iirc - flatpak and rpm (through rpmtree). ideally - do not install anything through latter. that is the one that requires the cli. if you can not find a package on flatpak (very common if you want a cli thing, or a niche gui software, or some browsers), then try to find if it is served elsewhere. for example, as this post highlights very nicely - use distrobox. for example, use distrobox and add arch (for example), and you can get new cli stuff.
for chrome, if you have 2 versions, either you have 2 different flatpaks, or 1 from rpmtree, for that, try using rpm-ostree search chrome (or some other package name, for example chromium). you may also just want to do chro (in a terminal window) and then press tab (once or twice) to get completion options. that should help you with name of package most likely.
do a lot of magic with the kernel on Bazzite
sadly, no. there are not any special kernel parameters or compilation difference. at best, these things can bring +- 5% difference (assuming a general benchmark, instead of a special synthetic benchmark). if there was some major switch you could hit which would increase performance, most distros would just press it. if most are not doing it, then it is likely because either their is not much to gain.
for example, cachy os compiles it's programmes for x86-64 v3/v4 as opposed to v1 or v2 for most distros. their have not been many extensions to x86-64 between v2 to v3, and most performance gain you get is in specific hashing benchmarks. on average, their is not much reason. as to why not all distros do it? because any software compiled in v1 runs on v4, but v3 can not be run on v2, v4 or v3, so if all distros would start doing it, then either they would have to stop serving users v3 or earlier versions (that is practically everyone with cpu before 2020, and new v3 cpus are still being made), or they would have to serve separate v3 versions for v3 folks, v2 for v2 folks and so on. that is a lot more costly, and increases software burden. even big company like microsoft is not serving different windows version forr different x86-64 versions (though they have different things, and windows 11 requires v2 or higher afaik)(it may even be v3, but not sure).
Am I hearing that I just need to switch to Bazzite and this problem disappears??
not a nvidia user, but answer is likely - no. bazzite has no special magic nvidia drivers. their special thing is that thy prepackage closed drivers (which most distros can but not do so that they are not in legal trouble of redistributiing closed software or the ethics of distributing closed software in a foss promoting environment).
I’d expect that Bazzite and Mint would use the same Nvidia proprietary drivers without much noticeable change in performance
precisely.
there is a possibility that maybe going to a newer kernel version or vulkan or other system libraries can get you some more performance, but on average, it is not going to be more than 5% (there may be some exceptional games which gain more, where there were game/engine specific bugs which were dealt in specified period). these are also the performance differences that people say that different distros (bazzite, cache, newer ubuntu release vs older) perform better than others, but you can check many places that assuming that game was already supported well.
general reason for bad performance is - nvidia on linux is bad. (period). their drivers are bad (closed or open, for a long time open drivers could not even regulate power, so they were stuck in lowest power mode).
you do not necessarily have to buy new amd hardware. try checking out protondb database, and see if the numbers you are getting are same or similar to others with similar hardware as you or not. if you are, then that is the best that can be done. if not, find what special sauce do they have (maybe a different library version, or some flags, or environment variable)
they still sponsors lots of youtubers and marketing now is on tab grouping (available on most browsers), theming (which is weaker than vivaldi, and maybe zen too) and ai (all browsers can open a web chat app). but they market to people who are using edge or chrome by default, and to them, it looks fancy.