Xlibre is better than both Xorg and Wayland. Change my mind.
>>108100398>doesn't work on nvidiai am filtered
>>108100419But we have AI now?? Can't it just make flawless drivers for any hardware on any platform??
>>108100398It's funny that right now this is only true because retards at Xorg literally want to destroy their own software.All that needs to happen for mass Xlibre adoption now is better Nvidia support and more OSs giving in.
>>108100457Yeah, and the people at GNOME and KDE force Gayland upon us, even though it isn't even ready yet.
>>108100419>Doesn't work on nvidia.Where did you get that from?It works with all versions of the proprietary drivers since 340, and newer cards also have full hw accel with the modesetting driver. This better nvidia support that what Xorg has.
>>108100431GPU manufacturers refuse to document the hardware.
>>108100603>muh intellectual propertyMost of it has been reverse engineered, the performance is just garbage. Might as well release the official docs.
>>108100500Not for me. GTX 1050 Ti here, 580 drivers installed, Cachyos. I get a black screen with just my pointer on after logging in.
>>108100603Never understood this. An almost infinite amount of trannies on typewriters are ready to improve your product for free and nvidia doesn't take the offer.
Xlibre makes the freedesktop people so mad because it highlights just how little work they actually do.Years, decades of bitching from them about how ohhhhhh nooooo xorg sucks, xorg is too big to work on, xorg is too oldThen some antimask heathen forks it, and within 8-12 months, we have legions of bugs fixed, new features added, sane defaults turned on (TearFree was always there to be turned on in Xorg, but we were all gaslit to believe it wasn't even possible!), and the features that Wayland compositors have (HDR, HiDPI, VRR) are all either done or being worked on as I type this.Literally it just shows people why freedesktop in its entirety should be thrown in the trash. Fucking Wayland itself is 17 years old now. These fucking people should be blacklisted from OSS for life for destroying the linux desktop.
>>108100676Totally agree. The same guys that force Gayland upon us also force Basedstemd, and keep saying that it is hard to support other init systems or that you can just not use the other parts of Basedstemd, which isnt even really true. Fucking hate the people at freedesktop, GNOME, Basedstemd, and Gayland.
>>108100717Why the fuck did it turn Systemd to Basedstemd? lol
>>108100657Can you open an issue on the Xlibre github? What version of Xlibre are you using? What ddx driver are you using (Nvidia or modesetting)?
>>108100657You need to add the line Option "IgnoreABI" "1" to your X configuration file to ensure compatibility with the NVIDIA drivers.
>>108100398still stuck outputting to the lowest monitor refresh rate on a multi monitor setup. if that ever gets fixed I'd move immediately.
>>108100676it's literally a conspiracy but thankfully enough people called them out the whole time and it looks like they've failed their main objectives at this point.
wait so is xlibre actually being developed? i was expecting it to be another political meme fork that just pulls upstream patches instead of doing anything
>>108100939We have been using it for 7 months now.
>>108100947that's not what i asked
>>108100953Why would we use something that doesn't have any advantages over the predecessor? Shiny thing syndrome?
>>108100939it was being intentionally ran into the ground by the illuminati and one of the people who were actually trying to maintain it forked it. It's actually interesting you should look into it or some other anon that knows all the details should post about it.>>1081009476/7 months
>>108100939>fork that just pulls upstream patchesThe thing is, upstream xorg doesn't do shit nowaday.
>>108100963/g/ is contrarian and runs meme forks just to show off to other autists, like how people use redot instead of godot despite being the same thing>>108100964i invite someone to redpill me since it would be nice to know the finished features that are missing in xorg
>>108100398Does it have real multi-monitor support capable of supporting mixed refresh rate/mixed DPI displays? Does it support adaptive sync without having to shut off all your other monitors?
>>108100939Pretty actively, yeah. I can't comment on a lot of things but I know they made it so tearing is optional (you can turn it on or off depending on your needs), they did "Xnamespace" to do some security improvements, and there was a recent spat where they announced starting HDR work and everyone in the usual places had a fit of confusion.
On the security front, have these guys reached out to the Xenocara people? That could be very fruitful.
horny
you should study architecture of both to say its worse or better or the same level. i browsed both and wayland design intent was a simplification. is rendering simplification better? sure, it is, though is it handled by the kernel like in windows? nope, its also some kind of client-server-kernel triplet, so im really frustrated about this tinkertroonerism. rendering is major but not the single thing, inputs is 50% of importance of the whole. inputs 50%, outputs 50%. display rendering, say 30% of importance.imo, NIX improvement should get back to basics - a new terminal project, similar to kmscon, then grow to GUIs.
>>108101197Did the improve the multiple screen support? I basically have the worst case for it, different screens with different resolutions and scaling, working at different refresh rates with different vcards (laptop with discrete vcard). I can make them match the refresh rate but the rest of the factors made it pretty awkward if I recall correctly.
>>108100829>>108101125>>108101932>multi monitornot a thing yet, no
>>108100728New here I take it? The word s o y gets filtered to onions if alone and to based if part of a word.
>>108101932Use xrandr to set a per-monitor scale. For example:xrandr --output lowDPImonitor --scale 2This has worked for decades.
xrandr --output lowDPImonitor --scale 2
>>108101244Probably not, but Xlibre is still better than Xorg in almost all ways. I dont know though.
>>108101508???
>>108104087That xlibre dev makes me honry.
>>108100398>no hdr supportuseless, only thing that'd make me switch from X.
>>108100666Screens can be used for mind control, anon. The big government chip maker probably has some dark shit in their firmware.
>>108105000>>108100666These digits are a problem.
>>108101932>scalingDPI scaling is unrelated to the display server. Its done by applications (unless you want blurry upscaled shit). Xlibre added a per monitor DPI property that applications can use to scale. Its up to gtk/qt to use that (or somebody fork qtk/qt and use that). Qt has always supported per monitor fractional scaling on x11, but it was done with an environment variable to set the scaling values. Its only relatively recently that gtk got true fractional scaling support (including for wayland), so until recently gnome wayland used to downscale the window texture making gtk applications blurry.
>>108104311Its being worked on. You can check out the code right now and try hdr https://github.com/orgs/X11Libre/discussions/251#discussioncomment-15585471
>>108105570Given that QT already does it via environment variables I imagine it should be fairly trivial to patch it to read a DPI property. It's just a no-op if it doesn't exist so it wouldn't even effect xorg users.
better how?>inb4 tearfreedont care for it, i use uncomposited x11 for its low latency
>>108105957then turn tearfree offYou can do that in Xorg(maybe) and in Xlibre.Wayland says lol fuck you, we know better than you
>>108105957uncomposited x11 does not work well on modern gpus, due to profound brain damage in the design of the scan out hardware.
>>108106567proof?
>>108105917Yep, it's not that hard. The only part that is a pain is dealing with Qt bullshit. Forking that huge fucking shit and then building it with their custom tools. Then doing that for multiple Qt versions.
>>108106567what do you mean?
>>108105957true, kde does let you block compositing on fullscreen but thats about it, i still have a seperate x11 session for games so i can use evdev which i find smoother than libinput>>108106567i cant find the gitlab issue but is this what you're talking about? i have linux-tkg compiled with the patch applied
>>108100829Run a separate xserver per screen.
>>108100603What needs documents for? Have AI do it.
>>108100939There are no upstream patches, that's the whole point.
>>108105570> DPI is unrelated> DPI was addedGPL-tards cannot even decide on and describe a simple thing.. the state of
>>108100398Why is this shilled so much lately? Very suspicious.
>>108100398Maybe, but X11 still works just fine on my machine.
>>108100717Call it cistemd and then the trannies will hate it and switch to openrc
>>108108333Yes, clearly the Jews don't want you to run Wayland
>>108100398>Xlibrethats cool, but phoenix is a better idea
>>108106785The hardware which reads the front buffer from VRAM and sends it to the video port has extremely aggressive buffering. The very instant the frame starts a high bandwidth read commences, copying frame data into a large fifo buffer, on recent amd gpus this buffer is 1536kb in size. The buffer fills up basically instantly, this breaks vsync, because since it happens before x11 can finish drawing, resulting in a tear which is like 2 inches from the top of the screen on my monitor. Older gpus had much less aggressive buffering, allowing time to actually do drawing during the vblank. Modern gpus are intended to work only with flip, and no consideration has been made for drawing into the front buffer.
>>108108743Yes, but right now it is more of an idea rather than an actual project. Right now it is best to use Xlibre, but maybe in 2 years Phoenix will be completed and it will be worth switching to. I hope that atleast Xorg apps will work on it properly, unlike on Wayland.
>>108100398>XLibre is better than what it started as>XLibre is better than something designed to be broken
>>108108802Source? Can't find anything.
>>108109983You won't find anything. Its driven my utterly insane that I am basically the only person on this entire planet who have seemed to notice the blaring issue. And the problem has been around for ages, Nvidia also has the problem, and it goes back to at least the 900 series. Back in like 2015 I got a GTX 960, and it was profoundly broken when running on XP, because its scan out hardware is fundamentally allergic to front buffer rendering.
>>108109050>Right now it is best to use XlibreXorg just works, I'm in no rush to use a fork from someone that barely knows C and can't be trusted to keep his project safe. I swear the XLibre shilling is getting to be as bad as the Wayland one.
>>108110124>I swear the XLibre shilling is getting to be as bad as the Wayland one.Ah yes, finally, X11 has aquired Wayland's main and most prominent feature: infinite restless shills
>>108110107why haven't you opened a bug report?
>>108110989Its not even a bug, its working exactly as designed.
