[HN Gopher] PCSX2 Disables Wayland Support
___________________________________________________________________
PCSX2 Disables Wayland Support
Author : MaximilianEmel
Score : 75 points
Date : 2023-11-26 20:23 UTC (2 hours ago)
(HTM) web link (github.com)
(TXT) w3m dump (github.com)
| iamcreasy wrote:
| > KDE isn't too buggy, GNOME is a complete disaster.
|
| What did the author mean by this? Don't know much about Wayland
| other than it is window system, replacing x11.
|
| KDE Plasma wiki[1] says, "KDE Plasma 5 is the fifth and current
| generation of the graphical workspaces environment created by
| KDE....KDE Plasma 5 uses the X Window System and Wayland."
|
| [1] https://en.wikipedia.org/wiki/KDE_Plasma_5
| lights0123 wrote:
| There is no de-facto Wayland software like there is the X.org
| server. Wayland is just a protocol--KDE and GNOME (through
| Mutter) both implement it. The author is saying that KDE did so
| better than GNOME.
| djbusby wrote:
| Briefly: it's a big lift for all the different desktop
| environments (DE) to move from X11 to Wayland. Lots of moving
| parts; re-implementing and re-factoring is needed.
| londons_explore wrote:
| The vast majority of users today run a hybrid of Wayland and
| XWayland... Getting the worst of both worlds, and extra bugs
| to go on top...
| nolist_policy wrote:
| That is not true though, XWayland is better maintained than
| native Xorg. It is here to stay for backwards
| compatibility.
|
| Disclaimer: I run ChromeOS and Wayland just works.
| jacoblambda wrote:
| KDE and GNOME roll their own wayland libs for their compositor.
|
| These have varying levels of quality with their own unique
| quirks and bugs. The "gold standard" for wayland at the moment
| is wlroots which is what basically every other wayland
| compositor uses. PCSX2 apparently works perfectly fine on
| wlroots based compositors but has some quirks with KDE wayland
| and is nigh unusable with GNOME's super janky wayland
| implementation.
| DistractionRect wrote:
| Basically Wayland only supports a subset of what X11 could do.
| It's more scoped and does a good job with what it's supposed to
| do.
|
| That said, the rest of the functionality is now an exercise for
| desktop environments, and they can (and do) things differently.
| OmarAssadi wrote:
| As noted by other commenters, there are several compositors
| available for Wayland (essentially the bit that actually
| implements Wayland).
|
| Some are compositors are developed independently of any
| particular DE, like wlroots, labwc, hikari, etc, but some are
| part of a larger project, as is the case with mutter and kwin
| (GNOME and KDE, respectively). And most of the time you install
| some sort of GNOME/KDE + Wayland distro, you'll usually also
| end up with their compositors.
|
| GNOME's implementation in particular has historically caused a
| lot of drama relative to some of the others. MPV used to have a
| pretty spicy wiki section dedicated to GNOME (it still kind of
| exists, though has been toned down a fair bit and addresses
| NVidia and some other specific issues more directly [1]):
|
| - https://github.com/mpv-
| player/mpv/wiki/FAQ/ddcbe1b88a99d2568... (2020)
|
| [1]: https://github.com/mpv-
| player/mpv/wiki/FAQ/a70c96040ad4fa374...
|
| -
| seeknotfind wrote:
| What a refreshingly honest commit message. The corporate over-
| professional dribble on work commit messages is so bland by
| comparison. That being said, I'm still left wondering if Wayland
| support on one of my projects is a problem.
| jdlyga wrote:
| Are we subjecting ourselves to the sunk cost fallacy with
| Wayland? It's been "almost ready" for 6 or 7 years at this point.
| OsrsNeedsf2P wrote:
| Ignoring the exponential pace Wayland has improved for end
| users in the last 36 months (see for example Plasma/Wayland
| Showstoppers[0]), would staying on X11 not be more of a sunk
| cost fallacy? There aren't even any Wayland alternatives with
| meaningful momentum
|
| [0]
| https://community.kde.org/index.php?title=Plasma/Wayland_Sho...
| barkingcat wrote:
| It's entirely possible that _both_ suffer from sunk cost
| fallacy, and from that analysis there isn 't a "better" or
| "more" sunk cost. It's already sunk, anything you sink
| further into it is a waste.
| rcxdude wrote:
| A way forward probably looks like a more feature-oriented fork
| of wayland than a do-over. The basic protocol is fine, it's
| just there's no sense of urgency to actually add important
| features to the standard, just a long design-by-comittee
| process which bogs down progress and frustrates developers and
| users.
| Gigachad wrote:
| It's been ready ready for a long time now. It's been on by
| default in Fedora for a long time now and I've not experienced
| issues. Or at least fewer issues than with X11.
| askonomm wrote:
| Ready ready on what? Select hardware? I've thrown 4 regular
| off the shelf laptops at various DE's using Wayland and they
| all suck in various ways to the point that they're not even
| usable. Meanwhile I can throw Windows on all of them, and
| they work flawlessly. I'm quite tired of Linux guys saying
| their stuff works just fine, without mentioning how extremely
| limited is the hardware support to actually run any of those
| things, and judging from various forums, even if you have
| supported hardware, things break.
| Laaas wrote:
| This is about X vs Wayland, not Windows vs Linux.
| askonomm wrote:
| It's about the GPU used in regular laptops, I'd imagine,
| for which Windows has drivers and support, and Linux does
| not. That's my point. X11 for those laptops works more
| reliably, but even then there's immense screen tearing,
| blurry texts, broken fractional scaling, etc. I just
| don't know why I would ever use software of this quality
| when even Windows XP had less issues like this in its
| hayday, and it released 20 years ago.
| bryanlarsen wrote:
| The large majority of laptops use Intel or AMD APU's,
| both of which work fabulously. Gaming laptops using
| NVidia chips aren't regular laptops.
| Gigachad wrote:
| It's ready for me in that it works better than X11 on my
| hardware. No it's not flawless like MacOS and to some
| extent Windows, but it's certainly an improvement over what
| we had.
| LtWorf wrote:
| Works flawlessly if you've never edited a Xorg config
| file... if you did, there is no way to keep using the
| same configuration
| agildehaus wrote:
| I'm using it full-time and the only issues I experience are
| with XWayland. But even those have been going away as native
| Wayland support lands in apps and nVidia slowly improves their
| drivers.
|
| Scaling and overall performance it's night and day versus X11.
| tannhaeuser wrote:
| If you mean wayland compositors in general, that's more like 15
| years according to [1].
|
| For gnome's Mutter, it's 12 years:
|
| > _Until they sort their s*t out, which is unlikely, since
| there 's been very little progress over the last decade, just
| keep it disabled._
|
| Begging the question if Wayland itself is already antiquated
| and if its maintenance perspective is any better than X.org's.
| When lack of developer motivation to work on X.org appears to
| be the sole point of leaving it behind since new apps haven't
| been created and (Nvidia) drivers for Wayland may still be
| problematic.
|
| I can totally understand the desire to start with a clean
| architecture, but chances are we have to basically start over
| at square one, with the problem of how to attract younger devs
| now who want to invest their time into newer languages and/or
| graphics pipelines rather than maintaining Wayland.
|
| [1]: https://en.m.wikipedia.org/wiki/Wayland_(protocol)
| haunter wrote:
| Further down:
|
| >... it is disabled on our release builds, because it's nothing
| but headaches for us, because of its broken by design nature
| causing issues for users. I listed a bunch of them in the OP as
| well.
|
| >We're sick of getting blamed for bugs in wayland compositors,
| while the various committees sit around arguing with each other,
| finally decide on standard ways of doing things after half a
| decade, then GNOME ruins it all by refusing to implement it.
| MuffinFlavored wrote:
| Linux in general...
|
| Think of the efforts wasted in duplicate/overlapping
| functionality between maintenance + development of
| distributions, package managers, shells, compilers, C
| libraries, desktop environments, etc.
| idle_zealot wrote:
| Unfortunately that's not avoidable. The duplication exists
| because of disagreement regarding how things _should_ be
| done. The alternative is not "my preferred project gets all
| the resources" but "my preferred project doesn't exist and
| all the resources go to some fundamentally flawed project
| that I don't want to use". See Windows or macOS for how that
| turns out.
| MuffinFlavored wrote:
| > The duplication exists because of disagreement regarding
| how things should be done.
|
| Is there too high a level of this disagreement in open-
| source software? Should it be called out from the
| perspective of "please try to use your time more
| effectively instead of contributing to fragmentation
| because it's harmful to the ecosystem from a macro
| perspective"?
| oynqr wrote:
| Wayland lacks a crutch to work around broken multi-monitor
| handling in composit
| oynqr wrote:
| Side note: The Glider HN client makes it real easy to fat
| finger comments since the rewrite. No editing or deleting
| either. Sigh.
| hashworks wrote:
| Actions like this will just hurt Wayland in the long run. It
| needs more users, and for that more applications that support it
| by default. No one will bring X11 to the level of Wayland, but
| someone might fix the existing Wayland problems.
| Gigachad wrote:
| Not sure it will do anything to hurt Wayland. The app will just
| run in XWayland which is mostly impossible to tell between a
| native Wayland app for users.
| rowanG077 wrote:
| Xwayland is instantly recognizable as shitty due to
| fractional scaling. Which almost everyone with a modern
| laptop uses.
| Gigachad wrote:
| I might not be up to date on this, but does fractional
| scaling work at all on Linux under any setup? Last I
| checked GTK just flat out didn't support it or something.
| imiric wrote:
| X11 today works better for more users than Wayland. No user
| will want to switch to something that provides a worse
| experience.
|
| Until these issues are ironed out, it's delusional to think
| that making Wayland the default will make users happy. Keep it
| as an experimental feature, and once it provides an objectively
| better experience for everyone, make it the default.
|
| These technical discussions by folks in the trenches often miss
| the forest for the trees. Users don't care that X11 is
| difficult to support, develop and maintain. They just want a
| working system. By the looks of it from this GH issue, Wayland
| is also a pain for developers. What a sad state of affairs for
| Linux.
| hashworks wrote:
| I guess I'm just hoping for some user that will be bugged
| enough by problems in Wayland that they decide to fix them.
| The chance of that increases with user count.
|
| I understand that the bulk "forest" of Linux users don't care
| about their window manager / desktop environment, but it's a
| higher number than in other operating systems. This approach
| wouldn't work in Windows or MacOS, in Linux it might.
| dinckelman wrote:
| As long as the option to keep it turned on exists, I have no
| issues with this. It's one thing to move forward with a new
| stack, but it's a different thing to have it as default when it
| doesn't fully work
| BaculumMeumEst wrote:
| I'm not sure if it's funnier to see popular projects shunning
| wayland because it sucks or Linus looking to move kernel
| development away from an email based workflow because "it really
| isn't working anymore". There is such an insane amount of coping
| about both of those topics online.
| Laaas wrote:
| FWIW all the reasons seem like an aversion to complying with how
| Wayland works.
|
| Yes, you can't position windows absolutely, but that's because
| you're not supposed to do that.
|
| In any case I don't see how being able to position windows is
| relevant at all to PCXS2's functioning on Wayland? Just don't
| position windows if you can't?
|
| It's an incredibly dumb reason not to support the protocol. Xorg
| is mostly unmaintained AFAIK.
|
| See https://gitlab.freedesktop.org/wayland/wayland-
| protocols/-/m... for the discussion on this
| fiddlerwoaroof wrote:
| One person's unmaintained is another's stable
| asveikau wrote:
| > Yes, you can't position windows absolutely, but that's
| because you're not supposed to do that.
|
| I think about 40 years of GUI APIs disagree with this.
| ryandrake wrote:
| > Yes, you can't position windows absolutely, but that's
| because you're not supposed to do that.
|
| Why not? This seems like a pretty opinionated policy from
| something that's supposed to be a platform to enable
| applications. Why should some other developer dictate how an
| application should work? I'd expect "My way or the highway"
| from Apple, but not from a linux API.
| Arnavion wrote:
| It's more nuanced than "should be able to do it" and
| "shouldn't be able to do it". The problem is covered in the
| wayland-protocols MR that the PCSX2 PR author also linked to
| (and failed to appreciate):
| https://gitlab.freedesktop.org/wayland/wayland-
| protocols/-/m...
| kcb wrote:
| Even if X11 goes away XWayland will be around.
| stefan_ wrote:
| That thread is a fantastic reason to never run wayland.
| bryanlarsen wrote:
| Another option might be to disable only on Gnome or Nvidia since
| those seem to be the source of most of the problems.
| apatheticonion wrote:
| As a naive user who largely hasn't had any issue with it. Is
| Wayland that bad?
| bryanlarsen wrote:
| In my experience Wayland is pretty bad on Nvidia binary driver,
| has some annoyances on Gnome and works fabulous on AMD hardware
| with KDE or a wlroots based compositor.
| LtWorf wrote:
| The internet is full of people that will inform you that all
| the stuff that stops working when switching to wayland is your
| own fault and you're a liar anyway.
|
| I'm quite convinced these people just use windows and
| occasionally open linux in a vm for 10 minutes.
| TillE wrote:
| Aside from bugs and driver issues, Wayland has some unfortunate
| design limitations. For example, Dear ImGui multi-viewports don't
| work because "Wayland doesn't let application read or write
| windows positions."
|
| https://github.com/ocornut/imgui/wiki/Multi-Viewports
|
| This is a feature available on Windows, macOS, and of course X11.
| Making choices like this means desktop Linux becomes even more of
| a weird island that nobody wants to support.
| reisse wrote:
| Quoting @stenzek
|
| > But Wayland is just broken, and everyone would rather sit
| around arguing with each other instead of actually addressing the
| design flaws.
|
| > It's not the first time such a proposal has been put forward.
| Something that developers need for their applications to work
| properly on WL (particularly multi-window applications), and it
| gets vetoed. Every other OS manages this fine. But apparently
| we're in the wrong for not conforming to some warped view of how
| applications should be, despite our applications working fine on
| every other platform.
|
| This is so damn true. Wayland lacks a Linus-style BDFL saying
| that the kern... the compositor is for applications, and not the
| other way around. We won't have nice things until Wayland
| maintainers stop thinking about what their end users must or must
| not do, and start closing feature gaps with X and
| compositors/window managers on other OSes. Right now they're
| reinventing the wheel while making it square.
| NekkoDroid wrote:
| My opinion on a few of the points they have:
|
| > Stupid obsession with CSD in Gnome
|
| I get why GNOME devs don't wanna implement it (most of the times
| looks out of place for the app, or the app looks out of place to
| the system if it doesn't use the system toolkit, so it's just
| better for them to roll something that looks good for the app)
| and the XWayland implementation for SSDs are still a holdover
| that would need to be ported that they don't want to do. But then
| again to my knowledge QT has a way to make a header bar (I
| _think_ even one in GTK style) and there is libdecor that handles
| headerbars.
|
| > Inability to position windows
|
| xdg-session-management is being worked to handle restoring of
| window positions. There are some contensious extensions being
| discussed about other window placements but they are still very
| much in flux. I personally am impartital for the actual use of
| the features and if they are really that necessary.
|
| > Hacks in render-to-main because WL craps itself otherwise
|
| This sounds more like a problem with the project itself than
| wayland considering the meriad of other things able to work, but
| then again its an emulator that might be doing some very weird
| stuff so I can't say much about it.
|
| > Despite said hacks, game list still glitches after stopping
| emulation, happens more often in gnome
|
| Probably a cascading problem from the previous
|
| > NVIDIA just crashes in swap chain creation under Wayland
|
| Not exactly _too_ surprised, although my laptop with Intel /NV
| hybrid has been working mostly fine for my mostly basic usage.
|
| > Broken global menus
|
| Those are a thing anywhere other than OSX? How do they handle
| this on Windows?
| anaisbetts wrote:
| Windows doesn't have global menus, it defines an API to get the
| workspace coordinates, which allows apps to know the area
| available to the window (i.e. not covered by the Start Menu)
| anaisbetts wrote:
| GNOME as a platform (and by extension, all of their associated
| projects which unfortunately includes the Wayland protocol) has
| such a bizarre worldview - it's like they think they can treat
| the Desktop OS, with all of its user expectations and existing
| working software, and remake it as a bigger iPadOS, to no one's
| actual benefit.
|
| What do I get in return for having to give up all of these things
| that used to work? It's not faster, it doesn't do extra things I
| couldn't do before - it's _strictly_ downsides all the way down
| shmerl wrote:
| _> NVIDIA just crashes in swap chain creation under Wayland_
|
| I'd say ignore Nvidia, their Wayland support is junk (until
| nouveau+nvk catches up). It's not a Wayland problem, it's Nvidia
| problem.
|
| For dealing with CSD mess, there is libdecoration that SDL
| started using.
|
| In general - just use SDL for for DE integration, instead of
| trying to reinvent the wheel.
| evanjrowley wrote:
| Due to the headline, it has dawned on me just now that the Steam
| Deck uses X11. Been using PCSX2 successfully on it for several
| months now. I never thought about whether it was using Wayland or
| X11 until reading this.
___________________________________________________________________
(page generated 2023-11-26 23:00 UTC)