Announcement

Collapse
No announcement yet.

SDL2 Reverts Its Wayland Preference - Goes Back To X11 Default

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #41
    Originally posted by RealNC View Post
    Wayland is still not viable, even with open source drivers. Look at this nonsense:

    https://gitlab.freedesktop.org/mesa/mesa/-/issues/6249
    That's a Mesa issue, not a Wayland one.

    Originally posted by abu_shawarib View Post
    Is it me or did everyone just ignored the mentioned bug in libwayland?
    It's more like a missing feature than a bug per se, though it makes avoiding accidental disconnects of unresponsive clients tricky for Wayland compositors.

    Originally posted by billyswong View Post

    CSD supporters probably don't think one may want to move / resize windows for hanged / busy applications.
    A decent Wayland compositor should allow moving surfaces of unresponsive clients, e.g. mutter does by dragging the mouse while holding the Super key (or via the Alt-Space menu).

    Resizing a Wayland surface requires active involvement of the client, so that couldn't work with SSD either.

    Comment


    • #42
      Originally posted by Myownfriend View Post
      Not even remotely. Plus you're really not qualified to say what design decisions are or aren't asinine. You don't code. You took credit for fixing some stuff because you pointed out some bugs but other people fixed them.
      Why does it matter if he codes? Design decisions that matter if you code are those internal to a project, not to an end user. If a design decision impacts end user then nobody gives a shit if he can code or not.

      In fact that's probably the issue here, since Wayland developers (those who develop the protocol I mean) are a bunch of retards who thinks the world revolves around their computing needs and reject any sensible feature request that other users want out of their display server/protocol because they use it on X11.

      Claiming "it's not needed" makes them a bunch of lunatic fringe fucks.

      Comment


      • #43
        Originally posted by mangeek View Post

        NVIDIA's Linux strategy has failed so spectacularly that it's causing collateral damage to the rest of the open-source ecosystem. I think NVIDIA really needs to cooperate with and support the Nouveau team, or get a fully-functional open-source Vulkan driver out there. We need something, they can still keep their proprietary driver for corporate support, GRID VDI, and other big money projects. I just want to be able to use their products in a reasonable way that doesn't make me jump through hoops or bend my system into some alternate universe.
        NVIDIA must be forced to embrace a dual support like the AMD one.

        Comment


        • #44
          Originally posted by Volta View Post

          It turns out mediocre nvidia driver is still POS and software like Steam were not updated to Wayland, yet. You could at least read the article.
          You could, but then you'd stumble upon "libwayland event overflow, libdecor not handling plugin load failures", which would interfere with ranting.

          Comment


          • #45
            Originally posted by Myownfriend View Post
            That person who created that issue even states that they only noticed the input lag in Retroarch and some in the discussion believe it's a Retroarch issue.
            The issue is that retroarch tries to reduce latency. Other games or game-like just use whatever they're given, in this case 4 swapchain images. In retroarch you can configure the swapchain size. But it doesn't work in this case. Even if you set it to 2, you get 4.

            Comment


            • #46
              Originally posted by Vermilion View Post
              Doesn't the linked page literally say: It works with amdvlk but not radv, with Wayland and on the same hardware? How is that a Wayland problem?
              It's a wayland problem in the sense that if you use wayland, you'll get that issue. The blame game is not really relevant. If you get more input lag in wayland, then who's fault it is doesn't fix the issue that you get more lag in wayland.

              Comment


              • #47
                Originally posted by -MacNuke- View Post
                Fun to see that the result of the tests linked in the issue shows that a modern, composited desktop on Xorg is way slower than a Wayland desktop.
                Yep. But you disable compositing on Xorg because of this for games. (You can set your window manager to automatically do this when you switch games to fullscreen.) You can't disable compositing on wayland though.

                Comment


                • #48
                  Originally posted by MrCooper View Post
                  That's a Mesa issue, not a Wayland one.
                  The end result is the same. Doesn't matter who's fault it is, really. Mesa has some issues, nvidia's blob has even more issues, so using wayland has issues and hurts adoption.

                  Comment


                  • #49
                    Originally posted by RealNC View Post
                    Yep. But you disable compositing on Xorg because of this for games. (You can set your window manager to automatically do this when you switch games to fullscreen.) You can't disable compositing on wayland though.
                    But even with compositing Wayland is as fast, apart from the single issue with radv that does not happen with amdvlk.

                    Comment


                    • #50
                      Originally posted by Volta View Post

                      It turns out mediocre nvidia driver is still POS and software like Steam were not updated to Wayland, yet. You could at least read the article.
                      Or you could enable emoticons to see when someone is being facetious.

                      Comment

                      Working...
                      X