Announcement

Collapse
No announcement yet.

XMir-Based Xubuntu Images Now Available

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

  • #51
    Originally posted by dh04000 View Post
    I think the XMir project has many benefits for users that everyone seems to not notice.....

    XWayland: Only applications can run. User must be using DE running natively under Wayland with applications for all non-wayland-native-DE (Xorg apps and pre-wayland gnome2, kde4, E17 apps, ect) can only be run as single applications.
    The fact nobody tried to use a DE on XWayland doesn't mean you can't. But since you can't run Wayland apps inside this layer, is saner to just use X.org.

    IE: if your DE of choice hasn't been converted to Wayland native yet, then too bad.
    Nobody forces you to switch.

    XMir: DE and applications can run (later application only mode will be supported). User can be running DE running natively under Mir or XMir for all DE's and applications.
    No Mir applications can run inside XMir, so, either you use only X.org apps, which would be better to just run natively on X.org, or you have the same scenario than Wayland. YAY.

    IE: Users have more choices. Use a Mir native DE or not, your choice.
    IE: Users have X.org.

    I see this as a HUGE advantage and great for users to assist with the post-Xorg times transitions that Wayland doesn't support. Its USE NATIVE or too bad with Wayland. There's a choice with Mir/XMir.
    The choice with XMir doesn't have the choice to use Mir applications, so you are better off with X.org. The idea of using either Wayland or Mir is using either Wayland or Mir, if your idea is to keep on X.org, well, then just keep on X.org and don't overcomplicate things.

    Comment


    • #52
      performance in Xmir should end up being almost identical to performance with just X
      No it won't and it never will. It's a very simple equation:

      DE > Xorg > Hardware
      vs.
      DE > Xorg > Xmir > Hardware

      There's no way you'll ever get less by stacking in more. Just not going to happen. All the composition still has to happen in Xorg, all the faults of X will still be there. Now it'd be a different issue if you were running a native DE on Mir, and then running legacy apps on top of Xmir. But running a whole DE under it will never give you good performance.

      Comment


      • #53
        Originally posted by mendieta View Post
        Boys and girls, what's with the noise. It seems like Xubuntu, an Ubuntu derivative, is going to evaluate whether the solution proposed by the parent distribution (Ubuntu) is technically viable at this point. This seems like a civil way to go about things, and more inline with Linus' approach of using what works, than with being opinionated or politically driven.

        Having said that, I still think Mir is a bad idea, but then again, that's an opinion, not a technical fact.
        XMir being an unneeded layer is a technical fact, too. The fact it doesn't provide any Mir features, is technical, too.

        Originally posted by dh04000 View Post
        Well, it sounds like it would be good for everyone if XMir and XWayland could be cross pollinated and yield both capable or DE and application only modes.

        I'm sure the DE mode patches will be ignored entirely without comment.
        Maybe it shouldn't be without a comment, but there's really no point on merging them. Come to think about it, all the benefits named until now are caused because an Intel specific feature produces tearing when not using a GL compositor, and it seems to be arguable, according to one user around here. In any other case: advantages over X.org == 0. Running a DE over XWhatever makes you unable to run native apps.

        Comment


        • #54
          Originally posted by bwat47 View Post
          It is not pointless, as I mentioned before it manages to solve the long running issue of video-tearing in XFCE.
          There are way better ways of solving it than by introducing an expensive compatibility layer.

          Comment


          • #55
            This whole exercise is part of "doing". It just so happens they decided to let this testing stuff go free in the wild instead of stuck in a hole someplace. I'm not sure why there's complaining. As far as I can tell that canonical with xmir is way more into "doing" than wayland is. It's important to get stuff into people's hands and not just sit on it. "Release early, release often".

            Comment


            • #56
              Originally posted by mrugiero View Post
              The fact nobody tried to use a DE on XWayland doesn't mean you can't. But since you can't run Wayland apps inside this layer, is saner to just use X.org.
              Nobody forces you to switch.
              No Mir applications can run inside XMir, so, either you use only X.org apps, which would be better to just run natively on X.org, or you have the same scenario than Wayland. YAY.
              IE: Users have X.org.
              The choice with XMir doesn't have the choice to use Mir applications, so you are better off with X.org. The idea of using either Wayland or Mir is using either Wayland or Mir, if your idea is to keep on X.org, well, then just keep on X.org and don't overcomplicate things.

              I'm thinking long term. When Xorg is no longer developed but still functional in XMir/XWayland(with full DE patch). There are still people running KDE3. My money is that after Xorg is no longer developed, there will still be people running KDE4 and other pre-Wayland/Mir DE's.

              Also, don't cut up my arguments into weak, easy to attack portions. Its not nice. I wrote it together, it only makes sense together.

              Comment


              • #57
                Originally posted by dh04000 View Post
                I'm thinking long term. When Xorg is no longer developed but still functional in XMir/XWayland(with full DE patch). There are still people running KDE3. My money is that after Xorg is no longer developed, there will still be people running KDE4 and other pre-Wayland/Mir DE's.
                Even in the long term X.org will be maintained. XMir will have to be updated to match each new X.org release regardless. And if in the super long term there are no new X.org releases, there are no new XMir releases either (except to match whatever changes there are in Mir). At that point switching to it might be an OK idea, as the X.org interface to the kernel might break at that point, while Mir would be updated... but at that time Xfce will already run natively on Wayland and/or Mir.

                Comment


                • #58
                  Originally posted by GreatEmerald View Post
                  Even in the long term X.org will be maintained. XMir will have to be updated to match each new X.org release regardless. And if in the super long term there are no new X.org releases, there are no new XMir releases either (except to match whatever changes there are in Mir). At that point switching to it might be an OK idea, as the X.org interface to the kernel might break at that point, while Mir would be updated... but at that time Xfce will already run natively on Wayland and/or Mir.
                  hmm i was told Xorg will become Xwayland so that will Kill Xmir? as the 2 will not have the same API's at all down the road?

                  Comment


                  • #59
                    Originally posted by dh04000 View Post
                    I'm thinking long term. When Xorg is no longer developed but still functional in XMir/XWayland(with full DE patch). There are still people running KDE3. My money is that after Xorg is no longer developed, there will still be people running KDE4 and other pre-Wayland/Mir DE's.
                    The thing is, if you run unsupported software, well, you run unsupported software. Nobody will stop you from using the older versions, along with an older X.org. Since running the DE over XMir or XWayland blinds you from them being there, and disables you from taking advantage of Wayland/Mir features, you are better off just using X.org. The only exception might be drivers support, which you can still get with X.org and an older DE now, and you will probably lose at some point when Mir or Wayland become the main platform. And I'm not even sure about that, since at least nVidia will probably keep X support for FreeBSD (since they support them now).
                    I do believe, too, that some people will stick to older DEs, but I believe they'll stick to X.org, too.

                    Also, don't cut up my arguments into weak, easy to attack portions. Its not nice. I wrote it together, it only makes sense together.
                    Arguments should be valid by themselves, to make a conclusion valid. I could attack them individually without cutting them into portions, but that would just make it harder to read.

                    Comment


                    • #60
                      Originally posted by LinuxGamer View Post
                      hmm i was told Xorg will become Xwayland so that will Kill Xmir? as the 2 will not have the same API's at all down the road?
                      No, X.org is and will continue to be its own thing. There's a reason why Wayland is not X12. X11 is still required by certain professional software, so X.org is not going anywhere. The users of said professional software also require security updates at the very least, so X.org will continue to be maintained for quite a while. XWayland and XMir are just specialised, integrated versions of X.org.

                      Comment

                      Working...
                      X