Announcement

Collapse
No announcement yet.

Wayland 1.0 Stable Release Is Imminent

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

  • Wayland 1.0 Stable Release Is Imminent

    Phoronix: Wayland 1.0 Stable Release Is Imminent

    Kristian Høgsberg spoke this week at XDC2012 about Wayland and Weston. Here's a short recap plus some videos that include new demos of this promising project...

    http://www.phoronix.com/vr.php?view=MTE5MDQ

  • #2
    Funny how krh is using weston for presentations and they haven't implemented a screen clone feature.




    On a serious note how much more development will wayland need after 1.0.

    Comment


    • #3
      So what's the big deal with Wayland?

      I'm honestly curious as to what this is going to bring to the Linux desktop. All I know is that it's supposed to be somehow better than X11. Can anybody define the "better" part for me and tell me why I should want to use it?

      Comment


      • #4
        If Weston compositor offers tear free desktop and low performance hit, it's a win...

        Xorg in 25 years haven't acomplished this
        Last edited by narciso; 09-21-2012, 05:34 PM.

        Comment


        • #5
          Originally posted by Larian View Post
          I'm honestly curious as to what this is going to bring to the Linux desktop. All I know is that it's supposed to be somehow better than X11. Can anybody define the "better" part for me and tell me why I should want to use it?
          You -as an end user- will get pretty much the same experience you are getting with X11. Wayland cuts what is not needed in order to have a graphic stack since all the functionality that was in X11 moved to other places.

          Some things are probabl;y more efficient but as an end user you might not feel them.

          Comment


          • #6
            Originally posted by Larian View Post
            I'm honestly curious as to what this is going to bring to the Linux desktop. All I know is that it's supposed to be somehow better than X11. Can anybody define the "better" part for me and tell me why I should want to use it?
            As an end user? Not a whole lot. X is being replaced for a lot of reasons but some of the big ones are: Break the forced backwards compatibility. X used to do a lot of things, many of those things (font rendering, memory management, and a few hundred other things) have since been moved to other, smaller, programs or into the kernel. BUT, some programs still depend on X being able to do those things so they have to stay in the server. Thats why you hear people talking about X12-- its an oppurtunity for the developers to just go "F*ck backwards compatibility" and break as much stuff as they want to modernize the X server.

            Unfortunately X is like a brand. And you have to support a lot of stupid stuff to be able to say you speak the "X protocol" so the protocol would either have to be updated (not likely as itd make BSD and Solaris no longer be X compatible) So the only option remaining is a new protocol: Wayland.

            Wayland is meant to do something X was never meant to do--- get out of the way. Let the kernel, let the toolkits, let the drivers do their thing and dont have the big bad bouncer that is X in the middle of things saying "No you cant do that because 20 years ago we hadnt thought of that"

            A nice little side affect of this will be more portable toolkits. GTK and QT and KDElibs and EFL will have to support X and Wayland at the same time because some users will be on X, some on Wayland, that means the libraries cant care about what protocol they are speaking--it has to be abstracted away. This way, if 10 years from now we decide that Wayland was great for 2012 but not so great for 2022, hopefully moving to a new backend will be easier because the toolkits ALREADY have to be fairly backend agnostic.

            Comment


            • #7
              Originally posted by Larian View Post
              I'm honestly curious as to what this is going to bring to the Linux desktop. All I know is that it's supposed to be somehow better than X11. Can anybody define the "better" part for me and tell me why I should want to use it?
              It is supposed to simplify things.

              Traditionally, X did everything, from graphics and input drivers, to drawing lines and fonts, and even network transparency. Things have changed in the meantime, and many things have moved out of X and into other places: linux kernel, window managers, GUI toolkits, Mesa, Freetype, etc.

              What Wayland is trying to do is to make a minimal display server which only offers what is needed, and interacts nicely with all these new technologies. Of course, X11 also interacts nicely with them now, but it is very large and complicated, difficult to maintain, and most of its functionality is not used on a modern desktop anymore. So the idea is to make the regular desktop run on top of kernel+mesa drivers and FreeDesktop.Org technologies, and use as light a display server as possible, while all the network transparency stuff X traditionally does is moved to an optional codepath which is only run when needed.

              It sounds quite nice in theory. The problem is that the reporting here is sensationalist, and a religious cult of Wayland has formed here on Phoronix, expecting the Kingdom of God as soon as Wayland descends from heaven, and are expecting stupid things (ridiculous speedups, better user experience). Wayland is an interesting technology, currently developed as a hobby by one guy aided by some occasional contributors, but it seems to have a lot of goodwill with many important developers in the Free Desktop ecosystem, so it is possible that in a few years, some people might replace X with it for some purposes.

              At that point, we'll have to see whether it's time to retire X, or what will happen. So far, X has adapted and triumphantly defeated every challenger, but it's showing signs of age. In the end, it will affect developers more than users.

              Comment


              • #8
                Originally posted by pingufunkybeat View Post
                Wayland is an interesting technology, currently developed as a hobby by one guy aided by some occasional contributors, but it seems to have a lot of goodwill with many important developers in the Free Desktop ecosystem, so it is possible that in a few years, some people might replace X with it for some purposes.
                You are severely mistaken: Wayland is and has been for a long time developed by Intel and Collabora among other companies commercially.Kristian Høgsberg works on Wayland for Intel for example.

                Comment


                • #9
                  Originally posted by 89c51 View Post
                  Funny how krh is using weston for presentations and they haven't implemented a screen clone feature.
                  He said that screen cloning is useless. He seems to live in an alternate reality in that everybody is fine with a black screen on the notebook…

                  Originally posted by 89c51 View Post
                  On a serious note how much more development will wayland need after 1.0.
                  Lots. Only the APIs will be stable, not the software itself.

                  Comment


                  • #10
                    Originally posted by Teho View Post
                    You are severely mistaken: Wayland is and has been for a long time developed by Intel and Collabora among other companies commercially.Kristian Høgsberg works on Wayland for Intel for example.
                    And pretty much every X developer is a Wayland developer because everyone knows how bloated the X protocol and X server is for the current style of computing

                    Comment


                    • #11
                      Originally posted by Awesomeness View Post
                      He said that screen cloning is useless. He seems to live in an alternate reality in that everybody is fine with a black screen on the notebook…
                      Apart of what he thinks the user MUST have: Cloned screen, extended screen, different workspace on different screen and maybe some more that i cant think from the top of my head.

                      All these should be also easily configurable with a well designed GUI.

                      Originally posted by Awesomeness View Post
                      Lots. Only the APIs will be stable, not the software itself.
                      I expect them to keep working on weston, and maybe have it as the one and only compositor with input from the DE guys instead of everyone doing its own thing, but the protocol itself??? Maybe they'll merge the things openismus does with the text but after that??

                      Comment


                      • #12
                        Originally posted by 89c51 View Post
                        I expect them to keep working on weston, and maybe have it as the one and only compositor with input from the DE guys instead of everyone doing its own thing, but the protocol itself??? Maybe they'll merge the things openismus does with the text but after that??
                        The developers have made it very clear: Weston is only a reference implementation. If you want to use it as THE compositor that is fine, but it is not what they are expecting or wanting people to do. Kwin is to be a Wayland compositor, Mutter is to be a Wayland compositor, Compiz is to be a Wayland compositor assuming the project isnt killed from now until Ubuntu ships Wayland by default. This way they can each do their own thing, they can each experiment in their own ways and see what works instead of just having ONE implementation that may or may not be the best way to do things.

                        Comment


                        • #13
                          Originally posted by Ericg View Post
                          The developers have made it very clear: Weston is only a reference implementation. If you want to use it as THE compositor that is fine, but it is not what they are expecting or wanting people to do. Kwin is to be a Wayland compositor, Mutter is to be a Wayland compositor, Compiz is to be a Wayland compositor assuming the project isnt killed from now until Ubuntu ships Wayland by default. This way they can each do their own thing, they can each experiment in their own ways and see what works instead of just having ONE implementation that may or may not be the best way to do things.
                          I expect the DEs to use weston with some added functionality at least in the beginning. If i remember well it suports plugins and DEs can use that to add stuff they want. That if they want to move to WL quickly. I expect ubuntu to follow that route.

                          We ll see. Nice and interesting things are coming for linux.

                          Comment


                          • #14
                            Originally posted by 89c51 View Post
                            Apart of what he thinks the user MUST have: Cloned screen, extended screen, different workspace on different screen and maybe some more that i cant think from the top of my head.
                            If I understood the talk correctly currently only cloned screens are not possible and even in that case he did not rule out that the feature will be added at some point.

                            Originally posted by 89c51 View Post
                            All these should be also easily configurable with a well designed GUI.
                            Not really a task for Wayland. GUIs are other people’s job.

                            Comment


                            • #15
                              Already got my Xorg ready for uninstallation!

                              Comment

                              Working...
                              X