Announcement

Collapse
No announcement yet.

The New X.Org Server Driver API Is Coming

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

  • The New X.Org Server Driver API Is Coming

    Phoronix: The New X.Org Server Driver API Is Coming

    The new driver API for the X.Org Server that would finally allow for the X.Org stack to better compete with modern desktop drivers on Windows and Mac OS X, may actually see the light of day, prior to the Wayland push...

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

  • #2
    X12

    http://www.x.org/wiki/Development/X12 ?

    Comment


    • #3
      Originally posted by uid313 View Post
      Time will show.

      Comment


      • #4
        Hey,as long as they keep their eyes on Wayland and feet on xorg we'll be fine.
        I hope this is true but I hardly doubt that this will make X12,it will probably have those features mentioned and other that help to transition to wayland (xwayland) .Just a mix of things so that we have stable ground.

        Comment


        • #5
          Not X12, a change in X.Org implementation

          X12 would be about an (incompatible) evolution of the X protocol to fix the issues of the X protocol.
          Here it's about changing the implementation of X.Org: contrary to what many believes, to fix many current issues of the X server, changes of the X protocol are not required but its a manpower issue: see David Airlie's end of the blog where he is (bitterly) joking that noone will port the drivers to the new API.

          And those who believes that Wayland will fix the manpower issue are just dreaming..

          Comment


          • #6
            Defines for multiple device driver APIs and USB hotplugging support in a 2D, network transparant, interaction driven canvas? Sorry but... Yuck! Blegh!

            Wayland is a more UNIX philosophy-like component; one thing for one job. Why not implement network transparant input devices where it's at? The kernel. And why not implement network stream of widgets where it's at? The widget toolkits.

            This whole X11 architecture is dead. Instead of defending it, why not evolve it? Plan9 did something similar; network the hell out of everything. If you van do it with storage and HTML, then you can do it with IO and widgets, too!

            Comment


            • #7
              Plan 9 from Bell Labs

              Yeah, look at what there is to learn from rio & blit in Plan 9 from Bell Labs.
              Plan 9 is Unix 2.0, its written by the guys who wrote the original Unix, these guys know what they're doing.

              Comment


              • #8
                Originally posted by uid313 View Post
                Yeah, look at what there is to learn from rio & blit in Plan 9 from Bell Labs.
                Plan 9 is Unix 2.0, its written by the guys who wrote the original Unix, these guys know what they're doing.
                Have they tested how rio behave in a WAN?
                That said, given the UI of Plan9, I'm not sure that the comparison is useful, Plan9's may have a nice design but its GUI is quite "different".

                Comment


                • #9
                  Cosmetics have nothing to do with the inner workings. Running on wireless of wired is completely irrelevant due to TCP/IP as opposed to UDP.

                  Are you even aware of what we're discussing here? (no offence )

                  Comment


                  • #10
                    Originally posted by V!NCENT View Post
                    Cosmetics have nothing to do with the inner workings. Running on wireless of wired is completely irrelevant due to TCP/IP as opposed to UDP.
                    Are you even aware of what we're discussing here? (no offence )
                    You should be more cautious when you're responding, especially since your answer is making me think that *you* have no clue.
                    WAN == Wide Area Network i.e. I was asking if rio worked well in a (wired) network which has high latency and small bandwidth: having a remote display working well in a LAN is easy, having a remote display working well in a WAN is hard..

                    And yes, the look may be important: as "cosmetic effects" can use a lot of bandwidth (which isn't so good in a WAN).

                    Comment

                    Working...
                    X