Announcement

Collapse
No announcement yet.

Apple Looks To Take Over X Server 1.9 Release Management

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

  • Apple Looks To Take Over X Server 1.9 Release Management

    Phoronix: Apple Looks To Take Over X Server 1.9 Release Management

    With yesterday's successful release of X.Org Server 1.9 on time by Intel's Keith Packard, most of the developers will now begin working towards X.Org Server 1.10. Like the past few X.Org Server releases, Keith Packard will go on to continue being the release manager for this new series. In the past there's been the input-expert Peter Hutterer of Red Hat to handle the stable release management duties for the point releases, but he will not be handling it for the 1.9.x series and it looks like Apple may be taking over...

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

  • #2
    I don't know if he should be the maintainer if he needs someone else to do the other half of the job. Why not let the other person be the maintainer and Jeremy can do his thing for Apple?

    Comment


    • #3
      As far as I know there *is* no other person right now, which is why Jeremy volunteered in the first place.

      Comment


      • #4
        What's the advantage to Quartz over X? Why do they even have it? Was it just too hard to do DRI on X at the time Mac OS X came out so they decided to write their own graphics server rather than using X11? Seems like an awful lot of trouble. Now none of the Mac stuff can be easily ported to X11, and they have to run and maintain two graphics servers.

        Comment


        • #5
          It might interest some people to know that he used to be a Gentoo developer.

          Comment


          • #6
            Originally posted by Smorg View Post
            What's the advantage to Quartz over X? Why do they even have it? Was it just too hard to do DRI on X at the time Mac OS X came out so they decided to write their own graphics server rather than using X11? Seems like an awful lot of trouble. Now none of the Mac stuff can be easily ported to X11, and they have to run and maintain two graphics servers.
            Quartz was initially released long before X.org even came into existence. We were all still using XFree86 back then. Quartz debuted with Mac OS X 10.0 in early 2001, and Xorg didn't split from XFree86 until around 2004. Compiz, XGL/AIGLX, and so didn't start to stabilize until 2006.

            DRI didn't exist when OS X came out. Mesa's/Xorg's support for OpenGL 2 didn't exist. Support for OpenCL _still_ doesn't exist in Mesa/X. Gallium3D wasn't even a twinkle in Brian's eye back then. XFree86, Mesa, and the related libraries were more or less barely useful pieces of crap compared to the state of the art Quartz. The gap has narrowed a lot since then, but it's still not even closed.

            Quartz also uses a vastly different rendering model than X11 does. Quartz is more closely aligned with Cairo or Direct2D, being based entirely on a Porter-Duff PDF-like compositing model, while XRender was not even in existence and X11 was still nothing more than ugly lines, rectangle fills, and software rendering, with no support for compositing a desktop.

            Apple did the right thing in terms of technology and competitiveness. They outdid X11 (and Windows) by a huge margin. It wasn't until Windows Vista that Microsoft began to close the gap (and not until Windows 7 and the release of DirectX 10 + Direct2D that they actually caught up), and Xorg is still trying to solve a few pieces of the puzzle that Quartz had figured out years ago.

            Pretty much the only reason for people like us to even consider Xorg superior in any way is the fact that Xorg is Open Source and Quartz is most certainly not.

            Wayland + Cairo-over-Gallium3D would be closer to a Quartz competitor than Xorg ever will be, honestly.

            Comment


            • #7
              Quartz is a proprietary technology which gives Apple the competitive advantage. MacOS X and Quartz are one of the reasons why people dish out huge amounts of money for Apple hardware (I know you can run it on a regular PC nowadays, but this is not supported, and something for experienced users, not typical Apple customers).

              Why on Earth would Apple want their apps to be easily portable to Linux/Unix? It's really bloody obvious why they're not using bog-standard X, but their own closed-source solution: They sell that for a living.

              Comment


              • #8
                DRI didn't exist when OS X came out.
                That's false, DRI was released with XFree86 4.0, which was released on March 8 2000.

                Comment


                • #9
                  Apple got the predecessor to Quartz when they bought NeXT. NeXT's DisplayPostscript exists since the 1980s. The Postscript part was ported to the related PDF technology to avoid licensing fees to Adobe.

                  When Apple acquired NeXT, Apple was almost bankrupt and Apple's Classic Mac OS technology wasn't going anywhere.
                  NeXT's OS was stable and included technology that was already years ahead of anything else *– including display tech.
                  Throwing that away just for the sake of using X11 would be madness from an economical standpoint.

                  It wasn't until Apple recovered and the more important compatibility task (“Carbon” compatibility layer for Classic Mac apps) was completed that they started to support X11.

                  Comment


                  • #10
                    Originally posted by curaga View Post
                    That's false, DRI was released with XFree86 4.0, which was released on March 8 2000.
                    OS X was deep into the development cycle by that time with 10.0.

                    Comment


                    • #11
                      Plus none of that stuff was around when OS X development truly started: With NeXTSTEP.

                      NeXT was the company that Steve Jobs founded after he was kicked out of Apple. They produced a Unix system that was aiming at the business productivity market. It had all the basic elements that still exist to this day with OS X, minus the MacOS legacy bits.

                      It did not use X Windows either.

                      Postscript-based display technology, hybrid microkernel, Unix/BSD-based, Objective-C programming language, ObjectOriented application development based around various 'kits' released by Next. It even had the 'dock'. blah blah blah

                      Sounds familiar? It should. It's basically everything that makes OS X, OS X. After NeXT's hardware failed to be competitive it was purchased by Apple and this is how Steve Jobs got back to being CEO.

                      They tried to standardize the API for Unix GUI application development and they called it 'OpenSTEP'. Linux has a open source OpenStep implementation called 'GNUStep', which is the 'other' GNU desktop.

                      If you look you'd notice that GnuStep and OS X Cocoa APIs are extremely similar.


                      Of course they dropped the Postfix-based display window used in early versions of OS X and replaced it with the somewhat similar Quartz stuff when they released the OS X 10.0 desktop version, but certainly going from what they had in NextStep to X Windows would of been a huge step back in terms of features and performance.


                      ------------------------

                      If you look at the historical rants against X Windows: people complained over and over again that it was too complicated, to slow, and had all sorts of hugely ugly problems and it was not particularly good at networking either. Mostly they said that it would be easier to drop it and do something else and that it would take decades to fix X's deficiencies.

                      Well, here we are. Decades later and much of X Windows has been improved massively through a combination of modernization and workarounds at the widget toolkit level. Still remains a lot of work to be done, of course.

                      Comment

                      Working...
                      X