Announcement

Collapse
No announcement yet.

Mir Development Stats Dominated By Canonical

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

  • Mir Development Stats Dominated By Canonical

    Phoronix: Mir Development Stats Dominated By Canonical

    For those curious about the Mir Display Server development but aren't actively following its Bazaar development repository, the development continues to be dominated by Canonical and here's some numbers looking at the current development statistics surrounding Mir...

    http://www.phoronix.com/vr.php?view=MTM5NTg

  • #2
    Shocking news, that only Canonical is working on Mir. But why did they clone Alan Griffiths, Alexandros Frantzis and Christopher James Halse Rogers?

    Comment


    • #3
      Originally posted by oleid View Post
      Shocking news, that only Canonical is working on Mir. But why did they clone Alan Griffiths, Alexandros Frantzis and Christopher James Halse Rogers?
      I thought the same in the first look. Forgot to proof-read Michael?

      Comment


      • #4
        The good news

        Canonical has already surpassed Wayland in lines of code and revisions and they also won nVidia's support. This shows that Canonical wants modern graphics features ASAP. I want them too. Hopefully this controversy will drum up more support for Wayland. There is always more than one solution to a problem.

        Comment


        • #5
          Originally posted by shaunehunter View Post
          Canonical has already surpassed Wayland in lines of code and revisions
          So Mir has more lines of code and more revisions but still does far less than Wayland? I thought Ubuntu fans have been saying that Mir was going to have a leaner, better-designed code-base. So why do they require more code to accomplish less?

          Originally posted by shaunehunter View Post
          and they also won nVidia's support.
          Citation needed.

          Originally posted by shaunehunter View Post
          This shows that Canonical wants modern graphics features ASAP.
          If that was really the case they would have contributed those lines of code and revisions to a nearly-finished product rather than starting over from scratch

          Originally posted by shaunehunter View Post
          I want them too.
          I don't. I want it done right. I don't want to be stuck with a poorly-thought out, bloated, and convoluted solution for the next decade or two. The graphics system is not something you can afford to rush or cut corners on, because it is not something you will be able to easily make radical changes to down the road. If you make fundamental mistake, you are stuck with it.

          Comment


          • #6
            Originally posted by shaunehunter View Post
            Canonical has already surpassed Wayland in lines of code and revisions [...]
            LOC isn't a valid measurement. Some say WTF/LOC is the only valid measurement of code quality


            Originally posted by shaunehunter View Post
            [...] and they also won nVidia's support.
            As stated multiple times by numerous people, supporting Mir means supporting Wayland and vice versa, as they only need OpenGL-ES... And NVidia eventually have to support Wayland, as RHEL will have to switch to Wayland at some point.


            Originally posted by shaunehunter View Post
            Hopefully this controversy will drum up more support for Wayland.
            I don't think much has changed due to Mir. The projects which announced Wayland support after hearing about Mir where working on Wayland anyway...

            Comment


            • #7
              And i think thats exactly why Canonical abandoned wayland. Except technical reasons on phones and tablets, i think the Wayland development rate is too slow for a company whis ambitious plans.

              Wayland is on development 2 years and seems that it will take another 1-2 till we see an end-user environment running smooth on a distribution, while Mir in less than six months is running in many phone devices, in less than a year should be ready for phones/tablets and in 13 months from the start of the project will be running as a default on Ubuntu 14.04 LTS/Unity 8!

              As an end user i see Canonical as the only company that can compete with google/android, microsoft and apple on OS terms. and in contrary to android, the so weird Linux community will be benefited from a quality display server with big support and developing rate, with a great, fast, productive environment (unity) and the better support from nVidia, AMD and Intel as the OS will become more popular and the gaming thanks to Canonical and Steam will be more popular.

              from a Windows user for a decade and Ubuntu user the last 2 years.

              Comment


              • #8
                Originally posted by oleid View Post
                LOC isn't a valid measurement. Some say WTF/LOC is the only valid measurement of code quality
                i dont get what that comparison is meant for at all?! Just another MIR vs. Wayland bit to keep the flamewars going.

                Originally posted by oleid View Post
                As stated multiple times by numerous people, supporting Mir means supporting Wayland and vice versa, as they only need OpenGL-ES... And NVidia eventually have to support Wayland, as RHEL will have to switch to Wayland at some point.
                i think there are still people out there who think ubuntu/canonical stole the drivers support. (like nvidia etc. would now only support MIR and not Wayland). But as you said its still the same for MIR and Wayland.

                Originally posted by oleid View Post
                I don't think much has changed due to Mir. The projects which announced Wayland support after hearing about Mir where working on Wayland anyway...
                There are alot of projects who didnt even talk about supporting wayland in near future and now kinda rush in and bring wayland support to the top of the todo to show that they already support wayland and it wouldnt make any more sense to work with MIR. Like KDE told "first we will make the transition to qt5 and then look into wayland" and now everyone at KDE seems busy to work on supporting wayland just to proof the point i meantioned above.
                So there is definitively an effect that the MIR announcment had on the wayland supporting.

                Comment


                • #9
                  Originally posted by TheBlackCat View Post
                  So Mir has more lines of code and more revisions but still does far less than Wayland? I thought Ubuntu fans have been saying that Mir was going to have a leaner, better-designed code-base. So why do they require more code to accomplish less?
                  Unit tests. Apparently they take a lot of space.

                  Originally posted by oleid View Post
                  As stated multiple times by numerous people, supporting Mir means supporting Wayland and vice versa, as they only need OpenGL-ES... And NVidia eventually have to support Wayland, as RHEL will have to switch to Wayland at some point.
                  EGL, not OpenGL ES.

                  Comment


                  • #10
                    Originally posted by k1l_ View Post
                    There are alot of projects who didnt even talk about supporting wayland in near future and now kinda rush in and bring wayland support to the top of the todo to show that they already support wayland and it wouldnt make any more sense to work with MIR.
                    I'm curious... do you have an example? For KDE...

                    Originally posted by k1l_ View Post
                    Like KDE told "first we will make the transition to qt5 and then look into wayland" and now everyone at KDE seems busy to work on supporting wayland just to proof the point i meantioned above.
                    So there is definitively an effect that the MIR announcment had on the wayland supporting.
                    They need to do the transition to Qt5 as this one includes a wayland backend. As for KWIN, according to the main maintainer, they are working on it since 2011... The work started by refactoring the code base in order to support it.

                    Maybe GNOME and KDE take the transition more important nowadays, however, claiming they didn't talk about it is not true.

                    Originally posted by GreatEmerald
                    EGL, not OpenGL ES.
                    Right, the OpenGL ES Native Platform Graphics Interface aka EGL.

                    Comment


                    • #11
                      Originally posted by oleid View Post
                      I'm curious... do you have an example? For KDE...
                      cant find the quote from back then (the KDE vs MIR issue brings alot of false positives)

                      Originally posted by oleid View Post
                      They need to do the transition to Qt5 as this one includes a wayland backend. As for KWIN, according to the main maintainer, they are working on it since 2011... The work started by refactoring the code base in order to support it.

                      Maybe GNOME and KDE take the transition more important nowadays, however, claiming they didn't talk about it is not true.
                      Yes, there was some preparation before MIR. But to say MIR didnt have an effect on the wayland transition isnt true, too. Like i said it got pushed to the top of the TODOs where it was not before MIR. The truth lies anywhere between us both, imho.

                      Comment


                      • #12
                        Originally posted by k1l_ View Post
                        Yes, there was some preparation before MIR. But to say MIR didnt have an effect on the wayland transition isnt true, too.
                        As I said: Maybe GNOME and KDE take the transition more important nowadays [...]

                        Comment


                        • #13
                          Originally posted by k1l_ View Post
                          Yes, there was some preparation before MIR. But to say MIR didnt have an effect on the wayland transition isnt true, too. Like i said it got pushed to the top of the TODOs where it was not before MIR. The truth lies anywhere between us both, imho.
                          More wayland coverage and more wayland communication because of Mir announcement, maybe. More work because of Mir, certainly not. It did happen at approximately the same time, but that's because wayland protocol was made stable at approx this time, RB OS (shipping wayland) was made available just a little before, Qt 5 (for which the wayland back-end was added) became ready at approx the same time (do you really think that the schedule of Qt 5 was influenced in any way by the Mir announcement?), etc...
                          Canonical announced Mir just when wayland pieces were coming together, and when most of the remaining work could be started. But just because things happen at the same time, doesn't mean there is a causality link between them.

                          Comment


                          • #14
                            Originally posted by oleid View Post
                            Shocking news, that only Canonical is working on Mir.
                            Yes, I brought that up in the previous Mir discussion concerning copyright assignment, are there any outside developers for Unity? As far as I can tell, Canonical aren't really interested nor expecting in any outside help, not for Unity or for Mir (or likely any other in-house projects).

                            The reason they are into duplicating effort is that of control and I suppose also strong integration into Ubuntu, the latter may end up making it hard to support Mir outside of Ubuntu's ecosystem, as seems to be the case with Unity.

                            Originally posted by oleid View Post
                            But why did they clone Alan Griffiths, Alexandros Frantzis and Christopher James Halse Rogers?
                            LOL yes, how could he miss this? I mean 'Christopher James Halse Rogers', in such a short list of 19 entries, how could he not see a long standout name like this be duplicated? So 19 developers should be 16.

                            Comment


                            • #15
                              Originally posted by BO$$ View Post
                              And they're progressing faster than wayland which has been in limbo for so many years.
                              Based upon what? Lines of code?

                              The increase in lines of code could just aswell be a ton of 'TO-DO' comments or as someone said, lots of different test code which will be pruned once the best approach has been found.

                              The interesting metric here is what they actually do, I suppose we will see some sort of comparison once Mir is up and running in a state to which it can be compared to Wayland, but until that happens we have no idea of where Mir stands in comparison.

                              Comment

                              Working...
                              X