Announcement

Collapse
No announcement yet.

Compiz Will Not Be Ported To Wayland

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

  • #31
    I applaud Sam's words.

    However, I wonder why it occurs NOW. In the past, I have had the feeling Compiz took years and years and years to stabilize, and never really reached a mature state. Now that it seems Canonical puts big $$$ into its development (and now that Compiz is an essential part of their desktop and plans), it seems weird.

    IMHO, Mutter has been in a better state for quite some time already. Why didn't they just wait a little more, before ? Now, it seems way too late to make such a decison. Better late than never ? Probably, but users will suffer from that, again.

    +1 about fragmentation issue (as a clueless user, granted). If think even a user can realize that, even more than a developer actually : he or she just have to USE several combinations of distros & environments to stumble into tons of different problems and notice that some components are not maintained properly here or there ; that some features are much more advanced here or there... And that, overall, everything works "ok-ish" which is OK for most of us because we are patient / use OSS out of principles etc. but not enough for the rest of the world.

    Canonical could have made a difference by supporting massively GNOME for instance. They chose to make their own shell and shopping lenses instead of fixing essential existing stuff (to be fair, it's not that black or white, they also do)
    Last edited by torturedutopian; 01-08-2013, 03:26 AM.

    Comment


    • #32
      Originally posted by ShadowBane View Post
      Yes, an open source (licensed for both GPL and LGPL) library is quite the walled garden... *insert remark about how GPL and LGPL both protect the freedom to do what you want with the code*
      I think that the issue is that if you want to contribute your code to Qt you have to allow Digia to sell your code in its proprietary version.
      This mandatory contributor agreement can be considered making Qt a walled garden, though with a very small wall as one can still fork LGPL's version of Qt if Digia misbehave.

      Comment


      • #33
        Originally posted by renox View Post
        I think that the issue is that if you want to contribute your code to Qt you have to allow Digia to sell your code in its proprietary version.
        This mandatory contributor agreement can be considered making Qt a walled garden, though with a very small wall as one can still fork LGPL's version of Qt if Digia misbehave.
        The agreement turns it into a better version of the BSD license, essentially. It can be used under a proprietary license, but only if you fund further development.

        I prefer the pure LGPL, but funky here obviously has other issues that bother him.

        Comment


        • #34
          Originally posted by Delgarde View Post
          Yeah, in hindsight, moving away from Mutter was a bad call for Ubuntu. True, Mutter wasn't performing well at the time, but it wasn't long before the Gnome devs addressed most of those problems.
          I'd like to see Canonical to contribute more upstream instead of forking and reinventing the wheel. However, they do need more control over their core parts, and the route the Gnome has taken, under the control of some zealots, removing features, without a clear roadmap and rejecting patches sent by Canonical that do not fit their vision, does not help the trust and communication between Canonical and GTK/Gnome devs, so they feel they need more control over their core parts. The bad thing is that the open source world is full of good guys, but it is too many times driven by pride and not by providing solutions to problems.

          Anyway, I think Sam Spilsbury is doing the right thing.
          Last edited by newwen; 01-08-2013, 07:26 AM.

          Comment


          • #35
            Originally posted by newwen View Post
            I'd like to see Canonical to contribute more upstream instead of forking and reinventing the wheel. Anyway, they do need more control over their core parts, and the route the Gnome has taken, under the control of some zealots, removing features, without a clear roadmap and rejecting patches sent by Canonical that do not fit their vision, does not help the trust and communication between Canonical and GTK/Gnome devs, so they feel they need more control over their core parts. The bad thing is that the open source world is full of good guys, but it is too many times driven by pride and not by providing solutions to problems.
            Maybe they should consider KWin then. KDE seems to go in the complete opposite direction on features, and for a generic compositor that might be just want Canonical wants. I just know KDE is pretty tightly integrated, so that might not be an option just on the required library support that would pull into Ubuntu base. Maybe they should just switch to the KDE stack entirely since they are pushing Qt on the phone anyway. They'd get Wayland support a lot sooner that way, I'd think.

            Comment


            • #36
              funkSTAR, i have a challenge for you: create a Qt (or KDE if you prefer) flame on some Justin Bieber forum of your choice!

              Comment


              • #37
                Until Microhard goes down I'd rather there be fragmentation/many distros - otherwise they will concentrate all their covert and overt war on linux on the one distro. If I'm remembering correctly they already attacked Redhat and Opensuse and won/perverted it. That is, I'd rather them shoot at a moving target or maybe they realized that already so they did that UEFI thing.

                I think of the various linux distros as kind of like the evolution of life forms - which ones will survive and which ones will die?
                Well evolution is based on "natural selection" which is a combo of forces outside of the lifeforms' control (environment) that exert an effect on what it needs to do to survive (example the bees and insects don't like your flower color therefeore you die cuz your seed does not spread or a predator can see you easier against the background than he can others or weather f's you up easy) and the other part of it is the random mutations in the DNA (or roll of the dice) and sometimes you just get lucky. The random DNA mutations you could assign as set up by a creator if you wish as an ingenius mechanism.

                I see the various distros differing details as DNA mutations and diversity is good and I see Microhard as a predator.


                Comment


                • #38
                  Originally posted by garegin View Post
                  the problem is that windows is like a petri dish for developers, while Linux distros are on of the worse platforms to develop for. there are no IDEs on par with Visual Studio or XCode and the base OS is fragmented between different glibc, gcc, kernel and x.org versions. the result is that it is more work to publish an app for Linux, than for Windows or OS X.
                  Code:
                  git clone xxx && cd xxx
                  ./autogen.sh
                  make
                  make install
                  Originally posted by Luke_Wolf View Post
                  I do have to wonder how many people who are complaining about the "Fragmentation" for development are actually developers... as I'm hearing a lot of lack of clue from that side here...
                  You're wondering about that? Jeez, it's so dead obvious: 1. Posts on phoronix. 2. Has no fucking clue what they're talking about.
                  What more, exactly, do you need to tell these shits have never contributed anything in their life while pretending to be open source fanboys,
                  mourning the downfall of their beloved gratis software due to "everybody doing what the fuck they want in their unpaid free time", aka "fragmentation", aka "use what works best for you".

                  It's again one of those days I'm wondering why I'm ruining my mental health browsing these threads..

                  Comment


                  • #39
                    Originally posted by zanny View Post
                    I want to reaffirm this, I much prefer QTCreator / MonoDevelop to VisualStudio / Xcode. Those latter two are slow as shit bloated messes that always crash when I use em.

                    And Linux as a platform is much nicer to develop for too because you can get almost all the info you need in the filesystem. In Windows you are stuck using WIN32_FUCKYOU_CAPSLOCKFUNC55(STUFF, STUFF, WHAT, MORE USELESS STUFF, WELCOME TO 1985), or WinRT if you really are masochistic.
                    Thanks. I have absolutely no experience with Visual Studio, but I absolutely love QtCreator, so every time someone praises VS
                    I'm left to wonder what makes it so much better than QtCreator.

                    Comment


                    • #40
                      If the fragmentation is so bad and it is so hard to release games for linux, how is it possible that every single humble indie bundle game does just fine while most of the developers hardly have the manpower to invest much time into that?

                      Comment


                      • #41
                        Originally posted by ChrisXY View Post
                        If the fragmentation is so bad and it is so hard to release games for linux, how is it possible that every single humble indie bundle game does just fine while most of the developers hardly have the manpower to invest much time into that?
                        What do you think you're doing here, with all your logic and reasoning?

                        Comment


                        • #42
                          Good. Just that - good. Go Weston, go Openbox. A project of critical system component should be allowed to struggle only for a short period of time. Too long - it should die, as either the team is not up to the task or it needs to be re-written from the ground up.

                          On a side note - overcomplicating a critical software layer just for the sake of visual effects should be punishable or made possible only via separate modules.

                          As for the fragmentation. While you may disagree with this being a problem, but the point about wasting resources is still valid and there's nothing you can say to undermine that. Let's be honest. Vast majority of OSS projects are just for the sake of fun and their authors don't have think about the end results, nor the place of the project in Linux desktop let alone end-user.

                          Canonical:
                          Let's look at it from a different direction. How do they distribute their efforts?
                          Upstream - near 0%
                          In-house projects that serve no one but Canonical - a lot

                          Comment


                          • #43
                            Originally posted by torturedutopian View Post
                            Canonical could have made a difference by supporting massively GNOME for instance. They chose to make their own shell and shopping lenses instead of fixing essential existing stuff (to be fair, it's not that black or white, they also do)
                            The problem is, of course, that GNOME upstream (whatever or whomever that might be; some people suggest it's actually just Red Hat ) did not agree with Canonical's vision on how Ubuntu/GNOME should work. So Canonical had 2 choices: following GNOME blindly (and listening to everybody complaining that they didn't contribute enough to GNOME while knowing that GNOME would likely turn down the patches they wanted included...) or implement their own desktop shell. And when they started to go for the second option, they found out mutter was not a usable choice back then (it also didn't become until several releases further), so they chose a solution that did actually work at the time... (but is presently proving to be less than ideal).

                            Now, looking back at things, it would probably have been better if they could have worked together, but working together has to come from two sides, and it seems like right now the GNOME (Shell) people are still not wanting to actively work together with others (or at least, I don't see any other reason why the Linux Mint people would want to fork mutter).

                            BTW: forks also happen in closed source software, usually when 2 partners disagree on the future of a project. Microsoft Windows NT & IBM OS/2 were forks of the same project, and so are Sybase's & Microsoft's SQL database servers (actually MS SQL Server forked from Sybase, itself a fork of the Ingres database which was also forked into several other commercial databases like Illustra/Informix and HP NonStop SQL). Illustrating that this is not only a problem with open source projects...

                            Comment


                            • #44
                              Originally posted by johnc View Post
                              It's kinda the same on Windows also. There are 400 million different applications that all do the same thing.

                              Granted there is usually one that is a cut above the rest... but a lot of times those are coming from a corporation.
                              OTOH, I know (knew) lots of open source & freeware applications that easily beat any commercial alternative for it on Windows (and Linux). Let's say that the software license does not imply any quality, and maybe that certain types of applications work better as open source (or freeware, for that matter) than others. Complicated specialist software means that the number of skilled developers is quite low, and as such a company might be able to ask lots of money for a working solution that is difficult to build by a couple of volunteers.

                              BTW: Maya was mentioned here somewhere as an example of closed source software being better, but Maya actually depends on several open source projects like Python, Qt, etc., and where they don't use open source, like in l10n of text strings, their own solution is inferior to the common open source solutions.

                              Comment


                              • #45
                                Originally posted by JanC View Post
                                OTOH, I know (knew) lots of open source & freeware applications that easily beat any commercial alternative for it on Windows (and Linux).
                                mplayer2 and XBMC being examples

                                Comment

                                Working...
                                X