Announcement

Collapse
No announcement yet.

KDE Neon Unstable Now Building With Qt 6 Frameworks/Plasma

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

  • KDE Neon Unstable Now Building With Qt 6 Frameworks/Plasma

    Phoronix: KDE Neon Unstable Now Building With Qt 6 Frameworks/Plasma

    For those feeling adventurous and wanting to see how things are coming along for KDE Plasma 6.0 and KDE Frameworks Qt 6 porting, the KDE Neon Linux distribution with its "unstable" edition has begun building against the Qt6 components...

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    They didn't finish fixing 1000s of wayland bugs but embarking on another 10 year journey of Plasma6 of fixing 500 bugs per month.

    Comment


    • #3
      To pat myself on the back, my main project works flawlessly with either 5 or 6 with exactly ZERO modifications. Not even preprocessor.

      It didn't come easy tho, I do painstakingly ensure to make usage as decoupled and peripheral as possible, the only viable strategy I managed to develop against Qt's even growing list of bugs over the years. It paid off for the bugs, it pays off for performance and maintenance, and now it pays off in portability.

      I could ditch Qt at any point, but it would probably cost a month or so to entirely uproot. Which is pretty good, I bet most big long term projects out there that use Qt are hopelessly stuck with it just like the guys at my last job. It would be equivalent to KDE removing Qt as a dependency.

      Comment


      • #4
        Originally posted by cl333r View Post
        They didn't finish fixing 1000s of wayland bugs but embarking on another 10 year journey of Plasma6 of fixing 500 bugs per month.
        A tad bit negative don't you think?

        Sometimes I feel some folks on this forum are against progress. You have to update your dependencies (e.g.: Qt 1 -> Qt6). And sometimes you have to rewrite a piece of software to fit with the current needs (e.g.: Xorg being build for different computers and use cases).

        I'm not saying previous upgrades of Qt by KDE did not hurt me: I too used buggy software for a while and even switched to Xfce for a while. But I do believe that Xorg needs to be replaced, Wayland is the way to go for now, KDE did well not sticking to (and thus forking) Qt 1.x, and over the years the Linux desktop experience improved.

        To come back to the negativity: Qt 6 latest fixes some Wayland bugs, KDE already fixed another few in their 6-cycle, the Qt 5->6 upgrade is not too involved.

        And that NEON comes with an unstable Qt6 release is just really cool + useful. Thanks Jonathan Riddell!

        Comment


        • #5
          Originally posted by ddriver View Post
          I could ditch Qt at any point
          Interesting. My experience with Qt is that the documentation got a lot worse (often merely out-of-sync), but all in all it's a pretty solid library.

          Do you know an (open source) UI toolkit that does a better job than Qt?

          Comment


          • #6
            Originally posted by rafanelli View Post

            Interesting. My experience with Qt is that the documentation got a lot worse (often merely out-of-sync), but all in all it's a pretty solid library.

            Do you know an (open source) UI toolkit that does a better job than Qt?
            Qt is "the best" - relative speaking - sadly there's nothing better to replace it with, sadly because "the best" does not guarantee quality. I use my own stuff, Qt is used in more of a front end capacity, with a minimal safe subset of its functionality.

            Documentation is fine IMO, well above average actually, I usually find what I need, my main issue is with bugs and design limitations. I owe Qt several significant full rewrite efforts due to hitting something unforeseen fairly down the dev cycle. So this whole design strategy doesn't work, back to square 1.
            Last edited by ddriver; 23 May 2023, 07:31 AM.

            Comment


            • #7
              Originally posted by cl333r View Post
              They didn't finish fixing 1000s of wayland bugs but embarking on another 10 year journey of Plasma6 of fixing 500 bugs per month.
              Actually, a long time ago I've seen some KDE conference in which one of the devs said that a Qt upgrade may regress their Wayland progress by a few months. Idk, maybe these days it's not the case anymore, but we'll have to see.

              Comment


              • #8
                Originally posted by cl333r View Post
                They didn't finish fixing 1000s of wayland bugs but embarking on another 10 year journey of Plasma6 of fixing 500 bugs per month.
                Well, FWIW, Qt 5 was designed pre-Wayland and Qt 6 is supposed to address some of the Wayland concerns. Personally, my only real concern is that KDE 6 (Is it still gonna be Plasma?) will mirror the KDE 5 early releases where it wasn't until around KDE 5.4 or 5.6 that it started to stabilize and become day-to-day usable.

                While I'm hopeful for the future, I still worry that 5 to 6 is going to be one of those times that I wish that Arch Linux offered both the Stable and LTS KDE releases.

                Comment


                • #9
                  I have a bad feeling.

                  Comment


                  • #10
                    I find it adventurous and complementary of them to attempt this at all!
                    At this point, with the Qt release being so open source hostile, it nearly seems to make more sense to just fork the darn thing again and go their own route. Perhaps call it "Kt" though pronouncing that might offend about half the planet's population

                    Qt6, the c++ side of things, is just as great or better as it's ever been.
                    QML on the other side is a downright nasty barely usable piece of crap in Qt6 that only really works if you customize every single darn thing (in other terms, use as little from Qt's stock components). Then again, plasma is super heavy on the customized components and essentially has it's own full fledged component stack so i'm guessing the porting on that area isn't "that" painful.
                    Last edited by markg85; 23 May 2023, 09:02 AM.

                    Comment

                    Working...
                    X