Announcement

Collapse
No announcement yet.

Digia Is Planning Enginio Cloud Data Storage For Qt 5.3

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

  • #16
    Originally posted by Honton View Post
    Claiming that free software needs CLA is a fallacy. Qt is nothing more than a MIR that did'nt die at birth.
    I wasn't saying it needed it, I was saying it helps. And what relevance does Mir have in this discussion?

    Originally posted by Honton View Post
    You should ask your self if a world without a CLAed tool kit would have produced a less fragmented stack. THAT is true value.
    I don't understand your point, what stack is fragmented?

    Comment


    • #17
      Originally posted by cbamber85 View Post
      I don't understand all the Digia hate - they're acting as a normal company! Shouldn't those in the FOSS community who vehemently hate CLAs be asking themselves - what happened to the alternatives? Specifically what happened to wxWidgets? It used to have near feature-parity, but now it's virtually unused; in fact a major new version of it was launched recently - to zero fanfare.
      Can't answer for wx since I never used it, but FLTK I do use and it's quite alive. Gtk has been losing momentum, but it too is alive.

      Perhaps it's because allowing CLAs struck a happy medium in the licensing of Qt: it made Qt Commercial's existence viable, which it turn funded the vast majority of the development in recent times - hence why it has bounded forward technologically and spread to more and more OSs.
      I would say Qt has been going in the wrong direction for ages now. It's been gaining bloat, and the dynamicness of v5 is very wrong direction too. Now to write a "proper Qt app" you need to write in three languages: QML, JS, and C++. That's coincidentally two too many. There's capabilities that you cannot use from C++, but only from QML now I hear.

      So for me, the vast majority of Qt development in recent times has been a negative.

      Comment


      • #18
        Originally posted by curaga View Post
        There's capabilities that you cannot use from C++, but only from QML now I hear.
        That's the first I've heard - you can write an app entirely in QML/JS, or entirely in C++, there's nothing forcing you to use either/both.

        I was very displeased when QtQuick came in v5, but I've recently started using it for a big project (with a C++ backend), and to be honest I wouldn't go back to the QWidgets library. Each to their own of course, but either way QtQuick/QML isn't going away.

        Though I have to say the whole Enginio thing obviously shite. What a waste of time.

        Comment


        • #19
          Originally posted by curaga View Post
          Can't answer for wx since I never used it, but FLTK I do use and it's quite alive. Gtk has been losing momentum, but it too is alive.



          I would say Qt has been going in the wrong direction for ages now. It's been gaining bloat, and the dynamicness of v5 is very wrong direction too. Now to write a "proper Qt app" you need to write in three languages: QML, JS, and C++. That's coincidentally two too many. There's capabilities that you cannot use from C++, but only from QML now I hear.

          So for me, the vast majority of Qt development in recent times has been a negative.
          QML is a descriptive language. Like XML, or CSS, or really, the list of widget in a toolkit (just to stress that point: to use QWidget it is not enough to know C++, you need to know which QWidgets are available and how they are supposed to be put together). How a descriptive UI language can be "one two many" language for building a user interface is quite beyond me.

          Comment


          • #20
            Originally posted by erendorn View Post
            How a descriptive UI language can be "one two many" language for building a user interface is quite beyond me.
            While QML's declarativness is without question especially nice for UI, using either QtQuick, QtWidgets or any other UI technology, it is also quite nice for other scenarios where one would otherwise have to create a domain specific language (DSL)
            Last year someone had a short talk about various QML use cases at FOSDEM:
            https://archive.fosdem.org/2013/sche...many_faces.pdf

            Cheers,
            _

            Comment

            Working...
            X