[PyQt] Qt Contributors Summit

Phil Thompson phil at riverbankcomputing.com
Thu Jun 9 09:47:32 BST 2011


On Thu, 9 Jun 2011 09:35:19 +0300, Attila Csipa <pyqt at csipa.in.rs> wrote:
> On Thursday 09 June 2011 00:37:38 you wrote:
>> implementation, ie. QML. What I certainly want is a public C++ API to
the
>> underlying code that QML sits on top of so that alternative declarative
>> solutions can be implemented.
> 
> Wouldn't that be the QtDeclarative module ?

No because that requires QML/JavaScript. People want Python/Python and
Ruby/Ruby and...

> I believe the issue with a
> lower 
> level access is that if a public API is provided, it must be supported.
> The 
> compatibility promise is currently on the QML level - and even there,
QML 
> moves so quickly that you have explicit versioning to avoid
> incompatibilities. 
> A good example is the current move from QML1 to QML2 - while the QML
> syntax in 
> pretty much the same, the underlying C++ code got a massive overhaul. If
> you 
> used the fact that it's on top of QGraphicsView, you could be in
trouble,
> as 
> scenegraph is a very different beast.

What's wanted is the ability to replace QML/JavaScript with something that
calls the same public C++ API that QML/JavaScript does. I don't have a
problem if, at this stage, it isn't possible to define that API, and I am
more than happy to contribute to the development of Qt5 to make sure it is
suitable for Python/Python. I do have a problem if, by the time Qt5 is
released, that API hasn't been properly designed and documented. If that is
too "difficult" then QML2 is being rushed - just like QML1 was.

Phil


More information about the PyQt mailing list