Qt 3.1b1

[Matthias Ettrich]

On Monday 09 September 2002 16:26, Neil Stevens wrote:
[snip]
>
> It *is* KDE's fault of KDE requires Qt 3.1 and the use of Qt 3.1 breaks
> compatibility.  If KDE chose to use Qt 3.1, despite having found out that
> Qt 3.1 is BIC, whose fault can it be?  Qt is free software, so nobody
> *made* KDE upgrade.

Neil, get the facts right, please. Qt 3.1 is _not_ binary incompatible to Qt 
3.0. Qt 3.1b1 had a few small issues, but those have been fixed already.

The moc change with qt_static_property is a different kind. 

Before I read any more nonsense like "KDE shall not upgrade", "kill 
Trolltech", "fork Qt", bla bla, can you all sit down for a minute, take a 
deep breath and relax and remind yourself that we are actually working on the 
same project?!

As long as I'm responsible for Qt development, and with Reggie, Lars and Harri 
on the team, we'll hardly do something that fucks KDE up. Ok? Everybody 
getting this? Thank you.

We are putting a great deal of work into Qt, among others for the sake of KDE. 
A lot of the work we do is inspired by the bug reports and suggestings we get 
from you guys. Suggesting not to make use of this work is like saying to the 
KTHML guys that you rather prefer not having their new code in KDE but stick 
with an older version of the engine. In case you wonder why I react 
personally, that's why.

Now back to the moc issue: we haven't decided yet, simply as that. We could do 
it differently and drop some features if the issue turns out to be too 
problematic. Personally I'm not yet convinced that the issue is a real one. 
Why would anybody who develops KDE applications (meaning: he subclasses KDE 
classes and puts Q_OBJECT macros in his subclasses) upgrade Qt from source 
without upgrading KDE? That just does not make sense to me.

Again: we won't do it in open disagreement with the KDE release team.

Simon already said that he doesn't see much point in reverting it, but I'd 
also like to see the opinion of other guys first, for example Dirk.


Matthias