[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-devel
Subject:    Re: Qt4 Themes
From:       Sandro Giessl <sandro () giessl ! com>
Date:       2006-07-03 17:18:13
Message-ID: 200607031918.13408.sandro () giessl ! com
[Download RAW message or body]

On Sunday 02 July 2006 15:40, Thomas Lübking wrote:
> Am Sonntag, 2. Juli 2006 13:28 schrieb Luciano Montanaro:
> > is it really needed?
> > We could just check if the style is a subclass of KStyle with
> > style->inherit("KStyle") or some such.
>
> ... qobject_cast<KStyle*>(style())
>
> good idea anyway.

Yes.

> however the other way is much more flexible and allows to bring up new
> elements even aside the KStyle base development (guess you wrote a widget
> for an app, release that on kde-apps and define an element (beyond a kde
> common base offset))

And if several apps do that and enums get mixed up? We would need a central 
register of custom base offsets for applications... sounds way too 
complicated to me.
I would be glad enough if only widgets residing in kdelibs would be well 
supported by KStyle.

> style developers could start to support this widget in their styles long
> before the next KDE release is out - that than can decide if the widget's
> worth to be included or at least officially supported in the KStyle enum.
>
> as the change itself (aside the binary compatibility stuff) is minor (no
> break for any existing app) and simple (due to the very specific way /all/
> draw* routines are implemented ;) and i doubt there're too many qt4 styles
> in the wild yet that'd needed to be adapted - hey why not =)

Well, due to the binary compatibility stuff, current method signatures cannot 
be changed, or am I wrong? Trolltech would have to overload each method with 
the bool return type. Not sure how their attitude towards bloating their API 
this way is, but I informed a troll about this thread.
 
>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<

[prev in list] [next in list] [prev in thread] [next in thread]