Drag and drop optimization

[Waldo Bastian]

On Tuesday 04 February 2003 13:42, Lubos Lunak wrote:
> On Tuesday 04 of February 2003 13:31, Waldo Bastian wrote:
> > On Tuesday 04 February 2003 13:08, Lubos Lunak wrote:
> > > One more thing to do could be to have larger
> > > 'sameanswer' area (see the XDND spec), but I'm afraid that wouldn't be
> > > that trivial in Qt.
> >
> > For kicker I commited the following:
> > -    if (QUriDrag::canDecode(ev))
> > -       ev->accept(rect());
> > +    if ((ev->source() != this) && QUriDrag::canDecode(ev))
> > +        ev->accept(rect());
> > +    else
> > +        ev->ignore(rect());
> >
> > Which disables dragging an item onto itself and sets an ignore-area.
> >
> > The latter should set the 'sameanswer' area, is that what you mean?
>
>  See http://www.newplanetsoftware.com/xdnd/ , section 'Client messages',
> description of 'XdndStatus'. The 'sameanswer' area is a rectangle where it
> doesn't make sense to send additional move events, because ... well, the
> answer from that area will be always the same. Sounds simple, but getting
> it from Qt widgets may not be as easy.

What I mean is that QDragMoveEvent specifically foresees in such rectangle. I 
would assume that Qt uses that information for this Xdnd 'sameanswer' area.

> > If that helps we might want to scan KDE source for this kind of usage and
> > set appropriate acceopt/ignore rects.
>
>  I actually don't find that necessary and worth it. Would you be really
> that happier to see that DND operations take only let's say 10% CPU instead
> of 100% for the short amount of time you do the DND operation?

I never say no to extra cpu cycles :)

Cheers,
Waldo
-- 
bastian@kde.org -=|[ SuSE, The Linux Desktop Experts ]|=- bastian@suse.com