On Thu, Jun 20, 2002 at 06:20:09PM +0100, Angus Leeming wrote:
> > Well I've only browsed your xforms email, but we need a sub-region not
> > the entire workarea object, unless you fancy copying a massive pixmap on
> > every cursor blink (didn't think so).
>
> That's what we do now. See the code in xscreen.C. Oh! you have ;-)
Eh ? No it isn't - without my patch, then we only XCopyArea() the amount
of stuff that actually changed - see the optimizations in screen.C (we
can do better, but that's by the by)
> void fl_redraw_form(FL_FORM * form)
> {
> for (FL_OBJECT *ob = form->first; ob; ob = ob->next) {
> if (withinBoundingBox(ob, BoundingBox))
> ob->redraw = 1;
> }
> redraw_marked(form, 0, 0);
> }
>
> Furthermore, each object could decide whether it used this info or just
> regenerate the entire pixmap.
But this is not really relevant - the problem is not drawing of
individual form elements (you mean xforms doesn't even do that ???) but
the fact the entire pixmap has to be copied each time.
> > xev != NULL, it's asking to draw the entire object. This is so broken I
> > don't know where to begin shouting ...
>
> Sure. But this will be equally bad for all other toolkits too unless we can
> think of some way of telling them which area to redraw.
Eh ? screen::expose is already generating correctly sized rectangles
(try it without my patch). This works great with Qt frontend.
> Here's example 1
> Start LyX -> xev == 0.
> presumably because LyXScreen::greyOut() doesn't call expose after it's called
> the painter.
>
> Example 2.
> Start LyX, open a buffer, close it. -> xev == 0 and
> the displayed document is not replaced by a gray pixmap.
> Same reasoning, since here the workarea paint method is called only if xev is
> non-zero.
Yes, this sounds like two bugs. Good call.
regards
john
--
"If a thing is not diminished by being shared, it is not rightly owned if
it is only owned & not shared."
- St. Augustine
