On Fri, Dec 06, 2002 at 02:36:15PM +, Angus Leeming wrote:
> Rubbish, it's telling the LyX core that a resize has occurred that's
> expensive. I'm suggesting doing that only when you've stopped changing the
> dimensions of the window, not for every blessed intermediate step.
I do not use re
On Friday 06 December 2002 2:22 pm, John Levon wrote:
> On Fri, Dec 06, 2002 at 10:09:02AM +, Angus Leeming wrote:
> > And a totally different idea. This is for xforms, but the idea should
> > also work for Qt. Here's the redraw function:
> >
> > Everything here is cheap except for the last cal
On Fri, Dec 06, 2002 at 12:08:42PM +0100, Lars Gullik Bjønnes wrote:
> How can this be true since the xforms is som much faster?
I don't really know. QFontMetrics::width() seems to be much slower, and
of course, we have to construct a temporary QString / QChar
> btw. did you profile scrolling?
On Fri, Dec 06, 2002 at 10:09:02AM +, Angus Leeming wrote:
> And a totally different idea. This is for xforms, but the idea should also
> work for Qt. Here's the redraw function:
>
> Everything here is cheap except for the last call, to the LyX core. So why
> not shove it on a timer and dis
On Fri, Dec 06, 2002 at 10:17:47AM +0100, Jean-Marc Lasgouttes wrote:
> If the functions is called several times in a row with same info (same font,
> maybe even same string!) you could cache data.
Well I considered that but am unable to come up with a really sane
scheme ...
> Instead of buildin
Andre Poenitz <[EMAIL PROTECTED]> writes:
| On Fri, Dec 06, 2002 at 10:09:02AM +, Angus Leeming wrote:
>> Everything here is cheap except for the last call, to the LyX core. So why
>> not shove it on a timer and discard FL_DRAW events that occur within 0.2 secs
>> in the same way as we alrea
John Levon <[EMAIL PROTECTED]> writes:
| I have abjectly failed. Everything I tried increased resize time.
>
| xforms is averaging 1.5 seconds for first-time-resize. Qt is averaging
| 5.5. I can not bring this down.
>
| 4.5 seconds of this is time as a result of qfont_metrics::width()
| (checked u
On Fri, Dec 06, 2002 at 10:09:02AM +, Angus Leeming wrote:
> Everything here is cheap except for the last call, to the LyX core. So why
> not shove it on a timer and discard FL_DRAW events that occur within 0.2 secs
> in the same way as we already fo for FL_DRAG events. Some thought may be
>
On Friday 06 December 2002 9:17 am, Jean-Marc Lasgouttes wrote:
> > "John" == John Levon <[EMAIL PROTECTED]> writes:
>
> John> I have abjectly failed. Everything I tried increased resize
> John> time.
>
> John> xforms is averaging 1.5 seconds for first-time-resize. Qt is
> John> averaging 5.5.
> "John" == John Levon <[EMAIL PROTECTED]> writes:
John> I have abjectly failed. Everything I tried increased resize
John> time.
John> xforms is averaging 1.5 seconds for first-time-resize. Qt is
John> averaging 5.5. I can not bring this down.
John> 4.5 seconds of this is time as a result of
I have abjectly failed. Everything I tried increased resize time.
xforms is averaging 1.5 seconds for first-time-resize. Qt is averaging
5.5. I can not bring this down.
4.5 seconds of this is time as a result of qfont_metrics::width()
(checked using an empty ::width() function).
Basically, our
11 matches
Mail list logo
