Abdelrazak Younes <[EMAIL PROTECTED]> writes:
| Lars Gullik Bjønnes wrote:
| > Abdelrazak Younes <[EMAIL PROTECTED]> writes:
| > | If I can achieve 16 seconds on my less powerful CPU, I reckon that,
| > | with a few X11 specific optimisation, you could achieve 8 seconds and
| > | be on par with LyX/xforms.
| > | | > The tests might however not be entirely fair since xforms is
| > quite bad
| > | > at fonts and anti-aliasing. But that difference should not be _that_
| > | > big.
| > | | I'll fix the last compilation bug so you can test with my branch.
| > I just did that and for the scroll-down test it is quite unusable,
| > not
| > because of speed, but because the screen is not updated. (And the
| > scrolling does not stop when I release the PageDown button).
|
| Interesting... What fix did you put in? The same as me (preeditString())?
| I don't understand why the redraw on screen doesn't work though...
I just updated and used your fix.
| > But even without seeing what goes on the scroll down takes ~23
| > seconds. (4.1.4)
| > So from that point of view trunk is working a lot better.
|
| That is an unfair comment Lars. Do you expect me to fix all issues
| without having access to an X11 platform?
No, but I would expect Qt4 to behave the same way on different
platforms. And if you want access... I can provide that... but the
network lag will probably make it unusable.
However if I turn on USE_KEY_BUFFERING¹ it behaves properly, updates
screen and stop when I release the button
After this change a scroll-down takes ~26 seconds... is this showing
that the painting of the screen does not really take much time?
¹ Which is not about keybuffering at all, but about event pruning.
--
Lgb
