Subject: Re: Release/Tree closing
From: Martin Sevior (msevior@mccubbin.ph.unimelb.edu.au)
Date: Thu Jul 12 2001 - 02:14:45 CDT
On Wed, 11 Jul 2001, Joaquin Cuenca Abela wrote:
>
> --- Dom Lachowicz <dominicl@seas.upenn.edu> wrote:
> >
> > Mike's bug could be Win32 specific or specific to
> > Mike's computer. It isn't
> > reproducible on Unix. But that's beside the point.
>
> nope, I reproduce it. I've using my sister computer,
> and I've just compiled the windows version, and run a
> little speed test between msword and abiword (we loose
> by little, we get 92% of msword speed, but the test
> was not very complete, I just wanted to get a rough
> approximation).
>
> The sad thing is that the repaint was completely
> broken. It's was a major pain to use abi (I was only
> seeing the first line of each paragraph, except when
> you change the size of abi's window).
>
> I vote to hold 0.9.0 until (at least) that gets fixed.
> (It's *really* unusable)
>
OK, could people who do win builds might like to try this. I put in some
code to speed up scrolling by limiting the requncy of redrawUpdates
following a lineup/line-down command. Anyway try commenting out these
lines in fl_DocLayout.cpp as a quick a dirty way that might fix the
problem.
//
// Check if we've been asked to wait for a while..
//
// UT_uint32 skip = pDocLayout->getSkipUpdates();
// if(skip > 0)
// {
// skip--;
// pDocLayout->setSkipUpdates(skip);
// return;
// }
//
These are lines 1299 - 1311 and this is in method:
void FL_DocLayout::_redrawUpdate(UT_Timer * pTimer).
Just a thought. This code works perfectly on Unix so there must be
something screwy happenning. In my experience this sometimes comes from
weird timer probllems so....
Cheers
Martin
This archive was generated by hypermail 2b25 : Thu Jul 12 2001 - 02:14:55 CDT