Re: Bug in GtkText

[Owen Taylor <otaylor redhat com>]

Nuno Ferreira <nmrf@rnl.ist.utl.pt> writes:

> > > /* Scroll the bottom of the line is on screen, or until
> > >  * the line is the first onscreen line.
> > >  */
> > > while (cache->next != text->line_start_cache && lph > height)
> > > {
> > >     TEXT_SHOW_LINE (text, cache, "cache");
> > >     TEXT_SHOW_LINE (text, cache->next, "cache->next");
> > >     scroll_int (text, LINE_HEIGHT(CACHE_DATA(cache->next)));
> > >     lph = pixel_height_of (text, cache->next);
> > > }
> > > 
> > > As you can see cache->next never changes so the loop can never finish.
> > 
> > Hmmm, actually it isn't that simple - cache->next doesn't change
> > but text->line_start_cache and lph _do_ change, so the loop
> > (in theory) can finish. 
> > 
> > I made some fixes to the Text widget recently related to problems
> > like this, so if you are using 1.1.1, you might want to
> > try grabbing the latest version from anoncvs.
> > 
> > Regards,
> >                                         Owen
> 
> This was with the latest version from CVS. I checked it with GDB and I'm
> almost sure that all values were not  changing. I'm certain that at least
> lph kept the same value. If you need further help I'll recheck when I
> manage to have some free time.

OK. As I said the _in theory_ lph should be getting changed
there. Having seen past bugs in this area, it doesn't shock
me to learn that it isn't. I'll take a look. The bug probably
lies in the scrolling code somewhere. 

Regards,
                                        Owen