Josh Andler wrote:
>
> It seems likely that it may have to do with something that has changed
> with
> GTK somewhere between what we had before and the version we have now. The
> only way I can think to test/check where things went wrong is to download
> other GTK binaries between what we had before and what we have now until
> you
> find one that doesn't show the issues and work up from there until you can
> identify when it broke.
>

Mmmh, actually I rely on the devlibs repository for all updates of such
libraries.
I can see that in rev.22 there has been an update to GTK 2.22 and Glib 2.26.
The preceding one was in January (update to Cairo 1.10.2) and I feel that
the problem was not present so far ago.
The rev.22 update has been quite massive so I don't think I'll be able to
make small steps back.

Correct, the problem was probably not present at that time. However, a minor graphical glitch which will be present on a minority of Windows boxes is far less of an issue than a security exploit with gtk that was present on all Windows boxes.

This may be one of those issues where if you're not willing to do the leg work and test, then no one probably will.

Cheers,
Josh