On Tuesday 03 February 2009, Johann Haarhoff wrote:
> Hi Samir/All
>
> >>> I found a quirk when trying to maximize vetically then horizontally
> >>> then unmaximize vertically and cycling so on.
> >>> This doesn't work all time not all combinations at least.
> >>> I obtain sometimes a different window size when going back to the
> >>> original size, pretty complicated to explain.
>
> OK, I believe I have fixed the inconsistencies here. Some testing and
> feedback by someone else would be useful.
>
> >> Yes, there is still some ironing out to do. I also discovered that
> >> my patch breaks the "Run" dialog, it comes up but does not accept
> >> any keyboard input.
>
> I managed to fix this, but I fixed the problem in a file unrelated to
> my patch. The offending lines were 1702 & 1703 in window.c
> (wManageInternalWindow):
>
>      if (wwin->flags.is_gnustep == 0)
>          wFrameWindowChangeState(wwin->frame, WS_UNFOCUSED);
>
> This causes internal windows (like the run dialog) to immediately lose
> focus, and thus lose keyboard input. I commented these and have seen no
> ill effects, but they have obviously been put there for a reason.

You've seen no ill effects because you do not use GNUstep, but people 
using it will see ill effects.

> According to "hg annotate" those lines were last touched by Dan Pascu
> in 2004 to convert tabs to spaces so the real purpose is hidden.
>
> Anyway, let me know if this breaks anything for anyone.

For people using GNUstep.

> I am getting to 
> know the code a bit better by now so future fixes should be a bit
> quicker :)
>
> Regards,
>
> Johann
> P.S. patch is against current tip, and not on top of my previous patch



-- 
Dan


-- 
To unsubscribe, send mail to [email protected].

Reply via email to