Re: amazing moving/growing windows!!

On Tue, Jul 05, 2005 at 03:32:43AM +0100, Renato Serodio wrote:
> Matlab opens a window if you issue a plot command. This one gets
> positioned nicely. You can also plot into new windows, obviously, but
> Flux deals strangely with that: the first window gets moved slightly
> upwards, and when you click the second, it might move up not so
> slightly; rather, I had a few escaping the screen!
> 
> Then, Ansys. This one not only moves windows, but starts enlarging them
> to a point your memory pops. Anyway, no one wants to pan 2km with the
> mouse.. Just imagine, you bring up the Save/Quit dialog and the window
> starts growing unboundedly (and quite fast) and you have no chance of
> clicking any button!

I've never experienced unwanted repositioning, but I have experienced
unwanted resizing. When I put two gvim windows together in a tabbed
group and then have one gvim instance split its display vertically
(by hitting C-W v), the windows progressively resize to a narrow strip.

> With Gnome there's no such magic behaviour. I won't go as far as saying
> Flux is to blame, because my experience with Ansys tells me they had
> mechanical engineers developing the interface, but something must
> account for this incoherent behaviour in the two WMs.

After some poking around, I decided that my problem was the result of
gvim trying to resize itself to not waste space: if it has a few extra
pixels on the edge that aren't enough to display a whole character, it
shrinks itself slightly to eliminate this space. When you put together
in a tabbed group two windows with different ideas of what the optimal
size increment is (the buffer divider in gvim takes less space than
would a regular character column), they keep "correcting" each other.

The programs that resize themselves automatically seem to have been
written without the possibility of tabbed window managers in mind.

I was satisfied to stop putting multiple gvim instances in tabbed
groups. I would rather like it, though, if there were some mechanism to
disregard non-user-initiated resize events for certain application
classes, as a sort of work-around.

-- 
Gregory K. Johnson


-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click