GNOME Bugzilla – Bug 93590
metacity should make up a saved size for automaximized windows
Last modified: 2004-12-22 21:47:04 UTC
Description of Problem: Metacity should remember to sizes when remembering application window sizes. One for a maximized window and one for a non maximized window. So that if I start an application that at the last start was maximized it should start maximized and then when unmaximized it should return to a resonable size instead of just becoming resizable and partly of screen. Steps to reproduce the problem: 1. Start gnome-system-monitor 2. Maximize it 3. Quit gnome-system-monitor 4. Start gnome-system-monitor 5. Un-maximize it. Actual Results: gnome-system monitor remains huge, but becomes a pain to resize because it's corners are off screen. Expected Results: gnome-system-monitor should go back to a resonable un-maxmized size, preferably the size it was before the initial maxmization. How often does this happen? everytime. Additional Information: Some apps start at the same size no matter what. Others such as nautilus and galeon exibit this problem.
metacity doesn't remember any window sizes - what you're seeing here is totally a function of the applications in question, and needs to be handled by the apps. (My personal opinion is that we need a protocol for metacity to handle it instead, but I have not been able to convince wm-spec-list@gnome.org of the need for this. There's an open metacity bug still to work on changing the WM spec I think.)
Look at this bug in redhat's bugzilla: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=82550 I attached a patch that works that around by setting up an acceptable unmaximized size for apps that don't work correctly. That's a hack but that makes your life easier when you use Mozilla or OOO for example. The patch is here: https://bugzilla.redhat.com/bugzilla/attachment.cgi?id=90034&action=view
Adding patch keyword, adjusting severity.
this patch seems like a pretty good idea to me; one caveat though: I suppose ideally we would ensure that the saved_rect we give it isn't smaller than the min_size hint, but the constraints code would enforce that anyway so it's not really a big deal.
We may want to base the default saved size on the work area size somehow, or it could still be larger than the screen. In any case, seems sane. I wouldn't worry about the minimum size issue, constraints will fix it.
the patch was out of date, so I had to reimplement it. Attached is the final patch committed to HEAD.
Created attachment 23204 [details] [review] final patch
*** Bug 123800 has been marked as a duplicate of this bug. ***
I hadn't seen it was fixed. Thanks Rob !