On Sat, 14 Jan 2006 18:28:40 +0900 Raphael Plasson <[EMAIL PROTECTED]> babbled:
> Carsten Haitzler (The Rasterman) wrote: > > >1. now u have to MOVE windows insdtead of show/hide - this is less efficient > >and causes more exposes and excess draws. 2. because of icccm and reparenting > >windows - moving a window with a desk switch creates lots of complications > >for apps that like to remeber their own geometry and restore it - if they > >exist while off screen (as they then believe they should start up at > >co-ordinates -3000, -1000 for example). 3. this makes descisions as to which > >desktop a window belongs to much harder as when does it change from > >belonging form one desk to another? you now have to put this logic in - > >everywhere 4. this makes shutdown and startup more complex as the wm has to > >move all widnows back to the current screen to avoid them possiblyt being > >lost 5. this is very hard to have BOTH methods (show+hide vs move) for how > >to handle desktop changes within the same code as they are quite different. > > > > > > > And wouldn't it possible to just consider to have a larger virtual > desktop, with windows in a fixed position, so that moving around the > desktops would mean to move the coordinates of the on-screen real > desktop (and not to move the windows) ? I actually assumed that it was > what was performed by fvwm2 in the (so old ?) time I used it - as you > were able to move everywhere in the virtual desktop, and not only to > slide by an integer number of screen size. no. fvwm never did this. this is possible - BUT the window still need to be informed of their moves - they also have massive problems with this. early e17 code did virtual roots this way and a lot of things just break - java, gtk did, and other apps, that are not able to understand virtual roots. icccm and clients place limits on what can be sanely done. > Is there a kind of functionality (directly from X server ?) that would > allow to have absolute coordinates for windows, with the possibility to > actually represent them with relative coordinates (i.e. absolute > coordinates + tunable offset) ? no. unfortunately. it sucks - i know - but after having written window managers for 10 years... i've made a lot of mistakes trying to be too funky and smart - and basically screwing up a lot of dumb client apps. sucks :( > Raphael > > > > ------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. Do you grep through log files > for problems? Stop! Download the new AJAX search engine that makes > searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! > http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click > _______________________________________________ > enlightenment-users mailing list > enlightenment-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/enlightenment-users > -- ------------- Codito, ergo sum - "I code, therefore I am" -------------- The Rasterman (Carsten Haitzler) [EMAIL PROTECTED] 裸好多 Tokyo, Japan (東京 日本) ------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click _______________________________________________ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
