[FusionDev] [Bug 545] Escape used as part of any key binding has no effect.

bugzilla at bugs.opencompositing.org bugzilla at bugs.opencompositing.org
Sat Apr 19 13:00:13 CEST 2008


http://bugs.opencompositing.org/show_bug.cgi?id=545


Mats Ahlgren <mats_a at mit.edu> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mats_a at mit.edu




--- Comment #5 from Mats Ahlgren <mats_a at mit.edu>  2008-04-19 11:00:12 ---
(In reply to comment #4)
> I want to be able to map [Ctrl]+[Alt]+[Esc] to xkill (like in KDE/kwin).

I was just about to report a new bug for this exact reason.

If one reserves the escape key to do an 'abort', from a usability standpoint
one should not let the user bind the key; a warning of "you can't use Esc in
Compiz, sorry" should be thrown.

In reply to comment #2)
> Escape is used as "abort" key for Compiz actions, so you can e.g. cancel window
> switching and have no focus change due to that.

It seems, in my humble opinion, that there is no great conflict here: If a mode
requests "abortability", then 1) pressing escape should trigger the abort and
2) pressing a modified-escape should trigger whatever binds to that -- it seems
to be an exclusive choice.

Also the reason for this design choice seems a bit awkward (no offense). I use
a decent number of plugins, but none of them requires an escape -- the switcher
doesn't, scale doesn't, deskop wall doesn't, etc. Granted I have seen some
plugins which put you into a "mode" (such as paint-fire), but hitting escape
has never worked. And in these cases, wouldn't the above paragraph hold? It
seems like an unmodified escape should be catchable in modes where
escapability-without-effect is desired--without sacrificing the ability to us
escape in keyboard shortcuts.


-- 
Configure bugmail: http://bugs.opencompositing.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.


More information about the Dev mailing list