Re: Design for Keyboard Shortcut Behaviour: Quasimode Command System

On Thu, May 27, 2004 at 12:31:35PM +1000, Karl O'Keeffe wrote:
> >Fortunately it doesn't have to be. I think this idea is nuts, but it's
> >also well-timed, because it validates some recent design decisions;
> >the YKB model can handle this just fine, without special application
> >support.
> 
> Does still still hold true?

Yes, of course. YKB forms a strong abstraction boundary between
applications and digital input devices.

> >Not a line of C code required; this can be implemented entirely in the
> >keymap. That's all you need, once widgets have been converted.
> >[...]
> >That's just an extended input method (and a fairly easy one to
> >implement, even if you do need a little ingenuity to compute the list
> >of available commands).
> 
> If this is still possible, do you think this is the best way to go 
> about things? I was just planning to implement QCS as a module. Whats 
> your take on this?

It's the only sane way, although you'd still need to load some
server-side code for the extended input method. Applications should be
completely oblivious to the existence of such a system.

-- 
  .''`.  ** Debian GNU/Linux ** | Andrew Suffield
 : :' :  http://www.debian.org/ |
 `. `'                          |
   `-             -><-          |

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFAtenBlpK98RSteX8RAv83AJ49IPqVvtThA7jIGirWcvyxtsbzQACePPNV
kqlrwhU35ztS3ZqiZx+0gng=
=pD8B
-----END PGP SIGNATURE-----