On Wed, Feb 23, 2000 at 12:37:16PM -0800, Jim Lynch wrote:
> I really don't understand how this key combination is generic. Branden,
> originally you said something like X offers mechanism, not policy, and
> this statement has been made before. I would add that it also does not
> specify UI.
That's what "mechanism, not policy" means.
> Please correct me if I'm wrong in the following statements, and refer
> me to some reading material in any case:
>
> - No key combination is special to an X server, none at all. (not
> true if for example ctrl-alt-backspace is special to the server)
True in general, with special exceptions. The one you mentioned, plus
CTRL-ALT-Fn for VT switching. Both these special sequences can be disabled
(and thus passed to clients) with appropriate parameters supplied in
XF86COnfig.
> - Key combinations that are not recognized by the frontmost client
> are always propigated rearward, (visual heirarchy) eventually reaching
> the background screen, ("object" heirarchy) AKA the x server itself.
I think this actually depends on proper client programming. A poorly coded
client can, I think, catch events it isn't going to use instead of passing
them down the list.
> - The specific combination ctrl-shift-1 (isn't this the same as ctrl-! ?)
> is not handled specially by the x server; therefore, in order for it
> to be handled specially at all, a client would have to handle it in some
> specific way, and in addition, the key combination would have to (somehow)
> reach the specific client, being passed to it if necessary by some other
> client.
Yes.
--
G. Branden Robinson | Religion is something left over from the
Debian GNU/Linux | infancy of our intelligence; it will
branden@ecn.purdue.edu | fade away as we adopt reason and science
roger.ecn.purdue.edu/~branden/ | as our guidelines. -- Bertrand Russell
Attachment:
pgpprS6M6Molg.pgp
Description: PGP signature