Keysym additions.

[Jim Gettys]

On Thu, 2006-09-21 at 16:04 -0700, Keith Packard wrote:
> On Thu, 2006-09-21 at 21:28 +0300, Daniel Stone wrote:
> 
> > Dude, you can't add 200 keysyms for languages.
> 
> Actually, there's no particular reason not to; keysym space is enormous,
> we added all of Unicode without even feeling it.
> 
> If you really have keys that say "Chinese (language)" on them, is there
> some reason not to report that correctly?
> 
> ISO639-2 defines somewhere around 420 languages; that gives us well
> defined names and reasonable semantics as well. No Klingon though; we
> may have to make something up for that.
> 

Of course, maybe it is time to adopt the ISO/DIS 639-3, instead, which
is in finalization. If we're going nuts, let's really go nuts.

See:
http://www.sil.org/iso639-3/default.asp

> > Could you not use XK_Mode_switch and have the language smarts in
> > userspace?
> 
> If you look at how the 'keysym' spec reads, it's interested in correctly
> and unambiguously defining the engravings on the top of the keys. While
> it doesn't go so far as to require every unique symbol to have a unique
> keysym (enter and return are 'the same' keysym), it does encourage us to
> report what the actual keys are rather than try to insist that
> applications 'guess' what the 'language_1' key is actually painted with.
> 

And we can have an additional keysym  on the keycode to give a hint it
is a language key.
                                     - Jim

-- 
Jim Gettys
One Laptop Per Child