On Wed, 4 Apr 2001 [EMAIL PROTECTED] wrote:
> So, we have some broken technology. One way to fix it is to add a
> presentation form to Unicode. But isn't there another fix -- to build into
> X the ability to map a key to a sequence of characters (which surely would
> be useful in other situations)? Of course, adding a presentation form in
> Unicode is the easier thing to do. I wonder if it's the right thing to do.
The problem is that it's among the few things on key caps that do not have
a character assigned yet. We will push that in X and other places that are
broken, and I'm sure we will have a hard time there, simply because
changing lots and lots of things only to support a key on petit Iranians'
keyboards may not be economically correct...
Please note that the main problem is not broken technology, but round-trip
with legacy charsets.
> Would you propose that it have a canonical or compatibility
> decomposition? If so, I believe that would be problematic with respect
> to stability of normalization forms.
We are proposing a compatiblity decomposition. Why would this cause
problems?
--roozbeh
