Re: Roadmap for the future allocation of UCS characters w/r

From: Martin J. Duerst (mduerst@ifi.unizh.ch)
Date: Fri Jan 10 1997 - 03:58:33 EST


On Thu, 9 Jan 1997, Alain LaBont/e'/ wrote:

> What an excellent idea! I applaude to it and am ready to support it...
> however doing this is as much of a monk's task than the nightmare I was
> talking about, with the burden of making it *in addition* for current
> characters, but if the IRG does the task, I am ready to gladly accept it
> immediately.

Well, it's a monks task, but IRG has already done it. Otherwise, they
wouldn't have been able to order the characters they way they are now.
So the main problem is to make the data available and to make it
"official" for those who need such stamping.
As for "chunks", I think the idea of leaving open spaces for future
allocations so that the characters stay in order is interesting, but
unworkable. A better thing to do is to assure that the additional
allocations get made in big chunks, i.e. not 200 for vertical extension
B, and then extensions C-Z likewise, but 5000-10000 for each extension.
That will help at least those that have to search a character manually
in the standard documents: The number of ranges they have to search
is small.
It should also be noted that while the Radical-and-stroke-count ordering
is widely accepted and useful, there exist many other orderings, some
of which need additional, context-dependent information. Also, for
some characters, the question of which part is the radical, and how
many strokes are used in the rest of the character, is answered
differently by different dictionaries. So the importance of the
ordering as used in the standards should not be overweighted; good
applications will provide various ways to search and sort CJKV text.

Regards, Martin.



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:33 EDT