Re: [Fwd: Re: [idn] Requirements I-D]

[James Seng <jseng@pobox.org.sg>]

Mark Davis wrote:
> I want to point out that the algorithm to case-fold, while not free, is pretty fast. The two issues are:
> 
> - using a large, fairly sparse, mapping table
> - sometimes mapping one character to multiples*

While the 1->N mapping algo is going to be a bit of challenge, algo isnt the
biggest problem here.

It is in what goes into the mapping table, which can get very troublesome for
a registrant if folding it not done properly. On the other hand, if we do it
too much, then it probably screw up the integrity of some languages.

For example, we can strike a point where we have enough folding to make life
easier for the registrant. But keep an option for registrant to do per domain
folding requirements not in the standard. 

Hmm, come to think about it, this may in fact solve the problem of
'scripts/languages' not handled by Unicode yet.

-James Seng