Casefolding Sigma (was: Re: IDNAbis Preprocessing Draft)

Gervase Markham gerv at
Wed Jan 23 16:46:25 CET 2008

Michael Everson wrote:
> The idea that Greek users should not be allowed to use final sigma is 
> shocking to me. There must be a technical solution found for them.

As always when I jump into this debate, I risk making a fool of myself, but:

Given that this is a unique or near-unique case, can't you push the 
problem upstairs to the application level? Have IDNA only permit 
non-final sigma, even in the final position, and make applications 
translate a typed final sigma to the non-final form before doing a 
lookup, and back again for display as necessary.

(Note that I'm not promising anything on behalf of the Mozilla project 
here :-)


More information about the Idna-update mailing list