U-labels, NFC, and symmetry

Patrik Fältström patrik at frobbit.se
Fri Apr 15 19:27:40 CEST 2011

On 15 apr 2011, at 18.36, Peter Saint-Andre wrote:

>> But, modulo a potential issue with characters newly-added to
>> Unicode, I still don't see the case for NFD: it certainly
>> doesn't make string comparisons any easier.
> Thanks for the input. I shan't post further in this thread until I've
> had a chance to think about things some more and check with some of the
> implementers in the XMPP community.

A suggestion, just because I know this discussion was ongoing in the XMPP community many many many moons ago, I think you have to find the *real* issues and reasons why NFD was chosen instead of NFC.

I would like John say that the difference between NFC and NFD is so small that the real cost will be to be different.

What you need are the arguments, that I do not remember, why someone suggested NFD in the first place.

Maybe it is worth it, else I would say, NFC is a better choice.

For Punycode/IDN, as John and others said, compactness was important. With different punycode etc, NFD could have worked as well.

But we picked NFC...


More information about the Idna-update mailing list