Q2: What mapping function should be used in a revised IDNA2008specification?

[Michel SUIGNARD]

> From: Vint Cerf
> What characters should be mapped into what other characters in a  
> revised IDNA2008 specification?
>
> Can we describe succinctly and precisely what these mappings are? How?

> What should they be?

I would use Martin Duerst formulation, eliminating completely NFKC, just
adding back a much simpler function just doing <narrow>, <wide> folding
because of East Asian IMEs.
There is really no point in adding back all other compatibility mapping.
The main purpose of this mapping is to allow simple Case mapping. The
rest are confusing to most users and safety risks.

Michel