Consensus Call on Latin Sharp S and Greek Final Sigma

[Vint Cerf]

Michel,

I have been interpreting the "no mapping" rules as suggesting that  
under at least IDNA2008, any mapping functions outside the protocol  
should avoid mapping PVALID chars into other PVALID chars. I think not  
mapping "undefined" chars into PVALID chars is also, arguably, a good  
rule.

vint


On Dec 1, 2009, at 1:31 PM, Michel SUIGNARD wrote:

> Maybe surprisingly, I would support option 1) PVALID
> It may have to do more with the case of the final sigma which is the
> forgotten case here. I think long term we are better with having them,
> even if the transition is painful, and I am confident that vendors  
> will
> create strategies to cope with it.
> Finally I don't understand discussion about disallowing mapping of
> PVALID or UNASSIGNED characters. If IDNA 2008 does not include  
> mapping,
> any discussions about it is out of scope, and any layers above it is
> free to do whatever it pleases before presenting it to the IDNA  
> protocol
> layer. That many of us think that these above layers should also  
> have a
> somewhat predictable behavior is again out of scope in this WG.
>
> Michel