06FD and 06FE should be PVALID for Sindhi

[Paul Hoffman]

At 9:25 PM +0100 4/1/08, Michael Everson wrote:
>That is fine. The important thing is that they are PVALID; whether 
>that happens with a change from "So" to "Lo" is irrelevant so long 
>as there is an exceptions list.

It is probably quite relevant to the IETF, which updates its 
standards *much* more slowly than the Unicode Consortium. It is also 
possibly relevant to the Sindhi community, given that the miscasting 
in the Unicode Standard could affect other standards in the future.

If we (for a value of "we" that includes active Unicode Consortium 
people) want IDNA200x to rely heavily on the Unicode Standard for our 
decisions, then the Unicode Standard should be corrected wherever 
possible instead of asking all the protocols to make exceptions. If 
we want to view this more as a process where the Unicode Standard is 
taken as mostly-but-not-completely correct, then relying on other 
standards to have actively-maintained exception lists is reasonable. 
I prefer the former; others might want the latter. But we need to 
decide sooner rather than later which model is most appropriate. From 
the messages on the thread so far, these characters seem like 
excellent test cases.