Re: [idn] prohibit simplified code points

["Adam M. Costello" <idn.amc+0@nicemice.net.RemoveThisWord>]

> Are you proposing that "mixed case notation" shoule be a mandatory
> feature of Punycode when used in IDNA ?

No.  The S->T mapping would be required in nameprep, but mixed-case
annotation would remain an optional feature of Punycode, and T->S
mapping would be an optional feature when displaying ACE names.  When
converting an ACE to Unicode for display, software that supports T->S
mapping could do it for the entire domain name (if the user requests
it) or for individual characters using mixed-case annotation (if it is
supported).

Software that doesn't support T->S mapping would simply display
the Unicode as-is, which means the user would see only traditional
characters (and simplified characters that have no one-to-one mapping).

Chinese users would tend to demand software support for T->S mapping.
Other users would tend not to care.

I think this proposal technically works, but I think it's probably too
much trouble for an incomplete solution (only one-to-one T-S matching is
handled).  A complete solution (for many languages, not just Chinese)
would have to be done at a higher layer.

AMC