Re: [idn] proposed i18n naming rules

["Eric A. Hall" <ehall@ehsco.com>]

Eric Brunner-Williams in Portland Maine wrote:

> > maximum label length of 63 UCS character codes
> 
> q: are you confusing octets and characters?

No. The characters/codes are the canonical representation of the IDN, not
the characters which are encoded and encapsulated in a data stream. The
encoded values will have their own length restrictions when they are
encapsulated. EG, ACE will provide a maximum of 59 octets when the
four-octet prefix is subtracted from the maximum encapsulation length of
63 octets, but this is a secondary restriction apart from the canonical
representation restrictions.

> >      only those UCS characters from the "safe-set"
> >
> >      [version of] http://www.unicode.org/Public/UNIDATA/Scripts.txt
> 
> q: is there mechanism for extension? is it in the ietf, or in the utc?

Undefined in the rules. Got any ideas here?

> >      minimum label length of 2 UCS characters
> 
> q: is two really better than one?

Not a question I can answer, alas.

> >  U+002E (FULL STOP) is ONLY valid for use as a separator when
> >    the IHI is written out
> 
> i don't know what this means.

It means that "." is legal as a label separator, but not as label data.

-- 
Eric A. Hall                                        http://www.ehsco.com/
Internet Core Protocols          http://www.oreilly.com/catalog/coreprot/