Re: Some issues with the IRI document [legacyNFC-06]

At 5:22 PM -0400 4/17/03, Martin Duerst wrote:
>I have changed variant B) in step 1) of section 3.1 from
>
>    If the IRI is in some digital representation
>    (e.g. an octet stream) in some non-Unicode encoding:
>    Convert the IRI to a sequence of characters from the UCS
>    normalized according to NFC.
>
>to
>
>    If the IRI is in some digital representation
>    (e.g. an octet stream) in some *known* non-Unicode encoding:
>    Convert the IRI to a sequence of characters from the UCS
>    normalized according to NFC.
>
>(emphasis only here).
>
>Do you think that this helps, or do you think that other changes
>are needed?

It certainly helps. That, plus the other changes you have made about 
spelling out UTF-8, makes it clear.

--Paul Hoffman, Director
--Internet Mail Consortium