OFF TOPIC normalization

Simon Josefsson simon at josefsson.org
Tue Oct 14 13:49:29 CEST 2008


"Mark Davis" <mark at macchiato.com> writes:

> The UTC *did* what you say it should have done.
> It did not revise Unicode 3.2; it changed a *later* version of the Unicode
> Standard, *and* issued a corrigendum so that people could either still claim
> conformance to U3.2 as it was, or could claim conformance to U3.2 with the
> corrigendum. From http://www.unicode.org/versions/corrigenda.html - An
> implementation claiming conformance to any version of Unicode in that range
> may
>  also claim conformance to one or more of the corrigenda applicable to
> that version.

Right, that is the correct approach.

The problem I see is in modifying IDNA2003 implementations rather than
fixing the IDNA2003 specifications.

> Until and if IDNA2003 was revised, it referred to U3.2 without the
> corrigendum, and was unaffected by any later corrigendum.

Exactly.

> I agree with Martin that in an ideal world, IDNA2003 should have been had an
> erratum/corrigendum for that, but none was issued. And now, of course, this
> is completely moot, since the effects of the normalization changes from the
> UTC are stunningly small compared to the other changes from IDNA2003 to
> IDNA2008.

We agree.

The issue is small, but it needs to be considered in the IDNA2008
specifications because there never was a revision of IDNA2003 that
incorporated this particular fix.

/Simon


More information about the Idna-update mailing list