New version: draft-ietf-idna-tables-01.txt

Andrew Sullivan ajs at
Wed May 7 15:54:40 CEST 2008

On Tue, May 06, 2008 at 10:37:36PM -0400, Vint Cerf wrote:

> Can we also hear from others on this please?

My (very strong) opinion is that, if we believe that moving from
DISALLOWED to PVALID is a big deal, then it is really critical that we
always err on the side of not putting things in DISALLOWED in case
there is any doubt.  If we can come up with some use case where
someone really might want to use a code point in some language, then
there is no justification for putting that code point in DISALLOWED
unless it actually breaks the Internationalized LDH model.  

I haven't seen an argument yet that the historic scripts break that
model, so I don't think they should be DISALLOWED. 

It might be that in the target BCP document that's been floated,
advice to zone operators could include such remarks as, "Since some
PVALID code points are only valid in contexts, you need to be able to
evaluate context for any code point you would publish in the DNS."
(This is approximately equivalent to John's remark to the effect that
you shouldn't accept anything you don't understand.)  This area is a
good place for registry restrictions, in my opinion.


Andrew Sullivan
ajs at
+1 503 667 4564 x104

More information about the Idna-update mailing list