Objection to draft-klensin-idna-5892upd-unicode70

Andrew Sullivan ajs at anvilwalrusden.com
Wed Dec 31 16:55:57 CET 2014


Dear colleagues,

On Mon, Aug 11, 2014 at 06:37:20PM +0000, Shawn Steele wrote:
> I think that's the problem in the opposite direction.  IDN folks are causing 'drama' about a non-issue of a Unicode character without participating in Unicode where such things are discussed.
> 

I'm not sure where to discuss this draft, so I'm doing it here for the
time being.  But it might be we need to take this to precis@ or the
apps area discuss list or something, because the particulars in the
current draft have wide implications.

The current draft-klensin-idna-5892upd-unicode70 (-02) offers four
options to be considered.  In my opinion, these in fact reduce to two,
because options 3.1 and 3.2 will automatically result in special local
mappings in some locales, and that effectively means "normalization
form IETF" (option 3.4), only perhaps less interoperable.  I don't
think that's going to fly, because it amounts to, "Upgrade every
computer that might get Arabic-writing-system input," and that seems
like a flag day.

Therefore, it seems to me that option 3.3 ("warn") is the only option
open to us.  But that option ought to boil down to, "In a global
context, the Hamza Above is likely to be at least dangerous, and its
use is not recommended."  Note that a _global_ context, so it would be
a recommendation for the root zone and so on.

Recall once again that the DNS does not need to be able to write
"words", and some pretty big classes of perfectly acceptable word are
already disallowed in the root (anything with an apostrophe, for
instance, which automatically disqualifies lots of words in English).
This is strictly speaking policy, not protocol, but it has been baked
in so many places that it is quasi-protocol anyway.

More generally, this is a problem for identifiers, and I suspect that
the advice for identifiers needs to be made clearer for the future
cases (that discussion probably does need to go to precis@).

Best regards,

A

-- 
Andrew Sullivan
ajs at anvilwalrusden.com


More information about the Idna-update mailing list