Charter changes and a possible new direction

[Andrew Sullivan]

Dear colleagues,

On Tue, Jan 13, 2009 at 01:39:15PM -0800, Paul Hoffman wrote:
> 
> That's not at all true. Unicode has been updated many times since
> 2003, and there has been no pressing need to update IDNA for each
> one.

The above seems to me to be a critical premise for this discussion,
and I'd like to understand the claims and counter-claims better.

I have been very much persuaded in the past by the argument that tying
IDNA2003 tightly to a particular version of Unicode is bad, because
applications using the system are unlikely to try to find out what
version of Unicode they're using (in the event that they even can find
that out).  It seemed obvious to me that this state of affairs could
lead to very bad effects.  The discussion of bad effects is in RFC
4690, section 3, I think.

The more I think about it, however, the more I think that the answer
to whether Unicode version tying is bad is entirely a matter of
engineering trade-offs.  Now the question is whether the actual
trade-off we appear to be trying to make is worth it.  I used to be in
the "obviously yes" camp, and now I'm in the "I dunno" camp.

So, I have two questions:

1.  Just how bad is it to tie IDNA to a particular version of
Unicode, and why?  (Ok, maybe this is two questions.)

2.  Given that our current plan is "protocol action for any change",
how are the current IDNA drafts not themselves tied to a particular
version of Unicode?  (Do I misunderstand that current plan?)

Best regards,

Andrew
-- 
Andrew Sullivan
ajs at shinkuro.com
Shinkuro, Inc.