Re: Re: stability

--On Tuesday, March 15, 2005 5:53 PM -0800 Erik van der Poel=20
<[email protected]> wrote:

> Martin v. L=F6wis wrote:
>> What is much more relevant is how further constraints in the
>> registry (beyond those imposed by IDNA) get implemented. Only
>> when that is sufficiently settled and deployed, considering
>> *updates* to IDNA should start.
>
> I disagree. The IETF should not wait for any of the registries
> to do anything before publishing new drafts or RFCs. The
> registries are not the only other players here. We have
> application developers and zone administrators depending on
> our work too.

Yes.

And let me add one observation.  At the moment, "the registries" =

(remember that there are around 275 of them) are going in every=20
direction possible.  ICANN's guidance, based to some extent on=20
the advice that IESG gave them, is so vague as to permit almost=20
any interpretation, clearly doesn't work for some important=20
cases (encouraging even more interpretations), and is being=20
outright ignored by some ccTLDs.   If we want the registries to=20
do something, we are going to need to provide some more specific =

guidance and we are going to need to document why it is a good=20
idea.  The same observation applies to application writers:=20
unless we actually change the standard (the algorithms or the=20
tables), all we are doing is offering advice.  That advice will=20
probably be taken if it is clear and well-reasoned and the=20
reasoning is well- explained justified.  And, in both cases, it=20
will be easier to persuade people to make changes before they=20
have an installed base that has to be forced to adjust to new=20
ways of doing things.  The more deployment occurs before we make =

suggestions or change things, the more likely whatever we do=20
will be ignored.

   john