Re: [wg-c] straw vote -- question one results & call for votes on remaining questions

[John Charles Broomfield <jbroom@manta.outremer.com>]

> On 19 August 1999, John Charles Broomfield <jbroom@manta.outremer.com> wrote:
> >
> >We all seem to agree that the wording of the option 1 & 2 is not something
> >that any of us like. Propose some way of starting to add TLDs where it can
> >be done in a slow fashion so that the effect can be observed, and that
> >nobody suddenly starts waving arms around saying "it's not being done as per
> >the schedule". One thing I think we've ALL learnt, is that despite Internet
> >going 10 times faster than any of us had expected, adding TLDs seems to find
> >it's way to add delays... Pre-adhering to any time-table (one per month for
> >example) would seem to be foolish at least in the light of what we've seen
> >over the past few years. You don't like the word "stop"... dunno how to
> 
> Wait...there've been stability issues with adding new TLDs to the root
> over the past few years?  When did this occur?

Lots of them. Or do you think that all ccTLDs are in a stable and uniform
situation? Many of them are in a mess. ".tm" is on hold... No registrations
are in ".re", ".ht" was thrown around for a while, questions were placed
about the Libian TLD... etc etc etc...
Stability is not just the question of what is in the root-zone, but a lot
more than that. If any of what has happened to those ccTLDs I have just
mentioned had happened to com/net/org/edu/mil/arpa... there would have been
outrage.
We are looking at adding a system that has to be workable, not something
that is delegating responsability and authority for everything and then
looking the other way with the excuse of "it's not our problem anymore".
Will registries work in a cooperative way with registrars, will the
registry<->registrar software work, will coordination on the zone files of
those TLDs be done in a timely manner? (as any ccTLD zone contact can tell
you, getting an update to the nameservers of the ccTLD that you are contact
for is a royal pain in the butt...). Would any of this give merit to stop
everything? Probably not, but it mike slow down the idea of suddenly adding
another 2000 in one go. The reason you do a test is because you forsee
different situations that you're not entirely sure of what will happen (if
you were 100% sure, you wouldn't need to test, would you?).

Yours, John Broomfield.