[registrars] Verisign change to operation of the .com DNS lookup service

[Ross Wm. Rader]

On 9/16/2003 9:32 AM Michael D. Palage noted that:

 > Listed below are some of the
 > questions that I have identified as relevant in my personal analysis, 
 > and
 > which I believe it would be in the registrar constituency's best
 > interest to address:

These are pretty loaded questions, but nonetheless, thanks for putting 
them forward - the answers are quite clear.

> (1) Does the use of wild-cards threaten the stability of the Internet, if so
> how (specific examples not generalization)?

Yes. All applications that require authoritative answers to A records 
are suddenly broken (score 1 for the spammers), there was no reasonable 
advance notice to the DNS community regarding something that can only be 
characterized as a major operational chance, there is little, if any, 
notion of "authoritative" being returned - the closest we get now is 
"maybe", not too mention that email is fundamentally broken - there can 
no longer be a reasonable expectation of delivery because messages being 
sent to non-existent domains aren't being retried and therefore not 
reported.

And in the interest of avoiding generalities, lets ignore the impact 
that this has on ICANN's credibility, operations support costs (with end 
users), the value of the UDRP and the hard-fought gains of the IP lobby 
in trying to control typo/cybersquatting.

> (2) Does the use of wild-cards adversely impact competition in the
> marketplace, if so how?

Microsoft, AOL, Opera, MelbourneIT etc. would be in a better position to 
answer this question than I could, but anytime multiple providers of a 
service are replaced with a single-source provider *solely* by virtue of 
that single-source provider having a contract to run a database and 
provide resolution for the names in that database, competition will be 
adversely impacted. This is the same argument that I put forward in the 
WLS debate and it is equally applicable today.

> (3) Can these stability/competition issues be addressed by best practice
> standards that would still permit the use of a wild card service by a
> registry operator?

That's not for me to establish, but in my opinion no. Until and unless 
such a best practices is proposed and adopted by the gTLD registry 
operators, then any such service should be immediately taken out of 
operation.

> (4) How is the use of wild-cards similar or different to the WLS service
> previously proposed by VeriSign?

It is an abuse of the single-source control of the registration database 
and gTLD resolution contract that each registry has.

> (5) What protocols does the use of wild-cards violate? Specifically, does
> the use of Wild Cards violate RFC 1034, 1035, and 2182 as VeriSign is
> required to comply with these RFCs in Appendix C of its Registry Agreement.

I haven't looked. I'm sure that someone more qualified to answer this 
will probably go on the record with comments at some point - like say 
Klensin or Cerf.

> (6) If the use of wild-cards threaten the stability of the Internet why did
> ICANN allow the incorporation of a wild card service into the .MUSEUM
> registry contract?

Why did the staff make an arbitrary decision without duly consulting the 
community? Perhaps thats a question the board could ask. Just because 
the decision was made doesn't make it a valid decision - its just the 
one that we got stuck with. Even if a valid consultation were to have 
occurred - I'm not sure that the feedback would have necessarily been 
representative; there aren't too many operators effected by decisions 
made by Musedoma.

> (7) If the use of the wild-cards threaten the stability of the Internet, why
> do several ccTLD operators currently utilize this feature?

Roughly 12 if I remember correctly - ac, cc, cx, mp, museum, nu, ph, pw, 
sh, tk, tm & ws. All high volume, all high impact if something screws 
up. With respect to the operators of these zones, they could all 
disappear and the DNS community would barely notice. They are irrelevant 
to this discussion. Decisions made in .com effect *every* DNS operator 
and every registrar globally.

> (8) If the use of wild cards threaten the stability of the Internet, should
> their prohibition be enforced across gTLD and ccTLD registries, if so how?

Yes, via contracts with the gTLD operators. What happens with the cc's 
can and should be sorted out between the operators and the local 
communities.



-- 


                        -rwr