<<< Date Index >>>     <<< Thread Index >>>

.net: Thick vs. thin registry, WHOIS models



As you may have noticed, the GNSO's advice on .net is explicitly
undecided on whether the .net registry should change from the
current thin model (as known from .com) to the thick model (as in
.info).

With the thin model, registrant data remains with the registrar; the
registry only knows the mapping from a domain name to a registrar,
and knows about the name servers. WHOIS services operated by the
registry publish that mapping; the registrant's identity is then
published by the registrar.

With the thick model, registrant data is also kept by the registry,
and the registry's WHOIS services publish these data.

With the .org reassignment, a transition from a thin registry model
to a thick one took place.


There are tradeoffs between the two.

The thick model is, on its surface, better for data users who get
uniformly formatted data from a single source.  Of course, this is
nothing that couldn't be achieved by deploying properly standardized
software -- just think about the DNS itself, which is
decentralized.

From a registrant perspective, the thick model takes care of escrow
concerns -- the assumption here is that the registry keeps its data
more secure than "some registrar," so it becomes easier to deal with
business or physical failures of registrar businesses.

A centralized storage of registrant information also makes transfers
and registry-level transfer enforcement easier, which generally is a
good thing in my book.


On the other hand, a thick registry model implies transfers of
registrant data across the boundaries of jurisdictions with
different privacy regimes, and it takes away control over the
publication of WHOIS data from registrars who may be subject to
relevant national law in a more direct way, and who may be more
accessible to registrants attempting to enforce their privacy
rights. (Think registrar in EU, registry in US.)


Personally, I certainly prefer a thin WHOIS model, or at the very
least a thick model that quacks like current thin WHOIS, i.e.,
leaves the choice of data elements published by registries to the
registrar.

What do others think about this?  Should we give independent advice
to the ICANN board on this topic?  (If so, we would have to finish
this advice ASAP, since the RFP is supposed to start by 1 October
2004.)


References:

* .net procedure
  http://www.icann.org/announcements/announcement-29jun04.htm

* Announcement of GNSO's .net criteria
  http://icann.org/announcements/announcement-10aug04.htm
  

Regards,
-- 
Thomas Roessler  <roessler@xxxxxxxxxxxxxxxxxx>
At-Large Advisory Committee: http://alac.info/