Re: Comments on Crispin draft

[Kent Crispin <kent@songbird.com>]

Hank Nussbacher allegedly said:
> 
> --=====================_849533274==_
> Content-Type: text/plain; charset="us-ascii"
> 
> 
> --=====================_849533274==_
> Content-Type: text/plain; charset="us-ascii"
> Content-Disposition: attachment; filename="CRISP2.TXT"
> 
> >     Technical Support for Shared Top Level Domain Registries
> >
> >                 <draft-iahc-stldtech-crispin-00.txt>
> >
> >    RESERVE reserves the domain name for a short period of time, and
> >    returns an error if the name is already in use, either as an
> >    existing operational domain, or through a prior reservation.  A
> >    "reserved" name is not functional, in the sense that no IP address
> >    or other information can be retrieved for it.
> 
> How long?  A day? 

I was thinking more in terms of an hour or less.

> A week?  What is to stop the registry from
> automating a process that resubmits the reserve every expiry period?

Yes, such an abuse would be possible, but it would be quite 
noticable by other competing registries, and possibly cause a loss of 
registry license.

There is nothing in the mechanism that pervents domain name hording, 
because I believe that problem cannot be solved at the mechanism 
level.

The problem of an unethical registry hoarding names is different than
the problem of some arbitrary person hoarding names, of course, and 
more serious.  I would expect that clear evidence of such would be 
grounds for losing a registry license.

> I remember back in my mainframe days, I used a system that kicked
> idle users off the system if they were idle for 60 minutes.  So the
> hackers just put together a program that made the user look active,
> thereby causing more of a memory and CPU load on the system, when in
> fact the interest was to reclaim memory which was in short demand at
> the time.

Been there, still there :-)

> Perhaps limiting the number of reserves a registry can
> have open at any one time and perhaps a second limit on the number
> of reserves a registry can issue per month would solve the problem.

Possibly, though I'm not convinced that in practice it would be a
problem.  The hoarding we see now in .com is at a different level
(with paid for, "confirmed" names, if you will).  And it might be
quite reasonable for a registry to renew a "reserve", if the customer
is delayed for some reason.  As I have mentioned before, name
conflicts in real life must be very rare. 

> 
> >    A CONFIRM is done by updating the domain with a longer timeout (a
> >    year or more -- basically a time reflecting the amount of time
> >    that has been paid for), adding NS records for the domains
> >    nameserver, and potentially other RRs.  After the longer timeout
> >    expires the domain is eligible to be deleted from the primary
> >    nameserver.  It is the responsibility of the end customer and the
> >    registry to renew the name.
> 
> You should make mention of grandfathered .com DNs, which were
> registered years ago and don't have to pay nor be renewed.  A very
> large TTL should be used here.

I didn't know such things existed, to tell you the truth -- I thought 
NSI was making everybody pay...

-- 
Kent Crispin				"No reason to get excited",
kent@songbird.com,kc@llnl.gov		the thief he kindly spoke...
PGP fingerprint:   5A 16 DA 04 31 33 40 1E  87 DA 29 02 97 A3 46 2F