[ga] "...no one -- absolutely no one -- really understands the behavior of DNS in the large..."

["Jim Fleming" <JimFleming@ameritech.net>]

http://www.icann-ncc.org/pipermail/discuss/2002-August/002766.html
kent@songbird.com kent@songbird.com 

"...no one -- absolutely no one -- really understands the behavior of DNS in the large..."

=========================================================================

People with DNS laboratories do....also, people who write DNS software understand how the DNS works...

1. Does ICANN have a DNS Laboratory ?

2. Does ICANN have any people who write DNS software ?

3. Do ICANN employees have offices and report to work each day ?

4. What do ICANN employees do each day ?


Jim Fleming
2002:[IPv4]:000X:03DB:...IPv8 is closer than you think...
http://www.iana.org/assignments/ipv4-address-space
http://www.ntia.doc.gov/ntiahome/domainname/130dftmail/unir.txt


http://www.icann-ncc.org/pipermail/discuss/2002-August/002766.html
[ncdnhc-discuss] Names Council agenda item request: discussionof wholesale price for names 
kent@songbird.com kent@songbird.com 
Wed, 28 Aug 2002 23:39:12 -0700 
--------------------------------------------------------------------------------

On Wed, Aug 28, 2002 at 09:32:04PM -0700, Dave Crocker wrote:
> At 11:26 PM 8/28/2002 -0400, Milton Mueller wrote:
> >Until we add more than 500 TLDs a year there is
> >no issue with BIND and DNS. None.
> 
> BIND, no.  DNS root access? Maybe no, maybe yes.  The effect of traffic 
> patterns on root access is a matter of continuing debate among the DNS 
> technical community, as recently as the past few weeks.

Yes, that has been a most interesting discussion.  It has reinforced my
belief that no one -- absolutely no one -- really understands the
behavior of DNS in the large.  One thing seems fairly clear -- the root
zone really is different than a TLD zone. 

In any case, it's moot: the issue isn't what the software will support;
the issue is the far more complex matter of the procedures of the
registry/registrar.  Verisign and most other registrars, for example,
have largely automated procedures, and those procedures are not perfect
-- there are many horror stories about lost domains, customer service
screwups, and the like.  The idea that the root zone should be run under
a similar model is entertaining, but it is simply never going to happen,
regardless of the most fervent libertarian fantasies.

This has nothing to do with ICANN, per se -- any agency with serious
pretentions of management of the root zone would inevitably be drawn
into the same position. 

Kent
=============================================================================


--
This message was passed to you via the ga@dnso.org list.
Send mail to majordomo@dnso.org to unsubscribe
("unsubscribe ga" in the body of the message).
Archives at http://www.dnso.org/archives.html