Shared TLD Daily Digest, Aug 26, 1996

-> Ooops
     by Kent Crispin 
-> Re: DNS based proposal for shared tlds
     by johnl@iecc.com (John R Levine)
-> Re: DNS based proposal for shared tlds
     by Simon Higgs 
-> Re: DNS based proposal for shared tlds
     by johnl@iecc.com (John R Levine)


----------------------------------------------------------------------

Date: 25 Aug 1996 07:55:45 -0700
From: Kent Crispin 
Subject: Re: New Non-Shared TLD's Create More Monopolies

John

- -- totally different subject from shared TLDs --

It was 4 am, and the baby woke me up, so naturally I went into my
office, to do some work for a new customer, but I got distracted, and
started roaming around the net.  I was reviewing the shared TLD stuff,
and, on a whim, I happened to check out jlc.net.

It seems like you are already doing what I am starting out with
songbird -- except I run linux and you run freebsd, and I am
concentrating on the Castro Valley, California area, and you are in
New Hampshire :-)

Now it so happens that this customer would like his own domain name,
and I was thinking "Gee, I'm going to need another nameserver for the
NIC", and the far-out thought occured to me -- maybe jlc.net would be
interested in setting up a reciprocal arrangement with songbird.com
for name service as far as NIC domain registrations are concerned.
Now, you must have already set up such a thing, so there is no real
gain for you.  So no problem if it's too much trouble.  But I thought
it would be kind of interesting to set up such an arrangement with
someone way back east -- I was back in DC a few years ago, going to a
Cray User Group meeting, and my wife and I took a couple of weeks and
went to Boston and various other places, but we didn't spend any time
in New Hampshire.

Anyway, just a curious thought.  If it's too much trouble to set up
such an arrangement, perhaps you could tell me how you arrange your
nameservice for the InterNIC.

- --
Kent Crispin				"No reason to get excited",
kent@songbird.com,kc@llnl.gov		the thief he kindly spoke...
PGP fingerprint:   B6 04 CC 30 9E DE CD FE  6A 04 90 BB 26 77 4A 5E


----------------------------------------------------------------------

Date: 25 Aug 1996 08:07:32 -0700
From: Kent Crispin 
Subject: Ooops

Sorry guys -- I was sending a private message to John Lelsie early in
the morning, and accidentally sent it to the list instead.  I must
remember to get some sleep...

- --
Kent Crispin				"No reason to get excited",
kent@songbird.com,kc@llnl.gov		the thief he kindly spoke...
PGP fingerprint:   B6 04 CC 30 9E DE CD FE  6A 04 90 BB 26 77 4A 5E


----------------------------------------------------------------------

Date: 25 Aug 1996 09:14:14 -0700
From: johnl@iecc.com (John R Levine)
Subject: Re: DNS based proposal for shared tlds

[plausible sounding DNS update scenario omitted]

>> If you want to treat an incrementally updated DNS as the CDB, I still
>> suppose that's possible but I'd be a lot more comfortable with a
>> multi-user database with a track record.
>
>DNS has a pretty long track record.

DNS has a swell track record, and and does what it does just fine.  But
incrementally updated DNS has no track record at all and as far as I can
tell, exists only as a paper proposal or at best a lightly tested
prototype.  As way to distribute the combined domain info, DNS is clearly the
way to go, but as a way to merge incremental updates, it's a complete
unknown.

In BIND, the DNS server that everyone uses, you update a primary server's
zone info by editing a text source file that contains all of the zone's info,
then send a signal to the running copy of BIND which tells it to slurp up
the new zone file and completely replace the previous info for the zone.
Secondary servers do a clock-based slurp over the net, but it's still the
whole zone at a time.

This is why I'm proposing to divide the work between two known working kinds
of software: a database for the CDB, since there's lots of databases that
have a track record handling multiple updates from multiple sources.  Then
once a day (or whatever) it dumps out the entire zone info as a zone file to
be slurped up by BIND for DNS distribution.

Like I said in previous messages, maybe someday incrementally updated DNS
will have enough of a track record that it'll make sense to use it as the
CDB.  But now, you'd be nuts to depend on it.


- --
John R. Levine, IECC, POB 640 Trumansburg NY 14886 +1 607 387 6869
johnl@iecc.com "Space aliens are stealing American jobs." - Stanford econ prof


----------------------------------------------------------------------

Date: 25 Aug 1996 12:56:00 -0700
From: Simon Higgs 
Subject: Re: DNS based proposal for shared tlds

At 12:15 PM -0400 8/25/96, John R Levine wrote:

>In BIND, the DNS server that everyone uses, you update a primary server's
>zone info by editing a text source file that contains all of the zone's info,
>then send a signal to the running copy of BIND which tells it to slurp up
>the new zone file and completely replace the previous info for the zone.
>Secondary servers do a clock-based slurp over the net, but it's still the
>whole zone at a time.
>
>This is why I'm proposing to divide the work between two known working kinds
>of software: a database for the CDB, since there's lots of databases that
>have a track record handling multiple updates from multiple sources.  Then
>once a day (or whatever) it dumps out the entire zone info as a zone file to
>be slurped up by BIND for DNS distribution.
>

I like the idea. It would be able to handle the distribution of a master
whois as well. The trick is to make the compiled zone file (or whatever
other data is being distributed) as dynamic as possible. I don't see why a
zone transfer request couldn't get it's info from the database instead of
pre-compiled file. This way if anyone updated at 6 am or 6pm, it has the
latest info. Otherwise you could get out of sync by almost 2 update cycles
(like buying something between the time the credit card cycle closes and
when the bill is printed). If there were conflicts, it would be sorted out
by the next cycle.


_____S_i_m_o_n___H_i_g_g_s_________________H_i_g_g_s___A_m_e_r_i_c_a_____
... "I'm fine - it's the others" ......... President/CEO ................
_____e-mail: simon@higgs.com _____________ http://www.higgs.com/ ________
... http://ds.internic.net/internet-drafts/draft-higgs-tld-cat-02.txt ...




----------------------------------------------------------------------

Date: 25 Aug 1996 18:13:38 -0700
From: johnl@iecc.com (John R Levine)
Subject: Re: DNS based proposal for shared tlds

>>This is why I'm proposing to divide the work between two known working kinds
>>of software: a database for the CDB, since there's lots of databases that
>>have a track record handling multiple updates from multiple sources [and
>>BIND to serve DNS].

>I like the idea.

Thanks!

>I don't see why a zone transfer request couldn't get it's info from the
>database instead of pre-compiled file.

That's a clever idea, but you could get almost the identical effect with
much less work by doing the database dump/BIND slurp cycle on the primary
server every hour or two.

- --
John R. Levine, IECC, POB 640 Trumansburg NY 14886 +1 607 387 6869
johnl@iecc.com "Space aliens are stealing American jobs." - Stanford econ prof