Nominum GNS Server and BIND

[Erik Aronesty]

Jim,

How do you decide which features you put in the free software and which get into BIND?

For example: If someone made a version of BIND that was faster then GNS, or supported new features, would you be open to modifying BIND?

Verisign still uses some (perhaps modified) version BIND for the root servers, right? 

It seems like a bit of a conflict of interest for a company to be maintaining free software that hosts and service providers use - and to be also developing a patented version as well.

	- Erik


-----Original Message-----
From:	Jim Reid [SMTP:jim at rfc1035.com]
Sent:	Wednesday, February 28, 2001 3:50 AM
To:	Peter Rose
Cc:	comp-protocols-dns-bind at moderators.isc.org
Subject:	Re: Hosting DNS

>>>>> "Peter" == Peter Rose <peter.rose1 at ntlworld.com> writes:

    Peter> Er ...  isn't this a bit of a vote of no-confidence in BIND
    Peter> 9.1, considering Nomimum wrote it :-)

No. But we would say that, wouldn't we? :-) The initial plan was to
use the GNS infrastructure as a showcase for BIND9. But then an even
better server came along, so we decided to use that instead. There's
nothing wrong with BIND9, though its performance isn't yet good enough
for root servers that get thousands of queries a second. For everyone
else it should be just fine. The world would be a much better place if
people used BIND9 instead of BIND[48]. Compare the code quality.

    Peter> I was also curious why your secondary.com is not using BIND
    Peter> - do you guys know something the rest of us don't ?

Yes. The technology in the GNS server is far better, partly because it
doesn't have to carry around the legacy cruft and baroque features in
named.conf. We could tell you more about it, but we'd have to kill
you afterwards. :-) There are patent issues for instance.