At 14:40 +0100 12/21/06, Stephane Bortzmeyer wrote:
On Wed, Dec 20, 2006 at 10:35:22AM -0500,
Edward Lewis <[EMAIL PROTECTED]> wrote
a message of 104 lines which said:
This effort could be quite large or quite constrained. I'm
indifferent in how large a scope is desired, but whatever scope is
desired, it has to be made really clear.
Giving your wishlist, I believe you are more a supporter of a large
scope? Because your list is long and it seems that some items (such as
health reporting) may require a strong coupling with the nameserver.
I didn't mean it as a wish list. What was my motivation?
Well, the one way to defeat a case of creeping features is to start
with a fully featured spec. 'Course, that'll make it too heavy to
ever get going. What a dilemma.
What I like to do is imagine the fully-fledged, impossible to attain
system first and then cut back from there. It's find to declare
parts of the problem space out of scope. But it's also helpful to
know what you've cut out in case it can be accommodated at a later
date.
My personal inclination goes toward a constrained scope. A protocol
more like ISC's metazones or PowerDNS' supermaster. I agree that
"views" are may be outside of a really constrained scope, I'll suggest
solution for them. But, in general, I would prefer to stick with zone
management (provisioning, configuring, reloading), not general
nameserver management.
It's okay to pick a small goal set, but then the solution ought to be
built knowing it is a partial solution (or a full solution to a
partial problem).
BTW, speaking of nameserver management, what's the opinion of the WG
on RFC 3197?
It was needed to cap off an idea that never flew, a procedural need
to stamp "HISTORIC" on the DNS MIBs. (Oddly, we moved more/as many
DNS Proposed Standards to Historic than to Draft Standards since,
well, at least the turn of the century plus some of the last.)
--
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Edward Lewis +1-571-434-5468
NeuStar
Dessert - aka Service Pack 1 for lunch.
_______________________________________________
DNSOP mailing list
DNSOP@ietf.org
https://www1.ietf.org/mailman/listinfo/dnsop
