On Sat, Apr 4, 2015 at 12:37 AM, Andrew Sullivan <a...@anvilwalrusden.com>
wrote:
> On Fri, Apr 03, 2015 at 02:23:52PM -0400, Casey Deccio wrote:
> > I've re-read the definitions of zone apex and delegation point in RFC
> 4033
> > (in conjunction with review of draft-hoffman-dns-terminology-02) and it
> > seems to me that they are not consistent with RFC 1034 notions.
>
> When I was working on this in the -00 with Paul, I came to believe
> that 1034 was not always totally careful about distinguishing theory
> and practice.
>
> To me, the "delegation point" is a logical question: where in the tree
> does a new authority take over?
>
> The "zone cut" is in fact a question of implementation: delegation
> points are implemented at zone cuts, which correspond to the point
> where the parent gives away its authority over the namespace.
> Therefore, _I_ think the zone cut is the place where the NS records
> are. But to me the delegation point is a logical construct, and as
> such represents a point in the tree. 4033's definition, however,
> ought to trump for the purposes of the terminology doc unless it's at
> odds with other docs, because the terminology doc is supposed to
> reflect known usage.
>
> I believe quite strongly, however, that the zone apex is in fact a
> logical point in the tree, and the zone origin is the implementation
> of that. I think that is just entailed by the way STD13 is written.
>
> I think errata would be incorrect: if this is a change, it's a
> clarification, not an erratum in any previous document.
Just to clarify, my original issue with the definition was not the location
of records, zone cut, or delegation point (per se), but more the
implication from the definition that there was more than one "node" or
(correspondingly) "name" in the namespace tree. For example:
Term used to describe the name at the parental side
of a zone cut. That is, the delegation point for "foo.example"
would be the foo.example node in the "example" zone (as opposed to
the zone apex of the "foo.example" zone).
To me, "...the name at the parental side of a zone cut" implies a
parenthetical "as opposed to the name somewhere else". Same with "...the
foo.example node in the 'example' zone".
The location of the zone cut is an interesting discussion, particularly in
the context of these definitions. In fact, that's why the text of my
"proposed fix" included that language, but I can see now that it derailed
things. For the purposes of this document the issue was the multiple
names/nodes discrepancy.
Regards,
Casey
_______________________________________________
DNSOP mailing list
DNSOP@ietf.org
https://www.ietf.org/mailman/listinfo/dnsop
