Re: in support of axfr-clarify

[bert hubert <ahu@ds9a.nl>]

On Wed, Nov 27, 2002 at 04:58:02AM +0000, Paul Vixie wrote:
> axfr-clarify should go forward, for two reasons.
> 
> first, bind has done wrong.  early versions were only barely within rfc1035,
> forcing other implementors (notably microsoft) to bend to match field
> behaviour.  bind9 is entirely within rfc1035 regarding axfr, and current
> bind8 is reasonably close.  we need to clarify the spec so that all future
> implementations can know what the right thing is.

Indeed. PowerDNS had a very hard time implementing AXFR reasonably correctly
until we stumbled on the axfr-clarify draft. Without it, there basically is
*no* specification. RFC1035 has two paragraphs on it on flowery language
which try to pass themselves off as a specification. So 'being within
rfc1035' strikes me as a statement more like 'we spent heaps of time
interpreting the meaning of 1035 and we think we are within it'.

In any case, we need a clarification. On a related note, for interop
interest, PowerDNS just went open source, initial drop is on
http://ds9a.nl/pdns

> second, i've seen comments here to the effect that a name node's ownership
> was zone-independent, and that's just plain wrong.  again, early bind got
> this horribly wrong and caused a lot of pain and confusion, and we need to
> retain what's been learned (which is foundational to rfc2181 and rfc2308),
> which is that a delegation point is owned by the child zone and the parent

How is this related to a proposed RFC about zone transfers? 

Regards,

bert

-- 
http://www.PowerDNS.com          Versatile DNS Software & Services
http://lartc.org           Linux Advanced Routing & Traffic Control HOWTO

--
to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.
archive: <http://ops.ietf.org/lists/namedroppers/>