Re: Call for Votes (Re: glibc's getaddrinfo() sort order)

[Ian Jackson <ian@davenant.greenend.org.uk>]

Anthony Towns writes ("Re: Call for Votes (Re: glibc's getaddrinfo() sort order)"):
> AFAICS we should be making a definitive statement wrt both Rule 9 and
> IPv6 [...]

My proposal makes a definitive statement to our libc maintainer about
Rule 9 and IPv6: we disapprove of rule 9 but do not overrule the
maintainer.  So that part is clear, I think.


I assume you mean that we should make a definitive statement to the
IETF about Rule 9.  I don't think that the TC is going to be able to
get enough of an understanding of the situation in IPv6 to come to a
clear conclusion about what the standard out to be.

If we were to try to come up with some clear statement we'd really
need input from all of the IPv6 experts in the IETF.  Effectively,
we'd duplicate the work of the IETF IPv6 working groups.

Perhaps a clearer statement would be something like this:

  3. We recommend to the IETF that RFC3484 s6 rule 9 should be
     abolished for IPv4.  The IETF should publish a standards-track
     RFC documenting the existing de-facto standard DNS round robin
     and deprecating RFC3484 s6 rule 9 for IPv4.
  4. We recommend to the IETF that RFC3484 s6 rule 9, and other
     relevant parts of RFC3484, should be revisited by appropriate
     IETF working group(s), with a view to clearly defining
     operationally sound rules for source and destination address
     selection.  Specifically, consideration should be given to the
     changes to the IPv6 addressing architecture since the publication
     of RFC3484, and to the common implementation strategies for
     adding IPv6 capability to existing applications.


Steve Langasek writes ("Re: Call for Votes (Re: glibc's getaddrinfo() sort order)"):
> I also think that if we're going to make a recommendation to the IETF, it
> behooves us to provide a coherent rationale for that recommendation, and
> vote on that rationale.

I don't think this is a good approach but I'm willing to try it.

I think we would be better off doing what I've seen done in some
courts: we vote on the decision but each write our own rationale.

Otherwise we'll get bogged down in unnecessary arguments over the
rationale - and what if we can agree on the decision but there is no
rationale that will command a majority ?


Ian.