Re: Review Requested: SMTP IPv6 operational experience document

[itojun@itojun.org (Jun-ichiro itojun Hagino)]

> > >    draft-ietf-ngtrans-ipv6-smtp-requirement-08.txt

	thank you for numerous comments.  i will try to integrate those
	into the document and submit another i-d (under individual name
	instead of "ngtrans", as suggested by ADs).

> Now, a dual-stack client has two sorting parameters. Maybe it would be
> nice to specify in this note that this IPv6 preference may take
> precedence over the randomization specified above (so first all IPv6
> addresses are tried, then IPv4).
> 
> Currently, my own implementation lets IPv6 take precedence. I'll make
> this configurable soon (mostly because people may want to have it
> otherwise).

	from IPv6 transition POV IPv6 should take precedence.
	having knob would be good for early transition period, mainly due to
	the problems documented in
	draft-morishita-dnsop-misbehavior-against-aaaa-00.txt,
	but please make IPv6 to take precedence by default.

itojun