Hi M. Bagnulo,
I read draft-ietf-shim6-locator-pair-selection-04, and I feel that it
overlaps with on-going work on address selection in a critical way.
Also, SCTP, MPTCP and MIF would need something like this as well, and it
doesn't make sense to define it separately for everything.
Some months ago, I proposed creating an API for performing address
selection that could be called from anywhere for any purpose:
http://www.ietf.org/mail-archive/web/ipv6/current/msg10357.html
http://www.ietf.org/mail-archive/web/ipv6/current/msg10367.html
I think such an API should benefit traditional use cases, multi-homing
use cases and multi-pathing use cases. Traditional TCP would call it
just once upon connect(), but something like SCTP or SHIM6 would keep
calling it a lot more often.
The special needs of the different kinds of users could be dealt with by
defining constraints and options usable in that API.
SECURITY ISSUES
I think the most compelling reasons to have a centralized API used by
all parties are detailed in sections 3, 2.3 and 2.7 of RFC5221. If all
working groups went ahead and defined their own stuff, getting those
kinds of issues fixed in all of them would be a hassle. But if they all
used the same API, the issues could be dealt with inside that API.
I'm not saying that you should cease and desist, or anything. I would
rather want your opinions and maybe even cooperation in creating a new
API. Rémi Denis-Courmont already also almost volunteered O:-)
(P.S. I'm not on the shim6 list, so please forward this message there,
if you feel it should be discussed there as well.)
--
Aleksi Suhonen
Department of Communications Engineering
Tampere University of Technology
--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
