On Mon, 29 Jun 2009 14:40:53 +0300, Aleksi Suhonen <aleksi.suho...@tut.fi>
wrote:
> Thoughts? Comments? Or does it exist already and I'm just ignorant?

No. There is no standard API for this, and it is a major road block to
implementing dual-stack correctly in some scenarios.

The problem is simply that getaddrinfo() ties the method to obtain the
address data (typically DNS and local IP stack configuration), with the
algorithm. However, it is not as simple as creating a function that takes
two lists of addresses and returns a list of pairs of addresses. Most
commonly, the caller still wants to use the locally assigned addresses (or
a subset thereof) for the source addresses. I considered writing one spec
for that, but I got side-tracked with other things...

So... For plain RFC3484, the input data could simply be a chained-list of
struct addrinfo, and the output data another chained-list of struct
addrinfo. But I can only hope that RFC3484 update won't render this
inadequate :-/ Who knows what people will want to put into source address
selection algorithms in the future?

-- 
Rémi Denis-Courmont

--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------

Reply via email to