Re: [mif] DNS selection with HE-MIF

[Brian E Carpenter <[e-mail]>]

On 03/02/2013 18:54, Ted Lemon wrote:
> On Feb 3, 2013, at 12:12 PM, Keith Moore <[e-mail]> wrote:
>> My proposal is to document the problems as best you can, and make concrete suggestions for areas in which the Internet architecture might be tweaked, but don't try to resolve the conflicts within MIF.   Instead, ask IAB to look at the matter and provide IESG with advice as to how to proceed.
> 
> This isn't bad advice.   It's worth noting that at least one member of the IAB has in fact been participating in the discussion already, however, so it's been my ongoing assumption that this issue is already on their radar, and that if we were doing the wrong thing, we would have heard something about it already.   I say that not to express resistance to your proposal, but simply to observe that I don't think we are _quite_ as far off the rails as you are suggesting.

Although it isn't about this precise problem, I have been banging on
about referrals for several years now (the most recent attempt being at
http://tools.ietf.org/html/draft-carpenter-referral-ps-02).

I mention this because despite a BOF at IETF76 and at least two IAB members
having showed interest in the problem, no traction has been gained whatever.
I think there's a strong temptation to put this whole topic on the "too hard"
heap, so we (collectively) continue to fiddle with the edges of it,
e.g. expecting DNS server selection to solve the problem.

   Brian