Re: [HOKEY] Inter-authenticator communication (was 2nd draft of new hokey charter)

[Tom Taylor <tom.taylor at rogers.com>]

As the person who presented the authenticator interaction proposal at the 
meeting I'd better answer this. At the meeting I heard Glen mutter "funny, there 
seemed to be interest on the list". That may be where the "definite interest" 
impression came from.

I agree with you, Katrin seems to have suggested it should be part of the 
architecture work. I suppose we can take the AD's advice on whether we need a 
separate work item just in case the architectural discussion leads to protocol work.

As for external SDOs, Simon Mizukovsky (hope I spelled that right) stated quite 
firmly at the meeting that other SDOs have already specified protocols in this 
area. I myself am unaware of such work (though 3GPP and 3GPP2 may have done the 
logical equivalent), but it seems reasonable to allow for its existence. One 
interesting point: Simon said the solutions were technology-specific. That 
suggests that there is work needed to develop a cross-technology optimization.

As the charts themselves indicate, that could be based on 802.21 signalling. If 
so, the protocol work obviously belongs to 802.21. However, are there scenarios 
where an IP-based inter-authenticator protocol would be useful?

I seemed to have reached a conclusion that we'd better do the architectural work 
first to find out.

Tom Taylor

Rafa Marin Lopez wrote:
> Tina TSOU wrote:
> > Hola Rafa,
> Hola Tina, thanks for your answer.
>
> First of all, any comments about my question regarding charter point 4).?
> > One person made strong representations against the work as presented 
> > in the
> > meeting. On the other hand, there was definite interest in this topic 
> > on the
> > list before the meeting.
> Secondly, I have just checked the e-mails in the ML but I failed to see 
> the interest as a work item. In fact, the only thing I have seen is one 
> comment in this e-mail:
>
> http://www.ietf.org/mail-archive/web/hokey/current/msg01623.html
>
> where it is said that this should not be a separate work item (If I 
> correctly understood).
>
> Am I missing something? Could you point me the e-mails I missed about 
> this interest?
>
> Btw, the charter text says:
>
> "Related work in external SDOs on authenticator interaction for 
> re-authentication may be taken into consideration."
>
> I would be interested in the related work in external SDOs, could you 
> point out some references?.
> > As to the second question: one would have to look at the scenarios on the
> > slides, but I think they are still valid.
> If they are still valid, then I don't completely understand the general 
> purpose/goal of this particular aspect. What are we trying to solve here?.
>
> Best Regards.
> >