Re: [MEXT] new charter proposal

[[e-mail] (Arnaud Ebalard)]

Hi,

<[e-mail]> writes:

> 1. Proposed rephrasing for the following paragraph:
>
>>The MEXT WG will also explore experimental alternative security
>>mechanisms. The security >mechanism specified in the existing standard
>>track RFCs (RFC3775bis, RFC4877) remains the >mandatory to implement
>>mechanism that guarantees interoperability between different
>>>implementations. The MEXT WG is chartered to deliver one or more
>>experimental alternative >mechanisms. All the alternative solutions
>>will be published as experimental RFCs. 
>
> The MEXT WG will work on alternative (to IPsec/IKEv2) security
> solutions. These alternative security solution(s) will be considered
> as experimental to begin with and evaluated for standards track at a
> latertime depending on implementations, adoption and usage.  

I think the first version was ok, i.e. I don't see the point in adding
that to the charter now but if others agree ...

> The security mechanism specified in the existing standards track RFCs
> (RFC3775bis, RFC4877) remains the mandatory to implement protocols in
> order to ensure interoperability. The MEXT WG will deliver one or more
> alternative security mechanisms. 

It is already stated in the "Goals and Milestones" section of the
charter ("Aug 2011 Submit I-Ds on alternative security mechanisms to the
IESG for publication as experimental") but this is fine with me.
 
> 2. The milestone of Aug 2011 is okay. But there could be proposals
> that are already quite mature and could be taken up by the WG now and
> not have to wait until mid-next year. 

Just a thought: regarding draft-korhonen-mext-mip6-altsec-05 you are
pushing, I think it would be interesting to get some thoughts (about the
way it reuses TLS and elements of IPsec) from people of ipsecme and tls
WG. The TLS renegotiation issue (corrected by RFC 5746) has shown us it
is not trivial to get things right, even when eveyrthing looks ok. I
think hld-sec [1] is also an another good example.

To be honest, I don't care about pushing or not the document as a WG
item. AFAICT, at the moment, it is the only alternative security
solution propose: you have no competition, so there is no rush. IMHO,
the important thing is to get *technical* reviews from security people,
familiar with the security protocols at hand. And that part takes time.

Cheers,

a+

[1]: http://tools.ietf.org/html/draft-ebalard-mext-hld-security-00