Re: [Dime] rfc3588bis version number

[Sebastien Decugis <sdecugis at nict.go.jp>]

>
> I don't mind going with Glen's approach. It would work in our give
> case. Any other thoughts ?
Hi,

In my humble opinion, saying that an implementation MAY do A, but SHOULD
do B, when A and B are contradictory, is equivalent to saying nothing at
all. So basically you end up not specifying how TLS should be handled,
which is very not what I would expect from the spec...

My 2 cents
Sebastien.

-- 
Sebastien Decugis
Research fellow
Network Architecture Group
NICT (nict.go.jp)