>>but, at the risk of looking academic, what is the point of basing the interop communication on a protocol that is going to be contaminated?
>
>contaminated? In wich way? I don't see your point.

Use of HTTP headers to carry information that should go in the SOAP envelope (cf. Michel's first article on this issue).

>Michael have show a way to have the auth withouth a stateless, and if SOAP3 allows to this in a better way, this is welcome.

I wouldn't insist in SOAP3 designation. There is not such a thing, even in MS terminology.

Is this a better way? I don't think so. If we do not aim at interoperability, why would we use SOAP in the first place?