Broadly agreeing with Dave here.

The point of an IETF working group is not to do design, research or
development. If you want to get the best spec possible there are much better
forums. For example six people working together in a closed room.

The alpha and omega of standards work is to establish a support constituency
for deployment. It is easy to develop a great spec, much harder to get
people to adopt it. 

Measured by the standard of getting agreement to deploy the DIX protocol the
BOF did not make much progress. That is the wrong measure to use. Protocols
are a dime a dozen here. 

Measured against the standard of focusing attention on a particular set of
use cases and the longstanding failure to get Internet wide single sign on
to deploy the BOF was a considerable success. That is the hard part -
getting people to believe that the area 1) is important enough to work on 2)
is capable of being solved.

With so many Identity 2.0 protocols on offer some form of requirements
definition / bake off was inevitable. The winner of the bake off is going to
be the protocol that demonstrates it best meets the proposal and requires
the least new mechanism.

There are some new use cases here that are not addressed in SAML. The
dispute is whether an extension to address them within the SAML framework is
or is not possible given that a possible requirement is 'do not use XML
angle brackets'.

The best way to win such bake offs is to align with an existing constituency
that has a different deployment base, provided of course this can be done
without unacceptable cost (i.e. a spec that looks like a dog's breakfast).

Attachment: smime.p7s
Description: S/MIME cryptographic signature

_______________________________________________
dix mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/dix

Reply via email to