RM again

[jacques durand <jacques@savvion.com>]

Hi  POCers:

Just to keep the steam on in the lively RM debate,  :)
we will send out tomorrow for your review,
a set of reliability Use Cases we have worked on (Fujitsu - Savvion).

I would add my two cents here: having been involved  in developing a
prototype of RM service
as an extension of our ebXML handler prototype, I personnally believe
the specs are (and have been for
a while) stable and detailed enough.
Of course, the spec is not expected to be final yet, but which one is?
If we were to use only the
"Approved ebXML specs", as published by the ebXML Web site, then we
would only demo
the Requirement Specs.... Yet I believe we are about to use for  Tokyo
some other spec
material that is certainly "greener" and  less concrete (and more likely
to change) than the RM one.
But I think one of the roles of POC  is precisely  to "test" the specs -
and possibly point at some
enhancements along with shortcomings. (This does not mean implementing
"ahead" of
the spec if it is deficient.)

Also, it may help to better define the  "consensus" process at some
point.
When it comes to backing a demo feature, should we only  count the vote
of those who can/will commit
to implement it? I believe that we need indeed to have a general
consensus on the spec,
but it would be sufficient to get  just a few vendors to commit to
implement (say, one more than
required to demo the feature). Open to discussion...

Jacques Durand
Savvion