Cameron Morris wrote on 6/1/2005, 5:14 PM:
> All day we've explored different ways of solving the same use cases.
> I'm certainly open to discuss them. However, I do not understand what is
> broken or wrong with the current proposal.
The note you're responding to now is the note that started the
discussions all day... Not sure why we're out of sequence here.
Anyway, the reasons why I think we should consider the alternative
proposal I made include:
a) It is a more general solution rather than one tied pretty
closely to the Liberty EP/PP services.
b) It supports the full, rich structure of the data that may
be present in the service (including any attributes on the
elements
c) it makes the data passed in the attribute match the format
that the consumer would normally be used to if they were
to access the service directly (e.g. we aren't making up
a new format for placing the data in the element)
d) it enables full use of XML capabilities to represent
the data that would be returned to the SP
The alternative proposal is sufficiently different that I think we
should fully discuss it before releasing the current spec as it's
a pretty significant change.
My $.02, of course.
Conor
