Alex Rousskov wrote:
Is the concept of a "processing point" application-agnostic? If it is
application-specific, it should be removed from the
application-agnostic spec (draft-beck-opes-irml-03).
I think the concept of processing points is indeed application agnostic,
even though each application protocol may have different processing
points. This depends, for example, on whether or not an OPES processor
can implement a cache for the application protocol and on the
request/response scheme of the application protocol.
If we believe that every application protocol has some "processing
points", then the definition of HTTP-specific points should probably
be moved from draft-beck-opes-irml-03 to an HTTP binding for IRML. It
may also be more appropriate to migrate from numbers to tokens/strings
when describing application-specific points.
Yes, I think application-specific processing points should be defined in
the corresponding application protocol binding for IRML, but I still
think that it makes sense to define a general mechanism of how to
specify processing points in IRML rules in the IRML spec. Is this what
you had in mind as well?
-Andre