It's not really independent of the protocol. There's an
assumption in iCAP that the information is related to the
transit protocol. Here, we have environmental information
not necessarily tied to any proxy protocol processing
Hm, maybe I'm missing something, but I still cannot see how that
depends on the callout protocol.
For example, I could define a rule at any processing point, which says
that messages from a certain user have to be processed by a certain
service plug-in every thirty minutes. I could then use an appropriate
callout protocol to forward this message to a callout server. But I
cannot see why the callout protocol has to know anything about such a
rule, or why there has to be anything in the callout protocol to
enable such a rule.
Of course, the rule language would have to support specification of
such a time-based rule. Based on a discussion at the NY Workshop, for
eample, we already added system time information to the set of
possible conditions (i.e. environmental information). The same could
be done for above example, without having to change anything at any
callout protocol. Am I missing something?
Per user state is definitely possible and scalable in OPES
Good to know :)