ietf-openproxy
[Top] [All Lists]

Re: P work in new charter

2004-06-30 22:07:17


On Wed, 30 Jun 2004, The Purple Streak, Hilarie Orman wrote:


 The whole reason we want to have callout servers is to move services
 out to a separate server. Are we now trying to move some of the
 processing complexity back into the OPES processor in form of a highly
 programmable, compelx rules language that allows "programming" of actions?

 -Markus

Yes, I've always believed that some actions should be done on the
OPES processor because of the efficiency of the data locality.

I agree that there are use cases where services implemented internally
by the OPES processor make a lot of sense, at least for performance
reasons. The trivial example is a "drop this message" service.

Fortunately, as far as P language is concerned, there is no
significant difference that would affect our charter:

        if (message meets criteria)
            services.drop(message); // call "global/remote" service

versus

        if (message meets criteria)
            processor.services.drop(message); // call embedded service

We should probably discuss this later, but it may be possible for OPES
processor to register its own service in the global services array so
that the user will always write code in the first example and get
efficient execution if the processor has an embedded drop service.

Alex.


<Prev in Thread] Current Thread [Next in Thread>