Re: Taxonomy Term closure needed

From: Joe Touch (touch@ISI.EDU)
Date: Sat Sep 11 1999 - 10:42:44 MDT


Oskar Batuner wrote:
>
> I'm afraid the definition "response messages" is somewhat misleading.
> We are interested not in messages but in data entities (or resources)
> carried in these messages. Here you get this discussion if we should
> store error replies. Well, we should not store some success replies
> too, like 200 reply to POST or PUT.

Agreed - the response isn't a pure replay, as you indicated later,
either. So the cache:

        - stores responses to earlier requests
          (it has to be a response, not a resource or anything else.
           a response is all the cache sees at this point)

        - look for incoming requests for which it can use
          stored data to generate an appropriate response

                - (define appropriate however you like - it's
                   whatever the cache _wants_ to send in response
                   to the new request)

        - generate that response in deference to forwarding the
          request, if so indicated

Joe



This archive was generated by hypermail 2b29 : Thu Nov 18 2004 - 11:21:27 MST