I would prefer if it was clearer from the text that the definitions
were copied while the "Note" sections weren't. It might just
be indentation, e.g., indent this document's "note" sections
*less* than the quoted text rather than more.
Some of your basic terms and notes belong in the revision of RFC 2616 as
it goes to "Standard"; you might consider submitting them as
errata ( http://purl.org/NET/http-errata ).
> Note: There is widespread colloquial mis-use of the term "cache"
> to mean "caching proxy".
I think you can be less pejorative:
<<The term "cache" used alone often is meant as "caching proxy".>>
but I don't know how widespread this is, actually. Personally, most documents
I've seen are careful about this.
> Ed.Note;IAN: This is the actual definition from RFC2616, but now
> excludes consideration of a reduction in server load. Do we
> wish to comment on that, is is response time related to server
> load in such a way that the comment is unnecessary?
We should acknowledge there are other motivations for caching other than
reducing response time, including reducing server load and enabling
disconnected operation. But in HTTP
"A cache stores cacheable responses in order to reduce the response
time and network bandwidth consumption on future, equivalent
requests."
could become
"A cache stores cacheable responses; for example, this may be
done to reduce the response time and network bandwidth consumption
on future, equivalent requests."
It's not clear you want to elaborate all of the purposes of a cache
in the definition for a cache.
> 2.2 First order derivative terms
>
> Using the base terms above as a foundation, we construct the
> following terms
Usually "we" is meant as "the authors of this document", but the
constructed terms were constructed elsewhere. How about
"The following terms are defined in terms of the base terms above:"
> caching proxy
> A proxy with a cache, acting as a server to clients, and a client
> to servers
This definition is just missing from RFC 2616, since it's used there.
I'd suggest submitting this as an errata.
> 2.3 Second order derivatives
>
> The following terms further build on first order derivatives
>
> authoritative reference
> The owner of data; content production system; possibly an origin
> server; the overall master copy of the content, if any
Why is this second-order? The definition is a bit unsatisfactory
since "owner" and "data" aren't defined, nor is "content
production system". I'm not sure what you mean.
> content consumer
> The user or system that initiates requests of an origin server
> (which may in turn be handled by a proxy).
I'm not sure how this is distinguished from "user agent" or
"client". (Actually, the line between "user agent" and "client"
isn't described very well in RFC 2616.)
> browser
> A special instance of a user agent that acts as a content
> presentation device for content consumers.
My guess is that both of these should be present in the update to
RFC 2616, and that we should more carefully distinguish between
those constraints that apply to a 'user agent' and those that
apply to a 'browser'. HTTP/1.1 makes some requirements on how
the "user agent" interacts with the "user" in the case of
a "browser" that is inappropriate for other kinds of clients,
e.g.,
... user agents SHOULD present to the user the entity returned
with the response, since that entity is likely to include human-
readable information which will explain the unusual status.
Larry
-- http://www.parc.xerox.com/masinter
This archive was generated by hypermail 2b29 : Thu Nov 18 2004 - 11:21:27 MST