Re: ISSUE VARY: Proposed wording
Henrik Frystyk Nielsen (frystyk@w3.org)
Tue, 08 Jul 1997 14:31:59 -0400
At 10:04 PM 7/7/97 +0200, Koen Holtman wrote:
>>An HTTP/1.1 server MUST include a Vary header field with any cachable
> ^^^
> why did you leave out `an appropriate'? I think that leaving this
> out makes the text worse. `an appropriate' signals that is the
> responsibility of server to ensure that the client does the right
> thing based on the vary header field contents.
To me, "appropriate" doesn't make a lot of sense in a spec - it is a
nothing but a fill word. The whole spec is about describing the "right
thing" or what's appropriate compared to what is "unappropriate", "bogus"
or "brain dead".
>>response that is subject to server-driven negotiation. Doing so allows a
>>cache to properly interpret future requests on that resource and informs
>>the user agent about the presence of negotiation on that resource. A server
>>SHOULD include a Vary header field with a non-cachable response that is
>>subject to server-driven negotiation, since this might provide the user
>>agent with useful information about the dimensions over which the response
>>vary at the time of the response.
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> Again, I liked the old text better. The addition of `at the time of the
> response' only begs questions which need not be answered.
There is an important note to this in that it points out how the vary
header allows a cache to handle _future_ requests based on what a server
knows _at the time_ of the response. This was one of the things that was
confusing in the current wording in the spec.
The time of the response is well-defined - it's the time at which the
message was originated and which may be put into a Date header. This is
described in section 14.19.
Thanks,
Henrik
--
Henrik Frystyk Nielsen, <frystyk@w3.org>
World Wide Web Consortium
http://www.w3.org/People/Frystyk