W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > May to August 1996

Re: Last gasp terminology issue

From: Roy T. Fielding <fielding@liege.ICS.UCI.EDU>
Date: Tue, 04 Jun 1996 13:56:20 -0700
To: Koen Holtman <koen@win.tue.nl>
Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
Message-Id: <9606041356.aa10716@paris.ics.uci.edu>
>>> The If-Modified-Since request-header field is used with the GET method
>>> to make it conditional: if the requested variant has not been modified
>>> since the time specified in this field, an entity will not be returned  ...
>>
>>The above is a protocol error.  IMS applies to the resource as a whole,
>>not to any single variant.
> 
> I do not agree that this is a protocol error.  I always assumed that
> the Last-Modified header applied to the variant, not to the whole
> resource.

It applies to the resource as a whole, which explains why the definition
was the way it was (i.e., specified in terms of how the recipient will
interpret it) and should be.

> It is very time-consuming, if not plainly infeasible, for a server to
> compute a collective last-modified value for all variants of a
> negotiated resource, especially if these variants are not all stored
> on the same origin server.  

Our (Apache and CERN) experience has been that it is quite easy.  For 
greater efficiency, cache the value and always report the latest one
of all the entities sent so far (which also explains why it is defined
in terms of what the server believes, rather than any absolute quality).

> I would agree to your proposed change if servers were allowed not to
> implement IUS for a resource if they never return Last-Modified dates
> in responses for the resource.

That is already true.

>>We (Henrik and I) spent a long time getting people to understand and
>>accept that definition of the semantics of last-modified, and that
>>work must not be thrown out at the last minute.
> 
> I am not aware of ever having been educated by about the fact that
> that last-modified applied to the negotiable resource and all of its
> variants as a whole.  I always assumed it applied to the individual
> variants.

The discussion occurred when the WG first started -- it was probably
before you got involved.  In any case, it doesn't need to be reopened.

 ...Roy T. Fielding
    Department of Information & Computer Science    (fielding@ics.uci.edu)
    University of California, Irvine, CA 92717-3425    fax:+1(714)824-4056
    http://www.ics.uci.edu/~fielding/
Received on Tuesday, 4 June 1996 14:10:54 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:32:02 EDT