W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > May to August 1997

Re: cache-busting document

From: Larry Masinter <masinter@parc.xerox.com>
Date: Sat, 7 Jun 1997 09:36:34 PDT
Message-Id: <33998E12.4A01@parc.xerox.com>
To: Martin Hamilton <martin@mrrl.lut.ac.uk>
Cc: http-wg@cuckoo.hpl.hp.com, ircache@nlanr.net
I'd like to see more analysis (or references to it) associated
with each individual piece of advice.

>   Use a server which supports HTTP 1.1 - this has a number of
>     additional features to support caching.

I've not seen any studies or reports that document the effect
on effectiveness of caching for the HTTP/1.1 features that
were added to support caching. Documenting the results would
be very useful. It would be my guess that, after such documentation,
you'd have more specific advice than 'use HTTP/1.1'.

>  Use the Expires header on documents and images where feasible
>     - this will help caches to decide when your objects are stale.

When is it feasible and when is it not? While we've conjectured
the applications of Expires for sites that do dynamic content
generation from static sources, for example, is that actually
feasible? Do sites with planned expiration set expires dates?
Is it feasible to, for example, declare that '/images' at a site
never changes (if you need to modify an image, give it
a new name and change all the references), and then set it
so that the embedded images never expire from caches even if
the documents are dynamic?

>   Use an HTTP server which supports the GET method with the
>     If-Modified-Since header - this will help browsers and proxy
>     caches to figure out whether their cached copy of a file is
>     out of date.

Don't they all? At least for files? 
...

I think you get the idea. Documenting what works and what
doesn't work in practice would be very useful. A checklist
for 'things to try' sounds more like a plan for a research
program, though, rather than practical advice for site
administrators, especially since the list isn't accompanied
by any advice on how to measure the effectiveness of the
trial.

Regards,

Larry
-- 
http://www.parc.xerox.com/masinter
Received on Saturday, 7 June 1997 09:39:04 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:32:44 EDT