W3C home > Mailing lists > Public > www-talk@w3.org > May to June 1999

Re: Static? Dynamic?

From: Dan Connolly <connolly@w3.org>
Date: Wed, 30 Jun 1999 10:46:54 -0500
Message-ID: <377A3BEE.F7C3FDE2@w3.org>
To: John Martin <jmartin@netapp.com>
CC: roconnor@uwaterloo.ca, www-talk@w3.org, jigsaw@w3.org
John Martin wrote:
> 
> Just to be clear on what we are discussing.
> 
> At 10:35 PM 29/06/99 -0500, Dan Connolly wrote:
> >As a community, I feel we have the following choice:
> >       -- provide TTL metadata in those cases where
> >               we don't want clients (incl. proxy caches)
> >               to assume a default TTL around a day or a week
> >               or whatever, so that folks can implement caching per
> >               the specs and get reasonable application behaviour
> 
> For client (user / browser) caches this may indeed be true but all the
> proxy caches I know of err on the side of caution. In other words, the
> non-existance of a 'Expires' or 'Last Modified' header ensures that an
> object will never be cached.

I've seen caches (e.g. the Hensa cache, if I recall correctly)
that cache for 24 hours in the absence of
Expires... hmm... maybe lack of Last-Modified causes even
the ones I've seen to be more conservative.

> >       -- live with the cache-busting techniques that providers
> >               are forced into due to the current lack of discipline.
> 
> This would be regrettable. Sadly, I dont know of many web server
> implementations which give information providers the choice to use / set
> Cache-Control or Expires settings.
> 
> Is W3C coordinating any work on cacheability or freshness issues like this?

Er... yes. I think the HTTP 1.1 design covers these cacheability
and freshness issues, and we did a lot of "coordinating" in
the development of the HTTP 1.1 draft standard.

We host the
	HTTP/1.1 Implementor's Forum
	http://www.w3.org/Protocols/HTTP/Forum/
(excerpt: "Agranat Systems EmWeb
    Server: The clockless server does not send Date headers,
but does provide enough information for caches to operate correctly. ")

A search for "cache" on our web site gives 3407 hits;
382 pages we maintain, plus 3025 hits from mailing lists archives.

http://search.w3.org/Public/cgi-bin/query?mss=simple&pg=q&what=web&fmt=.&filter=lists&q=%2Bcache

spread over these forums that we host and/or archive:
-www-disw -w3c-news -w3c-wai-hc -www-webdav-dasl -xsl-editors
-www-international -ietf-tls -www-logging -www-validator -w3c-wai-gl
-wai-wcag-editor -ietf-http-wg -ietf-http-ng -www-multicast
-ietf-discuss -ietf-http-ext -pics-interest -w3c-wai-er-ig -w3c-wai-ig
-w3c-wai-wg -www-annotation -www-dom -www-lib -www-jigsaw -www-talk
-www-proxy -www-wca -www-html -www-amaya -www-style -www-dist-auth
-w3c-dist-auth -www-email-discuss -w3c-wai-ua -ietf-dav-versioning
-w3c-sgml-wg -www-font -uri@bunyip -www-tv -www-svg -www-push

(see http://lists.w3.org/Archives/Public/)


We provide two implementations:

libwww: a C client library
	http://www.w3.org/Library/
	(integrated into Amaya, our browser/editor client)

jigsaw: a java server with proxy features (and a separable client
library)
	http://www.w3.org/Jigsaw/


I'm pretty sure Jigsaw has an administrative interface
for giving TTL info... lemme see...

this looks relevant:
	Configuration of Attributes 
	http://www.w3.org/Jigsaw/Doc/User/attributes.html
but not quite specific to TTL info...

but yes, Expires is one of the attributes you can set:
http://www.w3.org/Jigsaw/User/Reference/w3c.jigsaw.resources.HTTPResource.html#expires

Hmm... as an administrator, what I'd want to do is set
the TTL and have the server compute
expires = last-modified + ttl on each request.


-- 
Dan Connolly, W3C
http://www.w3.org/People/Connolly/
tel:+1-512-310-2971 (office, mobile)
mailto:connolly.pager@w3.org (put your tel# in the Subject:)
Received on Wednesday, 30 June 1999 11:47:12 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 27 October 2010 18:14:24 GMT