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

Re: Fact-checking: do any in-service proxy caches ever ignore

From: Donald Neal <d.neal@waikato.ac.nz>
Date: Thu, 20 Mar 1997 08:56:18 +1200
Message-Id: <3.0.1.16.19970320085618.479f979e@mailserv.waikato.ac.nz>
To: http-wg@cuckoo.hpl.hp.com
At 09:30 AM 19/03/97 +0100, Koen Holtman wrote:
[...]
>What I remember (though my memory may be faulty) about New Zealand
>educational caches is this:  at the http BOF at www5, when we were discussing
>the HTTP/1.1 caching design, someone connected to the New Zealand caches
>wanted it to be possible for a compliant HTTP/1.1 cache to ignore all
>attempts at a `reload' if the user was accessing some sites with a `low
>educational value'.  The cache would just keep returning the old cached
>object (with a warning) instead of revalidating with the origin server.
>
>I somehow assumed that this meant you would also ignore very short expires
>times from from these sites.  Or are very short expires times from sites
>with a `low educational value' not yet a problem?

  The question to which my posting was addressed referred to current
practice. I don't know of anyone now doing this, and the software I'm
working with now does not allow Expires headers to be ignored.
  Yes, I would still wish to have the option of setting a cache to limit
the bandwidth expended on sites unconnected with the purposes of the
organisation operating the cache.

- Donald Neal
Received on Wednesday, 19 March 1997 13:22:39 EST

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