W3C home > Mailing lists > Public > ietf-http-wg@w3.org > April to June 2009

Re: Proposal for i23: no-store invalidation

From: Henrik Nordstrom <henrik@henriknordstrom.net>
Date: Mon, 08 Jun 2009 14:57:49 +0200
To: Mark Nottingham <mnot@mnot.net>
Cc: HTTP Working Group <ietf-http-wg@w3.org>
Message-Id: <1244465869.25790.21.camel@localhost.localdomain>
mån 2009-06-08 klockan 20:57 +1000 skrev Mark Nottingham:
> Ping?
> 
> I'm still struggling to understand what you want here; see also Roy's  
> comment on the issue in the tracker.
> 
> Are you talking about 304 responses, 200 responses, *any* response to  
> a GET...?

It's about generalizing the rule to get a consistent cache behavior
independent of if the request was a GET or HEAD.

To answer the comment from Roy there is other conditions as well where
the request may need to be forwarded even if the cached entry still
looks valid as such. For example if the request contained Cache-Control:
max-age=N, Cache-Control: no-cache or at the extreme Cache-Control:
no-store (note: in all these cases as request headers, not response
headers). Without having the statement generalized there is corner cases
where a GET could still leave old entries in the cache but a HEAD with
the exact same conditions would force them to be invalidated. Mainly
when the cache contains a still seemingly valid response but the
resource has changed on the server and no longer results in a cacheable
response.

Regards
Henrik
Received on Monday, 8 June 2009 12:58:27 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 06:51:03 GMT