W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > October to December 2003

Re: Redirection

From: David Woolley <david@djwhome.demon.co.uk>
Date: Wed, 12 Nov 2003 23:13:28 +0000 (GMT)
Message-Id: <200311122313.hACNDSx01096@djwhome.demon.co.uk>
To: w3c-wai-ig@w3.org

> I consider that proxy behaviour, if not an abuse of HTTP then a failure to 
> implement it as well as is possible. Admittedly it's probably a failure based 
> on adapting to practical experience with sites which in turn fail to implement 
It is not an abuse of HTTP 1.1 as it is based on a requirement of HTTP
1.1 (I'm not sure at what conformance level).  It is in HTTP 1.1 because
GET mode forms were abused for applications that made significant updates
to the server state and should have been done with POST (the problem now
tends to be that forms with pure function semantics are tend to be done
as POST).  I think the HTTP specification says that there must be explicit
cache control information, not just a correlator (e.g. Last-Modified-Date)
before a proxy can assume it is safe to cache a URL containing ?

More precisely, bypassing upstreams proxies is an optimisation based on
the likelihood that a non-cachable resource will be returned.  As caches
are never required to cache any particular page, it can't be a violation.

HTTP 1.1 introduced workarounds for mis-use of GET forms for unsafe operations.
Received on Wednesday, 12 November 2003 18:19:18 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 19 July 2011 18:14:13 GMT