W3C home > Mailing lists > Public > ietf-http-wg@w3.org > October to December 2007

Re: i37 - Vary and non-existant headers

From: Henrik Nordstrom <henrik@henriknordstrom.net>
Date: Thu, 15 Nov 2007 02:55:26 +0100
To: Adrien de Croy <adrien@qbik.com>
Cc: "'HTTP Working Group'" <ietf-http-wg@w3.org>
Message-Id: <1195091726.30372.71.camel@henriknordstrom.net>
On ons, 2007-11-14 at 19:29 +1300, Adrien de Croy wrote:

> Just basically that since the encodings are optional, and you can't
> guarantee a client can use any content other than unencoded, then the
> only safe option where there is no C-E header is to send unencoded or
> 406 if you don't have an unencoded one (when will that ever happen on an
> origin server?).

When the server only has an encoded variant. Consider a request for
http://www.example.com/some.file and the server only has some.file.gz
and the request has "Accept-Encoding: [blank]".

Clients not sending Accept-Encoding is assumed to accept any encoding
which may exists now or in future. Not sending Accept-Encoding is
basically saying "I don't care how it's encoded, just send me whatever
you have, but preferably unencoded if possible".

> A 406 certainly
> would prompt UA developers to start advertising capabilities.

So would getting incompatible media.. but in real life it's not such big
deal. All the server need to do is to include User-Agent (or whatever
header identifies the client in question) in Vary.


Received on Thursday, 15 November 2007 01:55:43 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 11:10:43 UTC