Re: ETag & Content-Encoding

[James Aylett <james@xxxxxxxxxxxx>]

On Sun, Sep 30, 2007 at 06:44:53AM -0400, Sam Ruby wrote:

> >Is the gzip-ed content a different version of the original content?
> 
> My read of the spec is that the act of compressing the representation of 
> the resource does not affect the identity of that resource:
> 
> http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.11

That was always my reading of things, but as Julian points out, it's
not how apache are now reading it (as in, they behave like that but
there seem to be enough people acknowledging a bug to accept they're
wrong).

See 7.2.1 of RFC2616, where entity-body is laid out as the result of
the Content-Encoding (which itself is defined as an entity header in
14.11). So the *entity* changes with Content-Encoding; it is natural
that the ETag changes as well.

James

-- 
/--------------------------------------------------------------------------\
  James Aylett                                                  xapian.org
  james@xxxxxxxxxxxx                               uncertaintydivision.org