Re: HTTP If-* headers, etags from Erik Seaberg on 2002-04-29 (w3c-dist-auth@w3.org from April to June 2002)

From: Erik Seaberg <erk@flyingcroc.com>
Date: 29 Apr 2002 10:49:03 -0700
To: w3c-dist-auth@w3c.org
Message-ID: <86bsc2e6pc.fsf@unx51.staff.flyingcroc.net>

Jason Crawford <ccjason@us.ibm.com> writes:

> One thing I was concerned about is resources whose GET output
> changes often in hard to predict ways[....] But I think the answer
> to that is, you don't do a PUT against the live resource.

Are servers commonly expected to keep GET on a source resource bitwise
identical with the last PUT?  It seemed transcoding, canonicalizing,
or other postprocessing at PUT time would be expected to routinely
change strong ETags (in fact we do that sort of thing to deter abuse
of a free hosting service).  Are common clients unable to accomodate
this?

Received on Monday, 29 April 2002 13:50:28 UTC