W3C home > Mailing lists > Public > www-talk@w3.org > May to June 2004

Re: HTTP 1.1 pipelining

From: A Bagi <ahmed.bagi@virgin.net>
Date: Fri, 28 May 2004 21:14:54 +0100
Message-ID: <002001c444f0$71e724f0$0301a8c0@sn023784320093>
To: "Ian Clelland" <ian@veryfresh.com>, <www-talk@w3.org>
Cc: "S. Mike Dierken" <mdierken@hotmail.com>

Idempotence: the ability of a Document to be transmitted and accepted more
than once with the same effect as being transmitted and accepted once. This
somehow does not mean no side-effects (web applications, GET)!
Only idempotent requests can be pipelined, such as GET and HEAD requests
with maximum scucess.  POST and PUT are dodgy business!!
Ahmed Bagi
Manchester

----- Original Message ----- 
From: "Ian Clelland" <ian@veryfresh.com>
To: <www-talk@w3.org>
Cc: "S. Mike Dierken" <mdierken@hotmail.com>
Sent: Friday, May 28, 2004 7:34 PM
Subject: Re: HTTP 1.1 pipelining


>
> S. Mike Dierken wrote:
>
> > I'm looking into request pipelining & had a question about the kind of
> > requests allowed.
> > The RFC says only idempotent requests should be pipelined.
> > This FAQ
> > (http://www.mozilla.org/projects/netlib/http/pipelining-faq.html) from
> > Mozilla says PUT should not be used because it isn't idempotent.
> > Except that it is. PUT is idempotent (repeatable with deterministic
> > results).
> >
> > <>Which is it?
>
> I'm pretty sure that the Mozilla FAQ is wrong in this case. Perhaps the
> author is confusing being idempotent with having no side-effects. You're
> right that the semantics of PUT do make it acceptable for pipelining.
>
> Of course, given the number of web applications out there these days
> which break idempotence even for GET requests, I'd be worried about
> assuming that property for anything on the web.
>
>
> Ian Clelland
> <ian@veryfresh.com>
>
Received on Friday, 28 May 2004 16:16:54 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 27 October 2010 18:14:28 GMT