W3C home > Mailing lists > Public > ietf-http-wg@w3.org > July to September 2011

Re: Expect header 'understand' vs 'meet'

From: Jan Algermissen <jan.algermissen@nordsc.com>
Date: Wed, 7 Sep 2011 22:35:05 +0200
Cc: HTTP Working Group <ietf-http-wg@w3.org>
Message-Id: <14AE9EE8-A1BF-4A30-8F30-A51737B0BA3A@nordsc.com>
To: Julian Reschke <julian.reschke@gmx.de>

On Sep 7, 2011, at 10:24 PM, Julian Reschke wrote:

>> Can't I have a mechanism that allows the client to ask specifically the origin server to only work on my request if it exhibits a certain behavior? And isn't that (also) the idea behind 'Expect'?
> I think lots of people do not want "must understand" extensibility.

Consequently, the only way then to ensure[1] the client that some resource exhibits behavior X is to provide an appropriate 'link', yes?

So, instead of

POST /foo
Expect: exhibits-feature-x

one must do sth like

>> GET /start

<< 200 Ok
<< Link: </foo>; rel=exhibits-feature-x

POST /foo


All fine, just try to understand.


[1] Ignoring the time gap between the client learning and the client doing the request

> See also the various proposals for the "Prefer" header, which defaults the other way...

Received on Wednesday, 7 September 2011 20:35:43 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 1 October 2015 05:36:47 UTC