W3C home > Mailing lists > Public > www-tag@w3.org > September 2005

Re: Fwd: I-D ACTION:draft-maes-lemonade-http-binding-02.txt

From: Mark Baker <distobj@acm.org>
Date: Tue, 27 Sep 2005 20:11:26 -0400
To: "Roy T. Fielding" <fielding@gbiv.com>
Cc: W3C TAG <www-tag@w3.org>
Message-ID: <20050928001126.GW8497@markbaker.ca>

On Tue, Sep 27, 2005 at 01:47:19PM -0700, Roy T. Fielding wrote:
> There have been discussions here in the past about the applicability
> of RFC 3205 to various protocols that tunnel through HTTP.
> Interestingly, I just noticed this RFC that attempts to standardize
> an IMAP over HTTP that deliberately (and knowingly) violates every
> piece of advice in RFC 3205 and even manages to require that the
> server *not* implement GET.
> The saddest part is that the correct way to implement IMAP via
> HTTP is to simply use HTTP on a virtual mapping of resources.
> The two protocols have almost identical capabilities, so the
> only translation needed is to map the stateful interaction of
> IMAP to appropriate resource states in HTTP and then extend
> HTTP's authentication to include IMAP's mechanisms.


> Anyone have time to deliver a clue-stick to the LEMONADE stand?

As a veteran of several attempted "Web clue-stickings", in and out of
the IETF, I'd recommend restraint.  By the time the effort has a cool
moniker like "CRISP", "MIDCOM", or "LEMONADE" (not to mention "Web
services" 8-), IME, the commitment to developing a new protocol has
already been made.

Remember the APPLCORE discussions?  From what I could extract from them,
it appears as though the IETF community's idea of protocol reuse is
focused on layer 6 ala BEEP, not layer 7 ala HTTP.

Mark Baker.  Ottawa, Ontario, CANADA.          http://www.markbaker.ca
Coactus; Web-inspired integration strategies   http://www.coactus.com
Received on Wednesday, 28 September 2005 00:09:24 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:32:46 UTC