W3C home > Mailing lists > Public > www-ws-arch@w3.org > September 2002

Representing HTTP in arch diagrams

From: Mark Baker <distobj@acm.org>
Date: Thu, 19 Sep 2002 17:07:40 -0400
To: "Champion, Mike" <Mike.Champion@SoftwareAG-USA.com>
Cc: www-ws-arch@w3.org
Message-ID: <20020919170740.H1346@www.markbaker.ca>

Hi Mike

Just wanted to respond, but I'm not re-raising my objection FWIW.

On Thu, Sep 19, 2002 at 02:33:55PM -0400, Champion, Mike wrote:
> When I was working on the intro to the WSA document, I was struggling with
> how to illustrate the evolution of the web/web services from "HTTP does it
> all" to the view expressed in the triangle/3 stack diagram.  I didn't get
> very far, but this discussion is helping to crystallize it in my mind.
> 
> I don't think the "HTTP is the circle wrapping the whole thing" picture is
> quite right.

Yah, you're right.  I don't know that there's a reasonable answer
within the structure of that diagram.

>  Perhaps we have a series of pictures: start from the "picture"
> that Mark draws of HTTP being the interaction, the discovery mechanism
> (Google?), and the publish mechanism.  Discuss the reasons why many people
> (not ALL of course!) found that inadequate for machine-machine interactions
> (limits of URI encoding, need for rigorous description of the invocation
> mechanism, need for more sophisticated queries than search engines allow).
> Then show how the HTTP-is-everything diagram "blows up" into the 3-stack
> diagram.

Hmm, that would be interesting.  But like I say, perhaps this can wait
until later.

> Of course, HTTP-is-everything is still back there somewhere for those who
> know a priori or hard-code the knowledge of the interaction mechanism,
> location of important resources, etc.  That is a perfectly respectable thing
> to do, but not general enough to meet the WSA requirements, IMHO.  We need
> to show it in some diagram ... maybe the "where does this come from" view or
> the "assembly language" view, but not in all views.

I don't know what you mean by "knowledge of the interaction mechanism".
And why would you need to hard code important URIs?  I don't follow.

MB
-- 
Mark Baker, CTO, Idokorro Mobile (formerly Planetfred)
Ottawa, Ontario, CANADA.               distobj@acm.org
http://www.markbaker.ca        http://www.idokorro.com
Received on Thursday, 19 September 2002 17:07:44 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 3 July 2007 12:25:06 GMT