W3C home > Mailing lists > Public > xml-dist-app@w3.org > January 2001

Re: Integrating XP Into Web Infrastructure

From: Aaron Swartz <aswartz@upclink.com>
Date: Tue, 09 Jan 2001 08:52:02 -0600
To: Joshua Allen <joshuaa@microsoft.com>, "'Kurt Cagle'" <cagle@olywa.net>
CC: XML-DIST-APP <xml-dist-app@w3.org>
Message-ID: <B68081B2.1FD3D%aswartz@upclink.com>
Joshua Allen <joshuaa@microsoft.com> wrote:

> Extracting the information from the HTML would be pretty tedious, though, and
> prone to breaking.  Eventually you would want a generic sort of technique that
> service providers and developers could all count on so that each "hookup"
> wouldn't have to be negotiated individually.  At that point you would just
> invent XML-RPC :-)

Or just plain XML? Or plain text? We can still make HTTP GET requests for
XML, you know. ;-)

> So SOAP will often be used for synchronous calls to an individual
> method, but I also hope it will be used just as much for asynchronous
> message passing that may or may not invoke one or more actions on the
> receiving end.

I totally understand this -- SOAP had benefits and plain HTTP has benefits.
That's not the question. The question is what is the XP's position on the
issue.
 
> By lack of URIs do you mean the inability to call through an HTTP-GET (via
> querystring)?

Pretty much, yeah. But also the fact that most SOAP implementations live at
a specific URI. While SOAP may be able to be sent in many different ways,
the fact of the matter is its generally being done using POSTs to a single
URI receiver.

Many folks feel this is wrong. What does the WG feel?

> I think even synchronous RPC over HTTP is pretty incredible.  There are tons
> of possibilities, but so far the job of screen scraping and
> reverse-engineering forms has prohibited me from doing all but the things I
> really, *really* needed to do...That's what excites me about
> SOAP/XML-RPC or whatever else.

That's what excited me about XML on the Web, long before I ever heard of
SOAP or XML-RPC. ;-)

>>> 2) Do you expect to preserve the semantics of HTTP requests (HEAD, GET,
>>> POST, PUT) with the HTTP binding?
> I'm not sure I understand here

HTTP has some built-in semantics: a GET request is a specific request for
information; a POST request is a submission of information to a specific
URI; a PUT command is a request to place information at that URI. Most SOAP
implementations are set up so that every request goes to the same URI using
a POST command. The semantics of HTTP are thus lost.

> BTW, I'm not on the WG; this is just a focus and interest for me..

I understand, I hope my explanations have helped you and the others who are
confused about what I'm talking about. ;-)

-- 
Aaron Swartz <me@aaronsw.com>|               RSS Info
  <http://www.aaronsw.com>   |   <http://www.blogspace.com/rss/>
AIM: JediOfPi | ICQ: 33158237| news and information on the RSS format
Received on Tuesday, 9 January 2001 09:53:40 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:58 GMT