xml-dev - Re: XSLT with a longer wire == SOAP?

[Mark Baker <mark.baker@canada.sun.com>]

"Box, Don" wrote:
> Then why did HTTP/1.0 and greater support POST, which inherently rendered
> URI as user-clickable entities fairly useless.

URIs identify resources.  POSTed-to URIs identify containers.  That some
containers choose not to reveal a representation of themselves on a GET
is no fault of URIs or HTTP.

(btw, please don't bring up namespace URIs, especially not here 8-).

>Most HTML FORM posts use
> HTTP-based URI, yet clicking one embedded in an email message is unlikely to
> yield the expected results. That is no different than the URI I listed here.
> 
> > SOAP shouldn't use http URLs, SOAP shouldn't use HTTP on port 80.
> 
> I assume that you view CIM Over HTTP/WBEM, which is functionally a close
> cousin of SOAP over HTTP, as equally wrong? Or the various ORB vendors'
> (including MS and Sun) attempts to shoehorn all or part of their RPC
> protocols through HTTP as pretty bad too?

I don't know about Larry, but I do.

> I believe we've all heard the arguments against using HTTP for non-hypertext
> applications (if not, the SOAP archives have several old/dead threads on
> this one that go nowhere). The argument is largely academic given the
> propensity of developers with free will and free software to support them.

No, it's not academic.  Using HTTP on port 80 in this manner is a bad
idea for more reasons than because it's not hypertext.  See below.

> The reality is that Apache became the "application server" of choice whether
> you or I choose to use it or not.
> 
> At the end of the day, the case against HTTP is similar to the case against
> fatty food. No matter how many pundits and experts trot out pictures of the
> fat deposits harvested from the autopsies of overweight dead people, people
> will give into their desires and indulge in various guilty pleasures. I
> applaud your long-standing opposition to developers using HTTP for anything
> other than surfing web pages in HTML and supporting multi-media formats, but
> I am afraid you will have to convince far more people than just the readers
> of this list to stop the adoption of HTTP in ways you may not like.

It's one thing for hackers to use it to get something done.  It's an
entirely different thing to encourage wide spread deployment of a
"standard" (SOAP) that solves no known problem (modulo section 8), and
whose main commercial proponent is on the record as saying that they
want to lock-out competitors from the market by releasing stuff just
like it;

 http://www.opensource.org/halloween/halloween1.html#quote9

MB

***************************************************************************
This is xml-dev, the mailing list for XML developers.
To unsubscribe, mailto:majordomo@xml.org&BODY=unsubscribe%20xml-dev
List archives are available at http://xml.org/archives/xml-dev/
***************************************************************************