RE: "Uh, what do I need this for" (was RE: XML.COM: How I Learne d to Love daBomb)

["Simon St.Laurent" <simonstl@simonstl.com>]

On 19 Aug 2001 20:25:21 -0400, Champion, Mike wrote:
> For what it's worth, I was talking about SOAP 1.x itself.  I personally
> agree that it could be considerably simpler and still hit the 80:20 point
> ... but the same can be said for XML itself, XSLT, etc.  SOAP -- like the
> predecessors of most successful W3C specs ... and unlike the W3C specs that
> are causing most of the confusion -- does have a real track record. The
> folks on the soapbuilders list have been working on SOAP 1.1
> interoperability issues, and SOAP 1.2 is being defined by a largely public
> process that incorporates the SOAP 1.1 experience and should help ensure
> that it really does reflect best practices in  the field, not just best
> guesses in the committee room.  

I wish I had that confidence.  I'm afraid I see SOAP as having overbuilt
on rather rickety foundations in the first place, overstressing a
protocol that wasn't designed for (or even accidentally suitable for)
the kinds of operations SOAP is proposed to handle.

While I'm happy to see that public input and implementation experience
is having a substantial impact on the surface characteristics of SOAP, I
can't help but wonder why its underlying premises have been so rarely
challenged.  

The IETF has a draft[1], but I'm not sure it's made it to Best Current
Practices. (Last call went out 17 October 2000, so it may have been
dropped.)  The more I've thought about this issue, the more I think
Keith Moore is right.  XML-RPC is as far as I'm willing to push HTTP,
and that may be reasonably considered to be too far.

[1] -
http://search.ietf.org/internet-drafts/draft-moore-using-http-01.txt