On Mon, Sep 10, 2001 01:42:37PM +0100, Lloyd Wood allegedly wrote:
As reference (1) indicates, SOAP is documented in the W3C. Why is this
work being done as an IETF draft and not in the W3C? BEEP is RFC3080,
but SOAP in BEEP is a SOAP-specific problem, which afaik means it's a
W3C problem.
Good point. Let's take a parallel with how the IETF divides work with
groups working on lower layers. When we formed the sub-IP area the
philosophy was that the IETF would take on sub-IP issues iff ...
- The work depends on IP expertise, which is here,
A good application protocol is far less dependent on the format and structure
of the data being carried than it is on the proper design of the application
protocol itself. So, while it is true that data format and structure is the
W3C's domain and not the IETF's, the IETF is where the expertise to do
application protocol design lies.
We have plenty of examples of poor application protocol design, and we also
have plenty of examples of poor data bindings to application protocols, so we
know this isn't trivial stuff anyone can just slap together.
- That it is critical to the development of the IP infrastructure, and
A way to do the things SOAP does is critically needed at the application layer.
And while I personally do not like the approach SOAP uses, it does represent a
solution, and moreover it seems to be the solution peopel are pursuing.
- That it directly or indirectly affects operations or routing at the
IP layer.
SOAP is going to see widespread use no matter what. The other application
transport for SOAP is HTTP. Use of HTTP for things it isn't well suited for is
something we want to discourage. So doing this correctly is of considerable
concern to the application layer.
Ned