Re: rpc/literal

Paul Johansen wrote:
> In this article:
>
> http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dnsoap/
> html/understandsoap.asp
>
> the author makes the quote, "Rpc/literal [is] possible but [it doesn't]
> make much sense"
>
> Can someone expand on why that is the case?

rpc/literal is a tough one.  Even the WS-I during their basic profile
1.0 work failed to kill it, much to the chagrin of many doc/lit advocates.

They did "kill" rpc/enc though for good reasons.  Why have 2 different
wire encodings especially when one (literal) has so much industry
support.  I think its the same thinking that gets you to a philosophy
that rpc/literal is just confusing the situation.

If you want to stay code centric it can look attractive.  The parameters
are nice literal schema defined data types and the operations follow a
comfortable method calling arrangement.  But why?  You are restricting
clients to those that know how to handle rpc/lit.  Which is not very
many.  The case is strengthened by the fact that wrapped doc/lit looks
the same.  So lets all simplify and move on!

Jim Murphy
Mindreef, Inc.