At 11:04 AM 12/15/2004, Mike Taylor wrote:
>[...]
>We have learned -- the hard way -- that SRU is _much_ easier to
>implement than SRW.  On the client side, it is a literally trivial
>matter of an HTML form to generate the URL; the server is slightly
>less trivial, but really, a 20-line Perl script can implement an SRU
>server.
>
>By contrast, it seems that _no-one_ (with the possible exception of
>Matthew) like working with SOAP.

I think you'd have to include in that exception an awful lot
of folks who use .Net and other SOAPish tools.

Also, from a "political marketecture" perspective, we really
do need to be able to say that ISO 23950 is accommodating
the future trend toward Web services as well as the more
traditional CGI (apparently created by NCSA circa 1994?)

Perhaps what is needed is a short specification for the behavior
of the SRW/SRU service at the location given by the base url:

When an SRW/SRU service at the location given by the base url
receives a message that is an XML structure, it should pass
control to an appropriate SOAP processor capable of processing
SRW. Otherwise, the service should process the message according
to the appropriate process for HTTP GET or HTTP POST.

Do we need to have a constraint raeding whether a session must
to be consistent in its style of POST SOAP / POST CGI / GET CGI
across multiple request/response pairs?

Eliot