Been away for a week so I am just catching up.

> But it doesn't. Not the one at
> http://www.lib.ox.ac.uk/jafer/zng/zng-p1.wsdl.html
> which is the current file (isn't it?)  It just points to a schema.

It is the current file at present - I did ask the list for a summary of
any decisions made at the April ZIG as to any revisions I need to put in
(alongside some of the implementation issue that have emerged)
 
> Can't the wsdl define the response rather than point to a schema?

The WSDL can include the response rather than point to it - but in both
cases the response is defined by an XML Schema (whether embedded in the
WSDL or pointed to). Having a separate document is one of the WSDL
authoring styles and in any case means that the Response schema can be
imported into non-WSDL compliant tools easily enough for use in say SRU

 
> Is the reason that the wsdl doesn't do that now, that we thought there
> would be
> more than one possible response format, and wsdl can't handle that?
If
> that's
> the reason, and if we now agree that there won't be more than one
response
> format (do we agree?) then can we now define the response in the wsdl
and
> get
> rid of rs1?

At one point it look like we may have two schemas for SRU and SRW but if
the consensus is one then that is fine. In which case RS1 would be
replaced by a XML Schema for the SRU/SRW types.

Is this the only change - if so I'll put up a revised WSDL file.

Matthew