Re: WSRF: inconsistencies in the Primer document

[Tim Banks <tim_banks@uk.ibm.com>]

Hi Mark,

See below for responses.  I will propose to the TC that the Primer is
refreshed when the specifications are adopted as OASIS standards, and can
incorporate changes to answer your comments then.  I think the points about
missing pieces of the examples would be best answered by putting the files
containing the examples on the TC web site and hyperlinking to them from
the text (assuming there are no technological obstacles)

Thanks for the useful comments.

Tim Banks.

"Mark Wilkinson" <markw@illuminae.com> wrote on 11/02/2006 01:06:44:

> Hi Tim!
>
> I'm going through the WSRF Primer document
> (http://docs.oasis-open.org/wsrf/wsrf-primer-1.2-primer-cd-01.pdf) and I

> find things that are either inconsistent, or perhaps I am just
> misunderstanding...
>
> Page 15/42:  My understanding is that GetResourceProperty is a required
> method to be implemented in any WSRF service, but you don't have it as
one
> of the methods in your pseudo WSDL.

Right - it's in the real WSDL that accompanies the Primer, but it should
also be in the pseud-WSDL.  The  files that contain the wsdl, schema and
message samples are here:
http://docs.oasis-open.org/wsrf/wsrf-primer-1.2-examples-cd-01.zip

>
> Page 15/42:  Actually, the complete WSDL, with all XML Schema, would be a

> useful thing to have here.  You have fragments of it in isolation, with
> much of it replaced by "..." in various parts of the Primer, but some of

> the "..." is replacing things that are not entirely obvious.

Many of the fragments are extracts from the xml files that accompany the
Primer, but they could be made more accessible by hyperlinking from the
fragments to the complete example.

>
> Page 16/42:  The SimpleShoppingCartCreation port is not at all an obvious

> function, since it isn't returning what one would have expected from the

> previous WSDL example.  I'd really like to see the *full* WSDL and the
> *full* XML schema declaration for the input **and output** of the
> CreateCart operation, however neither of them are provided in this
> document.

Ok - they are in the files that accompany the Primer, but that could be
made more obvious.

>  Is the XML Schema representing the EPR of the response
> somethign that is automagically imported by your wsdl:import statement,
or
> does it have to be explicitly described by the service provider in their

> wsdl? (or does the service provider somehow extend an existing EPR
> specification to add their own service-specific parameters?)

The WSDL uses the wsa:EndpointReference element from the WS-Addressing spec
in the description of the response message. That element declares
optional ReferenceParameters of any xml type, so the service implementation
just includes what it needs to in the repsonse.

>
> Finally, it would be very helpful to have the SOAP Header as well as the

> SOAP Body for every message at every stage of every transaction.  It
isn't
> clear which messages contain the header information and which do not.
> This is particularly true for the Checkout function - what is in the
> Header and what is in the Body for that function??  Is the Body empty?
(as
> far as I can tell the Shopping Cart example doesn't get as far as the
> Checkout function... which is the critical one of course :-))

OK  - I will add those details in hyperlinked files.

>
> I guess what I am saying is, too much is taken for granted in that
> primer.  It really needs to have less editing and more complete, explicit

> examples.  A Primer has to assume that the person (like me!) is new
enough
> that you cannot take for granted that what you have edited out is
> obvious.  It isn't :-)

No problem - the feedback will help improve things.  Thanks!

>
> Is there any chance of getting a more complete primer any time soon?
> There are *zero* other examples of the usage of WSRF anywhere on the net,

> so until I can see a complete example I'm stuck not being able to use it

> :-/
>
> Any advice is welcome!

Yes, there's a good chance of a new primer version soon.

>
> Best wishes!
>
> Mark
>
> --
> --
> Mark Wilkinson
> Assistant Professor, Dept. Medical Genetics
> University of British Columbia
> PI Bioinformatics
> iCAPTURE Centre, St. Paul's Hospital