Re: [cam] WS-I and Schema profiles

["David RR Webber" <david@drrw.info>]

Farrukh,

The problem with a minimalistic approach
to schema is that it does not address the
right problem IMO.

We've said all along that simply apply
an XML veneer to eBusiness transactions
does not make for interoperability and
ease of integration - this was a myth created
by the hype around XML.

The other problem is that XML itself is the
exchange format - not schema.  Schema is
merely syntax about syntax.  And schema
was *never* designed for eBusiness
transaction handling.  This was missing from
day-one in the W3C requirements of schema, and
so it was never engineered to solve this task.

There is no means to share vocabularies and
dictionaries built right in - so that means
every piece of schema is a new piece - hence
we have hundreds of thousands of schemas;
all different.

Notice that CAM includes the ability to
link to registry semantics and vocabularies.
That is key to being able to both create
alignment and simplify building consistent
components with shared semantic
representations.  If you dumb down schema,
it is just that - dumb markup.

Another vital missing piece from schema
is the ability to state context.  That leads
to confusing constructs and complexity
in structure definitions in schema that
users cannot then understand why and
when they should apply what with what.

Anyway - I guess this is getting into a bit
of a rant ; -)

The main point still holds - if they are trying
to solve this problem in a real way - to
create easier means to ensure interoperable
business transactions - then they should be
looking at tools like CAM - and not trying
to dumb-down schema!

Bottom line is you can create very simple
and clear XML - then associate a CAM
template with that - to capture the
business use rules and apply context to
that structure.  Then you can associate
registry references to each part of the
XML to enable full semantic linkage
to the facets and other information
automatically.  And these can then
become fully functional sub-assemblies
that can be simply included in, with
context driven conditionals.

Imagine a world where you could not use
part numbers to lookup product details;
instead you had to print all that stuff on
the product itself.

I believe the WS-I folks can find
real solutions to this by taking advantage
of CAM and Registry mechanisms.

I've heard it said that schema as it stands
is like writing assembly code - what
people need is a "COBOL" to make
the task accessible.

I believe we can provide them with
those tools.

Thanks, DW.


----- Original Message ----- 
From: "Farrukh Najmi" <Farrukh.Najmi@Sun.COM>
To: "CAM TC" <cam@lists.oasis-open.org>
Sent: Wednesday, June 16, 2004 6:20 PM
Subject: Re: [cam] WS-I and Schema profiles


> David RR Webber wrote:
>
> >Monica,
> >
> >Reference the OASIS News item below.
> >
> >I'm not sure who "Glover" is - but I'm
> >guessing - probably chair of co-chair of WS-I?
> >
> >Could you bring to the WS-I teams
> >attention that OASIS already have the
> >CAM specification and open source jCAM
> >implementation that is perfectly suited to
> >creating Schema profiles with.
> >
> >There's also a tutorial available on how you
> >can augment schema definitions in this way
> >with several examples in the ZIP file
> >that is downloadable from the OASIS CAM TC
> >website.
> >
> >
> David,
>
> By schema profile they mean defining subset of XML Schema spec and
> specific contraints on use of XML Schema in order to improve
> interoperability. I do not think that CAM is suited for that task. No
> technology solution is. What is needed a a profile spec that describes
> the constraints on and subset of XML Schema.
>
> -- 
> Regards,
> Farrukh
>
>
> >Thanks, DW
> >
> >----- Original Message ----- 
> >
> >
> >>Time to Rethink XML Schema?
> >>Jack Vaughan, Application Development Trends
> >>
> >>Profiles are useful as they distill the experiences of bleeding-edge
> >>implementers. To make one Web service talk to another is still
> >>something of an accomplishment, and using WS-I profiles can save a
> >>lot of time on the old learning curve. But what of schema? The World
> >>Wide Web Consortium (W3C) has worked hard to standardize this part
> >>of XML. Apparently, even though XML is five years old, this is still
> >>a maturing area. Glover told us that his group is organized to help
> >>users surmount problems, and that feedback indicates schema
> >>representation is one of them. "A lot of people are having problems
> >>using XML schema representation. They are building schemas that
> >>define payload. The problem is that the spec [from the W3C] is
> >>extremely large. At times, it offers fundamentally different ways
> >>to describe the same thing," said Glover. He noted that the W3C is
> >>working right now to fix "errata," but suggests that WS-I may be
> >>able to be helpful in establishing useful profiles for schema. "There
> >>are areas where it could be made more precise," Glover said.
> >>
> >>http://www.adtmag.com/article.asp?id=9589
> >>See also XML Schema Languages: http://xml.coverpages.org/schemas.html
> >>
> >>
> >>
> >>
> >
> >To unsubscribe from this mailing list (and be removed from the roster of
the OASIS TC), go to
http://www.oasis-open.org/apps/org/workgroup/cam/members/leave_workgroup.php.
> >
> >
> >
>
>
>
>
> To unsubscribe from this mailing list (and be removed from the roster of
the OASIS TC), go to
http://www.oasis-open.org/apps/org/workgroup/cam/members/leave_workgroup.php.
>
>