[wg11] Part 28 teleconferences

[Ed Barkmeyer]

Heidi L Preston wrote:

>  Here is a summary of what I have heard from the last few
> days of exploder comments:
> 
> REQUIREMENT 1:
> Consensus: We need a simple XML Schema derived from the EXPRESS model by
> simple rules with no options.  It should eliminate restrictions and
> features that make the XML Schema difficult for people to read and
> difficult for tools to use.  This binding needs to be short, clear and
> concise so that any errors in this specification can be rapidly determined
> and eliminated.

agree.

> Open for discussion:
> 1.  The AP binding needs to follow the mainstream conventions for XML data
> (no surprises to the end user).
> 2. The AP binding must have sufficient legibility/understandability to
> allow the AP developers to add definitions to the schema that may come from
> the AAM, ARM, mapping tables and AIM rules.
> 3. The AP binding must support conformance checking so that vendors and end
> users can easily determine if an XML data set conforms to the AP.
> 4.  The AP binding should be at least as easy to implement as Part 21 and
> should make STEP data easier to understand.

In bullets 1-4, the term "AP binding" refers to the "simple XML Schema 
derived from the EXPRESS model by simple rules with no options", right?

In bullet 3, what is meant is:
"The AP binding must permit EXPRESS-based conformance checking, so that 
..."  (It may not have to do anything special to support it, but it must 
not do things that make EXPRESS-based conformance checking difficult.)

Also, I think we need a bullet 5:
5.  The XML schema, and the conforming XML documents, should not 
*require* pre-processors, post-processors, or their exchange libraries 
to have any knowledge of EXPRESS.  They may use other documentation to 
understand the semantics of the exchange, and the corresponding rules 
and constraints.

> REQUIREMENT 2:
> An XML Schema designed for type-safety validation of the constraints of an
> EXPRESS information model, which can be accomplished without the use of
> EXPRESS-based tools.

I agree with the intent here, but I think the wording is strange.  We mean:
"An XML schema designed for XML-schema-based validation of the exchange 
documents, including all constraints of the EXPRESS information model 
that can be reasonably mapped to XML Schema.  This binding should allow 
fairly strong validation of the exchange document without the use of
EXPRESS-based tools.  It should not require senders, receivers or their 
exchange libraries to have any knowledge of EXPRESS."

The last sentence is like bullet 5 above, but here it is *not* an issue; 
it is a requirement.

> REQUIREMENT 3:
> A configurable XML Schema defined by taking the STEP EXPRESS Schema
> explicitly/precisely into account.  P28Ed2 is used to formally describe the
> generation of that XML Schema out of the (original) AP data model.  A
> pre-existing XML Schema is defined somewhere and Part 28 is used to specify
> the relationship between it and a STEP EXPRESS schema covering the same
> scope.

In the second sentence, replace "AP" with "EXPRESS".
In the third sentence, replace "pre-existing" with "normative", and 
delete "STEP".

> The purpose of the configuration language is to define the mapping between
> the data elements modeled by the normative EXPRESS schema and the data
> elements specified by the XML schema.  The expectation should be that there
> is a one-to-one mapping for all the conceptual data elements in the EXPRESS
> ARM, but they may be somewhat reorganized in the XML.  In effect, the
> configuration language is the language of a "mapping table" between the
> EXPRESS schema and the XML schema.

I think Germany might have some problem with this view.
 From Guenter, I understand two issues:

1.  Whether the normative XML schema is "pre-defined", and the 
configuration file is documentary (like mapping tables), or the 
normative XML schema is defined as the one that results from the 
configuration file (whatever that XML schema is).  (OTOH, the structure 
of the configuration file may not depend on which of these is the case.)

2.  Whether there is a need for a conformance class for post-processors 
that can accept an arbitrary EXPRESS schema and an arbitrary 
configuration file for that schema and correctly process the 
corresponding XML data.  (No one has questioned whether we need the 
"schema generator" conformance class.)

[I think we continue with the forward mapping approach in the N229 CD, 
and you can delete the following:
> The current configuration language approach is "forward" and "by
> exception".  That is, it assumes the default mapping (one of (1) and (2)
> above) and specifies only the mapping for elements that do not match the
> default mapping.
]

> SIDE ISSUE:
> The Part 28 Base XML Schema is an accident of the design of the standard.
> We factored all of the stuff that is invariant over all Part 28-conforming
> schema into a single XML schema and made that schema a normative resource.
> We could have created 3 separate normative resource schemas, or simply
> required a conforming XML schema to contain all of those declarations, or
> at least the ones it uses, verbatim.  The three "separable base schemas"
> are:
> - the configuration language schema
> - the document structure schema
> - the data types and elements that may be used in a derived XML schema.

This issue is not a *requirement* for Part 28.  The requirement for 
runtime independence from EXPRESS is (now) given above.  This is about 
the document and schema structures that support that requirement.
It can be condensed to two issues:

1.  Should the Part 28 Base XML schema be divided into separate schemas 
for the three separable content areas:
 > - the configuration language schema
 > - the document structure schema
 > - the data types and elements that may be used in a derived XML schema.

2.  Should the last one -- "the data types and elements that may be used 
in a derived schema" -- be required to appear directly in the derived 
(Tns:) schema, so that no "Part 28 Base XML Schema" (ex:) is needed *at 
run time* to validate a uos?

David, do you agree with the addition of bullet 5 above?  And do you 
agree that this issue relates to that bullet (in Requirement 1 and 2)? 
Is there a UK issue that matches this "side issue"?  There is a related 
U.S. issue (US-4) but it only asks for "conformance of a uos only, 
independent of the iso-10303-28 document structure".

Thanks,
-Ed

-- 
Edward J. Barkmeyer                        Email: edbark at nist.gov
National Institute of Standards & Technology
Manufacturing Systems Integration Division
100 Bureau Drive, Stop 8264                Tel: +1 301-975-3528
Gaithersburg, MD 20899-8264                FAX: +1 301-975-4694

"The opinions expressed above do not reflect consensus of NIST,
  and have not been reviewed by any Government authority."