Terminology II

[Lorenz Däubler <l.daeubler@tu-bs.de>]

   [[[ For technical reasons, this message is forwarded only today.
       --PNX moderator ]]]

Dear all,

this posting represents merely some impressions about the exchange
discussion, so sincerely I hope I do not bother you all.

The problem of the exchange of model information in neutral data formats
is well-known especially in the field of CAD. Thus, an elaborate ISO
standard (ISO 10303) for the exchange of product model data arose in the
mid of the 1990s. This standard separates the techniques of the
representation of product information from the implementation methods
used for the data exchange. The representation techniques provide a
single representation of product information common to many
applications. In contrast the implementation methods support the
exchange of the product data. Additionaly to this ISO 10303 defines its
own formal language (called EXPRESS) to specify the representation of
the product information.

Transfering this approach to the exchange problem of petri nets this
would results in seperate exchange steps:

1) Formal specification of all elements that are needed to build any
kind of  petri nets. The description method for this specification is
arbitrary (for example class diagrams, ER diagram or something else).
The ISO/IEC 15909 of Mr. Billingtion is one step towards such a
specification. He uses set theory for his net specification.

2) With this formal representation of the petri net model information,
arbitrary implementation methods can be defined by mapping the formal
representation on the language used for that mapping. XML for example is
such a language. But maybe somebody is interested in sharing different
high level net models over data bases or between two applications on
runtime. This would call for alternative exchange structures.

It seems to me that the step 2 has been done before step 1. This  leeds
to informal discussions about basic structural properties of petri nets,
e.g.:

> Petri nets may be structured.  We distinguish two kinds of
structuring.
> The first kind organizes a Petri net in pages.  Thus, a Petri net may
> consists of pages. A page may be also substructured in pages.

I read this as: "Each net and each page may contain net elements
and pages, possibly both." Is this correct?

This could be discussed in a formal manner based on the ISO/IEC 15909.
Pages, hierarchies, reference nodes and so on could exclusively be part
of the ISO15909 discussion. Afterwards, the XML representation of (high
level) petri nets would be much more easier - I think. On the other hand
such an far-reaching normative reference restricts tool programmers in
their software design.

This discussion has been going on for years in the field of CAD, but now
their exchange problems seem to be solved.

Best regards
Lorenz

--
*****************************************************************
Dipl.-Ing. L. Daeubler          Langer Kamp 8, 38106 Braunschweig
Institut fuer Regelungs- und    Tel.: 0531/391-3312 Fax.: -5197
Automatisierungstechnik         mailto:L.Daeubler@tu-bs.de
TU Braunschweig                 http://www.ifra.ing.tu-bs.de
*****************************************************************