Page 370 -

368 G. Decket et al.
The chapter discusses an open issue with the underspecification of the
OR-join in BPMN, which has given rise to many proposals outside the standard-
ization process [83]. More open issues with BPMN are described in a number of
papers [78,202,264]
Finally, it is worth mentioning that, while this chapter discusses the relation
between BPMN and YAWL, relations between BPMN and other process model-
ing languages and formalisms have been discussed in the literature. For instance, a
transformation from BPMN diagrams to plain Petri nets is presented in [78], while a
transformation from BPMN to the process algebra (CSP) is presented in [267]. The
purpose of these transformations is to provide an execution semantics of BPMN and
to enable the analysis of BPMN diagrams using existing static analysis techniques.
Transformations from BPMN to BPEL have also been studied in prior work [187]
and they have been implemented in various tools, including commercial tools. The
purpose of the BPMN to BPEL transformation is to enable the execution of BPMN
diagrams using existing BPEL orchestration engines. However, in some cases, this
transformation may lead to complex and unreadable BPEL process definitions that
may subsequently be hard to debug and to maintain. In contrast, the transformation
from BPMN to YAWL discussed in this chapter is very direct. One could conceive
collaborative modeling environments in which some stakeholders will see a BPMN
view of a process model, while others would see an equivalent YAWL net. Given the
direct relations between BPMN and YAWL, it would be possible to maintain these
views “in sync.” The only issue would be to restrict changes in the YAWL view so
that any change can be propagated into the BPMN view.