Sebastian Schneider wrote:
Hello folks,
I am not sure if it was mentioned already but it is certainly a point I
forgot when sending my feedback to the list. One thing I pretty much
like about jBPM 4.x is its clear and easy-to-understand API. The API is
pretty well designed and should certainly serve as a template for a
future one. I don't know anything about the API of Drools Flow. Maybe
somebody can shed some light on this aspect.
Drools Flow uses a knowledge-based API (so the API is similar, whether
you're using
processes or rules), but the basic API to interact with the process
engine can be found here:
http://anonsvn.jboss.org/repos/labs/labs/jbossrules/trunk/drools-api/src/...
An additional aspect which came to my mind recently is that the
process
engine should be kept DBMS-independent and have a QA-matrix which takes
into account most of the widespread DB-systems.
That's exactly what I was thinking when I was writing about continuous
integration and QA.
With respect to embeddability one could even think of using the JPA
to
enable the usage of compatible object-relational mappers. If I remember
correctly there have been request or questions on the forum regularly
about using a different persistence framework than Hibernate.
Yes, the use of Hibernate directly instead of JPA is a known issue.
Kris