I've recently conducted due diligence studies on commercial and
open-source BRMSs for three large companies in Asia-Pacific (and am
likely to start another one in a few weeks).
Whilst your requirements are quite high level, drools probably meets
most of them. Likely you'll end up with a handful that will, and from
that point your specific requirements will dictate which one is the
Contact me directly if you want to discuss further.
Quoting Mark Proctor <mproctor(a)codehaus.org>:
Stephen Kestle wrote:
> Has anybody tried to evaluate the different rules engines in the
> interests of using it in their companies' technology stack? I've
> got to do some due diligence, so need to evaluate the others.
> Specifically, we're looking at a dynamic data model (not OO), and
> so that needs to be supported in some way (which I am heavily
> evaluating for drools :)). There are plenty of other criteria, but
> the show stoppers are:
> * unable to support a dynamic data model (being able to type
> different structures with the same object base, and to a lesser
> degree being able to compose those types into other dynamic types)
in drools 5.0, along with XSD/JAXB definitions and dataloaders and
Smooks ETL dataloader.
> * Lack of support (either community or commercial)
You can get support from JBoss, a Division of RedHat:
> * Lack of rule management (activation and expiry dates for rules)
rules have attributes to specify dateEffective and dateExpires as attributes.
> Also high on the list are:
> * DSL support
> * rule composition and reuse (I don't really know much about how
> this this concept plays out with rules yet)
> * Testing: the person creating the rules can write the tests, in
> the same designer
> * Auditing and understanding how the engine got to the state it's in
> So far I've found about 15-20 engines. Hopefully most of them will
> fall out of the early criteria :)
This message was sent using IMP, the Internet Messaging Program.