ofbiz has, though not implemented well, a rules engine as well as a workeffort and event(ECA, SECA, controller events) Model. the Project management also lends itself to rules.

I would suggest we work out a model rules engine that works around the way ofbiz works. That was the short comings of the other rules engines was the integration into the ofbiz model.


Maybe doing a review of rule engines in ofbiz now and why they did not get implemented would be a good exercise.

My efforts have been how to have a ui that a business person would understand, and generate ECA and SECA to accomplish Thier logic. The top level of this would be when they generate a business plan, it sets up the rules on how the business flow is done. This actually is base on compiler theory and some program that create web code for a particular type of web. I estimated to be about 1.5 man year.



Sam Hamilton sent the following on 7/28/2010 8:12 PM:
Hi Guys,

I have been thinking about problems we are soon going to start facing in 
marketing our OFBiz website and some of them can be solved using a rules 
engine, this then spilled over into other departments once I had read through 
the Drools website and saw how powerful it could be once implemented inside 
OFBiz.

I was thinking Drools http://jboss.org/drools/ as its ASL2 so we can have in 
trunk http://jboss.org/drools/downloads.html and also because of the GUI which 
could be plugged into OFBiz where ever we need rules - a few cool examples are 
here 
http://downloads.jboss.com/drools/docs/5.0.1.26597.FINAL/drools-introduction/html/ch02.html#d0e166
 and for the developers or data mining people using Birt etc they can use 
Eclipse.

The uses I can think of right now are:
1. Sending emails to customers - similar to auto responders or triggered emails 
.
2. Stock reordering - allowing different stock to be treated differently 
depending on for example how fast its selling or out of which warehouse the 
most is shipped.
3. Inserting marketing material into parcels - picking different customer 
segments and targeting them differently.
4. Shipping carrier upgrades / changes depending on either destination or price 
- like zappos do when they secretly upgrade customers to over night delivery.
5. Refunds to customers if something goes wrong (could automate some of the 
process) and then if its outside the rules it move to a human for review.

The real aim is that I want to move the process of rule changing out of the 
programers hands and put them into technical business people so that we can 
make fast changes to the business all done by the business users.

What does everyone else think?


Cheers
Sam


Reply via email to