[rules-users] RuleFlows and Working Memory..

[mmquelo massi]

Hi there,

I have got a couple of questions about ruleflows and woking memory.

What does happen each time I modify a fact within a ruleflow node without
calling update(fact)?

What does happen between the node where I
apllied the modifications and the next activation node?

I guess that between one ruleflow node and the next one, all the modified
facts (in the previous node) are automatically updated (without explicitly
call "update()")
so that once the next node has been activated,
the corresponding shadow facts result synchronized with the actual facts.

Am I wrong?

Should I call "update()" instead?

My scenario...

I have got one node where I modify a fact, but I do not
want to call the update() because otherwise I would generate
a "ping-pong" loop between two decision tables that are
placed in the same rule flow node. I think I could use
the "no-active" attribute but I can't.... cause I can't specify
it in a Decision table.

Another workaround could be the "clearAgenda()" method
but It is not accesible from the "drools" (the helper) object...
How can I call it from the decision table or within an "action" ruleflow
node?

Maybe there is a third solution.... I should "update()" all the
fact from an action node ...but in this case I would need
to call an "update()" method that allowed me to update
"all the facts of a given Class", Is That Possible in drools?

I took a look on the API docs but some methods are not explained.

Looking forward to hearing from you!

Thank You again!

Massi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.jboss.org/pipermail/rules-users/attachments/20080219/5e2f9790/attachment.html