[rules-users] drools quirks

Sat Sep 18 15:15:28 EDT 2010

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Thanks Wolfgang,
  The no-loop attribute fixed my crazy looping problem and your answer
on knowing EVERYTHING the rule might ever have to do is a good
answer.  I guess there will be a certain amount of trust I have to
place in the users.  In the meantime, are there any books on drools
you reccomend (other than the manual)?

On 09/18/2010 01:24 PM, Wolfgang Laun wrote:
>
> On 18 September 2010 17:19, Matt Young <solid at youngdev.net
> <mailto:solid at youngdev.net>> wrote:
>
> ksession.execute(document);
>
>
> I suppose this inserts a single EventRulesDocument object.
>
>
> Here is the rule:
>
> package sample1
>
> import sample1.bean.EventRulesDocument; import
> sample1.bean.Transaction; import java.util.Date;
>
> rule "Standard processing rule"
>
> when $doc : EventRulesDocument( ruleRun == false &&  # <<<
> Wolfgang event != null && event.eventType.eventCode == "testevent"
> ) then
>
> Transaction t = new Transaction(); t.setEvent($doc.getEvent());
> t.setCategory($doc.getEvent().getEventType().getEventCode());
> t.setTransactionAmount($doc.getEvent().getEventAmount());
> t.setCreateTs(new Date());
> t.setUsername($doc.getEvent().getCreateUsername());
>
> modify( $doc ) {setRuleRun(true)};   # Wolfgang <<<<<< modify( $doc
> ) {addSaveTransaction(t)} ;
>
>
> Here it is: You modify the EventRulesDocument, which causes
> another activation of the same rule. One way of avoiding this is to
> add the rule attribute no-loop true; another one is what you've
> done.
>
> This isn't going to be your only rule?? This one alone is scant
> reason to use an RBS.
>
> fireAllRules( 1 ) isn't a good remedy, because (normally) a
> Production System relies on repeated firings, depending on rules
> and facts.
>
>
> end
>
>
> Additionally If any code in the world can be executed in the RHS,
> then why shouldn't I just use groovy as my rule language?  At
> least then, I can write math like such: def num1=95; def num2=90;
> def num3=num2/num1;
>
> and all of the math is done with big decimal objects.  IF there is
> no restriction on what can be done in the RHS what is the advantage
> to using a rules engine over a general purpose scripting language?
>
>
> There is a good section in the Drools Expert manual: "Why use a
> rule engine?". Don't do it just to be trendy ;-)
>
>
> The whole purpose for me choosing drools over Groovy eval() was
> that I didn't want to give my users the ability to spawn threads,
> allocate massive amounts of memory etc.
>
>
> Seriously now, a Domain Specifiy Language is the maximum corset you
> can force upon your users. But you'll have to know /all/ they /may
> /do, and all the conditions.
>
> -W
>
>
>
> On 09/18/2010 10:59 AM, Wolfgang Laun wrote:
>> On 18 September 2010 16:15, Matt Young <solid at youngdev.net
> <mailto:solid at youngdev.net>
>> <mailto:solid at youngdev.net <mailto:solid at youngdev.net>>> wrote:
>>
>> I just started using drools and this is my first time
>> implementing a rules engine.  Everything has been pretty smooth
>> but I have some quirks that I am not sure I can live with. 1) for
>> some
> reason,
>> if I execute the Knowlege session against an object, the
>> knowlege session never returns.
>>
>>
>>> I don't understand the term "execute...against an object". Do
>>> you mean that you have just one fact inserted befor you call
>>> fireAllRules()?
>>
>>> But anyway, this call not returning is almost certainly due to
>>> a loop in your rules, or have you made sure that they don't?
>>
>>
>> The only way I can get the ksession to return is to make IF
> (obj ==
>> null) part of the LHS and make
>> modify($input){setObj("complete")} part of the RHS This seems
>> like a deficiency since I have users writing their own rules, I
>> can see them forgetting this requirement.
>>
>>
>>> Writing rules is programming, no holds barred :-)
>>
>>> Look into Domain Specific Languages (DSL) as a cushion for the
>>> unwary.
>>
>>
>>
>> IS their a way to get the ksession to firerules only once?  If
>> so what does that look like?
>>
>>
>>> kSession.fireUntilHalt( 1 );
>>
>>
>> 2) It seems that any code I want can be executed in the RHS.  I
>> could literally execute something like the following in the RHS.
>>
>> byte[] b= new byte[10000000000000]; // Really big memory waste
>>
>> I also could just start a bunch of threads.
>>
>> The point is that I am intending to let the users write their
>> own rules but I can't do that if there are no restrictions on
>> how/what can be done inside the rules.  Any suggestions?  Are
>> there sandboxes or filters I can activate to restrict the RHS?
>>
>>
>>> Again: DSL.
>>
>>> Also, Rule Templates might be the starting point; it'd depend
>>> on what the may be allowed to do.
>>
>>> And what about good old-fashioned training?!
>>
>>> -W
>>
>>
>
> _______________________________________________ rules-users mailing
> list rules-users at lists.jboss.org
> <mailto:rules-users at lists.jboss.org>
> <mailto:rules-users at lists.jboss.org
> <mailto:rules-users at lists.jboss.org>>
> https://lists.jboss.org/mailman/listinfo/rules-users
>
>
>
>
>> _______________________________________________ rules-users
> mailing
>> list rules-users at lists.jboss.org
> <mailto:rules-users at lists.jboss.org>
>> https://lists.jboss.org/mailman/listinfo/rules-users
>
> - -- Cheers, Matt Young solid at youngdev.net
> <mailto:solid at youngdev.net> http://youngdev.net
>
> -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
>
> iQIcBAEBAgAGBQJMlNibAAoJEO5jycTTPEzc5uUQAJ1ku8gHoMLKVf8Y64llpAIK
> n0jKIkUbvs44AKEIq3bv6hM9x1sqskJ8zXqV1FmrJ6FLeILPpqc2El/QVQkY5YVh
> xKbFU6FOEhjhR8lcOCLu9Pt7Bxh0eWcdpJsGN/4PAaUDmS4f4UEj1OH5TRdxKWvn
> sw4nR3N+KxJKIOSsvW2IXcULdiy3rQm3YYs7GvQmiPbrgLzSYxGhPBsy/W18P75f
> xefctOjsuFt89BP3BUjzut3Hu7Q/1bhCETwV0sGg2c7oynbgiRY96/tGCvdVYroM
> 1h0vcv2zwnxmJcWx0ipzf8Tmwkliy2MY2ITcgIrcZP21FY3AxcvU2SvKgrFj9C+N
> 68c8PM0m0ZMHB5AKNB+2AeI5/52AP9K9LwLxAZu0EaXpS62E/ZIQ0F7TSGNmZi2o
> 7bYLl1C7BH2JV9+Zfxy4B0kfOzMeoT+cBscA8YUQ8twmy2QHGQ5CqUdRvlaL1Ktc
> WfWFLjsuJm09D+iCS8wg+V4DcqeTH7NMPicodUFeXtxzvsZzmXoXRv+OIYGp2zla
> uBuMgQGYMI651RTWWvLN4IyUlbwJPWMg974LcBir1SOogyfdU/2A4s4kPMh2CK4v
> X/UCK4J/dX1DYGmrpEUqOkyHiajPqfbA7yed+K+SbfUbNwGvzl6q4gaE2eM+w1pz
> bT9rOgr9PFpyiL2EPdut =UoOJ -----END PGP SIGNATURE-----
>
> _______________________________________________ rules-users mailing
> list rules-users at lists.jboss.org
> <mailto:rules-users at lists.jboss.org>
> https://lists.jboss.org/mailman/listinfo/rules-users
>
>
>
>
> _______________________________________________ rules-users mailing
> list rules-users at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/rules-users

- -- 
Cheers,
Matt Young
solid at youngdev.net
http://youngdev.net

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
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=T8nI
-----END PGP SIGNATURE-----