Damn, I can't! Simple rule rule "TEST" dialect "mvel" when s : String() then System.out.println(s); end Java Test StatefulKnowledgeSession s = ... s.insert("ONE"); s.insert("TWO"); s.fireAllRules(); s.dispose(); Output TWO What the...??? Using v5.4.0.Final -- View this message in context: http://drools.46999.n3.nabble.com/Evaluate-rules-for-multiple-facts-of-th... Sent from the Drools: User forum mailing list archive at Nabble.com. _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

Can you suggest a way to diagnose what's going wrong? I added DebugAgendaEventListener and DebugWorkingMemoryEventListener, but the output is too confusing for me -- View this message in context: http://drools.46999.n3.nabble.com/Evaluate-rules-for-multiple-facts-of-th... Sent from the Drools: User forum mailing list archive at Nabble.com. _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

I'll try to make it; in the meanwhile, do you find this meaningful? ==>[ActivationCreatedEvent: getActivation()=[Activation rule=TEST, act#=0, salience=0, tuple=[fact 0:1:30100654:78406:1:DEFAULT:ONE] ], getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c ] ==>[ObjectInsertedEventImpl: getFactHandle()=[fact 0:1:30100654:78406:1:DEFAULT:ONE], getObject()=ONE, getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c, getPropagationContext()=PropagationContextImpl [activeActivations=0, dormantActivations=0, entryPoint=EntryPoint::DEFAULT, factHandle=[fact 0:1:30100654:78406:1:DEFAULT:ONE], leftTuple=null, originOffset=-1, propagationNumber=3, rule=null, type=0]] ==>[ActivationCreatedEvent: getActivation()=[Activation rule=TEST, act#=1, salience=0, tuple=[fact 0:2:30649454:83500:2:DEFAULT:TWO] ], getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c ] ==>[ObjectInsertedEventImpl: getFactHandle()=[fact 0:2:30649454:83500:2:DEFAULT:TWO], getObject()=TWO, getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c, getPropagationContext()=PropagationContextImpl [activeActivations=1, dormantActivations=0, entryPoint=EntryPoint::DEFAULT, factHandle=[fact 0:2:30649454:83500:2:DEFAULT:TWO], leftTuple=null, originOffset=-1, propagationNumber=4, rule=null, type=0]] TWO ==>[BeforeActivationFiredEvent: getActivation()=[Activation rule=TEST, act#=1, salience=0, tuple=[fact 0:2:30649454:83500:2:DEFAULT:TWO] ], getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c ] ==>[ActivationCancelledEvent: getCause()=CLEAR, getActivation()=[Activation rule=TEST, act#=0, salience=0, tuple=[fact 0:1:30100654:78406:1:DEFAULT:ONE] ], getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c ] ==>[AfterActivationFiredEvent: getActivation()=[Activation rule=TEST, act#=1, salience=0, tuple=[fact 0:2:30649454:83500:2:DEFAULT:TWO] ], getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c ] Direct Line, l'assicurazione online con il famoso Telefono Rosso e il Mouse sulle Ruote. Nel 1985 Direct Line ha rivoluzionato il mercato nel Regno Unito lanciando per prima la vendita diretta di polizze auto. In Italia ha vinto nel 2008 il primo premio per i clienti piu' soddisfatti. www.directline.it RESTRICTED: This e-mail (including any linked or attached documents) is intended for the addressee/s only and may contain confidential, proprietary or legally privileged information. Please do not distribute further without prior permission from the author -- View this message in context: http://drools.46999.n3.nabble.com/Evaluate-rules-for-multiple-facts-of-th... Sent from the Drools: User forum mailing list archive at Nabble.com. _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

Actually, incorrect, I misread the output, sorry. On Fri, Feb 8, 2013 at 9:12 AM, Jeremy Ary <jeremy.ary(a)gmail.com&gt; wrote: > It appears you have a rule named CLEAR not included in the original post > that's cancelling the activation that should println ONE, can you share > this rule with us? > > > On Fri, Feb 8, 2013 at 9:10 AM, pdario <dario.piantanida(a)gmail.com&gt;wrote: > >> I'll try to make it; in the meanwhile, do you find this meaningful? >> >> ==>[ActivationCreatedEvent: getActivation()=[Activation rule=TEST, >> act#=0, >> salience=0, tuple=[fact 0:1:30100654:78406:1:DEFAULT:ONE] >> ], >> getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c >> ] >> ==>[ObjectInsertedEventImpl: getFactHandle()=[fact >> 0:1:30100654:78406:1:DEFAULT:ONE], getObject()=ONE, >> getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c >> , >> getPropagationContext()=PropagationContextImpl [activeActivations=0, >> dormantActivations=0, entryPoint=EntryPoint::DEFAULT, factHandle=[fact >> 0:1:30100654:78406:1:DEFAULT:ONE], leftTuple=null, originOffset=-1, >> propagationNumber=3, rule=null, type=0]] >> ==>[ActivationCreatedEvent: getActivation()=[Activation rule=TEST, >> act#=1, >> salience=0, tuple=[fact 0:2:30649454:83500:2:DEFAULT:TWO] >> ], >> getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c >> ] >> ==>[ObjectInsertedEventImpl: getFactHandle()=[fact >> 0:2:30649454:83500:2:DEFAULT:TWO], getObject()=TWO, >> getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c >> , >> getPropagationContext()=PropagationContextImpl [activeActivations=1, >> dormantActivations=0, entryPoint=EntryPoint::DEFAULT, factHandle=[fact >> 0:2:30649454:83500:2:DEFAULT:TWO], leftTuple=null, originOffset=-1, >> propagationNumber=4, rule=null, type=0]] >> TWO >> ==>[BeforeActivationFiredEvent: getActivation()=[Activation rule=TEST, >> act#=1, salience=0, tuple=[fact 0:2:30649454:83500:2:DEFAULT:TWO] >> ], >> getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c >> ] >> ==>[ActivationCancelledEvent: getCause()=CLEAR, >> getActivation()=[Activation >> rule=TEST, act#=0, salience=0, tuple=[fact >> 0:1:30100654:78406:1:DEFAULT:ONE] >> ], >> getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c >> ] >> ==>[AfterActivationFiredEvent: getActivation()=[Activation rule=TEST, >> act#=1, salience=0, tuple=[fact 0:2:30649454:83500:2:DEFAULT:TWO] >> ], >> getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c >> ] >> Direct Line, l'assicurazione online con il famoso Telefono Rosso e il >> Mouse >> sulle Ruote. Nel 1985 Direct Line ha rivoluzionato il mercato nel Regno >> Unito lanciando per prima la vendita diretta di polizze auto. In Italia >> ha >> vinto nel 2008 il primo premio per i clienti piu' soddisfatti. >> www.directline.it RESTRICTED: This e-mail (including any linked or >> attached >> documents) is intended for the addressee/s only and may contain >> confidential, proprietary or legally privileged information. Please do >> not >> distribute further without prior permission from the author >> >> >> >> >> -- >> View this message in context: >> http://drools.46999.n3.nabble.com/Evaluate-rules-for-multiple-facts-of-th... >> Sent from the Drools: User forum mailing list archive at Nabble.com. >> _______________________________________________ >> rules-users mailing list >> rules-users(a)lists.jboss.org >> https://lists.jboss.org/mailman/listinfo/rules-users >> > >

Jeremy is right. In the log you can see 2 activations (one for each fact you insert), then the execution of one of the activations and then the cancellation of the other. How many rules do you have in your kbase? Is your rule as simple as the one you posted? Do you have more than just 1 condition in your LHS? Are you doing anything else other than the system.out in your RHS? Best Regards, XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX Esteban Aliverti - Blog @ http://ilesteban.wordpress.com On Fri, Feb 8, 2013 at 4:17 PM, Jeremy Ary <jeremy.ary(a)gmail.com&gt; wrote: > ==>[ActivationCancelledEvent: getCause()=CLEAR, > getActivation()=[Activation > rule=TEST, act#=0, salience=0, tuple=[fact 0:1:30100654:78406:1:DEFAULT: > ONE] > > There's definitely something cancelling the activation, but it's not > apparent to me right-out what's causing it based on the information > provided so far. To Estaban's comment, would it be possible to share an > isolated example as I can't help but to think there's something missing > from what we've seen so far that would cause this to occur. > > > On Fri, Feb 8, 2013 at 9:14 AM, Jeremy Ary <jeremy.ary(a)gmail.com&gt; wrote: > >> Actually, incorrect, I misread the output, sorry. >> >> >> On Fri, Feb 8, 2013 at 9:12 AM, Jeremy Ary <jeremy.ary(a)gmail.com&gt; wrote: >> >>> It appears you have a rule named CLEAR not included in the original >>> post >>> that's cancelling the activation that should println ONE, can you share >>> this rule with us? >>> >>> >>> On Fri, Feb 8, 2013 at 9:10 AM, pdario >>> <dario.piantanida(a)gmail.com&gt;wrote: >>> >>>> I'll try to make it; in the meanwhile, do you find this meaningful? >>>> >>>> ==>[ActivationCreatedEvent: getActivation()=[Activation rule=TEST, >>>> act#=0, >>>> salience=0, tuple=[fact 0:1:30100654:78406:1:DEFAULT:ONE] >>>> ], >>>> getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c >>>> ] >>>> ==>[ObjectInsertedEventImpl: getFactHandle()=[fact >>>> 0:1:30100654:78406:1:DEFAULT:ONE], getObject()=ONE, >>>> getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c >>>> , >>>> getPropagationContext()=PropagationContextImpl [activeActivations=0, >>>> dormantActivations=0, entryPoint=EntryPoint::DEFAULT, factHandle=[fact >>>> 0:1:30100654:78406:1:DEFAULT:ONE], leftTuple=null, originOffset=-1, >>>> propagationNumber=3, rule=null, type=0]] >>>> ==>[ActivationCreatedEvent: getActivation()=[Activation rule=TEST, >>>> act#=1, >>>> salience=0, tuple=[fact 0:2:30649454:83500:2:DEFAULT:TWO] >>>> ], >>>> getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c >>>> ] >>>> ==>[ObjectInsertedEventImpl: getFactHandle()=[fact >>>> 0:2:30649454:83500:2:DEFAULT:TWO], getObject()=TWO, >>>> getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c >>>> , >>>> getPropagationContext()=PropagationContextImpl [activeActivations=1, >>>> dormantActivations=0, entryPoint=EntryPoint::DEFAULT, factHandle=[fact >>>> 0:2:30649454:83500:2:DEFAULT:TWO], leftTuple=null, originOffset=-1, >>>> propagationNumber=4, rule=null, type=0]] >>>> TWO >>>> ==>[BeforeActivationFiredEvent: getActivation()=[Activation >>>> rule=TEST, >>>> act#=1, salience=0, tuple=[fact 0:2:30649454:83500:2:DEFAULT:TWO] >>>> ], >>>> getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c >>>> ] >>>> ==>[ActivationCancelledEvent: getCause()=CLEAR, >>>> getActivation()=[Activation >>>> rule=TEST, act#=0, salience=0, tuple=[fact >>>> 0:1:30100654:78406:1:DEFAULT:ONE] >>>> ], >>>> getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c >>>> ] >>>> ==>[AfterActivationFiredEvent: getActivation()=[Activation rule=TEST, >>>> act#=1, salience=0, tuple=[fact 0:2:30649454:83500:2:DEFAULT:TWO] >>>> ], >>>> getKnowledgeRuntime()=org.drools.impl.StatefulKnowledgeSessionImpl@4703c >>>> ] >>>> Direct Line, l'assicurazione online con il famoso Telefono Rosso e il >>>> Mouse >>>> sulle Ruote. Nel 1985 Direct Line ha rivoluzionato il mercato nel >>>> Regno >>>> Unito lanciando per prima la vendita diretta di polizze auto. In >>>> Italia >>>> ha >>>> vinto nel 2008 il primo premio per i clienti piu' soddisfatti. >>>> www.directline.it RESTRICTED: This e-mail (including any linked or >>>> attached >>>> documents) is intended for the addressee/s only and may contain >>>> confidential, proprietary or legally privileged information. Please do >>>> not >>>> distribute further without prior permission from the author >>>> >>>> >>>> >>>> >>>> -- >>>> View this message in context: >>>> http://drools.46999.n3.nabble.com/Evaluate-rules-for-multiple-facts-of-th... >>>> Sent from the Drools: User forum mailing list archive at Nabble.com. >>>> _______________________________________________ >>>> rules-users mailing list >>>> rules-users(a)lists.jboss.org >>>> https://lists.jboss.org/mailman/listinfo/rules-users >>>> >>> >>> >> > > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users >

Can anybody help? -- View this message in context: http://drools.46999.n3.nabble.com/Evaluate-rules-for-multiple-facts-of-th... Sent from the Drools: User forum mailing list archive at Nabble.com. _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

Well, so how can I make this rule fire for each fact? My scenario is: - different rules in the same activation-group to assign discount to the shopping cart - different activation-group to manage different discount policies - each rule has different salience to be sure only the most relevant one fires in each group I would like to apply discounts to several carts with just one call to the rule engine. Is it possible or do I have to fireAllRules with one one insertion at a time? -- View this message in context: http://drools.46999.n3.nabble.com/Evaluate-rules-for-multiple-facts-of-th... Sent from the Drools: User forum mailing list archive at Nabble.com. _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

@FrankVhh It looks ok to have a "no other reductions are applied" as a blocker, I find cumbersome to repeat all the conditions of other rules negated. Anyway, I'll try to summarize my situation: rule "20% discount" salience 0 effective 1st Feb -> 28th Feb activaction-group "online discounts" when paying with credicard cart contains special product A then cart.setDiscount(20) rule "15% discount" salience -10 effective 1st Feb -> 28th Feb activaction-group "online discounts" when paying with credicard cart contains special product B then cart.setDiscount(15) rule "10% discount" salience -20 activaction-group "online discounts" when paying with credicard then cart.setDiscount(10) rule "No discount" salience -30 activaction-group "online discounts" when paying not with creditcard then cart.setDiscount(0) I thought activation-group / salience combination was the easiest and more natural way to follow business rules. How would you suggest to change these?

@ part was just for your remark on cumbersomeness... I'm now reading the linked whitepaper. It's not easy for me to see how your example is easier than mine solution. I acknowledge salience redistribution would be a problem, that's why I kept them separated and kept salience values meaningful only within each activation-group. At the moment, anyway, the activation-group is stopping me from evaluating rules for a bunch of facts in only one run, so I have to find a different solution (supposing that a multiple single fact activation is not efficient, something I'm not sure). I'll read the whitepaper and see if agenda-group would better fit in my case. Thank you all. :-) -- View this message in context: http://drools.46999.n3.nabble.com/Evaluate-rules-for-multiple-facts-of-th... Sent from the Drools: User forum mailing list archive at Nabble.com. _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

I'm trying to understand this "Param" pattern, but I miss the Java calling for it: will I need to put the effective values in the Java code instead of the rule file? Can you elaborate a little about this? Take the Example 3 of your whitepaper: where are literals 0, 10 and "red" from the Example 2?

-- View this message in context: http://drools.46999.n3.nabble.com/Evaluate-rules-for-multiple-facts-of-th... Sent from the Drools: User forum mailing list archive at Nabble.com. _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

Hello, the "tinting" example from the whitepaper assumes each Tint fits exactly one Param, right? This is not my case, as a Cart can match different discounts and we want to assign the highest possible.

Is the pattern still suitable?

Another thing I see is Param is evaluating plain attributes from the Tints, but (not sure if this is the correct pattern) I'm using this to find if a Cart contains a specific item when Cart ($items : buyingItems, payment == Card....) i : Item (code == XXXXX) from $items then

laune wrote > There is always just one fit; it just seems that there are multiple > fits - when you define your individual cases sloppily. Well, this is an assumption to get a completely disjointed classification, but this is not my case. I have a first level discount if you pay online and a second level discount if you pay online AND buy a special product. So, the second is true only if the first is.

laune wrote >> when >> Cart ($items : buyingItems, payment == Card....) >> i : Item (code == XXXXX) from $items >> then > > I don't see why "XXXXX" shouldn't come from a parameter fact. > > As a general strategy, I try to avoid the "on-the-fly" objects > available with "from" and prefer these subordinate items to be > properly inserted facts. Usually this simplifies writing rules. If I simply put several carts and the whole items form them into memory, it would create a mess of carts and items, won't it? What will keep each item with its own cart?

laune wrote >> I have a first level discount if you pay online and a second level >> discount >> if you pay online AND buy a special product. >> So, the second is true only if the first is. > > Not *quite* correct: you have 1st level discount if you pay online and > DO NOT buy a special product. Well, this is strictly true, but it seems to me it is easier to have a fallback rules instead of specifying in the first rule all the negated conditions for more specific rules... Otherwise, I'll have to add to the first rule "AND boughtProduct does not contain spcial product A, B, C, D...."

and have to maintain this as we change the special products every month. What's the best practice for this? -- View this message in context: http://drools.46999.n3.nabble.com/Evaluate-rules-for-multiple-facts-of-th... Sent from the Drools: User forum mailing list archive at Nabble.com. _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

laune wrote > I'm confident that there's a clean and manageable way of preparing > this data to work in combination with a small and constant set of > rules. I refrain from elaborating this in all details without knowing > all of the requirements. I'll try to summarize them as properly as I can... Rule 0: default condition If cart is paid "offline", it has 0% discount Rule 1: online payment discount If cart is paid "online", it has 10% discount These two are fixed and always valid. Then we monthly decide a list of items that grant different discounts on the whole cart over the 10% of Rule 1 (i.e. Rule 1 must be true to get further discount). For example: Rule 2: valid for February 2013, item A If cart is paid "online" and contains item A, it has 15% discount Rule 3: valid for February 2013, item B If cart is paid "online" and contains item B, it has 17% discount Next month, both items and discount levels will change We need the firt Rule 0 because the user can change the payment method, so we have to reset the discount to 0. Ideally, we would like to left the Rule 0 and 1 unchanged (as they are basic rule) and manage the monthly rules without impact on them. To make things worse, our business people not only rely on items contained in the cart, but each month they decide a different attribute from the cart or user to be taken into account for special discount, so that the examples above may read like: Rule 2: valid for February 2013, item A If cart is paid "online" and contains item A and the user is older than 30, it has 15% discount Rule 3: valid for February 2013, item B If cart is paid "online" and contains item B and the cart is shipped within Europe, it has 17% discount As you remember, we used salience and activation groups, but this prevents us from applying rules to a bunch of carts, as the rule activation cancels other activations. Do you see a cleaner design? Thank you a lot for support. -- View this message in context: http://drools.46999.n3.nabble.com/Evaluate-rules-for-multiple-facts-of-th... Sent from the Drools: User forum mailing list archive at Nabble.com. _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

A lot to think about, it will take a little time to dissect this. Thank you! -- View this message in context: http://drools.46999.n3.nabble.com/Evaluate-rules-for-multiple-facts-of-th... Sent from the Drools: User forum mailing list archive at Nabble.com. _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

I'm sorry, it's not clear. Do you mean I have to add a condition in each rule to check whether other discount have been already applied? Even so, I can I control the preferred order of the discounts without salience? Do I have to put negated conditions of each rule?? It looks cumbersome... -- View this message in context: http://drools.46999.n3.nabble.com/Evaluate-rules-for-multiple-facts-of-th... Sent from the Drools: User forum mailing list archive at Nabble.com. _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users