Thank you Davide!

For now, as a work around I split this into two rules, 
1. rule : register the initial and second event and after (timer:2m) fire an 
internal event
2. rule : register internal event and then do the look behind (last 2 minutes) 
--> assert that the original event was valid

--> this is not nice code and might cause further problems, so I am really 
looking forward to get a bug fix ;) 

@W: I call "fireAllRules" every time I insert a new event. (is this bad? This 
seemed to works better (in terms of predictability (unit tests...) of outcome) 
than all my attempts to use "fireUntilHalt", w and w/o new Thread ...) 
especially with rules that use cron timers.



On 19.09.2013, at 10:23, Davide Sottara <[email protected]> wrote:

> It is a bug.
> The 2s delay is computed from the moment the rule is activated (by E2 at
> 12:01:50).
> rather than taking into account the correct "zero" timestamp ($event1).
> Thanks for reporting this, I'll open a JIRA
> Davide
> 
> On 09/18/2013 12:29 PM, Wolfgang Laun wrote:
>> Is the session being run via a single call to fireUntilHalt()?
>> -W
>> 
>> On 18/09/2013, Alexander Wolf <[email protected]> wrote:
>>> [Drools Version 5.5.0 Final]
>>> 
>>> Hey -
>>> 
>>> I got an event E1 that is only valid, if the latest event E2 following E1
>>> within 2 minutes has the value 1
>>> --> I want have the rule fire 2 minutes after E1 but only if E1 is valid
>>> 
>>> This is my rule:
>>> 
>>> rule "inform about E1"
>>> when
>>>     //event (T1) is the initial trigger
>>>     $event1 : Event(type == EventType.T1)
>>>     //there is an event (T2) with value 0 between 0,2m after doorClosed
>>>     $event2: Event(type == EventType.T2, value == 1, this after [0, 2m]
>>> $event1, $timestamp : timestamp)
>>>     //there is no newer event (T2) within the timeframe
>>>     not Measurement(type == EventType.T2, this after [0, 2m] $event1, 
>>> timestamp
>>>> $timestamp)
>>> then
>>>     //print info
>>>     log(drools, "E1 valid");
>>> end
>>> 
>>> An example of Events:
>>> 
>>> 12:00:00  - E1
>>> 12:01:00  - E2 ( value = 0 )
>>> 12:01:10  - E2 ( value = 0 )
>>> 12:01:40  - E2 ( value = 0 )
>>> 12:01:50  - E2 ( value = 1 )
>>> 12:02:10  - E2 ( value = 0 )
>>> 
>>> I would expect the output:   [log() does log the clock-time when my rule
>>> fires)
>>> 
>>> 12:02:00 E1 valid
>>> 
>>> But what I get is:
>>> 
>>> 12:03:50 E1 valid
>>> 
>>> 
>>> So I see that the late coming E2 (@12:02:10) is correctly ignored, the rule
>>> result is basically correct, but the rule is fired much to late.  (actually
>>> 2 minutes after $event2 and not as expected 2 minutes after $event1).
>>> 
>>> Could someone give me a hint what I did wrong?
>>> 
>>> Regards,
>>> Alex
>>> 
>> _______________________________________________
>> rules-users mailing list
>> [email protected]
>> https://lists.jboss.org/mailman/listinfo/rules-users
>> 
> 
> _______________________________________________
> rules-users mailing list
> [email protected]
> https://lists.jboss.org/mailman/listinfo/rules-users

Attachment: signature.asc
Description: Message signed with OpenPGP using GPGMail

_______________________________________________
rules-users mailing list
[email protected]
https://lists.jboss.org/mailman/listinfo/rules-users

Reply via email to