headers
Mark Proctor | 1 Oct 2005 15:29

[drools-user] Drools 2.1 final Released

The Drools <http://drools.org> development team is proud to announce the 
release of Drools 2.1 final.

Highlights include:
Excel Decision Tables
Spring and POJO integration
Xor Group
Improved exception handling via JDK1.4 Nested Exceptions

Please see the Release Notes <http://drools.org/Release+Notes> for more 
details on this release.
The Project Downloads <http://drools.org/Project+Downloads> page 
contains full information on obtaining Drools.

Happy Drooling,

The Drools Development Team.
Permalink | Reply | 1 comment |
headers
Niclas Hedhman | 2 Oct 2005 07:25
Picon
Favicon

Re: [drools-user] java.net.MalformedURLException in org.drools.io.RuleSetReader.read(...)

On Friday 30 September 2005 17:16, Francesco Russo wrote:
> So it makes me think something different happens when compiling from
> within Eclipse and with Ant.
> Should I post this problem on the Ant mailing list?! It seems to me I
> should...

If you have not located the problem, the Ant peeps will not be able to help 
you. And personally, I doubt that this is an Ant problem (> a million 
users...) and more likely an oversight at your end (1 user...).

1. Narrow down the file it happens in. I.e. what you think is being read may 
not be what is being read. (Common mistake). I get the impression from the 
stacktrace that you rely on some other package to assemble the rulesets, so 
perhaps you are getting more than you think.

2. The following 2 lines seems to indicate that Entity and/or DTD processing 
is going on when it happens. The drools ruleset you provided should probably 
not get to this section at all, which would re-inforce the assumption above. 
com.sun.org.apache.xerces.internal.impl.XMLEntityManager.setupCurrentEntity() 
com.sun.org.apache.xerces.internal.impl.XMLVersionDetector.determineDocVersion()

Good Luck
Cheers
Permalink | Reply | 1 comment |
headers
Francesco Russo | 3 Oct 2005 09:23
Picon
Favicon

Re: [drools-user] java.net.MalformedURLException in org.drools.io.RuleSetReader.read(...)

Thanks for your hints.
As you assumed it was not an Ant problem: I found out that the classes 
generated by Eclipse 3.1 differ from the ones generated by my java 5 
compiler. Now I'm pretty sure the problem resides there, in the 
compilation step carried out by Eclipse. I've already found other posts 
on the Internet talking about problems concerning the classes generated 
by the Eclipse compiler (I even didn't thought Eclpise used a compiler 
of its own!). 

I'll try and follow down this path...

Thanks,
Francesco

Niclas Hedhman wrote:

>On Friday 30 September 2005 17:16, Francesco Russo wrote:
>  
>
>>So it makes me think something different happens when compiling from
>>within Eclipse and with Ant.
>>Should I post this problem on the Ant mailing list?! It seems to me I
>>should...
>>    
>>
>
>If you have not located the problem, the Ant peeps will not be able to help 
>you. And personally, I doubt that this is an Ant problem (> a million 
>users...) and more likely an oversight at your end (1 user...).
>
(Continue reading)

Permalink | Reply | 0 comments |
headers
Geoffrey Wiseman | 3 Oct 2005 16:27

RE: [drools-user] [jira] Created: (DROOLS-408) Xor Group so only one rule can fire for a specific group

Pardon me if this has already been discussed, I'm still catching up on
some old emails.  If it has, ignore this, and I'll find the answer
later.

Looking at XOR Groups, I've skimmed the release notes and this feature
description, but I'm still missing one fact -- does this just cancel
existing activations, or also prevent rules from being activated in
later assertions if they're part of an XOR group for which a rule has
already fired?

	- Geoffrey

Geoffrey Wiseman
direct: 416.646.7124

main: 416.646.7000 
fax: 416.646.7050 

Exchange Solutions Inc.
250 Yonge Street, 18th Floor
Toronto, ON M5B 2L7
www.exchangesolutions.com

-----Original Message-----
From: Mark Proctor (JIRA) [mailto:jira <at> codehaus.org] 
Sent: Sunday, September 25, 2005 1:52 PM
To: user <at> drools.codehaus.org
Subject: [drools-user] [jira] Created: (DROOLS-408) Xor Group so only
one rule can fire for a specific group
(Continue reading)

Permalink | Reply | 1 comment |
headers
Mark Proctor | 3 Oct 2005 22:16

Re: [drools-user] [jira] Created: (DROOLS-408) Xor Group so only one rule can fire for a specific group

Use the source Luke:
http://cvs.drools.codehaus.org/viewrep/drools/drools/drools-core/src/main/org/drools/reteoo/Agenda.java?r1=1.58&r2=1.59
As you can see it simply cancels any activations for that group.

Mark
Geoffrey Wiseman wrote:
> Pardon me if this has already been discussed, I'm still catching up on
> some old emails.  If it has, ignore this, and I'll find the answer
> later.
>
> Looking at XOR Groups, I've skimmed the release notes and this feature
> description, but I'm still missing one fact -- does this just cancel
> existing activations, or also prevent rules from being activated in
> later assertions if they're part of an XOR group for which a rule has
> already fired?
>
> 	- Geoffrey
>
> Geoffrey Wiseman
> direct: 416.646.7124
>
> main: 416.646.7000 
> fax: 416.646.7050 
>
> Exchange Solutions Inc.
> 250 Yonge Street, 18th Floor
> Toronto, ON M5B 2L7
> www.exchangesolutions.com
>
>
(Continue reading)

Permalink | Reply | 0 comments |
headers
Mark Stang | 5 Oct 2005 06:06

[drools-user] Drools as an IOC

Hi,
I posted this on the IRC, but it was late at night, so I thought I would
ask the list.

<mstang> hi
<mstang> I am considering using drools as an IOC device.
<mstang> One way to use drools, seems to be a kind of engine where you
feed the working memory data and then say "go"
<mstang> What I was thinking of was having events "happen" and then have
those events added as facts and then letting drools respond
<mstang> Drools never returns, it just sits there and processes.
<mstang> Anyone have any thoughts?
<mstang> I was looking at the petstore application and it seems to
"gather" data and then feed it into drools.
<mstang> It seems to me that it could react to button clicks and process
the data real-time, rather than only at the end.
<bob> it should return
<bob> particularly from assertObject(...)
<bob> fireAllRules() could go circular, depending on what your rules do

My basic question is can I have rules that are running permanently.  It
seems that the petstore demo shows that once the data has been acquired,
it is processed.  What I want to do is have the rules process
continually as events arrive.  It seems that the rules engine is outside
of the process.  Only occasionally, is data or the working memory
updated and then all the rules fire.   I guess I would like the rules
engine to be the event processor. 

Thoughts?
(Continue reading)

Permalink | Reply | 0 comments |
headers
Smith, Philip | 5 Oct 2005 07:06
Favicon

RE: [drools-user] Drools as an IOC

The application I am working on does this. The drools memory exists for
a very long time and I assert sessions and retract sessions. I have a
mechanism for the rules consequences to send events through the asserted
session objects which are then handled by the rest of the application.

It is feasible. I had worried about memory leaks in drools but I haven't
found any. I found plenty of memory leaks in my own application. But
that is another story.

Philip Smith
pasmith <at> ea.com
(650) 628-7941 (x87941)

-----Original Message-----
From: Mark Stang [mailto:mark <at> markjstang.com] 
Sent: Tuesday, October 04, 2005 9:06 PM
To: user <at> drools.codehaus.org
Subject: [drools-user] Drools as an IOC

Hi,
I posted this on the IRC, but it was late at night, so I thought I would
ask the list.

<mstang> hi
<mstang> I am considering using drools as an IOC device.
<mstang> One way to use drools, seems to be a kind of engine where you
feed the working memory data and then say "go"
<mstang> What I was thinking of was having events "happen" and then have
those events added as facts and then letting drools respond
<mstang> Drools never returns, it just sits there and processes.
(Continue reading)

Permalink | Reply | 1 comment |
headers
Michael Neale | 5 Oct 2005 11:37
Favicon

RE: [drools-user] Drools as an IOC

Yes WorkingMemory can certainly be long running - most rule engines would
consider that a likely use case. 

Also note that working memory itself is serializable - may come in handy for
long running workingmemory "sessions" so you can pick it up and continue at
any time without keeping it in memory.

-----Original Message-----
From: Smith, Philip [mailto:PASmith <at> ea.com] 
Sent: Wednesday, 5 October 2005 3:07 PM
To: user <at> drools.codehaus.org
Subject: RE: [drools-user] Drools as an IOC

The application I am working on does this. The drools memory exists for a
very long time and I assert sessions and retract sessions. I have a
mechanism for the rules consequences to send events through the asserted
session objects which are then handled by the rest of the application.

It is feasible. I had worried about memory leaks in drools but I haven't
found any. I found plenty of memory leaks in my own application. But that is
another story.

Philip Smith
pasmith <at> ea.com
(650) 628-7941 (x87941)

-----Original Message-----
From: Mark Stang [mailto:mark <at> markjstang.com]
Sent: Tuesday, October 04, 2005 9:06 PM
To: user <at> drools.codehaus.org
(Continue reading)

Permalink | Reply | 0 comments |
headers
Joel G. Rivera-González | 5 Oct 2005 22:53
Picon
Favicon

[drools-user] java.lang.OutOfMemoryError problems...

Hi...
i'm new to the group but not to drools...

we have been using drools since 2.0 beta 13...
we stop on beta 17 when we had too many memory
problems with it...

We have been using beta 17 since it came out las year,
and basicly we have been working arround the
java.lang.OutOfMemoryError erros by restarting the web
server every day.

as a backgroud of out project:
we are developing using java 1.4 on eclipse.
it is a web-based project and we are using struts.
last year we added drools to develop the business
rules.
the production machine is running SunOS 5.8 and as web
server we are using WebSphere 5.

The rules have grown to almost 800k (will be more that
800k by next release).

yesterday i decide to upgrade to released version 2.1
to see if the out of memory problem was related to the
beta version 17...but just after a few hours up..the
development server went down...the dev server is the
same but with less ram (1 gig of ram + vm)

we have tweek the java vm, and it helped a little but
(Continue reading)

Permalink | Reply | 0 comments |
headers
Geoffrey Wiseman | 5 Oct 2005 22:56

RE: [drools-user] java.lang.OutOfMemoryError problems...

Out of curiosity:  Have you done any profiling to find the source of the leak?  Are you caching the rule base, so
that it doesn't have to re-interpret and re-compile the rules all the time?

In essence, I haven't seen any of those problems, but since our largest rule deployment was with a large
array of annotated pojos, I can't easily compare the 'size' of the rules. 

	- Geoffrey

Geoffrey Wiseman
direct: 416.646.7124

main: 416.646.7000 
fax: 416.646.7050 

Exchange Solutions Inc.
250 Yonge Street, 18th Floor
Toronto, ON M5B 2L7
www.exchangesolutions.com

-----Original Message-----
From: Joel G. Rivera-González [mailto:el_pio <at> yahoo.com] 
Sent: Wednesday, October 05, 2005 4:54 PM
To: user <at> drools.codehaus.org
Subject: [drools-user] java.lang.OutOfMemoryError problems...

Hi...
i'm new to the group but not to drools...

we have been using drools since 2.0 beta 13...
we stop on beta 17 when we had too many memory problems with it...
(Continue reading)

Permalink | Reply | 8 comments |

Gmane