jason-users Mailing List for jason

Hi Vincent,

This isn't really a Jason question, it's just a matter of programming 
(so I probably shouldn't be answering). You could do it in various ways, 
including using imperative-style if-statements that are also available 
in Jason. One way to do it is as follows.

Best,

Rafael


/* Initial beliefs and rules */

// distance thrasholds for each goal
dtg(0,5,stop).
dtg(6,10,swerve).
//...

// fake, just to test
info(10,20,a,b,c,d,e).
onScanTank(a,b,12,22,c).

/* Plans */

+onScanTank(_,_,XT,YT,_) : info(X,Y,_,_,_,_,_)
   & D = math.abs(X-XT)+math.abs(Y-YT) // assuming Manhattan distance
   & dtg(L,H,G) & L <= D & H >= D
   <- !G.

+!stop <- .print("stopping").

+!swerve <- .print("swerving").

// ...



On 25/07/2010 16:40, Vin Baines wrote:
> Hi all,
>
> I'm sure this is an easy question, but I've run out of steam and can't
> solve it.
>
> I have two tanks in an environment, and depending on the distance
> between them, I want different plans to be selected.
>
> So, for the first 'self' tank, I have:
>
> +info(PosX, PosY, PosZ, Health, MainGunBullets, ChassiHeading,
> MainGunHeading) : true<- .print("info updated").
>
> Where PosX and PosY are the values i'm interested in, and this update
> occurs every half a second.
>
> When this guy detects another tank, then
>
> +onScanTank(T, E, X, Y, Z) : true<- .print("woo seen another driver").
>
> Which also gives me a X, Y values. Here I was planning on extending so
> that PosX-X and PosY-Y is calculated, and an appropriate rule is
> activated depending on this distance (e.g. stop, swerve, slow down).
>
> I'd really just like to get that logic working for testing. I think it
> would make more sense (once I've checked that works!) to create an
> agent for each detected other tank, where I would then put this
> distance checking logic, and those other agents can then use the
> send/tell KQML stuff to inform the 'master' agent of plan selection,
> or maybe just beliefs.
>
> So, can anyone offer some help? Its been a long day!!
>
> Thanks,
>
> Vincent
>
>
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Jason-users mailing list
> Jas...@li...
> https://lists.sourceforge.net/lists/listinfo/jason-users

Hi Herve,

If this is because there is a new version of Eclipse, I'm afraid you'll 
have to wait until somebody else fixes that and post i on the list; I 
rarely use the Eclipse plug-in and Jomi is on holidays without any 
internet connection for a week. In fact this plug-in was created by a 
former student of Jomi, not us, but I remember last time there was a new 
release of Eclipse Jomi managed to fix the Jason plug in.

Cheers,

Rafael

Hi all,

I'm sure this is an easy question, but I've run out of steam and can't  
solve it.

I have two tanks in an environment, and depending on the distance  
between them, I want different plans to be selected.

So, for the first 'self' tank, I have:

+info(PosX, PosY, PosZ, Health, MainGunBullets, ChassiHeading,  
MainGunHeading) : true <- .print("info updated").

Where PosX and PosY are the values i'm interested in, and this update  
occurs every half a second.

When this guy detects another tank, then

+onScanTank(T, E, X, Y, Z) : true <- .print("woo seen another driver").

Which also gives me a X, Y values. Here I was planning on extending so  
that PosX-X and PosY-Y is calculated, and an appropriate rule is  
activated depending on this distance (e.g. stop, swerve, slow down).

I'd really just like to get that logic working for testing. I think it  
would make more sense (once I've checked that works!) to create an  
agent for each detected other tank, where I would then put this  
distance checking logic, and those other agents can then use the  
send/tell KQML stuff to inform the 'master' agent of plan selection,  
or maybe just beliefs.

So, can anyone offer some help? Its been a long day!!

Thanks,

Vincent

Hi Jomi,
Thanks a lot for the reply. I tried to use the getPercept() method just to
see what's in the agent's percept list because addition of some percepts did
not work for me. Now I have found the problem with my code and it has got
nothing to do with adding percepts.
Thanks once again for the explanation, I did not know about it.

/Suralk


Jomi Hubner wrote:
> 
> Hi Suralk,
> 
> The getPercept is supposed to be called by the agent reasoning process and
> not to "consult" the current perceptions. It implements an optimisation
> strategy that returns null if nothing has changed since last call (so the
> reasoning does not need to revise beliefs based on perception). If you
> call this method in your environment class, you are indeed changing the
> way the reasoning "sees" the environment. 
> 
> I wish you have a good continuation of your experiments with Jason.
> 
> Jomi
> 
> ps.: I'm adding a method for environment programmers to consult the
> perceptions.
> 
> On 22/07/2010, at 02:29, suralk wrote:
> 
>> 
>> Hi All,
>> I am new to Jason programming and was playing with the examples first
>> which
>> worked fine. I have started to use Jason for my real research work and
>> have
>> faced a problem with the addPercept() method.
>> 
>> In the RetrievePercepts() method below, I try to call the addPercept()
>> method, using the parameters received. As shown, I try to look into the
>> set
>> of percepts of the agent before and after adding a percept. also I make
>> sure
>> i clean the percept list at the beginning. As expected, the call to
>> getPercepts after calling addPercept() shows the added percept. 
>> I want to change some of the perceived percepts in my architecture
>> customization class, and when I look at the percept list inside the
>> perceive() method, the percept list is empty!
>> At the moment, I try to add two percepts in different occasions and
>> adding
>> one of them works while the other one does not.
>> 
>> It could be some small mistake that I am doing, but I just can't figure
>> it
>> out. Hope someone would be able to help me out.
>> 
>> thanks
>> 
>> /Suralk
>> 
>> public void RetrievePercepts(Literal percept, String agName){    	
>>    	clearPercepts(agName);    	    	
>>    	List<Literal> per =this.getPercepts(agName);
>>    	try {
>>    		if (per != null) {
>>    			Iterator<Literal> ip = per.iterator();
>>    			while (ip.hasNext()) {    				
>>    				Literal p = ip.next();
>>    				String  ps = p.toString();   				
>>    				   				   				
>>    			}
>>    		}    		
>> 	    }  catch (Exception e) {
>> 	        
>> 	    }
>>    	addPercept(agName, percept);
>>    	
>>    	List<Literal> per1 =this.getPercepts(agName);
>>    	try {
>>    		if (per1 != null) {
>>    			Iterator<Literal> ip = per1.iterator();
>>    			while (ip.hasNext()) {    				
>>    				Literal p = ip.next();
>>    				String  ps = p.toString();		
>>    			}
>>    		}    		
>> 	    } catch (Exception e) {
>> 	        
>> 	    }
>>    }
>> -- 
>> View this message in context:
>> http://old.nabble.com/addPercept%28%29-method-does-not-update-the-percept-list-tp29233392p29233392.html
>> Sent from the jason-users mailing list archive at Nabble.com.
>> 
>> 
>> ------------------------------------------------------------------------------
>> This SF.net email is sponsored by Sprint
>> What will you do first with EVO, the first 4G phone?
>> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
>> _______________________________________________
>> Jason-users mailing list
>> Jas...@li...
>> https://lists.sourceforge.net/lists/listinfo/jason-users
> 
> -- 
> Jomi Fred Hubner
> Federal University of Santa Catarina
> Department of Automation and Systems Engineering
> PO Box 476, Florianópolis, SC
> 88040-900 Brazil
> http://www.das.ufsc.br/~jomi
> 
> 
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Jason-users mailing list
> Jas...@li...
> https://lists.sourceforge.net/lists/listinfo/jason-users
> 
> 

-- 
View this message in context: http://old.nabble.com/addPercept%28%29-method-does-not-update-the-percept-list-tp29233392p29239593.html
Sent from the jason-users mailing list archive at Nabble.com.

Hi Suralk,

The getPercept is supposed to be called by the agent reasoning process and not to "consult" the current perceptions. It implements an optimisation strategy that returns null if nothing has changed since last call (so the reasoning does not need to revise beliefs based on perception). If you call this method in your environment class, you are indeed changing the way the reasoning "sees" the environment. 

I wish you have a good continuation of your experiments with Jason.

Jomi

ps.: I'm adding a method for environment programmers to consult the perceptions.

On 22/07/2010, at 02:29, suralk wrote:

> 
> Hi All,
> I am new to Jason programming and was playing with the examples first which
> worked fine. I have started to use Jason for my real research work and have
> faced a problem with the addPercept() method.
> 
> In the RetrievePercepts() method below, I try to call the addPercept()
> method, using the parameters received. As shown, I try to look into the set
> of percepts of the agent before and after adding a percept. also I make sure
> i clean the percept list at the beginning. As expected, the call to
> getPercepts after calling addPercept() shows the added percept. 
> I want to change some of the perceived percepts in my architecture
> customization class, and when I look at the percept list inside the
> perceive() method, the percept list is empty!
> At the moment, I try to add two percepts in different occasions and adding
> one of them works while the other one does not.
> 
> It could be some small mistake that I am doing, but I just can't figure it
> out. Hope someone would be able to help me out.
> 
> thanks
> 
> /Suralk
> 
> public void RetrievePercepts(Literal percept, String agName){    	
>    	clearPercepts(agName);    	    	
>    	List<Literal> per =this.getPercepts(agName);
>    	try {
>    		if (per != null) {
>    			Iterator<Literal> ip = per.iterator();
>    			while (ip.hasNext()) {    				
>    				Literal p = ip.next();
>    				String  ps = p.toString();   				
>    				   				   				
>    			}
>    		}    		
> 	    }  catch (Exception e) {
> 	        
> 	    }
>    	addPercept(agName, percept);
>    	
>    	List<Literal> per1 =this.getPercepts(agName);
>    	try {
>    		if (per1 != null) {
>    			Iterator<Literal> ip = per1.iterator();
>    			while (ip.hasNext()) {    				
>    				Literal p = ip.next();
>    				String  ps = p.toString();		
>    			}
>    		}    		
> 	    } catch (Exception e) {
> 	        
> 	    }
>    }
> -- 
> View this message in context: http://old.nabble.com/addPercept%28%29-method-does-not-update-the-percept-list-tp29233392p29233392.html
> Sent from the jason-users mailing list archive at Nabble.com.
> 
> 
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Jason-users mailing list
> Jas...@li...
> https://lists.sourceforge.net/lists/listinfo/jason-users

-- 
Jomi Fred Hubner
Federal University of Santa Catarina
Department of Automation and Systems Engineering
PO Box 476, Florianópolis, SC
88040-900 Brazil
http://www.das.ufsc.br/~jomi

Hi All,
I am new to Jason programming and was playing with the examples first which
worked fine. I have started to use Jason for my real research work and have
faced a problem with the addPercept() method.

In the RetrievePercepts() method below, I try to call the addPercept()
method, using the parameters received. As shown, I try to look into the set
of percepts of the agent before and after adding a percept. also I make sure
i clean the percept list at the beginning. As expected, the call to
getPercepts after calling addPercept() shows the added percept. 
I want to change some of the perceived percepts in my architecture
customization class, and when I look at the percept list inside the
perceive() method, the percept list is empty!
At the moment, I try to add two percepts in different occasions and adding
one of them works while the other one does not.

It could be some small mistake that I am doing, but I just can't figure it
out. Hope someone would be able to help me out.

thanks

/Suralk

public void RetrievePercepts(Literal percept, String agName){    	
    	clearPercepts(agName);    	    	
    	List<Literal> per =this.getPercepts(agName);
    	try {
    		if (per != null) {
    			Iterator<Literal> ip = per.iterator();
    			while (ip.hasNext()) {    				
    				Literal p = ip.next();
    				String  ps = p.toString();   				
    				   				   				
    			}
    		}    		
	    }  catch (Exception e) {
	        
	    }
    	addPercept(agName, percept);
    	
    	List<Literal> per1 =this.getPercepts(agName);
    	try {
    		if (per1 != null) {
    			Iterator<Literal> ip = per1.iterator();
    			while (ip.hasNext()) {    				
    				Literal p = ip.next();
    				String  ps = p.toString();		
    			}
    		}    		
	    } catch (Exception e) {
	        
	    }
    }
-- 
View this message in context: http://old.nabble.com/addPercept%28%29-method-does-not-update-the-percept-list-tp29233392p29233392.html
Sent from the jason-users mailing list archive at Nabble.com.

Hi Felipe, 

it will be very useful if you can share your code. The console is still the same (based on Java Logging).

Best,

Jomi

On 20/07/2010, at 13:14, Felipe Rech Meneguzzi wrote:

> If this is not a feature already in Jason, I remember that I implemented my own coloured console a couple of years ago, and if my code is still compatible with the current MASConsole I am happy to contribute it.
> 
> Regards,
> 
> Felipe
> 
> On Tue, Jul 20, 2010 at 3:16 AM, Rafael H Bordini <r.b...@ac...> wrote:
> 
> From: Numan Deonarine <num...@ho...>
> Date: Mon, 19 Jul 2010 19:09:31 +0000
> To: <jas...@li...>
> 
> Hi,
> 
> In the MAS console how do you change the colour/font- e.g. so that
> output from each agent is displayed in a different colour?
> 
> Thanks in advance,
> 
> Numan.
> 
> 
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Jason-users mailing list
> Jas...@li...
> https://lists.sourceforge.net/lists/listinfo/jason-users
> 
> 
> 
> -- 
> ___________________________
> Felipe Rech Meneguzzi
> fel...@gm...
> http://fmeneguzzi.blogspot.com
> ___________________________
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first_______________________________________________
> Jason-users mailing list
> Jas...@li...
> https://lists.sourceforge.net/lists/listinfo/jason-users

-- 
Jomi Fred Hubner
Federal University of Santa Catarina
Department of Automation and Systems Engineering
PO Box 476, Florianópolis, SC
88040-900 Brazil
http://www.das.ufsc.br/~jomi

If this is not a feature already in Jason, I remember that I implemented my
own coloured console a couple of years ago, and if my code is still
compatible with the current MASConsole I am happy to contribute it.

Regards,

Felipe

On Tue, Jul 20, 2010 at 3:16 AM, Rafael H Bordini <r.b...@ac...> wrote:

>
> From: Numan Deonarine <num...@ho...>
> Date: Mon, 19 Jul 2010 19:09:31 +0000
> To: <jas...@li...>
>
> Hi,
>
> In the MAS console how do you change the colour/font- e.g. so that
> output from each agent is displayed in a different colour?
>
> Thanks in advance,
>
> Numan.
>
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Jason-users mailing list
> Jas...@li...
> https://lists.sourceforge.net/lists/listinfo/jason-users
>



-- 
___________________________
Felipe Rech Meneguzzi
fel...@gm...
http://fmeneguzzi.blogspot.com
___________________________

From: Numan Deonarine <num...@ho...>
Date: Mon, 19 Jul 2010 19:09:31 +0000
To: <jas...@li...>

Hi,

In the MAS console how do you change the colour/font- e.g. so that 
output from each agent is displayed in a different colour?

Thanks in advance,

Numan.

Hi Tina, 

As far as I remember, to stop the system you are already using a good approach (an alternative is to put that functionality in the environment, if you have one in your application). To run several times the system with different parameter, I used shell scripts in another project. But it is also quite simple to write a Java program for that. Take a look at bin/build.xml (the run task) to see how the system is started in Jason.

HTH,

Jomi
On 19/07/2010, at 08:16, Tina Balke wrote:

> Dear Jomi and Rafael,
> I’d like to use batch runs with my simulation. Currently I’m using a simulator agent that gets a message from all agents once they have performed a certain number of actions. When all agents have sent him a message, he will stop the simulation and kill the agents. Even though this works with stopping the simulation, is there any way of automatically starting the simulation again (possibly with a different set of parameters – I’m using an XML file to set that is read at the beginning of a simulation in order to set certain simulation parameters)?
> Thanks a lot for your help in advance.
> All the best,
> Tina
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first_______________________________________________
> Jason-users mailing list
> Jas...@li...
> https://lists.sourceforge.net/lists/listinfo/jason-users

-- 
Jomi Fred Hubner
Federal University of Santa Catarina
Department of Automation and Systems Engineering
PO Box 476, Florianópolis, SC
88040-900 Brazil
http://www.das.ufsc.br/~jomi

Dear Jomi and Rafael,
I'd like to use batch runs with my simulation. Currently I'm using a simulator agent that gets a message from all agents once they have performed a certain number of actions. When all agents have sent him a message, he will stop the simulation and kill the agents. Even though this works with stopping the simulation, is there any way of automatically starting the simulation again (possibly with a different set of parameters - I'm using an XML file to set that is read at the beginning of a simulation in order to set certain simulation parameters)?
Thanks a lot for your help in advance.
All the best,
Tina

 agents:
   c beliefBaseClass jason.bb.JDBCPersistentBB(
      "org.hsqldb.jdbcDriver", // driver for HSQLDB
      "jdbc:hsqldb:bookstore", // URL connection
      "sa", // user
      "", // password
      "[a(1,tablea,columns(runs(integer))),
        book(5),
        book_author(2),
        author(2,author,columns(id(integer),name(varchar(30)))),
        publisher(2)]");


Hi All,
I am trying to use a mysql db for an agent's belief base and came
across this example in the API. I am in need of a clarification.

#1 What is meant by arity here? it is not just a random number is it?

#2 if defined like publisher(2) what does it convey about the table
publisher?


Many thanks
Sachini

Hello Dimitris,

>> I noticed in the faq that someone can actually use just the BDI engine
>> for a custom infrastructure. As we are trying to run large scale
>> simulations are you aware of repast-jason interface implementations? Or
>> maybe someone else that has tried to use the BDI engine?

In my master thesis and later in various real-world robotic projects I 
used the stand-alone BDI engine to interface with Player/Stage [1]. I 
found it pretty straight forward to interface with a custom infrastructure.

Best,
Nico

[1] http://playerstage.sourceforge.net/

Hi Dimitris,

We are not aware of any repast-jason integration, but certainly quite a 
few people have used the stand-alone BDI engine, including student 
projects. We have had enquires about integration of Jason with social 
simulation platforms, but we don't know if any was made, and not with 
repast. Since Repast agents are programmed in Java, it should -- in 
principle -- be fairly easy to integrate Repast and Jason if you use the 
BDI engine only (of course it might turn out not that easy when you do 
it in practice...).

Best,

Rafael



On 14/07/2010 23:47, Dimitrios Traskas wrote:
> Hi again,
>
> I noticed in the faq that someone can actually use just the BDI engine
> for a custom infrastructure. As we are trying to run large scale
> simulations are you aware of repast-jason interface implementations? Or
> maybe someone else that has tried to use the BDI engine?
>
>
> Cheers
> Dimitris Traskas
> Bath University,
> UK
>
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
>
>
>
> _______________________________________________
> Jason-users mailing list
> Jas...@li...
> https://lists.sourceforge.net/lists/listinfo/jason-users

Hi Dimitris,

Last January we published a paper at ICAART about "Evaluating Jason for 
distributed crowd simulations" which might be of your interest. You cand find 
it at my homepage:  http://www.uv.es/grimo

We have continued working on how to tune Java to enhance Jason's performance 
over the centralised infrastructure and we have a paper waiting for 
acceptation on this topic. As soon as we have it accepted I will upload it to 
the previous web page.

Cheers,
Francisco Grimaldo

A Dijous, 15 de juliol de 2010 00:45:23, Dimitrios Traskas va escriure:
> Hi,
> 
> I am new to Jason - just a couple of months - and recently came up with a
> couple of questions. I will try to add them to the mailing list separately.
> First of all I would like to find out the scalability of Jason. Has anyone
> tried to max out the number of agents when running on the centralized
> infrastructure? I have tried to create 5000 agents for the sake of my
> simulation and the system cannot create a new thread and comes back with an
> out of memory exception. I tried increasing the jvm settings for eclipse
>  but I suspect there is a setting somewhere else. Any ideas?
> 
> Cheers
> Dimitris Traskas
> Bath University,
> UK
> 
> 
> PS: I am using the Eclipse Jason plugin.
> 

-- 
-------------------------------------------
Dr. Francisco Grimaldo Moreno (PhD)
(Id. GnuPG: 0xBAEC75A7)

Departament d'Informàtica
Escola Tècnica Superior d'Enginyeria (ETSE)
Universitat de València
Avda. Vicent Andrés Estellés, s/n
46100 Burjassot-València (SPAIN)

Tel: +34 96 354 44 87 (+34 96 354 45 65)
Fax: +34 96 354 47 68
Email: fra...@uv...
Web: http://www.uv.es/grimo
-------------------------------------------

Hi Dimitrios,

to increase the heap size, you can customise the Ant script used to run the system. 

detailed steps:

1. in the bin directory,m rename the build.xml to c-build.xml

2. edit the c-build.xml file importing the build.xml created by Jason, then  "override" the tasks you want and remove the others. For instance:

<?xml version="1.0" encoding="ISO-8859-1" ?>

<project name   ="myproject"    basedir=".."     default="run">

    <import file="build.xml"/>

    <target name="run" depends="compile" >
        <echo message="Running project ${ant.project.name}" />
        <java classname="jason.infra.centralised.RunCentralisedMAS"
              failonerror="true" fork="yes" dir="${basedir}" >
            <classpath refid="project.classpath"/>
            <arg line="${mas2j.project.file} ${debug} "/>
            <jvmarg line="-Xmx500M -Xss8M"/>    
        </java>
        <antcall target="user-end" />
    </target>
</project>

In the jvmarg line you can customise the heap size.


HTH,

Jomi

the content of the 

On 14/07/2010, at 20:11, Dimitrios Traskas wrote:

> wow that's a quick answer :). I tried it actually, I used a pool of 100 threads and very quickly managed to get 5000 agents up and running. Where would I increase heap size if I wanted to increase the threads even more? 
> 
> 
> Thanks for the help
> Dimitris
> 
> 
> 
> 
> On 14 July 2010 23:58, Jomi Hubner <jom...@gm...> wrote:
> Hi Dimitris,
> 
> you can configure the number of threads the system will use. Take a look at examples/game-of-life, it runs 3600 agents quite well.
> 
> HTH,
> 
> Jomi
> 
> On 14/07/2010, at 19:45, Dimitrios Traskas wrote:
> 
> > Hi,
> >
> > I am new to Jason - just a couple of months - and recently came up with a couple of questions. I will try to add them to the mailing list separately. First of all I would like to find out the scalability of Jason. Has anyone tried to max out the number of agents when running on the centralized infrastructure? I have tried to create 5000 agents for the sake of my simulation and the system cannot create a new thread and comes back with an out of memory exception. I tried increasing the jvm settings for eclipse but I suspect there is a setting somewhere else. Any ideas?
> >
> > Cheers
> > Dimitris Traskas
> > Bath University,
> > UK
> >
> >
> > PS: I am using the Eclipse Jason plugin.
> > ------------------------------------------------------------------------------
> > This SF.net email is sponsored by Sprint
> > What will you do first with EVO, the first 4G phone?
> > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first_______________________________________________
> > Jason-users mailing list
> > Jas...@li...
> > https://lists.sourceforge.net/lists/listinfo/jason-users
> 
> --
> Jomi Fred Hubner
> Federal University of Santa Catarina
> Department of Automation and Systems Engineering
> PO Box 476, Florianópolis, SC
> 88040-900 Brazil
> http://www.das.ufsc.br/~jomi
> 
> 

-- 
Jomi Fred Hubner
Federal University of Santa Catarina
Department of Automation and Systems Engineering
PO Box 476, Florianópolis, SC
88040-900 Brazil
http://www.das.ufsc.br/~jomi

Hi again,

I noticed in the faq that someone can actually use just the BDI engine for a
custom infrastructure. As we are trying to run large scale simulations are
you aware of repast-jason interface implementations? Or maybe someone else
that has tried to use the BDI engine?


Cheers
Dimitris Traskas
Bath University,
UK

Hi,

I am new to Jason - just a couple of months - and recently came up with a
couple of questions. I will try to add them to the mailing list separately.
First of all I would like to find out the scalability of Jason. Has anyone
tried to max out the number of agents when running on the centralized
infrastructure? I have tried to create 5000 agents for the sake of my
simulation and the system cannot create a new thread and comes back with an
out of memory exception. I tried increasing the jvm settings for eclipse but
I suspect there is a setting somewhere else. Any ideas?

Cheers
Dimitris Traskas
Bath University,
UK


PS: I am using the Eclipse Jason plugin.

Dear users,

We have just released the version 1.3.3 of Jason! 

This release fixes some bugs and adds the feature of meta-events (details in the release-notes.txt).

Download:
	https://sourceforge.net/projects/jason/files/

Regards,

Rafael & Jomi

Why sometimes i have this error :
Exception in thread "nica" java.lang.OutOfMemoryError: Java heap space
   at java.util.Arrays.copyOf(Arrays.java:2882)
   at 
java.lang.AbstractStringBuilder.expandCapacity(AbstractStringBuilder.java:100) 

   at 
java.lang.AbstractStringBuilder.append(AbstractStringBuilder.java:390)
   at java.lang.StringBuilder.append(StringBuilder.java:119)
   at jason.asSemantics.Intention.toString(Intention.java:195)
   at java.lang.String.valueOf(String.java:2826)
   at java.lang.StringBuilder.append(StringBuilder.java:115)
   at jason.asSemantics.Event.toString(Event.java:89)
   at java.lang.String.valueOf(String.java:2826)
   at java.lang.StringBuilder.append(StringBuilder.java:115)
   at 
jason.asSemantics.TransitionSystem.applyRelApplPlRule2(TransitionSystem.java:279) 

   at 
jason.asSemantics.TransitionSystem.applyFindOp(TransitionSystem.java:347)
   at 
jason.asSemantics.TransitionSystem.applySemanticRule(TransitionSystem.java:146) 

   at 
jason.asSemantics.TransitionSystem.reasoningCycle(TransitionSystem.java:1021) 

   at 
jason.infra.centralised.CentralisedAgArch.run(CentralisedAgArch.java:197)
   at java.lang.Thread.run(Thread.java:619)

"nica" is my agent and i run my project in Eclipse. Do i must increase 
heap size?If yes, how can i do it?
Thank

Also, have a look at the logging system, it might help you e.g. writing 
the chosen plans (the first in the options list if you use the default 
selection) to a log file. See section 4.3 of the FAQ:

http://jason.sourceforge.net/faq

cheers


On 05/07/2010 14:27, Jomi Fred Hubner wrote:
>
> On 05/07/2010, at 08:35, Anselmi Biagio wrote:
>
>> Hi, how can i get at the end of execution of my project in Jason, the list of plan, that the agent haved choosed during running?
>
> You need to customise the Agent class. In your customisation, the method selectOption will receive as parameter all options of plans (Jason selects the first by default).
>
> e.g.:
>
> class MyAgent extends Agent {
>      public Option selectOption(List<Option>  options) {
> 	// do whatever you want with options here
> 	return super.selectOptions(options); // default implementation
>      }
> }
>
>> And why, when i use the internal action .time(15,_,_). and on my computer the hour is 15, MAS console give me a failure?
>
> It is a bug in the internal action .time: 15h is represented by 3 (pm). This was already fixed in the subversion and will be available in next release of Jason.
>
> HTH,
>
> Jomi
>
>
>
>> Thanks
>>
>>
>

On 05/07/2010, at 08:35, Anselmi Biagio wrote:

> Hi, how can i get at the end of execution of my project in Jason, the list of plan, that the agent haved choosed during running?

You need to customise the Agent class. In your customisation, the method selectOption will receive as parameter all options of plans (Jason selects the first by default). 

e.g.: 

class MyAgent extends Agent {
    public Option selectOption(List<Option> options) {
	// do whatever you want with options here
	return super.selectOptions(options); // default implementation
    }
}

> And why, when i use the internal action .time(15,_,_). and on my computer the hour is 15, MAS console give me a failure?

It is a bug in the internal action .time: 15h is represented by 3 (pm). This was already fixed in the subversion and will be available in next release of Jason. 

HTH,

Jomi



> Thanks
> 
> 

-- 
Jomi Fred Hubner
Federal University of Santa Catarina
Department of Automation and Systems Engineering
PO Box 476, Florianópolis, SC
88040-900 Brazil
http://www.das.ufsc.br/~jomi

Thanks Rafael.
Tested it on my application and your method works!!

I am in the process of integrating jason agents with OGCE WS-Messenger
infrastructure to model a pub-sub scenario.
The requirement i described in my previous mail was in fact a scenario in
that particular case study.

The pointer toward the jason API is much appreciated.

Sachini

On Fri, Jul 2, 2010 at 3:13 PM, Rafael H Bordini <r.b...@ac...> wrote:

> Hi Sachini,
>
> You don't need to create an internal action to generate random numbers,
> that's already available in Jason. In fact, it's always good to have a
> look at the stdlib in jason.sf.net/api -- very often people end up
> creating new internal actions for doing things that are already done in
> Jasons pre-defined internal actions.
>
> So your agent should look something like (untested):
>
> !keep_busy.
>
> +!keep_busy <- .random(X); .print(X); .wait(10000); !!keep_busy.
>
> Of course the agent can have other intentions in parallel, and every 10
> second it will come again to have the intention of printing a randon
> number. Note that the !! in the body instead of ! is very important in
> this case.
>
> Cheers,
>
> Rafael
>
>
>
>
>
> On 02/07/2010 10:03, sachini weerawardhana wrote:
> > Hi
> >
> > I have a scenario where i want an agent to generate a random number and
> > return the value once every 10 seconds.
> > I took the following approach:
> >
> > - Define a new InternalAction that generates the random number and
> > return it.
> > - use the InternalAction in a plan that execute once every 10 seconds.
> >
> > How do i make the agent to repeatedly execute a plan once every 10
> > seconds thoughout his lifetime?
> >
> > sachini
> >
> >
> >
> >
> ------------------------------------------------------------------------------
> > This SF.net email is sponsored by Sprint
> > What will you do first with EVO, the first 4G phone?
> > Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> >
> >
> >
> > _______________________________________________
> > Jason-users mailing list
> > Jas...@li...
> > https://lists.sourceforge.net/lists/listinfo/jason-users
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Jason-users mailing list
> Jas...@li...
> https://lists.sourceforge.net/lists/listinfo/jason-users
>

Hi Sachini,

You don't need to create an internal action to generate random numbers, 
that's already available in Jason. In fact, it's always good to have a 
look at the stdlib in jason.sf.net/api -- very often people end up 
creating new internal actions for doing things that are already done in 
Jasons pre-defined internal actions.

So your agent should look something like (untested):

!keep_busy.

+!keep_busy <- .random(X); .print(X); .wait(10000); !!keep_busy.

Of course the agent can have other intentions in parallel, and every 10 
second it will come again to have the intention of printing a randon 
number. Note that the !! in the body instead of ! is very important in 
this case.

Cheers,

Rafael





On 02/07/2010 10:03, sachini weerawardhana wrote:
> Hi
>
> I have a scenario where i want an agent to generate a random number and
> return the value once every 10 seconds.
> I took the following approach:
>
> - Define a new InternalAction that generates the random number and
> return it.
> - use the InternalAction in a plan that execute once every 10 seconds.
>
> How do i make the agent to repeatedly execute a plan once every 10
> seconds thoughout his lifetime?
>
> sachini
>
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
>
>
>
> _______________________________________________
> Jason-users mailing list
> Jas...@li...
> https://lists.sourceforge.net/lists/listinfo/jason-users