%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Thanks for your interest in eJason, our experimental implementation of
Jason in Erlang. Here you can find the commands necessary to run a MAS
using eJason (for UBUNTU Linux)

If you have any issues, please feel free to contact me at:
[email protected]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

To compile eJason sources, just type:

make

This will generate a set of ".beam" files inside "./eJason/ebin".

We also suggest copying into the "agents" folder all ".asl" files,
i.e. all Jason source code files, that will be run. This folder is
included in the path were eJason looks for Jason files.

To start an eJason system, execute the script 'ejason':

./ejason  [-c ContainerName]

where ContainerName is the name of the container that composes the
newly created system. As a matter of example, consider the command:

./ejason -c myContainer

Its execution will generate a container myContainer@userHost, where
userHost is the name of the host where the command was executed.

The ejason script starts an erlang shell invoking the command 'erl'.
Therefore, it accepts any parameter associated to 'erl' (see
http://erlang.org/doc/man/erl.html).

The ejason runtime system (the Distribution Manager and the
Supervision Manager) is started as well. 

----------------------------------------------------------------------------
The source code of the Jason agents (i.e. the '.asl' files) that run
in eJason must be placed in a directory belonging the to search path
of eJason. By default, the folders "ebin", "include" and "agents"
belong to it. Additional directories Dir1... DirN can be added via the
parameter "-pa Dir1 ... DirN" to the ejason script.

To parse and compile the eJason code of an agent, the following command
must be run:

> ejason:parse(agentName).
> ejason:compile(agentName).

To run this agent, type:

> ejason:run(agentName).

For simplicity, the three commands above can be consecutively run as:

> ejason:crun(agentName).


For instance, to run the "counter" agent within the tests provided the 
following sequence of commands suffices:

./ejason -c container1

> ejason:crun(counter).


Similarly, several agents (10 in this sample) can be started with the 
same code by executing:

> ejason:crun(counter, 10).

This agent will always be alive. If we want to stop it, we must call:

ejason:kill(counter).

If "N" agents with the same name (actually, they will be renamed
counter_2... counter_N) were launched, we can kill all them by
invoking:

ejason:kill(counter, N).


****** RUNNING TESTS  ****************

Check the README file for each of the tests, as some may require special
start parameters or the start of several Erlang nodes


****** NOTES ON THE LATEST GENERAL UPDATE  **************

The implementation of eJason now includes the monitoring and
supervision mechanisms described in the Ph.D. thesis of the developer. 

A formal semantics of these mechanisms is also available in the
aforementioned work.

(More information will be added after the formal defense).