This new editor will be available in Guvnor 5.4 

Change-Set Editor UI

To create a new Chang-Set you have to go to “Create New” -> “New Change Set” option in Knowledge Base section of Guvnor.
In the editor you have 4 items you can interact with:

1. Change-Set Editor canvas: Here you can edit your change-set by hand. You can add, remove or edit resources as you wish. Of course you will need to know the URL of each of the resources. You could get this information from the “Attributes” tab of each of the resources.
2. Change-Set URL: This is the URL you need to use in your applications to make reference to this change-set.
3. Package Button: This is an easy way to add packages to your change-set. You need to place the cursor where you want to add a new Package in the  Change-Set Editor canvas and then click this button. A Wizard will appear to help you choose the right package. See next section for further information.
4. Asset Button: Useful button to add a new Assets to the Change-Set using a Wizard. For more information please read “Adding Assets using a Wizard” section.

Adding Packages using a Wizard

Clicking in the “Package” button of Change-Set Editor, a wizard will be displayed to help you choose the package you want to add as resource.

The Package Wizard has 3 sections:

1. Name: Each resource inside a Change-Set could have a descriptive name. Use this optional field to set this property.
2. Description: Each resource inside a Change-Set could have a description. Use this optional field to set this property.
3. Packages: In this section you can select the package and the Snapshot version of a package you want to add as a resource of the Change-Set you are editing. If the Change-Set you are editing belongs to the Global Area, then you will be able to select the resource to add from all the packages defined in Guvnor and their snapshots of course. If the Change-Set you are editing belongs to a particular Package, you will only see the Snapshots of that package in this Wizard.

Adding Assets using a Wizard

You have 2 ways of adding assets as resources of a Change-Set: by hand using the Change-Set Editor canvas or using the Asset Wizard. This Wizard is invoked using the “Asset” button of the editor.

This Wizard has 5 sections:

1. Name: Each resource inside a Change-Set could have a descriptive name. Use this optional field to set this property.
2. Description: Each resource inside a Change-Set could have a description. Use this optional field to set this property.
3. Package: Use this field to filter the assets of a particular Package. If the Change-Set you are editing belongs to the Global Area, then you will be able to select any package, otherwise the package where your Change-Set belongs is automatically selected and you will not be able to change it.
4. Format: Use this field to filter the type of assets you are looking for.
5. Assets: The list of assets that match the selected package and format. You can select one or more assets using the check-box column and then click “OK” to add each resource to the Change-Set you are editing.

We will continue improving the demo to integrate it with more technologies and to add it more features. We are really looking forward to be in next year’s RuleML!

Emergency Service Demo Application

For the first time, Plugtree is going to organize an open training session. We are used to give private on-side sessions, but this time we wanted to make it open! We are expecting a mix of knowledge background and experiences that enriches us all.

The main focuses of these sessions are Drools-Expert and jBPM5. We are going to go from the most simple and basic concepts to the most advanced features of these frameworks. Of course we are not going to leave out Drools-Fusion!

If you are starting with these technologies, at the end of this training you will be able to understand and start using them. You will be familiarized will all the terms and concepts behind these technologies.

If you are already using some of these frameworks, you will have a great opportunity to clear all your doubts, ask about best practices, and to learn about projects’ internals.

It is very important to confirm your presence as soon as possible so we can find a place that meet our needs. We are looking for a nice hotel that allows us to have even coding nights for those who are interested!

If you are interested, please post a comment and get in contact with us writing a comment here or sending emails to:  info AT plugtree DOT com or my personal email: esteban DOR aliverti AT gmail DOT com, using the subject: Drools and jBPM5 Training in London.

You can register to this event here. Stay tuned for more information but please let us know if you are interested.

You can find more about this community training initiative @ilesteban and @salaboy

This week Mauricio Salatino presented a Drools/jBPM Demo application in Boston JUDCon.

Salaboy and I have developed this application at Plugtree to prove and show a lot of conceptual and architectural concepts behind JBoss Drools and jBPM projects.
The application is about an emergency service call center. It uses jBPM for define and run the company’s processes and Drools Expert to execute part of the involved Business Logic. The Demo application also uses Drools Fusion to correlate events coming form real-time sensors (a Wiimote or the keyboard) .

Prerequisites

Emergency Service (ES) is a multi-module maven project hosted in github. So, the first 2 prerequisites are git and maven.
You can get git binaries from the project home page: http://git-scm.com/download
In the case of Ubuntu, you already have git in the repositories.
In the case of maven, you will need to have at least version 2 installed. In the case of Ubuntu, you can directly get it from the package repository. You can also download its binaries from here: http://maven.apache.org/download.html

Getting the Sources

Getting Emergency Service sources is pretty straight forward. The only thing you need to do is to clone the project from its github location:

$git clone git://github.com/Salaboy/emergency-service-drools-app.git emergency-service

After cloning the repository, you will end up with this directory structure:

Demo Directory Structure

I don’t want to enter in details about the project structure in this post, however this is a brief explanation of each of the modules:

• emergency-service-core: this module contains the “server” application. This server is in charge of start jBPM’s Human Task Server, create a Drools Grid Node, add some entities to the distributed cache and initialize some of the services classes used by the application.
• emergency-service-model: this module contains all the model classes involved in the application. It also contains the classes used to reach the distributed cache and to send messages across different modules.
• emergency-service-tasklists-ui-swing: the task list swing application of the demo.
• emergency-service-world-ui-slick: the application containing the map of the demo.

Compiling the sources

Compiling the sources is a maven task. Make sure you have maven in your classpath and run:

$mvn clean install -DskipTests=true

*We are skipping test because they are kind of unstable yet

If everything works as expected, you will end up with the following message in your console:

A Successful Compilation

LWJGL Library Configuration

For the City Map application we are using Slick. Slick is a 2D java library that wraps LWJGL. If you want to use slick, you need to have LWJGL in your system.
These are the steps you need to do to get LWJGL running:

1. First step to get slick running is to get LWJGL 2.1.0. You can get it from here: http://sourceforge.net/projects/java-game-lib/files/Official%20Releases/LWJGL%202.1.0/lwjgl-2.1.0.zip/download
2. Download the zip file and expand it somewhere.
3. Edit  emergency-service-world-ui-slick/pom.xml file and search for -Djava.library.path configuration. Replace the current value to fit your configuration. Note that LWJGL has different directories for different operating systems, make sure you use the correct one.

Executing the Demo

When executing the demo, the order of execution of its modules is essential.
The first project you need to run is emergency-service-core:

$cd emergency-service-core
$mvn exec:exec

When emergency-service-core is running, you can start emergency-service-tasklists-ui-swing or emergency-service-world-ui-slick in any order. For example:

$cd emergency-service-world-ui-slick
$mvn exec:exec

The result should be:

World UI

and

$cd emergency-service-tasklists-ui-swing
$mvn exec:exec

The result should be:

Task List Application

Playing with the Demo

To start playing with the Demo, you need to start an emergency. To do this, you have to press the space bar key in the map. When you press that key, a new emergency will be displayed in the map.
In the first task list of emergency-service-tasklists-ui-swing you will see a task to be completed (you need to press the “Refresh” button.
At this point you can continue with the flow of execution: complete the first task, go to the second task list, complete it, etc.
Because the demo is not completed yet, it only supports HEART ATTACK emergencies for 1 person between 20 and 60 years old. We are working to support the all the possibilities soon.
To simulate patient heart beat, you can use a Wiimote (need to be configured in task list application) or ‘q’ and ‘w’ keys.

Conclusion

I tried to explain how to download, compile, configure and run the emergency service demo application.
Remember that this Demo is under heavy development, so expect changes and bugs!!
Feel free to play with it. to report any problem and to give us some feedback too!

This module covers the use of domain specific logic inside your processes. BPMN2 standard specify a special Activity to perform this: Task. Using it we can interact with external systems to create complex processes for real life situations.

The following topics are covered in this module:

Take a look at the slides and try to make the homework! I will be glad to help you if you find any problem!

P.S: This module is also published in salaboy’s blog.

All the material provided here is under the Creative Commons License.
Our aim is to lower Drools’ entrance barrier and to promote its adoption as well!. Feel free to join us in this journey, we will be waiting feedback, questions and criticisms!

What and When?

So, the 2 main questions are: What is going to be published? and When it is going to be published?

Salaboy and I made this roadmap to organize ourselves and to give more transparency to the whole process. So we will do our best to release the following material in the following dates:

Module #1 – February 28: Drools 5 and jBPM5 Theoretical Introduction:  This module covers a quick overview about all the modules inside the Drools 5 Business Logic Integration Platform. You will see that jBPM5 was also included here because it’s considered as one of the Drools Platform module. This module was conceived to cover all the conceptual background that you need to have in order to start working with Drools Expert.

Module #2 – March 3: Drools Expert Introduction + First Hands On Labs: This module gives us a quick overview about the needed Drools APIs to start interacting with the engine. A quick overview about how the engine works and how is the normal execution cycle for our rules is also discussed.

Module #3 – March 8: DRL Syntax + Two Hands On Labs: This module covers the DRL syntax in depth. Explaining you most of the features provided by the language. Two hands on labs will be provided to demonstrate the language functionality and the runtime behavior.

Module #4 – March 15: RETE Algorithm Introduction: A quick introduction to the RETE algorithm is discussed in this chapter. The main idea behind covering this algorithm is to understand the rule engine internals for performance tuning and optimizations that we can do over our rules.

Module #5 – March 20: DSL and Decision Tables + 2 hands on labs: This module is focused on user experience when writing business rules. This module will cover how to write Domain Specific Languages that will make the rules easier to write using the language that the people is used to use in their daily work. Decision Tables will be also covered to demonstrate how we can create template-like structures that will generate different rules using different set of parameters.

This is the tempted list of topics. If you feel we are missing something, please let us know and we will be glad to include it!

So, stay tuned!

Disclaimer: Plugtree is not affiliated with or sponsored by Red Hat®. Drools is a community project sponsored by JBoss®. All the material published during this training course is under the Creative Commons License.

• Create and Edit Spring Context Files inside Guvnor.
• Define Drools’ specific tags using a palette.
• Customize the elements in the palette to allow custom tags.
• Validate Spring Context xml.
• Expose Spring Context xmls through independent URLs.

This feature will be available in Drools 5.2.0M1. If you want to get more information, please read Nicola’s post.

Authoring rules without Custom Forms

Previous to this feature, when users want to add a restriction on a Fact field they must choose one of the three available options: Literal Value, Formula or Expression Editor.

Constraints Values Popup

These options are enough for create basic rules. But if you want to have more control over the possible values a field could have while authoring a rule you, then these three options are not enough.

One way to get more control on what users write is using Fields Constraints. But right now, they only work for post-authoring validation.

What is a Custom Form?

Custom Forms are for situations where selecting a value for a restriction involves some business logic. Suppose you have Address class with a country attribute. You might want to limit rules’ authors to enter valid names, and not just a plain string. Another case could be he use of internal codes for some other attribute that need to be retrieved from external systems.

For situation like those described above, Guvnor needs a way to communicate with external systems to get business validated values.

So, for the first example it would be nice to open a popup containing a Google Map where users could select a country name by clicking on it. This is one of the main purposes of Custom Forms. The following screenshot shows how this should look like

Custom Form Popup

So let’s go now to see how Custom Forms are created and how are they configured inside Guvnor.

Custom Form creation

Custom Forms are external web applications deployed in the same server Guvnor is. The Custom form must be accessible through an URL via HTTP GET method. You can use whatever technology you want for creating this form. Guvnor will open a popup containing an iframe and will load the form using its URL.

Guvnor / Custom Form Communication

The communication between Guvnor and the external application is performed using HTTP parameters (for Guvnor -> Custom Form communication) and DOM inspection (for Custom Form -> Guvnor communication).

Every time Guvnor invokes the Custom Form’s URL it will pass 4 GET parameters:

1. factType: the short class name of the Fact that is being edited.
2. fieldName: the name of the Fact’s field that is being edited.
3. cf_id: the id of any previously selected value. Guvnor doesn’t use this value at all. It is only used by the external application to restore any previous state when editing a value. This value is empty the first time a Custom Form is opened.
4. cf_value: The previously selected value. This value is empty the first time a Custom Form is opened.

The Custom Form could use this values to configure itself and must let users to select a new value for the field of the constraint that is being edited. Most of the times, the value itself is not meaningful for the external application. For example, if the external form let us choose a User name from a list, the external system might only cares about its ids. Guvnor in the other hand is only interested in the value of the name itself, because it is that value the one that will be used for the rule. But the next time a user edits the rule and wants to modify the User name of the constraint, the id of the User could be very useful for the External Form in order to restore any state. That is why Guvnor keeps 2 pieces of information for Custom Forms values: an internal id, and the value itself.

In order to get the selected value from the Custom Form, Guvnor uses javascript to introspect the external iframe DOM and retrieve the values. The external application must contain 2 <input> elements in the root container with ids: “cf_id” and “cf_value”. Guvnor will read the values of these input elements and store them. The value will be used in the rule (just like a literal value) and the id will be used in all the consequent invocations of the Custom Form.

Custom Form Configuration Inside Guvnor

Custom Forms are meta-data associated to a Fact’s field. So the better place to configure them are Working Sets.

The Working Set Management window now have an extra tab called “WS Custom Forms”. There you can configure different Custom Forms for different fields. A field can only have one Custom Form configuration in a Working Set.

Custom Form Configuration Panel

As you can see in the screenshot above, Custom Forms configuration consists of 5 elements:

1. Fact Type: You must choose from one of the working set’s fact types.
2. Field: This combobox is is filled in with the fields names of the Fact Type selected before.
3. UR: The URL of the custom form application that will be used for setting the field value in a restriction.
4. Width: The width of the opening popup.
5. Height: The height of the opening popup.

Using Custom Forms in Guided Editor

Once you have configured the Custom Forms inside a Working Set, you can start using them in your guided editor. To do this you must enable the Working Set first, and then start creating restrictions as you always did.

When you add a restriction to a field that has a Custom Forms configuration associated, the configured URL will be opened instead of the regular popup (shown in figure 1). The selected value will be added to the guided editor in a button widget. When the user clicks this button, the popup will be shown again.

An example

A screencast is worths 1000 screenshots, that is why I have created this little demo: http://plugtree.com/downloads/screencasts/CustomForms_demo.htm

There you can see how Custom Form are configured and used.

The External Application code could be found here: http://plugtree.com/downloads/screencasts/index.jsp.txt

Conclusion

Guvnor now support custom business logic to fill constraints values while authoring rules using Guided Editor. You can assign URLs to Fact fields using Working Sets; this URLs will be shown in Guided Editor when creating rules.

Custom Forms are external web applications deployed in the same server where Guvnor is. You can use any web technology you want to create this forms.

And, as I always say, this is an experimental feature. Feel free to use it and test it. Comments and constructive criticisms are always welcome.

What drool’s team needs now are users that test it. So, if you was one of the people who voted this issue or you was waiting it, you can start using it right now!

Adding DSL/DSLR Resources

Adding these type of resources to the knowledge agent’s kbase is as simple as adding any other resource type. You need to identify your resources with DSLR or DSL type:

<change-set xmlns='http://drools.org/drools-5.0/change-set'
	xmlns:xs='http://www.w3.org/2001/XMLSchema-instance'
	xs:schemaLocation='http://drools.org/drools-5.0/change-set drools-change-set-5.0.xsd' >
	<add>
		<resource source='http://localhost/rules.drl' type='DSLR' />
		<resource source='http://localhost/myExpander.dsl' type='DSL' />
	</add>
</change-set>

The order of the resources is not important. The agent will always treat DSL resources first.

Modifying DSL or DSLR Resources

DRLR and DSL resources are managed in different ways by the agent.
Whenever the agent notes that a DSLR resource has changed, it would be treated as a regular DLR resource: All its rules will be compiled (using the pre-existent DSL expander) and they will be added to the agent’s kbase.
The tricky part comes when you want to modify a DSL resource. Because DSL resources themselves don’t create any binary content (such as rules, queries, etc.), the only thing the agent does when a resource of this type is modified is to reload its definition. That’s it. No rules are created nor modified. If you want to recreate the rules that were created using the modified DSL resource, you have to modify the DSLR resource as well. If you don’t want to modify the DSLR content, you can just “touch” it (like the unix ‘touch’ command) this will trigger the change set appliance process just like if you were modified the resource’s content.

Conclusion

The Knowledge Agent now supports DSL and DSLR resources. You can add these type of resources using change sets, just like any other resource. When a DSLR resource is modified, the agent will treat it as a regular DRL file; the builder the agent uses will contain all the DSL resource you have previously added. If a DSL resource is modified, nothing will happen until the corresponding DSLR resource/s get modified too.

This feature will be available from 5.1 RC1. If you could test it before that you will be helping me (and all the drool’s team) a lot

Have fun!