Recently I had been searching for a good Bug tracking tool which is easy to learn and use. Bugzilla is something I had been used to for long. But this is a new and small company, and I want to introduce this application to the Project development team, testing/reporting team , designing team while they are working on mutiple projects. Even I want to monitor all the bugs in an easy and less time consuming way. I am worried only about reopened/non assigned bugs / assigned to me etc and I want to find it easily, not sitting and searching which is time consuming.

I first saw Mantis when my brother in law, showed it to me. I loved the simple and easy to use GUI. I took the application for a spin for one week,  installation was a breeze.  After 2-3 days of usage, I was sure,  I will be implementing this as our official bug tracking tool.

So officially we have started using Mantis for all our software bug tracking needs. It was just a matter of 10-20 minutes for all the people in my office to get a hang of it.  And they all started liking it, and I spend less than an hour importing all the existing bugs data to this application. Initially I was planning to have this use projects that just received the SOW, But seeing the simplicity of Mantis, I brought in data from all projects which are running right now.

Try out this tool when you have time. Go to http://www.mantisbt.org/

I am trying out another OSS app named Vtiger CRM, another Sourceforge project. Infact thinking of adding some custom modules to it for a customer. I will write an article about it  shortly… till then bye from me..

It all started with some collaborated work on Gantt project, years back. later I had worked with  ZK one of the robust Development tool I have seen in sourceforge. Eclipse plugin development importantly GMF and GEF also keeps me interested.

Though I am very busy with business commitments these days, still I find some time to test or look through some cool OSS tools.  The number of applications that gets started in SF shows how much passionate people are about technology. But there is always the plain old philosophy – “Survival of the fittest”. Some projects becomes world popular some vanishes with no one to further develop.

Sourceforge’s idea of showcasing a project per month based on their ranking mechanism is something truly great and a big boost to the developers who are contributing  to the projects. One of the projects that seriously got me interested is Cloud Computing application – eyeOS tool. After trying it out, I am so impressed. Hats off to the team that works for this project. Surely Cloud computing is the future , critics might say data security issues.

Have a look at the project here and why not try it out sometime..you just need apache server with PHP enabled.. if you have xampp for windows/mac its as easy as pie.  http://www.eyeos.org

We all run in to an issue with Plagiarism at least once in a life time knowingly and unknowingly.  It can be when we give the weekly assignments in school to a project report or matter content for the website. If I forget to add the quotes “..”  in the first paragraph of this blog entry well I am also in trouble- I will be using some one else s definition for Plagiarism in my own blog as if I wrote it for the first time.

So to keep out of trouble is  – Write your own contents if possible.  If you use someone else s ideas give them credits by clearly in writing. Use quotes “…” where ever applicable. Even when you do that make sure what you tell the masses is correct. Even if you give credits but send out wrong information say for eg: F=ma (Sir. Newton’s second law) was formulated by Sir. Thomas Alva Edison, that is also wrong. After all we want to make the online community better and more useful.

Lots of Plagiarism tools are available.  Check out this page:- http://www.shambles.net/pages/staff/ptools/ to find a wide array of tools.

Also while you take contents from a page and plan to give the credits to the original publisher, check out their terms and policies. Some of them require express written permission before being linked or copied to another page. A simple check will help you stay out of trouble.. Happy and safe blogging/Content writing.

Iphone is available in India and personally I love it like anything. Even developing for the same is a pleasure considering the support from Apple. Even if you pay 99$ for the developer program it is worth every penny but keep in mind you adhere to the terms and conditions  so that your apps wont be blacklisted from the apps store. I have seen many posts in the past where people complaining about their application being blacklisted or it was not a good decision so I am moving in to Android etc. But we need to understand the fact that in Business everyone need to stay. An application that allows an Iphone to make skype/vonage calls via WIFI will harm the mobile operator for sure. Some how Apple has control over these things while they have released the SDK to public.  In my whole hearted opinion,  it is Good.  Some of the Iphone developers/ users wont agree but the quality of the apps are not compromised since Apple has control over these things.

I am a great admirer of Linux and some what crazy about it right from school days. Today we have thousands of different flavors of Linux. When I heard about a common Linux platform based mobile phones I was so thrilled. I looked in to the Android stuff for past 3 months which made me understand the difference between developing for an open source based system and I phone which has still strings tied by Apple. Definitely I phone development is better.

For Apple with the way I phone sales are sky rocketing which will increase again with I phone 3 the near future seems good.When you develop an application for an I phone you are addressing a market of 3.9+ million.

For Android(G phone) – sticking with a single Mobile carrier will harm.. it needs to be open to more carriers and more geographical areas. The future will be too good for G phone, if they are ready to listen to what the users want. And some control on what is right and what is not while developing..Will Google do something like that to keep quality in check? or Is it the Mobile carriers who are reluctant seeing an open source platform where anyone can program stuffs which can be potentially bad for them? There is a recent buzz that recession hasn’t affected two fields – Telecoms and pro$   So we need to wait and see what all the changes coming ahead in Telecoms.

I hope to see some comments for this.

Two months back one of my key clients,  called me and asked me to change everything in the current ongoing project from prototype to Jquery. I was little shocked at the first moment because the application is already huge, Thousands if not hundreds of lines of code are written with Ajax every where. Now we need to change everything. My development team stood up and started the conversion learning the new technology in detail.

But I was thinking why such a requirement came up. I did some internet research and the client itself came up with some links showing why such a need was crucial.

See http://mootools.net/slickspeed/

Jquery is much ligther and faster compared to prototype. I agree these tests are not completely agreeable but still results are more favoring Jquery…

Keep a watch on Dojo while you run those tests.. its going to surprise you for sure…

Anoop Menon

Print all GMF diagrams with a single click

We know that GMF gives only three types of Print Options for the Print Range.

a. Current Diagram

b. All diagrams from open models

c. Selected diagrams.

Suppose consider your projects contain a large number of diagrams and you wish to Print all the diagrams. What are the options left? You need to open up each of the diagram and click print. This is a hard task when you have a large number of diagrams. I have found a better way by automating it.

Please note that there will be a lot of methods to accomplish this. This is one method I am going to detail. Since GMF doesn’t allow us to do this directly I am customizing four GMF internal classes. The classes we need to customize are

DiagramWithPrintGlobalActionHandlerProvider class

DiagramWithPrintGlobalActionHandler class

EnhancedPrintActionHelper class

DiagramPrinterUtil class

DiagramWithPrintGlobalActionHandlerProvider is the class that provides GlobalActionHandlerProvider for print functionality. This will be defined in our plugin’s plugin.xml

Before we start we need to make these four classes as custom ones. I am doing this because we cannot edit the source code of dependent plugins. If we have attached source for these classes in out target platform definition we can view the code while debugging but We cannot the edit the code for our own customization.

As the first step create four new classes in a package of your choice with names

CustomDiagramWithPrintGlobalActionHandlerProvider.java

CustomDiagramWithPrintGlobalActionHandler.java

CustomEnhancedPrintActionHelper.java

CustomDiagramPrinterUtil.java

Copy the original code itself in to these classes from its respective original counterparts. Once done that open up your plugin’s plugin.xml and change the GlobalActionHandlerProvider definition to as shown below.

<extension point=”org.eclipse.gmf.runtime.common.ui.services.action.globalActionHandlerProviders”>

<GlobalActionHandlerProvider

class=”mypackage.printing.CustomDiagramWithPrintGlobalActionHandlerProvider“

id=”mydomainPresentationPrint”>

<Priority name=”High”/>

<ViewId id=”mypackage.CustomDiagramEditorID”>

<ElementType class=”org.eclipse.gmf.runtime.diagram.ui.editparts.IGraphicalEditPart”>

<GlobalActionId actionId=”print”/>

</ElementType>

</ViewId>

</GlobalActionHandlerProvider>

</extension>

Once you are done with this, start your application using the Run configuration. Yep, We haven’t changed anything in our custom classes, but We want to make sure it works in the default way. If this works fine we can make sure that the print mechanism is actually referencing our custom classes instead of the GMF internal classes. Still having doubts? put some break points in the code and see whether our class is called while running in eclipse DEBUG mode.

Make sure this works fine. ie. The print options given above should work the same way as it existed because we have copied the original code itself to our custom classes.

So whats our ultimate AIM? We want to automate the printing of all diagrams with a single click. For that we need to make a lot of changes in code for these 4 classes that we have just made.

Step 1: Open up CustomDiagramWithPrintGlobalActionHandlerProvider.java and make changes shown below.

………………………

if (!getHandlerList().containsKey(context.getActivePart())) {

getHandlerList().put(context.getActivePart(),

new CustomDiagramWithPrintGlobalActionHandler());

………………………

return (CustomDiagramWithPrintGlobalActionHandler) getHandlerList().get(context.getActivePart());

………………………

Step 2:Open up CustomDiagramWithPrintGlobalActionHandler and make changes as shown below.

protected void doPrint(IGlobalActionContext cntxt) {

IPrintActionHelper helper = new CustomEnhancedPrintActionHelper();

helper.doPrint(cntxt.getActivePart());

}

…………………………

Step 3:open up CustomDiagramPrinterUtil.java

This is the class we are going to do a lot of coding to attain our goal.

What we are going to do is

#1.Get all the diagram files from the Workspace using our diagram file’s extension as an Array list.

#2.Ina loop we Will Open a diagram,Print the Diagram and Close the diagram. This loop will iterate for n times where, n is the size of the array list described in #1.

#1.I find all the diagram files present in my workspace using my diagram’s extension.

private static Collection getWorkspaceFiles(String extension)

{

ArrayList files = new ArrayList();

IWorkspaceRoot root = ResourcesPlugin.getWorkspace().getRoot();

IProject[] projects = root.getProjects();

for (int i=0;i<projects.length;i++)

{

try {

IResource[] resources = projects[i].members();

for (int j=0;j<resources.length;j++)

{

getChildren(resources[j], files, extension);

}

}

catch (CoreException ex)

{/*ignore*/}

}

return files;

}

private static void getChildren(IResource res, ArrayList files, String extension)

{

if (res instanceof IFile)

{

if (extension == null ||( res.getFileExtension()!=null && res.getFileExtension().endsWith(extension) ))

files.add(res.getFullPath().toString());

}

if (res instanceof IFolder)

{

try {

IResource[] resources = ((IFolder)res).members();

for (int j=0;j<resources.length;j++)

{

getChildren(resources[j], files, extension);

}

}

catch (CoreException ex)

{}

}

}

#2.Ina loop we Will Open a diagram,Print the Diagram and Close the diagram.

For anyone there will be a doubt where to put this loop and call the function calls. In my case I put this inside the code block that gets called when we select “All diagrams from open models” in the print dialog.

…………….

else if (helper.getDlgDiagramPrintRangeAll()) {

flag=1;

ArrayList files = (ArrayList)getWorkspaceFiles(”mydomain_diagram”);

System.out.println(”Number of diagrams to be printed is”+files.size());

for(int i=0;i<files.size();i++){

System.out.println(”Diagram opened is”+files.get(i));

openDiagram(files.get(i).toString());

PrintDiagram(diagramPrinter,helper);

closeDiagram();

}

…………

And the Functions for Opening and Printing and Closing are as follows:

//Method for Opening the Diagram

private static void openDiagram(String DiagramName){

URI diagramModelFilename = URI.createURI(DiagramName);

GMFResource myGMFDiagram = new GMFResource(diagramModelFilename);

try {

myGMFDiagram.load(Collections.EMPTY_MAP);

myDomainDiagramEditorUtil.openDiagram(myGMFDiagram);

} catch (IOException e1) {

e1.printStackTrace();

} catch (PartInitException e) {

System.out.println(”Error opening diagram editor”+e.getStatus());

}

}

……..

//Method for Printing the Diagram

……..

private static void PrintDiagram(DiagramPrinter diagramPrinter,IPrintHelper helper){

DiagramEditor dEditor = null;

List diagramEditors = EditorService.getInstance().getRegisteredEditorParts();

Iterator it = diagramEditors.iterator();

while (it.hasNext()) {

Object obje = it.next();

if (obje instanceof DiagramEditor) { //DiagramDocumentEditor

dEditor = (DiagramEditor) obje;

}

}

diagramPrinter.setDiagrams(Collections.singletonList(dEditor.getDiagram()));

printDiagrams(diagramPrinter, helper);

System.out.println(”File Printed”);

}

……..

//Method for Closing the Diagram

……..

private static void closeDiagram(){

IWorkbench workbench = PlatformUI.getWorkbench();

if (workbench != null) {

if (workbench.getWorkbenchWindows() != null

&&workbench.getWorkbenchWindows()[0].getActivePage() != null) {

IWorkbenchPage page = workbench.getWorkbenchWindows()[0].getActivePage();

if (page != null&& page.getEditorReferences().length > 0) {

ArrayList<Resource> resource = new ArrayList<Resource>();

for (IEditorReference editorReference : page.getEditorReferences()) {

TransactionalEditingDomain editingDomain = ((SgmoduleDiagramEditor) editorReference

.getPart(false))

.getEditingDomain();

if (editingDomain != null) {

ResourceSet resourceSet = editingDomain

.getResourceSet();

for (Resource element : resourceSet

.getResources()) {

resource.add(element);

}

for (Resource element : resource) {

resourceSet.getResources()

.remove(element);

}

}

((SgmoduleDiagramEditor) editorReference

.getPart(false)).close(false);

}

}

}

}

System.out.println(”Diagram closed”);

}

This works perfectly for me. But When you have lots of diagrams to be printed in a single stretch, make sure You allot more memory for your application else it may throw a java heap memory exception.

Hope this helps some one. Queries/Suggestions Welcome

Anoop Menon

But there is an alternative. I can copy the interesting functions from the ‘IDEApplication.java’ to my RCP application’s ‘Application.java’. And the results “BINGO” I got what I need.

You can find the full source to my Application.java from here. Please back up ur existing Application.java file as it will be easy to revert changes if anything goes wrong.

Hope this helps some one. Please leave comments/Suggestions/Queries

HTH,

Anoop Menon

Get the latest Eclipse from  here

Get the latest ZK from  here

I will come up with a good review after playing out with the new release of zk. Happy to know that firefox3 is supported.  Right now I am a bit busy with my RCP application.

Anoop Menon

What ZK is?

“ZK is an event-driven, component-based framework to enable rich user interfaces for Web

applications. ”-Definition from the creators itself.

Some stunning Features of ZK

* Rich GUI Experience

* Ability to Differentiate the Business Logic From Implementation

* Server Push Technology

* Internationalization

Rich GUI Experience

Everyone wants a rich GUI experience in their web applications. But the very need shouldn’t compromise the speed. Putting it in another way, the page should load fast. ZK includes an AJAX-based event-driven engine, a rich set of XUL and XHTML components and a markup language called ZUML (ZK User Interface Markup Language) which is not only easy to use, but also delivers a high quality GUI experience with out compromising speed.

One of the key aspect of ZK driven projects are the responsiveness of the application. At times you will scratch your head thinking whether its a desktop application!

Ability to Differentiate the Business Logic From Implementation

This one of the most important feature of ZK. It is easy to isolate the business logic completely from the User Interface. I can explain this with a simple example. I am using ZK to create a login screen which accepts user name & password from the user for authentication. First let me show you the real code and then explain how it works.

************************************Code**************************************

Login.zul

<?xml version=”1.0″ encoding=”UTF-8″?>

<div align=”center”>

<window id=”login” Title=”Login page” border=”normal”

width=”400px” mode=”modal”

use=”cyberpen.web.LoginWindow”>

<label id=”errLabel” style=”color: red”/>

<grid>

<rows>

<row>

<div align=”right”>

<label value=”User Name :”/>

</div>

<div align=”center”>

<textbox id=”username” constraint=”no empty”/>

</div>

</row>

<row>

<div align=”right”>

<label value=”Password :”/>

</div>

<div align=”center”>

<textbox id=”password” type=”password” constraint=”no empty”/>

</div>

</row>

<row spans=”2″ align=”center”>

<button label=”Login” onClick=”login.onLogin()”/>

</row>

</rows>

</grid>

</window>

</div>

LoginWindow.java

package cyberpen.web;

import org.zkoss.zul.Textbox;

import org.zkoss.zul.Window;

public class LoginWindow extends Window {

/**

*

*/

private static final long serialVersionUID = 8477939400437670121L;

public void onLogin(){

String User_Name = ((Textbox) getFellow(”username”)).getValue();

String PassWord= ((Textbox) getFellow(”password”)).getValue();

}

}

********************************End Of Code************************************

Lets see how this works. In the Login.zul file every distinct component has got a unique id. Please note that the id’s must be unique, if two components have the same id, ZK will throw an error during compilation.

<row>

<div align=”right”>

<label value=”User Name :”/>

</div>

<div align=”center”>

<textbox id=”username” constraint=”no empty”/>

</div>

</row>

As you can see the unique id for the textbox that takes the value of UserName is “username” . And We are getting this value directly in our java program. How is this possible ?

We have explicitly given in the Login.zul file that it should use the class “cyberpen.web.LoginWindow”

<window id=”login” Title=”Login page” border=”normal”

width=”400px” mode=”modal”

use=”cyberpen.web.LoginWindow“>

…….

</window>

Please note that full path to the location should be explicitly given.

Now how are we reading the value from textbox?

You can see that the Login Button onClick is mapped to the function “login.onLogin()”

<button label=”Login” onClick=”login.onLogin()”/>

Please note “login” is the unique id of the window component of our zul page which is asked to use the class “cyberpen.web.LoginWindow” and OnLogin() is the method defined inside this class.

Now we use ZK’s getFellow() method to read the value of username and password and store it in seperate variables.

String userName = ((Textbox) getFellow(”username”)).getValue();

String passWord= ((Textbox) getFellow(”password”)).getValue();

where “username” & “password” are the unique id’s of the textboxes for UserName and Password respectively.

So whats the exact point?

You can give the zul page to a web designer and ask him to decorate it while the underlying logic remains the same. Web designer need not know java or worry about any underlying stuff. Put it in converse, if you want to change the underlying business logic you don’t need to change anything in the front end.

This is worry free compared to having a long jsp file with lots of java scripts, and control flowing back and forth. It will be harder to make it and more harder to redesign later.

Server Push Technology

Server push means to put content from the server to the client actively. See http://www.zkoss.org/smalltalks/serverpush/ for more details.

Is it worth Learning?

I would say “YES” . At the end of the day, the question is “Did you make something simpler and easy to use”

ZK is a sourceforge project that is growing better day by day. See : http://www.zkoss.org/

Some Improvements required for ZK

1. The components needs to appear the same across different browsers. There are lots of hot fixes for some of the known issues please look through the ZK forum. But still there are some problems that I experienced while working on it, especially Mozilla Firefox failing to understand the height=”100%” , width=”100%” attributes.

1. Paint support is required which is due on the next release.

Viewer comments/suggestions are welcome.

HTH,

Anoop Menon

]]> http://confianzit.com/ceospeaks/2008/06/18/want-richstunning-web-applications-zk-is-right-for-you/feed/ 0 http://confianzit.com/ceospeaks/2008/06/16/convertion-of-a-plugin-to-a-standalone-rcp-application/ http://confianzit.com/ceospeaks/2008/06/16/convertion-of-a-plugin-to-a-standalone-rcp-application/#comments Mon, 16 Jun 2008 17:14:03 +0000 Anoop Menon http://cyberpen.wordpress.com/?p=5

The platform-independent Eclipse RCP architecture makes rich-client applications easy to write as the business logic is organized into reusable components called plugins. In this article, We are going to convert Mylyn 2.2 Release Plugin to a standalone RCP application. I choose Mylyn because it is one of the most popular eclipse plugin known to all. I assume that Eclipse IDE is already installed and running. I have used Eclipse Europa 3.3.0 in this tutorial.

1.Firstly we need to get the Mylyn 2.2 Release plugin and install it in our eclipse IDE. There are two different ways to have it done. One is to install the Plugin from eclipse itself using the Software update option. The procedure is nicely illustrated here.

Second method is to install the plugin manually. For that get the Mylyn 2.2 Release archive from here . Uncompress it and copy the contents of plugin directory to the eclipsebasedir/plugin and contents of features directory to the eclipsebasedir/features.

2.We need to restart the Eclipse IDE for the newly added plugins to work. For that close the eclipse IDE if it is already running and start it cleanly. We will use eclipse.exe -clean from command prompt in windows or eclipse -clean from shell prompt in Linux.

3.Now we need to Verify that the newly added mylyn plugin is working in our Eclipse IDE. Mylyn uses a perspective called “Planning”.In Eclipse, a perspective is a way to organize and view the files associated with our program. We can open the perspective from Window -> Open Perspective -> Other

We will select the perspective “Planning” and continue by clicking the Ok button.

Now Eclipse has opened the “Planning” perspective for us which shows a Task List view. To verify that the plugin is working properly, we can create a new Task using the New Wizard. For that Go to File ->New ->Task

A New Task will be opened for us and it will be listed in the Task List as shown below.

Now we have verified that the mylyn plugin is working properly. It is time to move on to our actual goal ie. Converting this Plugin to a standalone RCP application.

5.We start by creating a new Plugin project in eclipse using the New wizard.

For that go to File-> New-> Project

We will select Plugin Project and continue by clicking next.

6.We need to fill in a suitable name for our project for eg.“mylyn” and continue by clicking next.

7.In the next Screen we will see an option “would you like to create a Rich Client Application “ with radio buttons “Yes” or “NO”. Since we are going to create a Rich Client Application enable it by selecting “Yes” and continue by clicking the Next button.

8.Now Eclipse will show us a list of available Templates for our RCP. We will take the simplest of all “Hello RCP” template. Firstly we will enable “Create a plugin using one of the templates” by clicking the checkbox. We will select “Hello RCP” and continue by clicking Next. Please note “Hello RCP” is the simplest template available to which we add only those components that our application needs.

9.In the next screen Eclipse will allow us to name our application. Give an appropriate name

eg. “Mylyn Application”.This name comes on the title bar of our application. Give an appropriate package name too. I have used “mylyn” in this tutorial. Branding is an option provided by eclipse to brand our application. Branding includes adding splash image,custom icons etc. Enable branding as we will be branding our application in the final stage. Move on by clicking Finish.

If you had followed with me in this tutorial, eclipse will be in “Planning perspective” and will prompt to switch to “Plugin-in Development perspective”. Please select “yes” at this prompt.

6.Eclipse creates the Project for us and opens up the manifest file as shown below.

7.So far we have created an RCP application using the “Hello Template” and its time to see for ourself how it looks like. We can start the application from eclipse by Clicking the “Launch an Eclipse Application” link present in the manifest file. The application pops up as shown below. This is the basic building block of our Mylyn RCP application and we will add features to it in an iterative manner.

8.We can notice that the application screen area is small. This is because by default the size is set to (400,300) in a method preWindowOpen() inside ApplicationWorkbenchWindowAdvisor class.We will open up the ApplicationWorkbenchWindowAdvisor .java and set it to (800,600). Change the setShowCoolBar value to true as we will be adding some Coolbar menus to spice up our application.

9.Right now our application is a plain template. There are no menus or icons or Actions. Eclipse allows us to define and declare all the actions/menus that we need in our RCP application. All these are to be done in ApplicationActionBarAdvisor class. So we will open this file and define all the menus/actions/sub menus etc that we want for our entry level application programaticaly.

So after adding all the Actions/Menus etc we want for our Mylyn application, our ApplicationActionBarAdvisor.java looks like as shown below.

package mylyn;

import org.eclipse.core.runtime.IExtension;

import org.eclipse.jface.action.IContributionItem;

import org.eclipse.jface.action.ICoolBarManager;

………

public class ApplicationActionBarAdvisor extends ActionBarAdvisor {

//Declaring all the Workbench Actions

private IWorkbenchAction newAction;

private IWorkbenchAction saveAction;

private IWorkbenchAction openPerspective;

………….

public ApplicationActionBarAdvisor(IActionBarConfigurer configurer) {

super(configurer);

protected void makeActions(IWorkbenchWindow window) {

//All the actions are defined here.

newAction=ActionFactory.NEW.create(window);

newAction.setText(”New”);

register(newAction);

saveAction=ActionFactory.SAVE.create(window);

register(saveAction);

openPerspective= ActionFactory.OPEN_PERSPECTIVE_DIALOG.create(window);

register(openPerspective);

…………..

}

protected void fillCoolBar(ICoolBarManager coolBar) {

//Adding the Actions to the CoolBar.

IToolBarManager toolbar = new ToolBarManager(SWT.FLAT | SWT.RIGHT);

coolBar.add(new ToolBarContributionItem(toolbar, “main”));

toolbar.add(newAction);

toolbar.add(saveAction);

………….

}

protected void fillMenuBar(IMenuManager menuBar) {

//Adding the Actions to the Menu Bar.

MenuManager fileMenu = new MenuManager(”&File”,IWorkbenchActionConstants.M_FILE);

menuBar.add(fileMenu);

MenuManager windowMenu= new MenuManager(”&Window”,IWorkbenchActionConstants.WINDOW_EXT);

menuBar.add(windowMenu);

………….

fileMenu.add(newAction);

fileMenu.add(new Separator(IWorkbenchActionConstants.MB_ADDITIONS));

fileMenu.add(new Separator());

fileMenu.add(saveAction);

windowMenu.add(openPerspective);

………….

}

}

10.So we have completed adding the Menus/Actions for our application and it is time to define a product Configuration for our mylyn project. A product configuration is required to brand or package our final application. Product configuration manages information about splash screens,launcher icons,about text, plugin and features list,window images etc together for our application. To create a product configuration right click the mylyn project Go to New-> Product Configuration as shown below.

11.We need to give a name for the product configuration eg.“mylyn.product” and continue by clicking finish.

12.Eclipse will open the Product configuration file in a manifest editor for us. We can notice that eclipse has automatically selected “mylyn.application” as the application to run on launching this product. If it is not, select the same from the drop down menu.

13. So far so good. We need to check whether the application is working from our product configuration. For that click “Launch an Eclipse application” link in the product configuration file which will pop up the application. Make sure all the menus that we added in step 9 shows up but these menus wont work unless we include all the necessary Plugins. We can notice that the application comes up in (800*600) mode.

14.Now its time for us to include all the necessary mylyn Plugins and Fragments to our product configuration. We will do that by clicking the Configuration tab in the Product configuration file. We can see that all the plugins and fragments required for the basic application from template are already shown here.

15.For adding the mylyn Plugins and fragments to the product configuration, we will Click the add button which brings up a pop up list of Plugins and fragments. We need to scroll down and select only those mylyn Plugins which are required. Click the Ok button to continue. Now we need to resolve the plugin dependencies. Click “Add Required Plugins” link and eclipse will automatically add the dependencies for us. We need to save the project for the changes to take effect.

16.So far we have created a base application,defined the menus/actions, defined a product configuration, added the mylyn Plugins and fragments, resolved the dependencies. Now its time for us to check whether everything plug-in well and works well together. For that we need to get back to the “overview” Tab. We need to Click the Synchronise link. This ensures that all the changes that we have made in the product configuration and the mylyn.application synchronize with each other. We can start the application by clicking “Launch an Eclipse application” link present in product configuration.

17.We can see that the Mylyn Application starts in a new window. We can check whether the perspective “planning”,the views,creating a new task etc work successfully following the same procedure we have done in step 3.

18.We need to give a name for our application launcher and provide custom icons for our application. Eclipse provides a variety of tools to get it done. In the product Configuration click the “Launching Tab”. Give a Launcher name eg.“mylyn” and click save. We can provide the icon files required for branding here.(Note:-I have not done this below)

19.We want a standalone mylyn RCP application. So we will Export the application from our eclipse environment to our operating system’s directory structure. For doing this, in the product configuration file click “Eclipse Product Export wizard” which will bring up a pop up. We want the Root directory to be called “mylyn” and not the default name “eclipse”.Change the root directory name to “mylyn”,make sure the “synchronization” checkbox is checked. Browse through the directory structure to the folder to which we want to export the application and click finish to export.

20.Eclipse will export the application for us and if it is successful no errors will be thrown in the eclipse error log or problems view.

21.So everything done and its time for us to check out the final application. Browse through the directory structure to the folder to which we exported the application. We can start the application by clicking mylyn.exe. Make sure everything works creating a task,the views the planning perspective etc. Final application screen shot is shown below.

So thats it. We have successfully completed converting mylyn plugin to a standalone RCP application.

Please note the full code of ApplicationActionBarAdvisor.java can be obtained from here.

I hope this post would help some one.Please feel free to leave comments/Suggestions/Correct me.

HTH,

Anoop Menon

]]> http://confianzit.com/ceospeaks/2008/06/16/convertion-of-a-plugin-to-a-standalone-rcp-application/feed/ 16