Matt RaibleMatt Raible is a Web Architecture Consultant specializing in open source frameworks.


Over 10 years ago, I wrote my first blog post. Since then, I've authored books, had kids, traveled the world, found Trish and blogged about it all.

Maven Plugin for Running Integration Tests against Multiple Containers

Don Brown has created a pretty cool Maven plugin that allows you to run integration tests on multiple containers in one go. I learned about it on the Struts 2 Dev List:

I've started adding functional tests to Struts 2 by adding a few to the REST showcase application, running against Tomcat 5.x, Jetty 6.x, JBoss 4.2.x, and Resin 3.x. The magic happens through a new Maven 2 plugin I developed called maven-itblast-plugin, which enables multiple integration test runs against multiple containers in one go. For more info, see

You might notice that this plugin is hosted on GitHub. If you want to learn how to use GitHub with Maven, you might want to see Don's entry on Maven-enabled project hosting with GitHub. Thanks Don - good stuff.

Update: Don has posted more information about this plugin on his blog.

Posted in Java at Jul 31 2008, 05:39:23 PM MDT Add a Comment

Issues with AntRun Plugin and Maven

I started seeing the following error today when using Maven and the AntRun Plugin.

[INFO] [antrun:run {execution: default}]
[INFO] Executing tasks
[INFO] ------------------------------------------------------------------------
[INFO] ------------------------------------------------------------------------
[INFO] Error executing ant tasks
Embedded error: java.lang.IllegalAccessError: tried to access method;)Ljava/lang/String; 
from class

Searching the internet provided no results, so I was pretty stumped - especially since this error didn't happen on my MacBook Pro. It happened on AppFuse's Bamboo server (Linux), but not locally. Luckily, I was able to reproduce it on my Windows box and discovered the solution: upgrade to a newer version of Maven. I was using 2.0.6/2.0.7 and upgrading to 2.0.9 fixed the problem.

BTW, when is the Ant project going to release a new version of Ant? The current 1.7.0 version doesn't support spaces in path names, which seems like a pretty big issue to me (especially for Windows users).

Posted in Java at May 10 2008, 06:23:26 PM MDT 6 Comments

Running Spring MVC Web Applications in OSGi

For the past couple of weeks, I've been developing a web application that deploys into an OSGi container (Equinox) and uses Spring DM's Spring MVC support. The first thing I discovered was that Spring MVC's annotations weren't supported in the M1 release. This was apparently caused by a bug in Spring 2.5.3 and not Spring DM. Since Spring DM 1.1.0 M2 was released with Spring 2.5.4 today, I believe this is fixed now.

The story below is about my experience getting a Spring MVC application up and running in Equinox 3.2.2, Jetty 6.1.9 and Spring DM 1.1.0 M2 SNAPSHOT (from last week). If you want to read more about why Spring MVC + OSGi is cool, see Costin Leau's Web Applications and OSGi article.

To get a simple "Hello World" Spring MVC application working in OSGi is pretty easy. The hard part is setting up a container with all the Spring and Jetty bundles installed and started. I imagine SSAP might solve this. Luckily for me, this was done by another member of my team.

After you've done this, it's simply a matter of creating a MANIFEST.MF for your WAR that contains the proper information for OSGi to recognize. Below is the one that I used when I first tried to get my application working.

Manifest-Version: 1
Bundle-ManifestVersion: 2
Spring-DM-Version: 1.1.0-m2-SNAPSHOT
Spring-Version: 2.5.2
Bundle-Name: Simple OSGi War
Bundle-SymbolicName: myapp
Bundle-Classpath: .,WEB-INF/classes,WEB-INF/lib/freemarker-2.3.12.jar,
Import-Package: javax.servlet,javax.servlet.http,javax.servlet.resources,javax.swing.tree,

Ideally, you could generate this MANIFEST.MF using the maven-bundle-plugin. However, it doesn't support WARs in its 1.4.0 release.

You can see this is an application that uses Spring MVC, FreeMarker, SiteMesh and the URLRewriteFilter. You should be able to download it, unzip it, run "mvn package" and install it into Equinox using "install file://<path to war>".

That's all fine and dandy, but doesn't give you any benefits of OSGi. This setup works great until you try to import OSGi services using a context file with an <osgi:reference> element. After adding such a reference, it's likely you'll get the following error:

SEVERE: Context initialization failed
Configuration problem: Unable to locate Spring NamespaceHandler for
XML schema namespace []

To fix this, add the following to your web.xml (if you're using ContextLoaderListener, as an <init-parameter> on DispatcherServlet if you're not):


After doing this, you might get the following error on startup:

SEVERE: Context initialization failed
org.springframework.context.ApplicationContextException: Custom
context class []
is not of type [org.springframework.web.context.ConfigurableWebApplicationContext] 

To fix this, I change from referencing the Spring JARs in WEB-INF/lib to importing the packages for Spring (which were already installed in my Equinox container).

Bundle-Classpath: .,WEB-INF/classes,WEB-INF/lib/freemarker-2.3.12.jar,

After rebuilding my WAR and reloading the bundle in Equinox, I was confronted with the following error message:

SEVERE: Context initialization failed
org.springframework.beans.factory.BeanCreationException: Error
creating bean with name 'freemarkerConfig' defined in ServletContext
resource [/WEB-INF/myapp-servlet.xml]: Instantiation of bean failed;
nested exception is java.lang.NoClassDefFoundError:

As far as I can tell, this is because the version of Spring MVC installed in Equinox cannot resolve the FreeMarker JAR in my WEB-INF/lib directory.

To prove I wasn't going insane, I commented out my "freemarkerConfig" and "viewResolver" beans in myapp-servlet.xml and changed to a regular ol' InternalResourceViewResolver:

<bean id="viewResolver" class="org.springframework.web.servlet.view.InternalResourceViewResolver">
    <property name="prefix" value="/"/>
    <property name="suffix" value=".jsp"/>

This worked and I was able to successfully see "Hello World" from a JSP in my browser. FreeMarker/SiteMesh worked too, but FreeMarker didn't work as a View for Spring MVC.

To attempt to solve this, I create a bundle for FreeMarker using "java -jar bnd-0.0.249.jar wrap freemarker-2.3.12.jar" and installed it in Equinox. I then change my MANIFEST.MF to use FreeMarker imports instead of referencing the JAR in WEB-INF/lib.


Unfortunately, this still doesn't work and I still haven't been able to get FreeMarker to work with Spring MVC in OSGi. The crazy thing is I actually solved this at one point a week ago. Shortly after, I rebuilt Equinox from scratch and I'm been banging my head against the wall over this issue ever since. Last week, I entered an issue in Spring's JIRA, but thought I'd fixed it a few hours later.

I've uploaded the final project that's not working to the following URL:

If you'd like to see this project work with Spring MVC + JSP, simply modify myapp-servlet.xml to remove the FreeMarker references and use the InternalResourceViewResolver instead.

I hope Spring DM + Spring MVC supports more than just JSP as a view technology. I hope I can't get FreeMarker working because of some oversight on my part. If you have a Spring DM + Spring MVC application working with Velocity or FreeMarker, I'd love to hear about it.

Posted in Java at Apr 30 2008, 12:42:34 AM MDT 14 Comments

Maven Integration for Eclipse

If you're a Maven user and like Eclipse, you might want to checkout the new Maven Integration Plugin for Eclipse 0.9.0. Euxx has a couple blog posts talking about the new features - looks like pretty cool stuff to me.

The second feature is especially cool. If your dependencies supply SCM information - you can import the project from its source control system. Maven may have warts, but it also has incredible potential.

Posted in Java at Mar 12 2008, 04:22:43 PM MDT 2 Comments

Maven now supports attributes in pom.xml?!

In December 2005, I asked Is it possible to make pom.xml simpler?.

After seeing what the Spring Developers have done to simplify Spring context files, I can't help but think the same thing is possible for Maven 2's pom.xml. Is it possible to add namespaces and make something like the following possible?




<dep:artifact name="org/springframework/spring" version="1.2.6"/>

Or just allow attributes to make things a bit cleaner?

<dependency groupId="org.springframework" artifactId="spring" version="1.2.6"/>

At that time, the general response was "That's how Maven works. It's a matter of taste. You'll get used to it." It's been two years and sure, I'm used to it, but I'd still rather write less XML. That's why I was particularly pleased to see Brett Porter's write Maven now supports condensed POMs using attributes:

The issue is being tracked under MNG-3397.

The result is that something like this:


Halves in length to something like this:

<dependency groupId="junit" artifactId="junit" version="3.8.1" scope="test"/>
<dependency groupId="easymock" artifactId="easymock" version="1.2_Java1.3" scope="test"/>

Now that wasn't so hard was it? ;-)

Posted in Java at Feb 11 2008, 03:45:57 PM MST 4 Comments

Building a Better Maven with Ant

It looks like the Ant folks are thinking of building a better Maven.

I see many developers adopt Maven because they want a build system able to provide common features with no effort. Most of them don't want to spend much time writing an Ant script, or have seen or heard that maintaining Ant build scripts is troublesome. So they choose to use Maven only because it's easy to use for common use cases: install, write a simple pom of a few lines or generate it using an archetype, and you're ready to compile, test and package your new project following the Maven standard structure. They also get dependency management for free, and with only a few more effort they have multi module builds, and some nice features like code analysis, coverage, and a set of report gathered in a web site. That's really nice and that's what I like about Maven.

But Maven suffers from a lack of flexibility and robustness IMHO. And later the same people who first adopted Maven because of its perceived ease of use become frustrated when they need to tweek the system to their own needs or don't understand how the release plugin work. Then some of them go back to Ant, first having to go through a sometimes painful road to describe their whole build system in xml, especially if they aren't Ant experts. Others try to use new build tools like raven, buildr or others.

I really like Ant, and think it is a very good basis for robust and flexible build systems. People with enough knowledge of Ant can write very good build systems, testable, maintainable and adaptable. But you need to get your hands dirty, and you need to get a good knowledge of some of the mechanisms which can make an Ant based build system manageable: import, scripts and scriptdef, macrodef, presetdef, and so on. [Read More]

What do you think - is this a good idea?

I agree that Maven has its warts, but I don't think it's that bad. I've also heard that Maven has been successfully implemented at large companyies like eBay, Intuit and E*Trade[1]. Is the "Maven sucks" meme largely something that exists in the blogosphere, but not in the real world?

I think the biggest benefit of Maven is dependency management. I think it makes your code more modular and easier to build. Rather than having a monolithic source-code tree that depends on itself being built in a certain order, you can have individual modules that pull dependencies from a central location. This can be done with Maven's Ant Tasks as well. I don't see a problem with building a better Maven with Ant, but to try and build a better Central Repository sounds like a nightmare to me. The current repository has been improved for years and is much better than it was a couple years ago. That being said, I would love to see somebody build a more accurate Central Repository. Ideally, it'd be done sometime next week. ;-)

[1] I could be wrong about these companies. If you're a developer at one of these companies, please confirm or deny. Any comments on Maven's success at these companies would be great as well.

Update: Speaking of Maven, there's an interesting comment on a Javalobby post I wrote:

With all the critical remarks the Maven project is receiving, wouldn't it be time for some Maven project lead to step up and explain the team's position? Or is it completely deaf to the sentiments? How many builds have to fail, how much more headaches are needed before others start their own version of Maven and do it the right way (like Don [Brown])?

Seems like an excellent question to me. Guys?

Posted in Java at Feb 11 2008, 02:07:12 PM MST 18 Comments

Maven 2 Archetypes get a much needed improvement

Yesterday, a new version of the Maven Archetype Plugin was released. This release incorporates many of the improvements that were developed in a different project - code named "Archetype NG". The two major improvements are 1) you only have to use "mvn archetype:create" now and 2) you can create archetypes from existing projects.

I haven't tried #2, but #1 seems to work pretty well (especially since AppFuse archetypes are the first 9 ;-)).

[Read More]

Posted in Java at Feb 09 2008, 03:20:00 PM MST 15 Comments

RE: Why Grails doesn't use Maven

Graeme Rocher's in Why Grails doesn't use Maven:

In his post entitled "Grails - The Good, The Bad and the Ugly", Jonas has some nice praise for Grails, his main beef is that it is not built on Maven.

So I wanted to clarify why exactly we chose not use Maven (by default) and the explanation is there for all to see in Jonas' first example of creating a Grails application vs creating a Maven project:

Instead of

grails create-app name

could be just

mvn archetype:create -U\
-DgroupId=your.proj.gid -DartifactId=your-proj-id

My goodness, what a mouthful the Maven example is. There is a common acronym in the open source world called RTFM (read the *ing manual), when a user asks a question on a mailing list and the "experts" respond by pointing them to the place in the manual.

I think Maven's biggest problems are 1) poor metadata in the central repository and 2) the source of metadata in projects (pom.xml).

I believe #1 can be fixed if the Maven guys allow dependencies to be fixed based on user feedback. It's also gotten a lot better in recent years. In reality, maintaining transitive dependencies is hard and I believe Maven has done a good job. In reality, they're the only ones that slurp up transitive dependencies, so the only other option is to maintain the dependencies yourself.

To fix #2, I think the problem is mainly XML and the verboseness of the elements-only pom.xml that Maven requires. Most of the contents of a pom.xml are either dependencies, plugins or exclusions/variances of Maven's conventions. What if Maven's metadata was pluggable? What if XML was only one option? What if you could write a pom.groovy and describe your entire build process in 5 lines instead of 500? That would be very cool.

I'm still a Maven fan, mostly because it's greatly simplified the maintenance of and releasing of AppFuse. When I do GWT, Seam or Grails development in the future, you can be sure I'll try to use Maven to do the development. Why? Because I've learned how to use it and I don't feel the pain that so many others talk about. I also think it really shines on really large projects (builds that produce 30+ WARs for example). An Ant-based system on really large projects can become quite burdensome and difficult to maintain. Not only that, but it's very difficult to maintain a modular build system (where you can build/test/deploy a single WAR) with Ant. In my experience, really large Ant-based systems take forever to process that everything is up-to-date whereas Maven systems depend on each other and require you to keep them up to date. Sure it requires you to be smarter and run "mvn install" on your subprojects, but I'd rather do that than wait 5 minutes for Ant to process everything just to run a test.

You might remember that the main reason I used to prefer Ant over Maven was speed. That was in Maven 1 days. With Maven 2, speed is no longer a problem and I've found it much easier to run "mvn jetty:run" than "ant deploy" and wait for Tomcat to restart. IMO, the perfect development environment is one were you can run a command-line command (or use your IDE to start the server) and code away without worrying about restarts. Seam and Grails offer this environment, but it's unlikely your entire organization is going to use standardize on those frameworks and not have anything else. I think Maven and the Maven Jetty Plugin offer a nice alternative for the rest of those applications.

Posted in Java at Jan 16 2008, 10:49:35 AM MST 11 Comments

Upgrading AppFuse to Spring 2.5

Last night, I spent a few minutes upgrading AppFuse to Spring 2.5 RC1. According to InfoQ, Spring 2.5 is a drop-in upgrade for Spring 2.0. However, if you're using Maven, it's not quite that easy. The good news is it is easy, you just need to change your pom.xml a bit. The steps I used to upgrade AppFuse are listed below:

  • Add a repository for Spring's milestone releases:
  • Change artifactId of "spring-mock" to be "spring-test".
  • Change version to be 2.5-rc1.

At this point, if you're using "spring" as your artifactId (instead of the smaller fine-grained dependencies), you'll likely get the following error in a Spring MVC application:


This happens because Spring MVC is no longer included in the uber spring.jar. You'll need to add a dependency on "spring-webmvc" to solve this problem. Unfortunately, this JAR is dependent on the fine-grained modules, so you may have to modify your pom.xml to depend on the fine-grained modules - or exclude them all from spring-webmvc.

The good news is Spring has excluded all the invalid commons-logging dependencies for you so you don't have to anymore.

After getting all the dependencies straightened out - I ran the integration tests:

org.springframework.beans.NotReadablePropertyException: Invalid property 
'fileUpload' of bean class [org.appfuse.webapp.controller.FileUpload]: Bean 
property 'fileUpload' is not readable or has an invalid getter method: Does the 
return type of the getter match the parameter type of the setter?

Looking at uploadForm.jsp, I'm guessing the problem happens because of the following code:

<spring:bind path="fileUpload.file">
<input type="file" name="file" id="file" class="file medium" value="<c:out value="${status.value}"/>"/>

Confirmed - changing the "path" attribute to "file" fixes the problem. I also found out that setting the "value" on an <input type="file"> doesn't work, so wrapping the field with <spring:bind> doesn't make a whole lot of sense anyway.

To conclude, it doesn't look like the first release candidate of Spring 2.5 is exactly a drop-in upgrade for Spring 2.0, but it's pretty darn close. I'm sure by the time it's released, it will be. I'd encourage you to try 2.5 in your Spring-dependent projects to see if you find any issues.

Update: I was successfully able to migrate AppFuse from using the uber JAR to fine-grained JARs. However, I ran into a couple issues in the process. The first is that even though I'm including spring-aop in the appfuse-service module, it's not pulled in for the web frameworks (which depend on appfuse-service). Explicitly declaring spring-aop as a dependency for the appfuse-web module fixes this. Secondly, I had to modify my Acegi Security exclusions so it wouldn't include dependencies that no longer exist in 2.5.


Posted in Java at Nov 07 2007, 08:27:20 AM MST 3 Comments

Flex and Grails Made Easy

I love how easy it is to start new projects these days. It was very difficult when I started creating AppFuse way back in 2002. We've come a long way baby!

Here's a couple of easy ways to get started with Flex and Grails:

I hope to develop with Flex, Grails, GWT or YUI + Struts 2 in the next 6 months. These seem like the most exciting technologies for Java web development in 2008.

Posted in Java at Nov 01 2007, 11:00:38 PM MDT 5 Comments