A new release of Q4E, the Eclipse plugin for Maven, is out. I’m going to quote Abel announcement, only adding something that he forgot, the new 0.5.0 release we are working on has initial support for importing parent projects (pom projects).

After some delay (you can blame it on Christmas� we committers need to have some fun far away from the computer), there is a new version of q4e waiting for you.

This version is 0.4.0 and has some interesting new features. Probably the best one is that it is much faster.

On the fancier side of things, the dependency viewer has been cleaned a bit and now more cleanly draws the dependencies. Judge for yourself:

For 0.5.0, we have a new contributor, Jake Pezaro who is working on integrating his dependency analysis tool into q4e.

Also, we are introducing a new and noteworthy page for q4e, and you can follow the current status from the wiki (what�s pending, what�s done and previous releases� changelog). You can use the list of pending tasks to take a look at what we�re trying to get into 0.5.0� but also to help us! We always welcome helping hands (remember that you can look for easy tasks to get started).

If you like living on the edge, the development update site already has a preview of the upcoming 0.5.0 version. Come and try it!

After some time trying to figure out what the best conversion from Eclipse plugins / OSGi bundles naming conventions to Maven naming I had finally put many Eclipse plugins in the central Maven repository (see only the Eclipse plugins that got copied in this batch). It’s sometimes a bit tricky, but the main goal was a bidirectional automated translation from Bundle-SymbolicName and Bundle-Version to groupId, artifactId and version in Maven.

Basically you take the BundleId and use the dots to split it. The last section is the artifactId, the rest is the groupId. For the versions they are pretty much the same, only changing dashes and underscores. There’s a maven-osgi library very light that you can use to do these conversions, used by the Maven Eclipse and Felix bundle plugins.

This improves the support for building Eclipse plugins from Maven, although you can do some things already, or using Eclipse libraries from Maven projects. We’ll see more activity in these two fields soon.

Some people complained that you get weird jar names like net-1.0.0-I20070531.jar. Something I’d like to point out is that the repository is just an interface, the way artifacts are stored internally shouldn’t matter to the clients (Maven, Ivy,…). Plugins that generate a flat directory with jars (like assembly, war,…) should take care of renaming the jars to groupId.artifactId, to avoid clashes in the namespace, and so you get a meaningful name out of a hierarchical directory structure.

You are lucky, the Eclipse integration for Apache Maven (Eclipse IAM) proposal has been approved, and is the first step to have a project at the Eclipse Foundation.

The proposal is based on the work done at Q4E and we hope that by taking it into the next level at the Eclipse Foundation more people will be interested and will collaborate to make it completely successful and hopefully part of the Eclipse IDE distribution.

Please join us at the eclipse.technology.iam newsgroup at news.eclipse.org (web interface) for suggestions, discussions and information on the project.

Unfortunately the OSSummit Asia that I talked about last week has been postponed for 2008. It’s definitely a pity.

If you happen to be close by Hong Kong ( who isn’t? 😉 ), check out OSSummit, 26-30 November, the first joined conference between The Apache Software Foundation and Eclipse Foundation. I’ll be giving a talk Maven, Eclipse and OSGi working together.

We also hope to have a Birds-of-a-feather (BOF) session during the conference for those interested in Maven.

So Kepler is finally going through the creation review this Wednesday and hopefully in a matter of a couple of weeks it will be setup as a proper Eclipse project with source control, issue tracking and all those goodies that will allow us to commit our code and show the world what we’ve been working on.

And some more pictures…

Greenland

Grand Canyon

Grand Canyon panorama

Grand Canyon

Grand Canyon

Brett has posted the interview where we talked about the projects I’ve been working with lately, including Kepler and Q4E, so he saved me from writing a new blog entry 😉

And finally found some time to upload more pictures from my trip to China.

Shanghai skyline

Oriental Pearl Tower

Dragon in Yuyuan Garden

Yuyuan Garden

China World Trade Center

Starbucks in Beijing

Drum and Bell towers

Beijing

Beijing

Tian’an Men Square

Tian’an Men Square

Forbidden City

Forbidden City

Forbidden City

Forbidden City

Forbidden City

After a week in Cebu and Manila preparing the launch of a new project (more details to come…) and having more fun than ever thanks to the team there and the Exist people, a 24 trip to Frankfurt via Kuala-Lumpur, and driving as-fast-as-you-can in the Autobahn, I’m finally at the Eclipse Summit Europe 2007 in Ludwisburg with Brett Porter. BTW I added to the right a Dopplr banner so you can see my future trips in case we coincide, for instance I surprisingly met Ben Alex from Acegi security (Spring-security) in Manila who I knew for 3+ years by collaborating in Acegi but never met in person, even when he traveled to Spain or the USA.

Back to Eclipse Summit, this morning was dedicated to Equinox provisioning and afternoon to Server Side Equinox, learning what’s going on in the other projects and talking about Kepler. For those interested I have posted in the Kepler wiki as example on how Kepler and P2 metadata of an Eclipse plugin look like

• org.eclipse.jdt.junit
• in Kepler http://wiki.eclipse.org/Org.eclipse.jdt.junit_in_Kepler
• in P2 http://wiki.eclipse.org/Org.eclipse.jdt.junit_in_P2
• org.eclipse.equinox.common
• in Kepler http://wiki.eclipse.org/Org.eclipse.equinox.common_in_Kepler
• in P2 http://wiki.eclipse.org/Org.eclipse.equinox.common_in_P2

There are some things in the P2 Instalable Unit model that are not yet in Kepler but could easily be implemented by adding a new facet (extension points)