Eclipse on a Destroyer

Every once in a while, my Google alerts brings up something that makes me say ‘that is kind of cool’.  For instance, today this announcement that a fleet of US Navy destroyers will be using a Java rules engine, called Jesse, that is built on top of Eclipse.

I realize there are lots of rules engines built on top of Eclipse but having Eclipse deployed on a destroyer sounds kind of neat.   I also didn’t realize that Sandia National Labs was doing Eclipse development; kind of cool too.

Make your vote count: Vote today

For those getting bored about the ongoing blah blah about Clinton vs Obama or McCain vs Romney or you think US politics just doesn’t matter, I give you something that matters and where you can have a direct impact. The voting for the Eclipse Community Awards is now open.

We have a strong slate of candidates in the three different categories.

Beginning with the Top Ambassador, the nominees are:

• Jeff McAffer
• Ed Merks
• Tim Barnes
• Phillippe Ombredanne

For Top Committer, a full slate of nominees include:

• David Williams
• Mik Kersten
• Paul Webster
• Nick Boldt
• Chris Aniszczyk
• Trip Gilman

And finally the nominees in the Top Contributor category are:

• Jacek Pospychala
• Remy Chi Jian Suen

Good luck to all the nominees. From what I can tell, they all are great community members. Now it is your turn to vote and have a say in recognizing the top individuals in our community. Voting ends February 21 at 5pmEST.

An Eclipse IRC Superstar - Remi Suen

As Chris pointed out the Eclipse community is really embracing IRC. I am a lurker on the #eclipse channel and the one thing that has always impressed me is the awesome support provided by the community.

In particular, Remi Suen is machine in responding to questions. Yesterday, I found this site that logs the #eclipse irc channel, including stats. Sure enough Remy is the top contributor; almost 5x the number of lines.

It is people like Remy that really make Eclipse a great community! Thanks to Remy, #eclipse IRC is an important part of that community.

Understanding how people use Eclipse?

Understanding how and what people use in Eclipse has always been a challenge. To help address this issue, we are planning to introduce the Eclipse Usage Data Collector (UDC) in the upcoming Ganymede release. The intention of the UDC is to track how people use Eclipse, so as Chris has already pointed out, making it easier to know which features are being used and which aren’t.

The plan is to include the UDC in the packages created by the Eclipse Packaging Product (EPP). The UDC code is therefore being contributed and maintained as part of EPP. If you would like to test out the UDC, there is a version included in the Ganymede M4 packages. We would really appreciate your feedback now, so we can make it better for the final release. Please use bugzilla for any problems or feedback.

As with any tracking feature, concerns about privacy are always top of mind. To address these concerns we are doing two things: 1) all information collected by the UDC and sent to the Foundation server is sent anonymously, 2) individuals will have to opt-in before any information is sent. A version of the terms of use can be found here.

It will be great to hear people’s feedback about this feature and any ideas on how we can improve it.

Testing Ganymede M4

For those that like to be using the latest and greatest, the Ganymede M4 packages are now available from the Eclipse Packaging Project (EPP) download site.

This year EPP is going to produce packages for each Ganymede milestone and release candidate.  This will hopefully make it easier to test the packages before the final release and find cross-project issues.  Of course if you find issues with the packages, please report them in bugzilla.

Open Source Models

Dana Blackenhorn recently wrote two interesting articles about different approaches to open source development. The first article ‘The open source development incline‘ identifies three basic models:

• At the top of the incline are what I will call proprietary open source projects, owned by one company. Sun works on this model.
• In the middle of the incline we might put shared corporate projects, independent of any one sponsor, like Eclipse.
• At the bottom are true community projects, which don’t seem to be owned by anyone. Drupal is an example.

He then follows up with a good example of how two companies, Sun and IBM, leverage open source.

I think what Dana is capturing is the fact that there are different models for open source governance; something that I have mentioned before.

In my opinion each has its strengths, for instance:

1. Proprietary open source projects tend to have the benefits that typically come from commercial software companies. They tend to have a bit more polish, well define support options, executive escalation paths (sometimes referred to as ‘one throat to choke), and even indemnification.

2. Shared corporate projects, like Eclipse, Linux and even in some instances Apache, tend to have a more diverse community of contributors and participants. This diversity often helps foster wider adoption, creates a wider ecosystem of technology providers around the project, and tends to be more organic in its evolution. In fact, I would claim this model enables more innovation around a certain technology platform.

3. True community projects have the raw ability to ’scratch an itch’ and prospers if others share in the vision.

All three models are valid and needed. However when thinking about adopting or participating in an open source project, understanding the model chosen is very important.

Last week for nominations of the Eclipse Community Awards

The deadline to send in a nomination for the Eclipse Community Awards is next Friday, January 25.

In the individual category, we already have lots of great nominees.  However, I know there are others  that deserve recognition, so please take the opportunity to nominate someone you think deserves to be the Top Ambassador, Top Committer or Top Contributor.

In the technology category, we traditionally get 80% of the nominations in the last 2 days.  Why do people procrastinate so much?  Therefore, I expect to see a lot of products and open source projects added to the list of nominees.  As a reminder, we have added a new Equinox category this year, so if you have built an open source project or commercial product on top of Equinox, this is your chance for fame and fortune; well at least fame.

Is Industry Specific OSS the Next Big Thing?

It has been my belief that the next big thing for open source software (OSS) will be creating industry specific tools, frameworks and platforms. OSS is well established in the tools and middleware stacks but a lot of the benefits are also applicable to software higher up the stack.

Recently some other people have started talking about this, Bob Sutor made the point in his 2008 priorities, Alex Fletcher makes a strong point how OSS serve as organic collaboration networks, and new companies are being established to promote the concept of industry specific OSS. At Eclipse we also have a number of new projects that are industry specific, such as OHF, OSEE and a proposal for a Financial Market platform. There are also some successful open source projects like Sakai in education. However, I think we are just scratching the surface?

How can IT departments benefit from OSS?

OSS has demonstrated an effective process of allowing individuals and organizations to collaborate on software development, including competing interests. Linux, Apache and Eclipse are great examples of individuals and competing organization coming together to collaborate on the development of technology.

IT organizations are always looking for new ways to drive efficiencies and speed innovation in their organization. Henry Chesbrough has coined the term ‘Innovation Networks’ to discuss R&D departments that treat their R&D systems as an open system that include internal and external parties. It would seem to me that OSS provides a great model for creating software innovation networks for IT organizations.

What are the logical areas of collaboration?

Open source software implementing open standards is a well established. Apache implementing the Java JSRs and http; JBoss implementing EJB specs, Eclipse Modeling Project and the OMG, etc. Vertical industries also have their own industry specific standard, such as in Finance FpML, FIX Protocol, etc., in Healthcare HL7, or Automotive STAR standards. If an IT organization needs to implement a vertical standard, why not collaborate with others in the industry to implement it once. This would certainly drive cost efficiencies, quicker adoption and most likely better interoperability of the standard between the different consumers.

Creating a common software platform that supports an industry supply chain is another area that seems ripe for collaboration. A common industry software platform that allows many different players to participate and add value can be an engine for innovation in the entire industry sector. Eclipse is a great example in the software tools industry. I think there are other compelling examples in vertical industries.

Factors to consider when starting industry projects

Starting an open source project is non-trivial, especially if there are different organizations and even competing organizations involved. I actually think the Eclipse Foundation has established some best practices that can benefit any industry group that was interested in starting an project. Here are some factors that I think need to be considered to establish successful software innovation networks.

1. How will the IP be shared and managed amongst the participants? OS licenses and contribution agreements need to be put in place to ensure effective IP sharing.

2. What is the governance model that drives the policies and strategies of the group to ensure equal opportunity for collaboration?

3. There are many different styles of open source development. An open development process that stipulates transparency and openness to all potential participants is important.

4. Establishing and supporting the IT infrastructure, like the project web site, code repository, mailing list, etc.

5. How will promote and build the community? If the goal is to have many contributors and collaborators, then community building is an important factor to consider.

Will this ever work?

I must admit that I have been saying vertical frameworks are the next big thing for at least two years. Some progress is being made but we still have a long way to go. In my discussions, it seems the most IT organizations are rightly focused on their business requirements. However, a few are seeing the practices learned in open source could very well be applicable to their industry.

Most IT organizations have figured out how to be consumers of open source software. To drive the next level of efficiencies and innovation they will need to become active contributors to open source software. I believe this will happen, I think Eclipse is well suited to help make it happen and if it does it will be the next big thing in OSS.

 

 

 

 

 

 

 

 

 

A story of migrating a Java application to Equinox and OSGi

Bill Kayser has written an interesting summary of his experience migrating a 600,000 line multi-tier swing based application to run on Equinox. He begins the story with a very good explanation of OSGi as the platform for applications with a service oriented architecture.

It is interesting to see the benefits Bill achieved as the result of migrating to Equinox.

The improvements included but were not limited to:

• Going from 25,000 lines of ant code required for a full build down to about 200 lines of boilerplate configuration, plus about 200 lines of custom callbacks,
• We eliminated about seventy class file catalogs used to ensure extra classes were not inadvertently shipped in the wrong jar file or duplicated unnecessarily,
• Reducing the size of application distros by eliminating unused dependencies
• Surfacing previously unknown bugs based on dangling references to missing classes and libraries,
• Eliminating a large body of code devoted to managing extensions with segregated class spaces using custom class loaders.
• Going from managing four different runtime configurations for each application–the IDE classpaths, the IDE launchers, the runtime script classpaths, the build script classpaths–down to a single feature descriptor listing the OSGi bundles comprised by each application.

And probably the biggest improvement was in the introduction of the extension mechanism provided in Equinox to extend applications with a sort of dependency injection. This allowed us a new lease on life as a platform for custom configuration and specialized implementations instead of a product trying to be all things to all customers.

Well worth the read!

Equinox and OSGi Video Series

Michael Cote from Redmonk has been busily producing a series of videos about Equinox, OSGi and the different Eclipse projects that are building on top of Equinox. If you are new to Equinox and OSGi, this series should provide a great introduction.

Video 1 - Jeff McAffer introduces Equinox and Component Oriented Development

Video 2 - Jochen Krause demonstrates how Eclipse RAP can be used to build AJAX applications based on Equinox bundles.

Video 3 - Ricco Deutscher from the Swordfish project provides an interesting overview of how the Swordfish project plans to implement a SOA Runtime that support JBI and SCA running on Equinox.

Video 4 - In this video, Cote talks with Andy Smith from the Eclipse SODA project, a component of the Eclipse OHF project. SODA is a great example of using Equinox to implement dynamic deployment and management of remote devices.

Video 5 - Is an interview with Eric Newcomer, chair of the OSGi Alliance Enterprise Expert Working Group. The work of this OSGi Alliance working group is important for the future adoption of Equinox and OSGi in the enterprise. Therefore, it is interesting to hear Eric’s vision and opinions on this topic.