Hudson’s move to the Eclipse Foundation marks the beginning of a new, exciting chapter of possibilities for the open-source continuous integration server project and its community of users. While some question the sincerity and motivation behind Oracle’s decision to move Hudson to the Eclipse Foundation, and others react to perceived poor communication or licensing issues, the majority has been positive and encouraging, a sentiment I hope will continue to resonate and multiply within the Java developer community and bring us closer to reconciliation.
The Eclipse Foundation is an ideal home and should be very attractive for those interested in reuniting the community, mending the fork and advancing open-source software innovation. As SD Times’ David Rubinstein pointed out in his blog, “Eclipse has a great record of building community around open-source projects (rapid releases) that are then productized (stability), and both sides of the Hudson/Jenkins fork should be satisfied.”
I couldn’t agree more. At Sonatype, our perspective regarding the Hudson project has always been a simple one: We have been and will continue to be committed to advancing Hudson and making better software available to the community of Hudson users.
As a Strategic Member at Eclipse, we know first-hand that the Foundation’s infrastructure, processes and oversight are excellent. Eclipse is the place for a popular project like Hudson to mature, and now is the time for the Hudson and Jenkins communities to reunite. More resources than ever will be poured into Hudson. Oracle will continue to be a leading contributor with several full-time engineers on the project. Oracle and Sonatype have been working diligently to add fundamental architectural improvements to Hudson to improve stability and performance, and help spur continued innovation.
Early interest and commitment of resources from other industry-leading companies like VMware and Tasktop illustrate the industry’s interest in moving the project to a vendor-neutral ground, and the importance of Hudson as a key foundation for agile software development. In fact, a recent survey of more than 1,600 Java developers found that 80% of those surveyed have adopted CI and 75% of those CI users are using or experimenting with Hudson.
The Eclipse community has a long, established history of supporting a wide variety of programming languages and platforms. This experience, support and flow of contributions will help advance Hudson and appeal to organizations using Python, Ruby, C/C++ and others, not just Java. In fact, as the Eclipse Platform will be built using Hudson, it will need to support the Eclipse SWT project, which is a very important segment of native code, as well as the CDT project, which deals with development tooling for C/C++ projects. Hudson will also benefit from the Ruby knowledge and support Sonatype brings to the project, as it’s used internally to manage the Maven central repository.
At the end of the day, the best way for the Hudson project to attract the external community and broad-based usage support is by having a highly stable platform that plug-in authors can rely on. To this end, Sonatype has elected to open-source both its Maven 3.x and Eclipse integration for Hudson, and donate them to the project as a sign of support for the advancement of Hudson. We hope that by giving these valuable commercial features back to the community, we can help make the project even stronger, motivate plug-in developers to work with the Hudson platform, and continue to grab the attention of enterprise-class users.
It’s a new day and an excellent time to put egos, and the past, aside, and work together to build a community-led project that successfully combines the interests and contributions of both corporate-sponsored and independent developer-driven work. We sincerely hope that the Jenkins community and its leadership will consider rejoining the project at Eclipse.
Jason van Zyl is CTO and founder of Sonatype, and the creator of Apache Maven, the Plexus IoC framework and the Apache Velocity projects.