Wednesday, May 4, 2011

Oracle Proposes Hudson Becoming an Eclipse Foundation Project

In the recent highly opinionated post ("Ellison's treatment of open source since the Sun acquisition has been nothing short of reprehensible, and it made me really mad.") Oracle's fight with open source, the good guys won -- this time, Bill Snyder wrote that the Hudson/Jenkins fork is another example of "tension between [Oracle] and the open source community."

Today's major announcement, of course, is that Oracle is proposing transfer of Hudson to the Eclipse Foundation. There are several observations and questions that naturally arise from this announcement.

One of the things I first wondered when reading about this was if Hudson and Jenkins might merge back together. Paul Krill's article Oracle hands Hudson to Eclipse, but Jenkins fork seems permanent seems to answer, "Probably not," and backs that answer up with quotes from pertinent individuals.

If Hudson and Jenkins are not to reunite, then which one is more likely to enjoy long-term success? Hudson has name recognition, potential Eclipse community support, and major vendors behind it. Jenkins enjoys the support of the majority of the committers to Hudson, including its creator. Jenkins appeals to community members who want to support (or just use) freely available open source products, but Hudson will soon have the same advantage if it is part of the Eclipse Foundation. It sounds like Hudson will have an Apache 2 or Eclipse license, both of which are "friendly" to corporate users.

If Hudson and Jenkins continue as separate projects, there are advantages and disadvantages for the rest of us. The obvious disadvantage is the loss of additional testing and community involvement for either project due to the existence of the other. The obvious advantage is that competition is often good for the consumer and should help drive innovation. We have seen this effect between Java and C# and between the Eclipse IDE and NetBeans. Although the communities may be split, they can still borrow from each others' ideas.

One observation/question I have is if Oracle's handling of OpenOffice and Hudson are indicative of a pattern. In both cases, Oracle proposed donation of the product to an independent open source repository/provider after a fork occurred. It was LibreOffice in the case of OpenOffice and Jenkins in the case of Hudson.

The final question that won't be answered for some time to come is the long-term success and viability of these donated products and their forked alternatives. My belief is that both Hudson and Jenkins are likely to enjoy some success (though each product's success will be less than it would be if they were one), but I'm less certain about the future of OpenOffice/LibreOffice.