In July 1998, the Frenchman GaÃ«l Duval released his new GNU/Linux distribution called Mandrake-Linux. It was a fork of Red Hat using the KDE desktop – something that Red Hat itself was unwilling to provide because at that time the underlying Qt libraries were not open source. In 2005, the company set up to develop Mandrake-Linux further, MandrakeSoft, purchased the Brazilian open source company Conectiva, and the resulting distribution of the combined forces was re-named Mandriva. And now Mandriva is returning to its roots as a fork by being forked, as a new distro called Mageia:
Forking an existing open source project is never an easy decision to make, and forking Mandriva Linux is a huge task.
It was not an impulsive decision. We all spoke a lot before: former employees, Cooker contributors and users' communities. We collected opinions and reactions in the past weeks as we needed to get some kind of global agreement and to gather, before going ahead.
We believe a fork is the best solution and we have decided to create a new distribution: Mageia.
Here's the background to a step not undertaken lightly:
Many things have happened in the past 12 years. Some were very nice: the Mandriva Linux community is quite large, motivated and experienced, the distribution remains one of the most popular and an award-winning product, easy to use and innovative. Some other events did have some really bad consequences that made people not so confident in the viability of their favourite distribution.
People working on it just do not want to be dependent on the economic fluctuations and erratic, unexplained strategic moves of the company.
And then there is this other recent fork, OpenIndiana:
We are proud to announce OpenIndiana, an exciting new distribution of OpenSolaris, built by the community, for the community – available for immediate download! OpenIndiana is a continuation of the OpenSolaris legacy and aims to be binary and package compatible with Oracle Solaris 11 and Solaris 11 Express. Our first build is from our development branch suitable for testing. We are also working on a stable branch, will receive regular security and bug updates making it suitable for use in production completely free of charge.
Nor is this the only casualty of Oracle's takeover of Sun: another product that is no longer being supported by the company is OpenSSO, which has now been forked as OpenAM:
ForgeRock today announced general availability of OpenAM 9.5 software, the latest release of the OpenAM access management product, part of the I3 Platform. This represents the first full release of OpenAM since ForgeRock commenced sponsorship of the open source OpenAM project and provides a smooth migration option for Sun OpenSSO Enterprise 8 users. The latest version can be obtained from http://www.forgerock.com/downloads.html
"This is an important milestone for the OpenAM community," said Simon Phipps, chief strategy officer at ForgeRock. "This achievement marks the first fully community-sourced release of OpenAM. We're very pleased that users of OpenSSO Enterprise 8 can easily and freely migrate to OpenAM 9.5 now that the updates have been made."
So what's going on – why the sudden flood of forks? I think this indicates that we are entering a new phase in open source, and that the multi-year honeymoon for companies seeking to make money from free software is over.
Of course, making money from free software is perfectly legitimate, as Richard Stallman has emphasised many times. But that does not mean that such companies do not have responsibilities towards the coders and communities that support them. Moreover, simply abandoning software projects because they no longer fit into the latest flavour of corporate strategy is not a good way to win friends in the open source world. If such projects don't fit with that strategy, the solution is to help them to become independent, not simply to chuck them away like old boots.
There are two reasons why this is a wise course of action. First, it sends the right signal to the open source community – including those who might be engaged on other projects that are currently supported by the company in question. Oracle's high-handed approach to open source is fast making it Public Enemy Number 1 as far as free software is concerned (yes, even relegating Microsoft to second place). This means that people working on the MySQL or OpenOffice.org projects are going to be far warier, and more distrustful of the company's moves in future.
The second reason why those who seek to make money from free software would do well to treat their projects and their supporters with more respect is this latest wave of forks: people are beginning to realise that they don't have to put up with this stuff. Forking is certainly a big step to take, but the more people that decide to do it, the more acceptable it will become, despite its well-known risks.
Unfortunately, I think it is going to take a while for this message to sink in, and so I predict we are going to see plenty more forks in the near future as the community starts to re-assert itself. I also think that this tendency will lead to more independent foundations being set up to oversee the development of free software, since these provide a fairer and more balanced way for companies and coders to work together for mutual benefit. Interesting times...