From the outside, the world of open source software can appear turbulent, disorganised and disparate. But when it comes to the creation of software, these consequences work to its advantage.
Open source development is a democracy. These attributes make Linux effective, interesting, diverting and productive. You can do what you want with the source code behind the software. You can add features, remove broken ones, fix problems, bolt on attachments and even copy an entire project verbatim.
Article continues below
You can sell the software, use it as part of a commercial endeavour, rebrand it, copy it, redistribute it and charge to support it. You can make it your own.
The set of free software licences used by most open source software guarantees these freedoms, and for most of them, your only obligation is to share your changes to the source code. It's still the best reason for using free software, and it's something none of its competitors can touch.
This is why there's so much choice and diversity, and it's from this chaos of unstructured development that order eventually coalesces, forming some of the major open source projects we all love, from OpenOffice.org and GIMP to Fedora and Ubuntu.
It's a competitive environment, where the low cost of software means features and usability are what bring a project to dominance, rather than a great piece of marketing or an influential sponsor (even if both of these are having an increasing effect). In a world that's become used to six-month release cycles and free updates, it's easy to forget that the random fluctuations of free software development are never far from the surface, and that's definitely a good thing.
This fluctuation is the best evidence we've got that the developers still care, and that apathy isn't taking a project to the retirement village of an inactive SourceForge account. Dramatic changes still happen, and two influential free software projects have recently undergone the ultimate metamorphosis – they've been forked.
Forks are cataclysmic
A fork is the complete duplication of a project from one host to another, usually by a group of developers and users who have become so dispossessed and disenfranchised by the original project that they have no other option.
Outwardly, it's a sign that a project's leaders aren't listening to its contributors. Inwardly, it's a sign that nothing is changing and that the project may have lost its purpose.
A fork is a cataclysmic event. It's never taken lightly, and usually only as a last resort. It's happened many times in the past, most notably with XFree86, the X Window system used by the vast majority of Linux distributions and many other free operating systems.
X is one of the most important components in Linux. It handles screen updates and the way the user interacts with the system. The old guard at XFree86 had become unpopular – in a world where other GUIs were rapidly gaining functionality, such as anti-aliased fonts, shadows and compositing, X was going nowhere with XFree86. It was forked and became X.org.
At the time, many of us thought this was a mistake. Why risk a vital OS component with an untested new group? But the results were staggering.
X.org implemented almost all the features we'd wanted for years, and went from strength to strength. Mandriva, the Linux distribution, has become Mageia. Some of its core developers and community members have taken the brave decision to go it alone.
Like XFree86, it's a fork that's come after many years of apathy, with too little innovation and financial reward. A fork in this case has nothing to lose, because the alternative is nothing. A fork for those people who still care means Mandriva's survival, and that's exactly what can't happen in the world of proprietary software.
Even more dramatic is the latest fork of OpenOffice.org, the essential office suite that gives many people the only real alternative to Microsoft Office, and is now owned by Oracle. Under its previous owner, Sun Microsytems, change happened too slowly. There was a particularly draconian grip on what changes were allowed, and even the copyright over those changes.
Things haven't improved quickly enough at Oracle either, and this has given the community enough momentum to create the LibreOffice fork. LibreOffice and Mageia are important parts of the continued success of free software, but what's more important – and quite unlike the proprietary world – is that free software vendors need reminding that they're stewards for people's software and ideas. And Linux is the best possible ark for bringing those ideas home.