November 03, 2010, 12:30 PM —
This past weekend, while the US was ramping up for yet-another contentious election, politics of another sort were happening in the land of open source software.
Specifically, 33 members of the Germanophone project within the larger OpenOffice.org community gave Oracle and the OpenOffice.org team leaders notice that they would be walking away from the project and working for the new LibreOffice project, now being managed by The Document Foundation.
The reasons for the walkout are clearly stated: the developers are unhappy with the OpenOffice.org stance that any current OpenOffice.org project or team leader who is also working on LibreOffice should withdraw from their position in the OpenOffice community.
Oracle's stated reason for this requirement is their belief that such team leaders working on both projects represents a conflict of interest. LibreOffice developers, they feel, shouldn't be leading OpenOffice.org teams.
At first blush, this may seem like the Oracle/OpenOffice.org team are just a bunch of snobs who don't want LibreOffice developers taking the OpenOffice.org code off in some direction they [Oracle] doesn't want to go. There is some validity to that presumption, though knowing several of the key members of the OpenOffice.org team, I hesitate to actually call them snobs. They are really quite decent people, so instead, let's call them pragmatists.
The thought has occurred to me, though, that this may be about much more than Oracle not sharing its community. In fact, given the past history of how Oracle treats open source projects in general, I believe any strengthening of the LibreOffice community, whether through new developers or from developers migrating from OpenOffice.org, will ultimately benefit OpenOffice.org far more than a weaker LibreOffice.
An outcome, by the way, I believe Oracle planned all along.
Here's my thought process: Oracle is trying to keep OpenOffice.org going, but only in a certain direction. No one if quite sure what that direction is right now, but I think it's fair to assume there is a definite plan.
LibreOffice, for whatever reason, does not fit in that plan. Or Oracle is worried that LibreOffice is in sync now, but won't be later.
The solution? Invite LibreOffice contributors with OpenOffice.org leadership roles to choose between the two projects, which is exactly what happened last month. This accomplishes two things: those who leave can be replaced by more Oracle-friendly developers. Those who stay will have re-committed themselves to the success of OpenOffice.org, just by staying.
Ah, but there's another benefit here left unsaid: the people who leave--as well as any protesting developers who follow them--will very likely go work on LibreOffice. They will build LibreOffice to be a better, stronger office suite, using better, stronger code.
Code that OpenOffice.org will be more than able to cherry pick right back into their project.
The outcome of the LibreOffice fork is pretty much on a course to set up the same conditions Oracle has tried to use to its advantage with Red Hat Enterprise Linux. Initially Oracle and Red Hat were great partners, until Larry Ellison figured out that instead of partnering with Red Hat, they could repackage RHEL as Unbreakable Linux and get all of the service and support revenue they used to share with Red Hat for themselves.
Indeed, Unbreakable Linux was almost an exact clone of RHEL when it started, though there are some unique features within it now. But the bulk of the product is still developed and maintained by Red Hat.
Now, apply this to LibreOffice/OpenOffice.org. Right now, OpenOffice.org is the "canonical" suite between the two. But what if Oracle wants that to change? Let LibreOffice be "The" open source office application, they may be thinking, and let most of the developers work on that project instead. Heck, Oracle can even treat them a bit shabbily to hurry the walk-outs along.
Then you would have a scenario where LibreOffice was the upstream canonical source, not OpenOffice.org. And, as with RHEL and Unbreakable Linux, Oracle would be able to pick and choose all of the new LibreOffice innovations for its own office suite and be free of any poltical entanglements that would prevent them from adding the features Oracle wants to see in OpenOffice.org.
Granted, this means they would be "giving up" the upstream spot, but really, what would they lose? The overhead of duplicating the functions of all of those community members, for one, as well as the potential headaches of dealing with a recalitrant community.
Put in this light, it becomes a bit clearer why Oracle is insisting it doesn't need the community to keep OpenOffice.org going. They geniunely believe they don't. Like Oracle Unbreakable Linux, all OpenOffice.org needs to survive is the LibreOffice codebase and a much easier-to-manage, smaller in-house staff, in their opinion.
This is the same plan they used with RHEL, only now they're letting an upstream provider be created first, instead of using a preexisting provider.
All very pragmatic, isn't it?