1 2 3 next »

OpenOffice at the crossroads

A conversation with Michael Meeks

by Richard Hillesley

OpenOffice.org is a flagship for free and open source software, released under free software licenses and achieving downloads in the hundreds of millions. OO.o is a success by most measurements, but there have long been murmurings of discontent among developers resulting in complaints of "non-responsiveness and lack of leadership" on the project. The argument is not that the project is a failure, but that OpenOffice.org could be so much more, given a less top down approach to project management and a looser rein on developers' ability to get involved.

The code of OpenOffice.org is released under free software licenses but the copyright for all internal and third party contributions are assigned to Oracle/Sun, and the OpenOffice development team within Oracle/Sun dictates the rate of progress.

The most vocal critic of the process has been Novell employee and long time GNOME and OO.o developer, Michael Meeks. Meeks argues that copyright assignment discourages external contributions, and that over zealous control of the project inhibits developer initiative.

Meeks has long contended that OpenOffice.org has failed to attract and keep individual and corporate developers due to "a half-hearted open-source strategy that is not truly 'Open'" and lacks transparency. This has inhibited the potential of OO.o to be "even greater" than it already is.

The question why

"In a healthy project we would expect to see a large number of volunteer developers involved," Meeks has written. "In addition - we would expect to see a large number of peer companies contributing to the common code pool; we do not see this in OpenOffice.org. Indeed, quite the opposite we appear to have the lowest number of active developers on OO.o since records began."

Sun's grip on the development process has inhibited the creative hubbub common to most open source projects, and developers are reluctant to contribute because the incentives haven't been there. The virtue of successful open source projects is supposed to be that they are developer driven - open, democratic, noisy, argumentative, divisive, chaotic and productive. And they work.

So for Meeks the issues are simple. OO.o needs revision of the ownership and licensing structures. "The better place to be is just to get people to work together, so we can focus on the really interesting technical arguments, rather than squabbling over the basics."

This promises to be an interesting topic for Oracle, for as Meeks points out, "I use the name Sun advisedly. The organisation there still has not been merged into Oracle, so it remains to be seen what Oracle will do when that happens."

Come together

The underlying problem for Meeks and other developers is the slowness of development and fixing of bugs in OO.o, and the need to encourage Oracle / Sun to stimulate the responsiveness, innovation and speed of development that are the characteristics of other successful free and open source software projects. Some participants in OO.o development have pushed for the creation of an independent foundation for OpenOffice, on the model of Mozilla, KDE or Gnome, but Meeks is ambivalent about the advantages of such an approach.

"An independent foundation leads you to several questions," he says. "You can go for a Mozilla-like Foundation. They do a lot of good work, but there's an unfortunate instability in having a single large organisation doing all the work, and there are funding issues. Firefox has a fairly obvious funding model, but that wouldn't be true of an OpenOffice Foundation, so you wind up asking companies to chip in to pay the developers. It's an inelegant solution to the problem. It seems to me far better for each company to employ developers that work on the project in the common good, as is done with Linux, GNOME or KDE."

"Independence is good, but getting the asset out of Oracle would be very difficult. I personally would just settle for people working together, reasonably, and making some compromises that suggest it's worth having people work together."

"The ideal would be to get a little less heavy handed control form up-stream, and a more enlightened approach to working together. There's no business reason for separation and there are some very strong business reasons for working together, obviously - in terms of improving the product, of moving faster, in terms of not paying people to duplicate features that have already been written under some liberal license. Its absolutely nonsensical to have people in two companies writing essentially the same stuff. Words fail me to express how pointless it is."

"The ideal everyone wants is to work together. The question is why aren't they?"

Next: Every bug is a feature

1 2 3 next »