The world's lamest brand
As a reaction to this phenomenon, Meeks and Novell created Go-oo.org, of which Meeks says: "The primary function of go-oo is to act as the world's lamest brand. It's really up there in the terrible brand category, but it turned out to be a good place to host all the things that weren't being done by Sun. We initially added tools and features such as search and control of the source code, so we could actually find code that does what we want within this vast and sprawling code base. We added code to unify all the Linux distributions which were heavily patching OO.o to try and get it to build on their distributions. And that's where we are today. Lots of distributions, Debian and Red Hat, Ubuntu and Mandriva, take their release versions of OO.o from go.oo, and package them for their distributions, and there are features you don't really get in upstream OpenOffice, and some features the developers or Novell don't want to assign to Sun."
At the same time IBM have released a new version of Symphony, which is based on OpenOffice, but has a proprietary front end. Earlier versions of Symphony were based on an early release of OO.o which carried an obscure SISSL license. Meeks says "I commend Symphony to you. It's an unfortunate fork in some ways of OpenOffice, but the new version starts faster than it did, looks better, and is based on a much more recent OO.o code base. We're working closely with IBM on a number of features. It's a really good product, but therein lies the rub."
"In some places they do feed stuff back. We see their changes, but parts of Symphony are not open source, and we don't have the code for them, and interestingly, there is no source code available so far as I am aware of the version of OO.o that IBM is shipping inside their product, so clearly they're not shipping this under the LGPLv3. IBM have a fairly public antipathy towards the GPL unfortunately, and as a consequence you have to wonder what terms are they shipping OpenOffice under - and as there is a lot of my code in there, not only my code but Novell's code and a lot of other people's code, you have to wonder 'What were the terms and what was the deal? That's a shame, and would really help improve the transparency and confidence in Sun's stewardship around these things. The code was assigned to Sun, and I have no doubt there is no legal problem at all, but a lot of people have assigned their code to Sun in good faith, believing them to be good stewards. Maybe they are but its impossible to tell without knowing the terms under which third parties are shipping the code."
Up and down
Meeks blames much of the lack of developer interaction with OO.o on the bureaucratic rigidity imposed on the project by Sun, and consequent complaints about the slow rate of progress and slow turnaround of bug fixes, occasioned by Sun's fixation with quality control. Despite this, he has "a lot of sympathy with Sun's position. If you reduce the number of changes going into a piece of code in theory you can improve the quality, no doubt about it. But if you want an incredibly high quality piece of software there are several theories about how you get there."
"The problem is we're not starting from square one. OO.o has quality issues, maybe fewer than we did have, which is encouraging, but they're certainly there. And in consequence if you slow down changes you may improve quality, but from what base? Interestingly, the Linux kernel quality metric is improving and yet the rate of change is accelerating, so there are other ways to work. Maybe you have to go down hill a bit before you can find a better way up, so its possible if we relax QA just a bit, we make things worse, but if we increase the number of commits very radically we move to a different place in the graph, and bugs get traced and fixed much more quickly."
Meeks relates the problems of OO.o to other companies who have struggled with the concept of free software development. If you own and over-control the development process you lose the advantages of free and open source software. Organisations participate in free software projects for selfish reasons, because it works, and because it brings twice the resources at half the price. But they are most likely to succeed and satisfy the more selfish aims of the organisation if projects are overseen with a light touch and the developers are allowed the freedom to use their initiative.
"I'm optimistic that when Oracle complete their acquisition of the German piece of Sun there will be some more helpful input there, because what we ask is relatively simple and has no impact on their business. We don't want to hurt their business. Quite the opposite. We want to grow it, and grow ours at the same time."
Michael Meeks worked with Miguel de Icaza on the GNOME project, eventually joining Ximian, the company founded by Icaza and Nat Friedman, full-time. Ximian was acquired by Novell in 2003 and Meeks is now a Distinguished Engineer at Novell where he works on OpenOffice.org development.
For other feature articles by Richard Hillesley, please see the archive.