Last week I blogged about the features of the upcoming LibreOffice 4.1. I tried to explain the way we work and why LibreOffice should not be seen as a product, but as a community.
The post had some good success and a few comments as well. One of these attracted my attention. The poster “jsc” – Jürgen Schmidt if I’m not mistaken – is obviously an Apache contributor and an IBM engineer who in a previous life was also a long time StarOffice/Sun employee . I’m grateful for his comment as he’s tried to present the work on the sidebar from his perspective and that’s of course always interesting to understand his points.
To start with, let me stress that while I do recall the issues with the IAccessibility2 Framework, the sidebar was already part of the code coming from IBM at that time . The side bar was shipped with Lotus Symphony which was (which is) the fusion of the former Openoffice.org code base with the Lotus framework. IBM didn’t just donate the accessibility framework; at that time the story was that was IBM was essentially committed to contribute to OpenOffice.org regardless of how it translated in reality. Politically it was true, in terms of code it didn’t go well and I won’t be pointing fingers at anyone in that story.
The comment then goes on to explain how big of an integration work it is and points out that while LibreOffice works on its own integration of the sidebar in the code, it is copying the code from IBM and that all this is wasted time. I think we all agree that this integration work is quite important and of course we must appreciate that; but calling what the LibreOffice project does a “copy” seems wrong to me at two levels. First, LibreOffice is not a project that spends its time copying IBM ideas and features: LibreOffice integrates new features coming from people joining its community everyday and is busy scaling up its own processes and infrastructure in order to accomodate the flow of new code and volunteers. Contributors to the LibreOffice projects take good ideas wherever they come from, even when the come from IBM (and why wouldn’t they?) and at this stage it seems important to remind everyone that not only does the Document Foundation have the greatest respect for IBM but we would also love to work more closely with them. Second, I’d like to remind a very simple truth even at the risk of breaking into open doors: this is Free and Open Source Software. IBM releases code -in this case under the Apache Licence- and the code is made available to anyone who wants to use it, modify it and distribute it. To complain that this or that Free and Open Source Software project is copying a feature amounts to deny the Free and Open Source nature of the code and the whole release process surrounding it.
In this regard, the fact that IBM might have intended to have the Lotus side bar integrated in AOO first makes no difference: you cannot release Free and Open Source licensed software in public and then channel the code to only one implementation, regardless of the license you use. Then comes the argument that it’s all wasted time. It seems to imply that there’s some useless duplication of work on the sidebar done on at least two sides: the LibreOffice project and Apache. I don’t deny duplication, but the waste of time is an interesting comment. When the LibreOffice project was born, its founders explained that their goal was not so much to secede from the OpenOffice.org project but rather to propulse the existing community and project into the “next decade” (cf the text that was called the “Next Decade Manifesto”). With the decision to start the Apache OpenOffice project afresh the situation changed and all of a sudden there were two separate projects. This should not be labeled duplication; the two projects support two different licensing models, different ways of working, different visions of where they could go while sharing a common codebase for the first months of existence . When it comes to Free and OpenSource Software, there seems to be duplication everywhere: over a hundred Linux distribution, at least four front-ends to the bluetooth stack for Linux desktops, two native GTK web browsers running the webkit engine, five Twitter clients… and I could go on with an entire list. Somehow it just works, probably because this duplication is the expression of individual and collective freedoms. The failure or the success of each project falls ultimately on the shoulders of its contributors, and by contributors I mean its community. I guess what it means is that the success of a project is defined by a few complex factors that can then be brought back to whether the community is solid and growing. Code development, quality assurance, and code “duplication” is thus not wasted time provided that the contributors do it for a reason and are happy to do it: it’s how community and ultimately Free/Open Source Software works.
Last but not least, the comment ends up with a reminder that the hosting of OpenOffice at Apache comes with important costs and that donations there would be welcome. I surely sympathize with that: projects of this size come with with quite an infrastructure to run, maintain and ultimately pay for. It would be a pity to see the Apache Office project go because of its operational costs running too high, but ultimately I think the Apache Software Foundation has been running its own infrastructure for years in a very effective way and therefore these costs might be subsidized in the end.