A few minutes ago, Oracle has announced that Openoffice.org, the ODF-based Free Software suite for office productivity, will become a community developed project. In plain English, no more dual licensing, no more proprietary version, go ahead to incoming patches. Woot!
I am very happy to hear about this move, which was not entirely unexpected (by me, at least). To tell all the truth, reaching this point was my secret plan when I have started helping Oracle in the merger control procedure opened by the European Commission last year, where the acquisition of Sun was under scrutiny. I was telling everybody that the dual licensing approach was going to die, that id did not make much sense anymore, that it was "moot" – I actually mentioned MySQL there, but the same applies to Openoffice.org, actually. As it turns out, I was right.
The announcement is clear to reaffirm the pledge by Oracle to support open standards and to work hard on various important pieces of the Free Software portfolio it stewards – which includes also Java, in addition to the already mentioned MySQL, as well as a number of smaller projects. At the same time it is not entirely clear what in particular the new strategy implies in practical terms. My not-entirely-uneducated guess is that the company does not want to commit to one path or the other because there are some factors that can weigh in. One of them can be the presence of some industrial partner which might want to give its patronage to the project and steer it into another kind industry-lead project (but the obvious candidate seems unlikely to jump on this). Another scenario is the one in which a more grassroots approach is taken, and therefore the project lands in the hands of a foundation, a committee or something similar.
The elephant in the room at that point is obvious. There is a foundation that already provides a stewardship to a codebase which is almost overlapping with that of Openoffice.org, and which, albeit being a fork, can still be easily merged into the originating branch: Libreoffice by The Document Foundation. Will the two converge? Hard to say. Will there be some degree of code exchange? That seems the most natural path of evolution. Because the dual licensing encumbrance is soon to die – never too early – there is nothing preventing it, apart from historical frictions between the recently "divorced" developers groups.
So,what's the implication of this move by Oracle, will it make any difference? My answer is "yes". It is highly symbolic. It is a step I have been advocating with them for ages. It reduces friction and removes inefficiencies that the need to have copyright assignment implies.
Or, to make a long story short, good news!