fosdem.org

Michael Meeks will give a talk about "Liberating Open Office Development " at FOSDEM 2011.

Could you briefly introduce yourself?

I'm Michael Meeks, Christian, Hacker, Husband, Irritation - that sort of thing. I enjoy working for Novell who pay me to make great Free Software even better for our customers - currently I'm working on LibreOffice most of the time, though I've enjoyed working on a string of pieces of other projects over the years: principally MeeGo and GNOME.

What will your talk be about, exactly?

Lots of things about LibreOffice - briefly about why and how it exists I suppose, and about how we are trying to make LibreOffice a really well run, open, vendor-neutral, true Free Software project. Of course, I want to give people an update on where we have got so far - the huge progress we have made, and what they can expect next.

Mostly, I'd like people to understand more about the LibreOffice codebase, its scale, and features, its design and problems, so that if they want to start help out it is easy for them to do so. There are lots of great opportunities for people to get involved, and more to make a real impact in the project, even with little prior experience, I'd like to highlight some of them.

Finally, I suppose I want to try to explain to people the emerging threat to Free Software projects that comes from companies aggregating copyright, and encourage them that this is something to avoid.

What do you hope to accomplish by giving this talk? What do you expect?

Well, first I want to inspire people to get involved; we have some of the best developers at FOSDEM and I'd love to get some of their help with LibreOffice. Perhaps more important I want to give those who perhaps are looking for a home in Free Software somewhere to make a significant blow for freedom. LibreOffice, with its substantial user-base, and relevant feature-set is perhaps the best project to get involved with in this regard.

Secondly - I'd like to persuade people to think carefully about more than just the license of a corporately sponsored Open Source project - and to dig deeper into the governance and ownership models to make sure they are sound - before contributing. Hopefully if I do that, at worst it will avoid unpleasant surprises in future, and at best help other truly open projects flourish.

Thirdly, I hope to entertain and inform - we have an exciting world full of strange, wonderful and delicious things, and we can enjoy that I suppose.

One of the things that LibreOffice does different than OpenOffice.org is with respect to copyright assignment: the former doesn't require it. Can you explain the rationale behind this?

Sure - that is easy. If a single corporation requires copyright assignment, they create a very asymmetric situation whereby (only) they are not bound by the project license. This has the potential to create many problems, quite apart from the obvious corrosion of trust, and deterrence that mailing signed paperwork before having your code included can create.

I think Lennart Poettering explained it most succinctly, talking about systemd (which is a sexy project incidentally):

We value your contribution, and hence do not require copyright assignment.

Is your code only useful if it can be owned by a controlling commercial entity ? I would argue that, ideally, there should be a level playing field for all contributors, with complete reciprocity between them. Luckily, this is how LibreOffice, and the vast majority of successful Free Software projects operate. But come to my talk for some more details.

With LibreOffice, you wanted to lower the barrier for new contributors, e.g. by offering 'easy hacks' on the website and by not requiring formal paperwork such as a copyright assignment. At the end of 2010, the project had a successful start with a lot of new contributors, but did it keep up that pace? How many new contributors and translators did you attract?

I mainly watch two metrics to try to keep us on track (as kindly produced by Cedric Bosdonnat). A breakdown of code contributors each week, and also the overall number of entirely new code contributors to the project since we started. We publish statistics on that from time to time, and hopefully in the future will be able to automate that.

In the first three months - we attracted more than ninety entirely new code contributors, in addition to our existing contributor base. These guys have code committed now having never committed anything before. In addition, we have over fifty translators - that growth rate of around one new code contributor per day, and one new translator every two days is most encouraging. Can we keep that pace up ? Luckily it is more of a linear than exponential growth, so perhaps it is possible, let's see. Of course we still have a huge amount to do, and an insatiable appetite for people to adopt orphaned pieces of code, and parent them into beautiful symphonies of excellence, new languages to reach, better artwork and user interface to design, so there is always a place for new people.

The rate of contribution and growth dipped somewhat at Christmas - it seems people have families that enjoy seeing them, but it has bounced back afterwards. We have around sixty different individuals contributing code each week, two thirds of them entirely new to the project.

What drove that? Well, I really believe that the activating event was moving to a vendor-neutral, open, and reciprocal development model. Once that happened, most of the extremely obvious, best-practice of building developer communities was applied: including actively welcoming new participants, from whom we have had some simply excellent contributions.

Will LibreOffice be able to track OpenOffice.org development? And which changes will you incorporate? Is compatibility with OpenOffice.org important?

Thus far much of the work we have been doing has been code cleanup, integrating existing features, and fixes. Tracking Oracle's development seems quite do-able to me. Our biggest problems in the past were unwinding conflicts caused by merging our own code up-stream into OpenOffice.org and reconciling it with our (usually somewhat newer) versions. That should no longer be an issue.

Clearly we want to incorporate any good code that makes sense and is cleanly written from Oracle - just as we would from any other contributor, and give credit where it is due.

Compatibility with Oracle's OO.o is an interesting transitional concern, at least until LibreOffice rules the world. As an example, for now we want to be backwards compatible with their extensions - so we can only expand the low-level component APIs compatibly. Unfortunately, that ties us to a number of poor design decisions that we now have the manpower to undo, so this will not always be the case.

What will the differences between LibreOffice and OpenOffice.org for end users in the longer term?

That is really hard to say - primarily because there is almost no information about Oracle's plans for OpenOffice.org longer term. My passionate hope is that, in the end, there is no difference. That vision is of the two projects re-uniting as a single truly open, community project in which Oracle is an immensely respected and influential peer contributor. We have made every effort to structure the project and ownership to make that possible, and even desirable for Oracle, and the door is wide open.

Of course, in the meantime - I expect the feature and quality gap to grow incrementally. For developers, LibreOffice will have cleaner, more readable, and higher quality code, with unit tests, incremental re-writing and improvement. I expect the existing feature edge to grow incrementally over time, and our product and process to mature rapidly. Most of all, I anticipate LibreOffice becoming increasingly fun and friendly, as more talented people join the project, and are nurtured into effectiveness. Exciting times.

Have you enjoyed previous FOSDEM editions?

Emphatically yes ! FOSDEM is an outstanding conference - and more than that a strategic meeting place for the best European Free software developers. If I could go to only one European conference per year it would be FOSDEM.

Each year, I see old friends, make new friends, and talk until late into the night / morning. It is a fantastic place to convert IRC nicks into names, to add a voice and depth of character to future e-mail collaboration, and to deepen your friendships and passion around Free Software. It is also a great place for people to get started and involved with Free Software, I'd love people to come and meet up with the teams I care about: LibreOffice, openSUSE, GNOME, RepRap, whatever takes your fancy - learn, and become more effective as we change the world together.

This interview is licensed under a Creative Commons Attribution 2.0 Belgium License.