State of the Project, Year 6

-Louis Suárez-Potts

Originally presented 12 September 2006 at OOoCon 2006

Well over 650 people have registered for this OOoCon, the greatest number yet and and an indication both of the increasing importance of OpenOffice.org and of course of the excellent work of the French group,which I would like to thank: for hosting and managing a great conference. Thank you, Sophie, Cedric, Isabelle, Marie, Charles, Jean-François and every one else who contributed and who is continuing to contribute.

I'd also like to thank the INSA staff for their patience and hospitality.

This is an extraordinary conference and not just because of those here who have come from every curve of the globe. The event marks a year in which the OpenOffice.org project long passed 350,000 registered community members and is approaching at least 50 million (and probably twice that) users. Downloads over the last year have testified to the popularity of 2.0 and its upgrades. In an average week, about 250-300,000 people visited the download page and about 100,000 registered their copy of OpenOffice.org. Those numbers do not count all the copies installed through the many Linux distributions, from which, of course, most users obtain their Linux. The data only count page accesses to the download pages. But the numbers are eloquent.

They speak a truth that goes beyond the fact that OpenOffice.org is free: OpenOffice.org is extraordinarily good, and it's open in ways our chief competitor—I don't have to mention its name—cannot quite match. Not only are we open source but we use the Open Document format, or ODF, the first fully certified--by OASIS and the ISO, no less--open standard for office documents. As you all must know, it's even gotten to the point where that company has had to subsidize an open source project to build a "translator" for ODF/Open XML files. For them, the ODF closes doors, or opens doors they'd rather keep closed; for us, it simply opens doors—to government, to be sure--but also to any entity concerned about the preservation of its intellectual property and about the widescale availability of its documents. In the last year, we created a wiki page to track major deployments of OpenOffice.org, and it lists many, chief among them French government offices, but also government and corporate offices in Brazil, Norway, Denmark, Australia, Germany, Italy, India, Spain, Cambodia, Thailand, and others. And that does not even count those offices in Sweden, Canada, the United States, Venezuela, South Korea, Japan, China, and elsewhere who are very seriously considering migrating to the ODF and often OpenOffice.org.

Saying that .doc or .xls or .ppt are de facto industry standards is simply not good enough to bank on and hardly a guarantee. In contrast, our users have the guarantee of permanence. International bodies, not self-interested companies subject to the whims of the marketplace, maintain the file format. The importance of this freedom—and that's what it is—cannot be underemphasized. It also applies, even more so, to bridging the digital divide. Probably every country recognizes the importance of giving its residents the tools to participate in the modern economy. It used to be that learning how to read and write was good enough. Now, one must also be able to create a spreadsheet. Only OpenOffice.org, which, because it is free and will remain so, can uniquely give any user with access to a computer the tools to participate fully and completely in the world's new economy.

It's no surprise then, that when I asked the project leads about the most important accomplishment of the last year, everyone agreed that it was the ODF.

But there were many other points mentioned, too, and they were important. They ranged from technical improvements of interest especially to developers, such as the significant enhancements, according to Andreas Martens, lead of the Word processing project, to the quality of the code base. As he wrote, "root causes" of problems have been fixed and the SW project has "started to modularize the Writer core module". As he explained, "The benefit for users will be improved stability and performance of Writer, the benefit for developers will be easier development and less regression risk. This should lower the barrier to contribute code to our project." A similar point about code improvement was made by Stephan Bergmann of UNO and Mathias Bauer, of Framework who both emphasized the important improvements in code quality that make it easier for developers. The SDK has been further added to, as has the apparatus for extensions—all of which make it easier for developers to work on the code and contribute to the project. (We've always had the capability of accepting extensions, plugins, what have you; it's just easier now.) Last year we heard, clearly and loudly, that a major shortcoming of the project was that it was notoriously difficult for developers to work on the code and get their patches accepted. We heard you. And in the last year there have been profound changes made to the process and code itself. It is now much easier and more efficient for developers to learn how to code for OpenOffice.org, to write code, and to get their code reviewed; and if it is accepted, developers need not wait years; releases with new features are coming out every 6 months.

Without question, Michael Bemmer, the Director of Software Engineering for OpenOffice.org at Sun, has done an extraordinary job this last year. We, the users of OpenOffice.org the product and the members of OpenOffice.org the project, owe him our thanks.

The improvements in code, the advances, ultimately help endusers—like me. Eric Bachard's team, for instance, has done terrific work in advancing OpenOffice.org for the Mac; I use it daily and love it for its speed and capability; so do the 20,000 or so who visit the Mac download page every week, I am sure. And I am especially excited by the glimpses I have seen of it running without the X11 windowing system. I am sure this sentiment of gratitude for having an OpenOffice.org that works for you is shared by not only those in this room who speak any of the nearly 100 languages to which it has been ported but also those using platforms besides the 32-bit standard, including, as the indispensable lead, mentor, developer par excellence and friend to so many, Pavel Janík pointed out, to GNU/Linux on 64-bit processors.

Informing and helping people how to use OpenOffice.org, how to migrate to using it, how to do things with it, is the domain of the Documentation Project, led by Scott Carr and Gerry Singleton and including members of many, many projects. The last year has seen the remarkable work of the Authors group and the creation of documents that cover, in numerous languages, nearly every contingency a user might face. No other open source project is as well provisioned. The Documentation Project also contains templates, and I should underscore that it will be housing a great many more shortly, as well, I hope, macros, clipart, and other things that people need, use, and want.

Alerting people of what the product is, what it does and who can use it is of course the job of Marketing. I think we owe John McCreesh our thanks for doing a brilliant job as marketing lead. The project today hardly resembles that of last year. There have been campaigns—the Why one comes to mind—that have placed OpenOffice.org squarely in the minds of journalists throughout the world, and kept us sexy and interesting. Not a day goes by without a story on OpenOffice.org by some journalist, often enough exploring an angle that John has dreamt up.

The marketing efforts have naturally included more than John's and Cristian Driga's project. They've included the Native Language Confederation, its leads and members. Under the direction and assistance of Charles Schulz, the NLC has not only expanded to complement the l10n project but has taken OpenOffice.org to local communities. OpenOffice.org is used by tens and tens of millions and we owe our thanks to the NLC leads who, working sometimes with government, sometimes with small companies, have helped to make OpenOffice.org a part of the lives of millions of people. Furthermore, the contributions of the NLC go well beyond information and support. They include, increasingly, localization and documentation; and QA.

Neither Andre Schnabel nor Joost Andrae respond to my query on the list;s they were probably too busy. As the leads of the QA Project the work they and their teammates have done has been astonishing. Recall that a short while ago it was very difficult for involved users to file issues, let alone for the QA team to coordinate activity. That is much less the case today. The interface is much better but more importantly, the QA process is now very very good. OpenOffice.org works and the community works together in making sure that what people download is reliable. .

Making that download possible has been the work of our Distribution team. I technically lead that project, but everyone knows that the real lead is Stefan Taxhet; the Coordinator for OpenOffice.org on Sun's side. In the last year, he has made some fairly dramatic changes to how a user downloads OpenOffice.org. Gone are the days of total confusion. We now use a system maintained by the Oregon State University Open Source Lab, that makes downloading OpenOffice.org easy. Complementing that, users—at least thousands per week—can use our expanding P2P network.

The full list of accomplishments is long and wonderful, far longer than this document can track.

The ODF is, as I mentioned, important. But it's not our destiny to be a reference implementation of the ODF; we make the best implementation, but OpenOffice.org is a lot more than its file format. So where does the future lie? It does not lie in splashy releases, like 2.0, but in extensions, new features added progressively, and in making the code easier to work with. If there a 3.0 waiting, it will have to do, I imagine, with a significant re-architecture of the codebase, to make it more open and modular. But getting there is not trivial and will not happen soon, though I am sure it will happen. In the meanwhile, we are working, as you have heard tonight, to making OpenOffice.org processes more transparent and the code better.

How do I see, then, the next year? Let's imagine a fully developed Extensions Project, which Laurent Godard introduced last year and which he leads. We would then have a minimal (non-bloated) installation set a user could initially download. The set would have everything the user needs to get started—the application, some templates, some clipart, dictionary, and so on. But then, let's imagine that after downloading the program tells her of extensions, and on visiting the site, she learns of some that allow her to do a lot more than she could otherwise do. She also learns of the support offerings; of the templates, of macros: she discovers all that makes OpenOffice.org the initial and central node of an expanding universe.

What else do we need to do?

We need to continue with the mentoring and transparency programs, and we need to work more with developers throughout the world. Not everything can be done via the Web or email or IRC; sometimes, in-person meetings are necessary. You wouldn't be here if you didn't believe that. This is especially the case for core developers, who must often travel to Hamburg for direction and mentoring.

So, I'd like to propose we see about creating a regional developer network, where trusted developers in regions throughout the world can help other local developers develop for OpenOffice.org. The ultimate effect would be an expanding network of developers involved in OpenOffice.org. I would ask of our primary stakeholders—Sun, Novell, Intel, Google, Red Hat and others to assist; and I would also ask companies who have the resources to consider helping build such a network of developers.

I also propose that we aggressively push OpenOffice.org (both product and project) to endusers.

This means making it easier for endusers to get OpenOffice.org; it also means advertising OpenOffice.org and what it can do. Benjamin Horst, working on his own initiative and doing quite well indeed, was able to get funding for an ad in the New York City subway paper the Metro. But that had fairly small effect, as the audience was relatively small. I mean for bigger ads and a coordinated approach. Historically, we have focused on free marketing—we have no real budget—but that has got to change. People have to be informed. Yes, more applications are using ODF, yes, even Google's Writely uses it; yes, governments are trending towards it; and yes, OpenOffice.org is used by uncounted millions. But we have not reached the tipping point. We may be close, but until we are at the point where adoption is inevitable and where to use our competitor's product is seen as an act of wilful and selfish madness, we really cannot rest. The stakes our too high.

I would love that this time next year, at our next OOoCon, we can boast of having not just a 100 million or more users, but of being able to walk into a library or cybercafe and seeing OpenOffice.org on the screens of the machines, of having Google tell us that ODF documents represent a significant portion of documents found, and of having not just passive consumers happy to have such a great product for free, but users involved in their project, however minimally, because they believe in it and wish to further it.

Using something because you believe in it and participating in its creation represents a change in the way people think about things, and it's a change for the better. OpenOffice.org is helping to lead that change—for peace, for a better world. You should all be proud.

Thank you.