OpenOffice: A Project in Search of an Exit Strategy

Off the Beat: Bruce Byfield's Blog

In theory, I should be all over the story about Apache OpenOffice's struggle for survival. Over the years, I have written dozens of articles about OpenOffice.org, OpenOffice, and LibreOffice, and, although I titled my book Designing with LibreOffice, it includes some mention of OpenOffice, too. Yet now as OpenOffice tries to revive itself, my main thought is:

Could someone please put OpenOffice out of its misery? The time for heroic measures is past. All that is left now is to shut down with whatever grace can be mustered.

I understand the historic reasons why OpenOffice and LibreOffice exist. I understand that the OpenOffice.org developers who wanted a faster development pace and distrusted Oracle acted hastily in founding LibreOffice, even if they have succeeded brilliantly. I am aware, as well, of the bad blood between the two projects, and I imagine that OpenOffice members must pride themselves on being OpenOffice.org's direct legal successor, and on having survived the past six years with a constant scarcity of resources. And I would be missing the point if I didn't realize that one benefit of free software is that people can work on whatever they choose.

All the same, I keep asking myself what OpenOffice can do that LibreOffice cannot do better, and I keep coming up blank. When I suggested on an OpenOffice mailing list that question needed to be answered, one or two replied that the question was irrelevant, and a few others thought the question worth asking -- but no one could come up with an answer.

The only implied answer is that an unknown number of people -- probably less than a dozen -- do not want to see the project close, especially not on their watch. The sentiment is admirable, but hardly, I think, enough of a reason to continue fighting against the odds. Yet for no other apparent reason, the handful of people who are OpenOffice are preparing one last rear guard action, hoping a recruitment drive will produce enough developers to keep the project crawling along.

It is a last effort that, barring the unexpected, will almost certainly fail. So far, by my count, the discussion of OpenOffice's future has resulted in the resignation of two full time contributors, an interest in volunteering from half a dozen who sound inexperienced, and two possibly dodgy schemes to help.

Possibly, a more formal campaign through the Apache Foundation might have better results. Yet why would any developer with enough experience to be useful choose to work on the struggling OpenOffice when they could work on a thriving project like LibreOffice? Aside from OpenOffice inheriting the old OpenOffice.org name, LibreOffice seems to hold every advantage.

Fighting in the Rear Guard
Even more to the point, LibreOffice has always held every advantage. Whatever the ins and out of the LibreOffice fork in 2010, the majority of OpenOffice.org contributors forked with it, including many of the most ambitious and thoughtful developers.

Just as importantly, LibreOffice's multiple licensing means that it can borrow freely from OpenOffice, while OpenOffice, thanks to the restrictions of the Apache 2 license could not borrow from LibreOffice. Given that OpenOffice has struggled in recent years just to handle bug fixes, this situation has only been a benefit once or twice for LibreOffice in such features as the sidebar. By contrast, however, given the obvious pressure to match major features, it greatly handicaps OpenOffice, especially when it is already suffering from a lack of resources.

The number of releases alone tell the story: in the best of times, LibreOffice has had several releases to OpenOffice's one.

However, the full story is hard to realize until you see the difference line by line. Last year, I got hold of a comparative analysis of LibreOffice and OpenOffice produced internally by The Document Foundation, the organization that manages LibreOffice. Its fifty-five pages consist largely of a feature by feature comparison of the two office suite, both generally and module by module. While I have not confirmed the complete comparison, the checks I have made suggest it is accurate (nor, of course, would there be any point in it being false or misleading, since its usefulness depends on its accuracy).

For an OpenOffice supporter, the comparison makes for dismal reading. Here and there, OpenOffice has a new feature that is also included in LibreOffice. Mostly, though, the comparison goes for page after page, listing changes that LibreOffice has made, but that OpenOffice has not. By its end, readers can be in no doubt that LibreOffice is the more innovative of the two office suites by a wide margin. The gap is not even close. With an average of ten items per page, that means around 500 features LibreOffice has that OpenOffice lacks.

I have yet to see a comparable list for OpenOffice, but, assuming one exists, it could only be considerably smaller. In fact, although I regularly compared OpenOffice and LibreOffice between 2013-16 while I was writing my book, I only found one major advantage for OpenOffice: the ability to set line-spacing to one-tenth of a point, a useful feature for experts, but hardly enough to make the average user prefer OpenOffice.

Admittedly, LibreOffice has made major changes to the code, so the claim might be made that OpenOffice is at least sometimes more stable. The trouble is, such a claim is hard to quantify. Moreover, so far as the claim was valid, it would be valid largely because of inaction. OpenOffice.org's code was notoriously inefficient and one of LibreOffice's major ongoing accomplishments has been to clean it and reduce redundancies. Under such circumstances, a claim to stability would probably be a claim to obsolescence and lack of improvement.

I have to admire the quixotic determination of OpenOffice to continue against such odds. Yet when OpenOffice cannot even keep its code properly patched due to a lack of developers, the time has come to admit defeat. I hope the surviving project members will have the grace to admit defeat and find another project -- not necessarily LibreOffice -- where their determination and effort actually contribute to free software.