In September 2014, rumors were flying that Apache OpenOffice was floundering and might soon merge with LibreOffice. The rumors were denied, but revived in March 2015 when Jonathan Corbett used development activity statistics to show that OpenOffice was seriously short of developers, and had corporate support only from IBM. Now, OpenOffice’s most recent report to the Apache Foundation appears to reinforce these previous reports, and then some.
To be fair, the report is listed as “a working copy and not to be quoted.” However, I am discussing it anyway for two reasons. First, much of the report was mentioned in earlier reports, which suggests that its information is accurate. Second, when I contacted Jan Iversen, the new OpenOffice Chair, three weeks ago, he gave the same warning even more strongly. Since then the contents has gone through at least one more draft, but with little change of content, which makes me suspect that the excuse is an effort to delay discussion of the content. If I am mistaken, the fact will eventually become obvious, since the report is, after all, a public document.
Portrait of a Project
The report begins with the announcement that Jan Iversen has replaced Andrea Pescetti as project Chair. Accepting the report’s claim that the change was made to rotate the Chair and that the transition was “smooth,” this change is practically the most upbeat news given in the report.
By contrast, OpenOffice is trying to replace the release manger who recently resigned. A cynic might wonder about the implications of having to replace two key positions. Unsurprisingly, the report notes that work on the next release is “progressing slowly.”
In fact, if the report has a theme, it is the lack of developers. More specifically, many project activities are progressing extremely slowly or else are delayed. These activities include:
- “The 2 Mac (buildbot) delivered Q3 2014 from Infra are not operational, but a little group of volunteers work on installing the AOO development platform.”
- “Proof of concept for Digital Signing was made mid. 2014, no further work has been done. Once version 4.1.2 is ready for internal testing, work will continue.”
- “New volunteers for development show up regularly, but our lack of mentors has made it very difficult to keep them active. The level of commits on trunk remain low, only a few simple fixes have been committed.”
Contributors are at such a low that the report expresses concerns over the Apache Foundation’s proposed changes to its current content management system. According to the report, OpenOffice’s web pages depend on features that “cannot be easily ported to other tools,” which would “would cause a significant problem for AOO.” If the proposed changes are made, then “AOO will need to either only maintain only a few pages or run CMS on a project VM.”
The report does its best to express optimism by mentioning the consistent levels of activity on OpenOffice’s mailing lists, and the project’s presence at conferences, but the general picture is of a project that struggles to function from day to day, much less progress towards a release.
Reaching out, its way
OpenOffice is putting on a brave front, but even its leaders imply that changes are needed. One paragraph in the report urges about cooperation with “derived projects,” and how “talks in particular with one project have been more intensive, but has currently not led to any agreements” — which, given the lack of candidates, probably refers to LibreOffice.
However, OpenOffice’s approach to solutions may only add to the crisis. When the report talks about how “derived projects benefit from a AOO with a high level of development,” it is positioning itself as the direct successor to OpenOffice.org, with a paternal relationship towards projects like LibreOffice and NeoOffice.
This attitude is spelled out in a blog post entitled “Collaboration is in our DNA.” “We are really glad to be able to provide other open source projects as well as proprietary products with a rock solid platform to build on top on,” the blog proclaims, mentioning that LibreOffice and others are “benefiting from our work.”
The blog also argues the superiority of OpenOffice’s Apache Licenses and invites related projects to co-operate with it for the benefit of all.
Technically, of course, Apache OpenOffice is OpenOffice,org’s legal successor, Oracle having donated the code to the Apache Foundation. But what the project’s attitude fails to take into account is how little that matters. Forks are legal with free licenses, which makes LibreOffice every bit as much OpenOffice.org’s successor as OpenOffice is.
More importantly, contrary to the blog’s suggestion, Libre Office has not just contributed “added value” to the code. The truth is, LibreOffice has not only done more to improve the quality of the OpenOffice.org code than OpenOffice, but more in five years than OpenOffice managed in ten.
Under these circumstances, OpenOffice’s assumption of leadership appears arrogant, and is unlikely to encourage collaboration. The harsh truth is that OpenOffice needs LibreOffice more than LibreOffice needs OpenOffice.
No one wants to see OpenOffice humiliated. But unless the project can conceive of collaboration as work done by equals, its efforts at outreach are less likely to revive the project than to destroy it all the quicker.