We had an amazing success with the release of LibreOffice 4.0. New website pages, a flurry of articles (Time Magazine, ZDNet, TechCrunch Ars Technica, Computerworld, Slashdot, just to name a few), and a generally good feedback. We’re collecting as much data as we can to see how far we went in terms of downloads, but empirically we can already say that it was a success. The infra team worked hard to handle a huge load of visits and downloads; a major “Tweetstorm” that lasted for about 9 hours, and web trends that now show that this release was a major milestone in pushing the brand “LibreOffice” across the Internet. One thing is sure: we went out of this release in a different state we entered it.
The process the contributors
survived experienced has however shown its limits. There’s nothing fundamentally wrong with any of its steps: it’s pretty much that all the steps are not well coordinated, and the end result is that we can do much better to streamline the release process in terms of coordination across the board, from the estimated release date, the ongoing RCs to be tested, to the marketing and the design team deliverables. I realize that I might be popping the bubble of many of our loyal followers, but we need to improve the communication across the teams. And if you thought that the LibreOffice release process worked a bit like this:
Well, it works rather like this:
We are thus launching a discussion at various levels of the community to see how we can improve the way we work together, and this is already giving some results. But we need to make sure each team is fully included and synced to the release, and that the issues appearing on one end of the community get properly handled and communicated to the other relevant parts of the community. The silver lining, I think, is clear. Taking a step back it’s obvious we have grown so much in terms of community and project that we can no longer make sure just a few of us push on all the right buttons: there are simply so many of them now, and the reason there are so many is that we have grown dramatically in two years, even more so than what we had thought. Time for a scale up.