Caveats
The caveats. As to my motivation (please remember to play the man not the ball): I do not intend to make anyone afraid, uncertain or doubtful. If graphs scare you - please look away at this point. These graphs are built from estimates, hopefully they are fairly un-controversial ones, I detail them at the bottom. This is probably misleading in all sorts ways I didn't discover yet. My hope is that it provides a more helpful picture of the world today than this history graph that gets a frequent airing. By rendering only the last two years, we de-clutter lots of lapsed projects, and by not rendering version numbers we can use perceptual area for showing something more useful: an estimate of user-base. As/when I discover major bugs I'll update this, it is a work in progress:
A user-base & release picture
Notes:
This is clearly where it gets hairy. I've tried not to offend people, though I've tried to round against my obvious bias. Also, drawing very thin boxes in draw gives some rounding imprecision.
- OpenOffice.org was commonly supposed to have around 100m users, the vast majority on Windows. I collapse the (sweet) Portable variants, as well as Linux distribution versions of both LibreOffice and OpenOffice into their parent project to simplify and clarify.
- IBM announced that Lotus Symphony had 12 million users; I rendered it as 15m, and include it as an open source project on the basis that its opening is promised.
- OOo4Kids has a download counter that reads 1.05m downloads today, so I turned that into 1.5m to be safe.
- LibreOffice has a conservatively estimated user base of 25m (cf. Gartner's 35m Linux Desktop users alone).
- NeoOffice had around three million downloads per year; though they claim a user-base of over half a million. I put them down as 1m.
- OxygenOffice has a couple of million downloads, over 500k in the last year (vs. 2.2m over all time), so I put their users at 0.5m.
- RedOffice is omitted, mainly because I found it hard to find release date information, nevermind estimating user-base. The best stats we have don't suggest a vast Chinese user-base though.
Everything is meaningless
Of course user-base is a poor heuristic for future relevance. That would be driven by features and marketing, as provided by developers and contributors. Potentially graphing contributors in a similar way would be possible, but that was harder to acquire in half an hour. It would be ideal to render more than a rectangle - showing the evolution of the user-base over time, but I don't have that either, and there is some considerable estimation here. Finally, this doesn't include AOOI, which has yet to release a product, it'll be interesting to see how things change when they do.
