(this would have appeared on my blog on Nature Network, but the pulled the plug the day before. Sometimes correlation does not mean causation)
Fans of R and mixed models are aware of the lme4 package. This started out as Doug Bates re-writing the lme package using the new capabilities in R (S4 objects, for those who care about such things). It goes back to at least 2006, but isn’t stable yet: a source of mild amusement for me over the last few years. In software development, an un-stable version has a number starting with 0 (e.g. 0.4), and once the developers are happy with it, it gets upgraded to v1.0. The core R developers released R1.0-0 on the 29th February 2000, citing it as the nerdiest date possible, being an exception to an exception.
Anyway, the version numbers of lme4 show the problems: v0.999375 was released in 2008. I just checked and the latest version is 0.999999-0-1. This is more compactly (and more confusingly) written as 1-1e06-0-1.
I have been worried that lme4 will never become stable, but this latest version mollifies me with the thought that the developers can’t go on forever, so eventually lme4 will become stable when the machine precision forces it to be rounded up to 1.0.
