Git's avalanche
loudthinking.com
loudthinking.com
Git was the leading contender. Yes, Git rocks, but there's more to it than that. When your SCM of choice also hosts the Linux kernel (and, indeed, was developed for that very purpose), you know it will never be abandoned. No serious bug can last for long. That risk mitigation is worth a lot, and it's one of Git's biggest strengths.
I realized about three months ago that Git was going to win. I'm not sure what tipped it, but it made me happy: partially because I felt proud to have figured it out, but mostly because it meant I could stop agonizing about which SCM to use.
Viva Git!
The one issue about git that I'm not comfortable with is the whole explicit branching part of the the design. While for a large project this is pretty easily the "right" way to design a DVS (if for a large / professional grade project you're not explicitly planning everthing, you should be worried), for smaller projects I think darcs' patch dependency tracking approach is much better (or at least much lighter weight).
Tools like Github and Gitorious make Git palatable, and are also interesting from a 'social graph' point of view, where relationships amongst Geeks are measured by the lines drawn between the projects they work on or are interested in...
I think the relative simplicity and originality of git concepts is fueling the avalanche. Scott Chacon has been doing some cool stuff implementing Git in Ruby (http://github.com/schacon/git-ruby/tree/master) and possibly ActionScript.