I released the package that I developed as a part of my PhD in Machine Learning at Stanford. The idea is to be able to boost beyond trees, with more algorithms than the usual Gradient Boosting.
It works in v0.6. Has no documentation except my thesis. Unfortunately I will be away from computers at a yoga retreat for a few months (to unwind from my PhD); so I will not be actively maintaining it, but a friend may handle pull requests.
I forgot to acknowledge the awesome Julia community in my dissertation. You guys are great, and it was a real pleasure working in Julia. I did not even have R on my computer!