thanks for these references all.
As some folks who help with GSOC mentoring have pointed out offline, this summers work is not to be a research project, but a concretely achievable over the summer by a single student project. if we hit hard obstacles i'll help sort out a concrete path that maintains a path to success, but research here isn't the goal. rather "lets make something that WORKS WELL". Sometimes the novelty requirements for research are contrary to the best tech choices for building robust usable tools.
point being, thanks for sharing the fun reading, if any can help the student along, i'm happy to pass it along, but lets not nerd snipe students into other projects. (i'm bad enough with that for myself as is :) )