Making the documentation process slicker could be a GSoC project? Perhaps improving the documentation for some existing Boost libraries (s) en route might be good too.
This kind of thing might be better done outside of boost, our tools tend to be too focused on our own needs. For example, one of clang's open projects is to implement a code documentation tool.
Yes, and now Clang has Python bindings to make that easier. I wonder who wrote those? ;) We can't actually accept proposals to work on documentation, since it's Summer of /Code/. This came up a couple of times on Google's extremely high traffic, high annoyance mentor mailing list last year. Actually, writing Doxygen for Clang would be a /great/ project, but I don't think we could mentor it. Could we? Plus, I don't think Clang is quite stable enough in the areas where we need it to be stable. It's pretty close, though. Andrew Sutton andrew.n.sutton@gmail.com