
What do you think of the job of gathering, completing (i.e. with docs and tests) polishing, and submitting all that work as a GSOC project?
Ambitious? Seriously, though... The amount of work required to implement a Boost-quality library is way more than a 3 month project. Even if the library consists of a few relatively simple abstractions, I think it's unlikely that a Boost-ready library can be completed by the end of the summer. This sentiment has been echoed a number of times on the list, already. Successful projects, those that actually make it into the Boost release seem to take about year, in some cases 2 or 3. I don't think that's the for maintenance or extension projects, but for new development, certainly.
Note: most (probably all) STL implementations use RB trees in their implementations, but they don't expose an RB-tree interface, which, had it been available, could have been used to implement e.g. interval sets). So I don't see that as redundant
That's a good point. Doesn't the intrusive library have an RB-tree with an binary tree interface, though? Andrew