Below is a draft proposal for this. I didn't know what you would expect from such a proposal (talked to grafikrobot on freenode and he redirected me here), so the structures that are to be implemented are rather vague. If the proposal is on track, I will detail them. I have also talked so some guys on the irc channel about a more generic approach to trees and tries. What's your opinion on that (given that tree structures are already implemented in /intrusive)?
Hi Alexandru, My initial impression of the proposal is that it's too big. You won't be able to finish all of those data structures by the end of the summer, not with the extra overhead that's required for Boost. Second -- and this is a problem with a lot of proposals -- you need to show how your work would improve Boost. It's easy to list a lot of data structures. How do they fit with existing libraries or data structures? Best regards, Andrew Sutton andrew.n.sutton@gmail.com