Pavel Vozenilek wrote:
Would you mind to write review or just list of problems and whether/how they can be solved?
First sentences of my review are in my drafts folder. I need more time to finish it.
What role do you see for your system from: http://aspn.activestate.com/ASPN/Mail/Message/boost/2176732
It's more close to Dave's and Alexey's fsm example rather then to proposed Boost.FSM. So, it's not yet a system, it's only an example of metaprogramming on overload sets. Compared to MPL fsm, in my example (a) states are _types_, not numbers, (b) inheritance is used to model states hierarchy and (c) transition table is represented by a set of call operators. There is no history, orthogonality and all these cool features implemented by Andreas. Currently, I'm working on mpl::set interface for overload sets. New version is almost ready, I hope to get my state machine working tomorrow. -- Alexander Nasonov