From: jp (jp_at_[hidden])
Date: 2004-03-17 01:15:58
i was searching for finite state machine ideas and i came across some
old posts from Bryan Ross and Andreas Huber regarding a boost fsm.
i love boost, even tho i havent really used it much. i intend to get to
grips w/ boost::graph soon.
i dont kno if there is now a boost::fsm. the above messages were in 2002!
anyway, i have been using an object oriented fsm pattern for years now
which uses virtual function dispatching in place of the old function
pointer table thingy. i find it very robust & easy to understand. i
have made a few preliminary ventures into boosting it into a generic
template, and i wonder if theres still any interest in this idea.
i have a tendency to see everything in computer science as either a
graph or a proposition, so i find fsm's almost universally applicable...
another offspring of my fevered brain is a little lisp (or maybe its a
scheme..?), & i wondered about adding something like this to boost. it
is already amenable to templated type extensions and *i* like it. i
call it terp, so maybe it could be called boost::terp.
jp the a/p
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk