It's an interesting debate John! I take the point about reducing the mechanical and electrical complexity, however its not just complexity and reliability that are the players, its also the consequence of a failure - sorry, I know that you of course know that anyway!
I would rather have any number of failure of actuators etc than one complete FBW failure! Designing systems with true isolation / no single point failure modes is quite hard. Yes you can make fancy and very clever software and certify it to level one, but it can still have bugs (trust me, I know, though fortunately non-critical ones!). Once you get beyond a few thousand lines of code, I remain unconvinced its possible to guarantee bug-free!
By the way, by "bugs" I don't necessarily mean only where the software does something not intended by the writer, but also where the writer/system designer specifies the behaviour inappropriately - after all these guys sometimes live in the virtual and theoretical world of the lab, rather than the real world of flight.
Last edited by HeliComparator; 22nd April 2013 at 21:21.