> You mean asynchronous like one thread can generate an exception in > another? Or some sort of alarm signal. > Don't get me wrong, this sounds fesible enough ... but it also sounds > messy and ever so slightly big. Yes, it could potentially get big. But I'm guessing that the common case is for finally blocks to be small and use few locals, few of them being modified in the try block. This together with better register allocation that tends to leave variables in particular registers cuts down on the number of cases that would have to be handled significantly. But perhaps you're right though that this is too involved. Unfortunately I see slim point to finally/catch blocks if somewhat strict and "unsurprising" semantics are not guaranteed. Anyone know how the current crop of C++ compilers is handling this issue? Surely they must have the same problems. r~