Arithmetic overflow is unhelpful?

[srmaxwell3]

I've written a shift-and-add multiply example and, in the absence of a carry flag (might've been a good idea, folks) I am left shifting a positive value <= 32767 with the intent of discovering whether the result has become negative. No such luck, however, as the simulator complains "OverFlow Error: the value NNNNN is beyond the calculable range...".

Making the simulator more forgiving of overflow would enable more sophisticated programming.

Make the simulator even more PDP-8-like would be pretty awesome, too!

-+- Sid Maxwell