Closed dbarowy closed 10 years ago
Actually good counterargument: we are not able to identify any errors in non-formula inputs. Thus if we count these non-inputs against a user's effort, we would also need to count our non-finding of them against ourselves, too.
Because a user would have to examine all cells, not just inputs. They don't magically know dependence graphs.