hutchphd said:
But doesn't classical mechanics (historically at least) assume that there are an infinity of significant figures to be measured? Always worried me...
The set of real numbers is mathematically a trickier concept that you might initially imagine. For example, althought there is an uncountable infinity of real numbers, there is a countable subset called comptutable numbers, that are the only numbers that you can describe - and work with computationally.
This means that you couldn't have a real number lottery, for example. You can have a lottery where everyone can choose a number from a countable set (e.g. any integer, or any computable real), then the lottery organiser chooses one as well, and if there is a match, then you have a winner or winners. But, if you try to do this with an uncountable set, there is simply no way to describe the chosen number; and, there is no way to compare the lottery organiser's number with anyone else's.
I'd don't know that it makes much difference whether we consider QM or CM, but you get the same problem if you say that a measurement can be any real number. Leaving aside the physics, there is no way to describe mathematically a specific number from an uncountable set.
If, therefore, we assume some physical quantity can take any real value, then the actual specific quantity is ultimately indescribable.