[eridius]

15 decimal digits around is the approximate limit for a 64-bit IEEE floating-point number, which is what the built-in Calculator uses (most non-integral math done on computers uses either 32-bit of 64-bit IEEE floating-point numbers). The exponential notation is because floating-point numbers can represent absurdly large values, it just is limited on precision. So you can represent 10^31, you just can't represent 10^31 + 1.

[cbhl]

Some users do expect arbitrary precision for basic operations (addition, subtraction, multiplication, division) since the Calculator app that comes with Windows does it[0].

[0] http://blog.codinghorror.com/if-you-dont-change-the-ui-nobod...

[sneak]

Why should the user of a calculator care how the numbers are represented internally?

More importantly, why should they be forced to?

It's a bug.

[0xdeadbeefbabe]

Well, what if these users you speak of are trying to accomplish more than just homework?