News for hackers. Stuff that matters.
Re:Windows Calculator invalid input with 10^x function by zamadatix on Thursday August 02, 2018 @04:38PM (#17675257)
If I had to take a first guess since 1e10000 is an overflow in the current version therefore x = 1e9999 is the last valid 10^x input. If that holds true my guess would be someone added a check for that input but did it as "5 or more digits" rather than "10,000 or more".

My second guess is they are using some weird algorithm that hits an error case due to the limits of whatever number type they are using.

Re:Windows Calculator invalid input with 10^x function by rspeer on Thursday August 02, 2018 @06:36PM (#17676049)
I wonder if they're using a fraction type to store decimals (a reasonable way to avoid floating-point error), and when the 10^x function is implemented naively on this type, 10^(3.1415) becomes 10^(31415/10000) = (10^31415 / 10^10000) = (overflow / overflow).
Re:Windows Calculator invalid input with 10^x function by istjohn on Thursday August 02, 2018 @09:34PM (#17676733)
> 10^(31415/10000) = (10^31415 / 10^10000)

If n^(a/b) = (n^a)/(n^b) then 2^(1/2) = (2^1)/(2^2) = 2/4 = 0.5 but 2^(1/2) = sqrt(2) = 1.4...

Re:Windows Calculator invalid input with 10^x function by rspeer on Friday August 03, 2018 @02:35AM (#17677691)
Dang, you're right, I should have known that. I apologize for my bad math.

It would be the 10000th root of 10^31415, and at this point what I'm saying is probably entirely unrelated to how it's really implemented.