Page 1 of 1

SIN 45°

Posted: Fri Dec 22, 2017 4:56 pm
by toml_12953
It's nice to see the sine of 45° is properly calculated in 3.1. Thanks, Thomas!

Re: SIN 45°

Posted: Fri Dec 22, 2017 6:33 pm
by Thomas Okken
toml_12953 wrote:
Fri Dec 22, 2017 4:56 pm
It's nice to see the sine of 45° is properly calculated in 3.1. Thanks, Thomas!
You're welcome -- but it was calculated correctly before 3.1 as well.

Here's what happened:

In Free42 1.5.2, I introduced a special case for SIN and COS of 45° / 50 grads. This case was implemented incorrectly, but in a way that somehow ended up doing what I wanted it to anyway, except in the Decimal build for Windows. The Windows Decimal build would return a value that was correct in only 16 digits, instead of all 34.

I fixed the offending code, so it does the right thing on all platforms now, but again, the only versions that ever returned an incorrect result for SIN and COS of 45° / 50 grads are 1.5.2 through 2.0.7, Windows Decimal. The DM42 was never affected.