No joke, I actually hit this condition in a test suite and ended up stumbling across the October 1582 date in a Ruby library. It wasn't until I searched "October 10 1582" on the Web that I learned the significance.
https://gist.github.com/abachman/f97806e1c0fe8e4e1849e5f8412...
tl;dr - MySQL uses 1000-01-01 as the minimum value for a datetime field. Different Ruby libraries use different methods to represent dates, which can lead to situations that appear to claim that 1000-01-01 != 1000-01-01.
And so 15 October 1582 used to be the 0 for some COBOL date functions.
Later that was changed to Jan 1 1600. In IBM's systems you can control what you prefer by a switch: https://www.ibm.com/docs/en/zos/2.4.0?topic=services-date-li...