[tz] testing tzcode etc. with negative leap seconds

On 7/23/21 10:41 AM, Kerry Shetline via tz wrote:

in a few years we might experience the first negative leap second

If that happens, I expect that a lot of software won't be ready for it. In the interest of preventing time software bugs, I'm attaching a copy of a file that is like tzcode's leap-seconds.list copied from NIST, except that it contains several negative leap seconds. It's a bit of an extreme case. I used this file to test tzcode, and found one (verrry unlikely) tzcode bug that I plan to fix soon. Others are welcome to do further tzcode testing. You might want to give the attached file a try with your own software.