This PR is an expansion on #79 - the Julian start time is constant, since everything is in UTC.
Additionally, as far as I can tell, there is no need to do any fancy time conversions in predict_from_julian(). We can effectively do the reverse of predict_to_julian(): multiply the Julian date by the seconds per day, then add JULIAN_START_DAY.
We are able to remove the mktime_utc() function after these changes, which I have found returns incorrect values in some edge cases. This works since since everything is in UTC, and removes the dependency on time functions or timezone handling.
Thanks for the consideration, and your work on this great library!
This PR is an expansion on #79 - the Julian start time is constant, since everything is in UTC.
Additionally, as far as I can tell, there is no need to do any fancy time conversions in
predict_from_julian()
. We can effectively do the reverse ofpredict_to_julian()
: multiply the Julian date by the seconds per day, then add JULIAN_START_DAY.We are able to remove the
mktime_utc()
function after these changes, which I have found returns incorrect values in some edge cases. This works since since everything is in UTC, and removes the dependency on time functions or timezone handling.Thanks for the consideration, and your work on this great library!