For any DGGS that spans from past time to future time, the effect of leap seconds will need to discussed. This is because leap seconds aren't known in advance. So if a future measurement (a modelled prediction} is assigned a time and stored in the DGGS at that time. Then by the time we get to the assigned time leap second(s) may have been inserted, changing the time at which this is recorded.
Is this only an issue for DGGS with temporal axes based on the Gregorian calendar?
Can we identify an actual use-case where this is critical?
Can it be handled by appropriate use of Error budgets?
How do current systems that span leap seconds currently handle them?
.. this may spin out into a number of issues.
For any DGGS that spans from past time to future time, the effect of leap seconds will need to discussed. This is because leap seconds aren't known in advance. So if a future measurement (a modelled prediction} is assigned a time and stored in the DGGS at that time. Then by the time we get to the assigned time leap second(s) may have been inserted, changing the time at which this is recorded.
Is this only an issue for DGGS with temporal axes based on the Gregorian calendar?
Can we identify an actual use-case where this is critical?
Can it be handled by appropriate use of Error budgets?
How do current systems that span leap seconds currently handle them?
.. this may spin out into a number of issues.