Having looked at https://climate4impact.eu/impactportal/downscaling/downscalingdocs.jsp?q=alternative_model_calendars
it seems very different from finance

From a technical point of view the GCM calendar is determined by the repetition period associated with the variation of solar radiation input. This repetition period can be set to any value but in this context the values 360 days, 365 days and 365.2425 days are relevant. The latter case results in that the GCM simulates the Gregorian calendar.

It doesn't really matter though, from the isodatetime end our job is to provide people with a repating system of years, months, and days, it's down to the user to work out what they want to do with it.

ah, Cylc as LEGO: Lots of useful bits - painful when trodden on.

Just to summarise some (hopefully) Correct information about isodatetime:

• The 360 day calendar is literally just a calendar where a "year" is made up of 12 "months" of 30 "days" each
• Each day is 24 * 60 * 60 seconds long still, but the "real world" duration of the fundamental unit of second is whatever the user wants it to be
• That 2nd point is also true of the Gregorian calendar

• We provide convenience functions for getting the current Gregorian datetime based on real world unix time, but there isn't a canonical 360 day calendar in the world and no-one should try to convert Greg <-> 360 day unless they somehow supply some context for how the two relate for their specific 360 calendar

• We provide convenience functions for getting the current Gregorian datetime based on real world unix time, but there isn't a canonical 360 day calendar in the world and no-one should try to convert Greg <-> 360 day unless they somehow supply some context for how the two relate for their specific 360 calendar (which we don't provide for?)

Also:

• Ignore whatever it is the finance people are doing

Ignore whatever it is the finance people are doing

Probably good life advice

That 2nd point is also true of the Gregorian calendar

Not quite as we provide the ability to represent the current time using the gregorian calendar (though sadly it would seem this is also true for 360 day 🤦‍♀️).

duration of the fundamental unit of second is whatever the user wants it to be

Isodatetime does not handle leap seconds (which effectively means we use a leap-slur)..

Ignore whatever it is the finance people are doing

Ignore what the climate people do too! It's not isodatetime's job to work out what people want to do with the coordinate system we give them.

That 2nd point is also true of the Gregorian calendar

Not quite as we provide the ability to represent the current time using the gregorian calendar (though sadly it would seem this is also true for 360 day 🤦‍♀️). But fair enough, you could just use it as a coordinate system ignoring its relevance to the real world.

duration of the fundamental unit of second is whatever the user wants it to be

Isodatetime does not handle leap seconds (which effectively means we use a leap-slur)..

Ignore whatever it is the finance people are doing

Ignore what the climate people do too! It's not isodatetime's job to work out what people want to do with the coordinate system we give them.