On Wed, 2010-01-27 at 12:36 -0800, Dennis E. Hamilton wrote:
Rob, I mainly agree that there is no "type," Serial Day, but there are
an
> incredible number of places where the current OpenFormula functions
produce
> and accept Number values that are interpreted as serial-day values as
part
> of the definition of those functions.
>
Are there really? Aren't most simply conversions? Off the top of the
head I can't come up with one function that really depends on this
conversion!
THE USUAL ADDITIONAL THOUGHTS
>
> Many OpenFormula functions depend implicitly on the global
origin-date/epoch
> value and the calendar mapping that goes with that origin date. This
is
> true for anything involving a start-date and end-date parameter,
especially
> YEARFRAC and anything that has time-period basis adjustments that
depend on
> the actual calendar periods involved.
>
I don't see how this depends (indirectly or otherwise) on the global
origin-date/epoch value and the calendar mapping.
Could you explain please
We can't define these functions
> without acknowledging the existence of a mapping known to the
implementation
> of the functions and talking about the dates those mappings correspond
to.
> The DATEDIF function is a particularly interesting case of this
> interdependence.
>
The definition of datedif is:
Returns the difference in years, months, or days of two date numbers.
[plus a table to handle left overs]
How does this depend on the number<->date mappping?
Andreas
>
--
Andreas J. Guelzow
Concordia University College of Alberta
