[Data-modeling] Intervals as a primitive data type

[Tom Morris]

We discussed intervals in the context of uncertainty for points in
time, but intervals have a fundamental use for encoding time periods
independent of their use in supporting indefinite times.

There are a bunch of types with intervals specified using individual
properties like start_date/end_date, birth_date/death_date, etc.  If
these were encoded using a single property with an datetime_interval
as the type, they could be manipulated and compared much more
directly.  This would obviate the need for hacks such as used by the
timeline view. https://bugs.freebase.com/browse/CLI-7459

I know it's a little late in the game to be changing something so
fundamental, but from another point of view it's early enough to avoid
all the pain that the current scheme is going to cause in the future.

Is it at least a possibility to consider?

Tom