RE: [boost] Date/Time formal review

From: Joachim Achtzehnter (joachim_at_[hidden])
Date: 2002-04-17 12:57:50

Jeff Garland wrote:
>
> > The library suffers from a fundamental design flaw: it doesn't separate
> > the concepts of "time" and "calendar". Time intervals and points in time
> > are natural physical quantities that should be manipulated independently
>
> The civil calendar (in particular local time) is governed by the rules
> of legislators not physics.

What Ross was proposing was a different way to partition the model behind
the library. As the core of such a model he proposes the purely physical
aspects which are *not* governed by legislators. This would deal with
things like points in time and distances (durations) between them. This is
Physics.

Another distinct layer would be to represent (label) the physical time
points in a broken-down form. Here politics and human convention enters
and there are multiple different ways of doing this. Irrespective of how a
calendar/time system labels a physcial time point, the underlying concept
is always a physical time point. Some calendar/time systems (those most
influenced by politicians) are so screwy that it is difficult to avoid
ambiguity in mappings, but that is another matter.

> GDTL doesn't try and 'force' radically different date/time systems into
> the exact same mold.  Rather it is an attempt to build constructs that
> enable rapid construction of new and consistent date/time systems.  I
> believe sub-libraries is exactly the right way to handle this because
> the rules, constraints, and usage of these libraries is sufficiently
> different that it demands independent documentation.

Sub-libraries are fine for the calendar/time layer, but for the physical
core there are simple concepts, not many radically different ones,
especially at the interface (if actual internal representation is hidden).

> Well we think about it a bit differently.  The calendar or time_system
> provides the arithmetic rules, types, and broken_down types for each
> system.

The calendar/time system does not provide the rules for physical
calculations. The distance between two physical points in time does not
depend on the calendar used. It depends somewhat on one's relative motion
and location, but I think we can ignore relativistic effects... :-)

> Physical concepts are one domain and civil time concepts are another.
> Date-time systems must handle both.

I think we can all agree on this.

Joachim

-- 
work:     joachima_at_[hidden]   (http://www.netacquire.com)
private:  joachim_at_[hidden]          (http://www.kraut.ca)