Re: [Chicken-users] srfi-19 time, flonums and s11n

[Shawn Rutledge]

On 9/3/07, John Cowan <address@hidden> wrote:
> So go with the flonums; they're simpler.

If I take that approach maybe it makes more sense to use srfi-18 time
objects.  They consist of only a single flonum, right?  rather than
seconds (which ought to have been a fixnum but instead must be a
rounded-off flonum because of size) plus nanoseconds (which is just
the fractional part of what could have been in the flonum in the first
place).  And they can be converted to srfi-19 times if necessary,
right?  I see that when I do that I get a time of type time-duration.
But if I add it to the date Jan 1 1970 and then convert the date to a
time... geez what a long way around.  Anyway for events in a queue the
srfi-18 times would be good enough, if I accept the inefficiency
involved in using double-precision floating point.