Developer Notes for Coq-au-Vin
==============================

2013-09-15: Hand-rolled date processing routines?
    The original impetus for this is that SRFI-19's time->second function
    outputs a number that can't be processed without the numbers egg. So the
    dependencies start to look rather heavy. And it turns out that,
    according my time tests, using just Posix time functions + hand-coded
    parsing runs about 10x faster than the SRFI-19 version. It should also
    be noted that using posix string->time is a bit faster than my
    string-split + string->number version. However, that difference is only
    around 20-30%. And if I'm not using SRFI-19, the next-day function needs
    to operate on Y-M-D tuples, so it seems preferable to keep it simple by
    using Y-M-D for everything.