On Mon, Oct 30, 2006 at 10:34:56AM +0900, Gavin Kistner wrote:
> From:	Robert Conn [mailto:bob.conn / btinternet.com]
> > I haven't had time to do any "extra credit" stuff, or
> > even test this fully, but here goes -
> 
> For what it's worth, here's the test code I've written to test
> submissions. I'm not 100% sure that it's correctly testing things yet,
> but I think it is :)
> 
> The interesting bit for me has been trying to identify any solution
> that ignores requirement #2, such as solutions that always show the
> correct time, or that are always a bit fast or slow. The statistics
> are my best attempt at that so far.
> 
> To use, if you have defined your class name as "FuzzyTime", just
> include this file after your class definition. (Beware long
> incorrectly wrapped lines ahead.)

Woo Hoo! I ran it on mine and it passed.  My variation was as follows. Although
I'm not sure what it really means.  That I'm roughtly 80% the same as the
'actual' time in fuzzy format, 2% behind and 18% ahead ?

    Variation: 21.68% ahead, 1.41% behind
    Variation: 19.69% ahead, 2.47% behind
    Variation: 18.85% ahead, 2.85% behind
    Variation: 19.76% ahead, 1.94% behind
    Variation: 19.14% ahead, 2.19% behind
    Variation: 18.67% ahead, 2.52% behind
    Variation: 19.06% ahead, 2.24% behind
    Variation: 18.83% ahead, 2.46% behind
    Variation: 18.71% ahead, 2.74% behind
    Variation: 18.42% ahead, 2.92% behind
    Variation: 18.70% ahead, 2.60% behind
    Variation: 18.56% ahead, 2.66% behind
    Variation: 18.33% ahead, 2.90% behind
    Variation: 18.30% ahead, 2.95% behind
    Variation: 18.78% ahead, 2.72% behind
    Variation: 18.73% ahead, 2.76% behind
    Variation: 18.67% ahead, 2.79% behind
    Variation: 18.49% ahead, 2.94% behind
    Variation: 18.47% ahead, 3.03% behind
    Variation: 18.35% ahead, 3.08% behind

enjoy,

-jeremy

-- 
========================================================================
 Jeremy Hinegardner                              jeremy / hinegardner.org