John,

The talk I gave on Oudemans' curve will show up in a future issue of The
Compendium - so I'd rather not rehash it here before it has a chance to see
the light of day.  But I'll be glad to respond to your note.

> the shape of these "Oudaman's curve".  If I recall his talk, he said that
> the arrow shape I came up with is wrong, and that a proper Oudaman curve
> produces a dial which is more eliptical in shape.

Please note to begin with that I have made no comment about the design you
came up with - which I have never seen.  The approach you outline is
essentially the approach I recommended for approximating Oudemans' curve.  I
simply used a computer program that did things in much smaller increments:
one second of time for an increment instead of your hour of time.  The
smaller increment gives a much better approximation to the theoretically
correct Oudemans' curve - which is simply somewhat more rounded than the
arrow shape that had been published earlier in the BSS Bulletin.  Presumably
the reason you want to have equal intervals (of length, not angle) between
the hour points is so that when you interpolate, say for the half or quarter
hour,  your estimate is a good one.  Your technique doesn't give what I
would consider a good interpolation estimate.  It gives the aesthetically
pleasing appearance that the hour points are all equidistant from their
adjoining hour points - but I don't view this as going far enough to assure
a good value when you interpolate between them.  It's simply a matter of how
precise you want to be.

> I don't understand why my drawing is wrong if it obviously works!
It's not - and had you contacted me to discuss the question first, I would
have told you so.

Fred

Frederick W. Sawyer III   -    Sciatherics    'Tis nothing but a Magic
Shadow-show,
8 Sachem Drive, Glastonbury CT 06033     Play'd in a box whose Candle is the
Sun.
[EMAIL PROTECTED]     [EMAIL PROTECTED]

Reply via email to