NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Bowditch 2002
From: Jared Sherman
Date: 2002 Oct 3, 16:36 -0400
From: Jared Sherman
Date: 2002 Oct 3, 16:36 -0400
Mitch- <1. Always round up, like a modern electronic calculator> But they don't. "Modern" calculators for at least 20 years have had the option of round-up, round-off, or round-down available with sometimes varying results. And the option sometimes to simply truncate the digits, i.e. truncate at 2 places, 3 places, etc. And there's an additional source of error I suppose for degrees+minutes figures as opposed to degrees+decimal. The latter is now the international engineering standard for lat/lon positioning apparently, minutes and seconds are another obsoletism we practice here.AFAIK any sine conversions break down the minutes into decimal degrees first and then compute, but one could certainly prepare a look-up table with 60 values in it (rounded properly or not) and then add them. Another table, another opportunity for a typo? In all cases, I've seen some LONG decimals come out of simple calculations, like converting 7 minutes (from your 32d07m example) into decimals. Long enough to exceed capacity and cause rounding routines to kick in on many little creatures. Somewhere, something, I'll bet is running the wrong roundings--but the practical difference in position? Have you tried reducing a sextant sight yet to see what difference it makes?