XEphem uses polynomial series for positions. Even a little beyond their specified range they tend to blow up and become very inaccurate. So I would not trust any info earlier than 1500 B.C.
Hello, after reading the essay Hamlet's mill (De Santillana, Von Dechend - 1969) that investigates myths and supposed astronomical knowledge of ancient civilizations, I opened XEphem to have a look mainly to the precession phenomenon and the constellation where the vernal equinox was during ancient times. The XEphem manual reports that " 2.3.2 Time [...] The algorithm uses values tabulated in the Astronomical Almanac for years 1620.0 through 1998.0, and is accurate to within a few seconds over that interval. Dates prior are from Stephenson and Morrison or K. M. Borkowski, with an estimated error of 15 minutes at 1500 B.C". I wonder what precision one can expect for dates earlier than 1500 BC. Does Sky view is expected to show "valid" positions for stars - zodiac constellations - and vernal equinox? I am not interested to precision in the order of hours. Just wonder how far I can set a date in the past and obtain a "probably correct" result for the vernal equinox and the constellation where it was (just out of curiosity). Best regards Stefano
