UMBRAPHILE 2.4.0 ADVISORY FOR 23 NOVEMBER 2003 TOTAL SOLAR ECLIPSE

1) - The element set is based on the current lunar and solar ephemerides, and assume a delta-T of 68.0 seconds.  The aggregate error in predictive extrapolation to 23 November 2003 may be as much
as about half an arc minute in latitude (about 1km) and an arc minute in longitude (also about 1km at these latitudes) near the extreme ends of the path.

2) - Fourth contact times near the end of the path for 0h or later are on 24 November 2003.

3) The computation of circumstances for "Local Noon" does not work and generates an error.  I am aware of this bug (specific for the case of a polar latitude day-crossing eclipse).  I will fix this in
the next release of Umbraphile.