[Celestia-developers] Fixes to the star database

[Andrew T. <ajt...@go...>]

Following discussion with Fridger on the forums, I have revised the
rejection criteria used in building the star database. Instead of
trying to use a graded system to reject the stars, I am now using hard
cutoffs: stars are now rejected if they fail any of the following:

1. Magnitude information must be present.
2. The parallax must be greater than 0.2 mas.
3. The error in the parallax must be less than the parallax itself.
4. The error in the position, given by Sqrt((e_RA*cos(Dec))^2 +
e_Dec^2) must be less than 20 mas.

The values of 0.2 mas and 20 mas were chosen because those values were
used as grade boundaries in the original database generating code
(buildstardb.cpp).

If anyone wants to suggest revisions to the criteria, it would be
better to have this sooner rather than later.

Andrew