NavList:

Rafal, you wrote:

"Below you can see 0.1' accuracy version of Pub. 249 Vol. 1 for epoch 2020."

Just remember: Tenth of a minute of arc accuracy for an annual volume doesn't make sense in practice. You can use it for computational testing, but those tenths --and even the rounded whole minutes-- are not correct for the whole year. The stars move around. This is why the Nautical Almanac has daily tables for the stars (monthly or at best weekly would be enough, but daily was apparently chosen in order to make all data available at one "opening" of the volume). 

This also raises an interesting issue for those creating homebrew versions of pub.249 vol.1. What does it "mean" to create tables for "epoch 2020"? If you calculate exact values for Jan 1, 2020 and then round to the nearest minute, you will find differences from samples of the published volume. That's not how they do it! Why give preference to the beginning of the year when there are cyclic, annual variations in position that cancel out over the course of one year? If you prepare tables using exact star data for a single day, like Jan 1, 2020, then you're not duplicating Pub.249. You're already creating something new. And given that --if you (any of you) decide stick with that choice-- then go ahead and throw out the whole lot, all of the un-neccessary features of Pub.249 vol.1, as well as the pecularities and the out-right errors in the star selections (there ARE errors). Most significantly, make them available as monthly tables. For Rafal, let's see "Pub.249-RO vol.1 for epoch Nov2018". And when you have things in final form, zip 'em and upload every month for the next twenty years. Why not?! Or... turn it into a web app that allows a sort of dial-a-table with multiple options available, suited to each navigator's tastes.

Frank Reed