I have a 14 kb pole.exe and two versions of source code in QuickBasic = 4/14 kb from July 1995. I don't remember if I wrote one or both (or even none!), but I know I = did experiment with it. The extended version calculates the best fit to several predicted points = (assuming two or more). Note that for best accuracy for a slow satellite, the program should use = az/elev or Hour angle/decl. data, since siderial time/R.A. changes during the pass, but the telescope = polar axis doesn't move.