Where are you looking for the database file? The program is old enough that it probably doesn't have the modern default file location under My Documents, but rather whatever directory you happen to be in when you run it, or perhaps wherever the program is installed. Do you get the "Welcome to Predict!" message every time you run it? If so, it's not finding the database either, so maybe there is a problem with the file system and the file isn't being created.
Greg, et al,
Problem was on my end. A not-so-senior moment, a brain fade, I was not paying attention. I was not using ver 2.2.1D, but rather an older version that I had modified back around 2000. Downloaded the latest zip archive from KD2BD's website, and lo, there was the missing .DB file, as well as new docs, and a fresh .EXE with all the latest SGP4/SDP4 improvements (2002).
Time really got away from me here. Back in 2000 I was having a lot of fun modifying the C source and building my own DOS and Linux versions of predict that displayed many more satellites on the screen, as well as calculating rise times for more than 3 satellites at a time on the (M)ulti-sat real time display. I was also working with the Fodtrack interface, and following along with AO-40 preparations, culminating in modifying a 10 ft satellite dish to receive telemetry from AO-40 before it went silent (read - s9 w/no preamp!). I also remember the day in 2002 when Oscar 7 came back to life somewhat. Glad to see 7 is still surprising us.
Then I had to put all things AMSAT onto the back burner until recently. So I'm playing a lil catch up with software and bringing my brain up to speed again.
Getting back to Predict for a moment. I had noticed that when I installed my old software setup (on a now much faster computer than I had 10 years ago), Predict was doing fine with several satellites, as I compared it's real time output with that of satpc32 and various online resources. But the output for several older satellites was *way* off. When I installed the latest package, this problem went away. Since I never touched the prediction model code when I was doing my mods, I can only assume that the old orbital dynamic model John was using in the program had begun to act really quirky. The new SGP4/SDP4 code is spot-on to the tenth. The ability to enter bahn coordinates and schedules is a plus also.
All in all, I like predict because it offers a minimalist approach to providing satellite data. It's snappy text based display is just right for my needs. Soon I will begin to rediscover it's operation under Linux, and all that the new distros have to offer... not to mention the quantum leap in hardware since I was last programming!
Thanks guys! tu de Rich, N8UX