2008 TC3 impact determination in the new Find_Orb is troublesome #66
Description
I wanted to test your 2008 TC3 example. So I grabbed 883 observations from MPC, and only two of them were toggled on (this is probably expected). I toggled all of them on, but here lies the issue: the new find_c64 lacks the "Auto-Solve" feature. It was only in the older GUI, which is really hard to find online anymore (https://www.projectpluto.com/find_o64.zip).
Because there's no Auto-Solve, I have to juggle between Full Step, Herget and Väisälä. The residuals are in the ballpark of thousands or tens of thousands of arcseconds, and pressing the buttons may or may not decrease or increase them. It's all trial and error. Sometimes I reach a dead end where Full step fails, Herget fails and Väisälä changes nothing. I have to be very lucky to refine the orbit to the same residual Auto-Solve achieves (1.88).