Less a specific bug, and more a checklist of things to do when your final position of the frb looks wildly off!
Check the flag files. If you must do things manually, you want to flag harshly with 3 second voltage data - hundreds of channels is reasonable.
Check the calibration solution files (.ps files in output/xxxxxx/fluxcal). Compare both the calibrated and raw files, and look at individual baselines. You may need to identify a bad antenna (or few) to additionally flag out. Flat phases and amplutides (~15 Jy for the usual bandpass) is desired.
Check the finder image files - both the .fits produced, and the .jmfit files.
Check the image file, and the sources identified within it. The offset calculation may have been hampered by this.
Less a specific bug, and more a checklist of things to do when your final position of the frb looks wildly off!