This gets to be 10 times worse when you are downloading over serial, and it blows its guts on the last synch stage....

Checkpointing is a good idea. I would not be too upset if say, on track 8 of a 10-track download something went wrong, but at least the database had been kept up to date to that point, leaving me with *something* on the disk, rather than *nothing* as at present.