I noticed that LSS daily code that uses the read_targets_in_tiles function was proceeding rather slowly, so I tried a module swap to 3.0.0, restarted my run, and found that the code ran much faster. Later, running the code for 1b ran more slowly per tile. (I have been meaning to do more tests, but have not yet.) Possibly, this is related to reports of the ledger initialization becoming much slower.