"Anti-retroratchet" from old data entered after derailment

I know @bee is having some Tagtime issues right now, but looking at mit-todoist, is the sort of “anti-retroratchet” that looks to be happening with that graph intended behavior? It seems to come from old data being entered that is dated before a confirmed derailment.

I’m not sure exactly what behavior I’d expect in that situation, but this approach at least looks weird.

1 Like

My tagtime data went missing (still missing a big chunk ~30 hours(?) from
my b/meta graph). So that is absolutely expected. The road was dipping down
to the new (wrong) lower value, so the recommit moved the road down to
where the data was. This is what happens if you delete old data from a
kyoom goal.