Post batch race condition

[austinabell]

If a batch proof starts now and is frontrun, the batch proof will fail a few times and then exit. This should likely cycle back to the start of the service polling, check when the contract has been updated, or listen to update events from the contract and cancel/restart any in-progress proofs.

[austinabell]

oops I was wrong, I forgot I set it up to handle this. It is a bit naive in the sense that if the batch post fails due to a temporal unrelated reason, it will start from the beginning again.

This could be improved to handle cases more effectively, but likely not necessary. Probably better to just cache completed snarks/session IDs in memory or sth to avoid reproving if anything #10