It seems like we will not ever fully win our crusade to prevent users from relentlessly hammering to solve all problems so I would like us to go back to defaulting to hard hammer attempting to dump logs if possible. Before trying to dump it can check if there's an epics connection by pulling a simple PV (like fpga temp) if it successfully reads that PV then dump the epics tree if it fails with an epics timeout then right that out instead of the logs before proceeding to the hard hammer.
It seems like we will not ever fully win our crusade to prevent users from relentlessly hammering to solve all problems so I would like us to go back to defaulting to hard hammer attempting to dump logs if possible. Before trying to dump it can check if there's an epics connection by pulling a simple PV (like fpga temp) if it successfully reads that PV then dump the epics tree if it fails with an epics timeout then right that out instead of the logs before proceeding to the hard hammer.