I don't know if we have an exact definition, pixi , but using a current example there's a node that's currently flooding the mesh with packets, usually a result of the device in a weird state and needing a reboot, software upgrade or there's a hardware failure.
It's impacting the performance of routers in such state.
Because we don't know who the owner is or where the node is identified, till the problem is resolved, blocking the node seems like a fair action to prevent everyone else from being impacted.
@MTEK has tested blocking using the meshtastic CLI tool, which would be easy to script around, where it loops over a list of provided node IDs.
We'd likely need to maintain two lists as the block list wouldn't be implicit, so if a previously known-bad node goes into a good state, removing it from the list and looping over the list again won't remove that from the block, therefor we'd require problematic list and no-longer-problematic lists to loop through for unblocking.
