Our UR5e monitor adds a completion after every reconnect while the retained PLC bit stays high. Same coupon, same result. The local history already contains it. I can reproduce this in the saved input playback. I want to fix the count without clearing the PLC state just to make the screen behave.
What identity comes with that high bit? If you can show it's the completion already in the saved history, the monitor should recognise it. But don't make it ignore every first reading after reconnect; that first reading could also be work it hasn't seen.
Does the monitor actually load its saved completions before it starts listening? We had a display rebuilding itself while new messages arrived. It was remarkably good at forgetting things for a fraction of a second.
Chen, it subscribes first. The developer reproduced the extra row before history finished loading. Isabel, we added an unseen completion to the playback as a separate case.
Chen's timing catch is useful, but fixing startup order alone leaves the duplicate-message case. I'd want one durable rule for recognising an already-recorded completion, regardless of when it arrives
It did. The revised handler now checks the run and attempt against saved completions. Startup waits for reconciliation. Both duplicate timings pass in playback; an unseen completion is recorded once.
Unresolved, with no count change or resend. The installed monitor passed the retained-input checks with the maintainer. Covering technician distinguished that case from the known duplicate and the newly recorded result.