The arbitration trace is the part I'm most proud of — every resolution
returns exactly which signals were used and why, so it's fully auditable.
No black box decisions.
The race condition between disconnect and reconnect events is one of the
nastiest bugs in IoT — device sends offline then online milliseconds later, they
arrive out of order, and your system marks a live device dead. Timestamp
checks help but clock drift makes them unreliable. Curious how others are
handling this at scale.
reply