Hacker News Reader

1 in 106 is a strange ratio. It seems that this article was reformatted, particularly because of all the random hyphens. Is it possible that the original said "1 in 10⁶"? I would be a little concerned if I was in the military dealing with bombs that had around a 1 in 100 chance of exploding (or even arming) unexpectedly.

View on HN

I'm hoping that's the failure rate for detonation and not unintentional detonations.

Yeah 1 in 106 seems like an unbelievably high allowable failure rate.

I think "functional reliability" of > 0.95 means that a failure of 1 in 20 to detonate could be acceptable, whereas "safety reliability" of > whatever refers to the chance of improper arming of a weapon. False negative vs. false positive, if you will.

While both could be catastrophic under some circumstances, it seems reasonable that you would expect "safety reliability" to be much higher because it's tested more often. Near the end of the document it talks about how safety reliability should be very high.

I'd strongly suspect it was OCR'd 10^6 into 106.

Specifically as they talk about using multiple systems in series to increase safety reliability, where every system must function correctly in order for the weapon to be armed.

Reply on news.ycombinator.com