Hector Martin

I actually don't know how SuperSpeed negotiation works, looking it up now.

Ugh, this is worse than USB 2.0. I don't know what I expected. So there's some low-frequency signaling they use to negotiate, but it's completely silly (and they didn't even future-proof it so they had to tack on another modulation scheme to advertise SuperSpeedPlus).

Well the U is for unintelligible, clearly.

But we're stuck with it.

Anyway it looks like negotiation is actually intiated by *both* sides at once. The transmitter is supposed to detect the receiver's termination impedance (!) and then initiate LFPS polling, and then both sides kind of meet in the middle. I don't even.

This feels like a standard compromised into existance by host and device manufacturers arguing that silicon is expensive, and they don't want to implement that, let the other end do it. Now both ends do it.