davepage

Could obtain better quality at the higher channel counts by phase shifting the audio for each channel such that the modulation peaks do not exactly align for each (as they do now). Even inverting the audio for half the channels would help.

The Boeing/Ethiopian ETH-2 bulletin quoted in the above linked AAIB preliminary report on page 33 says "Initially, higher control forces may be needed to overcome any stabilizer nose down trim already applied. Electric stabilizer trim can be used to neutralize control column pitch forces before moving the STAB TRIM CUTOUT switches to CUTOUT ..."

It does not appear to be "precisely" the procedure, in that sufficient nose up trim was not selected by the yoke manual trim switch prior to pulling the trim cutouts. And then they did not work the mechanical trim wheel hard enough to reset the residual nose down trim.

And, the crew left the power setting at climb throughout almost the entire sequence, so they oversped the airframe.

And, when they (against procedure) (apparently) turned the electric trim system back on, they did not use the yoke manual trim switch to fix the nose down trim problem -- which is a big mystery.

The MCAS design inhibits its FCC trim down output when manual trim is utilized. So, if they had just continuously pushed the nose up manual trim switch on the yoke until they got the trim to neutral, then pulled the cutout (either the first time or second time), the accident could have been avoided.

History shows this is not a good human factors design on the part of Boeing, but the crew does not look good, either, IMO.

After the scandal, they added an opt-out setting (but on the privacy policy page, where the nominal user will not think to look for it).

It did not

FWIW, I set an ATM PIN for a US issued CC and found the PIN worked for retail terminals in NZ. YMMV

The hardware looks quite good, but do I want to give Xiaomi (an entity controlled by the PRC) access to my web browsing history, identity, contacts, calendar, files, location, calling history, SMS, etc? (these permissions are requested on Android by the associated app).

Keeping things simple, if the input data SNR is, say, 16 bits, and the PID coefficients are 16 bits, then there is 32 bits of significand at the output of the control gain multipliers. For single precision float, this is truncated to 23 bits. So, the system may settle when the transducer input is near zero (and the exponent can scale down to maintain sufficient precision), but then fail to settle well at the ends of the transducer range (where the transducer input is near +/-2^15 and the exponent in the single needs to stay high to avoid saturation). Also, consider the numerical integrator is running an infinite series of adds, so therein lies a trap for error growth, as well.

I have made this mistake before, and the sometimes strange control behavior can be frustrating. Considering the small performance penalty for using doubles on modern hardware, it is excellent advice.