[layla5alive]

It doesn't feel like you're very interested in solving this problem - you seem more interested in defending the status quo

[dredmorbius]

In fairness, an adversarial challenge can be useful in pointing out weaknesses (and possible mitigations) to a particular technical approach.

It's not clear that all of those objections are substantive or insurmountable.

"The USB port may have died" might be one possible response. (Not technically a lie, and hence defensible in court.) Or just silence.

Alternatively, some way of directing such probes to the decoy partition and presenting a sufficiently coherent impression of a valid partition might be another approach.

Much of this comes down to risks presented and costs of mitigation (or of getting mitigations wrong).

[gruez]

>"The USB port may have died" might be one possible response. (Not technically a lie, and hence defensible in court.) Or just silence.

That's about as convincing as "wow this phone just decided to experience catastrophic hardware failure after entering your totally-not-duress pin". Not to mention there's wireless adb.

>Alternatively, some way of directing such probes to the decoy partition and presenting a sufficiently coherent impression of a valid partition might be another approach.

That won't work because they'd notice the adb logs don't correspond to actions taken on the actual phone.

>Much of this comes down to risks presented and costs of mitigation (or of getting mitigations wrong).

Right, which is why grapheneos didn't bother implementing it, because it's a huge effort and it's not worth giving users a false sense of security (eg. thinking that the decoy works when it doesn't), and them getting sent to prison for it.