2 ms·

And when someone connects to a Commbox, how can they tell (across the network) whether it's operating as described, or an imposter box reporting to the bad guys

by ds9 13y ago

And when someone connects to a Commbox, how can they tell (across the network) whether it's operating as described, or an imposter box reporting to the bad guys while pretending to be trustworthy? This is a problem for a lot of these ideas, not only yours and mine (see my other post where the difficulty comes down to verifying the server). Logically, the solution would seem to be a tamper-proof (or tamper-evident) hardware subsystem, working much like the evil, so-called "Trusted computing" scheme, except with the owner having full control of all the keys.

samstave 13y ago
See my other post in this thread, where I posted the following. Which, please, tell me if this would work or not: >You run two instances of your VM. One is run locally to you - another is hosted. When a message is sent to you from another peer - it must reach both the hosted and the local one to be opened. When both instances are online at the same time - a hash is shared of the message between both instances and the local master can trust and open the message. When you're locally off line and a message is sent - Thus, if your locally offline - your hosted instance receives the message. When you come online - your local instance will receive the message from your hosted instance AND a hash from the senders hosted instance. If the hash from the sender matches the message hash from your hosted instance - the message is trusted and can be opened. Else it is dropped. The problem is that all these p2p secure connections can still always be slurped. The addresses are only within the system and one would never know which hosted instance belongs to what actual address.... (This area needs a lot more thought)