Threat model -- what we aim for

Protect against somebody watching Alice

Protect against curious Bob

Protect against `some' curious nodes in the middle