i think lennart was literally just riding the coattails of the good poettering though. gotta be convenient to have a creative and successful scientist improving your surname SEO when the truth is "i literally backdoored everyone and then blatantly killed the followup effort from florian wiemer to close the dlopen() attack vector by extending POSIX/the C standard, i am a comic book villain irl"
he didn't link to that exact 2018 bertram poettering paper i just linked above though and that one is the exciting one
We consider a communication model between two parties, Alice and Bob, as it occurs in real-world instant messaging (e.g., in smartphone-based apps like Signal). A key principle in this context is that the parties are only very loosely
synchronized. For instance, a “ping-pong” alteration of the sender role is not assumed but parties can send concurrently, i.e., whenever they want to.
again trashing the status quo of security proofs in the field because literally most professed "we proved TLS secure under X assumption" assumes a laughably primitive and bounded degree of adversarial inputs and generally just restates the the description of the system's operations with tons of extra jargon added and no attempt to characterize what abstract mechanisms are at play. bertram poettering keeps interrogating and rephrasing the problem statement like an actual mathematician
