It's time we accept that trusted computing is a fantasy. Rather, we need reliable computing whose security model results in an adversarial cost too high to practically exploit. That is the only feasible model for the future of secure computing.
@donb "trusted computing" as far as I can tell is a euphemism for "owners of the device are untrusted and should not be able to access all data flowing through the device, so we need to build devices that the copyright industry can trust to be controlled by them and not their owners".
@Kensan @HalvarFlake @donb Was re-reading "Principled Assuredly Trustworthy Composable Architectures" last night (PGN 2004, http://www.csl.sri.com/users/neumann/chats4.html#introx). Still relevant definitions I'll keep as reference.
@HalvarFlake @donb well, it is definitely the current use of the term. Whether it is a definition which I am willing to accept is a different matter ;)
It is also a mark of great stupidity that all initial efforts on trusted computing were essentially digital rights management and thereby ensured that there would be a huge reaction to it.
@cynicalsecurity @HalvarFlake @donb Still shaking my head at Intel Protected Audio/Video Path PAVP...
@HalvarFlake @donb Trusted Computing is indeed a loaded term. It is one of the reasons why I prefer the term trustworthy computing: it encompasses the notion that the system has to present a convincing argument to earn/justify the trust of the user.