> you are vanishingly unlikely to do a better job of designing an encrypted channel than every cryptographer and protocol designer and software security person

> you are vanishingly unlikely to do a better job of designing an encrypted channel than every cryptographer and protocol designer and software security person who has poured time into TLS Given the overwhelming, repeated breakages in the SSL/TLS record protocol I'm really surprised you put the designers of it on such a high pedestal. The researchers who broke SSL/TLS should certainly get that kudos, but the result of this is that a competent designer should be able to do a better job than TLS now if they merely choose to fully research the problem domain.