On Fri, Mar 14, 2014 at 04:45:01PM -0500, John Adams wrote: > Granted, it provides a low level of encryption for clients but it does not > provide Non-repudiability to those users, opening them up to MitM attacks.
It is inappropriate to say "opening up to MitM" if the alternative is plain-text HTTP which can be MitM'd by anyone anytime. Noone has suggested that the user should be given the impression that an opportunistic https connection is safe: Were I a browser vendor I would not show any lock icon at all when using this mode of https operation, yet pervasive passive surveillance is hindered at least on the web and the attacker is forced to step out into the open. What we need from web browsers is: - a way to accept self-signed certs silently - do not show a lock, operate as if it was plain-text HTTP - implement pinning as with Certificate Patrol add-on, so at least we get to enjoy TOFU What we need from web servers is: - generate self-signed certs for any plain-text website and upgrade to TLS/DHE by default Maybe we should give these self-signed certs a standard CA name, like using "*" as the name for the CA. Sounds like a simple and viable band-aid all in all. The kind of things that get discussed on the perpass list all the time. I'm for the clean slate fixing-the-internet-for-real approach, so that's my $.02 contribution of the day on this topic. -- Liberationtech is public & archives are searchable on Google. Violations of list guidelines will get you moderated: https://mailman.stanford.edu/mailman/listinfo/liberationtech. Unsubscribe, change to digest, or change password by emailing moderator at [email protected].
