On Aug 10, 2013, at 10:52 AM, Scott Kitterman <scott@xxxxxxxxxxxxx> wrote: > Unless you're checking identification provided by sources all agree are > trustworthy, you've done nothing of the sort. You may be able to attach an > unverified identifier to a group of statements, but there's still no firm > connection to identity (I'm not arguing in favor of one, but it seems a bit > silly to expend resources to protect against something you aren't actually > protecting against). I think you misunderstand the threat model.