I’m Dan, and I work on Persona. Whether or not Persona is a single point of failure is a great question, and one that we’re trying hard to address. Fundamentally, Persona is built atop a distributed protocol that doesn’t rely on any central servers or authority.

Think of it like this: With Persona, you’ll be able to sign in to any supporting website by signing in to your email provider and passing along proof that you were able to do that. It gives you the ability to say to a website “Hi, I’m dcallahan@mozilla.com, and here’s the proof.”

To do that transaction completely autonomously, you only need a browser that supports Persona and an email provider that speaks the BrowserID protocol. But here’s where things get tricky: No browser understands Persona (though we’re working on it for Firefox 17!), nor do most email providers understand the BrowserID protocol.

So, what do we do?

As a fallback, we have a javascript library at browserid.org that takes care of Persona support on browsers that can’t do it themselves. If that site goes down, those users won’t be able to even attempt to log in to sites using Persona. We also host a service at that same site that signs credentials for users whose email providers don’t understand the BrowserID protocol. For example, until gmail.com speaks the BrowserID protocol, Gmail users will have their credentials validated by our fallback at browserid.org.

This second bit *is* a single point of failure in terms of security: if browserid.org gets compromised, then attackers will be able to masquerade as Persona users on other websites. All of our code is open source and subject to peer review at https://github.com/mozilla/browserid, and we can also mitigate the damage of a breach by closing the hole and invalidating the cryptographic keys that we’re using at browserid.org.

But remember, everything at browserid.org (soon to be login.persona.org) is a fallback. If your email provider speaks the BrowserID protocol (or if you add support to your own domain), then you’re no longer at risk by a potential compromise at browserid.org. You get to choose who you trust, and, if you want, to take direct control of the security around your identity. Awesome!

…But doesn’t that make your email provider a single point of failure? Yes, but it *already is.* If someone compromises your email address, it’s trivial to gain access to your accounts elsewhere through their “forgot password” links. The difference with Persona is that *those* sites are no longer storing any sensitive authentication information, so, say, a LinkedIn leak won’t put you at risk elsewhere.

Lastly, as to your question about having the browser proactively help users select and use better passwords, that’s an active focus of Paul Sawaya’s Watchdog project.

If you have any questions, please let me know! You can find the Identity team on irc.mozilla.org in #identity, or on our mailing list at https://lists.mozilla.org/listinfo/dev-identity.

Why not have the browser enforce unique passwords, if I signup on Twitter with the password 12345, the browser could salt/hash/etc it.
Also when signing up to Facebook later the browser should not allow me to use a password I’ve already used on another domain.