Comment by IgorPartola | Hacker News Reader

Comment by IgorPartola | Hacker News Reader

Parent Full thread

OK, finally looking at it on a desktop...

So my first question is this: if I know "maria" and I want to look her up to get her GPG key, how does keybase handle that? Does it just do an email address lookup, as in goes to, say, GitHub, grabs her email address, maria@example.com, then goes to a public key server and grabs the key that corresponds to maria@example.com?

If that's the case, there is a security issue: what if Maria never published a GPG key, but Chloe did using Maria's email address? Moreover, what if Chloe has access to Maria's inbox and can read these messages I believe to be only readable by Maria?

Edit: I see from responses below that various online presences of an identity tied to "maria" are checked. Is this not then susceptible to its own attack? For example, if Maria does not have a Twitter account and I create one, or compromise hers and post a different key, will I be able to at least introduce doubt into her identity, if not take it over outright?

No, there are no proofs based on e-mail addresses, because such proofs are not publicly-auditable. We could ask that maria prove to the server that she controls a given gmail account, but there's no way for the server to prove that to you.

We want the server to be untrusted, ideally just a dumb message router.

If Chloe wants to impersonate maria, she'll need to get control of maria's twitter and github accounts. Just claiming maria's email address won't get her anywhere. (Note that GPG keyservers are susceptible to exactly the attack you describe).

Hold on. First, GPG servers are susceptible to the same type of attack, except they would never be used that way. You never look up a person by email, the send them an encrypted message using the key you get. Instead, you verify their key and email address out of band: you meet them, check their credentials, then sign the key. Keybase is trying to get rid of the in-person verification, an effort I applaud, but in favor of a much weaker check: whether a few centralized accounts had been compromised.

The other part, where you check Maria's Twitter and GitHub accounts, means that a few things like Twitter, and GitHub are impervious to Chloe: a tall order and a centralized one at that.

Once again, is the point here for me to get a tuple of (email address, public GPG key) so I can email Maria securely? If so, then someone somewhere has to prove that this tuple fetched from the public key servers is valid.

If the point is to only communicate via keybase.io, then the service is centralized, and useless once actual sensitive info is exchanged, the US government takes notice and shuts it down at the DNS level.

Cool, I agree no one should use PGP servers the way I described, but you never know what people are doing out there. To do things the proper way, as you described, is difficult in practice for lots of people.

To answer the question, the point isn't to get an (email address, GPG-key) mapping. It's to get a (public-internet-identity, GPG-key) mapping. People sometimes do this today in an adhoc manner (e.g. tweeting your GPG fingerprint). We want it to be checkable by user-friendly software.

I see. That MO makes a bit more sense then, though is it not then limited to just Keybase.io and will no longer work if something happens to this service? Or more importantly, is there a way to make this distributed?

If the site went away tomorrow, you'd still have keys in your GPG keychain. You'd also have a local cache of the server-side data relevant to you.

All public server-side data is available as a dump (https://keybase.io/__/api-docs/1.0#call-dump-all). Private data like encrypted public keys and password hashes we of course will keep under wraps.

We don't have immediate plans to make the system distributed, but if someone did it, we'd find it very cool. It's just too much for us to do right now.

Could you effectively add an online "in person" check by having keybase.io send the person a large uniquely watermarked placard with a unique OCRable pin that the person holds up while taking of photo of themselves mugshot-style that must be sent in as a lossy image type (where image tampering is detectable). With this image publicly published online, it would be easy to visually verify the person yet another way. Obviously for people with pseudonym identities that are online only, these check would be irrelevant.

Reply on news.ycombinator.com