- cross-posted to:
- technology@lemmit.online
- cross-posted to:
- technology@lemmit.online
Passkey is some sort of specific unique key to a device allowing to use a pin on a device instead of the password. But which won’t work on another device.
Now I don’t know if that key can be stolen or not, or if it’s really more secure or not, as people have really unsecure pins.
Nope. Not going to have my entire digital everything depend on me not losing or breaking a single electronic device.
From Ricky Mondello, who works on passkeys at Apple: “If it’s device-bound, it’s not a passkey”:
https://hachyderm.io/@rmondello/111188643228872151
That’s a terrible take … He’s confusing “what it does and how it works” with “how you manage it”.
It’s like saying “don’t call it a password if you write it down”. It’s confusing and unhelpful.
No it’s literally in the spec. Passkeys are designed for cross device synchronization. You have to go out of your way to make it local only (or use a different webauthn spec like physical security keys)
They’re just private keys. By nature you can copy them wherever you want. I guess I don’t know why he’s making that distinction at all.
The original spec is resident keys including TPM protected or hardware token protected keys designed to be impossible to copy. That’s why there’s a distinction.
You won’t need to?
The key is for a single device. Logging in on another one is going to generate another key.
They key is secured with the pin of the device, so when you try to log in, you can use the pin to log in, and not the password.
https://youtu.be/6lBixL_qpro?si=wFFQwrfjQBKDHs5B
Would you have to set up multiple devices when making your account then, if you wanted more than just your phone?
Yes. Unless there was a way to share that key between devices. I know Bitwarden is working on a vault for keys but it’s not released yet. It was due out now so I assume it’s delayed.
Apple is quite dogmatic in that their implementation demands cross device syncing.
It’s definitely possible to have such an implementation and those will be the most common. You take a very small security risk for a huge convenience factor.
For the more paranoid, device-bound passkeys do not sync. If you lose the device, you will need to go through a recovery procedure.
Thanks for posting this video. Very good explanation for those unfamiliar with the technology.
Do you not use MFA at all then?
Authy on the phone while also using their desktop app. If you lose the phone you still have options.
Each to their own but cloud syncing and MFA are a bad mix in my eyes. It has a “who watches the watchmen” problem and it somewhat defeats the point of having a trusted factor when you have an untrusted one on “someone else’s computer”.
Authy have demonstrated why this is a problem (https://techcrunch.com/2022/08/26/twilio-breach-authy/), plus they’re closed source, so it’s a big no from me.
Vaultwarden, a FOSS Bitwarden server compatible with upstream clients, is able to store TOTP, and when self hosted, you are the watchmen.
Yeah, this is fine. It’s closed source, opaque cloud solutions that people should be wary of.
I have the KeepassXC database with the TOTP thoroughly backed up, so nope.
So you are using MFA but for the password manager? (Appreciate that you’re not the OP I was asking)
If your MFA is bound to a single device and you have no backup then you’re doing it wrong.
You can accuse companies of doing it wrong by often only providing a single additional factor, but I don’t see what that has to do with me.
What I’m hearing is that when an authenticator app is the only option, you’ll go with nothing over something.
Perfection is the enemy of good after all.
Looks like I replied to the wrong comment.
The person you had replied to originally commented on not wanting to have the possibility of everything being broken by losing a single device. I think that’s important that everyone realize that some sort of a backup plan is needed, whether that be back up codes, saving the original QR code, or being able to use multiple devices to authenticate.
At any rate, I should have replied to someone else. Sorry for any confusion.
No harm, no foul.
And you’re very much right. There’s a growing problem in consumer perceptions of modern security methods and how they should be applied. Much more time is spent justifying the convenience of proposed solutions, and not nearly enough time spent understanding what the scope is of the attack vectors that lead to these things needing to exist in the first place.
And further, why some of these conveniences - much like the reused-everywhere super-long password of yesteryear - are a danger in themselves.