5 ms·
I get relatively annoyed about the facile pedantry of this argument, but this is exactly the poster case for it, so here goes: fingerprints are usernames, not p
by saosebastiao 10y ago
I get relatively annoyed about the facile pedantry of this argument, but this is exactly the poster case for it, so here goes: fingerprints are usernames, not passwords.
- dragonwriter 10y agoFingerprints are neither really suitable as usernames (because you can neither guarantee that users have them or that thhey don't change outside of your and the user's control or, really, even that they are unique, though they are conventionally assumed to be so) nor passwords (because they aren't secret, can't be changed if compromised, etc.)
- darpa_escapee 10y agoEven if prints were unique, the heuristics upon which we match them to a given input isn't perfect. It is an guesstimation.
- dvdhnt 10y agoHuh, I've actually never looked at them from that perspective, but it makes total sense.
- cheald 10y agoJust make sure to use a fingerprint manager and rotate your fingerprints on a regular basis.
- Waterluvian 10y agoI disagree. Fingerprints are privacy locks, not security locks. The lock in your bathroom isn't meant to secure the bathroom. It's just a way to ensure that people get the message, "please don't enter". A fingerprint on a phone is a way of saying the same thing. This phone isn't meant for common use, please don't enter.
- yellowapple 10y agoTo expand on your analogy: a username says pretty much the same thing, except instead of a locked door, you have one of those "Do Not Disturb" signs on your door handle.
- saosebastiao 10y agoI mostly cringe at the argument I quoted because it is often misapplied due to how we have historically misapplied passwords. A username is an identity. Historically due to the difficulty of verifying identities online, we have used passwords as a way to do so. And when all we need to do is verify an identity or control basic access levels (the bathroom lock!), a fingerprint is absolutely good enough. But a password is more authorization than authentication: requiring a password is appropriate when you need a conscious decision, not mere identification. Such as for paying for Pokémon toys. So essentially what I'm saying is that I agree with you.
- Waterluvian 10y agoYes, absolutely. :)
- colah3 10y agoThe "deliberateness" or "conciencousness" of a lock is an interesting dimension. A laser cutter I used to work with had two switches, one of which was a safety switch (like [1]) to prevent accidental activation. The goal here isn't security. It's a design decision to prevent accidents. It's almost a kind of intentional inconvenience. [1] Safety switch with cap prevent accidental switching: http://acuteelectrical.com.au/safety-switch/ http://acuteelectrical.com.au/safety-switch/
- Waterluvian 10y agoI bet legal experts could speak to this very well. At what point is someone breaking in?
- justinkramp 10y agoIdentity <> Authentication <> Authority I deal with this in my industry (telecommunications). Just because you've provided proof of identity (eg your phone number, account number), there are still things you're not allowed to do until you've authenticated your identity--and the system determines your authority, to perform an action. This is accomplished through a password, a PIN, etc. On a phone, it's an interesting shift because with a PIN, we essentially bypassed the need for identity and used only a password; regardless of who you are, you can get in if you have the right key. With the move to identity being sufficient to unlock a device, we're saying that just on the basis of identity, the authority that used to come with a password (sans identity) can be granted. It's a 180 degree turn. I don't see a way on my iPhone 6s to require both Touch ID and PIN; it's one or the other. Very few interactions require both, i.e. after a restart it requires the PIN before Touch ID will work. edit: to clarify the difference
- SixSigma 10y agoMy passport is the username, my face matching the photo is the password.