4 ms·

You could consider it orthogonal before, but would a passkey store accumulate of many more critical accounts could be attached to that google/apple account. So

by dv_dt 3y ago

You could consider it orthogonal before, but would a passkey store accumulate of many more critical accounts could be attached to that google/apple account. So a mistriggered security or regulatory function could cut off access to the account, followed by being cut off from your bank or other accounts wrapped up on that same passkey store?

lazide 3y ago
And DMV and USPS and US State Dept, and a bunch of other places too.
m-p-3 3y ago
> So a mistriggered security or regulatory function could cut off access to the account, followed by being cut off from your bank or other accounts wrapped up on that same passkey store? In terms of key, it's either stored at the OS/browser level (Windows Hello), or at the hardware level (Yubikey). Imagine them as if there were SSH key files; You hold the private key(s), protected by a PIN/password and Google simply has the public key(s) to authenticate you. Even if Google disabled the account, it still doesn't block you from using other stored Passkeys that are tied to other providers. The only place where you'd get in trouble is if these third-parties like the DMV used Google as their SSO.
anonymouskimmer 3y ago
Just a pet peeve of mine directed at you and the GP. "Orthogonal" - the better word for most circumstances unrelated to math is "unrelated". Depending on the situation, "has nothing to do with" works pretty well too. And to the GP: Since passkeys (and passwords, and etcetera) are related to some extent to one's "digital life", then they cannot be "entirely orthogonal". To really, really nitpick, even in math, orthogonality does not imply unrelatedness. Two objects being orthogonal to each other means that they both are related to the extent they belong to the same coordinate system. I guess, by my standards then, using the phrasing "entirely orthogonal" indicates that you believe that the two things are related to the extent they belong to the same coordinate system, but never meet. I don't like this usage because it seems that "digital life" embodies the entire coordinate system in which "passkeys" exist. Nothing is "orthogonal" to its coordinate system, at most it's orthogonal to an axis in that coordinate system. I'm kind of tired of this nitpicking myself, though. Please don't borrow nomenclature from other fields when already existing nomenclature (or better yet, colloquial terms) work just fine. Thanks.
- jjnoakes 3y ago
  Language evolves and it can't be contained. It's a fact of life.
  - anonymouskimmer 3y ago
    Webster did a pretty good job of directing it his preferred way.
- dv_dt 3y ago
  Having worked with non-orthogonal coordinate systems, the use seems fine with me. “Straight line” movement along an orthogonal dimension usually induces a change in multiple non-orthogonal coordinates. So colloquial use of saying something isn’t orthogonal seems like a reasonable way of saying things are coupled in a way that doesn’t allow free change in one direction without dragging along other changes.
  - anonymouskimmer 3y ago
    I didn't get that intent from what the original user of orthogonal posted. Maybe I just didn't understand what they posted. > >Lose your (google) passkeys account for whatever reason (ai says no, wrongthink, itar ban), and you risk losing your digital life. > Isn't this entirely orthogonal to the usage of passkeys? You could "lose your digital life" if you forgot your password, or remembered your password but tripped their risk scoring system (eg. fresh login from an unusual location). So the person they responded to was saying that loss of (centrally coordinated) passkeys means potential loss of all digital life. They are responding that a loss of (centrally coordinated) digital life can happen for a variety of other reasons unrelated to passkeys. The first person is narrowly concerned with central coordination of a particular passkey system, and the second person is saying that if "loss of one's digital life" is the primary concern, then they should instead be concerned with broader central coordination and verification. In essence, their criticism is that cause and effect are being reversed, or at least mistaken. I dropped out of school in the middle of taking linear algebra, so you know more about math than I do. Maybe orthogonal was appropriate here, but I'm still going to push back against its colloquial use in most situations.

by dv_dt 3y ago

lazide 3y ago

And DMV and USPS and US State Dept, and a bunch of other places too.

m-p-3 3y ago

> So a mistriggered security or regulatory function could cut off access to the account, followed by being cut off from your bank or other accounts wrapped up on that same passkey store? In terms of key, it's either stored at the OS/browser level (Windows Hello), or at the hardware level (Yubikey). Imagine them as if there were SSH key files; You hold the private key(s), protected by a PIN/password and Google simply has the public key(s) to authenticate you. Even if Google disabled the account, it still doesn't block you from using other stored Passkeys that are tied to other providers. The only place where you'd get in trouble is if these third-parties like the DMV used Google as their SSO.

anonymouskimmer 3y ago

Just a pet peeve of mine directed at you and the GP. "Orthogonal" - the better word for most circumstances unrelated to math is "unrelated". Depending on the situation, "has nothing to do with" works pretty well too. And to the GP: Since passkeys (and passwords, and etcetera) are related to some extent to one's "digital life", then they cannot be "entirely orthogonal". To really, really nitpick, even in math, orthogonality does not imply unrelatedness. Two objects being orthogonal to each other means that they both are related to the extent they belong to the same coordinate system. I guess, by my standards then, using the phrasing "entirely orthogonal" indicates that you believe that the two things are related to the extent they belong to the same coordinate system, but never meet. I don't like this usage because it seems that "digital life" embodies the entire coordinate system in which "passkeys" exist. Nothing is "orthogonal" to its coordinate system, at most it's orthogonal to an axis in that coordinate system. I'm kind of tired of this nitpicking myself, though. Please don't borrow nomenclature from other fields when already existing nomenclature (or better yet, colloquial terms) work just fine. Thanks.

jjnoakes 3y ago
Language evolves and it can't be contained. It's a fact of life.
- anonymouskimmer 3y ago
  Webster did a pretty good job of directing it his preferred way.
dv_dt 3y ago
Having worked with non-orthogonal coordinate systems, the use seems fine with me. “Straight line” movement along an orthogonal dimension usually induces a change in multiple non-orthogonal coordinates. So colloquial use of saying something isn’t orthogonal seems like a reasonable way of saying things are coupled in a way that doesn’t allow free change in one direction without dragging along other changes.
- anonymouskimmer 3y ago
  I didn't get that intent from what the original user of orthogonal posted. Maybe I just didn't understand what they posted. > >Lose your (google) passkeys account for whatever reason (ai says no, wrongthink, itar ban), and you risk losing your digital life. > Isn't this entirely orthogonal to the usage of passkeys? You could "lose your digital life" if you forgot your password, or remembered your password but tripped their risk scoring system (eg. fresh login from an unusual location). So the person they responded to was saying that loss of (centrally coordinated) passkeys means potential loss of all digital life. They are responding that a loss of (centrally coordinated) digital life can happen for a variety of other reasons unrelated to passkeys. The first person is narrowly concerned with central coordination of a particular passkey system, and the second person is saying that if "loss of one's digital life" is the primary concern, then they should instead be concerned with broader central coordination and verification. In essence, their criticism is that cause and effect are being reversed, or at least mistaken. I dropped out of school in the middle of taking linear algebra, so you know more about math than I do. Maybe orthogonal was appropriate here, but I'm still going to push back against its colloquial use in most situations.