EDIT: seems like it’s both simpler and more complex than I thought to achieve what I want. https://codeberg.org/PassFF/passff/issues/581 suggests it is “just” a local tweak. At the same time, any search I make for finding an actual implementation guide for authenticators is buried under pages of results for using passkeys in your own web app. The ecosystem’s current state seems to be “don’t bother, just trust us!”. Even the damn “passkeys.com” site is a barely disguised ad for OwnID.
I dunno, the browser could open a file picker and let me point it towards the passkey, or perhaps let me paste the key contents into a field?
It’s been several months since I looked into it, so maybe things have changed since, but I don’t see why the browser has to be hooked up to my password manager through a mechanism that seemingly can’t be user-piloted nor user-intercepted. How does a browser API require some external third-party software to be installed and configured to work?! It’s a public/private key pair at the end of the day, right? Not some novel algorithm, protocol, or whatever. Why can’t I simply tell the browser “here is where my existing passkeys reside; place any newly generated ones here as well”? Or perhaps “here is the program to call and here are the arguments to supply when you need to create and/or retrieve a given passkey”?
I don’t think I have claimed the standard itself is locking anyone out - let me correct that if I did. I have just not found a way to not be locked out of the feature itself in the products of big players like Google - my Google account(s) being the one service where I feel like I could actually benefit from using passkeys over passwords (not to mention one of those I have seen pushing the “replace all your passwords with passkeys!” messaging).
To be specific, my password manager is http://passwordstore.org/, (“the standard Unix password manager”).
EDIT: seems like it’s both simpler and more complex than I thought to achieve what I want. https://codeberg.org/PassFF/passff/issues/581 suggests it is “just” a local tweak. At the same time, any search I make for finding an actual implementation guide for authenticators is buried under pages of results for using passkeys in your own web app. The ecosystem’s current state seems to be “don’t bother, just trust us!”. Even the damn “passkeys.com” site is a barely disguised ad for OwnID.
I dunno, the browser could open a file picker and let me point it towards the passkey, or perhaps let me paste the key contents into a field?
It’s been several months since I looked into it, so maybe things have changed since, but I don’t see why the browser has to be hooked up to my password manager through a mechanism that seemingly can’t be user-piloted nor user-intercepted. How does a browser API require some external third-party software to be installed and configured to work?! It’s a public/private key pair at the end of the day, right? Not some novel algorithm, protocol, or whatever. Why can’t I simply tell the browser “here is where my existing passkeys reside; place any newly generated ones here as well”? Or perhaps “here is the program to call and here are the arguments to supply when you need to create and/or retrieve a given passkey”?
I don’t think I have claimed the standard itself is locking anyone out - let me correct that if I did. I have just not found a way to not be locked out of the feature itself in the products of big players like Google - my Google account(s) being the one service where I feel like I could actually benefit from using passkeys over passwords (not to mention one of those I have seen pushing the “replace all your passwords with passkeys!” messaging).
To be specific, my password manager is http://passwordstore.org/, (“the standard Unix password manager”).