So an idea that has been floating around for some time is that you can verify that a link a person puts in their profile belongs to them, by checking if the linked site links back to the profile (with a special attribute that signifies intention), and that it can be used to, indirectly, verify that a profile is "real"
I wonder what the UX for that would look like. I doubt this can or even should be something displayed in-stream like the Twitter verified badge, because with custom emojis, the display name area is absolutely untrustworthy. More than that, the presence of a verified link is meaningless unless you trust the linked site (such as someone's official personal homepage)
So it would probably be something only displayed next to those links.
@dotUser @Gargron login/account/identity schemes frequently conflate identity, security, privacy and authority. they are very much not the same concerns and persisting in 2018 to use rocks to try and drive in screws looks ridiculous. a random number (which is all a pubkey is when you’re not using it to do work) doesn’t prove anything that a rel=“me” link doesn’t.
@zensaiyuki @dotUser While I am for rel=me based verification, public key based verification is not just putting your public key somewhere. You generate a signature of the link with your private key you never upload anywhere, and put that on that link, and clients confirm this signature matches up with your public key. No one else can replicate that.
@Gargron @dotUser so is the actual goal to prevent someone from impersonating a celebrity, journalist or politician? or just any joe shmo. the real question is how much proof of identity is sufficient for the actual goal at hand. in the case of the twitter verified mark- the point of contention is that public remarks by public figures, if taken as genuine, have potentially serious consequences.
The original server operated by the Mastodon gGmbH non-profit