• 0 Posts
  • 73 Comments
Joined 1 year ago
cake
Cake day: June 22nd, 2023

help-circle



  • The only thing we know without a proof is that they might be doing it. We don’t have a proof they do it but we also don’t have any proof they are incapable of doing so. A reasonable course of action would be to take precautions against it while not condemning them either, until they are either proven actually guilty or actively unwilling to up their security, which would also strongly imply the former.







  • If you’re asking about a personal opinion: any policy purely based on tradition is worthless. Tradition is just peer pressure from dead people. Just like any peer pressure, it’s highly unlikely to produce anything but grief. If something is based purely on tradition without any other reason to exist, it’s unlikely to be an optimal policy.

    Back to the initial question. I don’t think we can get infinitely progressive but we can keep subtracting the cruft of tradition until there is no necromantic peer pressure left at all. Mind that if something happens to be a tradition but still has a good reason to exist, it should be evaluated like any other idea in terms of being good or bad. I mean removing just one of the reasons to keep this idea. If it is left with zero reasons, it’s out. Otherwise it’s fair game.








  • It certainly feels dangerous if forced upon users not aware of the trade-offs. For people already accustomed to using hardware keys, it’s very much an improvement, as more services will support them too. The problem is in the awareness. On the other hand, people already treat regular passwords as throwaway data and expect services to just let them in, or even never log them out. In this scenario, maybe passkeys can still be an improvement: roughly just as much as enforcing using a password manager.



  • Federation combined with keeping the historical federated data consistent is certainly a bitch. We can’t have it all. It could be like email that only handles delivery at any point in time and history is purely local, but Mastodon specifically keeps the federated data public. Propagating the change on the historical data to the federated instances would be nearly impossible. I don’t see how it could have been done better without sacrificing something else.