Alt Text
Show parent replies
okay but credit where it's due this one was more well-cited and raised considerations that weren't objectionable on their face (they go hand in hand, naturally)
btw he actually handles a description of rights here to my pleasant surprise. of course it falls apart a bit when trying invoke the normative force but the thin descriptive attempt is something i suppose
i can see a vision for how you get to something useful with architectural rights though, you'd just have to be fastidious and consistent in its use. something like: [constitutional networking] protocols define a set of actions a user* is able to perform to interact with the network.
this definition is codified through the code that everyone actually runs, and is purely descriptive. it is not based on an actual constitutional document or written specification (the implementations are more authoritative than the spec!)
these set of actions one can perform are called architectural rights. btw this also means that anything malicious one is capable of doing is a right, which is fine! i just thought i'd point that out. it also means no one can really complain about their rights being violated, unless in case of an
attempt to change what code everyone runs, which would tautologically be true. updates would be about what rights you get swapped out.
so you have to catch yourself when you try to use rights like this, because there is no such thing as a basic right to infringe upon, because if you can infringe upon it in that protocol, it's not a right
you *could* try to lay out "here's my favorite pet rights" and then design a protocol that meets those criteria, but if they're abstract pet rights it's likely to be a failure, because that's not what rights are on this model. "data portability" is not a right, but
the ability to swap pds is a right
the * on users is really important and a little underdeveloped, because this is also going to be users in the broadest sense, as in any entity that interacts. differentiation seems pretty important but it's also not quite so clear how to actually divvy it up
each protocol might have some place it seems natural to carve nature at its joints and say here's a user tm and here's a feed operator and here's a pds operator and here's a... and that's okay, but im skeptical at the delimiting you can actually do while maintaining clearsightedness
tldr i like the above rendition as a philosopher for the same reason any political moralist will dislike it: it is theoretically tidy but practically completely toothless any time you try to extend it to achieve any of the original motivating goals of its invention
i still think it can accomplish a whole lot, just different things than originally envisioned