Governance & Standards at @protocollabs — Former NYT, W3C TAG, science.ai — Privacy, Web, Science, Politics, Philosophy. (he/him/Ishmael)

  • 0 Posts
  • 6 Comments
Joined 8 年前
cake
Cake day: 2016年11月2日

help-circle


  • @poVoq For instance, I think it would make *a lot* of sense to manage PDS infra with coops the way it’s done in plenty of places for energy provision. Things become a lot harder to manage when the people who are good at providing a commodity *also* have to be good at CoMo. For completely different topics. In completely different languages. Etc. Decoupling really helps here.


  • @poVoq I used to think that treating the server as a cityish thing made best sense. But cities are dense, they are used for everything including many things we often don’t think about (see Jacobs, etc.). The mapping doesn’t work very well, except perhaps for people who are very much in one community rather than overlapping ones.

    The ATProto approach is credible exit and all the properties that make servers into commodities. It means that you have better flexibility in dealing with infra.


  • @poVoq Except that there is no *necessary* requirement to reproduce the constraints of IRL infrastructure specifically at that location. A good question is why pick a server instead of, say, people who use the same undersea cable? Typically that’s because cables are a commodity whereas servers provide a single point of capture. But there are two options: make the server democratic or make the server a commodity (a real one, with no power and near-zero switching costs).