Interested in the intersections between policy, law and technology. Programmer, lawyer, civil servant, orthodox Marxist. Blind.


Interesado en la intersección entre la política, el derecho y la tecnología. Programador, abogado, funcionario, marxista ortodoxo. Ciego.

  • 0 Posts
  • 39 Comments
Joined 1 year ago
cake
Cake day: June 5th, 2023

help-circle
  • I do not think it is a very good analogy. I do not see how this would turn into a broadcast medium. Though I do agree it can feel less accessible and there is a risk of building echo chambers.

    Not so concerned on that–people being able to establish their tolerances for whom they want to talk to is fine with me. But if the system goes towards allowlists, it becomes more cliquish and finding a way in is more difficult. It would tend towards centralisation just because of the popularity of certain posters/instances and how scale-free networks behave when they’re not handled another way.

    It’s most likely a death sentence for one-persone instances. Which is not ideal. On the other hand, I’ve seen people managing their own instance give up on the idea when they realized how little control they have over what gets replicated on their instance and how much work is required to moderate replies and such. In short, the tooling is not quite there.

    I run my instance and that’s definitely not my experience. Which is of course not to say it can’t be someone else’s. But something, in my opinion not unimportant, is lost when it becomes harder to find a way in.


  • I’m concerned that people are already eager to bury the fediverse and unwilling to consider what would be lost. The solutions I keep hearing in this space all seem to hinge on making the place less equal, more of a broadcast medium, and less accessible to unconnected individuals and small groups.

    How does an instance get into one of these archipelagos if they use allowlists?

    Same thing with reply policies. I can see the reason why people want them, but a major advantage on the fedi is the sense that there is little difference between posters. I think a lot of this would just recreate structures of power and influence, just without doing so formally–after all the nature of scale-free networks is large inequality.







  • For me the weirdest part of the interview is where he says he doesn’t want to follow anyone, that he wants the algorithm to just pick up on his interests. It’s so diametrically opposed to how I want to intentionally use social networks and how the fedi tends to work that it’s sometimes hard to remember there are people who take that view.











  • Not that I expect a lot of consistency from imperialists, but essentially the same lines of argument can be used regarding the Russian Federation.

    An advisory opinion would effectively settle Israel’s “bilateral dispute” without the state’s consent.

    Ditto for .ru and .ua.

    The court is not equipped to examine a “broad range of complex factual issues concerning the entire history of the parties’ dispute”.

    Same thing, especially if we get back to the formation of the Soviet Union, independence referenda, and so on.

    An advisory opinion would conflict with existing agreements between the parties and negotiation frameworks endorsed by the UN.

    This would be Minsk I and II.

    The request is not appropriate as it asks the court to “assume unlawful conduct on the part of Israel”.

    Ditto.


  • The biggest issues for me are:

    1. No centralisation means there’s no canonical single source of truth.
    2. Account migration.
    3. Implementation compatibility.

    No single source of truth leads to the weird effect that if you check a post on your instance, it will have different replies from those on a different instance. Only the original instance where it got posted will have a complete reply set–and only if there are no suspensions involved. Some of this is fixable in principle, but there are technical obstacles.

    Account migration is possible, but migration of posts and follows is non-trivial, Also migration between different implementations is usually not possible. Would be nice if people could keep a distinction between their instance, and their identity, so that the identity could refer to their own domain, for example.

    Last, the issue with implementation compatibility. Ideally it should be possible to use the same account to access different services, and to some extent it works (mastodon can post replies to lemmy or upvote, but not downvote, for example).


  • Historically many if not most conflicts started with the breach of an agreement. Without getting bogged down in irrelevant detail, there are issue of self-determination of Crimea, which repeatedly in 3 referenda (2 if you wish to exclude the last one) pronounced in favour of either autonomy or being part of the CIS (effectively Russian Federation). Likewise, and setting aside the 2014 events for the moment, there also were agreements that, in principle, may have served as a valid status quo, such as Minsk II, and were not complied to by the parties.

    So, sure, some form of trust-building will be necessary. But what’s the alternative? Endless war?