1. 29
  1.  

  2. 13

    The counter argument would be Moxie of course:

    One of the controversial things we did with Signal early on was to build it as an unfederated service. Nothing about any of the protocols we’ve developed requires centralization; it’s entirely possible to build a federated Signal Protocol-based messenger, but I no longer believe that it is possible to build a competitive federated messenger at all.

    So the big challenge will come when users expect some new feature which ActivityPub currently does not provide.

    1. 15

      Mastodon and the ActivityPub community have been iterating and pumping out new features on a rapid basis. On a protocol levle, ActivityPub itself is an iteration on the Activity Streams and ActivityPump protocols; themselves an iteration on OStatus. And there are plenty of ActivityPub instances that weren’t initially envisioned: PeerTube, MediaGoblin, NextCloud, … and chess?

      I suppose moxie would argue that Mastodon isn’t or won’t be competitive.

      I argue Signal, just like Twitter, will run out of money.

      1. 4

        Signal will become what WhatsApp was meant to become. WhatsApp could have been a secure messaging layer for businesses and consumers but Facebook made them an offer they couldn’t refuse so that dream wasn’t realized.

        Signal now has a foundation and they have one of the original founders of WhatsApp bankrolling the operation. I don’t think they will run out of money and might even realize the original WhatsApp dream.

        1. 1

          Want to longbet?

          1. 1

            Sure.

      2. 10

        That quote is not really a good counter argument, it basically reads like “federation is bad because I said so.” You have to read the rest of his post to tease out his arguments:

        • federation makes it difficult to make changes
        • federation still favors a service single provider (e.g. gmail and email)

        (Note: I don’t agree with moxie, just posting his counter argument for others to read)

        1. 8

          The counter argument would be Moxie of course

          I’d have a lot easier time taking his arguments seriously if he hadn’t threatened legal action against a free software project simply for trying to build an interoperable client.

          1. 4

            Mastodon seems to cope quite well with this, possibly because there are few implementations and upgrading the server application isn’t too hard.

            But I think the counter argument is entirely correct - it’s not possible (or at least very hard) to build a competitive federated messenger - and that’s completely fine. Competition is one of the parts of the centralised model that leads to de-prioritising users needs so that platforms can be monetise to keep it alive and “competitive”.

            1. 5

              Wait, what about matrix though?

              1. 2

                To clarify my opinion a bit - I’m suggesting that federated networks won’t succeed by the metrics used to measure if something is “competitive”, not that federated networks don’t work. I think Mastodon and Matrix are both really good projects that will be much better than the alternatives long term, since there won’t be many incentives not to prioritise the needs of their users.

                1. 2

                  Matrix from what I heard has scaling issues; we’re talking “three people on a single server massively increases load” bad. I think it’s due to protocol flaws?

                  1. 5

                    Any of matrix’s scaling issues come from federation (trying to sync room state across many different homeserver instances) and the poor state resolution algorithm they were using up until this past summer. Three (or thousands) of users on a single server participating in a room is not a concern, as that is a centralized instance.

                    Highly recommend following the matrix blog and TWIM for project updates, especially for anything about synapse (their reference homeserver implementation). It was recently updated to python 3 and the memory footprint has drastically reduced. Keep a lookout for the “next generation” homeserver implementation, Dendrite, sometime after the Matrix 1.0 spec releases.

                    1. 2

                      I remember reading that this was because the current reference server implementation is simply not optimized. They’re rewriting it in Go (IIRC the new server is called Dendrite), but we’ll have to wait and see how performance changes.

              2. 3

                Is there any Disqus alternative based on ActivityPub for example? Something I could embed into my site, and preferably associate the coments with my mastodon account?

                I have created a Mastodon account just a few days ago, and so far so good!

                1. 1

                  Not that I know of, but it would certainly be possible to create.

                2. 3

                  Decentralization is a vacuum of power. If you don’t fill it, someone else will.