top | item 38659002

(no title)

SamBorick | 2 years ago

The biggest strength of the fediverse is that every instance operator can make their own decisions on this matter.

discuss

order

berkes|2 years ago

> every instance operator can make their own decisions on this matter.

Not entirely. "The fediverse" has this concept of "fediblocking". If a large enough group of admins decides they are very much against this, they will not only block Threads, but block anyone who does not block Threads.

Basically strong-arming every admin who does not follow suit, to do as "the hivemind wants it to".

This has some positive sides to it, but it's also a weak point in the idea of decentralization.

capableweb|2 years ago

> If a large enough group of admins decides they are very much against this, they will not only block Threads, but block anyone who does not block Threads.

I don't think this is true. The instance my group is using isn't participating in any of the fediblocking/blocklists stuff, and hasn't been blocked by it.

glenstein|2 years ago

Exactly. I just don't see this as an either/or. I do worry about embrace/extend/extinguish, and I don't think "well it's an open standard" is an answer. See gmail.

That said, it's better for them to be on it than not be on it, and it's great as a vote of confidence in favor of Activitypub over Bluesky's go-it-alone alternative that I find counterproductive.

A little of column A, a little of column B.

edavis|2 years ago

I have my foot in both the ActivityPub and atproto camps as nothing more than an interested observer/developer, but I want to defend Bluesky a bit here.

What Bluesky are really building is atproto. If bsky.app and atproto were in a house fire and the team could only rescue one, they'd rescue atproto. Throughout this year there was such demand for a Twitter alternative that wasn't Mastodon, people jumped aboard bsky.app despite it only being a testbed/PoC for the development of atproto.

The Bluesky team reviewed the existing protocols for distributed public conversation and none of them checked all the boxes. So they built their own. Their goals were [1]:

- Account portability. Being able move your entire social graph (identity, posts, follows, likes, etc) without the previous server needing to cooperate or even be online [2].

- Scale. A "big-world" view of the entire network to enable global conversation.

- Trust. Letting users build custom feeds so you can control what your timeline looks like and being open by default.

One complaint against atproto is: why not work instead to improve ActivityPub? As someone who has worked with both protocols, I agree with the Bluesky team that it would have been too difficult to retrofit these features into ActivityPub (portability and scale, especially). Plus, a lot of Mastodon power users have philosophical disagreements with the things required for "big-world" global conversation so it's unlikely Bluesky would have received a warm welcome anyway.

So, on one hand, "go-it-alone alternative" is not flagrantly wrong. Bluesky is the only app using the protocol and it does fragment the space. But it's important to remember there are reasons for why they're going alone on this.

(Sorry, I've had these ideas swimming in my head for the past few days and your comment spurred me to organize my thinking by writing it down!)

- [1] https://blueskyweb.xyz/blog/3-6-2022-a-self-authenticating-s...

- [2] https://atproto.com/guides/faq#why-not-use-activitypub

black_puppydog|2 years ago

I have personally yet to see something that got better by fb/meta being "on it" (for an of course also very personal definition of "better")

AlienRobot|2 years ago

I feel like it's also its greatest weakness.

CharlesW|2 years ago

> The biggest strength of the fediverse is that every instance operator can make their own decisions on this matter.

If you look at how Spotify applied this strategy to the once-open medium of podcasting and successfully redefined "podcasts" to mean shows on their closed and proprietary platform, this seems unlikely to matter.