- cross-posted to:
- fediverse@kbin.social
- fediverse@lemmy.world
- cross-posted to:
- fediverse@kbin.social
- fediverse@lemmy.world
The Fediverse is currently divided over whether or not to block Threads. Here are some of the things people are worried about, some opportunities that might come from it, and what we need to do to prepare.
not mentioned:
facebook will make the social web liable when federating their content and use us as fodder for their regulation grievances.
the whole point of threads is to work around international regulations.
Just reacting as I read below…
The theory goes like this: a company decides to support an open standard, because it’s a popular thing that people expect support for. Then, when that company’s offering hits a critical mass of users, they quietly kill off support for that standard and keep the people on their platforms.
Company decides to support an open standard. Once it hits a critical mass, they take control of that standards body. Then they quietly kill of their competition through the standard its self.
If anything, it may be that Threads is looking to be a major player in the space, and hopes to benefit from an ecosystem shift by being the biggest project out there. Maybe that situation looks like developing a better API and clients than what Mastodon has, and getting other platforms to use it over MastoAPI.
Now you’ve got the right of it. Activity hub represents a real existential threat to meta in a way that reddit and twittter as corporate entities don’t. Twitter, reddit, etc, they all have more or less the same incentives around what their ultimate goal is for users. This is fundamentally different in the fediverse. Its a difference of alignment in incentives.
The promise of the Fediverse is that individuals ought to be in charge of who and what they see on the network.
Its definitely more than that. Reddit didn’t make reddit, redditors did. Twitter didn’t make twitter, people tweeting did. Youtube with out people posting content isnt youtube. A major, if not the main point, of the fediverse is that its we own the network, not some third party. Its our content, our community, our network.
Either we’re all going to drown in the noise produced by Threads, or the networks will become increasingly isolated.
Yeah. I have a post on that some where else earlier today. The basic math and network mechanics of it mean that even with marginal adoption, Threads federating is an extinction level event for non-threads instances. If one in ten thousand adopt threads and have a roughly equal engagement rate as current lemmy.world users, there will be a 1:1000 ratio of Threads to non-threads content. Because of this the math involved in social networks will basically make it impossible for any non-threads content to bubble to the top.
A part of me remains incredibly optimistic: the Fediverse might actually hit an inflection point that transforms the entire Internet for billions of people. In our battle against social siloes and surveillance capitalism, we might find that the Fediverse won.
Appreciate the article, don’t necessarily agree with all of the conclusions, but really appreciate the work. My bigger concern isn’t even threads dominating and making the rest of the fediverse irrelevant. My concern is Meta taking control of ActivityPub through soft engagement/ coercion/ engagement. The protocol is whats important, and realistically, we’re still in the infancy of federation & the fediverse. Its underfunded/ a hobby project at best. But its also our best real shot at a free and open internet we can all be a part of. I think it the fediverse needs (no pun intended) some really significant ‘meta’ level improvements that deal with distancing and federation with more granularity. On/ Off federation prevented us from getting to one million users, and a lack of engagement is still whats holding us back.
So, as a rexxit refugee who doesn’t fully understand the details of federation, I have a question. When I make a post or comment and it gets shared out, is the actual content shared out, or just a link back to my content. Because I can see Threads issues either way: if copies of the content get shared and Threads’ userbase is so incredibly massive, then all the smaller instances are going to struggle with storage. And if links are shared, then smaller instances are going to struggle with traffic. [I suspect both storage and traffic will be issues anyway, but that each is more of a problem depending on how communication and federation is handled.]
It’s just something I’ve been curious about, because most of the comments I’ve seen have been about the difficulty of moderating the content, or what Meta will do with the data or your they’ll take over, but I don’t recall seeing anything about the strain on actual infrastructure and the additional cost to support the influx of users.
I ran an instance for a bit when I thought the migration would be more substantial. Basically, each instance gets a full comment of every comment and post, less the heavier data like images and videos (but maybe sometimes images).
Its also all publicly available. If meta wants it, its all out there in the open now, no federation required.
So yeah, it could become an issue for smaller instances, but honestly, its a problem that smaller instances would likely welcome. I’m by no means advocating for meta here, but content, submissions, engagement; they are the life blood of places like this and we just do not have enough. No parts of the fediverse do.
You’re completely right that there are likely to be major scalability issues, at this point I don’t think anybody fully knows what the implications are, and it’s not getting a lot of discussion. This is part of why Meta’s proceeding slowly and presumably we’ll see a lot of performance work over the next few months to deal with the expected onslaught.
I have actually seen people worrying about the computational burden of handling all the extra data this could produce. I seem to recall someone actually doing back of envelope math and concluding that federation with threads would be cost prohibitive for a lot of instances.
If that is true then the whole debate we are having is irrelevant because federation will basically not be possible for websites which are seemingly being operated without anything resembling a business plan.
The Activitypub protocol is nothing but a piece of paper. Meta can rewrite it if they want, but that doesn’t mean Lemmy, Mastodon or the many other platforms would automatically use it. Programmers have to implement it, and then instance admins have to deploy it. So by inertia it’s likely that changes would be ignored by most platforms. However they could easily bribe developers or admins to make certain changes.
My understanding is that corporations are constantly trying and sometimes succeeding at influencing w3c standards to go in a direction which is favorable to themselves. For example increasing the legitimacy of DRM and surveillance. Developers of non-profit software (eg mozilla) then have to choose whether to be out of compliance or support nefarious technologies. If they choose against supporting the standards, then all users notice is that the application “doesn’t work” on certain websites.
I don’t necessarily know if running away and hiding is along term solution to this though.
Federation is different from a browser. Even if Threads pushed through some protocol changes on their side, it would change anything for Lemmy, Peertube etc federating with the original protocol. They couldnt federate with Threads then, but clearly most users wouldnt mind.
Yeah I mean i’d maybe, just maybe consider allowing Threads to exist on the fediverse if not for the mountains of evidence and arguements put forth against it.
Meanwhile the arguments for federation are literally ‘you can message people across the fediverse’. That’s it.
Thank you for your response, I appreciate your insights!
I think if there were really serious problems with a future version of ActivityPub, we could feasibly do one of two things:
- Maintain a fork of the protocol - this has actually already happened once, with an implementation standard called Litepub.
- Move over to a different protocol, such as Zot.
The second route is probably much harder, but there’s no real reason why a zombified Meta version of the protocol would do much of anything to anybody running vanilla ActivityPub at this time. You’d probably have some feature incompatibilities and breakage, but…if you’re not going to federate with them anyway, what can they actually do?
1. Meta is going to Embrace, Extend, and Extinguish the Fediverse
Verdict: UnlikelyThanks, now I know not to take you seriously.
Edit to add: My uBlock Origin extension is blocking “
threads.net
” on his site. Perhaps he’s got some skin in the game.Thanks, now I know not to take you seriously.
It’s all good, I don’t even take myself seriously most of the time. Most of what I have to say is dumb shit anyway.
Real talk, though: I legitimately think that Threads is incapable of actually extinguishing a federated network powered by open standards. Yeah, the infighting might fragment us, and the influx of millions of activities and interactions might overwhelm servers that connect with it. To some extent, they can propose protocol extensions and features and even make an ecosystem push with tooling.
But, so long as servers are federating via an open protocol, no entity can truly snuff out the network in its entirety. An actual EEE move would not actually work here: if they ever made such an attempt, we’d just defederate them.
My article is not a point about how we all need to shut up and start worshipping Meta, but that the things we ought to be most concerned about are in fact the things we’ve always neglected: actual user control over data, the ability for people to decide for themselves on what to connect to, and dealing with the technical requirements of hundreds of millions of people worth of traffic. And that’s just to start! If we want to reach the masses, we have to prepare for these things.
we’d just defederate them.
It’s true that we can always choose to defederate from them. What’s to worry about is their meddling with the ActivityPub standard using their incomparably vast resources, and them making their own extensions to the standard in efforts to suck users back into the Borg. Things like that.
Unfortunately, even if instance admins were to unanimously defederate, Meta—or any social media corporation—could create white-label instances to take their place, and we might be none the wiser of their control of them.
Unfortunately, even if instance admins were to unanimously defederate, Meta—or any social media corporation—could create white-label instances to take their place, and we might be none the wiser of their control of them.
I don’t necessarily disagree with this idea, but they would have to justify the business case to their shareholders. As of right now, the idea of a whitelabel personal silo is a limited value proposition to people not already invested in the Fediverse. If it’s whitelabel, what will Meta do? Start a new company? Inevitably, people would figure it out, and go with something else.
It’s true that we can always choose to defederate from them. What’s to worry about is their meddling with the ActivityPub standard using their incomparably vast resources, and them making their own extensions to the standard in efforts to suck users back into the Borg. Things like that.
I said this a little further up in the conversation, but if Meta produces some horrendous, awful version of ActivityPub that only benefits them, what’s stopping the rest of us from forking the protocol or adopting a different one? If we never switch to their version of doing things, and there’s feature breakage between us and Meta, who actually loses here?
If it’s whitelabel, what will Meta do? Start a new company? Inevitably, people would figure it out, and go with something else.
I think what’s more likely to happen than my maximalist argument, is that some existing instances will get direct or indirect funding or other forms of support from Meta, and start influencing their direction.
right, just like the extended SMTP and now we are beholden to metas magnificent SMTP implementation that has extinguished all others.
pffft. the protocol can only be extended by consortium… thats where your worry should lie. everything ive read says they have no chance bustin into that circle.
so all thats left is 'but geez their users will get this bell and that whistle that other instances wont… yep. exactly as the 'verse intended. get off your fuckin ass and innovate. give those users a reason to not want an @threads account not based on pure fucking spite.
My uBlock Origin extension is blocking “threads.net” on his site. Perhaps he’s got some skin in the game.
That’s a post embed.
Right, a post embed that results in anybody visiting your site gets tracked by Meta (whether or not they have an account there).
Thank you for your feedback. In the interest of working around this, the embed has been replaced by a video that we downloaded from the announcement post (which we were trying to get anyway) and just used that instead.
what does defederation movement have to do with 2nd point
I’m not against federating with Threads. It is a net positive to the network in my opinion.