How to Kill a Decentralised Network (such as the Fediverse) écrit par Ploum, Lionel Dricot, ingénieur, écrivain de science-fiction, développeur de logiciels libres.
I don’t have sympathy to GAFA, but the article oversimplifies the reality and jumps to the conclusion here and there. Broken logic is dangerous…
Edit:
For example,
Since its inception, Facebook have been very careful to kill every competition. The easiest way of doing it being by buying companies that could, one day, become competitors. Instagram, WhatsApp to name a few, were bought only because their product attracted users and could cast a shadow on Facebook.
This is oversimplification. Facebook not only acquired WhatsApp, but wanted access to its user data. So, it’s not “only because” they wanted to control WhatsApp before they become a rival.
The article’s logic becomes sloppy like this every few sentences if not words.
If threads uses activity pub, most activity pub users will be meta users using the meta client. Meta will not feel the pressure to conform to the activity pub implementation. They could add features as they want since all their users will use their client. This will cause a sudden incompatibility and the fediverse will have to be the one to fix the problem.
If the fediverse wants to update the protocol to add a feature, we’d have to run it by meta first since they would have to update their client. If they drag their feet it would be hard to force the update knowing it will disconnect the majority of users from the fediverse.
It’s the same situation described in the article with Google and XMPP.
I don’t see any leaps or jumps. This could be how meta kills the fediverse and we’d be walking into it eyes wide open.
I have a different argument for why Meta could kill the Fediverse.
Even before they engage on Embrace Extend Extinguish strategy, the simple fact that Meta will have an extremely large user base from the start may kill financially the rest of Fediverse.
All projects that embrace ActivityPub are not prepared to deal with the volume of data that a Meta’s facebook-like or twitter-like project would bring.
In a best case scenario it would force the development of current Fediverse projects to focus on sustentability earlier than necessary, and the missing features would be delayed. And that alone would cause problems to the future of such projects.
On the other hand, the smaller nodes would see the storage costs rising fast and would be at a higher risk of dying simply because they would not be able to pay.
When a big corporation like Walmart moves into a neighborhood it kills the small stores because it delivers most of what people want more effectively. Then when Walmart closes shops to consolidate those neighborhoods don’t go back to the way they were, they now have no stores.
There is a lot of content in the fediverse that wouldn’t exist with meta, because meta users would provide better content, more discussion, and more votes would mean more granularity so better content rises higher. That would stop a lot of the people who post content on activity pub. They would be too late and have too little engagement to be relevant. Those people don’t magically reappear if meta decides that activity pub was just a bad mistake.
I know that you have put more thought into this topic than I did and you just might be right.
But dude that wallmart analogy misses the point by a mile.
Wallmart kills the small stores by simply undercutting them, this has nothing to do with what meta might do.
And you didn’t really argue my point that people like you and I could and probably would consume meta content if it was federated.
But why would anyone that is part of the fediverse right now jump ship if meta came and went away again?
I damn sure will never create another meta account or use any meta app. Would you?
You wouldn’t create a meta account. But I know I consume a lot more content than I create. Probably 1% of social media users create 80% of the content. If meta joined, the users that make most fediverse content now will see their engagement drop. There will likely not be a good reason for them to post at all since, in all likelihood, that content has already been posted by a meta user or reposted with more engagement.
Eventually they’ll stop posting because it won’t be fun. At this point almost all content will be meta content, and most activity pub clients will be “alternative meta clients” in practice. If/When meta leaves, the fediverse will likely have a fraction of the content it has now, it’ll be a ghost town and have a long and hard road to recovery.
That’s not to mention the other problems in the article.
It’s a numbers game. Getting engagement and knowing your audience are skills. The fediverse is a small place compared to meta. Being a big player in the fediverse for most posters is like being in the best team in a college league. Meta joining with 500-2000x the users is like suddenly having to compete at a national professional level. Certainly a few players have the skill, but most will get benched in no time.
Maybe I’m wrong and I hope that I am, but I certainly know most default sub comments at the other place had no upvotes, no replies, and were at the bottom of the thread never to be seen. On here, nearly every comment i see or post has SOME engagement (like this discussion!). It’s a different game when you have hundreds of millions to billions of users.
I don’t have sympathy to GAFA, but the article oversimplifies the reality and jumps to the conclusion here and there. Broken logic is dangerous…
Edit:
For example,
This is oversimplification. Facebook not only acquired WhatsApp, but wanted access to its user data. So, it’s not “only because” they wanted to control WhatsApp before they become a rival.
The article’s logic becomes sloppy like this every few sentences if not words.
I don’t see any large leaps.
If threads uses activity pub, most activity pub users will be meta users using the meta client. Meta will not feel the pressure to conform to the activity pub implementation. They could add features as they want since all their users will use their client. This will cause a sudden incompatibility and the fediverse will have to be the one to fix the problem.
If the fediverse wants to update the protocol to add a feature, we’d have to run it by meta first since they would have to update their client. If they drag their feet it would be hard to force the update knowing it will disconnect the majority of users from the fediverse.
It’s the same situation described in the article with Google and XMPP.
I don’t see any leaps or jumps. This could be how meta kills the fediverse and we’d be walking into it eyes wide open.
I understand all those points.
I just don’t quite see how that would kill the fediverse.
After all wouldn’t we be exactly where we are now if meta disconnected again?
It’s not like you and I would suddenly start using threads if it started federating.
I have a different argument for why Meta could kill the Fediverse. Even before they engage on Embrace Extend Extinguish strategy, the simple fact that Meta will have an extremely large user base from the start may kill financially the rest of Fediverse.
All projects that embrace ActivityPub are not prepared to deal with the volume of data that a Meta’s facebook-like or twitter-like project would bring.
In a best case scenario it would force the development of current Fediverse projects to focus on sustentability earlier than necessary, and the missing features would be delayed. And that alone would cause problems to the future of such projects.
On the other hand, the smaller nodes would see the storage costs rising fast and would be at a higher risk of dying simply because they would not be able to pay.
When a big corporation like Walmart moves into a neighborhood it kills the small stores because it delivers most of what people want more effectively. Then when Walmart closes shops to consolidate those neighborhoods don’t go back to the way they were, they now have no stores.
There is a lot of content in the fediverse that wouldn’t exist with meta, because meta users would provide better content, more discussion, and more votes would mean more granularity so better content rises higher. That would stop a lot of the people who post content on activity pub. They would be too late and have too little engagement to be relevant. Those people don’t magically reappear if meta decides that activity pub was just a bad mistake.
I know that you have put more thought into this topic than I did and you just might be right.
But dude that wallmart analogy misses the point by a mile.
Wallmart kills the small stores by simply undercutting them, this has nothing to do with what meta might do.
And you didn’t really argue my point that people like you and I could and probably would consume meta content if it was federated.
But why would anyone that is part of the fediverse right now jump ship if meta came and went away again?
I damn sure will never create another meta account or use any meta app. Would you?
You wouldn’t create a meta account. But I know I consume a lot more content than I create. Probably 1% of social media users create 80% of the content. If meta joined, the users that make most fediverse content now will see their engagement drop. There will likely not be a good reason for them to post at all since, in all likelihood, that content has already been posted by a meta user or reposted with more engagement.
Eventually they’ll stop posting because it won’t be fun. At this point almost all content will be meta content, and most activity pub clients will be “alternative meta clients” in practice. If/When meta leaves, the fediverse will likely have a fraction of the content it has now, it’ll be a ghost town and have a long and hard road to recovery.
That’s not to mention the other problems in the article.
That’s where you lose me again.
Why would the engagement drop if a ton more users could consume your content.
It’s not like the ratio of content creators to lurkers is any better on meta apps than here.
Btw thank you for the good faith discussion, I don’t think that would have happened at the old place.
Lol the place that must not be named.
It’s a numbers game. Getting engagement and knowing your audience are skills. The fediverse is a small place compared to meta. Being a big player in the fediverse for most posters is like being in the best team in a college league. Meta joining with 500-2000x the users is like suddenly having to compete at a national professional level. Certainly a few players have the skill, but most will get benched in no time.
Maybe I’m wrong and I hope that I am, but I certainly know most default sub comments at the other place had no upvotes, no replies, and were at the bottom of the thread never to be seen. On here, nearly every comment i see or post has SOME engagement (like this discussion!). It’s a different game when you have hundreds of millions to billions of users.
Explain yourself.
See my edit.