This blog post by Ploum, who was part of the original XMPP efforts long ago, describes how Google killed one great federated service, which shows why the Fediverse must not give Meta the chance
This blog post by Ploum, who was part of the original XMPP efforts long ago, describes how Google killed one great federated service, which shows why the Fediverse must not give Meta the chance
@MyMulligan @jherazob I disagree.
I think the key thing is to just make sure that you don’t use non #FOSS clients. #GoogleTalk started as a client for #XMPP, people migrated to it, and then #Google dropped support for #XMPP. If so many people didn’t use #GoogleTalk, the #XMPP network would have remained unimpeded.
At this point it wouldn’t matter, all they need to do is to mess with the protocol and it’d achieve the same thing, Meta and everything in it’s sphere would “work well”, but connecting with true ActivityPub servers would work just glitchy enough to annoy their users and point the fingers towards our side, just like it happened with XMPP
@jherazob
what the heck makes you think that all the #Fediverse users are just gonna leave for #Meta because federation with it is “annoying”? lol
I wasn’t talking about our users, i was talking about theirs, a direct mirror of what the author described with XMPP
@jherazob The #Meta situation isn’t comparable to the #XMPP situation though
Correct, it’s worse, you can very much argue that Google had good faith intentions, you cannot even pretend that Facebook does while keeping a straight face
@jherazob I care more about the effects than intent in this case.
#Meta’s #Threads / #Barcelona / #Project92 doesn’t have the ability to do anything actually negative to the #Fediverse except potentially overload small instances with a flood of traffic.
I don’t get the fearmongering; lots of talk about “breaking the #Fediverse” coming from people who aren’t really doing a good job of articulating how exactly a new #Fediverse software–because that’s all this is at the end of the day–will break an entire network of software that already works with each other.
Example: Meta federates with lemmy. Lemmy is small so it gets more feature requests than it can code up. Meta comes in and looks at the most requested feature that’s been put on lemmy’s backlog. Let’s say it’s some mod tool. Maybe even AI mod tool that sorts comments based on sentiment analysis. And they only implement that feature for Meta clients - not for lemmy. Suddenly mods have a choice - use lemmy and face flood of trolls in their communities or move to Meta and be able to properly moderate those troll waves. Some will stay, some will move. Another new cool feature for Meta, some will stay, some will move. Eventually most users will be on Meta client because it has all these useful features. OP’s article describes the rest.
@Hexorg
> Lemmy is small so it gets more feature requests than it can code up.
Why? From who? Are a lot of #Meta users who are on #Project92 / #Barcelona / #Threads *really* going to be submitting feature requests for a software that they don’t use?
> Meta comes in and looks at the most requested feature that’s been put on lemmy’s backlog. Let’s say it’s some mod tool. Maybe even AI mod tool that sorts comments based on sentiment analysis.
What are the chances that this is something so significant that people would be willing to switch software over it?
> use lemmy and face flood of trolls in their communities
Where are these users coming from? This is already a problem on the #Fediverse, and we already know how to deal with it.
This scenario you’re pitching seems wildly implausible.
You woke up and decided to speak facts, damn
Exactly, thank you
Exactly! Let them join…and be ignored.
But how do we ignore them? Can you block an entire instance?
Yes. A lot of that is going on with instances being blocked en masse for allowing too many spam/bot accounts without any corresponding high quality activity coming from users registered with that instance. It’s very much in flux right now, with instance administrators trying to figure out which metrics to use and what lists to trust, but I imagine a more mature/robust process will be used by most serious instances soon enough.
The only official way is for the admin of your instance to defederate from the instance you want blocked, but that affects every registered user on your instance, even if they’re against the action.
I just released a user script that lets you block instances yourself on the client-side as a regular user - basically just removes post and comments from the HTML if they match your block list.
I see the threat in the sheer developing power of these giants, making all the shiny tools people were wanting, making their service too attractive to be ignored.