While I fully, really, completely get all the arguments on every side around Meta implementing ActivityPub, I'll confess to not understanding the profound concern that this will kill the fediverse.
Federated social media never "belonged" to anyone. Feel free to block Threads. Build disparate communities!
Or federate with them!
The network and power dynamics are not & will not be any way worse than they are now, or before Threaderation, or before Musk bought Twitter. They are only better*.
For me, the project of federation wasn't to create a global utopia, or even a local one — that's a project that no protocol or technology could possibly enable, and anyone selling that is a fraud.
In goal of federation as I see it is to give people choices; choices that were taken away by Facebook, mostly. That freedom ALWAYS implied a lot of hard work. It's going to be hard, and it's going to take a lot of work. I hope Facebook & others offer N-S-A/anonymous grants to enable some of that work.
@blaine I wonder if the extra activitypub requests from threads could smash a small mastodon instance into the ground if a post went viral on threads due to orders of magnitude increase in potential audience.
@andrewnez@blaine maybe, but that's the same with anything that goes viral on the fediverse (especially links that trigger previews (I've accidentally taken down a few instances by linking to their posts, and I only have a 1.5 thousand followers)
@andrewnez@blaine That's not how ActivityPub works. The big issue right now for small instances is when they include links; we need to start attaching link metadata to posts before they're distributed.
@andrewnez@blaine that’s already an issue with ActivityPub which doesn’t scale very well. That’s why the Fediverse will probably currently reached a stalling point from a technical viewpoint.
the only exception to the above is if Facebook uses this as a way to regulate moderation requirements so that we end up with something like the phone network regulatory environment, where you basically have to be a giant company to run a service. Twitter's first implementation was illegal under FTC rules, and only managed to grow because we got lucky that the telcos were a bit daft. There's a real possibility that Facebook tries to enshrine a global pseudo-monopoly, and that would be terrible.
@blaine I mean, most of the moderation requirements come down to either legal basis (“you legally have to moderate this stuff or you go to jail or face lawsuits”) or stuff that is “you need to be compatible with our rules and we'll be friends, don't and we'll block you from federating with us”
@blaine Only danger is if we get a mono-culture where we instead of standardised behaviour have to implement ”whatever Threads" implements.
But – we're already there – ActivityPub is not enough to replicate Mastodon, so this if anything can actually push standardisation further by having two large interoperable yet independent implementations.
I have hopes that iterations of AP will be more flexible around media types, and that we'll be able to have diversity where email/newsgroups/irc largely failed. I also acknowledge that I am perhaps overly optimistic 😅
@blaine And if not through iterations of AP itself then maybe through interoperable standards, as eg https://fed.brid.gy/ does for the IndieWeb standards
I do hope that AP etc will really embrace progressive enhancement / graceful degradation between instances and implementations, where a rich representation can gracefully degrade to a less rich one where a rich isn’t supported
I tried doing that with “flattrs” back in the day, have them degrade to “donation” and “like”
@voxpelli@blaine that was the point of ActivityStreams having the name/summary/content as defaults so there was something to make sense of even if you couldn't parse all the details, but Mastodon ignores name and uses summary for CW so it already broke that convention.
@KevinMarks@blaine Yeah, and I think one would need to be able to degrade to another entity to really be able to iterate well on media types and activity types
Add comment