As much as there is plenty of new people joining the threadiverse, the real wave starts today, with thousands of subreddits going dark.
Existing Lemmy/Kbin instances get hammered with new user registrations and deploy different coping strategies. Some plead, some close registrations. New instances spring up.
Soon, mainstream media will discover Lemmy exists. They will probably miss Kbin entirely, and most will also be very confused about the federated nature of Lemmy. Some might be able to remember Fediverse exists.
When Kbin finally shows up on their radar, they will find it difficult to explain how it fits into the narrative they already spun. My money is on someone calling it a "fork" of Lemmy. ๐คฃ
Eventually, as more instances start turning off registrations, and as some buckle under the load temporarily, the narrative becomes "this is why Lemmy will fail." Threadiverse will get treated like a VC-funded walled garden. Media will be flabberghasted at how "poorly" Lemmy and Kbin were able to "capture" the people wanting to migrate off of Reddit. They will complain endlessly about how hard it is to choose an instance, "confusing interface", and ask "thoughtful" questions on "how will they monetize".
Eventually, the wave subsides. Maybe Reddit reverses their silly ideas, maybe people get tired. There is a drop in active user accounts on the Threadiverse, compared to the peak of the wave, which is then taken as "proof positive" that Lemmy and Kbin could never "succeed".
What they will ignore, of course, is that by then Threadiverse is several times bigger and more active than before all the Reddit insanity. Communities stay active, people stay active, and slowly Threadiverse grows, as (just like the broader Fediverse) it is not a VC-funded startup that needs a hokey-stick growth.
It's a long-term project of making community-run platforms work. And that takes time, and effort, and love.
Mainstream media will 100% catch up more by the reveal of Meta's Twitter alternative that implements the
ActivityPub
protocol.Yeah but you can be sure Mets/Zuck will either bastardize the protocol from the start, or embrace/extend/extinguish.
Why would the implement it in the first place then?
Why did Google Chat implement XMPP initially (with federation even), only to then destroy it later?
Google's internal structure rewards duplicate efforts and people switching projects.
To be able to ride the wave of popularity of the term "decentralization". Just like BlueSky.
Interesting, but don't you think they'll just wall it off like how truthsocial.com is walled off?
A lot of fedi instances will block any Meta-owned instances on sight. Some will not. How it plays out long-term depends a lot on how well Meta instances get moderated.
That's a really interesting point - they're not to be trusted and they've already shown their hand, so to speak.
I can't see many connections for them...
Yup. But I do see it as potentially enabling people to migrate towards fedi, off of Meta instances, more smoothly than now. Some fedi instances will probably federate with Meta's instances, so one could have an account on a non-Meta instance (thus having access also to fedi instances that block Meta), but stay in touch with contacts on Meta instances.
That just might be enough to pull people towards greener pastures over here. ๐
I am pretty sure that people who already migrated to fedi will mostly not want to migrate back to Meta-owned instances. So it seems to me like it might be a one-way street. Which would be good!
What I really worry about is two things: