This text made me realise something: "defed or not defed" discussions are ultimately rushed.
Because at the end of the day, most Mastodon instances might defed Threads. Not due to Facebook's help in genocides or because they're a big corp, but simply because admins will say "screw it, 90% of rule violations come from Threads users, I'm not dealing with this shit."
Most people don't even know what's a proprietary image format. From their PoV it would be "shitty broken Mastodon doesn't show images properly". And they would still pressure Mastodon users to switch.
if Threads won’t display in a browser they’ve just blown one of their legs off.
I'm not sure but I think that a similar strategy could work for browsers, using a web plugin.
But even if Meta decided that Threads is unavailable from browsers, it wouldn't be blowing one of Threads' legs off. There are far more mobile than desktop users nowadays; and if they want to EEE the Fediverse, they need numbers for that.
Note: I did read your comment fully, but I'm going to address specific points, otherwise the discussion gets too long. (Sorry!)
“Some data format” is still a pretty vague handwave [...]
It is vague because there are multiple ways for Threads to screw with the Fediverse through data formats. But if you want a more specific example:
Let's say that Meta creates a new image format called TREDZ. It fills the same purpose as JPG, but it's closed source. The Threads app has native support for TREDZ images, but your browser doesn't render it.
If you access a Mastodon instance through Threads, everything works well, since the Threads app has support for other image formats. However, since your browser and current Mastodon apps have no support for TREDZ, pics in this format fail to render. You get broken content as a result, and probably some Threads crowds screeching at you because you ignored their picture, saying "u uze mastadon? lmaaao its broken it doesnt even pictures lol", encouraging you to ditch your instance to join Threads instead.
And you might say "reverse engineer TREDZ, problem solved". However:
As such, on a practical level, it would be not feasible to reverse-engineer TREDZ. And even if it was, the time necessary to do so is time that Threads is still causing damage to Mastodon.
Of course, this is just an example that I made up on the spot. Meta can think on more efficient ways to do so.
I’m sure that Meta would just love to be able to push a button that made all their competitors die. [...]
Yup. As you said, everyone wants that button. But due to the difference in power, Meta is closer to get that button than Mastodon is.
the Fediverse seems pretty solid against attack to me.
The protocol might be solid, but the community isn't. Communities stronger than the Fediverse died; and the Fediverse has the mixed blessing of decentralisation - the death of a part doesn't drag the other parts to the grave, but the survival of the other parts doesn't help much the dying one either.
The difference is the same as between boiling a frog* by throwing it in hot water, versus throwing it in cold water and heating it slowly.
In the defederated scenario, people resist to ditch Mastodon and go to Threads, for ideological reasons. The only ones who'd do it are the ones who are pissed at Twitter alone, and short-sighted enough to not realise that the issue with Twitter applies to traditional social media as a whole.
In the federated scenario, however, that resistance has been slowly degraded. Because Mastodon users are already interacting with Threads users, forging social bonds with them, and they'll try to avoid to lose those bonds.
I’m more worried about the load if it truly gets big and mastodon and threads interact a lot, tbh.
I'm a bit worried about this, too. You toot something, it gets insanely popular, and now Threads users hug your instance to death, the old Slashdot effect.
*inb4 boiled frogs are bad science, but a good analogy.
I fully agree that it doesn't matter for Lemmy right now. The issue is mostly Mastodon and Kbin, as both compete directly with Threads; and in a smaller scale Friendica, Matrix and PixelFed as they compete with FB/WhatsApp/IG.
The main reason why I support defederation is to not have users in Mastodon relying on contacts and content from Threads at all. Because, once Threads pulls off the plug (eventually they will want to), Mastodon won't be some small but stable network; it'll be a shrinking one, and that's way worse.
Sorry for the wall of text.
What specific features do you have in mind that could be implemented in a closed-source manner that couldn’t be reverse-engineered and implemented by open-source instance software too?
The features don't need to be impossible to reverse engineer; they could be costly enough to do so, rely on other FB/Meta platforms, or demand server capabilities past what you'd expect from typical Mastodon instances. For example:
and it’s unclear what benefit it would serve Meta that they can’t accomplish by just not joining the Fediverse in the first place.
Killing a bird and a baby mammoth with a single stone, before they grow and invade your turf.
On one side you have Twitter/X; it bleeds money and Musk is an idiot, but he has enough money to throw at the problems until they go away, and he has a "vishun" about an "errything app" that would clearly compete with FB/IG/WhatsApp. On another you have the Fediverse; it's small and negligible but it has potential for unrestricted growth, and already includes things like Matrix (that competes with WhatsApp) and Friendica (that competes with FB).
From Meta's point of view, Twitter/X is by far the biggest threat. It could be addressed without federation, but by doing so would feed Mastodon, and a stronger Mastodon means a stronger Fediverse and this power would put Matrix, Friendica etc. in a better position. With federation however they can EEE one while killing another, and still advertise the whole thing as "I don't understand, why you say that we have a monopoly over online communication? We're even part of a federation? Meta plays nice with competitors. I'm so confused~".
They might not be inherently bad, but they'll be likely bad depending on how it's done, and Facebook isn't to be trusted.
Just for the sake of example:
Note that Facebook has a long story of user-hostile decisions; as in, this crap wouldn't be below its moral standards. So, while most of the time this would be FUD, in this case it's just F, no uncertainty or doubt.
I think that Facebook is trying to kill the Fediverse and Twitter, before either becomes a real competitor.
It makes sense when you look at the big picture; Facebook's power is mostly Facebook itself (connecting people), Instagram (sharing pictures), and WhatsApp ("private" [eh] messaging). Microblogging has a small market in comparison with those three, but it opens a door to them - so both the Fediverse and Twitter have room to expand right into FB's turf.
So in the case of the Fediverse, if my reasoning is correct (dunno), the third "E" would be the traditional "extinguish", not "exploit" as proposed in the OP.
I'm counting only monthly active users, for both sides. FediDB lists 1.2M for the Fediverse, your link lists 1.7M of them.
Their joining the fediverse will be more disruptive than their leaving it I think
Eternal September-like? It's possible.
They can pull it - most users in Threads will be interacting with other Threads users and content. Mastodon will be simply "that ideologically weird corner", and in practice they won't miss it.
For scale: Threads currently has 100M users. The Fediverse as a whole has 1.5M.
Damn, I was almost going to suggest some stuff. (chippym.uk - chippy, UK, a rodent...)
Serious now. I'm glad that federation means that people aren't putting all their eggs into the same basket; sure, feddit.uk going MIA is a loss, but just imagine if it was a non-federated platform. Hopefully the old users will be able to build their new home in the new instance.