Might be worth it to push through anyways. Twitter had its fail whales back in the day, I think a couple of temporary outages caused by demand might even be a good thing and add to the hype. Absence makes the heart grow fonder doesn't it? Seems silly to trade insane growth for perfectionism instead.

> The Bluesky infra would not be able to handle scale without invite-only clownshow. Meta can open the floodgates without breaking a sweat.

Honest question: are they emulating "classical Twitter" to the extent of going with self-hosted inefficient Rails implementation? Or what is their problem with scaling exactly in 2023 beyond ordering more instances in some AWS gui?

And as many of us have observed, the key isn't just the app or UI, it's that there are people there. With Threads, people are indeed there. Like it or not I think this means Bluesky is dead-on-didn't-arrive.

Is it any surprise Facebook got this right? Understanding "it's cool because there are people there" is part of their origin story, after all.

isn't Bluesky supposed to run on a federated protocol? They could just open source their reference implementation and let people run their own instances. Not even the protocol itself seems to be public.

That assumes that if open registration was enabled, people would be banging down the doors to create an account. Very unlikely unless Twitter straight up shuts down.

The threads app can pull in millions in a day because they can login with Instagram SSO, zero friction.