We all know you insist on having the last word so I'm going to post this, and then ignore any follow up by you regarding this subject. I don't have the time to get into one of your back-and-forth minutiae arguments.
It's CLEAR that the Twitter CODE itself made TWO (or more) calls to the 2FA system, and only one got disabled (hence the problem). That's Twitter CODE, not the 2FA (3rd party most likely) server module. Nuanced, perhaps, but one is under the control of Twitter's devs, the other is just something loaded up onto the server to accept calls from Twitter's code.
We can agree that Elon should have not wholesale fired everyone that knew about the code, but there is also serious doubts about how helpful anyone that remained behind would have been.
That's a lot of typing that effectively says nothing
Elon on the other hand was much more clear-- he turned off SERVICES... that were running
on the server side
One of which was the one that sends out 2FA codes.
Thus breaking 2FA for users.
Which if he'd bothered to find out WHY a service was running before turning it off
server side would not have happened.
You appear to also be operating as he did, making assumptions about how things work without checking first.
It's CERTAINLY possible the service in question was poorly documented.
But turning off things when you're unsure what they do is
terrible practice.
Doubtless there's poorly labeled things you CAN get rid of. But there's almost certainly ones you can't- and until you're sure of the difference turning ANY of them off is likely to break things in an important way.
Which is exactly what happened.
That's one way to find out how something works: smash it into tiny pieces and see if you can put it back together again.
--Humpty Dumpty.
Sure.
But in a normal company you do that in a
test environment- not in production.
Now if you tell me Twitter doesn't HAVE a useful test environment, I wouldn't even be shocked.
But if they've decided to not bother to build one NOW and just keep randomly trying changes in production first, that's...not great.