I wish there was an exam you could take that would get you a "computing license" that you could show to bypass a lot of this stuff.
I wish there was an exam you could take that would get you a "computing license" that you could show to bypass a lot of this stuff.
With 2FA, almost all of the traditional attacks are impossible and the only one left is specially crafted fake login pages.
You're right when you're dealing with large groups of people which is exactly what I said in my original comment. If you understand these issues passwords on their own are fine. 2FA has some nasty edge cases that make it very easy to lose access to your account (which is what everyone in this thread is talking about.) So as always you're balancing security and availability. For some people the proper balance means just password authentication makes sense.
The best IMO would be client side certificates but no one except maybe Github supports those.
I don't claim 2FA is perfect, it has its own host of issues as is being discussed. But it isn't just for "the masses that don't understand how to make strong passwords or how to be secure". Certs are good, but they need to be made simpler than they are now for the very people you claim 2FA is for or they won't be adopted (one of the reasons they aren't more ubiquitous, IMO). And some certs implementations have issues with account recovery as well if you lose access to them.
It's a really hard problem that you do a bit of disservice to when you dismiss it as "just a problem of the plebes".
Any service that is storing passwords in plain text is probably not implementing 2FA in an safe way, either.
Can you explain why that is an issue?
Can you elaborate on the specific risk here?