It's better to let them them leak my throwaway email and 9IB2?iDnDaFt0cf6jwXQp_ rather than accessing my tweets, or even my twitter id.
We've been doing passwords on the web since before 1999 and we weren't storing them in plaintext even then. I bet there is a proven authentication framework they could use in this project.
https was there in 1999 too but luckily we must use the current one, which is more secure.
Implying Twitter won't just hand it over to governments.
"Governments" aren't anywhere near the complete scope of risk you bring on yourself by storing passwords badly.
I'd rather do that than risk my identity being compromised in a place like Iran where my life is at stake if I speak out against the government..
I understand, but maybe if you're blogging for the purpose of political dissidence you could look elsewhere? Home-grown authentication adds a lot of overhead to web apps, so it makes sense to hand that off to one or more standardized identity management systems.