An old thread from me back in 2018 on the first five years of ethereum proof of stake research:

An old thread from me back in 2018 on the first five years of ethereum proof of stake research:

https://twitter.com/VitalikButerin/status/1029900695925706753

What was the main reason for choosing FFG in the end? Is it what you alluded to at the end - despite the better theoretical properties of CBC, FFG was the easier implementation?

Building ENS since forever.

Interested in identity, cryptography, climbing, health.

CBC always had all kinds of complexity that made it hard to work with a very high node count.

Do you think more research would have eventually lowered the cost? The last thing I did on it was work on a more efficient confirmation rule for instance, and it doesn’t feel like we were hitting fundamental limits yet

Interesting, so in a way CBC would become more centralised if it couldn’t be fixed and you had to limit node count.