Roll up centric paradigm was chosen not because it works but because it’s hard to update existing mainnet. 

Ps pls share your favourite articles on this topic, I’m interested reading about any position.

Alternatively, rollup centric paradigm was chosen because plasma centric paradigm didn’t work (or was too early)

The history of plasma is an important read regardless:

Full-time builder. @pimlico | @33bits | @66bits | farcaster-solidity | x.com/potekhin_sergey. Zero-knowledge and asynchronous transactions evangelist.

Alternatively, rollup centric paradigm was chosen because plasma centric paradigm didn’t work (or was too early)

The history of plasma is an important read regardless:

https://haseebq.com/the-life-and-death-of-plasma/

Oh what a memories, thank for sharing! Great article

> Thus, all exits can only be processed after the challenge period is over, which takes up to 1 week in some proposals.

Another reminder for me on how important to read such retrospective articles. A system does not have to be perfect, many things that look like a problems, can be safely classified as an attributes or TBF at worse.