I understand that the generic blockchain faces a problem where it can choose to optimize between scalability, security and decentralization. I can see easily why scalability and decentralization are at odds (for those who haven't seen it, a great blog post here explains this).

However, what does it mean to have a blockchain that is scalable and decentralized but insecure? No matter how I formulate the problem, I always end up at a compromise between block size/rate and decentralization.

  • A directed acyclic graph DAG accomplishes to be decentralized and scalable, but fails the finality.
    – Ismael
    Dec 24 '18 at 19:21
  • Maybe you could comment on how finality = security? Dec 24 '18 at 20:48
  • Security is protection against blockchain attacks like double spends. Finality is good against double spend attacks, but it is not the only attack possible. So not finality implies not security, but reverse is not true.
    – Ismael
    Dec 26 '18 at 0:00

Neither the hundreds of blogposts around the so-called Trilemma nor Zooko’s original triangle depict reality and are far from mathematical proofs.

Just like you, I cannot picture good cases, where security and decentralization are at odds and I would say in most cases they actually go hand-in-hand. If Bitcoin gets too centralized (mining pools) the risk of a double spend attack originating from one pool increases. If EOS’ block producers need to agree on blacklists via e-mail, it does not sound too secure either, but at least they’re scalable.

Eventually I would say it depends on the strict definition of those terms, in order for such triangle to be a meaningful concept. If security is exchanged with consistency, the storyline changes. And security itself is more of a base layer, because without security, do you have anything of value?

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.