Jarvis and Friday: STARK-Friendly Cryptographic Primitives

Research
1

New research to be presented tomorrow about STARK friendly cryptographic primitives. Read more here: https://www.esat.kuleuven.be/cosic/jarvis-and-friday-stark-friendly-cryptographic-primitives/

1 Like
2

Did you hear the joke about the UFO?

Nevermind, it’s way over your head.

3

I saw this too! While I don’t fully understand the crypto, yet, from an implementation standpoint it seems like STARKS allow public blockchains (esp L2 platforms) to aggregate tx in a way that would only require publishing a proof of validity and the final tx vs making all tx public knowledge. Lots of talk on this on the Ethereum Research forum. Seems like with an L2 solution (Plasma, Raiden, whatever) have a similar effect to cut-through, with various pros and cons

  • pro: STARKS generalize to various usecases that don’t require a full blockchain, for publishing tx as well as anonymizing any computation performed
  • con: whoever computes the STARK would be able to log all the data inputs (although maybe some of the work in MPC or HE could help?)

I dunno, what were your thoughts on STARKS and did you see similarities to the protocol of MW/Grin?

4

Yes, more broadly the idea is to pass around proofs of valid state-transitions instead of having the whole network re-execute these.