Unless Tony changed it since I discussed it with him in his official thread in Q4 2016, in my opinion Byteball has at least four major flaws:

1. Afair, the transaction fee mechanism doesn't scale properly with the appreciation in the price of bytes. I don't remember all the details, but this seemed to me at the time to totally break the design going forward.

2. My analysis is the 12 witnesses can't realistically be changed without a hard fork (the algorithm for doing so won't work in reality unless the whales coordinate to make it so, i.e. not decentralized) if 50% of them stop functioning or colluding. In other words, another CartelCoin or hardforking chaos in the making.

3. Afaik, the transaction confirmations are fast but not sub-second (because you have to wait for 7 of the 12 witnesses to sign a new stability point tip), thus not optimal for apps that need very low-latency for onchain actions. Steem's (Bitshares') Graphene (DPoS) has I believe faster 1-confirmations (due to ability of whales to monitor and replace witnesses at will and because each witness produces its own block) but also not sub-second (especially if security against orphans requires more confirmations). Graphene has no asynchrony, so I am not arguing it is better than Byteball. But neither of them met my stringent design requirements. Byteball is interesting, which is why I dedicated an entire section of my whitepaper to discuss its stability point algorithm. I applaud Tony on his clever DAG consensus algorithm.

4. Afaics, his distribution model totally ruined any chance for a funding model to onboard the app developers, content providers, and users. This is really the killer mistake.

Also I don't think the use of JSON as a smart contract language is any tremendous innovation. He appears to be putting a lot of energy into that, so that is an entirely different focus from mine. One of the experiments I am working on is I am trying to create a better programming language for programming apps (a statically typed derivative of JavaScript that transpiles initially to TypeScript), not just smart contracts.


You don't see me going around in every Byteball thread bashing it. I will not do that. Good luck with it.

Btw, the Byteball logo/avatar is quite unique. Spartan (a la Google) and grabs the eye (personally it makes me think of Hal's camera eye in Space Odyssey).