Well, I said that I should post here...My primary objection to Ethereum is two-fold: first, the power of the smart contract language, and second, the fact that validation of the too-powerful smart contract language must be done onchain.
Given that the language is fully capable of supporting Layer 2 projects (and IIUC, such projects do exist), the problem there is more that the development and market culture suffers a sort of “everything on-chain!” brain-damage.
Arguably, there may be a place for a blockchain which
can execute complex contracts on-chain for use cases in which it is desirable for execution to occur when some parties are offline. (I am not saying that that’s Ethereum! I am not saying that “complex” means “Turing-complete”, either, though I take a somewhat nuanced view of that question.) The problem occurs when that’s the way that
everything is done, even when it makes no sense—which is
usually.
I think that the vast majority of “defi” use cases can have contracts executed fully off-chain, using the blockchain only as a synchronization and enforcement mechanism. It would result in much faster, more private, more efficient trades. Why not do it that way?
Compare and contrast (somewhat different concept, but very close):
http://langsec.org/occupy/I don’t think that a Turing-complete execution environment is
necessarily fatal. After all, if you use a non-Turing-complete smart-contract language, you are executing it on a Turing-complete computer at
some level; and for those willing to put serious CS expertise into writing contracts (
i.e., not average Solidity developers!), formal verification has made some great advances in recent years. Attempts to fix Ethereum’s notoriously weak security, such as KEVM, are interesting in that regard;
cf. IELE.
For a new project that is not constrained by backwards compatibility with a blockchain full of shitty code, I would hope that some concise explanations would be provided of how they address these issues.
Thanks. Will look into this...
Quick question: Why not aim for Spectrum/RGB? That’s a smart-contract platform which will support defi on Bitcoin. It will use the Simplicity language, which is designed from the ground up for security verification; and Spectrum on Layer 3 (atop Lightning) will have the privacy, speed, and efficiency advantages of off-chain contract execution.
Overall, I would suggest that if you want to gain people’s trust, do fewer manifestos (preaching to the choir here) and more informative explanations about just how this works, what useful properties the system has, and why it should be trusted to not fail catastrophically.