X

[X @big_s0sa_]

@EdFelten Agreed. Where do you think ZKRs fit into the future? Rollups that need instant composability?

[X @big_s0sa_]

@terorie_dev @bqbrady Totally. When execution becomes the bottleneck this is why ZK rollups will probably have much lower throughput that ORs because ZK proving is a large constant to superlinear discount on the underlying execution itself. Thus the throughput will be discounted as well.

[X @big_s0sa_]

@bqbrady @dystopiabreaker Agreed! VMs built for parallel execution might be able to outcompete a parallel EVM on needed throughput (although they have usually compromised on dev experience). Do we yet have estimates on how parallelizable is the EVM is?

[X @big_s0sa_]

@sreeramkannan @apolynya Using this logic, seems like DataLayr could be used as an efficient means of storing all the data involved in stateless validation further plummeting requirements on OR watchers

[X @big_s0sa_]

@sreeramkannan @apolynya Yes but as blocks start expiring they need honest parties to store state in order for whatever merkle proofs (or other) to be assembled when users want to withdraw

[X @big_s0sa_]

@sreeramkannan @apolynya Why must this data be made available? Doesn’t the honest watcher assumption imply that an honest party has possession of the state? Or is this a means of guaranteeing such a property for a zk rollup when state becomes too large?

[X @big_s0sa_]

@toghrulmaharram So, ideally, any vm would allow trustless access to them :)

[X @big_s0sa_]

@toghrulmaharram I also failed to mention that blockhashes (headers) on one chain are used on other chains for light client bridges. The blockhash opcode is how you would commit to the header on the sending chain in a nomad style bridge. State and tx commitments enable a host of powerful apps.

[X @big_s0sa_]

@toghrulmaharram Not that I know of, but a rollup might use something like that for proving forced inclusion. As for your second point, you only need to reveal the branch at which your desired state/tx lies, or am I missing the point?

[X @big_s0sa_]

@toghrulmaharram A couple that I’m working on :) But here’s Lido’s curve-merkle-oracle for retroactively proving the stETH-ETH price though MPT proofs of state https://t.co/0G7izxrcyj

[X @big_s0sa_]

@toghrulmaharram Glad to hear! Any numbers on what TPS the prover can handle?

[X @big_s0sa_]

@aztecnetwork Ayyyy, now it’s just that sweet sweet DA :)

[X @big_s0sa_]

@aztecnetwork Don’t think anyones beating you to the race on privacy :) Appreciate the transparency

[X @big_s0sa_]

@vdWijden Composability. You can use assets from one protocol to interact with another (e.g. depositing LP tokens into a farm, fractionalizing NFTs and putting them into liquidity pools, etc.). Also liquid staking. Not sure there are equivalents for these in CeFI.

[X @big_s0sa_]

@epolynya 7 day fraud proofs are not due to lack of awareness of fraud, honest watchers will always be verifying (OR assumption). It is to make it very expensive to censor a fraud proof tx during the interval = gas to fill up N blocks. With EIP1559 it gets pretty expensive, but still.

[X @big_s0sa_]

@coingecko why is the price of usdt so different on coigecko and coinmarketcap? weird

[X @big_s0sa_]

@toghrulmaharram Tru. This wouldn't happen on Avalanche because it's MEV resistant. Ideally fair ordering could be solved at the protocol layer. My hunch is that you can't get cryptoeconomically assured fair ordering bc it's not verifiable. I think it requires honest majority. Maybe that's fine.