Shanil Samarasinghe

wholeheartedly agree with point (1): data manipulation aka the oracle problem is overlooked, is considered by crypto academia and crypto VC a second-order problem, whereas in truth it is as challenging to solve compared to computation validity, the voodoo math of zkp notwithstanding, since there's no refutability/accountability: consider a rollup with a centralized entity operating the rollup<>L1 bridge. even in the extreme setup where the rollup's state can be fully misreported and arbitrarily declared by the operator, with no zk proofs or fraud proofs, the operator still has only a single chance to cheat---it is easy for everyone to see that the operator misreported the state and to abandon it forever. (which is arguably a sufficient guarantee in many sub 1B$ contexts. BTW, in such contexts many programming langs could do the job, eg python, and there's no need for the heavy lifting with a consensus-specialized VM. @JonaFunLevi note) compared to a zk-verified and governed bridge with a centralized oracle for data, eg price feeds -- it is impossible to hold the operator accountable to misreports of data, since outdated prices can always be attributed to actual or feigned internet latency, and the two options are indistinguishable (in part., verifiable attestations are irrelevant). i propose one route to address this, namely, to revert again to the PoW assumption of honest majority, and to design estimator mechanisms among miners (in the spirit of median or Huber's estimator, cf. Huber contamination model) that will output sufficiently verifiable input/data, @elimmea pls complete our paper :) admittedly honest majority is a weaker model compared to zkp's, but then again neither zkp nor tee contribute towards addressing the oracle problem finally, for the PoW route to be feasible, you must sample many miners frequently, at each internet RTT, which is one primary motivation to go beyond 10 blocks per second with an appropriate PoW-DAG protocol (@Avivz78 I think you might be pleasantly surprised by the value of your DAGs here <3)

The first kaspaweekly.com release is out! Featuring 👨‍💻 Progress on the Crescendo network upgrade 📔 Progress on the Igra Labs L2 🌏 Kaspa at the Crypto Expo Europe event in Romania 🧠 Kaspa's Eleven 📱 New Wallet Listing Kaspa Happy reading! Didn't subscribe yet? You can always read it at the archive (newsletter.kaspaweekly.com/archive/kaspa-…) Didn't get it? You likely haven't confirmed your subscription (you should have gotten an email to confirm after registering on the website, if not drop me a dm)