> i see so a minimal change to taproot would be a new version: “script spend only” or similar. BIP 360's P2MR (Pay-to-Merkle-Root) is P2TR but script spend only via a simple and minimal change. Thanks for working on commit-reveal -- also thanks to Tadge for lifeboat, I think this line of research is worthwhile and helpful. Correct me if you disagree, I think of commit-reveal as useful for the following case: Q-day happens AND we have P2MR AND OP_TXHASH, but not OP_SLH_CHECKSIG, commit-reveal might be the best option to migrate to a PQ signature scheme. I am pursuing the P2MR + OP_SLH_CHECKSIG approach this approach is the least disruptive to how people use Bitcoin today and would be simpler for wallets to support (SLH only requires signing a single transaction vs. two transactions and having to handle griefing/reset). In both cases we want BIP 360's P2MR. I'd also like OP_TXHASH and OP_SLH_CHECKSIG as well. I'm using OP_SLH_CHECKSIG as a placeholder here, if you like OP_SHRINCS_CHECKSIG better, you could also do P2MR + OP_SHRINCS_CHECKSIG On Wed, Feb 11, 2026 at 2:25 AM Erik Aronesty wrote: > i see so a minimal change to taproot would be a new version: “script spend > only” or similar. > > On Tue, Feb 10, 2026 at 6:41 PM Ethan Heilman wrote: > >> >> You'd still need BIP 360 P2MR (or P2TRD) since OP_TXHASH needs >> tapscript >> > false, covenant based multistep secret-reveal spending paths don't >> rely on signatures at all >> >> P2TR has a public key baked into it via the key spend path. This key path >> spend bypasses any covenant or script constraints. The attacker does not >> need to see a value spend to break the key path spend. This means a quantum >> attacker can break the nums point in key spend path of the initial output >> and the AnchorPublishTx output and then just taking all the coins [0]. >> >> > agreed. they have to spend resources to attack your private key and >> the only thing they can do is "grief" using a timing attack with the >> results, rather than steal outright. a massive incentive difference. >> >> Ok, so a core assumption you are making here is that a CRQC isn't >> powerful enough for recovering signing keys to be effectively free. This is >> likely to be true at the early stages of CRQC, but this assumption may not >> hold forever. If ECC is mathematically broken via a classical attack this >> assumption might not hold at all. I'm attempting to address both quantum >> and classical breaks. >> >> > TX_HASH is simple and generally useful and there is no guarantee that >> q-day will even come >> >> TX_HASH is great! >> >> [0]: As originally noted here: >> https://delvingbitcoin.org/t/a-quantum-resistance-script-only-using-op-ctv-op-txhash-and-no-new-signatures/2168/4 >> >> On Tue, Feb 10, 2026 at 7:19 PM Erik Aronesty wrote: >> >>> >>>> >>>> You'd still need BIP 360 P2MR (or P2TRD) since OP_TXHASH needs >>>> tapscript, and the only available tapscript supporting output type, P2TR, >>>> isn't quantum safe. >>>> >>> >>> false, covenant based multistep secret-reveal spending paths don't rely >>> on signatures at all >>> >>> >>>> >>>> I'm going to assume: >>>> - you mean to use this commit-reveal for migrating between signature >>>> algorithms, not for everyday use, >>>> >>> >>> it can be used if "q day" happens. otherwise ignored. >>> >>> >>>> - TXHASH is being used because you are waiting for the commitment to be >>>> confirmed on-chain rather than lifeboat's out-of-band commitment system >>>> >>> >>> it's used so you can commit to a spending constraint without committing >>> to the final "as yet to be determined" quantum-safe destination: >>> https://delvingbitcoin.org/t/a-quantum-resistance-script-only-using-op-ctv-op-txhash-and-no-new-signatures/2168 >>> >>> >>> >>>> Once you post your commit-txn, but before it confirms, other parties >>>> can post competing commit-txns that double spend your output. If one of >>>> malicious transactions confirm, you must now wait for a timelock to expire >>>> and then try to post your transaction. >>>> >>> >>> agreed. they have to spend resources to attack your private key and the >>> only thing they can do is "grief" using a timing attack with the results, >>> rather than steal outright. a massive incentive difference. >>> >>> >>>> They can block you again. Each time they burn some of you coins in >>>> fees. Miners get the fees, so they might be incentivized to do this. Thus, >>>> we must trust miners not to do this. Lifeboat doesn't have this issue since >>>> it uses out-of-band commitments, but out-of-band commitments have their own >>>> issues. >>>> >>> >>> each time you use the reset-path, they have to re-attack a new key. >>> very expensive just to grief a small amount of fees spread across all >>> miners. sounds like science-fiction levels of compute. >>> >>> >>> plus.... TX_HASH is simple and generally useful and there is no >>> guarantee that q-day will even come >>> >> -- You received this message because you are subscribed to the Google Groups "Bitcoin Development Mailing List" group. To unsubscribe from this group and stop receiving emails from it, send an email to bitcoindev+unsubscribe@googlegroups.com. To view this discussion visit https://groups.google.com/d/msgid/bitcoindev/CAEM%3Dy%2BUNum_%3D1Me0Q%3DfBc7oS6s4i_GwZEw1Ad60S_efHOVnqMw%40mail.gmail.com.