This week’s newsletter describes a proposal for an alternative design
OP_VAULT with several benefits and announces a new weekly Optech
podcast. Also included are our regular sections with the summary of a
Bitcoin Core PR Review Club meeting, announcements of new software
releases and release candidates, and descriptions of notable changes to
popular Bitcoin infrastructure software.
● Alternative design for OP_VAULT: Greg Sanders posted to the Bitcoin-Dev mailing list an alternative design for providing the features of the
OP_UNVAULTproposal (see Newsletter #234). His alternative would add three opcodes instead of two. To provide an example:
● Alice deposits funds in a vault by paying a P2TR output with a script tree that contains at least two leafscripts, one which can trigger the time-delayed unvaulting process and one which can instantly freeze her funds, e.g.
● The trigger leafscript contains her less-trusted authorization conditions (such as requiring a signature from her hot wallet) and an
OP_TRIGGER_FORWARDopcode. At the time she creates this leafscript, she provides the opcode a spend delay parameter, e.g. a relative timelock of 1,000 blocks (about 1 week).
● The freeze leafscript contains any authorization conditions Alice wants to specify (including none at all) and an
OP_FORWARD_DESTINATIONopcode. At the time she creates this leafscript, she also chooses her more-trusted authorization conditions (such as requiring multiple signatures from multiple cold wallets and hardware signing devices). She provides the opcode a commitment to those conditions in the form of a hash digest.
● Alice triggers an unvaulting by spending the output received to the above script tree (using it as an input) and choosing the trigger leafscript. At this time, she provides two additional parameters to the
OP_TRIGGER_FORWARDopcode, the index of the output which will receive this input’s funds and a hash-based commitment to how she wants to be able to spend the funds later. The opcode verifies that the indicated output of this transaction pays a P2TR output with a script tree similar to the one being spent except that the trigger leafscript is replaced with a script using an
OP_CHECKSEQUENCEVERIFY(CSV) relative delay equal to the delay specified previously (e.g., 1000 blocks) and an
OP_FORWARD_OUTPUTSopcode which includes Alice’s commitment hash. The method of reconstructing the script tree is similar to an earlier covenant proposal,
OP_TAPLEAF_UPDATE_VERIFY(see Newsletter #166).
● Alice completes the unvaulting by waiting until the relative timelock has expired and then spending the unvaulting output, choosing the tapleaf with the
OP_FORWARD_OUTPUTSopcode. The opcode verifies that the spending transaction’s output amounts and script’s hash to the same commitment Alice made in the previous transaction. In this case, Alice has successfully deposited funds to a vault, begun an unvaulting, been forced to wait at least 1,000 blocks to allow her monitoring programs to verify she really did want to spend the funds to the specified outputs, and completed the spend.
● If something goes wrong, Alice freezes the funds. She can do this at any time from the moment she deposits funds in the vault up until an unvaulting is completed. To freeze funds, she simply chooses to spend the freeze leafscript from the output of either the vaulting or trigger transactions. Recall that Alice explicitly placed the freeze leafscript in the vaulting transaction, and note that it was implicitly carried over by the trigger transaction which initiated the unvaulting.
One of the advantages to users of this approach over the original
OP_VAULTdesign is that the freeze leafscript can contain any authorization conditions Alice wants to specify. In the
OP_VAULTproposal, anyone knowing the parameters chosen by Alice could spend her funds to the freeze script. That wasn’t a security problem but it could be annoying. In Sanders’s design, Alice could (for example) require a signature from a very lightly protected wallet in order to initiate a freeze—this would perhaps be enough of a burden to prevent most griefing attacks but not enough of a barrier to prevent Alice from quickly freezing her funds in an emergency.
Several other advantages are aimed at making the consensus-enforced vaulting protocol easier to understand and verify as safe. Subsequent to our writing the above, the author of the
OP_VAULTproposal, James O’Beirne, replied favorably to Sanders’s ideas. O’Beirne also had ideas for additional changes which we’ll describe in a future newsletter.
● New Optech Podcast: the weekly Optech Audio Recap hosted on Twitter Spaces is now available as a podcast. Each episode will be available on all popular podcast platforms and on the Optech website as a transcript. For more details, including why we think this is a major step forward in Optech’s mission to improve Bitcoin technical communication, please see our blog post.
Bitcoin Core PR Review Club
In this monthly section, we summarize a recent Bitcoin Core PR Review Club meeting, highlighting some of the important questions and answers. Click on a question below to see a summary of the answer from the meeting.
Bitcoin-inquisition: Activation logic for testing consensus changes is a PR by Anthony Towns that adds a new method for activating and deactivating soft forks in the Bitcoin Inquisition project, designed to be run on signet and used for testing. This project was covered in Newsletter #219.
Specifically, this PR replaces BIP9 block version bit semantics with what are called Heretical Deployments. In contrast to consensus and relay changes on mainnet – which are difficult and time-consuming to activate, requiring the careful building of (human) consensus and an elaborate soft fork activation mechanism – on a test network activating these changes can be streamlined. The PR also implements a way to deactivate changes that turn out to be buggy or undesired, which is a major departure from mainnet.
Why do we want to deploy consensus changes that aren’t merged into Bitcoin Core? What problems (if any) are there with merging the code into Bitcoin Core, and then testing it on signet afterward?
Several reasons were discussed. We can’t require mainnet users to upgrade the version of Core they’re running, so even after a bug has been fixed, some users may continue running the buggy version. Depending only on regtest makes integration testing third-party software more difficult. Merging consensus changes to a separate repository is much less risky than merging to Core; adding soft fork logic, even if not activated, may introduce bugs that affect existing behavior. ➚
Heretical Deployments move through a sequence of finite-state machine states similar to the BIP9 states (
FAILED), but with one additional state after
DEACTIVATING(following which is the final state,
ABANDONED). What is the purpose of the
It gives users a chance to withdraw funds they might have locked into the soft fork. Once the fork is deactivated or replaced, they might not be able to spend the funds at all – even if they’re anyone-can-spend; that doesn’t work if your tx is rejected for being non-standard. The concern isn’t so much the permanent loss of the limited signet funds, but rather that the UTXO set may become bloated. ➚
Why does the PR remove
We don’t need a configurable interval between lock-in and activation in the new state model – with Heretical Deployments, it activates automatically at the start of the next 432-block (3 days) state machine period (this period is fixed for Heretical Deployments). ➚
Why is Taproot buried in this PR?
If you didn’t bury it, you’d have to make it a Heretical Deployment, which requires some coding effort; also that would mean that it would timeout eventually, but we want Taproot never to timeout. ➚
Releases and release candidates
New releases and release candidates for popular Bitcoin infrastructure projects. Please consider upgrading to new releases or helping to test release candidates.
● Core Lightning 23.02 is a release for a new version of this popular LN implementation. It includes experimental support for peer storage of backup data (see Newsletter #238) and updates experimental support for dual funding and offers. Also included are several other improvements and bug fixes.
● BTCPay 1.8.2 is the latest release for this popular self-hosted payment processing software for Bitcoin. The release notes for version 1.8.0 say, “this version brings custom checkout forms, store branding options, a redesigned Point of Sale keypad view, new notification icons and address labeling.”
Notable code and documentation changes
Notable changes this week in Bitcoin Core, Core Lightning, Eclair, LDK, LND, libsecp256k1, Hardware Wallet Interface (HWI), Rust Bitcoin, BTCPay Server, BDK, Bitcoin Improvement Proposals (BIPs), and Lightning BOLTs.