This week’s newsletter summarizes a paper about the PoWswap protocol and includes our regular sections with the summary of a Bitcoin Core PR Review Club meeting, announcements of new releases and release candidates, and descriptions of notable changes to popular Bitcoin infrastructure software. Also included is a short section celebrating five years of Bitcoin Optech and our 250th newsletter.

News

  • Paper about PoWswap protocol: Thomas Hartman posted to the Bitcoin-Dev mailing list about a paper he has written with Gleb Naumenko and Antoine Riard about the PoWSwap protocol first proposed by Jeremy Rubin. Powswap allows the creation of onchain-enforceable contracts related to the change in hash rate. The basic idea takes advantage of the protocol-enforced relationship between time and block production plus the ability to express time locks in either time or blocks. For example, consider the following script:

    OP_IF
      <Alice's key> OP_CHECKSIGVERIFY <time> OP_CHECKLOCKTIMEVERIFY
    OP_ELSE
      <Bob's key> OP_CHECKSIGVERIFY <height> OP_CHECKLOCKTIMEVERIFY
    OP_ENDIF
    

    Let’s imagine the current time is t and the current block height is x. If blocks are produced an average of 10 minutes apart, then if we set <time> to t + 1000 minutes and <height> to x + 50, we’d expect that Bob would be able to spend the output controlled by the above script on average 500 minutes before Alice can spend it. However, if the rate of block production were to suddenly more than double, Alice might be able to spend the output before Bob.

    There are several envisioned applications of this type of contract:

    • Hashrate increase insurance: miners must purchase their equipment before they know for certain how much income it will generate. For example, a miner who purchases enough equipment to receive 1% of the network’s current total of rewards might be surprised to find that other miners also purchased enough equipment to double the total network hashrate, leaving the miner with 0.5% of the reward instead of 1%. With PoWSwap, the miner can make a trustless contract with someone who is willing to pay the miner if hashrate increases before a certain date, offsetting the miner’s unexpectedly low income. In exchange, the miner pays that person an upfront premium or agrees to pay them a larger amount if the network-wide hashrate stays the same or decreases.

    • Hashrate decrease insurance: a wide variety of problems with Bitcoin would result in a significant decrease in network-wide hashrate. Hashrate would decrease if miners were being shut down by powerful parties, or if a significant amount of fee sniping began to suddenly occur among established miners, or if the value of BTC to miners suddenly decreased. Holders of BTC who wanted to insure against such situations could enter trustless contracts with miners or third parties.

    • Exchange rate contracts: in general, if the purchasing power of BTC increases, miners are willing to increase the amount of hashrate they provide (to increase the rewards they receive). If purchasing power decreases, hashrate decreases. Many people may be interested in trustless contracts related to the future purchasing power of Bitcoin.

    Although the idea for PoWSwap has been circulated for several years, the paper provides more detail and analysis than we’ve previously seen.

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.

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.

Add getprioritisationmap, delete a mapDeltas entry when delta==0 is a PR by Gloria Zhao (glozow) that improves the Bitcoin Core feature that allows miners to modify the effective mempool fee, and thus the mining priority (higher or lower), of particular transactions (see the prioritisetransaction RPC). The fee increment (if positive) or decrement (if negative) is called the fee delta. Transaction prioritization values are persisted to disk within the mempool.dat file and are restored on node restart.

One reason a miner might prioritize a transaction is to account for an out-of-band transaction fee payment; the affected transaction will be treated as if it has a higher fee when choosing which transactions to include in the miner’s block template.

The PR adds a new RPC, getprioritisationmap, that returns the set of prioritized transactions. The PR also removes unnecessary prioritization entries, which can arise if the user sets deltas back to zero.

  • What is the mapDeltas data structure, and why is it needed?

    It’s where the per-transaction prioritization values are stored. These values affect local mining and eviction decisions, as well as the ancestor and descendant feerate calculations. 

  • Do transaction prioritizations affect the fee estimation algorithm?

    No. Fee estimation needs to accurately predict the expected decisions of miners (in this case, other miners), and these miners don’t have the same prioritizations that we do, since those are local. 

  • How is an entry added to mapDeltas? When is it removed?

    It’s added by the prioritisetransaction RPC, and also when the node restarts, due to an entry in mempool.dat. They are removed when a block containing the transaction is added to the chain, or when the transaction is replaced

  • Why shouldn’t we delete a transaction’s entry from mapDeltas when it leaves the mempool (because, for example, it has expired or been evicted due to feerate dropping below the minimum feerate)?

    The transaction may come back into the mempool. If its mapDeltas entry had been removed, the user would have to re-prioritize the transaction. 

  • If a transaction is removed from mapDeltas because it’s included in a block, but then the block is re-orged out, won’t the transaction’s priority have to be reestablished?

    Yes, but reorgs are expected to be rare. Also, the out-of-band payment may actually be in the form of a Bitcoin transaction, and so it may need to be redone as well. 

  • Why should we allow prioritizing a transaction that isn’t present in the mempool?

    Because the transaction may not be in the mempool yet. And it may not even be able to enter the mempool in the first place on its own fee (without the prioritization). 

  • What is the problem if we don’t clean up mapDeltas?

    The main problem is wasteful memory allocation. 

  • When is mempool.dat (including mapDeltas) written from memory to disk?

    On clean shutdown, and by running the savemempool RPC. 

  • Without the PR, how do miners clean up mapDeltas (that is, remove entries with a zero prioritization value)?

    The only way is to restart the node, since zero-value prioritizations are not loaded into mapDeltas during restart. With the PR, the mapDeltas entry is deleted as soon as its value is set to zero. This has the additional beneficial effect that zero-value prioritizations aren’t written to disk. 

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), Lightning BOLTs, and Bitcoin Inquisition.

  • Bitcoin Core #26094 adds the block hash and height fields to getbalances, gettransaction, and getwalletinfo. These RPC calls lock the chainstate to make sure they are up-to-date with the latest block and so they benefit from including the valid block hash and height in the response.

  • Bitcoin Core #27195 makes it possible to remove all external receivers from a transaction that is being replaced using the bumpfee RPC from Bitcoin Core’s internal wallet. The user does this by making the only output of the replacement transaction pay the user’s own address. If the replacement transaction gets confirmed, this prevents any of the original receivers from being paid, which is sometimes described as “canceling” a Bitcoin payment.

  • Eclair #1783 adds a cpfpbumpfees API for CPFP fee bumping one or more transactions. The PR also updates the list of recommended parameters for running Bitcoin Core to ensure that creating a fee-bump transaction is a viable option.

  • LND #7568 adds the ability to define additional LN feature bits when the node is started up. It also removes the ability to disable any hardcoded or defined feature bits during runtime (but additional bits may be still be added and later disabled). A related proposal update in BLIPs #24 notes that custom BOLT11 feature bits are limited to a maximum expressed value of 5114.

  • LDK #2044 makes several changes to LDK’s route hinting for BOLT11 invoices, the mechanism that a receiving LN node can use to suggest routes for a spending node to use. With this merge, only three channels are suggested, support for LDK’s phantom nodes is improved (see Newsletter #188), and the three channels chosen are selected for efficiency and privacy. The PR discussion includes several insightful comments about the implications for privacy of providing route hints.

Celebrating Optech Newsletter #250

Bitcoin Optech was founded, in part, to “help facilitate improved relations between businesses and the open source community.” This weekly newsletter was started to give executives and developers inside Bitcoin-using businesses more insight into what the open source community was building. As such, we initially focused on documenting work that might affect businesses.

We quickly discovered that not just business readers were interested in this information. Many contributors to Bitcoin projects didn’t have the time to read all the discussions on the protocol development mailing lists or to monitor other projects for major changes. They appreciated someone notifying them about developments that they might find interesting or which might affect their work.

For almost five years now, it’s been our pleasure to provide that service. We’ve tried to expand on that simple mission by also providing a guide to wallet technology compatibility, an index to over 100 topics of interest, and a weekly discussion podcast with guests that have included many of the contributors whose work we’ve been privileged to write about.

None of that would be possible without our many contributors, who in the past year have included: Adam Jonas, Copinmalin, David A. Harding, Gloria Zhao, Jiri Jakes, Jon Atack, Larry Ruane, Mark “Murch” Erhardt, Mike Schmidt, nechteme, Patrick Schwegler, Shashwat Vangani, Shigeyuki Azuchi, Vojtěch Strnad, Zhiwei “Jeffrey” Hu, and several others who made special contributions to particular subjects.

We also remain eternally grateful to our founding sponsors Wences Casares, John Pfeffer, and Alex Morcos, as well as our many financial supporters.

Thank you for reading. We hope that you’ll continue to do so as we publish the next 250 newsletters.