/ home / newsletters /
Bitcoin Optech Newsletter #135
This week’s newsletter links to the summary of last week’s taproot activation meeting and announces another scheduled meeting for next week, plus it describes recent progress in discreet log contracts and a new mailing list for discussing them. Also included are our regular sections with the summary of a Bitcoin Core PR Review Club meeting, descriptions of releases and release candidates, and a list of notable changes to popular Bitcoin infrastructure projects.
● Taproot activation meeting summary and follow-up: Michael Folkson posted a summary of the scheduled meeting to discuss taproot activation methods (see Newsletter #133). There appeared to be support among participants for using the BIP8 activation mechanism with the earliest allowed activation time about two months after the first release of Bitcoin Core containing the activation code and the latest allowed activation of that initial deployment about one year subsequently.
More controversial was whether the
LockinOnTimeout(LOT) parameter should default to true (requiring miners to either eventually signal in favor of the new rules or risk a chainsplit) or false (allowing miners to signal however they’d like without immediate consequences, though some users could choose to enable
LOT=truelater). Folkson wrote a second post to the mailing list summarizing the arguments he’s seen for the two different options and announcing a follow-up meeting to discuss them (and some less controversial issues) in the Freenode ##taproot-activation channel on February 16th at 19:00 UTC.
● New mailing list for Discreet Log Contracts: Nadav Kohen announced the creation of a new mailing list for discussion of topics related to Discreet Log Contracts (DLCs). He also summarized the current state of development on DLCs, including multiple compatible implementations, the use of ECDSA signature adaptors, effective compression algorithms for DLCs contingent on numeric outcomes, and support for k-of-n threshold signing from oracles “even supporting numeric cases where some bounded difference is permitted between oracles”.
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.
Only support compact blocks with witnesses is a PR (#20799) by John Newbery that proposes to remove support for the non-segwit version of BIP152 compact blocks.
The review club discussion focused on understanding compact blocks, high versus low bandwidth modes, and versioning use and compatibility before diving into the changes in the code.
What are compact blocks?
Compact blocks, specified in BIP152, are a way to relay blocks across the Bitcoin peer-to-peer network with less bandwidth use. They take advantage of the fact that during transaction propagation, peers have already seen most of the transactions contained in the received block. Compact blocks can be relayed either in high bandwidth mode or low bandwidth mode. When relayed in high bandwidth mode, compact blocks can also speed up block propagation.
What is required for compact blocks to work?
The receiver must have a mempool containing transactions that are likely to be included in the block. Compact blocks are therefore only useful for relaying blocks at or near the tip of the block chain. For older blocks, the receiver won’t have the transactions in its mempool and would need to request them using a
getblocktxnmessage. In such cases, just requesting the full block is more efficient.
How do compact blocks save bandwidth?
Rather than using full transaction ids, compact blocks contain short ids that are smaller in size but sufficient to uniquely identify transactions. A node receiving a compact block matches each short id with a transaction in its mempool to reconstruct the full block. This greatly reduces block relay bandwidth. ➚
What is the difference between “high bandwidth” and “low bandwidth” modes?
In high bandwidth mode, compact blocks are sent unsolicited, trading higher bandwidth use for improved latency, whereas in low bandwidth mode they are sent only on request after an
headersmessage was received. In high bandwidth mode, compact block messages can be relayed prior to full validation, requiring only that the block header be valid before relay. ➚
How do we choose which peers are in high bandwidth mode?
We select up to three peers that recently sent us a novel valid block. This logic is in the net processing function,
Why can we stop supporting version 1 compact blocks?
BIP152 supports two versions: version 1 (without witnesses) and version 2 (with witnesses). Version 2 is required to reconstruct segwit blocks. Segwit was activated in August 2017, so providing version 1 pre-segwit compact blocks to peers is no longer useful; without the witnesses, peers are unable to verify that the blocks follow the consensus rules. ➚
Practically speaking, how can version 1 support be dropped?
By ignoring incoming
sendcmpctmessages with version equal to 1, no longer sending
sendcmpctmessages that signal version 1 support, only sending
NODE_WITNESSpeers, and responding to
blocktxnmessages with witness-serialized blocks and transactions.
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.
- ● LND 0.12.1-beta.rc1 is a release candidate for a maintenance release of LND. It fixes an edge case that could lead to accidental channel closure and a bug that could cause some payments to fail unnecessarily, plus some other minor improvements and bug fixes.
Notable code and documentation changes
Notable changes this week in Bitcoin Core, C-Lightning, Eclair, LND, Rust-Lightning, libsecp256k1, Hardware Wallet Interface (HWI), Rust Bitcoin, BTCPay Server, Bitcoin Improvement Proposals (BIPs), and Lightning BOLTs.
● Bitcoin Core #19509 adds per-peer message capture between nodes as well as the ability to produce JSON outputs from those logs. Using the newly introduced command line argument
-capturemessages, any message that the node sends or receives is logged. There is a long history of tools with similar functionality, but this addition provides a native alternative to less actively maintained options.
● Bitcoin Core #20764 adds additional information to the output produced using
bitcoin-cli -netinfo. New details include the type of peer (e.g. full relay or blocks-only), number of manually added peers, peers using BIP152 “high bandwidth” (faster) block relay mode, and any peers using the I2P anonymity network (being worked on in a separate PR).
● Rust-Lightning #774 adds support for fetching blocks and headers from Bitcoin Core’s REST and RPC interfaces. Additionally, a
BlockSourceinterface is provided and can be extended to work with custom sources.
● HWI #433 adds support for signing PSBTs with OP_RETURN outputs.
● BIPs #1021 updates the BIP8 soft fork activation mechanism, changing its behavior for nodes that choose to enforce the lockin on timeout feature. Previously, those nodes would reject any block that didn’t signal for activation during the mandatory activation period. After this change, the nodes will tolerate up to the maximum number of non-signaling blocks possible that still guarantee activation. This reduces the number of blocks that may be needlessly rejected and reduces the potential for miscommunication between miners and node operators.
● BIPs #1020 updates BIP8 to no longer require signaling during the locked in period now that activation can be assured using required signaling in the recently added must signal period.
● BIPs #1048 rewrites most of the BIP322 proposal for generic message signing as proposed a few weeks ago on the mailing list (see Newsletter #130). The changes allow wallets that don’t implement a complete set of checks to return an “inconclusive” state for signatures that use scripts they don’t understand. It also clarifies the implementation instructions and removes unnecessary data from the signature serialization.
● BIPs #1056 adds BIP350 with a specification of bech32 modified (bech32m) as previously discussed on the mailing list (see Newsletter #131). This modified version of bech32 addresses from BIP173 will apply to planned taproot addresses and later improvements using segwit witness scripts.
● BIPs #988 updates the BIP174 specification for PSBTs to require programs operating in the creator role to initialize empty output fields, similar to an existing requirement to initialize empty input fields. Existing PSBT creators such as Bitcoin Core were already doing this.
● BIPs #1055 updates BIP174 to clarify the format for proprietary extensions, extends the table of fields with rows indicating how they apply to different versions of PSBTs, and marks the BIP as final with regard to the original version 0 of PSBTs.
● BIPs #1040 updates the BIP85 specification for creating keys and keychains from a super BIP32 keychain. The updates describe how to use the super keychain to create keys for RSA-based PGP signatures that can be loaded on to GPG-compatible smartcards.
● BIPs #1054 updates the BIP141 specification of segwit’s consensus changes to clarify how the number of signature operations (sigops) is counted when
OP_CHECKMULTISIGVERIFYare used. The text previously noted that they were counted the same as when P2SH was used but the update describes the unintuitive particulars: “For 1 to 16 total public keys CHECKMULTISIG is counted as 1 to 16 sigops respectively, and for 17 to 20 total public keys it is counted as 20 sigops.”
● BIPs #1047 updates the BIP39 specification for deterministically generating BIP32 seeds from a phrase, adding a warning that the use of non-English word lists is not widely supported and so is not recommended for implementation.