This week’s newsletter includes action items related to the security release of Bitcoin Core 0.16.3 and Bitcoin Core 0.17RC4, the newly-proposed BIP322, and Optech’s upcoming Paris workshop; a link to the C-Lightning 0.6.1 release, more information about BIP322, and some details about the Bustapay proposal; plus brief descriptions of notable merges in popular Bitcoin infrastructure projects.
● Upgrade to Bitcoin Core 0.16.3 to fix denial-of-service vulnerability: a bug introduced in Bitcoin Core 0.14.0 and affecting all subsequent versions through to 0.16.2 will cause Bitcoin Core to crash when attempting to validate a block containing a transaction that attempts to spend the same input twice. Such blocks would be invalid and so can only be created by miners willing to lose the allowed income from having created a block (at least 12.5 XBT or $80,000 USD).
Patches for master and 0.16 branches were submitted for public review yesterday, the 0.16.3 release has been tagged containing the patch, and binaries will be available for download as soon as a sufficient number of well-known contributors have reproduced the deterministic build—probably later today (Tuesday). Immediate upgrade is highly recommended.
● Allocate time to test Bitcoin Core 0.17RC4: Bitcoin Core will soon be uploading binaries for 0.17 Release Candidate (RC) 4 containing the same patch for the DoS vulnerability described above. All testers of previous release candidates should upgrade. Testing is greatly appreciated and can help ensure the quality of the final release.
● Review proposed BIP322 for generic message signing: this recently-proposed BIP will allow users to create signed messages for all currently-used types of Bitcoin addresses, including P2PKH, P2SH, P2SH-wrapped segwit, P2WPKH, and P2WSH. It has the potential to become an industry standard that will be implemented by nearly all wallets and may be used by many services (such as peer-to-peer marketplaces) as well as for customer support, so Optech encourages allocating some engineering time to ensure the proposal is compatible with your organization’s needs. See the News section below for additional details.
● Optech Paris workshop November 12-13: member companies should send us an email to reserve spots for your engineers. Planned topics include a comparison of two methods for bumping transaction fees, discussion of partially signed Bitcoin transactions (BIP174), an introduction to output script descriptors, suggestions for Lightning Network wallet integration, and approaches to efficient coin selection (including output consolidation).
● C-Lightning 0.6.1 released: this minor update brings several improvements, including “fewer stuck payments, better routing, fewer spurious closes, and several annoying bugs fixed.” The release announcement contains details and links to downloads.
● BIP322 generic signed message format: since 2011, users of many wallets have had the ability to sign an arbitrary message using the public key associated with a P2PKH address in their wallet. However, there’s no standardized way for users to do the same using a P2SH address or any of the different types of segwit addresses (although there are some implemented non-standard methods with limited functionality). Picking up a Bitcoin-Dev mailing list discussion from several months ago, Karl-Johan Alm has proposed a BIP that could work for any address (although it’s not yet described how it would work for P2SH or P2WSH addresses involving an OP_CLTV or OP_CSV timelock).
The basic mechanism is that the authorized spender or spenders for an address generate scriptSigs and witness data (including their signatures) in much the same way they would if they were spending the funds—except instead of signing the spending transaction, they sign their arbitrary message instead (plus some predetermined extra data to ensure they can’t be tricked into signing an actual transaction). The verifier’s software then validates this information the same way it would to determine whether a spending transaction was valid. This allows the message signing facility to be exactly as flexible as Bitcoin scripts themselves.
Currently, discussion appears to be most active on the BIP proposal’s pull request.
● Bustapay discussion: a simplified alternative to the proposed Pay-to-Endpoint (P2EP) protocol described in newsletter #8, Bustapay provides improved privacy for both spenders and receivers—and also allows receivers to accept payments without increasing the number of their spendable outputs, a form of automatic UTXO consolidation. Although proposed to the Bitcoin-Dev mailing list a few weeks ago, several aspects of the proposal were discussed this week.
Although P2EP and Bustapay could end up being implemented by only a few wallets and services similar to the BIP70 payment protocol, there’s also chance they could end up being becoming as widely adopted as wallet support for BIP21 URI handlers. Even if they don’t see general adoption, their privacy advantage means they could end up well deployed among niche users. In either case, it may be worth dedicating some engineering time towards tracking the proposals and proof of concept implementations to ensure your organization can easily adopt them if desirable.
Notable commits this week in Bitcoin Core, LND, and C-lightning. Reminder: new merges to Bitcoin Core are made to its master development branch and are unlikely to become part of the upcoming 0.17 release—you’ll probably have to wait until version 0.18 in about six months from now.
● Bitcoin Core #14054: this PR prevents the node from sending BIP61 peer-to-peer protocol reject messages by default. These messages were implemented to make it easier for developers of lightweight clients to get feedback on connection and transaction relay problems. However, there’s no requirement (or way to require) that nodes send a reject message or an accurate reject message, so the messages arguably only end up wasting bandwidth.
It’s recommended that developers connect their test clients to their own nodes and inspect their nodes’ logs for error messages in case of problems (perhaps after enabling debug logging). Users who still need to send
rejectmessages can use the
-enablebip61configuration option, although it’s possible that
rejectmessages will be removed altogether in a release after 0.18.
● Bitcoin Core #7965: this long-standing issue tracked the removal of code in the libbitcoin_server component to handle whether the wallet is compiled or not. The issue was finally closed this week by the merge of Bitcoin Core #14168. This issue, along with a number of other issues such as Bitcoin Core #10973 (Refactor: separate wallet from node) and Bitcoin Core #14180 (Run all tests even if wallet is not compiled) are part of a long-term effort to disentangle the wallet code from the server code. Doing so provides a number of benefits including easier code maintenance, better opportunities for testing individual components, and potentially more secure software if the wallet component is moved to its own process.
● LND #1843: a configuration option intended only for testing (
--noencryptwallet) has been renamed to
--noseedbackup, been marked as deprecated, and had its help text updated and changed to mostly uppercase warning text. Developers are worried that ordinary users are enabling this option without realizing that it puts them only one failure away from permanently losing money.
● LND #1516: thanks to updates in the upstream Tor daemon, this patch makes it possible for LND to automatically create and set up v3 onion services in addition to its existing v2 automation. For the automation to work, users must already have Tor installed and running as a service.
C-Lightning: for testing, an RPC proxy is now used to simplify mocking responses to various RPC calls, making it easier to test lightningd’s handling of things such as fee estimates and crashes of bitcoind.