Published on: 2020-12-18T02:02:04+00:00
In a recent discussion on the Bitcoin mailing list, Pieter Wuille and Rusty Russell explored the implementation of changes to witness version 1 (v1) addresses. Currently, there are no v1 receivers, so the focus is on determining what software and infrastructure support sending to v1 addresses and whether they will upgrade. The first option discussed is to continue using v1, while the second option involves upgrading to a new version. There are concerns about trailing typos causing issues for non-upgraded software under option 1. Option 2, on the other hand, would likely lead to fixes when someone attempts a v1 send. However, it is possible that non-upgraded software may never get fixed, delaying the ability to use v1 addresses.Pieter is interested in understanding which codebases, services, and infrastructure currently support sending to witness v1 BIP173 addresses. Rusty suggests spamming an address from various sources to test compatibility with v1 addresses. They discuss the potential adoption of new technology based on the replacement of older infrastructure and codebases. Both Pieter and Rusty acknowledge that their thinking on this matter may change over time.Another proposal brought up by Rusty Russell involves two options for the future of v1 receivers. The first option is to continue using v1, while the second option requires upgrading to a new version. The support for sending to v1 BIP173 addresses is a key consideration in making this decision. Rusty favors the second option as it forces upgrades and breaks clearly. He believes that accepting v1 addresses without upgrades could create liabilities. David A. Harding agrees with the first proposal, highlighting the effort put into obtaining widespread support for bech32 addresses. Harding suggests that consensus should restrict v1 witness program size to maximize safety.In a separate thread, David A. Harding expresses his preference for using the backwards compatible proposal from BIPs PR#945. He suggests that consensus should restrict v1 witness program size by rejecting transactions with scriptPubKeys paying v1 witness programs that aren't exactly 32 bytes. Harding believes that deferring a hard decision is not useful, and he hopes that most software will have implemented length limits by the time segwit v2 is used.Rusty Russell proposes an alternative to the length restrictions discussed in a BIPs pull request. The alternative involves using a checksum change based on the first byte, unless the first byte is 0. There are two proposals debated in the discussion. The first proposal suggests length restrictions for future segwits, while the second proposal suggests a checksum change based on the first byte. Rusty prefers the second option as it forces upgrades and breaks clearly. However, David A. Harding argues that the second option does not force upgrades but creates another opt-in address format. Harding favors the backwards compatible proposal from BIPs PR#945 and suggests consensus restricting v1 witness program size for safety purposes.The author of the proposal suggests an alternative to length restrictions proposed by Rusty Russell. The alternative involves using a variant based on the first byte, unless the first byte is 0. There are two proposals discussed: length restrictions and a checksum change based on the first byte. The first proposal restricts future segwit versions, while the second weakens guarantees against typos. The author prefers the second proposal as it forces upgrades and addresses the length extension bug. They emphasize the need for a decision to be made promptly to begin upgrading software. It is also mentioned that Lightning uses bech32 over longer lengths but will follow Bitcoin's choice.Overall, the discussions revolve around the implementation of changes to witness version 1 addresses, the support for sending to v1 BIP173 addresses, and the options for future versions. Various proposals and concerns are raised regarding compatibility, upgrades, safety, and the simplicity of the changes. The participants acknowledge that their perspectives may evolve over time and emphasize the importance of gathering data and making informed decisions.
Updated on: 2023-08-02T02:47:10.950630+00:00