lightning-dev
Combined summary - [META] Organization of 1.1 Spec Effort
In a message posted on November 27, 2018, Matt Corallo expressed his preference for substantive discussions to happen on GitHub or the mailing list instead of live meetings.
While he was in favor of IRC meetings, he argued that finalization during a live meeting is unfair to those who cannot attend regularly or missed the discussion. Additionally, he pointed out that textual communication poses fewer barriers than audio/visual meetings as not everyone wants their face or voice to be seen or heard. This message was posted on Peter Todd's website, with his contact information provided in the signature block.In a mailing list discussion, Rene suggested creating a wiki page to make the process transparent and link it from README.md. He expressed that the current process is new to him and newcomers cannot be expected to read through the entire mailing list. Matt Corallo expressed his preference for substantive discussions to happen on GitHub or the mailing list rather than in live meetings, which he believes is unfair to those who cannot regularly attend or miss the meeting where the topic they care about is discussed.Rusty Russell proposed revisiting the current process rules as they approach 1.1, suggesting moving to an IRC meeting and having a more formal approval method for PRs. He also added milestones to PRs as 1.0/1.1 and aims to clear all 1.0 PRs this week for tagging in the next meeting. The discussion thread ended with a link to the Lightning-dev mailing list.The Lightning-dev mailing list is discussing the process rules for their 1.1 iteration, as they approach the end of 1.0. Currently, they have biweekly Google Hangouts, which require unanimous approval at the meeting before any substantive spec change can be applied. Implementation changes require two interoperable implementations before being considered final. Typo, formatting, and spelling fixes can be applied after two acknowledgments without a meeting necessary. As they approach 1.1, they are considering moving to an IRC meeting, which would be more inclusive and better recorded but could be lower-bandwidth. They are also considering having a more formal approval method for PRs with a "CONSENSUS:YES" tag that applies once they have acknowledgments from two teams and no Naks, followed by a meeting to review consensus, a "FINAL" tag, and commit the next meeting. This gives at least two weeks to comment on the final draft.Despite the benefits of live meetings, it is acknowledged that doing finalization in a live meeting is unfair to those who can't find the time to attend regularly or happened to miss the one where the relevant discussion occurred. Therefore, substantive discussions should preferably happen on GitHub or the mailing list. The current process rules will need to be revisited as they approach 1.1.Rusty Russell has added milestones to PRs as 1.0/1.1 and hopes to clear all 1.0 PRs this week for tagging in the next meeting, then start on 1.1 commits. The author of the message informs that for the 1.0 spec, a biweekly Google Hangout was held at 5:30 am Adelaide time, and provides a link to the minutes of all meetings. The current process rules are also listed in the message which includes unanimous approval for substantive spec changes, two interoperable implementations for implementation changes, and typo fixes with just two acks. As they approach version 1.1, the author suggests revisiting these rules. Two questions are proposed: whether to move to an IRC meeting as Bitcoin development does, and whether to have a more formal approval method for PRs with a "CONSENSUS:YES" tag. The author notes that milestones have been added to PRs as 1.0/1.1 and hopes to clear all 1.0 PRs this week for tagging in the next meeting before starting on 1.1 commits.