diff --git a/notes/2021/2021-01-14.md b/notes/2021/2021-01-14.md new file mode 100644 index 0000000..00f9eda --- /dev/null +++ b/notes/2021/2021-01-14.md @@ -0,0 +1,74 @@ +# 2021-January-14 ESLint TSC Meeting Notes + +## Transcript + +[`2021-01-14-transcript.md`](2021-01-14-transcript.md) + +## Attending + +* Brandon Mills (@btmills) - TSC +* Toru Nagashima (@mysticatea) - TSC +* Milos Djermanovic (@mdjermanovic) - TSC +* Nicholas C. Zakas (@nzakas) - TSC + +@nzakas moderated, and @btmills took notes and updated issues with resolutions. + +## Topics + +### [TSC/reviewer payment structure](https://github.com/eslint/tsc-meetings/issues/232#issuecomment-759768291) + +> @nzakas: I'd like to review the TSC/reviewer payment structure we have and see what people think now that we've been paying people for just about a year. I'm mostly interested in the limits: does the 20 hour per month limit still make sense? Does $50/hour still make sense? Is there anything we want to tweak at this point? + +* @nzakas has not come close to the limit. @mdjermanovic typically spends 35-40hr. @btmills usually hits around 20hr. @mysticatea hasn't had much time. +* The limit was originally to prevent burnout, and the $50/hr rate was what Webpack was paying and seems to be the going rate for frontend contract work. +* @nzakas likes to keep the hour limit even if it's more a reminder than a hard cap but feels uncomfortable that @mdjermanovic has been doing double the work unpaid and proposes keeping $50/hr up to 20hr/mo and adding $25/hr over 20hr/mo. :+1: from @btmills, @mdjermanovic, and @mysticatea. +* Should we have a hard limit at 40 total hours per month? That would help calculating budget reserves. @mdjermanovic could work more some months but is fine with a limit. @nzakas proposes starting with a 40hr/mo cap and revisiting in a few months. :+1: from @btmills, @mdjermanovic, and @mysticatea. + +**Resolution**: We will pay $25/hr for hours 21-40 each month and revisit in a few months. @nzakas will update the team repository. + +### [Process](https://github.com/eslint/tsc-meetings/issues/232#issuecomment-759768291) + +> @nzakas: What parts of our process are slowing/breaking down right now? It seems that everyone isn't around as much as previously, so it might be a good time to talk about how that's affecting the project. + +* @nzakas hasn't been able to do as much lately but is hoping that will change soon. +* @btmills feels like we're back where we were before we had a half-time maintainer. Latency has increased, and issues/PRs past the first page get less activity. +* @nzakas agrees and feels the logjam is when issues/PRs are opened and that PRs are still getting merged. The project board would help here. +* @btmills agrees that the beginning of the funnel is the slow part, and the release cycle acts to keep merges moving. +* @nzakas plans to finish up the [project board](https://github.com/eslint/tsc-meetings/blob/5a1f2107149119453b61d499a3cfeef34f8df712/notes/2020/2020-10-08.md#project-to-track-issue-triage) and setting up a [triage team](https://github.com/eslint/tsc-meetings/blob/5a1f2107149119453b61d499a3cfeef34f8df712/notes/2020/2020-09-24.md#triage-team) to help. +* @btmills suggests offering existing committers to formalize an hourly rate on the triage team, and @nzakas suggests "promoting" repeat contributor award recipients. +* :+1: from @mdjermanovic, @btmills, and @mysticatea. + +**Resolution**: @nzakas will prioritize finishing the project board and then follow up with a propsal for starting the triage team. + +### [Lock files](https://github.com/eslint/tsc-meetings/issues/232#issuecomment-759865391) + +> @nzakas: let’s talk about why we don’t use lock files in the repos and whether it makes sense. + +* Lock files are most useful for deployed applications to ensure versions are consistent between environments and across deployments. +* We could use a lock file to keep versions consistent in our development environments. +* End-user installs via `npm install eslint` or `yarn add eslint` do not use dependencies' lock files, so CI shouldn't lock versions either with `npm ci`. +* If we were to use `package-lock.json` and one of our dependencies accidentally breaks in a newer version, end users would see the issue, and CI could reproduce it, but we'd have to update our lock file to reproduce it locally. +* @nzakas suggests publishing this reasoning as a README FAQ. @btmills suggests linking to https://www.twilio.com/blog/lockfiles-nodejs. + +**Resolution**: @btmills will add a README FAQ explaining why we don't use a lock file. + +### [TC39 tools outreach meetings](https://github.com/eslint/tsc-meetings/issues/232#issuecomment-760440650) + +> @btmills: I've been attending monthly TC39 tools outreach discussions for about a year now. The time commitment is about 2 hours between pre-reading discussion topics, the meeting itself, and sending out notes to our team. +> +> - Should we continue having a team member attend? +> - Would anyone else like to do some? (Not saying I don't want to - it's interesting - just making sure I'm not excluding someone else who's interested.) +> - Is the notes format sufficient? Can I improve anything? + +* @nzakas feels it's important to be there, would like to go at some point, and enjoys the notes. +* @mdjermanovic says the notes are great. +* @btmills finds attending beneficial persionally for education and for the project when topics are relevant to ESLint. It gives advance notice of future language changes and a chance to help inform designs that affect ESLint. + +**Resolution**: Attendance is worthwhile, @btmills will continue going, and @nzakas hopes to attend in the future. + +### [Scheduled release for January 15th, 2021](https://github.com/eslint/eslint/issues/13982) + +* Jenkins is up, and it appears that @nzakas fixed the permissions error. +* Team members should add public keys to GitHub so @nzakas can add them to the Jenkins server. + +**Resolution**: @mdjermanovic will do the release.