From 8e35cdaab67f4cd7a40ed1adbb2c0bd6e9ef956d Mon Sep 17 00:00:00 2001 From: Lars Wirzenius Date: Sun, 20 Jun 2021 08:13:41 +0300 Subject: docs: add meeting minutes for yesterday's meeting Sponsored-by: author --- blog/2021/06/19/meeting.md | 108 +++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 108 insertions(+) create mode 100644 blog/2021/06/19/meeting.md diff --git a/blog/2021/06/19/meeting.md b/blog/2021/06/19/meeting.md new file mode 100644 index 0000000..0348dfb --- /dev/null +++ b/blog/2021/06/19/meeting.md @@ -0,0 +1,108 @@ +[[!meta title="Iteration planning: June 19"]] +[[!tag meeting]] +[[!meta date="2021-06-19 12:00 +0000"]] + +[[!toc levels=2]] + +# Review of actions from last meeting + +* There were no actions. The action about [[!issue 22]] was a mistake. + +# Review of the iteration that has ended + +[[!milestone 30]] was for this iteration. We closed all issues picked +for the iteration: + +* [[!issue 194]]—trial release +* [[!issue 195]]—`./check` to tail logs on failure +* [[!issue 197]]—`lib/daemon.py` to break long lines in log +* [[!issue 199]]—honour `JAVA_HOME` for Pandoc +* [[!issue 200]]—sort lists in metadata output +* [[!issue 201]]—move `bindings-ubm` into `tests` +* [[!issue 203]]—document `build.rs` + +All issues tagged `current-goal` are now closed, except [[!issue +125]], which is the tracker issue for getting people to try Subplot. + +# Review of the repositories + +We reviewed the `subplot-web` and `subplot-container-images` +repositories, and there were no open issues, no open merge requests, +no extra branches, and no unsuccessful CI pipelines. The `subplot` +repository had no open merge requests, no unnecessary branches, and +the pipeline for the `main` branch was successful. There is one +work-in-progress branch for changing the Subplot subplot to use Rust +instead of Python for step implementations. All good: no pending +cleanup work from this iteration. + +# Goal for this iteration + +We discussed the goal for this iteration, which is to get a few people +actually try Subplot. We reformulated [[!issue 125]] accordingly. We +think all code changes are ready for the first outside users, so the +work we want to do is to make a release for them to try, and then be +ready to react to any feedback quickly. + +# Issue review + +We reviewed all issues this time. Most issues are not relevant to +near-term goals, but to manage that better, we created a label for +"goal two", after the current goal ("goal one"). Goal one is about +getting others to try Subplot; goal two is to make Subplot nicer to +use. Once goal one is done, we'll rename the `current-goal` tag to +`goal:1`, and the `goal:2` tag to `current-goal`. + +We added the following issues to the `goal:2` tag: + +- [[!issue 20]]—conditional scenarios +- [[!issue 22]]—location of helper binaries +- [[!issue 74]]—environment variables for Bash +- [[!issue 120]]—list all matching bindings +- [[!issue 190]]—test suite is slow +- [[!issue 188]]—libraries using regex patterns +- [[!issue 196]]—Python runner stops on first failure +- [[!issue 198]]—unified Subplot libraries + +We decided to pick one difficult issue for each goal. For goal two +that is [[!issue 20]]. + +We closed [[!issue 154]]] as hopefully done, and [[!issue 191]] as +unwanted. + +After the review there were 33 open issues. + +# Current goal + +Our current goal is to get Subplot the software and project into such +a shape that we feel confident that we can ask a few people to give it +a proper try. This requires for Subplot to have releases that are easy +to install, tolerably easy to use, and to have at least some +rudimentary documentation aimed at beginners. + +# Plan for next iteration + +We created [[!milestone 31]], with two issues: + +- [[!issue 125]]—get people to try Subplot +- [[!issue 204]]—create release + +# Other business + +We discussed using GitLab issues for iteration meeting agenda. For +now, we'll stick to what we have, since it works for the two of us, +but as soon as others want to join, we'll re-evaluate. Using issues, +or possibly merge requests, might be useful if contributors aren't in +time zones close to the European one. + +We discussed using `cargo release` to simplify the release process, +but decided to evaluate it in more detail later. + +We discussed release cadence, and decided to start each iteration by +making a release, unless there is something to prevent a release. +Making the release at the start of an iteration means it's less +hurried, and so more likely to be correct. We also decided to automate +the release process as we go along. + +# Actions + +No actions. -- cgit v1.2.1