fix(release-nudge): stop mandating a NEWS.md changelog step - #67
Conversation
Release notes are written on the GitHub release itself across the org, so the nudge asserts a changelog convention it cannot verify and that several repos deliberately do not keep. Step 3 gains a prompt to read the published notes, which is the check the step was reaching for. Co-authored-by: Sam Abbott <contact@samabbott.co.uk>
There was a problem hiding this comment.
The PR removes the mandatory NEWS.md-update step from the release-nudge comment and replaces it with a prompt to check the published release notes, since not all repos keep a NEWS.md. The change is a clean, self-contained edit to a Julia string builder with no other references to NEWS.md left dangling elsewhere in the repo, but the new sentence added to step 3 has a grammar error.
Automated first pass by seabbs-review-bot (Claude sonnet), triggered by: first pass. Not a human review. Comment @seabbs-review-bot to ask for another pass: @seabbs any time, the author's agent once it has pushed changes. Add the no-review label to opt this PR out. Ping @seabbs with any questions.
| "`/register` on an issue or pull request, or run " * | ||
| "the Register workflow manually from the Actions tab." | ||
| "the Register workflow manually from the Actions " * | ||
| "tab, then check the published release notes read " * |
There was a problem hiding this comment.
issue This renders as "...run the Register workflow manually from the Actions tab, then check the published release notes read sensibly." — missing "that", so it's not a grammatical sentence. Since this is the whole point of the PR (replacing the NEWS.md check with a release-notes check), it's worth getting right: "...tab, then check that the published release notes read sensibly."
Closes #45.
What
Drops the
NEWS.mdclause fromsteps_textinrelease-nudge.ymland adds aprompt to step 3 to read the published release notes.
Step 2 now reads:
The load-bearing "the bump must land on the default branch first" clause is
untouched.
Why option 3 rather than the issue's preferred option 2
The issue preferred rewording toward "update
NEWS.mdif this repo maintainsone". Since it was filed, the one-repo decision became the org standard: the
kit seeds no changelog at all
(
EpiAwarePackageTools.jl/src/scaffold.jl:298, "No NEWS.md seed: release notesare written on the GitHub release itself"), and
src/scaffold.jl:445 _PROSE_CHANGELOGStreatsNEWS.md,CHANGELOG.mdandrelease-notes.mdas legacy to warn about. A conditional clause would keepasserting a convention the standard has dropped, so dropping the sentence is
now the more correct option.
On the JuliaRegistrator comment in the thread
That concerns the issue body, not this workflow.
release-nudge.ymlalreadyrecords at lines 22-25 that no handle is written anywhere in the file, not even
the registry bot's own, not even inside a code span, and
grep -n 'JuliaRegistrator' .github/workflows/release-nudge.ymlreturnsnothing. The issue body has been edited separately to defuse the trigger
string.
Validation
python3 -c "import yaml; yaml.safe_load(open('.github/workflows/release-nudge.yml'))"passes. Nudge staleness is keyed on the
release-nudge-statemarker (versionand unreleased commit count), not on the prose, so this text change does not
churn open nudge issues.
This was opened by a bot. Please ping @seabbs for any questions.