Skip to content

chore(deps): bump assertj across all eight smoke setups, and route them to develop - #517

Merged
DemchaAV merged 1 commit into
developfrom
chore/smoke-assertj-and-dependabot-scope
Aug 5, 2026
Merged

chore(deps): bump assertj across all eight smoke setups, and route them to develop#517
DemchaAV merged 1 commit into
developfrom
chore/smoke-assertj-and-dependabot-scope

Conversation

@DemchaAV

@DemchaAV DemchaAV commented Aug 5, 2026

Copy link
Copy Markdown
Owner

Why

Dependabot opened #514 against main. That is precisely the divergence the maven
block's target-branch: develop exists to prevent — its own comment records it being
fixed in v1.6.8 "after the #111 / #115 episodes". main is the release pointer, moved by
fast-forwarding develop onto it; a commit landing on main directly ends that, and the
next git push origin develop:main stops being a fast-forward.

The cause is scope, not the rule: scripts/release-smoke/ sits outside every directory
the block lists, so updates for those poms arrive through the default branch instead. The
commit-message prefix gives it away — build(deps-dev) rather than the configured deps.

#514 also bumped one project of eight. The smoke setups are deliberately identical
consumers, differing only in which artifacts they depend on. A version moving in one of
them is drift, not an upgrade.

What

  • All eight smoke poms go to assertj 3.27.7 together — the version the engine itself
    already uses.
  • The maven block's directory list covers /scripts/release-smoke/*, so the next update
    for them targets develop like everything else.

Tests

./mvnw -B -ntp clean verifyBUILD SUCCESS, and dependabot.yml parses with both
entries still routed to develop.

The change is in the smoke projects themselves, so the meaningful check is the smoke:
run against the published 2.1.1 train with the bumped version — 8/8 pass.

On #514

It should be closed rather than merged or retargeted: retargeting would still leave one
project of eight moved, and this supersedes it with the uniform bump plus the config fix
that stops the next one arriving the same way.

…em to develop

Dependabot opened #514 against main. That is the divergence the maven block's
target-branch exists to prevent: main is the release pointer, moved by
fast-forwarding develop onto it, and a commit landing on main directly ends that.
The cause is scope — scripts/release-smoke sits outside every directory the
block lists, so updates for those poms arrived through the default branch
instead. The directory list now covers them.

#514 also bumped one project of eight. The smoke setups are deliberately
identical consumers, differing only in which artifacts they depend on, so a
version moving in one of them is drift rather than an upgrade. All eight go to
3.27.7 together, which is the version the engine itself already uses.

Verified by running the smoke against the published 2.1.1 train: 8/8 pass.
@DemchaAV
DemchaAV merged commit 523cc99 into develop Aug 5, 2026
10 checks passed
@DemchaAV
DemchaAV deleted the chore/smoke-assertj-and-dependabot-scope branch August 5, 2026 08:50
DemchaAV added a commit that referenced this pull request Aug 5, 2026
The release-smoke note explained #514 landing on `main` as a consequence of
those paths sitting outside every listed directory, which reads as though
listing them stops it happening again. It does not: Dependabot raises
security updates against the default branch whatever `target-branch` says,
so the next advisory touching these manifests will open against `main`
again. Records the rule and the handling it implies — close the PR and carry
the bump to `develop`, as #517 did — plus the fact that a non-default
`target-branch` also exempts security updates from the `labels` and
`commit-message` settings.

Also drops an inherited claim that the eight release-smoke projects are
identical setups. They are deliberately different, one per published
coordinate combination; only their test scaffolding matches, which is why a
bump has to be applied eight times.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant