Skip to content

🐛 PR in progress chevron defers to running CI/Review and never speaks the CI verdict - #16644

Merged
clubanderson merged 3 commits into
v5from
fix/pr-progress-chevron-fallback
Nov 4, 2026
Merged

clubanderson merged 3 commits into
v5from
fix/pr-progress-chevron-fallback

Conversation

@clubanderson

@clubanderson clubanderson commented Nov 4, 2026 •

Copy link
Copy Markdown
Member

Fixes #16643

Post-roll screenshot (row for #16634): PR in progress ACTIVE "CI failing · unassigned …" beside CI ACTIVE "CI 4/11 · 1 failing (Sign-of…". Two chevrons active that are not the CI + Review pair, and the PR in progress lane speaking the CI verdict.

JS — tracker-row-strip.js activation resolution

  • CI / Review are resolved first (stageTrackerTrackRunning). If either runs (tracksRunning >= 1, was === 2), the stage chevron renders ham-settled-current and icon only — words move to the tooltip (Waiting on CI, running — …). A decision trigger keeps its button. So the active set is ⊆ {CI, Review} or exactly one chevron; the stage chevron is the ≥ 1 fallback.
  • tracker-row.js stageTrackerFixingWord: "CI failing · author-owned" / "Red · nobody on it" → "Fix owed · …".

Go — stage_tracker_remediation.go

  • Unpicked-up red PR: Unassigned · next: scanner (· fixer queued, N ahead kept); owner wait: Waiting for @author · fix owed. The red word (CI failing / Changes requested) rides only in Detail. Constants prProgressUnassignedWord / prProgressFixOwedWord.

Tests

  • Go: TestStageTrackerRemediationCaptionNeverNamesTrackVerdict (CI red / review red / both × unassigned / owner wait) + updated captions.
  • Node stage_tracker_remediation_chevron.test.js: #16634 shape — (a) CI 4/11 · 1 failing still running + unassigned ⇒ CI the only active chevron, PR in progress icon-only settled; (b) nothing running ⇒ PR in progress active "Unassigned · next: scanner", CI red icon; (c) CI + Review running ⇒ both active, stage icon-only. "three qualify" / one-track cases updated.
  • Prelude invariant over every row fixture (stage_tracker_ui_prelude_test.go): if > 1 lit, no stage chevron is lit; a lit stage chevron's words never match CI failing | CI n/m | Changes requested.

Docs: canonical "Chevron activation" rules added to docs/dashboard-sections/activity-monitor.md.

Note: #16637 (open) also edits tracker-row-strip.js / stage_tracker_activity.go from a pre-#16632 base; this PR is cut from current v5 and does not touch the same hunks, but #16637 will need a rebase regardless.

make dashboard-gen-check, go vet, go test ./pkg/dashboard/, docs citation check green locally.

Follow-up commit — one fill for an active track (same issue)
The darker amber on #16637 / #16567 was .ham-seg.ham-activity-track.active.ham-stalled (55 % amber mix + --text): the stalled-blocker fill from #16600, applied on 050 rows where the JS running predicate saw nothing moving (CI missing "not reported" / pre-#16632 unknown "no verdict yet"). Now: a text-bearing track has exactly the .ham-seg.active fill (stalled keeps stillness + person glyph); Go CI Active = running/queued regardless of poll age (age stays in the tooltip via observed_at); prelude invariant: every text-bearing track is plainly active with no pending/todo/done/failing/settled class; CSS test: no rule sets a background on .ham-activity-track.active.

…eaks only lifecycle

Activation resolves CI and Review first: if either runs on the head the
stage chevron (PR in progress, Merge queue …) renders settled-current and
icon only, its words in the tooltip, so CI + Review is the one pair that
can be active together and every other row has exactly one active
chevron. The unpicked-up and owner-wait captions drop the red word
("Unassigned · next: scanner", "Waiting for @author · fix owed"; the
external-PR fixing word reads "Fix owed · …"); the CI / review verdict
stays in the detail. Prelude invariant over every row fixture: active set
⊆ {CI, Review} or size 1, and no stage chevron names a track verdict.
Canonical rules documented in the Activity Monitor section.

Fixes #16643

Signed-off-by: Andrew Anderson <andy@clubanderson.com>
@kubestellar-prow kubestellar-prow Bot added the dco-signoff: yes Indicates the PR's author has signed the DCO. label Nov 4, 2026
@kubestellar-prow

Copy link
Copy Markdown
Contributor

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
Once this PR has been reviewed and has the lgtm label, please assign nicholasjackson for approval. For more information see the Code Review Process.

The full list of commands accepted by this bot can be found here.

Details Needs approval from an approver in each of these files:

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

@github-actions

github-actions Bot commented Nov 4, 2026 •

Copy link
Copy Markdown
Contributor

Open PR overlap check

Base branch: v5

Conflicts are checked with every PR merged onto the current tip of v5, so a PR that only needs a rebase is not reported as conflicting with its siblings.

PR Title Relation Files
#16637 🐛 Stage tracker: say CI running (not activity unknown) when checks are still pending CONFLICT src/pkg/dashboard/stage_tracker_activity.go
#16634 🐛 fix(dashboard): preserve settled review verdicts and explain findings CONFLICT src/pkg/dashboard/stage_tracker_motion_test.go
#16615 ✨ feat(ci): approve gated workflows for trusted PR heads same files src/pkg/dashboard/stage_tracker_activity.go

@hivecommons-hive

hivecommons-hive Bot commented Nov 4, 2026 •

Copy link
Copy Markdown
Contributor

Hive status — updated in place; edits do not notify. Full history is in the Activity Monitor.

clubanderson this PR is held at stage/071-needs-author and the hive will not merge it yet.

Failing checks:

  • test (dashboard shuffle 1/3)

Next: the hive does not push to your branch. Push a new head (or reply here); CI and the review swarm re-run automatically and the PR returns to the normal path.

@hivecommons-hive hivecommons-hive Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Read the activation code in tracker-row-strip.js, the remediation captions in stage_tracker_remediation.go and the lane-word fallback in tracker-row.js against v5; the change is coherent and looks correct to me. One small docs gap.

Correct as far as I can verify

  • tracker-row-strip.js:413-414: waitingOnTracks = tracksRunning >= 1 (was === 2) plus the icon-only content at :508 when waitingOnTracks && !sg.cap && !decision gives exactly the rule the new doc section states — the active set is ⊆ {CI, Review} or a single chevron. The tooltip prefix at :472 names the one running track correctly when only one runs.
  • stage_tracker_remediation.go:332,344: the PR-in-progress caption now carries fix owed / Unassigned and redWord is still prepended to out.Detail in both branches, so the CI/Review verdict is not lost, only moved to the detail. The remaining "CI failing · <check> red" at stage_tracker_activity.go:211 is the CI lane's own caption, which is where #16643 wants it.
  • static/index.html carries the same tracksRunning >= 1 / 'Fix owed' edits as static-src, so the regenerated bundle is in sync.
  • Checked for overlap with #16634 (also edits stageTrackerActivityTrackHtml in the same files): hunks do not touch and a three-way merge of the two heads is clean.

Finding (low, docs-currency) — src/docs/dashboard-sections/activity-monitor.md:505 still documents the lane word as Red · nobody on it; after tracker-row.js:291 that string is Fix owed · nobody on it (and Fix owed · bot PR). Line 85 of the same file was updated for the external-PR variants but this sentence was not. One-word fix.

— hive: agent=reviewer backend=copilot model=claude-fable-5.1 copilot=1.0.88

The darker amber the operator saw was .ham-activity-track.active.ham-stalled
(a 55 % amber mix with --text) — the stalled-blocker fill from #16600 on
rows where nothing moved (CI "not reported" / "no verdict yet" at 050).
An active, text-bearing track now has exactly the .ham-seg.active fill;
stalled keeps only its stillness and person glyph. Go CI Active follows
the state (running / queued) regardless of poll freshness; the
observation age rides ObservedAt for the tooltip. Invariants: every
text-bearing track is plainly active (prelude over all fixtures), and no
CSS rule restyles the active track background.

Refs #16643

Signed-off-by: Andrew Anderson <andy@clubanderson.com>
@github-actions github-actions Bot added the conflicts-with-open-pr Open PR has a merge conflict with another open PR label Nov 4, 2026

@hivecommons-hive hivecommons-hive Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Correctness review (read at head 72baa0b, which superseded the dispatched 923b33e)

The change does what #16643 asks: tracker-row-strip.js:413-415 resolves CI/Review first and demotes the stage chevron on tracksRunning >= 1, :505 drops its words to the tooltip, and stage_tracker_remediation.go:332-346 keeps the red verdict in Detail only. The dropped Go freshness guard (stage_tracker_activity.go:256,350) is safe: no Go code consumes CIActivity.Active, and the JS pulse is still gated by its own 2‑minute window (tracker-row-strip.js:273-278), while stageTrackerTrackRunning already counted running/queued state regardless of active (:238). Node suites stage_tracker_remediation_chevron and stage_tracker_activity pass at head (49/49); go build ./pkg/dashboard is clean.

One advisory, not blocking:

  • Decision rows with CI genuinely running lose their fill. stageTrackerTrackRunning only ignores decision_needed/human_attention on the inferred ring-not-full path (tracker-row-strip.js:249); an explicit a.active or state: 'running' still counts (:238). With tracksRunning >= 1 the "Needs human decision" chevron now gets ham-settled-current (:425) — grey surface-3 — while the CI chevron is the amber one. The button is kept (:505), so nothing breaks, and it follows the issue's rule 3 literally; just flagging that a row parked on a person now visually points at CI rather than the person. If you'd rather keep the decision chevron lit, the narrowest change is to exclude decision rows from deferral:
      const waitingOnTracks = tracksRunning >= 1 && !((item.decision_needed && item.decision_needed.question) || item.human_attention);

Limits: Go tests could not be run on this host (blocked for agents); CI test (dashboard */4) was still in progress when I read the head.

— hive: agent=reviewer backend=copilot model=claude-fable-5.1 copilot=1.0.88

@hivecommons-hive

Copy link
Copy Markdown
Contributor

CI triage (quality): test, coverage, test (dashboard 4/4) and test (dashboard shuffle 1/3) on 72baa0b are one failure, PR-local but not this PR's fault (job 126802141286):

stage_tracker_motion_test.go:501: node row stamp fixture failed: fork/exec .../node: argument list too long
--- FAIL: TestStageTrackerMotionRowStamp

The fixture passes the lifted stage-tracker row builder to node -e as one argv; v5 sits just under Linux's 128 KiB MAX_ARG_STRLEN, and any growth of the lifted strip (as here) tips it over. #16637 is red on exactly the same line.

Merge order: #16634 (green, waiting on tide) already converts all four -e sites in stage_tracker_motion_test.go to a temp-file runner (runMotionScript). Once it lands, merging v5 into this branch should clear the failure with no change of its own; it also removes the need to hand-apply the same fix here. Not pushing to your branch since the PR is human-authored.


🐝 Hive Agent: quality | Instance: hosted-available-oke-11-placeholder-r05x | SHA: unknown

— hive: agent=quality backend=copilot model=claude-fable-5.1 copilot=1.0.88

Refs #16643

Signed-off-by: Andrew Anderson <andy@clubanderson.com>
@clubanderson
clubanderson merged commit bc23607 into v5 Nov 4, 2026
81 of 82 checks passed
@kubestellar-prow
kubestellar-prow Bot deleted the fix/pr-progress-chevron-fallback branch November 4, 2026 19:35
@github-actions

github-actions Bot commented Nov 4, 2026

Copy link
Copy Markdown
Contributor

Thank you for your contribution! Your PR has been merged.

We'd love to hear how your experience was: share feedback

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

conflicts-with-open-pr Open PR has a merge conflict with another open PR dco-signoff: yes Indicates the PR's author has signed the DCO. size/L Denotes a PR that changes 100-499 lines, ignoring generated files.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

🐛 PR in progress chevron lights beside a running CI chevron and speaks the CI verdict

1 participant