Repository navigation
Description
Decision required: no
Summary: A routine tidy-up pass over everything currently open in this test repository, looking for duplicate or overlapping requests. Good news: the queue is clean — every open change already points at the issue it works on, and the two big planning issues deliberately split their work. The only loose thread is one review-backlog issue that was filed with an empty description, which a recent change may already address. Nothing here asks anyone to close or merge anything.
Summary
Triage pass over the 15 open issues and 6 open PRs in projectbluefin/testsuite as of 2026-10-29. Found 0 near-duplicate pairs, 0 same-root-cause clusters, 0 items that appear already resolved. One open question (#1053 ↔ PR #1065) and several verified-distinct pairs are recorded so the next pass does not redo the work. No actions have been taken; this is a checklist for maintainers.
Related but distinct: cross-link only
- Review backlog (intent-alignment): GNOME end-to-end test job can report success without checking any test passed #1053 and PR feat(gnome): add a GNOME OS desktop check with real keyboard input #1065 — open question. Review backlog (intent-alignment): GNOME end-to-end test job can report success without checking any test passed #1053's title is "GNOME end-to-end test job can report success without checking any test passed", but its body is empty, so the concrete job/line it reports cannot be verified. PR feat(gnome): add a GNOME OS desktop check with real keyboard input #1065 independently adds exactly that class of guard: "Apply the strict extension result gate to the baseline too. Require the qecore/Behave process to exit successfully as well: wrapper or final-hook errors cannot pass merely because scenario records say
passed." If Review backlog (intent-alignment): GNOME end-to-end test job can report success without checking any test passed #1053's (missing) evidence names thegnome-e2eresult handling, feat(gnome): add a GNOME OS desktop check with real keyboard input #1065 likely addresses it and should reference it; a human with the original review context needs to repopulate Review backlog (intent-alignment): GNOME end-to-end test job can report success without checking any test passed #1053's body first. - PR fix(e2e): default the test image to dakota:testing #1061 and fix(smoke): 16 Dakota smoke scenarios fail on current dakota:testing #1059 — related, already cross-referenced. fix(e2e): default the test image to dakota:testing #1061 fixes the stale
dakota:latestdefault ine2e.yml; fix(smoke): 16 Dakota smoke scenarios fail on current dakota:testing #1059 tracks the 16 product-side Dakota smoke failures that surfaced once a current image was booted. fix(e2e): default the test image to dakota:testing #1061's body already cites fix(smoke): 16 Dakota smoke scenarios fail on current dakota:testing #1059 and v1 tag held by the post-merge VM gate #1058. Distinct: config default vs. image regressions. - PR feat(e2e): keep Linux security protections on during desktop tests and fail on blocked actions #1037 and PR fix(e2e): default the test image to dakota:testing #1061 — both edit
.github/workflows/e2e.yml(disjoint concerns: feat(e2e): keep Linux security protections on during desktop tests and fail on blocked actions #1037 removesselinux=0and adds SELinux gates; fix(e2e): default the test image to dakota:testing #1061 changes theimageinput default). Hunks appear disjoint, but whichever merges second will need a rebase; both are fork PRs the authors must refresh. - Expose the unreachable suites via e2e.yml #704 and Lab-repo infrastructure prerequisites blocking test coverage lab#1182 — two stalled epics that reference each other and deliberately split scope: Expose the unreachable suites via e2e.yml #704 exposes the unreachable suites via
e2e.ymlinputs (testsuite-owned), Lab-repo infrastructure prerequisites blocking test coverage lab#1182 tracks the lab-owned GPU/TPM/second-image prerequisites. Not duplicates; keep both.
Checked and not duplicates
- PR feat(e2e): keep Linux security protections on during desktop tests and fail on blocked actions #1037 ↔ feat(e2e): qualify the GHA VM with SELinux enforcing instead of selinux=0 #1024 —
Refs projectbluefin/testsuite#1024, deliberately non-closing (hosted VM run and rollout decision design(ci): qualify trustworthy GHA regression results and roll out blocking promotion gates #1021 remain open). Link present. - PR feat(e2e): test that first-boot setup really installs the default apps and tools #1051 ↔ feat(e2e): exercise shipped startup services and system Flatpak initialization in GHA #1025 —
Refs projectbluefin/testsuite#1025, deliberately partial (scenario re-enables, greenboot Lab-repo infrastructure prerequisites blocking test coverage lab#1182, release wiring remain). Link present; issue already labelled partially-addressed. - PR feat(gnome): add a GNOME OS desktop check with real keyboard input #1065 ↔ feat(gnome): prove a genuine GNOME OS guest lane for extension validation #908 —
Refs projectbluefin/testsuite#908, body states "This PR must not close the issue". Link present. - PR docs: show how to run the GNOME 51 check on the published test image #1029 ↔ test(smoke): audit GNOME 50-specific workarounds against GNOME 51 and update gnome.md #827 —
Refs projectbluefin/testsuite#827with a "Still open on test(smoke): audit GNOME 50-specific workarounds against GNOME 51 and update gnome.md #827" section. Link present. - PR fix(lifecycle): fail update checks when required tests are skipped #1057 ↔ fix(lifecycle): require a real pinned-image upgrade and rollback in mandatory qualification #1023 —
Refs projectbluefin/testsuite#1023, partial per the maintainer's scoping comment; remainder waits on Lab-repo infrastructure prerequisites blocking test coverage lab#1182's second-image prerequisite. Link present. - PR feat(e2e): test that first-boot setup really installs the default apps and tools #1051 ↔ PR feat(e2e): keep Linux security protections on during desktop tests and fail on blocked actions #1037 — feat(e2e): test that first-boot setup really installs the default apps and tools #1051 injects
selinux=0in its newstartup-qualification.ymlwhile feat(e2e): keep Linux security protections on during desktop tests and fail on blocked actions #1037 removes it frome2e.yml: different workflows, and feat(e2e): test that first-boot setup really installs the default apps and tools #1051's body explicitly defers enforcing to feat(e2e): qualify the GHA VM with SELinux enforcing instead of selinux=0 #1024. No conflict. - Review backlog (correctness): Same retrying per-alias lookup pattern remains in other step modules #1038 ↔ Review backlog (intent-alignment): GNOME end-to-end test job can report success without checking any test passed #1053 — both "Review backlog" filings, different findings (retrying per-alias lookup pattern vs. job success reporting). Distinct.
Filed by dedupe agent (issues-only mode). Recommendations only — nothing has been closed, labelled, or merged.
🐝 Hive Agent: dedupe | Instance: hosted-projectbluefin-knuckle-gjvq | SHA: f950551a4bc18c73c58b681dce27699f28bf0724
— hive: agent=dedupe backend=copilot model=claude-fable-5 copilot=1.0.88
Activity
hivecommons-hive commentedon Nov 5, 2026
Closing: this advisory dedupe checklist expired after 7 days.