Repository navigation
Description
Decision required: yes
Summary: Two open pull requests propose the same improvement to how optional server add-ons are exported, and a separate pair changes the same installer typing logic. This checklist asks maintainers to decide which proposal to close and what order to use for the installer changes.
What I need from you: Decide whether to close PR #649 as superseded by #646, and review the merge order for PRs #644 and #647.
Summary
Triage based on the 2026-11-12 14:11 snapshot, which reports 22 open issues and 26 open PRs in projectbluefin/server. The issue search returned 22 rows, but the prescribed open-PR search returned no rows despite the snapshot listing 24 actionable PRs and one held PR; the PR inventory could not be fully reconciled. One verified near-duplicate pair and one merge-order conflict were found; no resolved candidates were verified. No actions have been taken.
Near-duplicates: close one
| Keep | Merge into it | Reason |
|---|---|---|
| PR #646 — exporting one add-on no longer deletes earlier exports | PR #649 — fold the copied export recipes onto one helper | Both change the same Justfile export recipes and export-contract tests: #649 factors five add-on recipes into a helper, while #646 covers that shared refactor and also fixes k0s wiping the other exports and duplicating checksum lines. PR #649 itself says it is superseded by #646. |
- Fold the copy-pasted export-*-sysext recipes onto one helper (#638) #649 — close as superseded by fix: exporting one add-on no longer deletes the add-ons exported before it #646 after confirming fix: exporting one add-on no longer deletes the add-ons exported before it #646 covers the shared export-helper change
Conflicting PRs: merge order matters
- fix: widen dogfood installer per-key gap under console noise #647 and test: install a network-booted server from its own screen, and check unknown firmware keys refuse to start it #644 — both change the installer's
type_keyslogic; PR fix: widen dogfood installer per-key gap under console noise #647 explicitly says it conflicts with test: install a network-booted server from its own screen, and check unknown firmware keys refuse to start it #644's rewrite. Land test: install a network-booted server from its own screen, and check unknown firmware keys refuse to start it #644 first, then rebase or fold PR fix: widen dogfood installer per-key gap under console noise #647's timing change into it.
Related but distinct: cross-link only
- ci: server's full image build never finishes since 2026-11-08, so the merge queue lands nothing #650 and [shared-ci] build failing across projectbluefin/server #652 — both concern the server image
buildcheck, but ci: server's full image build never finishes since 2026-11-08, so the merge queue lands nothing #650 documents the long-running build and cache-miss evidence while [shared-ci] build failing across projectbluefin/server #652 is the shared incident for the current red check. Keep their incident scopes distinct unless a maintainer confirms one tracker should replace the other.
Checked and not duplicates
- SBOM lists pcre2 once although the image carries two copies (FSDK 26.08.3 and fsdk-containers' 26.08.2) #599/fix: list both pcre2 copies in the SBOM instead of hiding the second #645 — the issue reports two pcre2 versions hidden from the inventory; PR fix: list both pcre2 copies in the SBOM instead of hiding the second #645 preserves the distinct copies and closes SBOM lists pcre2 once although the image carries two copies (FSDK 26.08.3 and fsdk-containers' 26.08.2) #599.
- dogfood-installer: type_keys drops a keystroke under console noise, first boot says the passwords did not match #600/fix: widen dogfood installer per-key gap under console noise #647 — the issue reports a dropped password keystroke under console noise; PR fix: widen dogfood installer per-key gap under console noise #647 widens the key delay and closes dogfood-installer: type_keys drops a keystroke under console noise, first boot says the passwords did not match #600. This pair is separate from the test: install a network-booted server from its own screen, and check unknown firmware keys refuse to start it #644 overlap noted above.
- Secure Boot auto-enrollment fails in virt-manager VM: .auth files signed with APPEND_WRITE (0x67), systemd-boot writes 0x27 #612/fix: Secure Boot key enrollment no longer refused in VMs #648 — the issue reports Secure Boot enrollment rejected because signed attributes differ; PR fix: Secure Boot key enrollment no longer refused in VMs #648 uses the supported symbolic attributes to match the required write flags and closes Secure Boot auto-enrollment fails in virt-manager VM: .auth files signed with APPEND_WRITE (0x67), systemd-boot writes 0x27 #612.
- [ci-maintainer] Weekly home-lab test throws away the boot log that explains its failures #623/ci: keep boot logs from weekly home-lab tests #627 — the issue reports missing per-machine boot logs; PR ci: keep boot logs from weekly home-lab tests #627 uploads the nested serial logs and closes [ci-maintainer] Weekly home-lab test throws away the boot log that explains its failures #623.
- test: prove native installer console entry and firmware fixture coverage #450/test: install a network-booted server from its own screen, and check unknown firmware keys refuse to start it #644 — PR test: install a network-booted server from its own screen, and check unknown firmware keys refuse to start it #644 says it only partially addresses test: prove native installer console entry and firmware fixture coverage #450; pre-boot menu capture, the external console-driver contract, and broader measurements remain outside that PR.
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: 31aaad65f304332ba3f3fe124c6bd6b38ab290c8
— hive: agent=dedupe backend=copilot model=gpt-6-luna copilot=1.0.88
Decision taken by hive: Hive did not park this: engineering decision taken by hive — [dedupe] triage: duplicate and overlapping open issues/PRs to consolidate. Owner may override with /hive hold or needs-human.
Activity
jeefy commentedon Nov 12, 2026
Issue triage
Next action: a triager or maintainer reviews this issue.
Reporter or contributor
kind/*label is set, categorize the issue with one of:/kind bug,/kind regression,/kind security,/kind feature,/kind documentationor/kind cleanup./area <area>:desktop,flatpak,gaming,hardware,installerordx.Triager or maintainer
/triage accepted. Only triagers and maintainers can do this; acceptance does not assign an implementer.needs-kindclears when a kind is set.needs-triageand this post clear when triage is accepted.See Triage & Prow Commands for all available commands.
hivecommons-hive commentedon Nov 12, 2026
🌱 Good first issue — bring your own coding agent
Hi! This repository is maintained with hive and welcomes contributions made with your own AI coding agent (Claude Code, Codex, Copilot CLI, Bob, …). hosted-projectbluefin-knuckle-gjvq picked this issue as a good place to start because:
How to land it (no waiting on a human):
git commit -s).Fixes #661— that closing keyword is how hive's contributor lane links the PR to your claim.Paste this to your agent
Prefer not to drive the loop by hand? Run your own hive (we call them clankers) — a relay on your machine that claims, opens the PR and handles review fixes for you, and earns you a contributor profile and a leaderboard place: see https://hosted-projectbluefin-common-nmq5.hive.hivecommons.dev/contribute.
If nobody claims it, hive may pick it up itself — claiming first prevents that. Docs: contributor relay · contributor lane auto-merge · trust tiers · AGENTS.md.
hivecommons-hive commentedon Nov 17, 2026
All 2 items resolved (#649 resolved, #647 merged) — closing this advisory.