NS7: V1 release hardening and rollout #7
Reference in New Issue
Block a user
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
OUTCOME:
mimcan be built, installed, tested, and dogfooded by actual users as a prod-ready v1 without surprising configuration or hidden dependencies.WHY NOW: The PRODUCT.md proof is real user choice over Neovim for small iPhone SSH fixes; rollout needs reliable builds, profile boundaries, tests, install path, and supportable diagnostics.
AUTHORITY: PRODUCT.md; req: governance/001-005; all v1 requirement rows as release gates
SLICES:
MEMBERSHIP MECHANISM: Gitea fallback via this parent checklist plus explicit parent link in each child issue body.
AGENT AUTHORITY DEFAULTS:
BLOCKERS: external real iPhone SSH dogfood receipt is required for #41. Local build/test/install proof is complete; closure needs either a pasted DOGFOOD.md receipt from an actual iPhone SSH edit or explicit tracker acceptance change.
PROOF: A fresh server checkout can build named profiles, run tests, install the binary, perform the mobile SSH dogfood scenario, and diagnose common failures without hidden setup.
TEST STANDARD: Each child slice must carry its test weight: regular tests prove the intended payoff works, and adversarial tests cover realistic failure, boundary, or regression cases. Prefer protocol/headless/replay/fixture tests over brittle UI snapshots unless rendering behavior is the claim. See req: testing/001 through req: testing/004.
NON-GOALS: No native app, cloud service, telemetry, package manager, runtime config ecosystem, or compatibility promise beyond documented v1 boundaries.
GOAL: Implement this Northstar by completing its linked child slices without leaving the product/refusal boundary in PRODUCT.md. Close this parent only when every child is closed and the PROOF above is demonstrated.
Tracker accounting update: NS7 is blocked only on #41 external dogfood evidence.
I changed the stale
BLOCKERS: nonefield on #7 and #41 to reflect the real blocker: the required DOGFOOD.md receipt from an actual iPhone SSH edit, or an explicit tracker acceptance change. No implementation blocker is currently known from local verification.Current local proof at
f5c81155earemains:Not closing #7 or #41: PRODUCT.md proof still requires the real mobile SSH dogfood receipt.