Fill gitea's pool keys in the provisioner client's gitea tests, which fail on main since 4becc7bd (issue 1966) #653

2026-11-10T11:59:53Z

mesh-tools commented

2026-11-10 11:59:53 +00:00

Issue 1966 (notes 178, 179). Since 4becc7bd (forge database pool caps are config values) modules/gitea/files/server.env has three config: lines (database-max-open-connections, database-max-idle-connections, database-connection-lifetime). mesh-provisioner-client's TestNatsAHandedGiteasServerEnvIsFilledAsTheControllerComposedIt (bound_gitea_test.go:67) and TestNatsAMovedGiteaAndClaudeCodeHaveTheirFilesFilledFromTheBus (moved_forge_test.go:103) fail on main: their test bus gave mesh-config no values for those keys, so the client rightly never writes server.env, and the expected text lacked the lines.

What changes: test files of modules/mesh-provisioner-client only. A helper giteaPoolConfig writes the three keys into mesh-config_resolved as mesh-config does (40, 10, "5m", the types modules/gitea/module.json:190-209 declares; read live with mesh-config.config); giteaRun and the moved-forge test call it; giteaBoundExpected gains the three filled lines and the template's two comment lines (the filler renders the whole text). No test function added, no module.json change.

Quick or thorough: quick, the instance. The kind (second today, after #646 broke gitea's whole test): a merge under 1966 broke another module's test on main because no check gates merges now; its thorough fix is decided and in the joint milestone (ADR 0666 §11, merge-pull refuses a head whose required step failed). Noted on issue 1966 (note 179).

Proof: beside mesh-sdk 31760596, mesh-contracts f5d76e84, mesh-delivery a6bba2af (as the checker places siblings), both tests fail without the change (each waits its 30 s) and pass with it; the module's package passes go vet both ways and go test with and without -tags integration (13 s, 17 s). TestNatsAnEntryWithoutAtOrFromLeavesGiteasServerEnvUnfilled still fails only for what it tests.

Reviewed-by: mesh-code-reviewer@shanks (for lead-47@shanks) at b373e34b29 (round 1: approve)

https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ

Issue 1966 (notes 178, 179). Since 4becc7bd (forge database pool caps are config values) modules/gitea/files/server.env has three config: lines (database-max-open-connections, database-max-idle-connections, database-connection-lifetime). mesh-provisioner-client's TestNatsAHandedGiteasServerEnvIsFilledAsTheControllerComposedIt (bound_gitea_test.go:67) and TestNatsAMovedGiteaAndClaudeCodeHaveTheirFilesFilledFromTheBus (moved_forge_test.go:103) fail on main: their test bus gave mesh-config no values for those keys, so the client rightly never writes server.env, and the expected text lacked the lines. What changes: test files of modules/mesh-provisioner-client only. A helper giteaPoolConfig writes the three keys into mesh-config_resolved as mesh-config does (40, 10, "5m", the types modules/gitea/module.json:190-209 declares; read live with mesh-config.config); giteaRun and the moved-forge test call it; giteaBoundExpected gains the three filled lines and the template's two comment lines (the filler renders the whole text). No test function added, no module.json change. Quick or thorough: quick, the instance. The kind (second today, after #646 broke gitea's whole test): a merge under 1966 broke another module's test on main because no check gates merges now; its thorough fix is decided and in the joint milestone (ADR 0666 §11, merge-pull refuses a head whose required step failed). Noted on issue 1966 (note 179). Proof: beside mesh-sdk 31760596, mesh-contracts f5d76e84, mesh-delivery a6bba2af (as the checker places siblings), both tests fail without the change (each waits its 30 s) and pass with it; the module's package passes go vet both ways and go test with and without -tags integration (13 s, 17 s). TestNatsAnEntryWithoutAtOrFromLeavesGiteasServerEnvUnfilled still fails only for what it tests. Reviewed-by: mesh-code-reviewer@shanks (for lead-47@shanks) at b373e34b29ec43722db6b04db78fe552fde62a20 (round 1: approve) https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ

mesh-tools added 1 commit 2026-11-10 11:59:53 +00:00

Fill gitea's pool keys in the provisioner client's gitea tests (issue 1966)

mesh/check checking b373e34b29ec on 3c0b4d29a8ea

mesh/delivery merged: source.main moved to novox/mesh-catalog@3c0b4d29a8ea

Details

b373e34b29

The template gained three config keys the tests' bus held no values for, so the client rightly never wrote server.env.

Claude-Session: https://claude.ai/code/session_01JG73KEq4zoDfM9WtYpcQgZ

mesh-tools commented

2026-11-10 11:59:56 +00:00

Delivery novox/mesh-catalog@b373e34b29ec — merged since 2026-11-10T14:49:04Z

Source — source.main moved to novox/mesh-catalog@3c0b4d29a8. Whether it reached the machines is mesh-delivery-view's, derived from source.main, the wants and the haves.

Transitions

2026-11-10 11:59 (new) → proposed (announced): novox/mesh-catalog#653's head announced
2026-11-10 11:59 proposed → ready (accepted): a verdict names this commit and the trunk's head as its trunk commit, the gate passed or warned, and the repository's own check did not fail; or the check gates no merge (check-gates-merge false)
2026-11-10 14:49 ready → merged (merged): on the trunk its modules follow: merged there

The commit's note under refs/notes/mesh-plan keeps every transition of this change up to its merge: git log --notes=mesh-plan.

**Delivery** `novox/mesh-catalog@b373e34b29ec` — **merged** since 2026-11-10T14:49:04Z **Source** — source.main moved to novox/mesh-catalog@3c0b4d29a8ea. Whether it reached the machines is `mesh-delivery-view`'s, derived from `source.main`, the wants and the haves. **Transitions** - 2026-11-10 11:59 (new) → proposed (announced): novox/mesh-catalog#653's head announced - 2026-11-10 11:59 proposed → ready (accepted): a verdict names this commit and the trunk's head as its trunk commit, the gate passed or warned, and the repository's own check did not fail; or the check gates no merge (check-gates-merge false) - 2026-11-10 14:49 ready → merged (merged): on the trunk its modules follow: merged there The commit's note under `refs/notes/mesh-plan` keeps every transition of this change up to its merge: `git log --notes=mesh-plan`.

mesh-tools merged commit 3c0b4d29a8 into main

2026-11-10 14:49:02 +00:00

mesh-tools deleted branch fix/1966-provisioner-client-pool-config

2026-11-10 14:49:02 +00:00

Sign in to join this conversation.

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: novox/mesh-catalog#653

Fill gitea's pool keys in the provisioner client's gitea tests, which fail on main since 4becc7bd (issue 1966) #653

Fill gitea's pool keys in the provisioner client's gitea tests, which fail on main since `4becc7bd` (issue 1966) #653