Skip to content

fix(build): the live proof's comment names the mechanism the drive actually runs #41

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
github-actions merged 1 commit into from
Oct 30, 2026
Merged

fix(build): the live proof's comment names the mechanism the drive actually runs #41

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
21 changes: 10 additions & 11 deletions candy/plugin-build/buildkit_live_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,13 +4,11 @@ package build
// `buildctl`, and a real image bridged into the run engine's store.
//
// ⚑ Why this lives in the module rather than as an assembled `charly box build`: the plugin is
// COMPILED INTO charly, so assembling charly requires every pinned plugin to compile against the
// new contract — and they have not all migrated yet (measured: `plugin-box` v0.2026272.2039 fails
// on `spec.BuildRequest.PodmanJobs`, `rt.BuildEngine`, `spec.BoxConfig.PodmanJobs`). A live proof
// that has to wait for four other repos would prove nothing about THIS change, so the drive is
// driven directly here: `driveConfig.buildImage` is the whole changed path — daemon ensure, the
// buildctl argv, the OCI export, and the bridge (load → tag → verify) — and it is invoked exactly
// as `runBoxBuild` invokes it.
// COMPILED INTO charly, so an assembled build couples this proof to every other pinned plugin's
// release — the assembled build answers "does the whole product assemble", while this proof
// answers "does THIS drive work". So the drive is driven directly here: `driveConfig.buildImage`
// is the whole changed path — daemon ensure, the buildctl argv, the OCI export, and the bridge
// (annotate → store → load → verify) — and it is invoked exactly as `runBoxBuild` invokes it.
//
// It SKIPS VISIBLY when the toolchain is absent (R8), and is never mocked: where the tools exist
// it runs a real build.
Expand Down Expand Up @@ -68,11 +66,12 @@ func TestBuildImage_LiveBuildsAndBridgesIntoTheRunEngine(t *testing.T) {
}

// ⚑ THE BRIDGE'S NAMED DUTY, asserted where it matters: the reference must RESOLVE in the RUN
// engine's store. A load returns success while landing an OCI-exported image untagged, so this
// is the assertion that fails if the tag re-application is ever dropped — the failure mode the
// plan calls out as breaking every downstream verb.
// engine's store. A load returns success while landing an image the engine cannot resolve by
// reference, so this is the assertion that fails if the export's NAMING ANNOTATION is ever
// dropped — the failure mode imagestore_live_test.go proves from the other side, where an
// archive with the annotation stripped imports DANGLING.
if !kit.LocalImageExists("podman", tag) {
t.Fatalf("the built image is not resolvable as %s in the run engine's store — the tag duty failed", tag)
t.Fatalf("the built image is not resolvable as %s in the run engine's store — the naming annotation did not take", tag)
}

// And the READ PATH the bridge exists for: labels, read through the run engine, which shares
Expand Down