type: decision
status: active
timestamp: 2026-07-02
tags: [ci, dagger, pipeline, migration]

Dagger keep + full sweep — confirmed 2026-07-02

After a re-grill on Dagger's disadvantages (Docker dep, cold start, ecosystem loss), the 2026-07-01 pipeline-stack decision is confirmed. Full retro-migration of all 20 own/* repos proceeds. Local pipeline runs justify Dagger over `act`.

Dagger confirmed 2026-07-02

The re-grill

User questioned Dagger 2026-07-02 after starting a retro-migration. Steel-manned both sides. Locked answers:

Question	Answer
Do you actually run pipelines locally on every commit?	Yes
Realistic chance of moving off GHA in 24 months?	Yes
Worth 4-8h to learn Dagger?	Yes
Full sweep or template-only?	Full sweep of all 20 own/* repos

Why Dagger, factually

GHA has no first-party local runner. act (nektos/act) is community; ~85% fidelity. Every OIDC-based action fails locally (deploy-pages, cloudflare wrangler on some configs).
Dagger runs 100% same code locally + in CI. Local dagger call ci = CI dagger call ci. This matches “every commit locally” workflow.
Portability is realistic hedge. GHA pricing changed 2024. Codeberg is our mirror-of-nine. GitLab is a fallback. Dagger runs unchanged on all.
Type-safe pipelines refactor cleanly. TS compiler catches broken shared helpers before push.

What was rejected

Reverting Dagger + returning to pure GHA-YAML. Considered. Rejected because local-run is real requirement.
Template-only (new repos start with Dagger, existing keep YAML). Considered. Rejected — divergence between old and new repos = drift-in-perpetuity.

Migration scope

All 20 own/* repos: sweep to Dagger-via-reusable-workflow model per reusable-workflows-layered-2026-07-02.
Forks (frk/*): SKIP. Per no-fork-divergence, forks stay byte-identical to upstream. If we want Dagger there, file an upstream PR — each upstream decides.
Scorecard, CodeQL, deploy-pages OIDC workflows: KEEP native. GH-integrated features Dagger can’t replicate.

Cross-refs

Edit on GitHub · Back to index