villebro commented on code in PR #44250:
URL: https://github.com/apache/superset/pull/44250#discussion_r4057521093
##########
.github/workflows/tag-release.yml:
##########
@@ -55,8 +55,7 @@ jobs:
contents: write
strategy:
matrix:
- build_preset:
- ["superset", "dev", "lean", "websocket", "dockerize", "py311",
"py312"]
+ build_preset: ["superset", "dev", "lean", "py311", "py312"]
Review Comment:
Closing the loop on the second half of this — exercising the ref swap in the
command check. That was still outstanding: the check asserted the release
*tags*, but never the checkout-master → snapshot-helper → checkout-release-ref
sequence the release job actually performs.
Added `assert_release_ref_swap` in f32361cfc2. Rather than fetching a real
tag, it simulates the swap directly against the property the release path
depends on — that the helper resolves `Dockerfile` from the working tree, not
from wherever the script lives. It snapshots the helper, points a synthetic
older tree at `ARG PY_VER=3.11.13-slim-bookworm`, and asserts release mode
picks up *that* tree's `PY_VER` and leaves its Chromium default alone, while
`ci` mode against the same tree still gets the explicit
`INCLUDE_CHROMIUM=false`, so the two modes cannot silently converge.
I checked it has teeth by breaking it both ways: resolving the Dockerfile
relative to the script (emits master's `3.11.14-slim-trixie` instead of the
release tree's value) and applying the CI override in release mode. Both are
rejected.
Also worth recording here, since it was the riskiest unknown in this area: I
verified the `DOCKERHUB_TOKEN=` blanking is safe on the `--push` path, which no
PR build exercises. In supersetbot 0.5.45 the variable is read once, at
`src/docker.js:125`, solely to decide whether to append `--cache-to`; it plays
no part in authentication. Push auth comes from `docker/login-action` in
`setup-docker` writing to the Docker config, so hiding the env var cannot break
release or master publishing.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
