Problem
PR #53's summary table frames finding F2 as Critical — Fixed, but in the default configuration (JOB_STORE_BACKEND=memory, ENQUEUE_ENABLED=false) submitted jobs still never complete end-to-end. The PR delivers the wiring to turn the pipeline on; it does not turn it on.
This is the correct backwards-compatible choice (new behavior is gated off so existing behavior is byte-identical), but the changelog/severity language reads as though merging #53 yields a working pipeline, which it does not by itself.
Requested change
Soften the F2 framing in CHANGELOG.md / release notes to something like: "end-to-end job processing is now wireable via configuration (redis backend + enqueue_enabled + a running ARQ worker)" rather than "Fixed", so the next operator does not assume the default config processes jobs.
Requires (and is gated by) the atomic-update fix in the companion issue before the redis backend is enabled in any shared deployment.
Problem
PR #53's summary table frames finding F2 as Critical — Fixed, but in the default configuration (
JOB_STORE_BACKEND=memory,ENQUEUE_ENABLED=false) submitted jobs still never complete end-to-end. The PR delivers the wiring to turn the pipeline on; it does not turn it on.This is the correct backwards-compatible choice (new behavior is gated off so existing behavior is byte-identical), but the changelog/severity language reads as though merging #53 yields a working pipeline, which it does not by itself.
Requested change
Soften the F2 framing in
CHANGELOG.md/ release notes to something like: "end-to-end job processing is now wireable via configuration (redisbackend +enqueue_enabled+ a running ARQ worker)" rather than "Fixed", so the next operator does not assume the default config processes jobs.Requires (and is gated by) the atomic-update fix in the companion issue before the redis backend is enabled in any shared deployment.