chore: remove stray empty uv.lock - #146
Conversation
uv.lock was committed by accident in #125 (a media-captions PR) on 2026-06-17. It is a 3-line stub: version = 1 revision = 3 requires-python = ">=3.12" Zero [[package]] entries — it locks nothing. It also declares requires-python >=3.12, contradicting the project's ^3.10. Nothing reads it: no workflow, script, or doc references uv.lock or any uv command. This project is Poetry-managed. It is worth removing on those grounds alone, but there is also a suspected link to a stuck dependency graph. GitHub has not re-parsed poetry.lock since 2026-06-17, and the timing brackets this file: 16:47 #124 poetry.lock changed -> parsed OK (still what the dependency graph displays today) 18:42 #125 uv.lock added Aug 12 #131 poetry.lock changed -> ignored Aug 18 #145 poetry.lock changed -> ignored uv.lock is the only manifest-like file added in that window. As a result the graph still reports pre-June versions (torch 2.7.1, cryptography 46.0.3, gitpython 3.1.45) and is missing wandb, pyarrow and optuna, which #131 added in August. 62 Dependabot alerts remain open against versions no longer present anywhere on main. This is a timing correlation, not a proven cause — GitHub's docs do not list uv.lock among supported Python manifests, so there is no documented mechanism for it to interfere. Deleting it is correct housekeeping either way, and it is the one lever the evidence points at. Deliberately scoped to this single deletion: touching poetry.lock in the same commit would itself force a re-parse and confound the test. If the graph refreshes after this merges, uv.lock was the cause. If not, this is a GitHub-side ingestion failure and the next step is a support ticket. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
Important Review skippedReview was skipped due to path filters ⛔ Files ignored due to path filters (1)
CodeRabbit blocks several paths by default. You can override this behavior by explicitly including those paths in the path filters. For example, including ⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Plus Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Deletes
uv.lock, which was committed by accident in #125 (a media-captions PR) on 2026-06-17.It is a 3-line stub with zero package entries — it locks nothing:
It also declares
requires-python = ">=3.12", contradicting the project's^3.10. Nothing reads it — no workflow, script, or doc referencesuv.lockor anyuvcommand. This project is Poetry-managed.Suspected link to the stuck dependency graph
GitHub has not re-parsed
poetry.locksince 2026-06-17, and the timing brackets this file exactly:poetry.lockuv.lockpoetry.lockpoetry.lockuv.lockis the only manifest-like file added in that window.Consequences visible today: the graph reports pre-June versions (
torch 2.7.1,cryptography 46.0.3,gitpython 3.1.45) and is missingwandb,pyarrow, andoptuna, which #131 added in August. 62 Dependabot alerts remain open against versions that exist nowhere onmain— #145 patched all of them.This is a timing correlation, not a proven cause. GitHub's supported ecosystems docs do not list
uv.lockamong Python manifests, so there is no documented mechanism for it to interfere. The deletion is correct housekeeping regardless, and it is the one lever the evidence points at.How to tell whether this worked
After merging, wait a few minutes for the dependency graph to re-scan, then run:
There are exactly two outcomes, and both are informative:
Count drops toward 0 →
uv.lockwas jamming dependency-graph ingestion, and the June timing was the tell. The 62 alerts were phantoms matched against a stale June snapshot, and they clear on their own once the graph re-parsespoetry.lock.Count stays 62 → the correlation was coincidence. Deleting the file was still worth doing on its own merits, and the stuck graph is a GitHub-side bug. The support ticket is then well-evidenced:
poetry.lockwas last parsed on Jun 17 2026; subsequent changes ine014ee4(Aug 12) and68bb94c(Aug 18) were both ignored; andwandb,pyarrow, andoptuna— added to the lockfile by feat(migrate): wandb → Pluto historical data migration #131 in August — are absent from the graph entirely.Either way, no vulnerable dependency exists on
main. Every one of the 62 open alerts was checked against the lockfile actually onmain: 0 are still vulnerable, 62 are already patched. This is a reporting problem, not a security exposure.Scope
Deliberately limited to this single deletion. Touching
poetry.lockin the same PR would itself force a re-parse and confound the test — we would not learn which change was responsible.Separately still outstanding (not fixed here, to keep this clean):
poetry.lockcontains onemarkers = "<empty>"entry fornvidia-cusparse, which is not valid PEP 508. Worth cleaning up on its own.🤖 Generated with Claude Code