Skip to content

docs: cudapp.lock is not on omega-mv main; name the real pins - #212

Merged
aien-spark merged 1 commit into
mainfrom
docs/cudapp-lock-status
Oct 31, 2026
Merged

aien-spark merged 1 commit into
mainfrom
docs/cudapp-lock-status

Conversation

@aien-spark

Copy link
Copy Markdown
Collaborator

Doc-only. AUDIT_SCOPE.md and docs/START_HERE.md described cudapp.lock in omega-mv as existing. It does not exist.

Evidence (MEASURED, 2026-10-31)

  • omega-mv main d39450a23b5936a0aafd35cc23ca70e5468fd4c0: git ls-tree -r --name-only origin/main | grep -i cudapp lists no cudapp.lock; git grep -E 'cudapp\.lock|check-cudapp|cudapp-stage' origin/main has 0 hits.
  • cudapp.lock (content 6bfd376064650e2188b6a08d7d0222074a5e0d1b), mk/cudapp-consumer.mk (check-cudapp) and tools/cudapp-stage.sh exist only on the open draft build(native): recover canonical CUDApp consumer cutover omega-mv#924 (head 87aba69488118ac230d3d1f37cc1b07570243d2c). aien-sovereign-core-mv#784 (draft) proposes another cudapp.lock; its main has none (contents API 404).
  • This matches what README.md (line 469), AGENTS.md (line 16), Makefile (line 13) and docs/NAMING.md (line 4) already say: the pin is planned, not created.
  • Every file the new AUDIT_SCOPE text names exists on CUDApp main edb9a70361c369180326c6836ab56da0e54e5c5b (git ls-files): om/omega-bridge.lock, om/pipeline/dependencies.lock, native/learn53/omega653.lock, native/learn53/moe-closure.lock, tools/c5/omc.pin.

Change

AUDIT_SCOPE.md line 141, before:

- Omega builds the same sources at a pinned commit (`cudapp.lock` in omega-mv).

after:

- Omega does not build these sources at a pin yet. omega-mv main has no `cudapp.lock`; it still
  builds its own diverged copy of the GPU code (#106) and takes the driver client from physics-mv
  (#107). The lock, with its `check-cudapp` target and `tools/cudapp-stage.sh`, is proposed in
  the open draft omega-mv#924.

AUDIT_SCOPE.md line 135, before: `native/learn53/omega653.lock` is a sub-component lock. after: lists all five sub-component pins above.

docs/START_HERE.md lines 65-68, before:

Canonical consumer integration uses the existing single-commit `cudapp.lock`,
not a second artifact or qualification registry. The isolated Omega consumer
cutover builds an authenticated committed source archive and links its actual
callers to that canonical library:

after:

Canonical consumer integration is meant to use one single-commit `cudapp.lock`
in Omega, not a second artifact or qualification registry. That lock does not
exist on omega-mv main yet: `cudapp.lock`, the `check-cudapp` make target and
`tools/cudapp-stage.sh` are proposed in the open draft
[omega-mv#924](https://github.com/aien-dev/omega-mv/pull/924). That isolated
Omega consumer cutover builds an authenticated committed source archive and
links its actual callers to that canonical library; with #924 checked out:

Checks

  • crumb verify . (~/.local/bin/crumb): OK, 33 crumbs current, before and after.
  • Pinned crumb d0f0b95 crumb status .: 1 STALE (third_party/nvidia-open-580.173.02) both on origin/main and on this branch; pre-existing, not touched here. CUDApp CI does not run crumb verify.
  • No code, gate, workflow or evidence file changed. Builds/tests NOT_RUN (doc-only); CI is the check.

AUDIT_SCOPE.md and docs/START_HERE.md described cudapp.lock in omega-mv as
existing. omega-mv main has no cudapp.lock; the lock, check-cudapp and
tools/cudapp-stage.sh are proposed in draft omega-mv#924. AUDIT_SCOPE.md
now also lists the sub-component pins that do exist in this repository.
@aien-spark
aien-spark merged commit c76412e into main Oct 31, 2026
15 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant