Skip to content

Dissolve std.baseboard's minted BaseboardManufacturer onto Vendor<Domain> - #7189

Merged
briansrls merged 4 commits into
mainfrom
session/keen-bat-128
Jul 25, 2026
Merged

briansrls merged 4 commits into
mainfrom
session/keen-bat-128

Conversation

@briansrls

@briansrls briansrls commented Jul 25, 2026 •

Copy link
Copy Markdown
Contributor

Summary

std.baseboard re-coined ASRock Rack as a freshly-minted single-variant
BaseboardManufacturer enum (dispatched via a match fn) instead of citing it
on the single Vendor<Domain> authority (DESIGN §3) all other hardware
vendors already ground on (Vendor<Hardware> — see CpuModelCatalogRow,
GpuFacts, DramModuleCatalogRow). Per
docs/plans/extdeps-std-grounding-inventory.md §1/§2 this is an s3-fork.

Because std cannot import extdeps.vendor (the layer DAG points imports
toward std, never away from it), grounding the vendor field forces the
whole baseboard domain shape — BaseboardModel, BaseboardCatalogRow,
ServerBaseboard — out of std/baseboard.dag and into
extdeps/boards/types.dag, matching the existing CPU/GPU/DRAM precedent
where the catalog-row shape lives in extdeps beside its cited vendor rows,
and std keeps only the frameworks (measure, types) those shapes build
on. BaseboardCatalogRow.vendor is now a plain Vendor<Hardware> field
(cited extdeps/vendor/asrock_rack.dag), not a derived-by-match enum —
dropping BaseboardManufacturer and baseboard_manufacturer() entirely.

Consumers updated: dag/extdeps/boards/asrock_rack.dag (catalog row now
carries vendor: asrock_rack), dag/gunbc/fleet_intent.dag (import
repointed), dag/test/claim/bmc_factory_credential_witness_test.dag
(assertion now compares .catalog.vendor against the cited vendor entity
instead of matching the deleted enum).

Test plan

  • claim_batch --source-root dag --source-root src/v2 --entry dag/test/claim/bmc_factory_credential_witness_test.dag --function bmc_factory_credential_holds → PASS bmc_factory_credential_holds (green-by-execution, exercises the new vendor field end to end).
  • claim_batch ... --entry dag/test/claim/fleet_intent_network_witness_test.dag --function fleet_intent_network_container_to_bmc_allowed → PASS (proves fleet_intent.dag's ServerBaseboard consumers, incl. srv1_baseboard..srv4_baseboard, still resolve and compile clean against the moved module).
  • cargo build -p v1-compiler --bin claim_executor / --bin claim_batch → clean.

…ain>

std.baseboard re-coined ASRock Rack as a freshly-minted single-variant
BaseboardManufacturer enum instead of citing it on the single Vendor<Domain>
authority (DESIGN §3), and std cannot import extdeps.vendor (layer DAG
points toward std), so the whole baseboard domain model — BaseboardModel,
BaseboardCatalogRow, ServerBaseboard — moves to extdeps/boards/types.dag
beside its cited catalog row, matching the CpuModelCatalogRow/GpuFacts
precedent (vendor is a plain Vendor<Hardware> field, not a derived-by-match
enum). Adds extdeps/vendor/asrock_rack.dag as the cited entity.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
@gunbai-bot
gunbai-bot Bot force-pushed the session/keen-bat-128 branch from 00034c1 to c5c7f66 Compare July 25, 2026 00:45
@gunbai-bot
gunbai-bot Bot marked this pull request as ready for review July 25, 2026 00:45
@gunbai-bot gunbai-bot Bot changed the title Extdeps/std §3-fork: consolidate minted-vendor enums onto the single Vendor<Domain> authority (DESIGN §3 exemplar) — per docs/plans/extdeps-std-grounding-inventory.md §1/§2: dissolve std.baseboard.BaseboardManufacturer (=AsrockRack) and any sibling freshly-minted single-vendor enums into the generic Dissolve std.baseboard's minted BaseboardManufacturer onto Vendor<Domain> Jul 25, 2026
@cursor

cursor Bot commented Jul 25, 2026

Copy link
Copy Markdown

Bugbot is not enabled for your account, so this pull request was not reviewed.

Enable Bugbot in the Cursor dashboard to get automatic reviews on future PRs.

@briansrls
briansrls merged commit c57c6e8 into main Jul 25, 2026
5 checks passed
@briansrls
briansrls deleted the session/keen-bat-128 branch July 25, 2026 05:48
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