docs: composite orchestration and the catalog cross-walk move here from UDLM - #122
Merged
Merged
Conversation
…om UDLM Two sections of udlm/docs/spec/foundations/template-composition-model.md described what the control plane DOES rather than what a composite IS, which puts them on this side of ADR-008. "What DCM Does vs What the Registering Provider Does" is a table whose left column is entirely control-plane behaviour — placement, dispatch sequencing, status computation, compensation execution, audit aggregation. "Interop Note — DCM Control-Plane Catalog Model" cross-walks this project's catalog model to UDLM's field by field, which is a statement about this project. Renumbered to stand alone: the sections carried §4 and §2.5 from their old document and read as gaps here. The data model stays in UDLM and is cited, not restated — constituent declaration, provided_by, the dependency graph, required_for_delivery, compensation declaration, the four-state representation, composition visibility, rehydration, nesting depth, the registration contract, and the CMP-* rules. Where the two disagree, UDLM is authoritative, and the document says so. The UDLM side leaves a pointer here in the same change set. All 6 validate.yml steps pass. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
croadfeldt
added a commit
to croadfeldt/udlm
that referenced
this pull request
Aug 9, 2026
The audit (#460) read all 522 lines section by section against three questions: is it stated elsewhere, does it describe something the model has moved past, and does it belong to this repo. Nothing was stale — scanning for provider_extensions, a stored entity_type, relationship_type, the six-field reference and a stored nature found zero hits. So the question was only whether each section belongs here. CUT — §1.2 "Why This Model Is Correct" ("Governance stays with DCM · Policy stays with DCM · Audit stays with DCM · Recovery stays with DCM") is reassurance that a principle holds, and §1.3 "The Practical Meaning" narrates what DCM then does, restating §7. Argument and runtime narration; AGENTS.md already excludes both. MOVED to dcm — §2.5's field-by-field cross-walk to the control plane's own catalog model, and §4's table whose left column is entirely control-plane behaviour: placement, dispatch sequencing, status computation, compensation, audit aggregation. That is the ADR-008 line. Both now live at dcm/docs/specifications/dcm-composite-orchestration.md (croadfeldt/dcm PR #122) and each leaves a pointer plus the one sentence a reader here still needs. TRIMMED — §7 listed eight pipeline steps of which one, expansion, is a data-model event. It now states that one and what it implies (Requested carries enumerated constituents and no dependency-resolved values, because a payload cannot bind from a sibling that is not yet realized) and cites request-realization for the rest. Kept as a data rule rather than sequencing: a policy decision rejecting any constituent rejects the composite. KEPT after checking, not by default — §6 rehydration reads as a candidate for merging into four-states.md §5, which does describe Faithful and Provider-Portable modes. It is not a duplicate: provider_portable, historical_exact and historical_portable appear in no other file in the repo. Merging it would have deleted the only definition of three policy values. 522 -> 466 lines. All 48 validate.yml steps pass; 238 files scanned, 0 broken links. Refs #460. Pairs with croadfeldt/dcm#122. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
croadfeldt
added a commit
to croadfeldt/udlm
that referenced
this pull request
Aug 9, 2026
The audit (#460) read all 522 lines section by section against three questions: is it stated elsewhere, does it describe something the model has moved past, and does it belong to this repo. Nothing was stale — scanning for provider_extensions, a stored entity_type, relationship_type, the six-field reference and a stored nature found zero hits. So the question was only whether each section belongs here. CUT — §1.2 "Why This Model Is Correct" ("Governance stays with DCM · Policy stays with DCM · Audit stays with DCM · Recovery stays with DCM") is reassurance that a principle holds, and §1.3 "The Practical Meaning" narrates what DCM then does, restating §7. Argument and runtime narration; AGENTS.md already excludes both. MOVED to dcm — §2.5's field-by-field cross-walk to the control plane's own catalog model, and §4's table whose left column is entirely control-plane behaviour: placement, dispatch sequencing, status computation, compensation, audit aggregation. That is the ADR-008 line. Both now live at dcm/docs/specifications/dcm-composite-orchestration.md (croadfeldt/dcm PR #122) and each leaves a pointer plus the one sentence a reader here still needs. TRIMMED — §7 listed eight pipeline steps of which one, expansion, is a data-model event. It now states that one and what it implies (Requested carries enumerated constituents and no dependency-resolved values, because a payload cannot bind from a sibling that is not yet realized) and cites request-realization for the rest. Kept as a data rule rather than sequencing: a policy decision rejecting any constituent rejects the composite. KEPT after checking, not by default — §6 rehydration reads as a candidate for merging into four-states.md §5, which does describe Faithful and Provider-Portable modes. It is not a duplicate: provider_portable, historical_exact and historical_portable appear in no other file in the repo. Merging it would have deleted the only definition of three policy values. 522 -> 466 lines. All 48 validate.yml steps pass; 238 files scanned, 0 broken links. Refs #460. Pairs with croadfeldt/dcm#122. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Paired with croadfeldt/udlm#460 (the audit) — this is the receiving half; the UDLM PR removes these sections and leaves a pointer here.
Why they move
Two sections of
udlm/docs/spec/foundations/template-composition-model.mddescribed what the control plane does rather than what a composite is. That is the ADR-008 line.What stays in UDLM
Everything about the composite as data — constituent declaration,
provided_by, the dependency graph,required_for_delivery, compensation declaration, the four-state representation, composition visibility, rehydration, nesting depth, the registration contract, and theCMP-*rules.The new document cites that and restates none of it, and says explicitly: where the two disagree, UDLM is authoritative.
Note
Renumbered to stand alone — the sections carried
§4and§2.5from their old document and read as gaps here.All 6
validate.ymlsteps pass.🤖 Generated with Claude Code