# Synthesis related-work ledger

Status: paper-local author ledger

Date: 23 July 2026

This ledger records which established results and system capabilities the
synthesis may attribute to prior work. It also records the boundary that the
manuscript must preserve. A citation supports a statement only under the
source's model and assumptions. It is never evidence that CatDB implements
the cited capability.

| Area | Primary or official keys | Established material used by the synthesis | CatDB boundary |
|---|---|---|---|
| Logical and conceptual data models | `Codd1970RelationalModel`, `Chen1976EntityRelationship` | Logical/physical separation and conceptual entities, relationships, and attributes predate CatDB. | CatDB may not claim first invention of a semantic or conceptual layer. |
| Categorical knowledge representation | `SpivakKent2012Ologs` | Ologs use objects, functional aspects, and commuting diagrams for knowledge representation. | Categories and functors as knowledge structures are established. |
| Functorial migration | `Spivak2012Functorial`, `SpivakWisnesky2015Relational` | Schema functors induce migration operators with distinct variance; selected fragments have relational foundations. | CatDB must name the data operator and cannot infer flow from schema-arrow direction. |
| Algebraic databases and integration | `SchultzWisnesky2017AlgebraicIntegration`, `SchultzEtAl2017AlgebraicDatabases`, `CQL2026` | Algebraic integration, categorical database semantics, and executable categorical workflows are established. | The candidate delta is operational composition and evidence discipline, not the categorical foundation. |
| Model management | `BernsteinMelnik2007Model` | Models and richer mappings can be first-class values with management operators. | "Mappings are first class" is not a novelty claim. |
| Data exchange and composition | `FaginEtAl2005DataExchange`, `FaginEtAl2005Composition`, `FaginEtAl2009Clio` | Data exchange, solutions, certain answers, mapping composition limits, and generated transformations are established. | CatDB must compare its admitted mapping language and generated plans directly. |
| Federation and mediation | `ShethLarson1990Federated`, `Wiederhold1992Mediators`, `Lenzerini2002DataIntegration`, `CalvaneseEtAl2017Ontop` | Federated systems, mediators, GAV/LAV integration, rewriting, and ontology-based access are prior art. | A shared semantic domain and mapping-aware query rewriting are not new by themselves. |
| ETL operations | `VassiliadisEtAl2001Arktos` | ETL modeling, control, and execution include physical and operational work. | CatDB must retain extraction, cleaning, transfer, loading, scheduling, recovery, and exception costs. |
| RDF, OWL, GQL, and semistructured models | `W3CRDF11Concepts`, `W3COWL2Overview`, `ISOIEC39075_2024`, `AlsubaieeEtAl2014AsterixDB` | Graph, ontology, property-graph, and typed semistructured systems have native semantics. | CatDB may connect models only with explicit preservation and residual reports. |
| Identity resolution | `FellegiSunter1969`, `BenjellounEtAl2009` | Record linkage and entity resolution treat identity as evidence-bearing inference and reconciliation. | Similarity or schema typing alone cannot authorize canonical identity. |
| Provenance | `buneman2001whywhere`, `green2007semirings`, `lebo2013provo` | Why/where provenance, provenance semirings, and PROV interchange are established. | CatDB's candidate boundary is cross-layer versioned composition, not first invention of provenance. |
| Bidirectional transformation | `foster2007combinators` | Lenses supply laws for bidirectional transformations and view update. | A readable mapping is not automatically a safe write path. |
| Dataset versioning | `MaddoxEtAl2016Decibel` | Relational dataset branching and version management are established. | Cross-object semantic versioning is the candidate question. |
| Query optimization | `SelingerEtAl1979AccessPath`, `Graefe1994Volcano`, `BegoliEtAl2018Calcite` | Cost-based access-path selection, extensible execution, and heterogeneous query frameworks are established. | CatDB is not a new general optimizer and has no current optimizer result. |
| Incremental maintenance | `GuptaMumick1995`, `McSherryEtAl2013`, `BudiuEtAl2023` | Materialized-view maintenance, differential dataflow, and DBSP provide mature incremental techniques. | CatDB's proposed semantic invalidation layer needs implementation and comparison. |
| Adaptive and learned planning | `LeisEtAl2015`, `AvnurHellerstein2000`, `MarcusEtAl2021Bao` | Optimizer error, adaptive processing, and practical learned optimization are established baselines. | Semantic metadata must improve constrained decisions after overhead is counted. |
| Consistency and convergence | `HerlihyWing1990Linearizability`, `GilbertLynch2002CAP`, `ShapiroEtAl2011CRDT`, `AmelootEtAl2013Transducers`, `BailisEtAl2014Coordination` | Linearizability, CAP, CRDT convergence, monotonicity, and coordination-avoidance boundaries are established. | Category theory does not remove communication, consensus, or noncommutativity. |
| Mechanized database semantics | `ChuEtAl2017HoTTSQL` | Machine-checked SQL semantics and rewrites predate CatDB. | CatDB may claim only exact named Lean declarations under audited assumptions. |

## Internal evidence inputs

`CatDBP01` through `CatDBP18` are repository artifacts, not external prior
literature. The manuscript labels P1 through P18 as internally accepted paper
artifacts. An internally accepted paper model is not a runtime result and
cannot supply an absent implementation or experiment.

## Mandatory comparison dimensions

Any future novelty review must compare:

1. admitted schema and mapping language;
2. operator direction and mapping composition closure;
3. coverage, partiality, loss, ambiguity, and identity authority;
4. semantic and physical versioning;
5. virtual, federated, generated, and materialized realizations;
6. provenance granularity and completeness;
7. workspace, merge, and distributed operation semantics;
8. implementation surface and formal coverage;
9. correctness, maintenance, and runtime baselines; and
10. measured metadata and governance overhead.

## Citation discipline

- All external capability statements are `supported by prior literature`.
- Official standards and project documentation support only their documented
  syntax or behavior.
- No prior source establishes a CatDB implementation or systems result.
- All descriptions are paraphrases; the manuscript contains no source quote.