Appendix C — References

The spec concordance. The axioms below are the book’s external dependencies — deliberately its only ones. Four lists follow, kept separate per the discipline of Appendix A: the witnesses, the candidates, the prior art, and the works the audit examines.

Axioms — definitions used as premises:

definition	source	first used
I — URI syntax; decentralized minting via the authority component	RFC 3986	Ch 1; B.2
Req, Resp — the message form; the safe/unsafe method split	RFC 9110 §6, §9	Def. 1.1
representations reflect resource state over time	RFC 9110 §3.2	Ch 4 (τ)
validators Last-Modified, ETag	RFC 9110 §8.8	Prop. 4.5
the caching calculus	RFC 9111	Prop. 4.5, corollary
origins — I partitioned into parties’ regions	RFC 6454	(17.1); B.9
triple, graph, merge	RDF 1.1 Concepts (2014)	Ch 8
blank nodes as existentials	RDF 1.1 Semantics	Prop. 9.1
datasets and named graphs	RDF 1.1; TriG (2014)	Prop. 9.2
the selection algebra, denotationally	SPARQL 1.1 Query §18	Ch 8; Prop. 8.1
the delta on the wire	SPARQL 1.1 Update	Ch 8 (Prop. 7.1’s reveal)
documents as named graphs, read-write	SPARQL 1.1 Graph Store HTTP Protocol	Ch 8; Ch 17
canonical labeling of unnamed entities	RDFC-1.0 (2024)	Prop. 6.1; Prop. 9.1
tree transformation	XSLT (1999; 3.0, 2017)	Ch 8
presentation	CSS (1996)	Ch 8
forms as the write instrument	HTML: forms	Prop. 7.2

Currency — checked July 2026: RFC 3986 remains Internet Standard 66 — updated, never obsoleted. The update is BCP 190 (RFC 8820): guidance on URI ownership, with no change to syntax. That is B.2’s minting doctrine, in BCP form. RFC 9110 and 9111 are the current HTTP standards. RFC 6454 stands unrevised since 2011; the HTML Standard restates the same scheme–host–port tuple for browsers. RDF 1.2 (Candidate Recommendation, April 2026) preserves every definition cited above: data conforming to 1.1 remains conforming. Its headline addition, the triple term, is the annotation syntax Chapter 9 contrasts and scores. And RDF 1.2 keeps named graphs — the fourth position, Chapter 9’s structural prediction, survives another revision. SPARQL is cited at 1.1 throughout, the current Recommendation.

Witnesses — norms and independent results corroborating, never premises:

• Architecture of the World Wide Web, Volume One, W3C Recommendation, 2004 (TAG):
  • §2.5 URI opacity — Thm. 8.2’s genericity.
  • §3.5 available representations — S4, minus the intermediates, and (17.1)’s minting obligation.
  • §4.3 separation of content, presentation, interaction — Def. 4.3.
  • §4.4 link identification, Web-wide linking, hypertext links — Ch 10’s scoring.
  • §5.1 orthogonality — S2/S3.
• The Rule of Least Power, TAG finding, 2006 — prefer the least powerful language that suffices: the norm behind Ch 12’s scoring.
• R. T. Fielding, Architectural Styles and the Design of Network-based Software Architectures, dissertation, 2000 — Ch 4’s payoff: the positioning of this book as the second half of a derivation whose first half Fielding wrote, and the uniform interface’s four clauses (§5.1.5) typed there. Also the discarded hypermedia constraint in Ch 10, and the property list of Ch 11.
• M. Shapiro, N. Preguiça, C. Baquero, M. Zawirski, Conflict-free Replicated Data Types, 2011 — the independent derivation of the merge laws from replication pressure (Ch 5’s corroboration; B.4).
• httpRange-14, W3C TAG issue, resolved 2005 — the name/address distinction the model types apart (R3 vs. S4); Ch 18’s encoding choices.
• Cool URIs for the Semantic Web, W3C Interest Group Note, 2008 — the deployed encodings (fragment, 303) of that distinction.
• Pappus of Alexandria, Collection, Book VII — the classical statement of the twin method of analysis and synthesis; Chapter 2’s name for the book’s shape.
• I. Newton, Opticks, Query 31 — “the Investigation of difficult Things by the Method of Analysis ought ever to precede the Method of Composition”; Chapter 2.
• T. Berners-Lee, Information Management: A Proposal (CERN, 1989) — the origin memo, and Mike Sendall’s cover note “vague but exciting”; Chapter 1’s epigraph, and the browser-editor bootstrap of Chapter 18.
• V. Bush, As We May Think (The Atlantic, 1945); T. Nelson, Computer Lib / Dream Machines (1974) — association over hierarchy, the graph refusing the tree, stated before the web; Chapter 6.
• T. Berners-Lee, J. Hendler, O. Lassila, The Semantic Web (Scientific American, May 2001) — the agent-over-machine-readable-data scenario Chapter 22 derives; written as fiction, now falsifiable.
• J. E. Labra Gayo, E. Prud’hommeaux, I. Boneva, D. Kontokostas, Validating RDF Data (2017), foreword by D. Brickley and L. Miller — Chapter 9’s epigraph.
• D. McComb, Software Wasteland (Technics, 2018) and The Data-Centric Revolution (Technics, 2019) — the data-centric case (Semantic Arts) for data over application code; Chapter 19’s corollary reached from enterprise waste rather than derivation.
• J. Somers, The Coming Software Apocalypse (The Atlantic, September 2017) — code grown past comprehension in safety-critical systems, and the remedies that move engineers above it (model-based design, Lamport’s TLA+); Chapter 19’s liability, reported from the field.
• J. Rayfield, BBC World Cup 2010 dynamic semantic publishing (BBC Internet Blog, July 2010) and Sports Refresh: Dynamic Semantic Publishing (April 2012) — 700-plus World Cup pages, then ten thousand Olympic pages, generated from an RDF triple store; Chapter 21’s first mover on the publishing side.
• M. Jusevičius, A. Smirnovas, J. Šėporaitis, Graphity – A Generic Linked Data Platform (position paper, W3C Workshop on Linked Enterprise Data Patterns, December 2011) — the comics site Helt Normalt on RDF, SPARQL and XSLT, the codebase an order of magnitude smaller than the relational system it replaced, ontologies reused down to a zodiac vocabulary; Chapter 21’s first mover on the application side (the author’s, per Chapter 18).
• A. Singhal, Introducing the Knowledge Graph: things, not strings (Google, May 2012) — the announcement that made knowledge graph the industry’s name for graph-shaped state; Chapter 21’s dating of the term.
• J. Walker, Is Linked Data the Future of Data Integration in the Enterprise? (NXP Smarter World Blog, January 2013) — product data scattered across systems, converted to RDF, dereferenceable per product; “the Linked Data is the API”; Chapter 21’s first mover on the integration side.
• M. Stonebraker, I. F. Ilyas, Data Integration: The Current Status and the Way Forward (IEEE Data Eng. Bulletin, 2018) — silo counts at enterprise scale: GE’s ~75 procurement systems, Merck’s ~4,000 databases; Chapter 21’s head of the curve, measured.
• Palantir, Ontology (platform page; on Foundry’s marketing since 2018) and the S-1 registration statement (SEC, August 2020) — “a data model that reflects the real world,” the word filed with the regulator; not built on RDF; Chapter 21’s vocabulary crossover.
• Google, A reintroduction to our Knowledge Graph and knowledge panels (May 2020) — “over 500 billion facts about five billion entities”; proprietary, no public endpoint or dump; Chapter 21’s closed giant.
• Gartner, Top 10 Data and Analytics Technology Trends for 2021 (press release, March 2021) — “by 2025, graph technologies will be used in 80% of data and analytics innovations, up from 10% in 2021”; Chapter 21’s wave, dated by its analysts; the figure spans all graph models, not RDF alone.
• J. F. Sequeda, D. Allemang, B. Jacob, A Benchmark to Understand the Role of Knowledge Graphs on Large Language Model’s Accuracy for Question Answering on Enterprise SQL Databases (2023) — GPT-4 at 16.7% over enterprise SQL, 54.2% over the same data as an RDF graph; Chapter 21’s grounding motive, measured.
• Microsoft, Fabric IQ ontology (preview, announced November 2025) — “an ontology is a shared, machine-understandable vocabulary of your business”; Power BI’s datasets became semantic models in November 2023; neither on RDF; Chapter 21’s vocabulary crossover.
• The UniProt Consortium, sparql.uniprot.org (release 2026_02, June 2026) — 232.5 billion triples behind a free public SPARQL endpoint, RDF distribution since 2008; the largest publicly queryable knowledge graph; Chapter 21’s open giant.
• Wikidata, statistics (2026) — ~123 million items, ~18 billion triples across the query service and its scholarly split (May 2025); where part of Freebase settled; Chapter 21’s open giant.
• The Linked Open Data cloud (version of 15 June 2026; diagram CC BY) — 1,360 interlinked open datasets; Chapter 21’s map of the open graphs.
• D. Huynh, Freebase Parallax: A new way to browse and explore data (Metaweb, 2008) — set-based navigation over graph data: a set carried to the set it relates to. Google acquired Metaweb in 2010, and Freebase was a seed of the Google Knowledge Graph. Chapter 23’s set-to-set capability, demonstrated.
• D. Siegel, Pull: The Power of the Semantic Web to Transform Your Business (Portfolio, 2009) — the personal data locker (a life held as one owner-controlled graph) and intentcasting (Searls’s coinage, below); Chapter 23’s personal dataspace, reached from the demand side. Video: Personal Data Locker Vision.
• D. Searls, The Intention Economy: When Customers Take Charge (Harvard Business Review Press, 2012; the term intention economy coined 2006) — customers broadcasting qualified intent for vendors to answer (intentcasting); Chapter 23’s stating-a-need capability, from the demand side. Talk: The Intention Economy book talk, 2012.
• Apple, HyperCard (1987) — a running application reshaped in place, its stacks editable documents, with no rebuild and no redeploy; Chapter 23’s live-morphability lineage, lacking only a cross-party substrate.
• D. Shea, CSS Zen Garden (2003) — one unchanged HTML document redressed by hundreds of stylesheets; Chapter 23’s presentation independence (S1), demonstrated at the presentation layer alone.
• The Semantic Web stack (the “layer cake”), W3C, early 2000s — proof and trust sketched as its top layers and left unbuilt; Chapter 23’s provenance capability (R4), the layer the deployed web skipped.
• T. Berners-Lee, Cool URIs don’t change (W3C, 1998) — persistence of names asked of the web and mostly declined; Chapter 23’s data-outlives-the-application (R3).
• The mashup era — P. Rademacher’s HousingMaps (Google Maps × Craigslist, 2005) and ProgrammableWeb (2005) — combination without permission, until the APIs metered and re-siloed; Chapter 23’s applications-nobody-planned (R2/S3).

Candidates — specified, not standardized; Part V’s seams:

• WebID — W3C Incubator, 2005–; identity as a dereferenceable URI. The identity seam (Ch 18).
• WebAccessControl — the acl ontology, grown on the W3C wiki; adopted by Solid. The access-control seam (Ch 18).
• RDF/POST — community spec, AtomGraph, building on Sergei Egorov’s original draft. The form-native write seam (Ch 9; Ch 18).
• SaxonJS 3 with IXSL — XSLT 3.0 evaluated in the browser; the interactive extension binds events to template rules. The arrange seam, occupied client-side (Ch 9; Ch 17).

Prior art — the formal neighbors of Appendix B, cited so the boundaries can be checked:

• The RDF foundations school — closest formal treatment; stipulates triples, derives nothing about their necessity: C. Gutierrez, C. Hurtado, A. O. Mendelzon, J. Pérez, Foundations of Semantic Web Databases (PODS 2004; JCSS 77(3), 2011); M. Arenas, C. Gutierrez, J. Pérez, Foundations of RDF Databases (Reasoning Web 2009); S. Muñoz, J. Pérez, C. Gutierrez, Minimal Deductive Systems for RDF (ESWC 2007).
• E. L. Robertson, Triadic Relations: An Algebra for the Semantic Web (SWDB 2004, LNCS 3372) — asserts triples-minimal for RDF as motivation, underived; B.2 supplies the derivation.
• The Peirce line — the arity theorem’s owners: R. W. Burch, A Peircean Reduction Thesis (Texas Tech UP, 1991); H. Herzberger, Peirce’s Remarkable Theorem (1981); F. Dau, J. Hereth Correia, Two Instances of Peirce’s Reduction Thesis (ICFCA 2006); J. Hereth Correia, R. Pöschel, The Teridentity and Peircean Algebraic Logic (ICCS 2006; Semiotica 186, 2011); S. Koshkin, Is Peirce’s reduction thesis gerrymandered? (TCSPS 58(4), 2022) and the relational-database formalization (Logic J. IGPL, 2024/25 — ternarity(R) = n − 2).
• The dyadic-reduction results that B.2’s scope note answers: L. Löwenheim, Über Möglichkeiten im Relativkalkül (Math. Annalen 76, 1915); W. V. O. Quine, Reduction to a Dyadic Predicate (JSL 19(3), 1954).
• The coordination-free line — programs characterized, data model left open: J. M. Hellerstein, P. Alvaro, Keeping CALM: When Distributed Consistency Is Easy (CACM 63(9), 2020); T. J. Ameloot, F. Neven, J. Van den Bussche, Relational Transducers for Declarative Networking (JACM 60(2), 2013 — the CALM proof); N. Conway et al., Logic and Lattices for Distributed Programming (Bloom^L, SoCC 2012 — B.4’s deployed witness); S. Laddad et al., Keep CALM and CRDT On (VLDB 16, 2023).
• CRDTs meet RDF, as engineering: L.-D. Ibáñez, H. Skaf-Molli, P. Molli, O. Corby, Live Linked Data: Synchronising Semantic Stores with Commutative Replicated Data Types (IJMSO 8(2), 2013 — SU-Set; the (triple, id) tags it needs for deletion are the erasure argument surfacing as an engineering symptom).
• The genericity family: A. Chandra, D. Harel, Computable Queries for Relational Data Bases (JCSS 21(2), 1980); S. Abiteboul, V. Vianu, Queries and Computation on the Web (ICDT 1997; TCS, 2000); G. Fletcher, J. Van den Bussche, D. Van Gucht, S. Vansummeren, Towards a Theory of Search Queries (ICDT 2009); A. Hogan, Canonical Forms for Isomorphic and Equivalent RDF Graphs (TWEB 11(4), 2017 — IRIs rigid, blank nodes renameable: the practice that B.8’s definition departs from).
• M. Franklin, A. Halevy, D. Maier, From Databases to Dataspaces (SIGMOD Record 34(4), 2005) — the word’s database-literature sense, pay-as-you-go integration; disambiguated from Chapter 17’s web-native sense.
• Formal separation, one seam, 2004: T. Parr, Enforcing Strict Model-View Separation in Template Engines (WWW 2004) — definitions and theorems for model-view separation (“there was no formal definition of separation”). The paper excludes XSLT from its scope and characterizes template power via the Chomsky hierarchy. It is the priority citation for formalized separation, and the single-seam treatment the factorization generalizes.

Audited — works the book examines to score, chiefly in Part IV:

• XML 1.0 (1998); Namespaces in XML (1999); XPath 1.0 (1999); XSD (2001); xml:id (2005); XLink; XPointer; XQuery 1.0 and XPath 2.0 Formal Semantics (2007) — Ch 8’s rarity remark, Ch 10.
• JSON — RFC 8259 / ECMA-404; JSON Pointer — RFC 6901 (2013); JSONPath — RFC 9535 (2024); JSON Schema (drafts) — Ch 10’s tooling table.
• GraphQL — Ch 14; Ch 16’s table.
• Linked Data Platform 1.0 (W3C REC, 2015) — Ch 18’s wrong-layer instance: containers as canned selections; subtract them and the Graph Store Protocol remains.
• Should we remove XSLT from the web platform?, WHATWG HTML issue, August 2025 — Ch 9’s third mismatch, with the removal underway. The stated grounds are unmaintained implementations, which is Ch 9’s maintenance-failure finding in the platform’s own words.