Origami EPUB lets an author publish not only formatted text but the structure of the thinking itself: addressable passages, explicit concepts, citations, equations, tables, spatial objects, relationships, and authored views. Crucially, Origami can be treated as a publication target rather than an authoring-application format, so the author does not have to adopt a particular editor in order to create a rich Origami publication.
• Paragraph-level addressability for paragraphs, headings, figures, equations, glossary entries, bibliography entries, and notes.
• Precise citation of passages rather than only whole documents or pages.
• Stable element identity through typo, punctuation, formatting, and non-substantive wording changes.
• New identity for substantively changed passages.
• Explicit lineage when paragraphs split, merge, or are replaced.
• Clear distinction between identity and lineage.
• Work identity distinct from a particular publication or edition.
• Multiple editions recognized as manifestations of the same intellectual work.
• Explicit relationships between editions.
• Authored concept metadata rather than concepts inferred only from word occurrence.
• A machine-readable glossary with stable concept identities.
• Concept definitions that can be exposed at the point of reading.
• Concept-to-citation relationships.
• Machine-readable document structure and semantic heading indexes.
• Self-contained canonical bibliography data.
• Copyable human-visible self-citation.
• Self-description: the document can tell humans, LLMs, and software that richer metadata exists.
• Human-visible instructions explaining where machine-readable metadata lives.
• Cross-document semantic links.
• Links that identify both the target publication and the target element.
• Quoted-text provenance alongside link targets.
• Explicit cites and transcludes relationships.
• Authored expandable text through stretchtext.
• Different levels of exposition in one work without overwhelming first-pass reading.
• Live computational tables.
• Published table values plus formulas.
• Reproducible computational explanation.
• Reader experimentation without destroying the author’s published values.
• MathML as authoritative mathematics rather than equation images.
• Optional LaTeX as a derived representation.
• Stable, directly citable equation identities.
• Stable, directly citable figure identities.
• True 3D scholarly figures packaged with the publication.
• Physical-scale metadata for spatial models.
• Explicit model up-axis/orientation.
• Explicit model extent.
• Editorial poster views for 3D figures.
• Reader inspection of the actual 3D object.
• Source provenance for 3D assets.
• Reduced publication assets pointing back to full-resolution originals.
• Authored spatial layouts as part of the publication.
• Published concept maps and spatial organizations as authored intent.
• Multiple authored views over the same content.
• Relationships declared independently of any one visual layout.
• Spatial publication without sacrificing conventional linear reading.
• Graceful fallback by design.
• 3D figures still have ordinary readable carriers.
• Tables retain static published values.
• Equations remain MathML.
• Glossary items remain ordinary XHTML.
• No Origami semantics need depend on JavaScript.
• Authorial semantics survive environments where scripting is disabled.
• Metadata survives application boundaries.
• Important semantic facts live in the publication rather than only in the originating app.
• Controlled metadata redundancy for resilience and discoverability.
• Human-readable and machine-readable copies can coexist with declared authority.
• Authored intent is separated from reader activity.
• Long-term inspectability if the original software disappears.
• The EPUB remains readable as EPUB while its augmentation stays documented.
• The publication can act as a portable knowledge object rather than only a presentation file.
• Origami widens the range of software that can be used to author academic work because conversion can start from structured source rather than from one proprietary document format.
• Semantic HTML/XHTML can be converted especially directly because headings, paragraphs, lists, figures, links, tables, and other structures are already close to EPUB’s native content model.
• Markdown and Pandoc-based workflows can serve as strong authoring inputs, retaining structure, citations, notes, tables, and mathematics where the source expresses them.
• JATS or other structured XML can serve as a high-quality scholarly input because article structure, references, notes, and semantic roles are already explicit.
• DOCX and ODT can be normalized into Origami when authors use real heading styles, tables, footnotes, and citation structures rather than purely visual formatting.
• LaTeX and Overleaf-style projects can be converted while retaining strong mathematics and bibliography information, with custom macros normalized during conversion.
• Jupyter notebooks can become Origami publications while preserving narrative structure, figures, tables, equations, and links to computational provenance.
• Long-form and source-oriented tools such as Scrivener, code editors, and HTML editors can participate when they export structured HTML, Markdown, XML, DOCX, or another converter-friendly format.
• Potential authoring environments therefore include semantic HTML/XML editors, Markdown editors, Quarto/Pandoc workflows, JATS-oriented tools, Word, Google Docs, Pages, LibreOffice, LaTeX editors, notebooks, and plain source-code editors.
• An Origami-native authoring tool can still provide the richest direct support for concepts, maps, spatial layouts, live interactions, and metadata, but it is not a prerequisite for publishing Origami EPUB.
• Existing source structure can be preserved rather than re-marked-up from scratch when the converter can recognize headings, citations, glossary terms, equations, tables, notes, figures, and bibliography data.
• The converter can generate stable Origami IDs, normalize source markup to EPUB-compatible XHTML, build the OPF and navigation, create semantic and interaction records, package resources, generate the Visual-Meta colophon, and validate the result.
• This gives publishers a common rich publication target while allowing authors to keep working in the software in which they think and write best.
• Accessibility metadata and alternative-text claims can be generated or validated against the actual publication content rather than asserted independently of it.
• External annotation architecture lets readers and reviewers annotate heavily without modifying the author’s published work.