Skip to main content

OMI-SPEC-230 — Publishing Model

Status​

Draft

Version: 0.1.0

Legacy identifier: OMI-SPEC-007

Purpose

The Publishing Model defines how a semantic manuscript is transformed into one or more publication formats.

Unlike traditional publishing workflows, OMI separates scholarly content from presentation.

A manuscript is authored once and rendered many times.

Design Principles

The Publishing Model follows these principles:

  • Single source of truth
  • Semantic rendering
  • Format independence
  • Reproducibility
  • Interoperability
  • Extensibility
  • Automation

Publishing Philosophy

Publishing is not document conversion.

Publishing is the process of rendering semantic knowledge into publication-specific representations.

The manuscript never changes.

Only its presentation changes.

Publishing Pipeline

Semantic Manuscript

        │

        ▼

Publication Profile

        │

        ▼

Rendering Engine

        │

        ▼

Publication Format

Every publication is generated from the same semantic source.

Publication Profiles

A Publication Profile defines how a manuscript should appear for a specific target.

Profiles may define:

  • typography
  • citation style
  • note style
  • layout
  • metadata mapping
  • accessibility requirements
  • branding

The manuscript itself remains unchanged.

Supported Outputs

OMI is designed to support multiple publication formats.

Examples include:

  • HTML
  • PDF
  • EPUB
  • DOCX
  • JATS XML
  • Crossref XML
  • DataCite XML
  • Markdown
  • Plain Text

Future formats may be added without changing the manuscript.

Multi-Target Publishing

One manuscript may generate several outputs simultaneously.

Example:

Semantic Manuscript

        │

 ┌──────┼──────────────┐

 ▼      ▼              ▼

HTML    PDF       JATS XML

 ▼      ▼              ▼

Website Print     Repository

Each output is derived independently.

Rendering Engine

The Rendering Engine interprets semantic objects and applies publication rules.

Examples:

Heading

↓

HTML

<h2>

↓

PDF

Large Heading

↓

JATS

<title>

The semantic object remains identical.

Notes

Annotations may be rendered differently.

Examples:

PDF

↓

Footnotes

HTML

↓

Tooltips

EPUB

↓

Endnotes

JATS

↓

Semantic elements

Citations

Citation rendering depends entirely on the publication profile.

Examples:

APA

↓

(Author, Year)

Chicago

↓

Footnote

IEEE

↓

[15]

The Citation Model itself never changes.

Figures

Publication Profiles determine:

  • placement
  • captions
  • numbering
  • cross references
  • accessibility

Tables

Profiles may define:

  • responsive HTML tables
  • printable tables
  • accessibility metadata
  • long table handling

Equations

Supported rendering may include:

  • MathML
  • LaTeX
  • SVG
  • PDF vector output

Metadata

Metadata is exported independently for each target.

Examples:

Crossref XML

↓

Crossref Schema

DataCite XML

↓

DataCite Schema

JATS XML

↓

JATS Metadata

No metadata is duplicated manually.

Accessibility

Publishing Profiles should preserve accessibility.

Examples:

  • semantic headings
  • image descriptions
  • table headers
  • keyboard navigation
  • tagged PDF
  • EPUB accessibility metadata

Accessibility requirements should be configurable.

Preservation

Publication formats intended for long-term preservation should maintain semantic integrity.

Preferred preservation formats include:

  • JATS XML
  • PDF/A
  • EPUB
  • Markdown

Future mappings may support institutional repository standards.

Automation

Publishing workflows may be fully automated.

Example:

Review Completed

↓

Accept

↓

Generate HTML

↓

Generate PDF

↓

Generate JATS XML

↓

Deposit Metadata

↓

Publish

No manual document recreation is required.

Plugin Architecture

New publication formats are added through plugins.

Examples:

OMI Core

        │

        ├── HTML Renderer

        ├── PDF Renderer

        ├── EPUB Renderer

        ├── JATS Renderer

        ├── DOCX Renderer

        └── Custom Renderer

The Core Publishing Model remains unchanged.

Interoperability

Future integrations include:

  • Open Journal Systems (OJS)
  • Open Monograph Press (OMP)
  • Open Preprint Systems (OPS)
  • Crossref
  • DataCite
  • Zenodo
  • Institutional Repositories

Versioning

Every publication is reproducible.

A published output should always be traceable to:

  • manuscript version
  • publication profile
  • rendering engine version
  • exact generated artifact
  • validation/release conditions that governed delivery

This ensures reproducibility and transparency.

Reference implementation note​

Open Manuscript Studio currently demonstrates this model with validated JATS, semantic HTML and paged PDF publication builds. JATS uses semantic-fidelity gates plus pinned offline JATS 1.4 Article Authoring DTD validation; PDF uses a pinned Vivliostyle renderer; and JATS/HTML/PDF artifacts can be paired with portable .omi-build.json provenance sidecars containing revision, profile, renderer and artifact digests.

See Publication Output Validation and Provenance for the implementation-level pipeline. This evidence is informative and does not by itself change the normative status of this Draft specification.

Future Work

Future specifications will define:

  • Rendering API
  • Template Language
  • Publication Profiles
  • Workflow Automation
  • Continuous Publishing
  • Dynamic Publications

Change history

  • 0.1.0 — Migrated from provisional OMI-SPEC-007 to canonical OMI-SPEC-230.

Summary

The OMI Publishing Model transforms a single semantic manuscript into multiple publication formats without altering the original scholarly content.

By separating semantic meaning from presentation, OMI enables automated, reproducible, and interoperable publishing workflows across journals, repositories, and future scholarly communication platforms.