Chapter 11. Phases 4–6 — When the Materials of a Book Become a Book

Imagine a print shop: the weight of paper in one hand, the smell of fresh ink, and a monitor showing each hour as four CMYK inks accumulate on the sheet. A run of 1,000 LOREBOOKs moves onto paper in about twelve to sixteen hours, then enters the bindery to become books. That final print run ends the first LOREBOOK cycle. Before it lies four months in which book materials become a book product.

At the end of Phase 3, the LOREBOOK is still material. Its 1,500 Markdown files contain prose and wikilinks, but consistency across Parts has not been integrated, the end matter is empty, and neither a print PDF nor a website exists. Three stages remain.

Phase 4 spends one month verifying consistency; Phase 5 spends one month producing end matter; Phase 6 spends two months creating print and public web editions. This chapter covers those final four months.

Phase 4 — Consistency among Materials

Phase 4 verifies relationships among all 1,500 completed entries. Phase 3 checked one Part at a time; Phase 4 first integrates checks between Parts. With 5,000–10,000 wikilinks, changing one entry can affect many others. The work asks whether an update in one Part matches citations elsewhere, whether a character's history agrees with stories, and whether Librarying and LOREBOOK frontmatter align.

The process has five steps.

First, dead-link checking confirms that every wikilink points to a real entry. If [[Paksu Mudang]] appears without a corresponding file, it is dead. Extract all wikilinks, compare them with filenames, then add missing entries, update renamed links in bulk, or remove mistaken citations.

Second, duplicate-information detection finds definitions scattered across entries. Keep one canonical location and replace repetitions with wikilinks. If the fictional IP “○○” defines the Sinnaenim ritual both in Part 2 §3 and in a Paksu Mudang entry in Part 9 §1, keep Part 2 §3 as primary and link to [[Sinnaenim Ritual]] from the latter.

Third, and most exacting, is the SSOT hierarchy check. Compare the LOREBOOK against higher SSOTs—rulebooks, character sheets, and stories—and update it wherever they conflict. Automation catches clear differences in names, years, and places, but people must judge subtle conflicts in personality, relationships, and motives.

Fourth, verify consistency of nouns, viewpoints, and faction systems. Confirm that Chapter 6's two-coordinate mapping and Librarying's official entry names are used throughout.

grep -r "old form" .  → PASS only at 0 results
grep -r "Baekryong-guk" . → discarded form, PASS only at 0
grep -r "shrine" . → inspect consistency of the entry name

Fifth, check Librarying frontmatter alignment. The eight fields inherited as described in Chapter 5 must match exactly. An automated script compares the thirteen-column Librarying CSV with LOREBOOK frontmatter and reports every mismatch.

Phase 4 passes when seven criteria are ✓: zero dead links; zero duplicates, except explicitly intentional ones; zero SSOT violations; zero noun/viewpoint/faction grep violations; 100 percent frontmatter alignment; synchronization with final character sheets; and consistency verification for all one hundred stories. The deliverable is _Checklist.md containing all results.

Automation takes about one to two hours; human judgment and review take two to three weeks, making about one month. Automation carries roughly 90 percent of the mechanical load while people resolve nuanced conflicts.

Phase 5 — One Month of End Matter

Phase 5 builds the appendices beyond the 1,500 body entries.

Eight appendices are recommended: four required and four optional. The required four are a glossary (5–15 pages, 200–500 entries alphabetically), index (3–10 pages, page references by person, place, and subject), map atlas (4–12 pages of world, regional, city, and dimensional maps), and master timeline (3–8 pages placing scattered events in chronological order). They enable search and reference after a reader finishes the volume.

The optional four depend on the IP: NPC minicards with one-line descriptions for 25–50 major figures; a one-line faction catalog; family trees and relationship diagrams; and a companion story-collection guide. Character-heavy artbook or comics-guide IPs emphasize minicards, while IPs with story assets almost require the companion guide.

The glossary and index can be automated. Extract definition fields from all frontmatter and sort them to make one line per entry—keyword, definition, and referenced Part. Extract citations from every body entry and classify them by person, place, and subject to build index references. What takes two to three weeks manually can finish in about one week with generation plus human review.

The six-step flow is: Part 12 §1 atlas, ordering maps and captions; §2 master timeline, extracted from Part 3 and human-sorted; §3 NPC minicards, extracted from Part 9 and reviewed; §4 glossary, generated from frontmatter and reviewed; §5 index, generated from the body with page numbers filled after typesetting; and optional appendices suited to the IP. It takes about one month.

The deliverable is Part 12 at ✓, with all four required and appropriate optional appendices complete. The 1,500 entries, up to eight appendices, and 100-percent-approved illustrations make the main volume nearly complete. Phase 6 carries it into two media.

Phase 6 — Conversion into Two Media

Phase 6 is the final stage. The finished body and appendices split into print and public web editions through three flows: web conversion, print typesetting, and print ordering.

Web Conversion

Chapter 1's single Markdown source principle now operates: the same 1,500 files create both media. MEJE IPs prefer Quartz, which converts an Obsidian Vault into a static site while preserving wikilinks and graph view.

Alternatives include Obsidian Publish at USD 8–16 per month, or Hugo, Eleventy, Astro, VitePress, and Docusaurus, which offer greater design freedom with higher learning costs. Deployment can use Cloudflare Pages, Vercel, Netlify, or GitHub Pages.

Begin _WebConversionGuide.md by deciding the generator, URL mapping from directories to paths, SEO metadata (title, description, keywords), palette, and fonts. Build a pilot site containing all 1,500 entries; verify body text, illustrations, cross-references, and graph view; configure a domain such as lorebook.example.com, hosting, and HTTPS; then perform the production build and deployment. This takes about two to four weeks.

Print Typesetting

The LOREBOOK workflow owns the Markdown; external tools such as InDesign or Affinity Publisher turn it into a printable PDF.

Choose a typesetting tool—InDesign for professional publishing, Affinity Publisher as an alternative, or Vellum mainly for novels—and apply a standard B5 hardcover template, five-color palette, and page templates T01–T17. Convert frontmatter to page metadata, prose to body copy, wikilinks to page-number references, and place illustrations while applying the palette.

Phase 5's index initially has empty page numbers. Once pagination is fixed, the numbers populate automatically, eliminating manual entry.

Finally, print one pilot copy and inspect color, fonts, margins, and illustration placement. Revise and reprint if defects appear. The flow takes about one to two months. An external publishing team may typeset, but the reader participates in review and decisions.

Print Ordering

The recommended checklist covers quotes from three printers at small, standard, and large quantities; paper samples for body, endpapers, and cover; cover finishing such as black or gold foil, UV, and embossing; measured spine width from the PDF; CMYK conversion and output profile checks; publisher ISBN; back-cover barcode; delivery and transport; inspection; and reprint terms.

CMYK—Cyan, Magenta, Yellow, and Key black—is the print standard. Its gamut differs from screen RGB, so conversion and proofing are mandatory. RGB illustrations often become muted, and the five-color palette's accents must be reviewed after conversion.

Phase 6 produces two deliverables: a public static site readers access by domain, and a roughly 450-page B5 hardcover on the shelf. Together they end the first LOREBOOK cycle.

Gates for Phases 4, 5, and 6

The Phase 3-to-4 gate opens when every entry in every Part is at least ◕ and illustration commissioning has begun, with automated and master review. The 4-to-5 gate requires zero violations in _Checklist.md and passed consistency checks. The 5-to-6 gate requires all four required and four chosen optional appendices at ✓ and illustrations 100 percent approved. The publication gate after Phase 6 requires the master to approve the pilot print, ISBN, barcode, and legal specifications.

Every entry condition is explicit. Failure returns the project to its current Phase; temporary bypasses are not recommended.

Phases 4–6 for the Fictional IP “○○”

Phase 4 took about one month. It found twelve dead links: five resolved by adding entries and seven by updating links, reaching zero. Duplicate detection found the Sinnaenim definition twice; Part 2 §3 became primary and Part 9 §1 became a link. One SSOT conflict with Story 023, “Dawn at the Shrine,” led to a LOREBOOK update. Noun checks found zero uses of discarded “Baekryong-guk” and consistent “Dongbaek-guk”; frontmatter reached 100-percent agreement. Final character-sheet synchronization incorporated Kim Miseon's updated Paksu Mudang sheet, and review of one hundred stories confirmed all thirty cited stories. The deliverable was a fully ✓ _Checklist.md.

Phase 5 took about one month. Part 12 §1 organized five maps and captions—Meryton, Old Mudang Street, Haewon Restaurant Quarter, Spirit Realm, and Road to the Underworld. §2 sorted eighty events; §3 made thirty-five NPC minicards; §4 generated and reviewed one-line definitions for 1,200 entries; §5 generated indexes by person, place, and subject. Optional matter included the minicards, five family trees, and the companion guide. Every appendix reached ✓.

Phase 6 took about two months. Quartz web conversion deployed to lorebook.example.com/○○ after a pilot build, taking about three weeks. An external team used InDesign with B5 hardcover specifications and T01–T17 templates to convert Markdown, populate index pages, and revise typesetting after a pilot print; production printing took about five weeks. After comparing three quotes, selecting paper, cover, and binding, and issuing ISBN and barcode, the project printed 1,000 copies. The deliverables were the deployed site and 1,000 printed books.

The End of the First Cycle

When Phase 6 lands, the first LOREBOOK cycle ends and the reader can place the volume on a shelf.

The LOREBOOK is not made once and abandoned. As the IP operates, new stories appear, characters change, the Librarying Vault receives incremental updates, and multilingual editions proceed. The LOREBOOK changes with them. Chapter 13 discusses operation as a living resource.

For now, the nine- to fourteen-month first cycle is complete: one roughly 450-page B5 hardcover reaches the shelf, and the same material is available as a static site on its domain. It is the first product that binds an IP's materials into one book.

The copyright in this work belongs to Kim Dongeun WhtDrgon. and MEJE Works Corp. No part of this work may be reproduced, distributed, transmitted, displayed, performed, broadcast, translated, adapted, or otherwise used by any means without prior written permission from the copyright holders.