For the case behind “turn manuscript into EPUB,” define the evidence first: assign one authoritative repair source to every defect class. The controlled case is four defects in body text, chapter order, author metadata, and display settings. A pass must demonstrate this outcome: complete the article-specific sample with separate evidence for body return path, order return path, metadata return path, display-setting return path, regeneration. The excluded range is stated separately: direct repair of an existing EPUB and editing CSS in an external tool.
Decision: Assign one authoritative repair source to every defect class
Make the authoritative input and its state visible before work spreads across copies. The register must still identify body return path and regeneration when the next session begins.
The article-specific evidence card has these fields: body return path, order return path, metadata return path, display-setting return path, regeneration. Do not count the following excluded range as evidence: direct repair of an existing EPUB and editing CSS in an external tool.
What the controlled material must reveal
Classify four defects—body text, chapter order, author metadata, and display settings—and declare one authoritative return path for each. Correct only those sources, regenerate from the same settings, and inspect both the four intended repairs and unrelated differences. Never repair generated XHTML or treat an author name typed into body text as metadata.
This case addresses the failure behind the search: losing the authoritative input or acceptance evidence while trying to assign one authoritative repair source to every defect class. Name the source, working copy, and delivered or reopened result so that another editor can locate each one without relying on a filename such as final.
Run order and evidence
Use the following product-independent sequence on the named sample. It is an acceptance method, not a claim that Rune Studio has already completed this particular case.
- 1. Classify the four defects. Attach the observed body return path evidence to this step.
- 2. Declare one source for each. Attach the observed order return path evidence to this step.
- 3. Edit only those sources. Attach the observed metadata return path evidence to this step.
- 4. Regenerate from unchanged settings. Attach the observed display-setting return path evidence to this step.
- 5. Inspect intended repairs and unrelated differences. Attach the observed regeneration evidence to this step.
Pass, revision, or not tested
Do not infer one row from another. Record pass, revision, or not tested beside the actual target, date, output path or artifact ID.
- Body return path — expected state: written before the run; hold the row if generated XHTML is edited.
- Order return path — expected state: written before the run; hold the row if an author name in body text is treated as metadata.
- Metadata return path — expected state: written before the run; hold the row if unrelated settings drift during regeneration.
- Display-setting return path — expected state: written before the run; hold the row if generated XHTML is edited.
- Regeneration — expected state: written before the run; hold the row if an author name in body text is treated as metadata.
Three failures that invalidate the run
A second pass is meaningful only when its source and changed condition remain identifiable. Keep the earlier result and append the retest instead of replacing the failed row.
- 1. Generated XHTML is edited; preserve the failed artifact.
- 2. An author name in body text is treated as metadata; change only the responsible condition.
- 3. Unrelated settings drift during regeneration; return to the named source.
Rune Studio as a bounded candidate
Current product documentation covers capabilities relevant when you need to preserve manuscript order and metadata, generate EPUB 3, and inspect the resulting structure. That scope can justify a trial, but it does not show that four defects in body text, chapter order, author metadata, and display settings passed this article's acceptance card.
Rune Studio is a plausible candidate when fewer source, setting, or reference handoffs help you assign one authoritative repair source to every defect class. Use another tool or destination check for the excluded range stated here: direct repair of an existing EPUB and editing CSS in an external tool.
First task
For the search phrase “turn manuscript into EPUB,” the conclusion is to assign one authoritative repair source to every defect class. The result is accepted only when the record establishes this outcome: complete the article-specific sample with separate evidence for body return path, order return path, metadata return path, display-setting return path, regeneration.
Begin here: make a four-row defect-to-source map. If generated XHTML is edited, stop at that row and return to its source. Check the Rune Studio product page for the current Mac feature scope before applying the same acceptance card to a product trial.
