No, I’m sorry, still problems with sorting / autonumbering. And when you put things in the right order through drag and drop it’s modifying Sequence numbers again.
But because I like the plugin so much for all the other possibilities and statistics like the Codex with the characters, locations and items etc, I will only be using the timeline view. That seems to be working well with the ordering. Also with drag and drop to modify the order.
It’s really only the Board that’s not working for me.
The rest of the plugin is something I would develop my own if I could program for Obsidian!!!
So, thank you very much for this!
Via BRAT I get a message about a description issue. It seems that version 0.9.1 has a typo that triggers this message. It says it encounters version 0.91?
Yes, there is a typo with commit version 1.9.1 that is tagged 1.91 so BRAT identifies an inconsistency and refers to it as the latest version. You must manually select 1.9.3 to install it.
@Hans1
You’re right — there were bugs in the Board view drag-and-drop. Thank you for the detailed report. I’ve now identified and hopefully fixed the root causes:
What was wrong:
Drag-and-drop used the wrong sort order internally. The Board view displays cards sorted by act → chapter → sequence, but when you dragged a card, it was computing the new position using only the sequence field. This mismatch meant cards would end up somewhere unexpected after the drop.
Drag-and-drop never updated the chapter field. When you dropped a card between scenes in, say, Chapter 5 and Chapter 6, the card kept its old chapter number. Since the Board sorts by chapter first, it would “snap back” to its original chapter group instead of staying where you dropped it.
Chapter corruption bug. Under certain grouping modes, the drag-and-drop was accidentally overwriting your chapter numbers with sequence numbers — which would scramble your entire ordering.
This explains why your data got progressively worse the more you tried to fix it with drag-and-drop, and why the Timeline view (which uses a completely different drag-and-drop approach) worked fine.
What’s fixed in the 1.9.4 update:
The Board view drag-and-drop now works like the Timeline — after a drop, it rebuilds the visual order and renumbers sequences cleanly within each (act, chapter) group. It also properly updates the chapter field when you move a card between chapter groups within an act.
Your existing scene data may have been corrupted by previous drag-and-drop operations. After updating, try clicking “Resequence all scenes” in the Board view to clean up the numbering — that logic was already correct.
Sorry for the trouble, and appreciate your patience!
Only small glitch: when drag and drop it renumbers the Sequence numbers instead of renumbering chapters, but that’s a small thing for me. I don’t mind if a piece of text is a chapter or a scene. The most important thing for me is the general order and that works great now! Also when renumbering.
Now I can drag and drop without being affraid of corrupting the whole order.
That’s by design — drag-and-drop renumbers the sequence (position within a chapter) rather than chapter numbers, because most users have multiple scenes per chapter and wouldn’t want their chapter assignments to change on every drag.
Quick question — are you grouping by Act or by Chapter in the Board view? If you’re grouping by Act, drag-and-drop only renumbers the sequence within each chapter (since one act column can contain multiple chapters). If you switch to Group by Chapter, dragging a scene between columns will update its chapter number too.
Act / chapter sort order with 10+ entries — Filter dropdowns, board columns, manuscript order, exports, navigator, plotgrid and stats no longer sort acts/chapters lexically (1, 10, 11, 2, 3, …). They now sort numerically (1, 2, 3, …, 10, 11). Also fixes scene order in exports when a project has more than 9 acts or 9 chapters.
New Features
Custom scene fields Define your own metadata fields on scenes (Story Grid functions, Truby aspects, beat-sheet labels, genre conventions, anything you need) without overloading the title or subtitle. Manage templates from Settings → Custom Scene Fields or inline from the Inspector. Supports text, textarea, dropdown and multi-select types.
Inspector — every scene now has a “Custom Fields” section to edit values.
Board → Group by — dropdown and multi-select fields appear as grouping options in the Kanban board (multi-select scenes show in every matching column).
Filters — dropdown and multi-select fields each get a chip group in the filter panel.
Scene cards — up to three populated values appear as small badges on cards, and hovering any card shows a full summary of every custom value.
Values are stored under universalFields: in scene frontmatter; templates live in <project>/System/field-templates.json and sync with the rest of the project.
Free-form act and chapter labels — Acts and chapters are no longer restricted to plain integers. You can now use:
Text labels: Prologue, Epilogue, Interlude A (sort numerically first, then text).
Plain integers: still work exactly as before.
Inspector — Act is now a free-text field — The Act control in the Scene Inspector is a text input instead of a 1..N dropdown, so you can type any of the values above. Illegal Windows path characters (< > : " / \ | ? *) are flagged with a warning.
Folder and filename safety — Scene folders (Act 1.1/) and filename prefixes (1.1-02 Title.md) are sanitized for non-numeric act/chapter values. Pure integers continue to be zero-padded as before for sortable filenames.
Safer next-act numbering — The auto-suggested next act number now ignores non-numeric values, so projects containing a Prologue no longer break the increment.
Series-shared characters and locations — per-book scoping — Series Mode now lets you mark which books a character or location appears in instead of always sharing every entity with every book. Right-click any character card or location/world row in the Codex views to:
Promote to series — move a book-local entity into the shared series Codex so sibling books can use it.
Demote to project — move a series-shared entity back into the current book’s local Codex.
Restrict to “” only / Add to / Remove from “” / Share across all books — toggle the new optional books: frontmatter list. Empty / missing means “appears in every book” (the historical behavior, fully backwards compatible).
A new All books / Showing: filter chip in the Characters and Locations views hides entries that don’t appear in the current book.
Series projects can now keep both a per-project Codex/Characters and Codex/Locations folder for book-only entities, alongside the shared series-level Codex.
Codex / character / location notes vanishing after a template insert(#74) — When a Templates / Templater plugin replaced the YAML block and dropped the type: discriminator, the affected note silently disappeared from the Codex, Characters or Locations view. StoryLine now falls back to folder-based detection: any note that lives under the canonical entity folder (Codex/<Category>/, Characters/, World/Locations/) is recognised even if type: is missing, wrong, or the frontmatter is empty.
Plot Grid sorting by Sequence / Chapter / Act now respects non-numeric values(#76) — Sorting rows by Act, Chapter, or Sequence used to coerce values via Number(...), which silently turned strings like Prologue, Epilogue, or 1.1 into NaN and produced an apparently random order. Plot Grid (and the shared SceneQueryService used by every other view) now uses the numeric-aware compareActChapter comparator and falls back through Act → Chapter → Sequence in reading order. Note: the second half of the issue — automatic sequence reflow when inserting a chapter mid-book — is intentionally deferred to a later release; it requires a project-wide renumber that we want to design carefully.
New Features
Multiple roles per character(#72, Tier 1) — A character can now hold more than one role at the same time (e.g. protagonist + narrator). The role field in the Character inspector accepts comma-separated values and renders one badge per role. Existing single-role characters keep working unchanged.
Role history across scenes / plotlines / books(#72, Tier 2) — Characters can now record a history of role changes in a structured roles: YAML array. Each entry has a role label and optional from (scene name or [[wikilink]]), plotline, and book (free-text label) anchors. A new repeating-row editor under the role field in the Character inspector lets you add/remove entries. Both the legacy role: field and the new roles: history coexist; if both are set, the structured history wins for display order. The book: anchor is plain author-facing metadata — no cross-book plumbing required.
Wikilink-aware scene references(#73) — Scene fields that reference other notes (pov, location, characters, setup_scenes, payoff_scenes) are now stored as Obsidian [[wikilinks]] so renames are picked up automatically by Obsidian’s metadata cache. Plain-text values continue to work — the readers accept either form. Toggle in Settings → Write scene references as wikilinks (on by default).
Custom fields visible to Properties / Bases / Dataview(#71) — Universal Field Templates gained an optional Top-level YAML key. When set, the field’s value is mirrored to a real top-level YAML key (in addition to universalFields:) so it appears in Obsidian’s Properties panel, Bases, Dataview and the graph. Reserved StoryLine keys (type, pov, act, chapter, tags, …) are blocked. Toggle in Settings → Mirror custom fields to top-level YAML (on by default).
Step-sibling and teammate relations(#70) — Added step-sibling to the family relation list and teammate to the professional relation list, with matching symmetric inverses.
Default scene frontmatter snippet(#77) — Universal Field Templates gained an optional Default value that is applied automatically when a new scene is created (multi-select fields accept comma-separated defaults). In addition, Settings → Default scene frontmatter accepts a free-form YAML block whose keys are merged into every newly created scene’s frontmatter. StoryLine-owned keys (type, title, act, chapter, sequence, status, wordcount, …) always win on conflict, so the default snippet can never overwrite the engine’s own metadata.
Wordcount can now ignore comments and checkbox lines(#78) — Two new toggles under Settings → Scene Cards:
Exclude %%comments%% from wordcount(default on) — Obsidian comment blocks no longer inflate the count, so author notes and TODOs stay invisible to the manuscript total.
Also ignore checkbox lines(default off) — When enabled, - [ ] and - [x] lines are also stripped before counting, which is useful for outline-style scenes that mix prose with task lists. Both toggles flow through the central MetadataParser.countWords helper, so they apply consistently to scene cards, the inspector, the Writing Tracker, and exports.
Internal
Series Arc View — foundation(#66 Phase 1, scaffold only) — A new SceneProvider indirection layer (services/SceneProvider.ts) was added so future view code can pull scenes from either the active book or every book in a series. Settings flags seriesArcView (default off) and warnOnCrossBookMove (default on) are reserved. The cross-book Manuscript / Kanban / Plot Grid rendering itself ships in a follow-up release.
Really looking forward to using it!