Skip to content

🗺️ Roadmap & Feature Backlog ​

This document provides transparent tracking of features currently in development, architectural plans, and features parked in the backlog for future releases (target: ZenTauri v2.0.0).


📌 Implementation Status Overview ​

CapabilityCurrent StatusTarget MilestoneDescription
QA Reference Dual-Pane📋 Backlog / Parked (v1.3.0)v2.2.0Text-based dual-pane comparison (parked in v1.3.0 in favor of Mode 3: Dual PDF QA)
Harvard-Kyoto & IAST Hotkeys✅ Active (v1.2.0)v1.2.0Bidirectional shortcuts ⌥⌘D & ⌥⌘H with tag-safe directive isolation
Academic Footnotes & Deflists✅ Active (v1.2.0)v1.2.0Standard [^1] footnotes and Term\n: Definition lists via markdown-it
Native Typst PDF Compilation⏳ In Developmentv2.0.0Pure Rust Typst compilation in src-tauri without browser/WebKit print engine
Dual PDF QA & Proofing (Mode 3)📋 Backlog / Conceivedv2.0.0Dual-vector PDF viewer with bounding-box annotations & AI correction prompt
Synoptic Master-JSON (Mode 2)📋 Backlog / Conceivedv2.1.0Split-pane editor for scanned editions (book.json AST format from Bṛhadāraṇyaka)
DOCX Bridge & CriticMarkup📋 Backlog / Conceivedv2.1.0Pandoc-powered DOCX ingest to clean Markdown, scholarly export template, and Track Changes
In-Editor Live IME (Stage 2)📋 Backlog / Conceivedv2.0.0CodeMirror live conversion inside 《...》 and compose sequences in prose
Knowledge Graph & Wikilinks📋 Backlog / Parkedv2.0.0Re-integration of interactive graph, [[ fuzzy autocomplete, and Cmd+Click

1. Native Typst PDF Compilation ​

  • Status: In Development (src-tauri via typst & typst-pdf crates)
  • Goal: Completely replacing web-based HTML print rendering with instantaneous, publication-grade academic PDF typesetting powered directly by the native Typst compiler.

Key Milestones: ​

  1. Direct AST Mapping: Compiling Markdown AST elements into semantic Typst markup.
  2. Academic Layouts: Automated page numbering, margins, footnotes, and Devanagari typography cascades.
  3. No Heavy Chromium/Node Dependencies: Lightweight, cross-platform binary builds.

2. In-Editor Live IME (Stage 2 of Sanskrit Hybrid Framework) ​

  • Status: Planned / Architecture documented in Sanskrit Input Architecture
  • Goal: Context-aware real-time transliteration directly inside the CodeMirror 6 editor without switching operating system keyboard layouts.

Planned Stages: ​

  • Stage 1 (Canonical Storage): ✅ Active — Markdown source always persists standard IAST or Unicode Devanāgarī.
  • Stage 2 (In-Editor Live IME): 📋 Planned
    • Scope-Based Live IME: While typing inside Sanskrit brackets (《...》 or ⟪...⟫), ASCII Harvard-Kyoto (kRSNaH) is converted to IAST (kṛṣṇaḥ) or Devanāgarī (कृष्णः) in real time.
    • Compose Sequences: Fast inline character sequences in running prose (.r => ṛ, -a => ā, ~n => ñ, etc.).
  • Stage 3 (Post-Hoc Transliteration): ✅ Active — Global shortcuts ⌥⌘D and ⌥⌘H for selection toggling.

NOTE

OS Keyboard Compatibility: Direct input via OS keyboards (such as EasyUnicode or Mac Option-keys) is always passed through unmodified and will never be constrained by Stage 2.


  • Status: Backlog (temporarily deactivated in active UI to reduce bundle size by ~600 kB)
  • Components: src/components/GraphView.vue, src-tauri/src/db.rs (SQLite Workspace Indexer)

Planned Enhancements upon Re-Activation: ​

  1. Editor Autocomplete for Wikilinks ([[): Fuzzy document search popup when typing [[ in CodeMirror.
  2. Direct Source Code Navigation (Cmd+Click): Instant jumping to linked target files.
  3. Backlinks Inspector: Dedicated sidebar panel for incoming cross-references.
  4. Graph Filters & Orphan Detection: Isolate unlinked notes and filter by tags or folders.

4. Dual PDF QA & Proofing (Work-Mode 3) ​

  • Status: Backlog / Conceived
  • Goal: Comprehensive side-by-side collation and AI-assisted proofing of re-typeset editions against historical reference PDFs.

Planned Stages: ​

  1. Normalized Overlay & Bounding Boxes: Click-and-drag box annotations on rendered SVG vector pages (hayro_svg) with relative percentage coordinates.
  2. Web-First AI Prompt Generator: 1-click clipboard prompt containing source Markdown and structured correction table for Claude/ChatGPT/Gemini Web.
  3. Annotated PDF Export & Automation: Direct export of flattened PDF with visual annotations and optional local Ollama diff-patching.

5. Synoptic Master-JSON Editor (Work-Mode 2) ​

  • Status: Backlog / Conceived (Inspired by the Bṛhadāraṇyaka-Upaniṣad edition project)
  • Goal: Segmented, split-pane editing of scanned historical editions (book.json / AST format) without requiring error-prone manual TEI-XML markup.

Planned Features: ​

  1. Split-Pane Scan Alignment: Synchronized historical PDF scan on the left, structured segment entry on the right.
  2. Field-Based Segments: Dedicated inputs for Devanāgarī, IAST, translations, critical variants (Slaje vs. Böhtlingk), and philological commentary.
  3. Automated Export Pipeline: Native Rust compilation from Master-JSON into standard TEI-P5 XML and Typst-PDF.

6. DOCX Interoperability & CriticMarkup (Word Bridge) ​

  • Status: Backlog / Conceived
  • Goal: Lossless bridge to Microsoft Word-centric academic publishers and reviewers under a strict Plain Text First doctrine.

Architecture: ​

  • Plain Text as Canon: Markdown is the persistent Single Source of Truth. .docx is strictly an ingest and export format.
  • DOCX Ingest: Automatic Rust/Pandoc conversion removing Word XML styling, normalizing Unicode diacritics, mapping footnotes to [^1], and converting native Word Track Changes to CriticMarkup ({++inserted++}, {--deleted--}).
  • Scholarly Export Template: Compilation to .docx using a pre-configured reference template guaranteeing correct academic font hierarchies (Gentium Plus, Noto Serif Devanagari).
  • In-Editor CriticMarkup: Visual rendering of inline edits and comments in CodeMirror 6 with 1-click accept/reject controls.

Zentauri • Native Scholarly Markdown Editor