Competitive Study — Scientific Editors and Notebooks

Goal: identify the strengths and weaknesses of each competitor to build the best scientific notebook editor on the market.

---

1. Market Overview

Product	Type	Open Source	Primary Focus
Jupyter Notebook/Lab	Computational notebook	Yes	Code execution + visualization
Notion	Collaborative docs/wiki	No	General productivity, blocks
Overleaf	Online LaTeX editor	Partial	Academic LaTeX documents
Mathcha	Visual math editor	No	Formulas + diagrams WYSIWYG
Observable	Reactive notebook	Partial	Data visualization, JS
Quarto	Scientific publishing	Yes	Reproducible multi-language documents
Typst	Typesetting language	Yes	Modern alternative to LaTeX
CoCalc	Collaborative platform	Partial	Jupyter + LaTeX + terminal in the cloud
Deepnote	Collaborative notebook	No	Jupyter in the cloud + collaboration
HackMD/CodiMD	Collaborative Markdown	Yes	Real-time Markdown

---

2. Detailed Analysis by Competitor

2.1 Jupyter Notebook / JupyterLab

Strengths:

• De facto standard for scientific computing
• Code execution in 100+ languages via kernels
• Massive extension ecosystem
• .ipynb format widely adopted
• Rich output: charts, tables, interactive widgets

Weaknesses:

• Outdated UX — text editing is clunky
• No WYSIWYG for Markdown (separate edit/preview)
• LaTeX only in preview, no visual formula editor
• No drag & drop to reorder cells (JupyterLab added it partially)
• No contextual floating toolbar
• Limited themes, manual CSS customization
• No real-time collaboration (JupyterLab has experimental extension)
• Slow startup (Python server + kernel)

Opportunity for sci-notebook:

• Modern UX with click-to-edit, floating toolbar, insert handles
• Visual formula editor that Jupyter doesn't have
• Instant startup (no server, pure frontend)
• Better mobile experience

---

2.2 Notion

Strengths:

• Exceptional UX — the gold standard in block editing
• Slash commands (/) to insert any block type
• Smooth drag & drop reordering
• Inline databases, tables, kanban, calendars
• Real-time collaboration
• Reusable templates
• Robust public API

Weaknesses:

• No native LaTeX support (only inline with $$, basic rendering)
• No code execution
• No visual formula editor (raw LaTeX only)
• Closed, proprietary, vendor lock-in
• No robust offline mode
• Slow formula rendering compared to native KaTeX
• No advanced syntax highlighting for code

Opportunity for sci-notebook:

• Adopt Notion's block UX but with a scientific focus
• Visual formula editor >>> what Notion offers
• Open source, no vendor lock-in
• Better code support with real syntax highlighting

---

2.3 Overleaf

Strengths:

• Market-leading collaborative LaTeX editor
• Real-time compilation with PDF preview
• Thousands of academic templates
• Integration with journals for direct submission
• Git sync, Dropbox sync
• Complete version history

Weaknesses:

• Steep learning curve (requires knowing LaTeX)
• No real WYSIWYG mode (the "rich text" is limited)
• No cells/blocks — it's a plain text editor
• No code execution
• No interactive embeds
• Slow for large documents
• Dated, non-modern interface

Opportunity for sci-notebook:

• Visual formula editor that eliminates the LaTeX barrier
• Cell-based structure more flexible than plain text
• Interactive embeds (charts, videos, iframes)
• Superior performance (no need to compile PDF)

---

2.4 Mathcha

Strengths:

• Most intuitive WYSIWYG math editor on the market
• Visual palette of symbols and structures
• Integrated diagrams (tikz-like)
• Export to LaTeX, PDF, image
• No need to know LaTeX to write formulas

Weaknesses:

• Only formulas and diagrams — not a complete notebook
• No code cells
• No Markdown
• No plugins or extensibility
• No programmatic API
• Closed, proprietary
• No real-time collaboration

Opportunity for sci-notebook:

• Combine Mathcha's formula UX with a complete notebook
• Our MathEditor already has a similar block palette
• Add: drag & drop formula blocks, real-time preview
• Extensible via plugins

---

2.5 Observable

Strengths:

• Reactive notebooks — cells update automatically
• First-class data visualization (D3, Plot)
• Notebook import between users
• Native JavaScript execution in the browser
• Active visualization community

Weaknesses:

• JavaScript only (no Python, R, etc.)
• No robust LaTeX support
• No visual formula editor
• Proprietary data model (not .ipynb)
• No offline mode
• Learning curve for the reactive model

Opportunity for sci-notebook:

• Multi-format support (Markdown + LaTeX + code + embeds)
• Visual formula editor
• Standard, portable JSON format
• Observable embeds as iframe

---

2.6 Quarto

Strengths:

• First-class scientific publishing
• Multi-language (Python, R, Julia, Observable JS)
• Output to HTML, PDF, Word, presentations, books
• Integration with Jupyter kernels
• Cross-references, citations, bibliography
• Extensible via Lua filters

Weaknesses:

• Not an editor — it's a build/publishing system
• Requires CLI + external editor (VS Code, RStudio)
• No interactive UX in the browser
• No visual formula editor
• No drag & drop
• No real-time collaboration

Opportunity for sci-notebook:

• Be the visual editor that Quarto needs as a frontend
• Interactive UX that Quarto doesn't have
• Possible integration: export to Quarto format (.qmd)

---

2.7 Typst

Strengths:

• Modern alternative to LaTeX with simpler syntax
• Ultra-fast incremental compilation
• Real-time preview
• Built-in functions and scripting
• Better error handling than LaTeX
• Open source

Weaknesses:

• Young ecosystem, few packages
• No visual formula editor
• No cell/notebook model
• No external code execution
• Limited adoption vs LaTeX

Opportunity for sci-notebook:

• Possible plugin to render Typst in addition to LaTeX
• Our visual formula editor is something Typst doesn't have
• More flexible cell model

---

2.8 CoCalc

Strengths:

• Jupyter + LaTeX + terminal + chat in one platform
• Real-time collaboration with CRDT
• Granular version history
• Support for Sage, R, Julia, Octave
• TimeTravel (edit replay)

Weaknesses:

• Complex, overloaded UX
• Slow (remote server required)
• No visual formula editor
• Outdated interface
• Requires account and connection

Opportunity for sci-notebook:

• Clean, modern UX vs CoCalc's complexity
• Works offline, no server
• Visual formula editor

---

3. Feature Comparison Matrix

Feature	Jupyter	Notion	Overleaf	Mathcha	Observable	Quarto	Typst	CoCalc	sci-notebook
Cells/Blocks	✅	✅	❌	❌	✅	✅	❌	✅	✅
Markdown WYSIWYG	❌	✅	❌	❌	❌	❌	❌	❌	✅
LaTeX rendering	✅	⚠️	✅	✅	⚠️	✅	✅	✅	✅
Visual formula editor	❌	❌	❌	✅	❌	❌	❌	❌	✅
Symbol palette	❌	❌	❌	✅	❌	❌	❌	❌	✅
Image cells	✅	✅	✅	✅	✅	✅	✅	✅	✅
Image drag & drop	❌	✅	❌	❌	❌	❌	❌	❌	✅
Embeds/iframes	⚠️	✅	❌	❌	✅	✅	❌	❌	✅
Floating toolbar	❌	✅	❌	❌	❌	❌	❌	❌	✅
Insert handles	❌	✅	❌	❌	❌	❌	❌	❌	✅
Click-to-edit	❌	✅	N/A	N/A	❌	N/A	N/A	❌	✅
Drag reorder	⚠️	✅	❌	❌	❌	❌	❌	❌	✅
Keyboard nav	✅	✅	✅	⚠️	⚠️	N/A	✅	✅	✅
Light/dark themes	⚠️	✅	❌	❌	✅	✅	✅	⚠️	✅
Code execution	✅	❌	❌	❌	✅	✅	✅	✅	✅
RT collaboration	❌	✅	✅	❌	❌	❌	❌	✅	🔜
Open source	✅	❌	⚠️	❌	⚠️	✅	✅	⚠️	✅
Serverless	❌	❌	❌	❌	❌	❌	❌	❌	✅
Plugin system	✅	❌	❌	❌	❌	✅	✅	❌	✅
Framework agnostic	❌	❌	❌	❌	❌	❌	❌	❌	✅
Presentation mode	⚠️	❌	❌	❌	❌	✅	❌	❌	✅
PDF/DOCX export	⚠️	❌	✅	✅	❌	✅	✅	✅	✅
Mobile support	❌	✅	❌	❌	❌	❌	❌	❌	✅

Legend: ✅ = Yes | ⚠️ = Partial | ❌ = No | 🔜 = Planned

---

4. Unique Competitive Advantages of sci-notebook

4.1 What NOBODY else has (combined)

1. Visual formula editor + notebook: Only Mathcha has a visual editor, but it's not a notebook. Only Jupyter has a notebook, but no visual editor. sci-notebook is the only one that combines both.

2. Framework agnostic + serverless: Works 100% in the browser with no backend. No competitor offers this with the same level of features.

3. Plugin system + modern UX: Jupyter's extensibility with Notion's UX.

4. Open source + scientific focus: Quarto is open source but not an editor. Jupyter is open source but the UX is outdated.

4.2 Key Differentiators

Differentiator	Description
Visual MathEditor	9 categories, 100+ blocks, real-time preview, dual mode (visual/raw)
Zero-server	Pure TypeScript, works offline, embed in any app
Micro-bundle	Core ~45KB, React adapter ~80KB, vs Jupyter (MB of Python + JS)
Instant boot	No kernel startup, no compilation, immediate rendering
Composable	Every feature is an opt-in plugin, tree-shakeable

---

5. Identified Gaps — What We Were Missing

5.1 Critical (necessary to compete)

#	Gap	Reference Competitor	Priority
G1	Slash commands (/ to insert)	Notion	🔴 High
G2	Drag & drop reorder of cells	Notion, JupyterLab	🔴 High
G3	Code execution (at least JS/Python via Pyodide)	Jupyter, Observable	🔴 High
G4	Syntax highlighting in code cells	Jupyter, VS Code	🔴 High
G5	Export to PDF, HTML, Markdown, .ipynb	Quarto, Jupyter	🔴 High
G6	Real-time collaboration (Yjs/CRDT)	Notion, Overleaf, CoCalc	🟡 Medium

5.2 Important (quality differentiators)

#	Gap	Reference Competitor	Priority
G7	Table editor interactive within cells	Notion	🟡 Medium
G8	Mermaid diagrams rendering	Jupyter, Quarto	🟡 Medium
G9	TOC sidebar (table of contents)	Notion, JupyterLab	🟡 Medium
G10	Find & replace across cells	Overleaf, Jupyter	🟡 Medium
G11	Cell output display (for execution)	Jupyter	🟡 Medium
G12	Version history / diffing	Overleaf, CoCalc	🟡 Medium

5.3 Nice-to-have (polish)

#	Gap	Reference Competitor	Priority
G13	Presentation mode (slideshow)	Jupyter RISE	🟢 Low
G14	Comments / annotations	Notion, Google Docs	🟢 Low
G15	Templates gallery	Notion, Overleaf	🟢 Low
G16	Mobile-optimized UI	Notion	🟢 Low
G17	Autocomplete for LaTeX commands	Overleaf	🟢 Low
G18	Citation management (BibTeX)	Overleaf, Quarto	🟢 Low

---

6. Strategy — How to Be the Best

Immediate Phase (v0.2) ✅ COMPLETE

• [x] G1: Slash commands — SlashCommand.tsx with 8 types, filter, keyboard nav
• [x] G2: Drag & drop reorder with top/bottom indicator
• [x] G4: Syntax highlighting with Shiki (lazy loading, 30+ languages)
• [x] G8: Mermaid diagrams — renderMermaidFallback() in pipeline

Short Phase (v0.3) ✅ COMPLETE

• [x] G3: JS code execution in sandbox + custom language executors
• [x] G5: Export to standalone HTML + Markdown + .ipynb + JSON
• [x] G7: Interactive table editor — TableCell.tsx
• [x] G9: TOC sidebar — TOCSidebar.tsx

Medium Phase (v0.4) ✅ COMPLETE

• [x] G10: Global find & replace — FindReplace.tsx with Ctrl+F
• [x] G11: Cell outputs — CellOutputDisplay.tsx (stream/display/error)
• [x] G17: LaTeX autocomplete — 120+ commands in 8 categories
• [x] G12: Version history — VersionHistory with save/restore/diff

Long Phase (v1.0) ✅ COMPLETE (partial)

• [x] G13: Presentation mode — PresentationEngine with 3 split modes, transitions, fullscreen
• [x] G16: Mobile-optimized UI — MobileAdapter with touch gestures, responsive CSS
• [x] G5 (extended): PDF/DOCX export — plugin-export with browser print and Office Open XML
• [ ] G6: Real-time collaboration (Yjs) — out of scope v1.0
• [ ] G14: Comments — out of scope v1.0
• [ ] G18: Citations — out of scope v1.0

---

7. Design Principles for "The Best Editor"

1. Instant feedback: Every user action produces immediate visual result (<16ms)
2. Zero friction: Never more than 2 clicks for any common operation
3. Progressive disclosure: Simple by default, powerful when you need it
4. Keyboard-first: Everything accessible by keyboard, mouse is optional
5. Offline-first: Works without connection, syncs when online
6. Composable: Every feature is a plugin, nothing is mandatory
7. Beautiful defaults: Looks professional without configuration
8. Accessible: WCAG 2.1 AA, screen readers, high contrast
9. Fast: <100ms for any operation, <1s for boot
10. Open: Standard JSON format, no vendor lock-in, MIT license

---

8. Conclusion

sci-notebook occupies a unique niche: scientific notebook editor with modern UX, visual formula editor, and plugin architecture — all serverless, open source, and framework agnostic.

No competitor combines these characteristics. The strategy is:

1. Maintain the unique advantages (MathEditor, zero-server, plugins)
2. Close the critical gaps (slash commands, drag reorder, code execution, syntax highlighting) — ✅ DONE
3. Polish the UX to surpass Notion in the scientific niche
4. Build community with excellent docs and working examples