Skip to content

Latest commit

 

History

History
733 lines (538 loc) · 26.7 KB

File metadata and controls

733 lines (538 loc) · 26.7 KB

Torus Documentation TODOs

This file contains comprehensive documentation tasks to bridge the conceptual gaps between the sophisticated Torus blockchain implementation and user understanding. The approach uses sequential learning paths - breaking complex topics into digestible, connected pieces that build on each other.

HIGH PRIORITY - Core Understanding

1. Economics Learning Path (4 Sequential Files)

Overall Goal: Transform user understanding from "simple staking" to sophisticated multi-stakeholder economic system with matrix-based reward distribution.

Learning Progression: Foundation → Distribution → Technical Depth → Advanced Economics

1.1 validator-miner-roles.mdx (Foundation - 5 min read)

Core Concept: Two distinct participant types in Torus with different functions and rewards.

Why This First: Users must understand the roles before they can understand the reward system.

Key Points:

  • Validators assess quality and assign weights (like judges)
  • Miners provide actual services via APIs (like service providers)
  • Different roles = different reward streams (dividends vs incentives)
  • Simple analogy: Validators are critics, miners are performers

Small Example:

Alice (Validator): Reviews and scores service quality
Bob (Miner): Runs Twitter memory API service
Carol (Miner): Provides GPU compute services

Alice assigns weights: Bob=100, Carol=200 (Carol provides better service)
Results: Bob gets incentives, Carol gets 2x incentives, Alice gets dividends

Code References:

  • torus-substrate/README.md - Terminology section
  • torus-substrate/pallets/torus0/src/agent.rs - Agent types
  • torus-substrate/docs/linear-emission.md - Validator/miner distinction

Leads To: "Now that you understand the roles, let's see how they get rewarded..."


1.2 emission-distribution-basics.mdx (Building on Roles - 7 min read)

Core Concept: 64,000 tokens per day distributed every 100 blocks based on validator weight assignments.

Why This Next: Users need to understand the basic flow before complex calculations.

Key Points:

  • Daily emission: 64,000 TORUS tokens (10,800 blocks, ~8 seconds each)
  • Distribution cycle: Every 100 blocks (~13 minutes)
  • Weight assignment drives distribution
  • Pending emission accumulates then distributes

Small Example:

Day 1, Block 1-100: Accumulate ~593 TORUS tokens
Block 100: Distribution time!

Alice's weights: Bob=100, Carol=200 (total=300)
Alice's stake: 1000 TORUS

Distribution calculation (simplified):
- Bob gets: (100/300) × Alice's_influence × total_pool
- Carol gets: (200/300) × Alice's_influence × total_pool  
- Alice gets dividends based on their performance

Code References:

  • torus-substrate/docs/linear-emission.md - Accumulation and distribution
  • torus-substrate/pallets/emission0/src/distribute.rs - Distribution logic
  • torus-substrate/pallets/emission0/src/lib.rs - Block emission parameters

Leads To: "But how exactly are these weights turned into precise token amounts?"


1.3 matrix-reward-calculations.mdx (Technical Depth - 10 min read)

Core Concept: Matrix mathematics ensures fair multi-party reward distribution.

Why This Next: Users who want deeper understanding need to see the mathematical foundation.

Key Points:

  • Sparse matrix multiplication: weights × stakes = miner ranks
  • Hadamard product: element-wise multiplication for validator bonds
  • Normalization: ensures fair distribution regardless of network size
  • Feedback loop: successful miners increase their validators' dividends

Small Example:

Simple 2×2 case:
Validators: Alice (1000 stake), Dave (500 stake)
Miners: Bob, Carol

Weight Matrix:    Stake Vector:    Results:
Alice: [100, 200]    [1000]    →  Miner incentives
Dave:  [50,  300]    [500]     →  Validator dividends

Matrix math ensures:
- Higher stake validators have more influence
- Better miners (higher weights) get more rewards
- Validators who pick winners get more dividends

Code References:

  • torus-substrate/pallets/emission0/src/distribute/math.rs - Matrix operations
  • torus-substrate/docs/linear-emission.md - Mathematical explanation
  • torus-substrate/pallets/emission0/src/distribute.rs - Implementation

Leads To: "These calculations create different income streams and fees..."


1.4 fee-economics.mdx (Advanced Economics - 8 min read)

Core Concept: Delegation fees and weight control fees create sustainable economic incentives.

Why This Last: Users need to understand the full system before grasping the fee structures.

Key Points:

  • Delegation fees: Validators charge stakers for using their validation services
  • Weight control fees: Validators can delegate weight-setting to specialists
  • Economic incentives: Fees align interests between all parties
  • Sustainability: Good performance attracts more delegation

Small Example:

Complete Economic Flow:
1. Alice (1000 TORUS) stakes with Dave (validator)
2. Dave charges 10% delegation fee
3. Dave delegates weight control to Expert Eve (5% fee)
4. Eve assigns weights: Bob=100, Carol=200
5. Distribution happens:
   - Bob/Carol get incentives
   - Dave gets dividends minus fees to Eve and Alice
   - Alice gets staking rewards minus Dave's fee
   - Eve gets weight control fee for her expertise

Code References:

  • torus-substrate/pallets/torus0/src/fee.rs - Fee mechanisms
  • torus-substrate/pallets/emission0/src/weight_control.rs - Weight delegation
  • torus-substrate/docs/linear-emission.md - Fee processing

Connects Back: "This completes the economic loop we started with validator/miner roles"


2. Advanced Permissions Learning Path (3 Sequential Files)

Overall Goal: Move from basic "delegation" understanding to sophisticated recursive permission trees with enforcement.

Learning Progression: Basic Trees → Control Mechanisms → Advanced Governance

2.1 basic-delegation-trees.mdx (Foundation - 6 min read)

Core Concept: Recursive permission chains enable specialization at multiple levels.

Why This First: Users need to grasp basic recursion before complex controls.

Key Points:

  • Delegation chains: Alice → Bob → Carol → Dave
  • Percentage compounding: 30% × 50% × 20% = 3% of original
  • Specialization benefits: Each level adds specific value
  • Positive-sum economics: Everyone benefits when the chain works

Small Example:

Delegation Tree:
Alice (Root emissions) 
  ↓ delegates 30% to Bob (Platform operator)
    Bob (Platform services)
      ↓ delegates 50% to Carol (AI specialist) 
        Carol (AI services)
          ↓ delegates 20% to Dave (Model trainer)

Result: Dave gets 3% of Alice's original emissions
But: Dave's specialization makes the whole chain more valuable

Code References:

  • torus-substrate/pallets/permission0/src/permission.rs - Permission structures
  • torus-substrate/docs/permission0.md - Recursive delegation concept

Leads To: "But what if you want more control over when payments happen?"


2.2 distribution-controls.mdx (Building Complexity - 7 min read)

Core Concept: Different distribution mechanisms (manual, automatic, timed, threshold) serve different business relationships.

Why This Next: Users need to understand timing/control options before governance mechanisms.

Key Points:

  • Manual: Recipient requests payment when work is done
  • Automatic: Payment when accumulated amount hits threshold
  • Timed: Payment at specific block or intervals
  • Different trust levels require different controls

Small Example:

Business Scenarios:
- Manual: "Pay me when I deliver the analysis" (project work)
- Automatic: "Pay me every 1000 TORUS accumulated" (ongoing service)
- AtBlock: "Pay me at block 500000" (milestone payment)
- Interval: "Pay me every 1000 blocks" (salary-like)

Same delegation, different timing = different business relationships

Code References:

  • torus-substrate/pallets/permission0/src/permission.rs - DistributionControl enum
  • torus-substrate/docs/permission0.md - Distribution mechanisms

Leads To: "Sometimes you need third parties to verify conditions..."


2.3 enforcement-authorities.mdx (Advanced Governance - 9 min read)

Core Concept: Off-chain verification with on-chain execution enables complex business conditions.

Why This Last: Most sophisticated concept requiring understanding of previous permission mechanics.

Key Points:

  • Bridge real-world conditions to blockchain execution
  • Multi-signature governance for permission control
  • Off-chain verification, on-chain enforcement
  • Enable complex business relationships that can't be fully coded

Small Example:

Scenario: AI Service with Quality Requirements
1. Alice delegates to Bob's AI service
2. Enforcement authority: Committee of 3 AI experts
3. Condition: Service must maintain 95% accuracy
4. Process:
   - Experts verify off-chain performance metrics
   - 2 of 3 experts must vote to continue payments
   - If performance drops, experts can pause accumulation
   - Blockchain enforces the committee's decisions

Code References:

  • torus-substrate/pallets/permission0/src/permission.rs - EnforcementAuthority
  • torus-substrate/docs/permission0.md - Enforcement system
  • torus-substrate/pallets/permission0/src/ext/ - Implementation

Connects Back: "These authorities can control the delegation trees you learned about"


3. Network Management Learning Path (3 Sequential Files)

Overall Goal: Show how Torus maintains network health through sophisticated agent lifecycle management.

Learning Progression: Individual Journey → Scarcity Economics → Selection Mechanisms

3.1 agent-registration-lifecycle.mdx (Foundation - 6 min read)

Core Concept: Agent journey from registration through participation to potential removal.

Why This First: Users need to understand individual experience before system-wide mechanics.

Key Points:

  • Registration → metadata setup → weight assignment → reward participation
  • Not all agents participate equally (consensus vs general participation)
  • Performance tracking and potential pruning
  • Economic incentives at each stage

Small Example:

Bob's Agent Journey:
Day 1: Registers agent, pays fee, sets up metadata
Day 2-7: Builds reputation, starts getting small weights
Day 8-30: Consistent service, weight assignments increase
Day 31+: Either thriving (more weights) or struggling (risk pruning)

Success path: Better service → higher weights → more rewards
Failure path: Poor service → low weights → potential removal

Code References:

  • torus-substrate/pallets/torus0/src/agent.rs - Agent lifecycle
  • torus-substrate/README.md - Agent terminology

Leads To: "But what happens when too many agents want to join?"


3.2 network-capacity-economics.mdx (Scarcity Mechanics - 7 min read)

Core Concept: Dynamic burn mechanism maintains network quality through economic scarcity.

Why This Next: Users need to understand capacity constraints before selection mechanisms.

Key Points:

  • Network has optimal size for efficiency
  • Burn costs increase when network approaches capacity
  • Economic barrier prevents spam/low-quality registrations
  • Market-based quality control

Small Example:

Network States:
50% capacity: Registration fee = 10 TORUS (base cost only)
80% capacity: Registration fee = 25 TORUS (burn mechanism activates)
95% capacity: Registration fee = 100 TORUS (high barrier to entry)
100% capacity: Registration requires someone else to be pruned

Result: Only high-value agents join when network is full
Quality increases as scarcity increases

Code References:

  • torus-substrate/pallets/torus0/src/burn.rs - Burn mechanism
  • torus-substrate/README.md - Burn terminology

Leads To: "Who gets to participate when space is limited?"


3.3 consensus-participation.mdx (Selection Mechanisms - 8 min read)

Core Concept: Not all agents become validators; selection based on stake and governance trust.

Why This Last: Most complex concept requiring understanding of agent lifecycle and capacity constraints.

Key Points:

  • Consensus members vs general agents distinction
  • Stake thresholds for validator participation
  • Allocator roles (governance-selected trusted validators)
  • Lazy registration during reward epochs

Small Example:

Agent Categories:
- General agents: Can receive weights, participate in economy
- Validators: Top staked agents, can assign weights to others
- Allocators: Governance-trusted validators, special permissions
- Consensus members: Active participants in current reward epoch

Selection Process:
1. All agents can register and provide services
2. Only top-staked become validators (limited slots)
3. Governance selects some validators as allocators  
4. Only consensus members participate in each reward cycle

Code References:

  • torus-substrate/pallets/governance/src/roles.rs - Allocator roles
  • torus-substrate/pallets/emission0/src/distribute.rs - Consensus member management
  • torus-substrate/docs/linear-emission.md - Consensus member aggregation

Connects Back: "This determines who can assign the weights we learned about in economics"


MEDIUM PRIORITY - System Understanding

4. Advanced Governance Learning Path (2 Sequential Files)

Overall Goal: Move beyond basic DAO voting to understand sophisticated governance economics and specialized roles.

Learning Progression: Economic Incentives → Specialized Governance

4.1 governance-economics.mdx (Foundation - 6 min read)

Core Concept: Quadratic proposal rewards and treasury mechanics create democratic governance incentives.

Why This First: Users need to understand the economic incentives before specialized roles.

Key Points:

  • Quadratic formula: √stake gives voting power (favors smaller stakeholders)
  • Treasury funding from fees and emissions
  • Proposal rewards incentivize participation
  • Economic democracy vs plutocracy

Small Example:

Proposal Voting Power:
- Whale (10,000 TORUS): √10,000 = 100 voting power
- Medium (1,000 TORUS): √1,000 = 31.6 voting power  
- Small (100 TORUS): √100 = 10 voting power
- 1000 tiny holders (1 TORUS each): 1000 × √1 = 1000 voting power

Result: Collective small holders > individual whale
Economic incentive: Everyone benefits from participation

Code References:

  • torus-substrate/pallets/governance/src/proposal.rs - Proposal rewards
  • torus-substrate/pallets/governance/src/voting.rs - Quadratic mechanics

Leads To: "But who can perform specialized governance functions?"


4.2 specialized-governance-roles.mdx (Advanced Roles - 7 min read)

Core Concept: Curator permissions and allocator roles enable specialized governance beyond general voting.

Why This Next: Builds on economic understanding to show governance specialization.

Key Points:

  • Allocators: Governance-trusted validators with special permissions
  • Curators: Specialized roles (application review, whitelist management, penalties)
  • Permission delegation for governance functions
  • Multi-signature governance mechanisms

Small Example:

Governance Specialization:
- General voters: Vote on proposals, get quadratic rewards
- Allocators: Chosen by governance, can assign weights to new agents
- Curators: Delegated specific powers:
  * Application review: Screen new agent applications
  * Whitelist management: Add/remove accounts from whitelist
  * Penalty control: Apply performance penalties

Result: Specialized governance scales better than direct democracy

Code References:

  • torus-substrate/pallets/governance/src/roles.rs - Specialized roles
  • torus-substrate/pallets/permission0/src/permission/curator.rs - Curator permissions

Connects Back: "These roles work within the economic incentives we learned about"


5. System Evolution Learning Path (2 Sequential Files)

Overall Goal: Help users understand current capabilities vs future vision and make informed participation decisions.

Learning Progression: Current State → Future Evolution

5.1 v05-current-capabilities.mdx (Current State - 6 min read)

Core Concept: v0.5 implements core building blocks but not full recursive competency vision.

Why This First: Users need to understand what's available now before future plans.

Key Points:

  • "Pragmatic, growth-oriented" implementation
  • Core mechanisms work: emission, permissions, namespaces, governance
  • Agent swarms possible but require manual coordination
  • "Battle testing" current systems

Small Example:

What Works Today:
✅ Alice can stake and earn rewards
✅ Bob can register agent and provide services  
✅ Carol can delegate permissions to Dave
✅ Network governance functions
✅ Economic incentives align participants

What Requires Manual Work:
🚧 Complex multi-agent coordination
🚧 Automatic swarm formation
🚧 Advanced competency discovery

Code References:

  • torus-substrate/README.md - v0.5 positioning
  • torus-docs/src/content/docs/getting-started/understanding-torus.mdx - Current description

Leads To: "How will this evolve into the full vision?"


5.2 evolution-roadmap.mdx (Future Vision - 7 min read)

Core Concept: Path from current building blocks to full self-assembling agent swarms.

Why This Next: Users need roadmap context to make long-term participation decisions.

Key Points:

  • Runtime upgrades preserve state while adding features
  • Storage migrations enable system evolution
  • v1 target: Full recursive competency networks
  • Current participation prepares for future capabilities

Small Example:

Evolution Path:
v0.5 (Today): Manual agent coordination
- "Alice manually finds Bob's AI service"
- "Carol manually delegates to Dave"

v1 (Future): Automatic swarm assembly  
- "Alice signals need for Twitter analysis"
- "Network automatically assembles: Bob→Carol→Dave chain"
- "Swarm self-optimizes based on performance"

Bridge: Current building blocks → Future automation

Code References:

  • torus-substrate/pallets/*/migrations.rs - Upgrade mechanisms
  • Storage version management across pallets

Connects Back: "The economic and permission systems you learned enable this future automation"


COMPANION HOW-TO GUIDES NEEDED

Several conceptual files would benefit from practical how-to companions that let users apply what they learned. These should be developed after the conceptual foundation is established.

High Priority How-To Guides (Phase 2)

From Economics Path:

validator-miner-roles.mdx needs: "How to Choose Your Role: Validator vs Miner"

  • Content: Role decision framework, resource requirements, setup steps
  • Why Important: Fundamental participation decision users need guidance on
  • Links to: Existing agent registration guides

fee-economics.mdx needs: "How to Set and Optimize Your Fees"

  • Content: Setting delegation fees, optimizing rates, fee market dynamics
  • Why Important: Economic optimization many users miss
  • New Guide Required: Not currently covered

From Permissions Path:

distribution-controls.mdx needs: "How to Choose Distribution Controls for Your Business"

  • Content: Manual vs automatic vs timed decision trees, threshold setting
  • Why Important: Critical for permission effectiveness, common confusion point
  • New Guide Required: Advanced topic not covered

enforcement-authorities.mdx needs: "How to Set Up Enforcement Authorities"

  • Content: Authority selection, multi-sig setup, condition definition
  • Why Important: Powerful but complex feature, enables sophisticated contracts
  • New Guide Required: Advanced governance topic

From Governance Path:

governance-economics.mdx needs: "How to Maximize Your Governance Participation"

  • Content: Voting strategies, proposal timing, reward optimization
  • Why Important: Economic opportunity users often miss
  • Enhancement Needed: Expand existing governance guides

Medium Priority How-To Guides (Phase 3)

From Economics Path:

emission-distribution-basics.mdx could use: "How to Track and Predict Your Emissions"

  • Content: Understanding your reward calculations, timing predictions
  • Why Useful: Helps users plan and optimize participation
  • New Guide Required: Analytics/tracking focused

From Network Path:

agent-registration-lifecycle.mdx could use: "How to Optimize Your Agent for Long-term Success"

  • Content: Performance optimization, reputation building, avoiding pruning
  • Why Useful: Agent sustainability strategies
  • Enhancement: Build on existing registration guides

Cross-Linking Strategy

Concept → Practice Flow:

Each conceptual file should end with:

## Ready to Apply This?

Now that you understand [concept], put it into practice:
- 📋 [Primary How-to Guide] - Step-by-step implementation
- 🔗 [Related How-to Guide] - Connected actions

## Continue Learning
📖 Next concept: [Next file in learning path]
🔄 Review: [Previous concept if needed]

Practice → Concept Flow:

Each how-to guide should include:

## Why This Works

For deeper understanding of the concepts behind these steps:
- 🧠 [Conceptual explanation] - Economic/technical background
- 💡 [Advanced concepts] - Next level understanding

Updates to Existing How-To Guides

Enhancement Opportunities:

Current agent registration guides → Add links to:

  • Agent lifecycle concepts (why registration matters)
  • Validator vs miner role concepts (strategic decisions)

Current delegation guides → Add links to:

  • Delegation tree concepts (economic reasoning)
  • Distribution control concepts (timing strategies)

Current staking guides → Add links to:

  • Economics concepts (how rewards really work)
  • Fee economics (optimization strategies)

Implementation Phases

Phase 1: Conceptual Foundation (Priority)

Write all 14 conceptual files first. Users need to understand "why" before "how."

Phase 2: High-Impact How-Tos

Focus on guides that address common confusion or unlock significant value:

  1. Distribution controls guide (biggest confusion point)
  2. Enforcement authorities guide (powerful but complex)
  3. Governance optimization guide (missed economic opportunity)
  4. Role selection guide (fundamental decision)

Phase 3: Enhancement and Polish

  • Cross-link existing guides with new concepts
  • Add medium-priority how-to guides based on user feedback
  • Create learning path navigation

Success Metrics:

  • Conceptual understanding: Users ask sophisticated questions instead of basic ones
  • Practical application: Increased usage of advanced features (enforcement authorities, complex distributions)
  • Economic optimization: Better fee strategies, governance participation
  • Cross-referencing: Users naturally move between concept and practice content

IMPLEMENTATION GUIDANCE

Sequential Learning Implementation

Directory Structure:

/explanations/
  /economics/          # 4-file learning path
    - validator-miner-roles.mdx
    - emission-distribution-basics.mdx  
    - matrix-reward-calculations.mdx
    - fee-economics.mdx
  /permissions/        # 3-file learning path
    - basic-delegation-trees.mdx
    - distribution-controls.mdx
    - enforcement-authorities.mdx
  /network/           # 3-file learning path
    - agent-registration-lifecycle.mdx
    - network-capacity-economics.mdx
    - consensus-participation.mdx
  /governance/        # 2-file learning path
    - governance-economics.mdx
    - specialized-governance-roles.mdx
  /evolution/         # 2-file learning path
    - v05-current-capabilities.mdx
    - evolution-roadmap.mdx

Each File Should Include:

  1. Single Core Concept - One main idea per file
  2. Why This Now - Progression rationale
  3. Concrete Examples - Real scenarios users can relate to
  4. Code References - Where to find implementation
  5. Forward Link - "Next, learn about..."
  6. Backward References - "Remember from previous file..."

Writing Style Guidelines

For Conceptual Documentation:

  • Start with "Why does this exist?" before "How does it work?"
  • Use concrete scenarios over abstract descriptions
  • Show economic flows and incentive structures
  • Connect to user goals and decision-making
  • Avoid technical implementation details
  • Use analogies from familiar systems

Visual Aid Requirements

Essential Diagrams for Each Topic:

  • System flow diagrams showing cause and effect
  • Economic relationship maps
  • Decision trees for user actions
  • Timeline visualizations for processes
  • Comparative tables for understanding trade-offs

Cross-References Strategy

Link Integration:

  • Connect new content to existing how-to guides
  • Reference related concepts in other explanations
  • Create learning paths from basic to advanced concepts
  • Maintain glossary updates with new terminology

Validation Approach

Content Accuracy:

  • Cross-reference with substrate implementation code
  • Validate examples against actual blockchain behavior
  • Test user comprehension with community feedback
  • Update when substrate code evolves

NOTES FOR IMPLEMENTATION

Priority Order Rationale

  1. Economics (4 files) - Most fundamental misunderstanding, sequential learning essential
  2. Permissions (3 files) - Core to Torus vision, builds complexity gradually
  3. Network Management (3 files) - Important for participation decisions
  4. Governance (2 files) - Build on existing foundation, shorter learning path
  5. Evolution (2 files) - Context setting, helps with expectations

Total: 14 digestible files instead of 5 overwhelming documents

Note: Namespace system is already well-covered in /explanations/v06/recursive-delegation.mdx and /explanations/v05/control-space.mdx

Resource Requirements

  • Technical writer familiar with blockchain economics
  • Graphic designer for system diagrams
  • Substrate developer for code reference validation
  • Community feedback mechanism for comprehension testing

Success Metrics

  • User questions shift from "how do I stake?" to "how do I optimize my delegation tree?"
  • Reduced confusion about validator vs miner roles
  • Increased sophisticated usage of permission system
  • Better informed governance participation

RELATED FILES TO UPDATE

When implementing these TODOs, also update:

  • /explanations/concepts-terminology.mdx - Add new sophisticated terms
  • Existing how-to guides - Reference new conceptual understanding
  • /getting-started/understanding-torus.mdx - Deepen the introduction
  • Navigation structure - Organize new explanations logically

The goal is transforming user understanding from "Torus is a staking network" to "Torus is a sophisticated coordination protocol with recursive economic incentives for autonomous agent swarms."