Polkadot Whitepaper - Read Through and Analysis

[NicolaBernini]

Read Through and Analysis of the Polkadot Whitepaper

[NicolaBernini]

Blockchain Technology Limitations

Main limitations of current Blockchain Technologies

• Scalability
• Isolatability
• Developability
• Governance
• Applicability

[NicolaBernini]

Scalability Issues

Resources

Mainly regarding Resources spent per Tx Processing

• At Node Level

  • Cost of Validation
  • Cost of Consensus (PoW, PoS, ...)

• At Blockchain Level

  • How much of the computation is replicated across the nodes

Design

• Having a state model which relies on a canonical global state only, means having no parallelization at all
  • In global state context, Block adding is a Blockchain wide synchronization point hence it slows up computation
  • Furthermore, Consensus like PoW make this event even slower in order to make it both expensive and random
• The idea is scaling performance by improving the parallelization at state model level, which introduces the idea of decoherence (relaxing the vinculus of evolving the BC state machine by global states only)

[NicolaBernini]

Tradeoffs

The following one vs the others ones

• Scalability
• Security
• Privacy

[NicolaBernini]

Polkadot Project Governance

• Open Source
• Community Driven
  • Based on RFC System

[NicolaBernini]

Polkadot Architecture

Idea

• Think about the Actor Model : each Blockchain is an actor which runs its own processing independently and in parallel with the respect to the other Actors, Polkadot provides essentially a Decentralized Message Passing Framework
• More Scalability is achieved by increasing Parallelism

Elements

• Relay Chain: the Polkadot Core
• Parachains (Parallel Blockchains): Blockchain connecting to Polkadot = Actors
• Bridges: allow to connect to Blockchain with their own Consensus

[NicolaBernini]

Fundamental Layers

• Focusing on Distributed Computation only (so considering Decentralization and Trustless Context at a later stage) there are 2 fundamental needs: Distributed Communication and Standardized Computation

Distributed Communication

Requirements

• Needs to be P2P
• Needs to provide enough flexibility to implement a huge amount of higher level protocols
• Needs to be computationally efficient as it is focused on running on multiple devices, with different HW and available resources in general (e.g. bandwidth, battery, …)

Candidates

• Libp2p

Standardized Computation

Requirements

• Needs to guarantee any operation gives the same result independently on the underlying computational infrastructure hence providing a full abstraction layer
• Needs to standardize the computation cost even considering a high level measurement unity (e.g. Ethereum Gas) : e.g. a certain operation needs to cost X gas independently of the platform
  -- Similar to C++ Specs specifying the computational complexity of operations on containers

Candidates

• Ethereum Virtual Machine

  • Good but a so far could be a little too specific

• WebAssembly

  • Very promising new standard
  • Possibly in the future of many Blockchains