Bitcoin Network, Forks & Proof of Stake
Bitcoin proved the concept. Now we explore how the network really works, how it evolves, and the revolution that replaced mining with staking.
Recap: Where We Left Off
In Module 4, you learned how Bitcoin moves from one machine to a decentralised network. Gossip protocol spreads transactions, Nakamoto consensus resolves disagreements, and proof of work makes cheating expensive. But we only scratched the surface.
This Module's Road Map
We'll explore the real Bitcoin network's architecture (full nodes, SPV, miners), how the protocol upgrades via BIPs, the critical difference between hard and soft forks, and then cross the bridge from Proof of Work to Proof of Stake — the consensus mechanism that now secures Ethereum and most modern chains.
Module 4 — PoW
P2P gossip, Nakamoto consensus, longest chain wins, 51% attack. Secure but energy-hungry.
Module 5 — Evolution
Node types, governance, forks, Proof of Stake, the Merge. How blockchains grow and adapt.
Module 6 — EVM
From Bitcoin Script to Ethereum's programmable virtual machine. Part 2 begins.
Bitcoin Network Architecture
Not every computer on the Bitcoin network does the same job. There are several types of nodes, each with different responsibilities and resource requirements.
Full Node
Downloads and validates every block and transaction since 2009. Stores the entire blockchain (~600 GB). Enforces all consensus rules independently. The backbone of decentralisation.
SPV / Light Node
Only downloads block headers (~60 MB). Verifies transactions using Merkle proofs. Your phone wallet is an SPV client — it trusts full nodes but still checks Merkle paths.
Mining Node
A full node that also competes to find valid blocks. Runs ASICs or GPUs. Collects transaction fees + block reward. Today dominated by mining pools.
How SPV Works: Merkle Proofs
Remember the Merkle tree from Module 2? An SPV node asks a full node: "Is my transaction in block #820,000?" The full node returns a Merkle proof — the minimal set of hashes needed to compute the Merkle root. The SPV node checks: does this Merkle root match the block header I already have? If yes → transaction is confirmed. No need to download all ~4,000 transactions in the block.
| Property | Full Node | SPV Node | Mining Node |
|---|---|---|---|
| Storage | ~600 GB | ~60 MB | ~600 GB + mempool |
| Validates all rules | ✅ | ❌ | ✅ |
| Trust required | None | Trusts full nodes | None |
| Creates blocks | ❌ | ❌ | ✅ |
| Est. count (2025) | ~18,000 | Millions | ~15 pools |
Other Network Participants
Archive nodes store every historical state (common in Ethereum). Relay networks like FIBRE and Falcon provide low-latency block propagation between miners — reducing orphan blocks. Block explorers (like mempool.space) are full nodes with APIs so you can query the chain from a browser.
BIPs: How Bitcoin Evolves
Bitcoin has no CEO, no board of directors, no update server. Yet the protocol has been upgraded dozens of times since 2009. How? Through Bitcoin Improvement Proposals (BIPs) — a transparent, community-driven process.
Notable BIPs
Key BIPs That Shaped Bitcoin
- BIP 141 — SegWit (2017): Separated signature data from transactions, fixing malleability and increasing effective block capacity from ~1 MB to ~2.3 MB.
- BIP 340 — Schnorr Signatures (2021): More efficient, smaller signatures. Enabled batched validation — verifying 1,000 signatures nearly as fast as one.
- BIP 341 — Taproot (2021): Made complex smart contracts look like simple payments on-chain. Privacy + efficiency improvement.
- BIP 32 — HD Wallets: Hierarchical deterministic key derivation. Generate unlimited addresses from a single seed phrase.
- BIP 39 — Mnemonic Phrases: The 12/24-word seed phrase you write on paper. Maps random entropy to human-readable words.
Key Insight: No One Controls Bitcoin
The BIP process is deliberately slow and conservative. Changing consensus rules requires convincing thousands of independent node operators to voluntarily upgrade their software. This "rough consensus" model is both Bitcoin's greatest strength (censorship resistance) and its biggest frustration (changes take years). Compare this to Ethereum's more agile governance in Section 6.
Hard Forks vs Soft Forks
In Module 4, we saw natural forks — temporary chain splits that resolve within a few blocks. But protocol forks are different: they're deliberate rule changes. And the type of fork determines whether the community stays together or splits in two.
Soft Fork
Tightens the rules. Old nodes still accept new blocks (backward compatible). New nodes enforce stricter rules. The chain stays as one. Example: SegWit.
Hard Fork
Loosens or changes rules. Old nodes reject new blocks. The chain permanently splits in two. Each side keeps its own history. Example: Bitcoin Cash.
| Aspect | Soft Fork | Hard Fork |
|---|---|---|
| Backward compatible? | ✅ Yes | ❌ No |
| Chain split? | Only if miners reject | Always — two chains |
| Upgrade required? | Optional (but recommended) | Mandatory for new chain |
| Risk level | Lower — network stays unified | Higher — community can fracture |
| Famous example | SegWit (BIP 141, 2017) | Bitcoin Cash (Aug 2017) |
Case Study: The Block Size War
In 2015–2017, the Bitcoin community fought over how to scale. One camp wanted bigger blocks (8 MB) for more transactions per second. The other wanted SegWit — a soft fork that improved capacity without changing the block size limit. The debate lasted two years, involved corporate agreements (New York Agreement), and ended with both: SegWit activated on Bitcoin, while the big-block camp hard-forked to create Bitcoin Cash (BCH) in August 2017. BCH itself later split into BCH and BSV. This is why hard forks are considered "nuclear options."
Ethereum's Hard Fork History
- The DAO Hack (2016): $60M stolen via smart contract bug. Ethereum hard-forked to reverse the hack. Those who refused the reversal continued as Ethereum Classic (ETC).
- Planned upgrades: Unlike Bitcoin, Ethereum uses hard forks routinely for upgrades (Byzantium, Constantinople, Istanbul, London, Shanghai...). The community expects them.
- EIP process: Similar to BIPs, but Ethereum's core developers coordinate via regular All Core Devs calls. Faster iteration, but more centralised decision-making.
Beyond Proof of Work: Proof of Stake
Proof of Work secures Bitcoin, but it has a well-known cost: massive energy consumption. Bitcoin mining uses roughly as much electricity as a small country. Proof of Stake (PoS) achieves consensus without mining — replacing computational work with economic collateral.
The Core Idea
In PoW, you prove honesty by spending electricity. In PoS, you prove honesty by locking up money. If you cheat, you lose your stake. The economic incentive is the same: make attacking the network more expensive than cooperating.
| Aspect | Proof of Work | Proof of Stake |
|---|---|---|
| Resource spent | Electricity (ASICs) | Capital (locked tokens) |
| Energy use | ~150 TWh/year (Bitcoin) | ~0.01 TWh/year (Ethereum) |
| Finality | Probabilistic (~60 min) | Deterministic (~13 min) |
| Attack cost | 51% of hashrate | 33% of total stake |
| Penalty for cheating | Wasted electricity | Stake destroyed (slashing) |
| Hardware needed | ASICs ($10K+) | Consumer laptop |
PoS Criticisms
PoS isn't perfect. Critics raise: (1) "Rich get richer" — those with more stake earn more rewards. (2) Nothing-at-stake problem — without slashing, validators could vote on every fork for free. (3) Long-range attacks — an attacker who once had a large stake could rewrite old history. Modern PoS protocols (like Ethereum's Casper) address these with slashing conditions and checkpoint finality, but the debate continues — especially among Bitcoin maximalists.
The Ethereum Merge
On 15 September 2022, Ethereum completed the largest protocol upgrade in blockchain history. The "Merge" transitioned Ethereum from Proof of Work to Proof of Stake — without stopping the network for a single second. Imagine changing the engine of a plane mid-flight.
Phase 1: Beacon Chain Launch
Dec 2020A separate PoS chain (the "Beacon Chain") launched alongside the existing PoW chain. Validators began staking ETH. The two chains ran in parallel for nearly 2 years.
Phase 2: Shadow Forks & Testing
2021 – Sep 2022Developers tested the Merge on multiple testnets (Ropsten, Goerli, Sepolia) and "shadow forks" of mainnet. Each rehearsal caught new edge cases. Meticulous engineering.
Phase 3: The Merge
15 Sep 2022At a specific Total Terminal Difficulty (not a block number), the PoW execution layer detached from mining and connected to the Beacon Chain's consensus. The last PoW block was mined, and the next block was validated by PoS. Zero downtime.
Phase 4: Shanghai/Capella
Apr 2023Staking withdrawals enabled. Validators who staked ETH since Dec 2020 could finally unstake. Yet most chose to stay — validating the economic model.
The Key Misconceptions
The Merge did NOT reduce gas fees. Gas fees depend on demand, not consensus mechanism. The Merge did NOT increase transaction speed. Block times went from ~13s (PoW) to exactly 12s (PoS) — almost the same. What it DID do: eliminate mining, reduce energy by 99.95%, make ETH slightly deflationary (issuance dropped ~90%), and set the stage for future scalability upgrades (sharding, danksharding).
The Modern Consensus Landscape
PoW and PoS are the two giants, but the blockchain world has invented many other consensus mechanisms — each optimised for different trade-offs.
Delegated PoS (DPoS)
Token holders vote for a small set of delegates (~21) who take turns producing blocks. Very fast (sub-second finality) but more centralised. Used by: EOS, Tron, Lisk.
PBFT / IBFT / QBFT
Classical Byzantine Fault Tolerance. All validators communicate in rounds to reach agreement. Instant finality. Doesn't scale beyond ~100 nodes. Used by: Hyperledger Besu (QBFT), Tendermint (Cosmos).
Proof of Authority (PoA)
Pre-approved, identity-verified validators. No staking, no mining. Maximum throughput. Zero decentralisation. Used by: Private networks, testnets.
Proof of History (PoH)
A verifiable delay function creates a cryptographic timestamp for every event. Combined with PoS for ordering. Used by: Solana (400ms blocks, 65K TPS theoretical).
Avalanche Consensus
Probabilistic consensus via repeated random subsampling. Each node polls a small random set of validators, converging quickly. Used by: Avalanche (sub-second finality, 4,500 TPS).
Raft / Paxos
Crash-fault tolerant (not Byzantine). One leader, followers replicate. Used where all nodes are trusted. Used by: Private Ethereum networks (Quorum), traditional distributed databases.
| Mechanism | Finality | TPS | Decentralisation | Energy |
|---|---|---|---|---|
| PoW (Bitcoin) | ~60 min | ~7 | High | Very high |
| PoS (Ethereum) | ~13 min | ~30 | High | Very low |
| DPoS (EOS) | ~2 sec | ~4,000 | Low (21 nodes) | Low |
| PBFT (Besu) | Instant | ~1,000 | Very low | Minimal |
| Avalanche | <1 sec | ~4,500 | Medium-High | Low |
The Blockchain Trilemma
No consensus mechanism optimises all three: decentralisation, security, and scalability. Bitcoin chose decentralisation + security (but low TPS). Solana chose security + scalability (but fewer validators). Enterprise chains chose scalability + security (but no decentralisation). Understanding these trade-offs is the key to choosing the right chain for your use case — and that's exactly what Part 2 and Part 3 will explore.
Exercise & Self-Check
Exercises
- Node comparison: You're building a Bitcoin wallet for a mobile app. Would you use a full node or an SPV client? List 3 reasons for your choice and 1 trade-off you'd accept.
- Fork analysis: Read the Wikipedia article on the Bitcoin Cash fork. Identify: (a) the specific rule change, (b) the block height where the split occurred, (c) which chain has higher hashrate today. Is BCH dead? Why or why not?
- PoS economics: If Ethereum has 900,000 validators each staking 32 ETH, what's the total value staked (at current price)? Now calculate: how much would an attacker need to acquire 33%? Compare this to the cost of a 51% attack on Bitcoin (hint: look up the cost per hour on crypto51.app).
- Consensus matching: For each scenario, pick the best consensus mechanism and explain why: (a) A global public payment network, (b) A hospital consortium sharing patient records, (c) A gaming platform needing 10,000 TPS, (d) A government-backed CBDC.
- The Merge deep-dive: Go to ultrasound.money and note: (a) current ETH issuance rate, (b) ETH burned vs issued since the Merge, (c) whether ETH supply is currently inflationary or deflationary. Write a 3-sentence summary.
Self-Check Questions
- What is an SPV node, and why does it need Merkle proofs?
- Explain the difference between a hard fork and a soft fork. Give one example of each.
- Why did the Bitcoin community refuse to simply increase the block size? What was the alternative solution?
- In Proof of Stake, what replaces the role of electricity in Proof of Work? What is slashing?
- Name three things the Ethereum Merge did NOT do (common misconceptions).
Module Summary
You've completed Part 1. You now understand: Bitcoin's full network architecture (full nodes, SPV, miners), how the protocol evolves through BIPs, the critical difference between hard and soft forks, why Proof of Stake exists and how it works, the historic Ethereum Merge from PoW to PoS, and the broader consensus landscape from DPoS to PBFT. Part 1 gave you the foundations. In Part 2, we cross the bridge from Bitcoin's limited scripting to Ethereum's fully programmable virtual machine — and you'll write your first smart contract.
Part 1 Complete!
You've finished all 6 modules of Part 1 — Blockchain Foundations. You can now explain how blockchains work from hash functions to global consensus. Part 2 begins with Module 6: we move from Bitcoin to Ethereum, from reading transactions to programming them.