Learn Solana through its history
Every simulation below reconstructs a real, source-cited event — and makes you operate the mechanism at its center. You don't watch the Wormhole exploit; you forge the receipt. You don't read about Proof of History; you tamper with a real SHA-256 chain and watch it break. Four tracks, sixteen playable teardowns, free and open source (CC-BY-4.0).
TRACK 1 Consensus, time & liveness
How a blockchain agrees on time and order — and what actually happens when block production stops.
- Why sequential SHA-256 hashing can act as a clock
- Tamper with one event and watch every later hash cascade-break
- Why "PoH as a strict VDF" is academically debated
- How validators coordinate a cluster restart from a snapshot
- Liveness vs safety: why an outage is not a hack and no funds moved
- Why restarts need a supermajority to agree on the restart slot
- Why one bug in one client can halt a single-client network
- Stake weight, not client count, decides liveness (~⅓ / ⅔ thresholds)
- Kill a client at different stake shares and see when the network survives
TRACK 2 Exploits, security & systemic risk
Four real incidents, four different failure modes: a forged verification, a manipulated price feed, a leaked secret, and a contagion that traveled the dependency graph — not the chain.
- Be the clerk: forge the receipt, not the signature
load_instruction_atvsload_instruction_at_checked— one deprecated call- Counterfactual: the checked reader rejects the same forgery
- Pump a thin market until fake PnL outweighs a real treasury
- Zero contract bugs used — the design was the vulnerability
- Why the legal aftermath (2025 vacated convictions) matters
- How plaintext logging turns an app into a key custodian
- Flip the scrub-sensitive-data flag and re-run the drain
- Forensics honesty: only ~15% of wallets were conclusively traced
- Audit each dependency on FTX — custody, liquidity, keys, or just narrative
- Detonate and watch contagion travel the graph while the chain keeps finalizing
- Fork the one hard break (Serum) into OpenBook under community keys
TRACK 3 Markets & liquidity
The market microstructure layer: how prices form, how orders route, and how launches bootstrap liquidity.
- Drag a real x·y=k curve and feel slippage vs depth
- Why the same swap costs 37% in a shallow pool and 6% in a deep one
- Illustrative model, labeled as such — the mechanism is exact
- Hand-route a swap across three venues, then lose to the optimizer
- Convex price impact: why the best split shifts with order size
- Why aggregation became Solana's invisible default
- Tap through the four frictions token launches used to require
- DEPLOY: watch launch cost collapse to near zero
- What friction-free launching did to token supply — documented as culture, not investment
- Drag BONK's price until the bundled airdrop beats the $599 phone
- Run the arbitrage loop yourself and clear the shelves
- Why demand switches species: product demand vs arbitrage demand
TRACK 4 Scale, load & distribution
What happens when a million people show up at once — congestion economics, RPC load, mass distribution, and community coordination.
- Spam harder, watch bots match you and your share snap back
- Tragedy of the commons in fee markets, played not told
- Congestion vs outage: the chain kept producing blocks
- Hold to claim while the whole crowd piles onto the same RPCs
- Why demand can look like a denial-of-service
- Pause, fix, ship: the one-day delay that saved the mint
- Why a real wireless network settles proofs off-chain
- Flip to "settle everything on-chain" and watch load explode
- Oracles as compression, not as price feeds
- Run a million-wallet distribution as a dry-run rehearsal
- Why WEN was the test flight for the JUP airdrop
- Claim windows, eligibility snapshots, and load shaping
- Act I: choose BONK's allocation — insiders or the ecosystem
- Act II: crowdfund the $690k WIF Vegas Sphere, then EXECUTE
- Why a crowd aggregates money easily but can't sign a contract
For educators: everything here is free to reuse under CC-BY-4.0 — simulations, landmark texts, and the open dataset. Every claim is cited (73 references, weighted toward primary sources) and every landmark carries a "sources last verified" date. Found an error? The repository has a factual-correction template — corrections with a primary source get priority.