Skip to main content

Smart Contracts

Which platform makes it hardest to write a dangerous smart contract?

DX is the northstar for crypto adoption. Writing safe smart contracts is beyond most developers — too costly, too time-consuming, too dangerous. The chain that engineers risk out at the platform level wins.

Where Work Lives

Smart contract writing is 20–30% of a production Web3 project. The contract is kept minimal by design — smaller surface, lower gas, fewer attack vectors. The heavy lifting migrates off-chain.

LayerShareTools
Smart contracts~20–30%Foundry, Anchor, Sui Move
Frontend / dApp UI~25–30%React, Next.js, Wagmi, Viem
Indexing & data~15–20%The Graph, custom subgraphs
Backend & APIs~15–20%Node.js, event listeners
DevOps & infra~10–15%RPC nodes, L2, rollup infra

Off-chain computation cuts gas costs 90–95% versus equivalent on-chain execution. The contract becomes a verification layer — it checks proofs, not does math. DeFi protocols compute complex pricing off-chain and submit only final values on-chain. The work is building trusted pipelines that feed the contract, not the contract itself.

Three layers, distinct failure modes:

LayerRoleWhere It Fails
ContractsTruth anchor — minimal, auditable, immutableLogic errors are permanent
Bridgeethers.js, Viem, event listeners, oraclesMost integration bugs live here
Off-chainAgents, indexers, APIsWhere value compounds, where scale breaks

The chain comparison below covers where you spend the 25%. The platform choice shapes the 75% too — composability primitives, indexing tooling, and wallet integration differ significantly across EVM, SVM, and Move.

Hyperlinks transformed information flow — anyone could connect any document to any other document. Smart contracts do the same for value. A hyperlink says "go here." A smart contract says "if this condition is met, value flows here, measured by these instruments, verified on-chain."

InventionWhat It PipesTransformation
HyperlinkInformationAnyone can publish → web grows → network effects
Smart contractValueAnyone can pipe value → mycelium grows → network effects

Smart contracts are the piping in Agent & Instrument Diagrams. They sit at the Protocols layer of the knowledge stack — sequencing primitives into value streams. Blinks make this tangible: a URL that carries a transaction wherever a link can go.

Quick Comparison

DimensionEVM (Solidity)SVM (Rust/Anchor)Move (Sui)
SafetyOpt-in (libraries, audits)Partial (runtime guards)Structural (compiler-enforced)
OnboardingDays (JS-like, Remix)Months (Rust learning curve)Weeks (novel but purpose-built)
Speed12s blocks, Foundry v1.0400ms blocks, slow deploys390ms finality, fast local tests
ComposabilityFlexible but dangerousCPI (depth limit 4)PTBs (1,024 ops) + full stack
Ecosystem31K devs, deepest tooling17K devs, fastest growing1.4K devs, completing full stack

Safety

The pit of success question: does the platform prevent vulnerabilities, or does the developer prevent them?

VulnerabilityEVM (Solidity)SVM (Rust/Anchor)Move (Sui)
Re-entrancyManual prevention (OpenZeppelin guard)Runtime restricts CPI depth to 4Impossible by design (no recursive calls)
Asset duplicationManual balance managementRuntime prevents same-slot double-writeLinear types: move only, never copy
Integer overflowChecked since Solidity 0.8Panics in debug, wraps in releaseAlways checked at bytecode verifier level
Access controlModifier-based (convention)Account ownership + manual signer checksCapability-based (compiler-enforced)
Formal verificationCertora, Halmos, Echidna, SlitherTrident (early)Move Prover + Sui Prover (open-sourced 2025)

Re-entrancy alone caused $325M in stolen assets in 2025. Move eliminates the class entirely. 5 of the OWASP Smart Contract Top 10 are structurally impossible in Move.

Onboarding

How fast from zero to deployed contract?

DimensionEVM (Solidity)SVM (Rust/Anchor)Move (Sui)
Language curveDays (JS-like syntax)Months (ownership, borrowing)Weeks (novel but purpose-built)
Concepts before deploy~5-6~8-10~6-7
Zero-install pathRemix (browser IDE)Solana PlaygroundCLI required
IDE supportMature (Hardhat LSP, Foundry)Excellent (rust-analyzer)Growing (Move LSP, trace debugger)
DocumentationMassive corpusStrong, improvingSmallest, growing fast
Test frameworkFoundry v1.0, HardhatLiteSVM + Anchorsui move test (built-in)
Tests in contract languageYes (Foundry)No (TypeScript/Rust separate)Yes (Move #[test])

Iteration Speed

MetricEVMSVMMove (Sui)
Testnet block time~12 seconds (Sepolia)~400ms (Devnet)Sub-second (Testnet)
Mainnet finality~15 minutes~13 seconds economic~390ms (Mysticeti)
Local test speedFoundry: 2-5x faster than HardhatLiteSVM: in-process SVMFast compilation, no validator needed
FuzzingFoundry built-in + Echidna (production)Trident (early)Move Prover + Belobog

Sub-second finality changes what you can build. Real-time agent workflows, on-chain IoT, HFT-style DeFi — these require ~400ms or less.

Composability

Can agents compose transactions without middleware?

PrimitiveEVMSVMMove (Sui)
Transaction batchingMulticall + EIP-7702 (2025)Instructions in one tx, CPI depth 4PTBs: 1,024 ops with data flow between
Native identityNone (ERC-4337 account abstraction)None (SPL Name Service)zkLogin (protocol-level, 7.6M txs)
Off-chain storageIPFS, Arweave (fragmented)Shadow Drive, ArweaveWalrus (same-stack, $140M raised)
Programmable accessSeal (identity-based encryption)
Verifiable computeNautilus (TEE attestations on-chain)

Sui's extended stack (Walrus + Seal + Nautilus) means identity, storage, encryption, and verifiable compute ship as one platform — not glue code across five vendors.

Ecosystem

MetricEVMSVMMove (Sui)
Monthly active devs~31,900~17,700~1,400
New devs (2024-25)16,181 added11,534 added (+83% YoY)219% growth (H1 2024)
Audit costBaseline+25-40% premium+30-45% premium
Production DeFiUniswap, Aave, CompoundRaydium, Jupiter, OrcaDeepBook, Cetus, Navi
Token standardsERC-20/721/1155 (mature)SPL Token, Token-2022Sui Coin/NFT (standardizing)

Smallest ecosystem, fastest percentage growth, highest developer commitment (>50% single-chain). The thesis: DX quality predicts developer migration. Developer activity is the leading indicator.

Standards War

Who controls what "standard" means?

DimensionEVMSVMMove (Sui)
GovernanceEIP/ERC — community-driven, anyone submitsSIMD — foundation-managed, 66.67% validator voteSIP — Mysten + community, fast-track option
Token philosophyStandard interface, any implementationShared program library, foundation-maintainedLanguage-embedded (compiler IS the standard)
FungibleERC-20 (each token deploys own contract)SPL Token (single shared program)sui::coin (typed object)
Non-fungibleERC-721 (interface standard)Metaplex NFT standardAny object with key ability
ExtensionsERC-1155, ERC-4337 (AA), ERC-3643 (RWA)Token-2022 (extensions on shared program)Composable objects + PTBs
MaturityThousands of EIPs, battle-testedFoundation-curated, growingSmallest — language does the work

EVM standardizes the INTERFACE — define the shape, let anyone implement. Solana standardizes the PROGRAM — one shared implementation, everyone calls it. Sui standardizes at the LANGUAGE level — there's no separate NFT standard because every object with key is already uniquely identified, owned, and transferable. Three philosophies. The question is whether making standards unnecessary beats having mature ones.

Standard Fittings

Language-level safety handles objects and ownership. It does not handle access choreography or overflow-safe arithmetic. OpenZeppelin Contracts for Sui proves the point — even Move needs a standard fittings library.

In A&ID nomenclature: standard fittings are instruments. In P&ID: ANSI-rated valve assemblies and certified pressure relief devices. You specify them by code. You do not build your own.

FittingP&ID EquivalentWhat It Does
two_step_transferTwo-key safety interlockOwnership handoff requires initiate + accept. Hot-potato guards enforce return.
delayed_transferTime-delay relayTime-locked wrapper. Schedule, wait, execute. Minimum delay enforced.
mul_divFlow computer(a * b) / c without intermediate overflow. DeFi's core calculation.
decimal_scalingUnit converterSafe precision conversion between token decimal scales.
u512Extended-range gauge512-bit intermediate math. Prevents precision loss in tokenomics.

Lindy Effect

The Lindy Effect applies to standard fittings: the longer a pattern survives without exploit, the more trust it accumulates. OpenZeppelin's EVM contracts are Lindy — battle-tested since 2016, hardened by a decade of audits and attacks. When they port patterns to Move, the implementation is new but the design is not. Trust transfers across platforms when the pattern is proven.

Trust LayerWhatLindy Age
Language (Move)Linear types, capability-based accessYoung, structurally sound
Standard Fittings (OpenZeppelin)Access patterns, safe mathPatterns Lindy from EVM, implementation new
Applicationidentity, mandate, tokenomicsInherits both layers

Each layer's survival compounds into the application's trustworthiness. Standards that get safer with age are the opposite of Wirth's Law — where software gets slower faster than hardware gets faster. Standard fittings are the exception. Time is their ally.

Depth: ERC Standards | SPL Token | Sui Development

Learning Path

Where do you start, and what's the sequence?

StageEVM (Solidity)SVM (Rust/Anchor)Move (Sui)
SetupNode + FoundryRust + Solana CLI + AnchorRust + Sui CLI
Browser IDERemix (zero install)Solana Playground— (CLI only)
First contractforge initforge testanchor initanchor testsui move newsui move test
Mental modelGlobal state + msg.senderAccounts + PDAs + CPIsObjects + ownership + capabilities
TokensERC-20 / ERC-721 / ERC-1155SPL Token / Token-2022sui::coin / objects with key
TestingFoundry (same language)LiteSVM (TypeScript)sui move test (same language)
Frontendethers.js/viem + wagmi@solana/web3.js + wallet-adapter@mysten/sui + dApp Kit + zkLogin
AdvancedUpgrades, AA, DAO, MerkleNFTs, DePIN, AI agentsPTBs, Walrus, Seal
AuditingSlither + Echidna + CertoraTrident (early)Move Prover (built-in)
Deployforge script --broadcastanchor deploysui client publish
Best courseCyfrin UpdraftBootcamp 2024Move Book
Time to productiveWeeksMonthsWeeks

Depth: EVM Development | Solana Development | Sui Development

Dig Deeper

Context

  • Smart Contract Standards — Standard fittings catalog: Lindy libraries, safety comparison, DeFi primitives
  • Essential Algorithm — Smart contracts bridge ROUTE → SETTLE: algorithms that enforce protocols on-chain
  • Protocols — Algorithms decide the route; protocols enable the handshake; smart contracts execute both
  • Matrices — Disruption scoring: actuator potential measures how directly predictions trigger on-chain action
  • Knowledge Stack — Primitives → Protocols → Standards → Platform: where smart contracts sit
  • Value Stream Map — The factory blueprint that smart contract piping makes executable
  • A&ID Template — Pipe = smart contract in the Agent & Instrument nomenclature
  • Developer Experience — DX scoring framework and crypto comparison methodology
  • Sui Technical — Object model, Move language, agent economy primitives
  • Crypto Problems — What keeps going wrong and why
  • Blockchain Decisions — Which chain for which use case
  • Agent Commerce — The standards war for agent transactions
  • Hacker Laws — Lindy Effect, Gall's Law, Wirth's Law — the laws that govern standard fittings
  • Pipe Dreams — Smart contracts as intelligent hyperlinks: the piping that makes pipe dreams buildable

Questions

Does the platform that makes standards unnecessary win against mature community-driven processes?

  • How do you weigh onboarding speed against safety depth when choosing a first chain?
  • What breaks when language-embedded standards need to evolve faster than the language itself?
  • If 95% of smart contract operations become commodity, does the standards layer commoditize too?
  • Which governance model — community EIP, foundation SIMD, or language SIP — produces better standards faster?
  • If standard fittings get safer with age (Lindy), does the chain with the oldest audited library always win on trust — even if a younger chain is structurally safer?