Ethereum

Ethereum.

Ethereum is an open-source, public blockchain platform for executing smart contracts and decentralized applications. Its Ethereum Virtual Machine (EVM) processes transactions and contract code according to common rules. Ether (ETH) is the network's native cryptocurrency, used to pay transaction fees and participate in proof-of-stake consensus. Gas measures the resources consumed by execution; it is a unit of work, while fees are paid in ether.

Vitalik Buterin proposed Ethereum in 2013. A bitcoin-funded ether sale supported development in 2014, and the Frontier network launched on 30 July 2015. Following The DAO exploit, a disputed 2016 fork led to the separate Ethereum and Ethereum Classic chains.

Ethereum mainnet changed from proof of work to proof of stake on 15 September 2022 through The Merge. Subsequent upgrades enabled validator withdrawals and expanded data support for rollups, which execute transactions on a separate layer and settle results through Ethereum.

History

Origins and funding

Vitalik Buterin published an initial Ethereum white-paper draft in November 2013, after exploring ways to build applications on shared blockchain infrastructure rather than separate systems. In a January 2014 public introduction, Buterin named the initial core team as himself, Mihai Alisie, Anthony Di Iorio and Charles Hoskinson; the post also listed an expanded group that included Gavin Wood and Jeffrey Wilcke. These are the names in that dated announcement, not a complete account of every later contributor.

The ether sale opened on 22 July 2014 and ran for 42 days. Its published terms set an initial rate of 2,000 ETH per bitcoin for 14 days, followed by a declining rate down to 1,337 ETH per bitcoin. A contemporaneous Foundation update reported that, during the first two weeks, the sale had received more than 25,000 BTC in exchange for more than 50 million ETH. Ethereum.org’s historical account gives the completed sale as about 31,000 BTC, then valued at approximately US$18 million. The sale announcement also described two intended endowment pools for early contributors and pre-genesis expenses, and for a long-term reserve; these were stated allocation plans, not an audited account of eventual spending.

The Genesis Sale terms named the Swiss company Ethereum Switzerland GmbH (EthSuisse) as the seller of ether. A separate Swiss non-profit, the Ethereum Foundation, supported the project's development. Gavin Wood wrote the Ethereum Yellow Paper, its formal technical specification, alongside work on the software implementations.

Launch, the DAO and early forks

Before launch, the Olympic public test program invited participants to stress-test the protocol and offered a total prize pool of up to 25,000 ETH. The Foundation described tests of transaction load, virtual-machine behavior, mining and chain reorganizations; later it announced distribution of Olympic rewards. “Up to” describes the announced prize pool, not proof that the full amount was paid.

Frontier began on 30 July 2015 as a live but deliberately bare-bones release intended primarily for developers and other technical users. It let participants mine ether and deploy contracts, but the Foundation described the release as early software with substantial risk. The Homestead upgrade followed on 14 March 2016 at block 1,150,000; its changes included fixes to contract creation and the addition of the DELEGATECALL operation, among other protocol changes.

The DAO incident on 17 June 2016 arose from a recursive-call vulnerability in the DAO’s own contract: an attacker used repeated calls to move approximately 3.6 million ETH into a child DAO. The Foundation’s contemporaneous notice characterized the flaw as specific to the DAO, not a defect in Ethereum’s base protocol. The community debated a soft fork and then a hard fork; the proposed response explicitly said it would not reverse prior transactions or blocks. At block 1,920,000 on 20 July, the hard fork applied an irregular state change that moved about 12 million ETH from the Dark DAO and Whitehat DAO contracts to a recovery contract. This figure is the amount transferred by the state change, not the amount originally drained by the attacker. Some miners continued the pre-fork rules, and that chain became Ethereum Classic.

A separate series of denial-of-service attacks in September and October 2016 exploited operations whose gas costs were too low relative to their resource use. Tangerine Whistle repriced several operations to address immediate network-health issues. Spurious Dragon followed as the second stage of the response, including state cleanup and replay protection through EIP-155. These were protocol and client upgrades responding to network attacks; they were not a continuation of the DAO recovery action.

Berlin in April 2021 adjusted gas costs for state access and added typed transactions and optional access lists. London followed in August with EIP-1559: transactions pay a protocol-calculated base fee that is burned, and may include a priority fee for the block proposer. This changed the fee mechanism; it did not set a fixed transaction price. These upgrades continued the execution-layer evolution while the separate Beacon Chain was already operating.

The Beacon Chain began on 1 December 2020 as a proof-of-stake consensus network running alongside mainnet. Its launch did not immediately change how mainnet executed transactions: the existing execution chain continued under proof of work until The Merge on 15 September 2022. The Merge connected execution to Beacon Chain consensus, replacing mining with proof-of-stake consensus while preserving Ethereum’s existing execution history and state. Withdrawals were enabled later, through the coordinated Shanghai and Capella upgrades in April 2023.

Protocol upgrades

Ethereum’s protocol has changed through coordinated network upgrades. The table lists completed protocol activations; prospective upgrades are discussed under the roadmap. Early upgrades changed execution rules in one layer; after the Beacon Chain launched, major upgrades also coordinated execution-layer and consensus-layer changes.

Date (UTC)	Upgrade	Selected changes
30 July 2015	Frontier	First live, bare-bones network release for technical users; followed the Olympic test phase.
14 March 2016	Homestead	First planned major upgrade; included EIP-2, EIP-7 (DELEGATECALL) and EIP-8 networking changes.
20 July 2016	DAO fork	Recovery-contract state change at block 1,920,000; the community remained split over the response.
18 October and 22 November 2016	Tangerine Whistle; Spurious Dragon	Repriced undercosted operations after DoS attacks; then added further state cleanup and replay protection (EIP-155).
16 October 2017	Byzantium	Added the REVERT instruction and other EVM and cryptographic changes.
28 February 2019	Constantinople and Petersburg	Included CREATE2, EXTCODEHASH, gas-cost changes and a delayed difficulty bomb; Petersburg followed at the same block to remove an element of Constantinople.
8 December 2019	Istanbul	Changed gas costs and added EVM capabilities, including changes intended to improve compatibility with the Zcash transaction format.
1 December 2020	Beacon Chain genesis	The proof-of-stake consensus chain started alongside the then proof-of-work mainnet; it did not yet replace mainnet execution.
15 April 2021	Berlin	Optimized selected gas costs and introduced typed transaction envelopes and optional access lists.
5 August 2021	London	EIP-1559 introduced a base fee that is burned, alongside a transaction fee structure with an optional priority fee.
15 September 2022	The Merge	Mainnet execution joined the Beacon Chain consensus system, ending proof-of-work mining on Ethereum mainnet. Existing transaction history and execution state were preserved.
12 April 2023	Shanghai/Capella (Shapella)	Enabled validator withdrawals, including partial withdrawals of balances above the effective validator balance and full exits.
13 March 2024	Cancun/Deneb (Dencun)	EIP-4844 introduced blob-carrying transactions, a separate temporary data space aimed at rollup data availability.
7 May 2025	Prague/Electra (Pectra)	Included EIP-7702 for persistent code delegation by externally owned accounts until changed or cleared, EIP-7251 raising the maximum effective validator balance, and EIP-7691 increasing the target and maximum blobs per block from 3/6 to 6/9.
3 December 2025	Osaka/Fulu (Fusaka)	Activated PeerDAS, a data-availability sampling design for blob data, plus blob-parameter-only (BPO) forks for later blob-capacity increases.

Architecture

Accounts and the Ethereum Virtual Machine

Ethereum is a public blockchain whose replicated state machine is the Ethereum Virtual Machine (EVM). Nodes apply valid transactions in block order to a shared state containing account balances, contract code and contract storage; each accepted block advances that state according to protocol rules. Transactions can transfer ether, create contracts or call code already deployed at an address. The EVM gives those operations a common execution environment, while the network’s consensus rules determine which ordered state transition becomes canonical.

Ethereum accounts occupy one address space but have two forms. An externally owned account (EOA) is controlled by a private key and can originate signed transactions. A contract account has code and persistent storage, but no private key; it acts when a transaction or another contract call reaches it. Account data includes a nonce, balance, storage root and code hash. The nonce orders an EOA’s transactions and helps prevent replay of an already used signed instruction; a contract’s nonce also tracks contracts it creates. Contract code is compiled to EVM bytecode, and execution can update the account state.

A full Ethereum node normally combines an execution client and a consensus client. The execution client accepts and checks transactions, runs the EVM and maintains the latest execution state. The consensus client implements proof-of-stake rules, communicates attestations and block proposals, and chooses the chain head using the protocol’s fork-choice rules. A validator client can be added to the consensus software to perform validator duties. Independently developed clients follow the same protocol specifications, and a validating node coordinates execution, consensus and validator software.

Independently developed clients provide software diversity: a defect in one implementation need not affect nodes running another. Concentrating validator stake in one client can turn a software error into a network-wide consensus problem. Client diversity therefore concerns the distribution of stake across implementations, not simply the number of available programs.

Ether, transactions and gas

A transaction is an instruction authorized by a cryptographic signature, usually created with the sender’s account key. It identifies a destination or contract-creation action and may include ether value and input data. Nodes check the signature and transaction constraints, then execute the request under the EVM rules; a block proposer includes valid transactions and other nodes re-execute them to verify the resulting state. Contract accounts cannot originate ordinary Ethereum transactions by signing with a private key, although code in a contract can initiate calls during execution. Signing authorizes a request to change state, without guaranteeing inclusion or a particular outcome.

Execution work is priced in gas, a protocol-defined unit that meters operations and limits the computation a transaction can consume. A sender sets a gas limit and fee parameters. Under the EIP-1559 fee mechanism, the protocol computes a per-gas base fee that moves with block demand; the base-fee portion actually paid is burned, while a priority fee (tip) can go to the block proposer. The effective price cannot exceed the sender’s maximum fee, and unused gas-limit capacity is not charged as used computation. Ether amounts are counted in wei, with 10¹⁸ wei per ETH; gwei is 10⁹ wei and is a common way to express a gas price. The fee is based on gas used and the effective per-gas price, so it changes with both execution complexity and network conditions.

Ether issuance and supply

Ether (ETH) is Ethereum's native asset, used for transaction fees and staking. Since The Merge, new ETH issuance comes from proof-of-stake rewards; proof-of-work block rewards have ended. Net supply change depends on new issuance and destruction of existing ETH, including fee burning. Validator issuance varies with the amount staked, while fee burning varies with network usage. The resulting supply can therefore increase or decrease; the older forecasts of a fixed long-term inflation rate do not describe the present mechanism.

The historical white paper named additional ether denominations: one szabo is 10¹² wei and one finney is 10¹⁵ wei. It suggested finney for microtransactions and szabo or wei for technical discussions. These historical conventions coexist with the gwei unit commonly used for fee quotations.

Consensus

Proof of stake

Ethereum uses proof of stake for block production and consensus. Time is divided into 12-second slots grouped into epochs of 32 slots. A validator is selected to propose a block in a slot, while committees of validators attest to the chain; clients use these votes and a fork-choice procedure to determine the current head. A validator deposit of 32 ETH is the minimum activation balance. Since the Pectra/Electra changes, a validator using compounding withdrawal credentials can have an effective balance up to 2,048 ETH, while 32 ETH remains the activation minimum. The higher ceiling allows excess balance and rewards to compound or permits consolidation of validator balances; it does not mean every validator begins with or must hold that maximum.

Validators are expected to propose blocks when selected and attest to valid blocks. Missed duties can reduce expected rewards or incur protocol penalties, while specified conflicting block proposals or attestations are slashable: the validator loses stake and is forced through an exit process. Withdrawals are governed by validator withdrawal credentials and protocol queues, rather than by an ordinary transfer from the consensus layer. Pectra introduced a compounding credential format that can retain rewards above 32 ETH up to the effective-balance ceiling; 0x01 credentials automatically sweep balances above 32 ETH to a designated execution-layer address. Validators retaining 0x00 credentials must upgrade their credentials before withdrawals are possible. Staking remains subject to these duties, penalties and withdrawal rules, and is not risk-free.

Energy consumption

Ending proof-of-work mining greatly reduced the electricity used to operate Ethereum. In a 2022 study, the Crypto Carbon Ratings Institute estimated that The Merge reduced annualized network electricity consumption by more than 99.988%. Its estimate used measurements of node hardware and software configurations and assumptions about the network before and after the transition. This is a dated study estimate rather than a fixed measurement of present-day consumption; electricity use and carbon emissions also depend on network size and the energy sources used.

Historical mining

Before The Merge, Ethereum used the memory-hard Ethash proof-of-work algorithm. Its growing dataset was intended to make specialized mining hardware less advantageous, although Ethash ASICs were eventually developed and GPUs also remained in use. Mining no longer produces valid blocks on Ethereum mainnet. Ethash remains relevant to the history of Ethereum and to other networks that retained proof of work.

Smart contracts and token standards

Developers usually write smart contracts in a higher-level language and compile them to EVM bytecode before deployment. Solidity is a contract-oriented language with syntax influenced by C-family languages. Vyper is a Pythonic EVM-targeting language whose documentation emphasizes simplicity and auditability. Earlier Ethereum development also included languages such as Serpent, LLL and Mutan; surviving project repositories document them as historical EVM-compilation tools, not as evidence that they remain current choices for new work. Whatever the source language, deployed code is executed as EVM instructions, so language syntax does not change the underlying consensus rules.

ERCs specify interoperable contract interfaces, rather than creating a single system-wide token ledger. ERC-20 defines a common interface for fungible tokens, where units of a given token are interchangeable. ERC-721 defines ownership and transfer operations for individually distinguishable, non-fungible tokens identified within a contract. ERC-1155 allows one contract to manage multiple token types and supports batch operations, reducing the need to deploy a separate contract for each type. A token’s behavior still depends on its particular contract implementation; conformance to an interface alone does not establish the quality, backing or legal status of an asset.

Account abstraction

Account abstraction allows account software to define more of the rules for authorizing and paying for operations. ERC-4337 specifies a higher-layer system in which users submit UserOperation objects to a separate mempool. Bundlers include these operations in transactions sent to an EntryPoint contract, which coordinates validation and execution; paymasters can sponsor fees. This design does not require a consensus-layer protocol change.

EIP-7702, activated with Pectra, instead introduced a protocol transaction type through which an externally owned account can authorize delegation to code at another address. Delegated code runs in the account's context, while the account can still originate transactions. This delegation mechanism is distinct from ERC-4337's bundler flow, although the two can be used together. Delegation persists until the account changes or clears it.

Scaling and roadmap

Rollups and data availability

Ethereum’s scaling approach has shifted toward a rollup-centric roadmap rather than the earlier plan to divide ordinary transaction execution across execution shards. Rollups execute batches of transactions on a separate layer and publish data or proofs to Ethereum for settlement and verification. This reduces the amount of per-transaction execution performed directly on layer 1, while Ethereum continues to provide the base protocol, consensus and data-availability services used by rollups. Rollups differ in how they prove state updates and what assumptions they make; they are not simply faster copies of the mainnet.

Optimistic rollups generally treat submitted state updates as valid unless a challenger proves an error during a challenge period. Zero-knowledge rollups use a validity proof that the rollup’s contract checks before accepting a state update. Despite the name, a zero-knowledge proof used for validity does not necessarily conceal transaction data: rollups must provide the data needed for their users or independent parties to reconstruct state, and publication methods vary. Optimistic designs depend on their challenge process and its actors; validity-proof designs depend on their proof system and verifier. Withdrawal timing and finality therefore depend on the rollup’s specific design, not solely on Ethereum’s slot timing.

The Dencun upgrade activated EIP-4844 blob-carrying transactions in March 2024. Blobs provide a separate, temporary data-availability space that rollups can use instead of placing all batch data in ordinary execution calldata. Ethereum’s protocol guarantees blob availability for a limited period—about 18 days—after which nodes may prune the data; blobs are not permanent archival storage. Fusaka went live on 3 December 2025 and introduced PeerDAS, which lets nodes sample data-availability columns instead of requiring every node to download every blob. PeerDAS expands the design’s capacity for blob data, but activation alone did not automatically raise blob limits; subsequent parameter changes and rollup adoption determine how much capacity is used.

Layer-2 systems have different operational and security arrangements. A rollup may depend on a sequencer for transaction ordering, upgrade keys or a security council, and the maturity of its proof and escape mechanisms varies by project. Posting data to Ethereum can support independent reconstruction, but users still need to understand the particular system’s contracts, data-availability design, proof assumptions and exit conditions. Performance figures such as transactions per second cannot be generalized from the Ethereum protocol alone: observed throughput depends on a layer’s workload, batching, data capacity, fees and implementation. Roadmap capacity goals are targets, not guaranteed service levels.

Earlier scaling research

Joseph Poon and Vitalik Buterin's August 2017 Plasma working draft proposed child chains anchored to a parent blockchain through commitments and fraud proofs. It addressed how an application could move computation away from the parent chain while using that chain to resolve disputes, including problems of unavailable data and exits. Plasma is part of the history of Ethereum scaling research and differs from rollups that publish transaction data to the base layer.

Glamsterdam

The Ethereum Foundation's September 2026 Glamsterdam announcement described enshrined proposer-builder separation (ePBS) and block-level access lists among the planned changes. It scheduled the Sepolia testnet upgrade for 6 October 2026 at 13:53:36 UTC; Hoodi and mainnet dates were still to be determined in that announcement. An announced testnet schedule does not establish completed mainnet deployment.

Governance and standards

Ethereum's protocol governance takes place through off-chain discussion and coordination among users, application developers, node operators, validators and client developers. It differs from on-chain voting used by some applications and decentralized autonomous organizations. A proposal does not automatically become a protocol rule merely because it has an EIP number.

Ethereum Improvement Proposals (EIPs) document proposed specifications and their rationale. Core proposals concern changes requiring a network upgrade, while other categories cover areas such as networking and application interfaces. Implementation, testing, coordination and activation are separate from documenting a proposal. Operators adopt protocol changes by running compatible client software; disagreements over rules can lead to a chain split, as in the DAO fork.

Applications and ecosystem

Ethereum applications use contracts for functions such as exchanging tokens, lending, payments and asset issuance. Decentralized finance (DeFi) is a collective name for financial applications whose rules and positions are handled through smart contracts. Different applications can call one another, but their governance, collateral and operational arrangements differ. Stablecoins are tokens designed to track a reference value, often a national currency. Their designs include issuer-held reserves and cryptoasset collateral; the intended price relationship is distinct from an unconditional redemption or value guarantee.

Non-fungible tokens can identify particular digital items, including collectibles, artwork and game assets. The token record and the rights to an associated item are separate questions. Tokenization also encompasses securities and claims on off-chain assets. The Bank for International Settlements distinguishes digital representations of existing assets from assets issued directly on programmable platforms, and discusses custody, legal and operational arrangements that remain necessary. A token transfer alone does not establish every off-chain ownership right or remove the need for intermediaries.

Contracts cannot directly fetch a changing external website or independently observe physical events. Oracle systems provide off-chain information to contracts, for example a price or the outcome of an event. This extends the range of applications, while introducing dependencies on how the information is collected, transmitted and checked. Ethereum consensus on a transaction does not establish that its externally supplied input was accurate.

Naming services

The Ethereum Name Service (ENS) provides a naming system with separate registry, resolver and registrar roles. The registry records a name's owner, resolver and time-to-live value; a resolver returns records such as cryptocurrency addresses or content hashes. Registrars determine how names are allocated. This architecture is more extensive than the simplified name-registry example in the original Ethereum white paper.

Historical applications and prototypes

Early projects explored uses beyond simple ether transfers. Their proposals, prototypes and later product launches need to be distinguished:

• Prediction markets: the authors of Augur described a decentralized oracle and prediction-market platform in a 2015 paper. Its design used participants to report event outcomes, making the reporting mechanism part of the application rather than a fact supplied by Ethereum itself.

Gnosis also describes its early work in decentralized prediction markets in its retrospective history.

• Gold tokenization: Digix's January 2016 white paper proposed Ethereum contracts and chain-of-custody records for physical assets, using gold as its principal example. Digix subsequently announced the April 2018 launch of DGX, described by the issuer as representing one gram of gold. These descriptions document the project's design and launch claims, rather than proving continuing custody or redemption arrangements.
• Music and royalties: Ujo and Imogen Heap demonstrated selling the song Tiny Human using Ethereum contracts and automated royalty distribution. Ujo's retrospective describes it as an alpha prototype, retired in November 2016.
• Energy and connected devices: Slock.it and RWE explored contract-based electric-vehicle charging in 2016. A 2017 project account described the Share&Charge contracts used on public Ethereum for charging services. These were particular projects, rather than evidence that Ethereum itself operates a charging network.
• Identity: Zug's 2017 digital-identity pilot used uPort and public Ethereum. It was a municipal pilot; a later city response states that Zug replaced this system with eZug. The pilot should therefore not be described as a current citywide Ethereum identity service.
• Employment: Ethlance's January 2017 launch announcement described a job market with its database on Ethereum and its frontend files on IPFS, including job applications and invoicing. Blocklancer's 2017 development diary described a proposed distributed job market. The announcements establish historical examples, not present-day operation or adoption.
• Document verification: Luxembourg's 2018 public-sector blockchain white paper described the Luxembourg Stock Exchange's Ethereum-based application for document authenticity and proof of existence.

• Games: Etheroll's February 2017 draft white paper described an Ethereum dice-betting application using ether. Its proposed randomness mechanism obtained data from Random.org through Oraclize, illustrating how a contract-based game can depend on an external oracle. The document records the project's design claims; it does not establish current operation or guarantee fairness.
• Options and governance: Etheropt's source repository describes an Ethereum options exchange using external ETH/USD price reports. Backfeed proposed decentralized cooperation and governance through contribution-based evaluation. These are different application designs, with their own rules and external dependencies.
• Ethereum and IPFS experiments: AKASHA's archived alpha repository documents a pre-release application using Ethereum and IPFS. Its README identifies the alpha as outdated and its AETH currency as a private-chain test token.
• Tokenized securities: Santander announced in September 2019 that it had issued a US$20 million bond using public Ethereum, with the bond and investment cash represented through tokens. The announcement records the issuer's dated account of this specific transaction.

The original Ethereum white paper also illustrated a small Namecoin-like name-registration contract. This was a simplified programming example, not a claim that a complete naming protocol, with all its ownership and operational rules, could be replaced by a fixed five-line program.

Wallets

Wallet software provides an interface for accounts, balances, signing and interaction with applications. A wallet is distinct from an account: the account and its assets are represented in Ethereum's state, while the software manages access and requests. Wallets can take the form of browser extensions, mobile or desktop applications, hardware devices, or contract-based accounts.

Examples include MyEtherWallet software and hardware products from Trezor, Ledger and KeepKey, whose providers document Ethereum support. Their listed integrations and features are product-specific. The older Jaxx example needs historical context: its developer announced that normal Jaxx Liberty operations would cease on 27 March 2023.

Enterprise software and permissioned networks

Public Ethereum is permissionless, while a permissioned ledger limits participation to selected organizations. An enterprise system may reuse Ethereum technology without using the public mainnet, and its access controls, governance and consensus arrangements can differ. NIST's blockchain overview describes this distinction between permissioned and permissionless systems.

Microsoft and ConsenSys announced an Azure-hosted Ethereum development environment in November 2015. The offering combined Ether.Camp tools with BlockApps environments for private or consortium networks and deployment to public Ethereum. This was an infrastructure offering, rather than evidence that every participating enterprise used Ethereum mainnet.

The Enterprise Ethereum Alliance (EEA) launched on 28 February 2017. Its announcement named organizations including Microsoft, J.P. Morgan, Intel and ConsenSys, alongside technology companies, start-ups and research groups. In May 2017, the alliance announced more than 86 new members; its July announcement reported total membership above 150. These are dated membership milestones, rather than counts of current members.

J.P. Morgan developed Quorum as an enterprise variant of Ethereum, launched in 2016. ConsenSys announced its acquisition of Quorum in August 2020. The announcement described Quorum's enterprise applications and its relationship to Ethereum, rather than a replacement for the public network.

In July 2017, S7 Airlines announced a ticket-agent settlement platform developed with Alfa-Bank. The issuer's release described a private blockchain based on the Ethereum protocol, with contracts automating payment requests, checks of available funds and transaction-status updates while the bank processed payments. This was a specific private-network workflow, rather than a public-mainnet deployment.

Royal Bank of Scotland's Innovation Engineering team also described an Ethereum-based Clearing and Settlement Mechanism in the September 2016 Emerald technical paper. It explored a private ledger and reported tests using six simulated banks. These were research and prototype results, not measurements of public Ethereum throughput or evidence of a production banking rollout.

Security and research

Public contract execution makes code and transactions inspectable, but inspectability does not establish that a contract is correct. Coding errors, permissions and interactions with other contracts can expose assets to loss. Deployed contract code is generally difficult to change; upgrade mechanisms may permit changes through designated contracts or authorities, introducing their own governance and access-control considerations. The DAO exploit is a historical example of a contract-level vulnerability with consequences beyond the application itself.

Formal verification checks whether a program satisfies a specified set of properties under a model and its assumptions. This differs from establishing that every aspect of an application is safe or that all external inputs are correct. A 2016 paper by Bhargavan and colleagues, including researchers from Microsoft Research, outlined analysis of Solidity and EVM contracts through translation into F*. A separate paper by Luu and colleagues introduced Oyente, a symbolic-execution tool for finding potential vulnerabilities. These papers document early research methods, without certifying arbitrary deployed contracts.

Academic research has also examined fraud implemented through contracts. Bartoletti and colleagues' study, first submitted in 2017 and revised in 2019, identified 138 Ponzi schemes among contracts with available source code and expanded its dataset to 184 by examining similar bytecode. The dataset describes a particular historical collection and does not measure the share of all Ethereum contracts or current transactions. The study distinguished execution of programmed payout rules from the legitimacy or sustainability of the financial scheme.

See also

• Ethereum history

• Altcoins
• Waves Platform
• Monero
• Bitcoin Cash
• Dogecoin
• Zcash
• Bitpay
• Mining
• ICO (Initial Coin Offering)

External links

• Ethereum community website
• Ethereum GitHub organization
• Ethereum YouTube channel
• Ethereum Reddit community
• Ethereum: the World Computer (historical video)
• Introducing ConsenSys Academy: Why We Code on Ethereum (historical video)