Industries / Technology / Crypto and blockchain
Ledgers and agreement
This is the work of keeping one shared record of who owns what, copied on many computers that agree on each new entry without a central keeper.
6
Milestones achieved, from 1991 to 2023.
3
To watch for: announced, not yet confirmed outside.
3
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1991
Records chained by hash
Stuart Haber and W. Scott Stornetta described a time-stamping service that hashes each new batch of documents together with the previous result, so that changing an old record would break every link after it. From 1995 the company Surety printed its latest value in the New York Times; the researchers who write Decentralized Thoughts call it "the longest-running chain of hashes" and note that the Bitcoin paper cites the work.
Credit
Stuart Haber and W. Scott Stornetta; Surety
2009
A ledger with no keeper
On 31 October 2008 Satoshi Nakamoto sent the Bitcoin paper to a cryptography mailing list, describing "a new electronic cash system that's fully peer-to-peer, with no trusted third party". Computers race to solve a puzzle, called proof of work, to add each block, and the longest chain counts as the record. The first block is dated 3 January 2009 and the software was released on 8 January.
Credit
Satoshi Nakamoto
2013
Two versions, two chains
In March 2013 a large block that Bitcoin version 0.8 accepted was rejected by older versions, because of a database lock limit that nobody had known was in effect a rule, and the ledger split with about 60% of mining power on the new side. The mining pools BTCGuild and Slush downgraded to the old version to rejoin the chains, "despite the fact that this caused them to sacrifice significant amounts of money", and at least one large double spend got through.
LessonEvery copy of a shared ledger must give the same answer, including on limits no one wrote down.
Credit
Gavin Andresen (post-mortem, BIP 50); Marek Palatinus and Michael Marsee
2021
A fast chain halts
Solana, a chain built for speed, stopped for about 17 hours in September 2021, when bot transactions overflowed validators' memory, and for almost 19 hours in February 2023. Helius, a firm that runs Solana infrastructure, lists ten incidents between December 2020 and August 2024. Solana resumes only when validators holding at least 80% of the staked coins are back online, and in December 2020 most of the downtime was spent waiting for them.
LessonWhen a ledger needs most of its validators to move, a bug becomes hours of waiting for them to restart together.
Credit
Helius (record); Solana's validators and core developers
2022
Stake replaces mining
On 15 September 2022 Ethereum moved from proof of work to proof of stake, in which validators lock up coins as a bond instead of spending electricity on a race. A report by the Crypto Carbon Ratings Institute (CCRI), commissioned by the Ethereum software firm ConsenSys, put the fall in electricity use at 99.988%, from about 22.9 million MWh a year to about 2,601 MWh. In July 2026 the Cambridge Centre for Alternative Finance put Ethereum's use at about 7.87 GWh a year, after later upgrades.
Credit
Ethereum's developers
2023
Speed claims measured
Researchers from the University of Sydney, EPFL and TUM ran six chains under five realistic applications on the same footing. Solana, claimed at 200,000 transactions a second, reached 8,845, and Avalanche, claimed at 4,500, reached 323; they found "none of the tested blockchains can yet support the load of these realistic DApps".
LessonA chain's own throughput figure says little until someone runs it under a stated workload.
Credit
Vincent Gramoli, Rachid Guerraoui, Andrei Lebedev, Chris Natoli and Gauthier Voron (Diablo, EuroSys 2023)
Watch for
Finality in 150 milliseconds
Solana's Alpenglow upgrade, which replaces its voting and finalization rules, went live on mainnet on 21 October 2026. The Solana Foundation aims to cut finality, the point after which a transaction cannot be reversed, from about 12.8 seconds to about 150 ms; its own dashboard showed 6.6 seconds at launch, by Cointelegraph's report.
Achieved whenan outside party measures mainnet finality near 150 ms over several weeks.
Claimed by
Solana Foundation and Anza, October 2026
Each block proven in time
A zero-knowledge proof lets anyone check that a block was processed correctly without redoing the work, and the goal is to make one within Ethereum's 12-second block time. Succinct said in May 2025 that 93% of blocks were proved in under 12 seconds "in our tests". Brevis reports an average of 6.1 seconds and 99.9% of 1,000 blocks on 16 RTX 5090 graphics chips, about $100,000 of hardware, while Ethproofs, a public tracker funded by the Ethereum Foundation, showed 91.73% for its older version, by EthPanda's account.
Achieved whenEthproofs or another outside party records the Ethereum Foundation's standard (99% of blocks within 10 seconds, on hardware under $100,000 and 10 kW) over a month of mainnet blocks.
Claimed by
Succinct, May 2025; Brevis, May 2026
Fewer qubits to break keys
Google Quantum AI estimates that a quantum computer with fewer than 500,000 physical qubits could break the 256-bit elliptic-curve keys that protect Bitcoin and Ethereum "in a few minutes", about 20 times fewer qubits than earlier estimates; it published a zero-knowledge proof of its result instead of the circuits. Bitcoin's response is still a draft: BIP-360 would add an output type without an exposed key but "does not include the introduction of post-quantum signature schemes", and BTQ Technologies runs it on its own test network.
Achieved whenthe estimate passes peer review or an outside group reproduces it.
research.googleraw.githubusercontent.comthequantuminsider.com
Claimed by
Google Quantum AI, March 2026
Open
01
A register of chain halts
An open dataset of every time a major public chain stopped adding or finalizing blocks, found from the chains' own block times (any gap far beyond the normal interval), for the 30 largest chains since 2015. Each halt is matched to the chain's post-mortem and its cause coded: software bug, spam, consensus fault, upgrade or other.
The briefCredit · unclaimed02
Proving claims, rerun outside
An outside rerun of the published real-time proving results. Rent the hardware a team names, run its released prover on 1,000 consecutive Ethereum mainnet blocks, and publish the share proved within 12 seconds, the slowest 1% of times, the power drawn and the hardware cost.
The briefCredit · unclaimed03
Forced exits, tried on testnets
Rollups are ledgers that run on top of Ethereum and post their records to it, and most promise that a user can force a withdrawal through Ethereum if the rollup's operator stops or censors them. Use that path on the public test network of each of the 15 largest rollups, without the operator's ordinary route, and time each attempt from request to funds back on Ethereum's test network.
The briefCredit · unclaimed
Open 01 · the brief
A register of chain halts
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
An open dataset of every time a major public chain stopped adding or finalizing blocks, found from the chains' own block times (any gap far beyond the normal interval), for the 30 largest chains since 2015. Each halt is matched to the chain's post-mortem and its cause coded: software bug, spam, consensus fault, upgrade or other.
Why it is still open
Histories exist one chain at a time. Helius's record of Solana lists ten incidents from 2020 to 2024 with causes and downtimes, but covers one chain, and this research found no register across chains.
Built with what exists
Public RPC endpoints and block explorer APIs for block times; DuckDB or pandas to find the gaps; an AI model to read post-mortems and propose cause codes, each checked by a person (see the Note "The jobs agents repeat").
First step
Pull Solana's block times since March 2020 and check that the ten incidents in Helius's history appear as gaps.
Done looks like
30 chains from 2015 to 2026, every gap above a stated threshold listed with start, end and duration taken from block data, at least 90% of halts longer than one hour matched to a post-mortem, a 10% sample coded again by a second person with at least 90% agreement, and the data published under CC BY.
Open 02 · the brief
Proving claims, rerun outside
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
An outside rerun of the published real-time proving results. Rent the hardware a team names, run its released prover on 1,000 consecutive Ethereum mainnet blocks, and publish the share proved within 12 seconds, the slowest 1% of times, the power drawn and the hardware cost.
Why it is still open
The figures come from the teams themselves. Succinct's are from "our tests", Brevis's 6.1-second average and 99.9% come from internal tests on 1,000 blocks starting at block 24,000,000, and Ethproofs tracks teams proving on their own hardware.
Built with what exists
The provers the teams have released as open source (Succinct says SP1 Hypercube's verifier and block-proving code are open, with a production prover to follow); rented machines with 16 NVIDIA RTX 5090 GPUs, the setup Brevis describes; a metered power supply; Ethproofs results as the reference.
First step
Check which provers are released in full, then prove 100 consecutive blocks with one and publish the timing log.
Done looks like
Two provers, 1,000 consecutive blocks each, results set beside each team's claim with any difference reported, and logs and scripts public so a third party can repeat one run.
Open 03 · the brief
Forced exits, tried on testnets
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
Rollups are ledgers that run on top of Ethereum and post their records to it, and most promise that a user can force a withdrawal through Ethereum if the rollup's operator stops or censors them. Use that path on the public test network of each of the 15 largest rollups, without the operator's ordinary route, and time each attempt from request to funds back on Ethereum's test network.
Why it is still open
L2BEAT grades rollups on whether such exits exist and how long users have to leave before an upgrade (at least 7 days at its Stage 1 and 30 days at Stage 2), and says its framework covers decentralization "and not in terms of security". A 2025 paper by Francisco Gomes Figueira, Martin Derka, Ching Lun Chiu and Jan Gorzny warns that if the operator fails, "users may lose access to digital assets on the rollup". This research found no published record of these exits being tried across rollups.
Built with what exists
Each rollup's documented forced-transaction contracts on the Sepolia test network; Foundry's cast tool for scripted transactions; test ether from public faucets, so no real money moves; L2BEAT's project pages for the paths to try.
First step
On Sepolia, send one forced transaction through one rollup's inbox contract on Ethereum and record how long it takes to be included.
Done looks like
15 rollups, three attempts each, with time and outcome for every attempt in a public table and scripts that others can rerun.