Update, 10 November 2026: Ting has its own website now, with the wiki, at
ting-project.com. A third-party pool is running, the Discord link has changed, and the network has grown to around 220 GH/s.
What is Ting?I've been working on Ting, a new proof-of-work blockchain that went live on 21 October 2026.
The idea is to keep the underlying consensus as small as possible, and freeze it permanently. Consensus handles transaction ordering, proof of work, deterministic program execution, state commitments and resource accounting. Everything else, including accounts, coins and signature schemes, is implemented as programs running on the chain.
The rules are written in a single specification and were frozen at launch. There is no upgrade mechanism, governance vote or foundation with the authority to change them. Changing the rules means starting a different chain.
Ting also has no native coin. The first coin is Daler (DLR), deployed using the same coin program that anyone else can use. Daler has no premine and is issued through mining rewards.
A few reasons to take a look:- CPU and GPU mining. Ting uses SHA-256, but not Bitcoin's block format or double-SHA-256 PoW. Existing Bitcoin ASIC miners cannot mine it. The network is currently small enough that an ordinary home computer can find blocks.
- No premine. Daler started with zero supply, no allocations and no minting authority. All issued coins have come from blocks mined since its deployment.
- Working software. There are node implementations in Rust and Go, a miner with a built-in wallet, an explorer, REST API, documentation, an Android wallet and a couple of example applications.
- Fixed consensus. The meaning and resource cost of a deployed program cannot be changed by a later protocol upgrade. Applications can still evolve without changing consensus.
- Open source. The protocol rules, implementations, tests and conformance vectors are all public. Anyone can inspect them or build another implementation.
This is not a Bitcoin fork, and there is no exchange listing or price to discuss. I'm posting it here because the network is running and I'd like more people to look at it, try mining and review the design.
Technical details| Consensus | Proof of work |
| PoW | Single SHA-256 over Ting's tagged block header, with a 16-byte nonce |
| Bitcoin ASICs | Not compatible; no merge-mining |
| Target block time | 60 seconds |
| Difficulty adjustment | Every block, ASERT-type schedule anchored at genesis, 12-hour half-life |
| Mainnet launch | 21 October 2026, 00:00 UTC |
| Smart contracts | Deterministic integer WebAssembly profile (115 instructions, 12 host functions) |
| Transactions | Deploy and invoke |
| Resource accounting | Compute, state access, data and storage growth |
| Block capacity | Adjusted by miner voting within fixed rules and floors |
| Node implementations | Rust reference node and Go implementation |
| License | MIT / Apache 2.0 |
There is no built-in ECDSA requirement, account format, currency or fee mechanism. Those belong to programs. The intention is that applications can introduce new cryptography or account models without requiring the underlying protocol to change.
The Go implementation was written from the specification rather than ported from the Rust code. It validated the existing chain on its first run, and both implementations have since passed the same conformance tests.
The complete specification is available
here, with more readable explanations in the
wiki.
Daler (DLR)Daler is the first coin deployed on Ting. It uses the standard coin program, which can also be used to create other coins with different supply and issuance rules.
| Symbol | DLR |
| ContractId | 305378d17f6af84a563d0f8a8455084948aff7330b62b4f2f98a44b82ba113c4 |
| Deployment | Block 443, 21 October 2026, 13:47 UTC |
| Maximum supply | 210,000,000 DLR |
| Block reward | 100 DLR |
| Halving | Every 1,050,000 blocks (about 2 years) |
| Tail emission | None |
| Premine | 0 |
| Initial allocations | None |
| Minting authority | None |
| Decimals | 8 |
Since Daler was deployed at block 443, earlier blocks earned no DLR. Rewards are claimed by a transaction in the mined block. If a block doesn't specify a payout account and claim its reward, nothing is issued for that block.
You can check the coin's configuration and current supply directly through a node:
Daler configuration |
Current issuance |
Coin program documentationMining so farAs of block 27,749 on 7 November 2026, the distribution of claimed rewards was:
| Miner | Blocks | DLR claimed |
| My original wallet | 2,440 | 244,000 |
| Another miner | 3,544 | 354,400 |
| My new wallet | 1 | 100 |
| My node, no payout | 21,322 | 0 |
At that point, there were two of us mining Daler. The other miner actually found the project on GitHub before I made any public announcement, contacted me, and started mining on 5 November. They had already mined more Daler than I had.
I've also had a node running to keep the chain producing blocks during development. It deliberately had no payout account, so 21,322 blocks went by without claiming Daler rewards. That's 2,132,200 DLR in unclaimed rewards.
All of this can be verified by going through the blocks using the public API. The numbers above are a snapshot, not live totals.
Since then, more miners have joined. By block 38,311 on 10 November, 1,653,500 DLR had been issued, and 19 different payout accounts found blocks in the 15 hours before that, none with more than about 30% of them.
Mining TingThe miner is a full Ting node with mining and a local wallet built in. It can use the CPU, GPU or both, and benchmarks the available methods to pick the fastest one automatically.
CPU mining uses SHA instructions where supported. GPU mining uses Metal on macOS, Vulkan on Linux/Windows or DirectX 12 on Windows.
On 10 November the network was around 220 GH/s, up from 0.7 GH/s when I first posted this. For comparison, the miner gets approximately 690 MH/s on a MacBook Air M3 using both CPU and GPU. That's just a benchmark, of course, and the network difficulty will adjust as miners join.
The official miner mines solo. A third-party pool runs at crypto-connection.net (Stratum, PPLNS, 1% fee, 100 DLR minimum payout). It needs a miner that speaks Stratum, which the official miner doesn't.
Others have started building miners and pools for Ting. They're run by their authors, and I don't review or endorse them. The only official releases are the ones under
github.com/ting-organization.
Getting started1. Download the miner from
GitHub releases.
2. Extract the archive and check its SHA-256 checksum against SHA256SUMS.
3. Run the following commands from the extracted folder:
./ting init
./ting wallet create
./ting miner payout <account shown by wallet>
./ting miner start --threads 4
./ting start
The wallet command creates a key and shows your account address. Use that address for the payout setting. You can also use an account from Ting ID instead.
The node will connect to the production network, synchronize and begin mining. The thread count controls how many CPU threads the miner may use.
On Windows, replace ./ting with .\ting.exe.
Supported platforms:- Windows x86_64
- Linux x86_64
- Linux ARM64
- macOS Apple silicon
- macOS Intel
Check mining status with:
The full instructions, configuration options and benchmarks are in
MINER.md.
WalletsBuilt-in miner walletThe miner has its own wallet, so you don't need a phone or another application to receive or send mining rewards.
./ting wallet account
./ting wallet balance
./ting wallet send <account> 12.5
The private key is generated locally and stored encrypted in wallet.json, using Argon2id and XChaCha20-Poly1305. The CLI never displays the private key.
Make sure to back up wallet.json and keep the password somewhere separate. Losing either means losing access to the funds.
Wallet documentationTing ID (Android)Download Ting IDThere's also a native Android wallet called Ting ID.
An identity has two accounts:
- Spending: Normal payments, available immediately.
- Savings: Outgoing payments are delayed by seven days and can be cancelled during that period.
It also supports recovery through a recovery phrase or guardians, contacts, QR payments and signing requests from applications.
Ting ID is still early software. I wouldn't recommend keeping anything significant in it until it has had more use and review.
ApplicationsTwo small applications are already running on Ting:
Ting SwapAn on-chain coin exchange. Users can list one coin in exchange for another, and accepting a listing settles the exchange in a single transaction.
Ting CanvasA shared canvas where anyone can paint. A new canvas starts every week.
These are mostly examples of what can be built using programs on Ting, rather than the main purpose of the project.
A few things worth pointing outI don't want to pretend this is a mature network. There are some obvious limitations and trade-offs.
- The network is small. It has grown from roughly one laptop to around 220 GH/s, but somewhere around 10 to 30 high-end GPUs could still match it, and someone with that much rented hashing power could potentially reorganize the chain. It's also concentrated: one pool and one other account each find about a quarter of the blocks. I would not treat payments as secure after just a handful of confirmations. More independent miners are needed.
- Frozen consensus is a trade-off. It means nobody can change the rules under existing programs, including me. But it also means consensus bugs cannot be fixed through an upgrade. If something fundamentally unsafe is discovered, the answer is a new chain, potentially carrying over the old chain's state. I know this is a controversial design choice.
- Both node implementations are mine. The Go implementation was written from the rules rather than translated from Rust, and both agree on the existing chain. That's useful, but it isn't the same as an independently developed third implementation. I'd very much welcome one.
- AI-assisted development. A substantial amount of the code was written with AI coding assistants under my direction. This is documented openly in the repository. The protocol is specified separately, with conformance vectors and tests whose expected results are derived from the rules rather than the implementation under test. The code and tests are there for anyone to examine.
- Unsigned binaries. Release archives include SHA-256 checksums and build attestations, but the desktop binaries are not code-signed. Verify downloads before running them, or build from source.
For anyone interested in the reasoning behind the design, there's a longer explanation here:
Why consensus is frozen |
AI usage and testing
Feedback and questionsThis is the first public announcement of Ting. It's still a very small project, but the network is live and the software is available to try.
I'm particularly interested in feedback on the protocol design, consensus specification, and anything that looks incorrect or questionable. If anyone wants to try mining, run a node or build something on it, that would be great too.
I'll be following this thread and answering questions.
GitHub:
ting-organizationWebsite:
ting-project.comDiscord:
discord.gg/aH32zkEN58Discussions:
GitHub DiscussionsSecurity reports:
SECURITY.mdVideos:
YouTube