Quarlcoin is a UTXO chain. It is Bitcoin Core's architecture with three things
changed, each for a reason that can be checked rather than argued about. It
launched today; the numbers below are the ones the chain is actually running.
**Source:**
https://github.com/quarlcoin/quarlcoin**Release 0.1.0:**
https://github.com/quarlcoin/quarlcoin/releases/tag/v0.1.0---
## The launch
- Genesis: `2f5ec90129e2cb5bbc26a2e972898e7b22a635344f0cf524117021bbb46cb321`,
timestamped 2026-08-03 13:00:02 UTC
- Binaries published: **2026-08-08 07:47:52 UTC**
- First block mined: **2026-08-08 08:03:05 UTC**
Fifteen minutes separate those last two, and they are stated rather than
characterised. The first blocks were mined by the developer on a laptop CPU
because nobody else had downloaded the release yet. Difficulty starts at 1, so
whoever arrives next is on the same footing within one block.
**There is no premine and no allocation to anyone.** The genesis coinbase pays
to `OP_RETURN`: the 50 QRL of the first block cannot be spent by anybody, which
is a property of the chain and not a promise. Every coin after it is mined.
There is no ICO, no sale, no fund, no team allocation, no fee to a developer
address. There is nothing to buy from anyone.
---
## What is different
**BLAKE3 instead of SHA-256, once instead of twice.**
Bitcoin hashes twice because SHA-256 is a Merkle–Damgård construction and leaks
a length extension: given `H(m)` and the length of `m`, anyone can compute
`H(m ‖ padding ‖ suffix)` without ever seeing `m`. The second pass repairs that,
at the cost of doing the work twice. BLAKE3 is a tree over a keyed permutation
with a finalised root, so there is nothing for a second pass to repair.
The block hash, the merkle root, every txid and wtxid, the sighash, the address
hash and the base58 checksum are one BLAKE3 each. The address hash is the same
root read out to twenty bytes rather than thirty-two, so `HASH160` is a length
and not a composition of two functions. RIPEMD-160 survives only as the
`OP_RIPEMD160` opcode.
libsecp256k1 is carried as a subtree with one change: its SHA-256 is BLAKE3, so
there is no SHA-256 anywhere in the system. The rename of `secp256k1_sha256` to
`secp256k1_hash256` throughout that subtree is part of that change — a struct
named for a hash it does not compute is how a reader comes to a wrong conclusion
about consensus code.
**ASERT instead of a retarget epoch.**
There are no difficulty periods. The target of a block is an exponential in how
far its parent is off schedule, measured from a single anchor — the genesis
block — so every target the chain will ever have is a function of three numbers,
and a node can compute the difficulty at any height without walking the chain to
it. Hashrate that arrives or leaves is answered on the next block rather than up
to two weeks later. The half-life is 12 hours.
**Two and a half minute blocks, 84 million coins.**
50 QRL a block, halving every 840,000 blocks, which is the same four years
Bitcoin halves on.
## What is the same, deliberately
The UTXO model, the script system, SegWit and Taproot, descriptor wallets, the
peer-to-peer protocol, the mempool, the block database, the RPC interface — all
of it is Bitcoin Core's, because all of it works and none of it was the thing
worth changing. Signatures are secp256k1: ECDSA for SegWit v0 and Schnorr
(BIP 340) for Taproot.
---
## Networks
| | main | test | regtest |
|---|---|---|---|
| P2P port | 9743 | 19743 | 19844 |
| RPC port | 9742 | 19742 | 19843 |
| address prefix | `Q…` `R…` `qrl1…` | `m…` `2…` `tqrl1…` | `m…` `2…` `qrlrt1…` |
| BIP32 keys | `43q…` | `42S…` | `42S…` |
| BIP44 coin type | 9743 | 1 | 1 |
None of these values is shared with Bitcoin, Litecoin or Dogecoin. The BIP32
version bytes and the coin type in particular were checked against those chains
and against the SLIP-0044 registry: a wallet or a block explorer shown a
Quarlcoin key must not report a Bitcoin address, and the way that goes wrong is
by inheriting a constant rather than choosing one.
---
## Download
Linux x86-64 and Windows x86-64, seven binaries each — the node, the Qt wallet,
a CPU miner and three utilities.
```
123d38b15b736ea2d4672f79290584416e27a9b99202f3cdb7d8a82bd8e2a58b quarlcoin-0.1.0-x86_64-linux-gnu.tar.gz
b74feefbd008e7d36ee9613cd6cef67361d1da546c886de6d57d37886bce25eb quarlcoin-0.1.0-x86_64-w64-mingw32.zip
```
Nothing is signed. A signature made on the machine that produced the binaries
proves that the machine produced them, which is what the checksum already says.
Signing belongs to a separate key on separate hardware, and pretending otherwise
would be worse than leaving it out. Build from source if that matters to you:
```sh
git clone
https://github.com/quarlcoin/quarlcoincd quarlcoin
cmake -B build -DENABLE_GUI=ON
cmake --build build -j"$(nproc)"
ctest --test-dir build
```
## Running
```sh
quarld -daemon -addnode=1.240.102.201
quarl-cli getblockchaininfo
```
The network is one seed old, so name it outright for now rather than relying on
discovery. More seeds is the first thing this chain needs from anyone who wants
to help.
## Mining
```sh
quarl-miner -address=<your QRL address> -threads=8
```
Roughly 37 MH/s on eight laptop cores, which is what the opening difficulty was
calibrated against. There is deliberately no GPU miner in this repository: a
slow official one would set a floor nobody has a reason to beat. The two RPCs it
uses, `getblocktemplate` and `submitblock`, are the same two anyone else's miner
would use, so a better miner is a straightforward thing to write and it is worth
more to this chain than another node would be.
---
## What is not here
This list is the reason to read the rest of it sceptically.
- **No second implementation.** Every node on the network runs this code, so a
consensus bug in it is a consensus bug in the network.
- **No security audit.** The consensus changes are covered by tests that check
them against independent references — the ASERT target against a separate
implementation in Python, BLAKE3 against the official vectors and its own SIMD
kernels against its portable one — but nobody outside this project has read
the code.
- **One seed node, and a chain hours old.** A chain this short reorganises for
free, and a single participant can outweigh it. Treat anything you mine as
unconfirmed for far longer than you would on a chain with history.
- **No hardware wallet, no PSBT tooling, no block explorer.** The signature
scheme is Bitcoin's, but the sighash is BLAKE3, so a device that computes the
sighash itself will not sign for this chain until it is taught to.
- **No exchange, no listing, no price.** Nobody is selling anything and there is
nothing to buy. If you see a Quarlcoin sale, it is not this project.
## Contact
shortvector@proton.meIssues and pull requests:
https://github.com/quarlcoin/quarlcoin
