[ANN][PoW] SHEAR (SHE) | SHE is private | CPU-mined ShearHash | Mainnet 21Dec 26

[rpAI]

SHEAR (SHE)
SHE is private.
A CPU-mined proof-of-work coin where amounts and links stay off the public record

Mainnet genesis: Monday 21st December 2026, 2pm UK time
v12 public testnet: early December 2026

Website | Docs | Whitepaper | Source | Miner | Pool

What Shear is

SHE is private, and we make it so. Shear is a new proof-of-work chain, built from scratch, for people who want money that doesn't broadcast what they do with it.

You mine it on an ordinary CPU. There's no premine, no founder reward, no sale of SHE by the developers, no ICO and no airdrop. Every SHE that exists was mined, or minted by the protocol's own Reserve rules described below.

When you send SHE, the chain doesn't publish how much moved, which coin you spent, or who received it:

• Amounts are hidden. Every note carries its value inside a cryptographic commitment with a range proof. Nodes check that the sums add up without ever seeing the numbers.
• Spends don't point at the coin they spend. A spend proves its note is one of this book's notes, across the whole set (ADMITv3 full-set membership), without saying which one. A one-time spend tag stops the same note being spent twice.
• Addresses don't sit on the chain. You give a payer a silent ID (she1...). The coin lands on a fresh revolving dest (ssa1...) that the book writes. Your rest-frame address (shear1...) never appears in a block.

An honest caveat: the membership set is only as large as the number of notes in the book. Early on, while the book is small, that set is thin. It grows as the chain is used.


Coin and mining

Ticker	SHE
Smallest unit	rho (1 SHE = 100,000,000,000 rho)
Algorithm	ShearHash-v3, a CPU hash built on RandomX (light mode)
Block time	90 seconds
Difficulty	aserti3-2d, network-wide and anchored to genesis. No pool or miner can set a private target
Block reward	Starts at 1.00 SHE per block. It falls 0.01 SHE each 400-day epoch, down to a floor of 0.20 SHE
Who gets the reward	Split between the hashers that proved work that round, not just whoever found the block
Hash bonus	Each proven share earns its own small bonus on the next sealed block. The size of the bonus is set by governance vote
Transaction fees	Paid in full to the finder of the block that seals them
Pool fee	The public pool takes 1% of the block pot. It's an ordinary node with a stratum, and you can solo mine against your own node instead

Mining uses the official ShearK-Miner on Windows, Linux (Debian/Ubuntu, Arch, Fedora, openSUSE) and macOS. Point it at stratum+ssl://pool.shear.digital:443, or run your own node and mine solo.


The Vault and the Reserve

The only way SHE is created outside mining is the Reserve. You lock SHE in the Vault for an epoch. When it matures, you withdraw it together with a Reserve reward that the protocol mints.

• A Vault lock is withdraw-only. It never counts as spendable while it's locked, and withdrawing is the only way out.
• The reward rate is fixed when an epoch opens. A vote can't change it.
• Holders vote on the hash bonus. Votes can't touch the block reward schedule.

Wallet

The Continuum wallet (GUI and CLI) runs on Windows, Linux and Android. A macOS build is still to come.

• Spendable is the first thing you see after logging in. It's exactly what a send can use, never a padded figure.
• Incoming payments show as Pending until the wallet has confirmed them against the chain.
• Backup is your shewall.bin file plus its password. There's no seed phrase. Keep both safe, and never give either to anyone. Nobody from Shear will ever ask for them.

Where things stand

We're in the final hardening stretch before the v12 public testnet. Over the last few weeks, a lot of that work has gone into finding and fixing problems before anyone's money depends on the chain. One of them was a block-integrity fix that makes sure every block commits to exactly the transactions inside it.

We moved mainnet to Monday 21st December 2026 at 2pm UK time, so v12 gets at least a week of real use on testnet first, with time to spare. We'd rather launch late and solid than early and broken.

The road to mainnet:
[list=1]

• Finish the remaining work below (through November)
• A fast-block private test network, where every feature gets used on desktop and Android and anything that breaks gets fixed (from around 27th November)
• v12 public testnet launch (around 3rd December)
• At least 7 days of real use on the v12 testnet
• Mainnet genesis: 21st December 2026, 2pm UK time

Still in progress, with the date each piece should be done:

• Network and block-integrity hardening: ~28 Oct
• Vault lock safety, so only a lock's owner can ever withdraw it: ~30 Oct
• Command-line wallet Vault locks: ~31 Oct
• Mining hash safety (no native crashes, and it works on older CPUs): ~2 Nov
• Correct network settings everywhere, so testnet, mainnet and dev nets can't mix: ~4 Nov
• Signed name registration: ~5 Nov
• Stricter checks on data arriving from the network: ~7 Nov
• Fairer per-share credit for miners: ~8 Nov
• The hidden miner payout: ~9 Nov
• Extra spend hardening: ~10 Nov
• Tests moved to real chain data on every consensus path: ~13 Nov
• User-created tokens. Anyone can create a token with a fixed, public total supply that can never be added to. Token amounts and holders get the same hiding as SHE. The design is reviewed first, then built: ~19 Nov
• The rest of the test suite moved to real chain data: ~21 Nov
• Five older pieces of work brought up to date (gas limits, reward claims, withdraw rules, Vault fees in the wallet, launch checklist): ~25 Nov
• Clearer wording and polish across the wallet: ~27 Nov

These dates are our best estimates, not promises. Everything is reviewed and tested before it lands, and if testing finds a problem, we fix it first. We'll post here whenever a date moves.


Testnet notice

Until mainnet genesis, everything is testnet. Testnet coins have no value, and testnet chains can be reset. Mainnet doesn't exist yet. Nodes and wallets refuse to produce mainnet coins before genesis.

This is new software. Run it, test it, break it and tell us what you find. Please don't treat anything here as financial advice.


Links

• Website: https://shear.digital
• Docs: https://shear.digital/docs
• Source (node, wallet, pool, site): github.com/rgsneddon/shear-testnet
• Official miner: github.com/rgsneddon/ShearK
• Pool: https://pool.shear.digital (stratum+ssl://pool.shear.digital:443)
• P2P seeds: to follow

Thanks for following along. Questions, bug reports and hard criticism are all welcome in this thread. Please donate to help towards the ongoing development of SHEAR: https://donate.shear.digital/

[/list]

[yoshikiazuma]

hi there, she.

good to see we are running tests before mains to the net

could you update here in a couple of days time to time how it's going

and inform us with the good news of the mainnet launch

ill will give it a go with the testnet and see how the software looks out to be overall

(maybe im glad to be able to check the testnet out for the first time look around and maybe the mainnet will more inviting)
 

[rpAI]

Further tests... I have deployed shear-testnet-v2

1hash=1tx is now active on v2 and fully confirmed coins are being minted within a target average of 9 seconds.
block rewards now 0.1 per 9 seconds (keeps the 1 SHE every 90 seconds) and block bonuses are fairer, with a 10ms miner hook into new blocks at the official pool.

This is the power of the Chronoflux mathematical equations at work.

amazing, ground-breaking.
If it wasn't for the privacy mechanisms, we could indeed push up to and beyond 4million TPS.

You can mine at shear-testnet-v2 here: https://pool.shear.digital/

1hash=1tx stays one bonus unit per hash, but the chain does not store one JSON object per hash.

Each 9s block rolls a Flow bundle: one sample row per hasher with count = N. Coinbase still pays N × 1 protocol unit. After 100 confirms, sample bodies prune; the header continuity root and coinbase stay sealed. That is how Shear stays lean while throughput rides the 9s pot.
watch it at work!

[rpAI]

Lots of work done last night - finalisation of the consensus rules on test-net.

Previously researched the possibility of 9 second blocks with 0.1 rewards. This proved a load on the scalability so we are reverting to the optimal 1Shear /90second block time and chronoflux still does the pruning work to keep the chain lean and usable.

Baked in 6 confirmations til spendable (policy recommendation of 12) and applied the tx fee structure (more on this later).

recutting clients to the new rules now... watch this space!

[rpAI]

IMPORTANT NOTICE: The Shear pool (https://pool.shear.digital) is now fully fledged, tested and working as it should. No further testing of the pool software is required for the SHEAR pool between now and mainnet launch.

I'll leave a thread on it so testnet continues to move blocks but for now, the work is done.

If you are mining on testnet, you are just burning money, so move off now please.

THANK YOU TO ALL WHO MINED TO HELP ME BUILD THIS POOL SOFTWARE - I WILL OPEN-SOURCE THE POOL AT MAINNET LAUNCH

[rpAI]

Good news!
The final touches are being implemented in testnet and we are on-schedule to launch in 3 days time.

With excellent privacy preserving mechanisms like silent addresses and stem/fluff at node level, no user info is broadcast and we have implemented some excellent security measures to ensure your coins are your coins with no exceptional vulnerabilities in the project.

We look forward to welcoming miners on Friday 11th September at 9PM BST.

Docs section has been added to the website, as well as the Shear white paper.

[rpAI]

Mainnet launch delay.

A quick note: due to the failure of my windows machine, I am forced to push back mainnet launch for another week. Launch will be Friday 18th September at 9pm UK time.

Apologies for this further delay. I am sourcing a new windows box and will have all work completed for next Friday, ready for a fair launch of Shear.

Thanks for your patience.

[rpAI]

UPDATE!

The first coin to mint and prove a separate hash bonus in the coinbase and make a missing dest invalid:
SHEAR on shear-testnet-v2
at testnet block height 1479.

Now it's time for the serious work.

Shear Admittance — naming lock for the full Shear anon set
=======================================================

Date: 2026-09-12

ONE-LINE
--------
Admit proves a spend is an admissible extraction from the committed
current J without revealing which flowline carried it.


CHRONOFLUX MAP (motivation only — not a consensus string)
---------------------------------------------------------
Principia primitive:  div J = 0,  J = rho_t * u
A configuration is admissible iff evolution preserves that relation.

  Principia                         Shear Admit
  --------------------------------  --------------------------------
  temporal density rho_t            live output density (the fluxset)
  flow u                            worldline of a spendable output
  current J                         committed spendable set
  admissible configuration          spend that verifies against the current
  commitment to J                   tree root (jroot)

LOCKED NAMES
------------
Layer                  Name                 Notes
---------------------  -------------------  ------------------------------------------
Public / user copy     Admittance           "Admittance hides which output moved."
Long form              Shear Admittance     spec titles, CCS-style write-ups
Consensus / tx type    AdmitV1              first live type; bump to AdmitV2 on
                                            breaking statement change
Crate / module         shear-admit          from the principals 'admissibility' fundamentals
Live output set        fluxset              the set Admit proves against
Accumulator root       jroot                commitment to J as of a reference height
Proof blob             admit_proof          opaque bytes; length from input count
                                            + tree depth
Construct / verify     admit_prove          FFI and daemon
                       admit_verify

SPEC HEADER (paste into protocol notes)
---------------------------------------
Admit v1 — Shear admittance proofs
Operator cites: Curve Trees (Campanelli, Hall-Andersen, Kamp,
              USENIX Security 23 / ePrint 2022/756)
Statement   : spend is admissible in J = rho * u over the fluxset
Auth        : Shear spend-key / EIP-712 intent
Linkability : Shear native (dest / spend-tag as already specified);

EXPANDED CITATIONS
----------------------------------------------------------
- Accumulator: Campanelli, Hall-Andersen, Kamp — Curve Trees: Practical
  and Transparent Zero-Knowledge Accumulators, USENIX Security 23,
  ePrint 2022/756.
- If divisor gadgets or Parker leaf/SA+L split are used: cite Parker
- Reviews of those gadgets stay attributed to their authors
  (Cypher Stack / Veridise / etc.). They are not "Admit audits" until
  they cover Shear's statement.

Allowed in prose:
  "Admit instantiates Curve Trees."
  "Motivation is Chronoflux continuity."


DOC SENTENCE BANK
-----------------
User:  Admittance hides which output moved.
Spec:  Admittance proves a spend is an admissible extraction from the
       committed current J without revealing which flowline carried it.
Audit: Admit v1 is a Curve Trees membership proof over Shear's fluxset
       plus Shear spend-auth.

IMPLEMENTATION NOTES
-----------------------------------------------------
- Helios/Selene exist to tower Ed25519. If Shear's base curve is not
  Ed25519, pick a cycle for that field or accept a worse in-circuit
  hash.
- Public amounts now depreciated (Shear privacy goal: amounts hidden, dests stealth, full anon set)
- Tree grow/trim/reorg, reference height, and unlock rules are Shear
  consensus — write them under Admit.

[rpAI]

How to solo mine Shear (testnet v4) — full setup

Solo = your own validating node + ShearK pointed at localhost.
Same node binary as everyone else. Local stratum is 127.0.0.1:1111 (no pool fee, no operator custody).
When you seal a block, ShearK prints BLOCKFOUND!!!

Pins (do not use older cuts)

• Book / magic: shear-testnet-v4
  
• Wallet: Continuum 0.45 packs (or 0.46 when APK/zips are on the tag)
  releases/tag/0.45
  releases/tag/0.46
  
• Miner: ShearK-Miner 2.5
  ShearK releases/tag/2.5
  
• Tree: github.com/rgsneddon/shear-testnet
  
• Public seeds (IBD): p2p.shear.digital:30303 , r2r.shear.digital:30303 , b2b.shear.digital:30303
  
• Public pool (not solo): pool.shear.digital:1111
  

You will need

• Continuum wallet — create/backup shewall.bin and Copy dest (ssa1…)
  
• Node.js 18+ (20 LTS fine), git, and a C toolchain for the native ADMIT addon
  
• ShearK-Miner 2.5 binary
  
• Disk + time for IBD (sync the tip before you expect finds)

A) Wallet — get your ssa1 dest

• Download Continuum for your OS from the official shear-testnet release tag (not a random mirror).
• Create / unlock the wallet. Backup shewall.bin + password offline. Never share shear1 (rest-frame), the password, or shewall.bin.
  
• On Continuum, use Copy dest — that is your owned ssa1…. Solo login must be that ssa1 (or she1 + --dest ssa1 on ShearK).
  
• she1 is what you share for payments. Mining login for solo should be the ssa1 dest so bonuses land correctly.
  

B) Dependencies — Linux (Debian/Ubuntu)

Code:

# Node 20 example (any Node >= 18)
curl -fsSL https://deb.nodesource.com/setup_20.x | sudo -E bash -
sudo apt-get install -y nodejs git build-essential python3

node -v   # >= v18
npm -v
git --version
gcc --version

Arch-like:

Code:

sudo pacman -S nodejs npm git base-devel python

C) Dependencies — Windows

• Install Git for Windows.
• Install Node.js 20 LTS (nodejs.org) — include npm.
• Install a C++ build toolchain for node-gyp / native ADMIT ("Desktop development with C++" / VS Build Tools).

PowerShell check:

Code:

node -v
npm -v
git --version

D) Clone & build the node

Code:

git clone https://github.com/rgsneddon/shear-testnet.git
cd shear-testnet
git checkout main
git pull

npm install
npm run build:admit

node node/src/node.js --print-config

Data dir default:

• Linux/macOS: ~/.shear/testnet-v4
  
• Windows: %USERPROFILE%\.shear\testnet-v4
  
• Override: SHEAR_DATA
  

E) Start a solo node (IBD + local stratum)

Code:

cd shear-testnet
npm run solo

That is node node/src/node.js --solo.

What it does:

• Validates the shear-testnet-v4 book
  
• P2P on :30303 (public seeds)
  
• RPC on 127.0.0.1:18332
  
• Solo stratum on 127.0.0.1:1111 only (not exposed to the internet)
  

Optional env:

Code:

SHEAR_SOLO=1
SHEAR_STRATUM_PORT=1111
SHEAR_STRATUM_BIND=127.0.0.1
SHEAR_SEEDS=p2p.shear.digital:30303,r2r.shear.digital:30303,b2b.shear.digital:30303
SHEAR_DATA=/path/to/datadir

Leave the node running. Wait until height catches the public tip (pool /api/stats).

Check tip:

Code:

curl -s https://pool.shear.digital/api/stats

F) Install ShearK-Miner 2.5

Release zip: github.com/rgsneddon/ShearK/releases/tag/2.5

Selftest (must pass, must not start hashing):

Code:

ShearK-Miner --selftest

Expected ShearHash-v3 light digest (2.5):

Code:

98818c31d739ef821db0242f76bd244b96f1fb5049d27ea9a192e95c67b39a8b

Do not use Shear-Miner 1.x / ShearK below 2.5 on this book.
Windows: Unblock / allow if Defender quarantines unsigned PE.


G) Point ShearK at YOUR solo stratum

Replace YOUR_SSA1 with Continuum Copy dest.

Linux / macOS:

Code:

./ShearK-Miner --pool 127.0.0.1:1111 --user YOUR_SSA1.solo --backend jit-full --threads 8

Windows:

Code:

ShearK-Miner.exe --pool 127.0.0.1:1111 --user YOUR_SSA1.solo --backend jit-full --threads 8

Notes:

• --pool MUST be 127.0.0.1:1111 for solo (not pool.shear.digital).
  
• Login is ssa1….worker (e.g. .solo).
  
• If you only have she1: --user she1….solo --dest YOUR_SSA1
  
• --backend jit-full = 2 GiB dataset path (recommended). --backend jit = 128 MiB cache.
  
• Rest-frame shear1 is refused as a login.
  

Share accepts are normal. When you seal a block you should see:

Code:

BLOCKFOUND!!!  *** SHEAR BLOCK SEALED ***  ... blocks=N ...

blocks=0 on share accepts alone is normal until a real seal.


H) Rewards & Continuum

• Solo has no 1% pool fee.
• Pot + hash bonus mint in coinbase; Continuum shows spendable after 6 confirmations.
  
• Keep / reopen the wallet so Shearview can collate landings.

I) Common failures

• need_dest — used she1 without --dest, or not an ssa1. Use Copy dest.
  
• stale / dropped — job restamp; not the same as Rejected.
  
• Continuum has SHE but miner blocks=0 — often prior pool mining, not this solo session. Solo seal must print BLOCKFOUND!!!
  
• IBD stuck — check seeds, outbound :30303, disk, magic shear-testnet-v4.
  
• Native build fails — need a real C++ toolchain. Never copy a macOS .node onto Linux/Windows.
  
• --pool pool.shear.digital:1111 — that is not solo.
  

J) Minimal copy-paste (Linux)

Code:

sudo apt-get update && sudo apt-get install -y git build-essential python3
# install Node >= 18, then:

git clone https://github.com/rgsneddon/shear-testnet.git
cd shear-testnet && npm install && npm run build:admit

# terminal 1 — node + solo stratum
npm run solo

# terminal 2 — after IBD catches tip; ssa1 from Continuum Copy dest
./ShearK-Miner --pool 127.0.0.1:1111 --user ssa1YOURDEST.solo --backend jit-full --threads $(nproc)

K) Safety

• Official links only: shear.digital / github.com/rgsneddon/shear-testnet / github.com/rgsneddon/ShearK
• Never paste seed, password, shewall.bin, or shear1 into Discord/Telegram "support"
  
• Testnet coins can vanish. Not financial advice.
• Mainnet is not live. Do not set SHEAR_MAINNET_EMIT.
  

she is private.
Solo = tip node + ShearK → 127.0.0.1:1111

[rpAI]

DINS-DAG and the Private Mining Corp

Mining should feel like joining a company that pays you for real work — not like hoping you were on the lucky template when the block closed.

That is the idea behind DINS-DAG, and the reason we talk about Shear's mining layer as a Private Mining Corp.

The problem we are fixing

On many pool-based chains, your shares live in one operator's memory. If another template wins the seal race, honest work can evaporate as "lost round." The pool that found the block gets to decide what counted. Community miners feel that as AFK unfairness: you hashed, you were eligible… then the books shrugged.

Shear already pays a scheduled block pot and a hash bonus for proven, destination-bound work. DINS is how we make that bonus behave like network law, not pool luck.

What DINS-DAG is

DINS stands for Deterministic Inclusive Network Settlement.
Your proven, destination-bound shares enter a shared network window everyone can re-check — not a private pool list that dies when someone else seals. Settlement is deterministic (any full node recomputes the same payroll) and inclusive (eligible in-window work still counts when a competing template wins the seal race).

DAG stands for Directed Acyclic Graph.
Here that means a share-DAG: mining proofs link as a growing graph of work on top of Shear's linear block spine (the familiar ~1 SHE per block continuity). Tips advance. At seal, the chain commits a deterministic "blue" view of that graph. Any honest seal of the window settles the same eligible work. Competing seals stop being a reason to zero people who already proved in-window hashes.

So DINS-DAG = fair, consensus-ruled mining settlement, organised as a share graph instead of a single pool's forgotten round.

In plain terms:

• You mine to your payout destination.
  
• Your work gets a network-visible receipt in the window.
• When a block seals, settlement is recomputable by any full node — no trusted "pool said so" oracle.
  
• Losing the template race should not erase eligible work that was already in the window.

The pool still matters. Stratum, difficulty smoothing, the HUD, Continuum — that operator UX stays. What moves is mint authority for hash bonus: from "this pool's open round at seal" toward consensus settlement from the share-DAG.

Why "Private Mining Corp"

Think of the network as a tightly run co-op that only admits cryptographically proven labour:

• Membership is your dest-bound proof, not a KYC form.
• The "payroll" is deterministic settlement at seal — Window credit becomes sealed landings, then spendable after confirmations.
• Books stay readable: Window → Sealed → Spendable, with your own blue detail in Continuum when you want it.
  
• The full network graph belongs on surfaces like dag.shear.digital; your wallet shows your books, not everyone else's leaves.
  

Private, because payouts stay tied to Shear's confidential note model and dest-bound proofs — not a public scoreboard of who owes whom in plaintext. A corp, because the rules are shared, capped, and boringly fair: same emission schedule, clearer who gets paid when the block lands.

What we are not claiming

This is the product direction we are building toward on Shear's testnet stack — design-hardened, spike-first, fail-closed. We are not promising a magic APY, an uncle-block casino, or a second full blockchain. Emission stays honest; DINS chooses eligibility, it does not print extra SHE for vibes.

No oracle mint. No double-pay paths. Caps so the graph cannot DoS the chain. Privacy hard lines stay: the window is not a place to dump raw addresses or open amounts.

Why it is worth the enthusiasm

If DINS-DAG lands the way we intend, Shear mining feels less like gambling on which pool won the race, and more like holding equity in a private corp that actually books your shift.

Hash. Prove. Land in the window. Seal. Get paid under rules everyone can re-run.

That is the Private Mining Corp. That is DINS-DAG.

— Shear

[rpAI]

Good evening everyone.

Work has been done to build the DINS-DAG and it is now waiting on enablement.

The fingerprint has ADMITv2, DINS-DAG, 1HASH=1TX and FULL MINER PRIVACY.

Shear includes a Full Community Governance Suite with moderator admin portal 'Mishi' - a built-into-fleet VPN to mask your IP when using a fast-sync wallet and Deterministic In-Node Shares (DINS) to fully spray tx data in fragmented bits across the full flux-set of membership notes, ensuring ultimate privacy in a hybrid linear-DAG Proof of Work architecture. PREPPED.

0.51 Node. 0.51 Wallet. 2.6 ShearK Miner
https://shear.digital/

[rpAI]

Last step done before Mainnet prep.

Chain. Testnet-v5 was reset and is producing.
Pool and the operator node fleet sit on the same tip: height, hash, ibd=false, miners on stratum 1111, sequential ingest 1 → 2 → … with full seal checks (ADMITv2 + DINS-DAG pot and hash bonus). Empty-batch genesis seals; hash bonus is share-batch units only.

Privacy. The book only records ssa1 dests. Continuum accepts she1, ssa1, and shear1 in To: full she1 becomes a one-time silent dest; ssa1 is itself; your own shear1 settles on destCommit (the spendable mailbox). Short she1 fingerprints stay unpaid. Rest-frame strings never go on chain.

Balances. Spendable is reconstructOwner on sealed notes after 6 confirmations.

Pending/unconfirmed is the pie until then. Owed-π is not spendable. Combined Continuum figure is the sum of those dest notes, so you can receive on any of the three identities and spend what the node actually verifies.

NODEv7 / Continuum 0.56. Source on main matches this chain. Users need a fresh node datadir.

@everyone - we have reset the chain again, so you will need to delete your old wallet versions and install 0.56 fresh.

NODE: https://github.com/rgsneddon/shear-testnet/releases/tag/v7

WALLET: https://github.com/rgsneddon/shear-testnet/releases/tag/0.56

[rpAI]

We will begin a 72hr soak test today on shear-testnet-v11 (pending genesis)

SHEAR-TESTNET-V11 — 72h SOAK WATCH LIST

STAMP AT GENESIS LIVE
- Start time (BST), genesis/tip SHA, height-0 baseline; planned end = start + 72h

BOOK / RELEASES
- Magic/network: shear-testnet-v11
- Continuum 0.71 + node v18 fully aligned
- ShearK 2.8 stays (no new miner cut unless hash/stratum job changes — already checked)
- GENESIS_BITS 17; ASERT_TAU_MS ≈ 1,440,000; median11 + ±1
- MAX_SHARES_PER_BLOCK 65536 (direct increase from the prior 8192); SHARE_FLOOR_BITS 8 (256 units/share); HASH_BONUS_NANOS 1
- Lean live: P2P frame carries full 65k; packed share bodies on wire+disk; no JSON~114B/share primary; no full-chain prune rewrite
- Pool fee dest: ssa1q5495…d8jj (stats feeDestTail only); mining dest later (Russell)
- Checkpoints/vault seal: first 1000, then every 400 (unchanged)

JUDGE
- From-start mean vs 90s after FULL 72h (not short windows)
- Early stretch under 17 bits must not replay v10 soft sprint (~59s first-100)

TIMING / ASERT
- Rolling / last-hour / last-N / median vs mean; τ close speed; ±1 behaviour

SHARES / 65k / LEAN
- Shares/block p50/p90/max + % at cap
- Verify ms/block (flag if p99 ≫ few seconds); body size; relay; orphans
- Pool collate / stratum lag; IBD/peer stalls; small-host memory / peak RSS
- Truncation fairness if cap binds (hash-share weight, not low-dest bias)
- Reject spikes: dup_share, hash_bonus, seal_failed, other share-law failures
- Bonus vs pot; small-miner PROP variance; stall-gap share counts

DISK / PRUNE (chronoflux degeneracy)
- Datadir du + chain.bin/(jsonl) breakdown; MB/block, GB/day
- Pre- vs post-sample-prune (~1000 conf) delta; samples shrink, money/hash spine stays
- Confirm segmented prune (not full-history rewrite)

PRIVACY / WALLET / SEND
- Bare ssa1 must NOT return balances/notes/history publicly
- Continuum send via node (pool not send authority)
- Spendable paints SYNCING→CONNECTED; no false 0 SHE when notes can still open
- Empty new fee mailbox OK until fees land

POOL / P2P
- /api/stats uptime; pool memory (no 1GB freeze class)
- Multi-host tip height/hash agreement
- skipFlow 0% under prune floor; zero hash-vout mismatches

HOLD MID-SOAK
- No τ/bits/max-shares/encode flip; no pool bounce/chain reset; no UI redesign / fee-emission rewrite without Russell GO

ON COMPLETE
- Full post-soak pack + recommended action list.

[rpAI]

Shear update: new dates for the v12 testnet and mainnet

Hi all,

A quick update on where we are and when things are landing.

The short version: mainnet is moving from 16th November to Monday 21st December 2026 at 2pm UK time. The v12 public testnet is now expected in early December.

Why the change? While hardening v12 we found and fixed more than we expected, and one of the things we found was serious enough that we want it fixed, tested and running on testnet for a good while before mainnet. We would much rather ship a chain that works properly than hit a date. The new date gives the v12 testnet at least a week of real use, and some room to spare, before mainnet goes live.

Here's the order things happen in:

1. Finish the remaining work below (through November)
2. Run a fast-block private test network, where we use every feature on desktop and Android and fix anything that breaks (from around 27th November)
3. Launch the v12 public testnet (around 3rd December)
4. At least 7 days of the v12 testnet running under real use (to around 10th December)
5. Mainnet genesis: Monday 21st December 2026, 2pm UK time

What's still being worked on, with the date each piece should be done:

- Network and block-integrity hardening, including a fix that makes sure every block commits to exactly the transactions inside it: around 28th October
- Vault lock safety, so a lock can only ever be withdrawn by its owner: around 30th October
- Command-line wallet fix for Vault locks: around 31st October
- Mining hash safety, so the native hasher can't crash and works on older CPUs: around 2nd November
- Correct network settings everywhere, so testnet, mainnet and dev networks can never get mixed up: around 4th November
- Signed name registration: around 5th November
- Stricter checking of data arriving from the network: around 7th November
- Fairer share credit for miners: around 8th November
- The private miner payout: around 9th November
- Extra privacy hardening for spends: around 10th November
- Moving our tests over to real chain data for every consensus path: around 13th November
- Private tokens. Anyone will be able to create their own private token, with a fixed supply that can never be added to and the same privacy as SHE. The design is written and reviewed first, then built: around 19th November
- The rest of the test suite moved over to real chain data: around 21st November
- Bringing five older pieces of work up to date (gas limits, reward claims, withdraw rules, Vault fees in the wallet, and the launch checklist): around 25th November
- Clearer wording and polish across the wallet: around 27th November

These dates are our best estimate, not promises. Each step gets reviewed and tested properly before it lands, and if testing turns up a problem we fix it first. If anything moves, we'll tell you.

Thanks for sticking with us. Every one of these steps is about making sure Shear is solid on day one.

The Shear team

[rpAI]

Shear (SHE) dev update — 30 Oct

• Privacy core (ADMITv3): full-chain membership proof. Every spend hides among every coin on the chain. The generator set is built and independently reviewed; the final consensus binding is in full review now. We're not rushing this one.

• Mining privacy: pool and solo payouts are being redesigned so payees and amounts stay hidden, with unlinkable per-share claims.

• Wallet: privacy hardening in progress, and the balance you see is always what you can actually spend.

• Testnet v12 is next once these land and pass audit; mainnet follows.

• Genesis target: 21 Dec 2026, a target, not a deadline. No gate is skipped to hit it.

• No premine, no dev fee, no coinbase tax. Community owned, funded only by grants, donations and crowdfunding.

Privacy is non-negotiable. It ships when it's right.