DEV Community: StoneVast

I withdrew $0.24 of mined crypto to an exchange — and it was the hardest part of the whole project

StoneVast — Mon, 19 Oct 2026 14:08:34 +0000

I withdrew $0.24 of mined crypto to an exchange — and it was the hardest part of the whole project

Mining was the easy part. Moving the coins out took a 3.5-year chain sync, a privacy-coin paradox, and one memo field that almost ate my money.

Context

I run a transparent mining workbench: one RTX 4070 Laptop, lolMiner, FishHash algorithm, mining IRON (Iron Fish) on HeroMiners pool. ~16 MH/s at ~49W. Daily net profit after electricity: about $0.075. Yes, seven and a half cents.

I publish every number publicly. Part 1 of this series covered the brutal economics. This post is about what happened when I tried to actually withdraw — the part most mining posts conveniently skip.

Step 1: The wallet that shows 0

My pool dashboard said it had paid me 1.8+ IRON. My Iron Fish CLI wallet said: balance 0.

Reason: Iron Fish is a privacy chain. Your wallet doesn't "query" your balance — it scans the chain and decrypts notes belonging to your keys. My node was 3.5 years of blocks behind. No sync, no notes, no balance.

Fix was boring but long: official chain snapshot + ironfish start --wallet --workers=-1, then wait for the wallet scan. Final result: 3.09185719 IRON visible.

Lesson: with privacy coins, "the explorer says you got paid" is not a thing. The explorer doesn't know who you are. Only your synced wallet knows.

Step 2: The memo that almost ate the deposit

Gate.io, like most exchanges, uses one shared IRON deposit address for all users, distinguishing accounts by a required memo field.

Iron Fish CLI transfers support memos, but the flow isn't obvious. If you send to an exchange deposit address without the memo, your coins arrive at the exchange — credited to nobody. Recovery means a support ticket and prayers.

I sent a 0.2 IRON test first. It confirmed on-chain (block 1846578) and showed up in the Gate app. Then I sent the remaining 3.0 IRON (block 1846594). Both arrived.

Test-first is the whole trick. The test cost me nothing but a few minutes; skipping it could have cost everything.

Step 3: Verify it yourself

Tx hashes — this is the point of the project, everything is checkable:

9281c0be78dfdb1d0196c832e4ea74afe80544d73b076aa1ecc6298ff335fb0a
c0092ac5633747806c0e222fac770ec81088add39d34557fb0db54a5240af23c

The privacy-coin paradox, in practice: the explorer can confirm these transactions exist and their status, but hides sender, receiver and amounts. So "proof" on a privacy chain is a layered thing — pool dashboard ↔ on-chain tx existence ↔ exchange credit. Three independent ledgers agreeing is the strongest verification available.

What $0.24 proves

3.2 IRON ≈ $0.24. Not coffee money. But the full loop now exists and is documented:

GPU → pool → wallet (synced) → exchange (memo'd) → sellable USDT

Every step reproducible, every number public. Dashboard, daily snapshots, withdrawal log with both tx hashes: https://haolinsong32-lang.github.io/stonevast/withdrawals.html

Honest FAQ

Q: Is laptop mining profitable?
A: No. $0.075/day net. My long-term hardware goal (an RTX 6000 Pro) would take ~30 years at this rate. I do it because a fully-transparent micro-rig generates data, and data compounds.

Q: Why Iron Fish?
A: FishHash is GPU-friendly (memory-hard DAG, ~4.6GB, Blake3 core, ASIC-resistant by design), and the privacy model makes the "prove everything" constraint interesting.

Q: Would you recommend the Iron Fish CLI flow?
A: It works, but budget an evening for first sync + wallet scan. And write down the memo requirement before your first exchange deposit, not after.

Part 1 (the economics): I mine IRON on a laptop and built a base-13 cipher — an honest numbers report

I Mine IRON on a Laptop and Built a Base-13 Cipher — An Honest Numbers Report

StoneVast — Mon, 19 Oct 2026 01:35:02 +0000

I Mine IRON on a Laptop and Built a Base-13 Cipher — An Honest Numbers Report

No hype. No "passive income" promises. Just a laptop GPU, real electricity bills, and every number verifiable on a public mining pool.

The honest numbers (updated daily)

Metric	Value
Hardware	RTX 4070 Laptop (8 GB)
Hashrate	14.5–16.7 MH/s (FishHash)
Power draw	~49 W
Electricity cost	¥0.55/kWh (~$0.076)
Daily net profit	~¥0.54 (~$0.075)
Total mined so far	1.8576 IRON ≈ ¥0.99

Yes, you read that right. About 7 US cents per day. At this rate, saving up for my dream card (an RTX 6000 Pro, ~¥70,000) would take roughly 355 years from mining alone.

I publish these numbers daily on my site — not to impress you, but because verifiable honesty is the entire point of the project.

Why mine at all, then?

Three reasons:

It's a live data source. Real hashrate, real payouts, real prices — a stream of ground truth I can build tools around.
It's a trust anchor. Anyone can check the pool dashboard and confirm I'm not making numbers up.
It's fun. And fun keeps the project alive longer than discipline does.

The side quest: a base-13 "mother codebook" cipher

While watching logs grow, I started wondering: could I design a tiny encoding scheme of my own? The idea:

Merge 26 letters into 13 symbols (a/b share one, c/d share one, ...)
Same trick for digits (0/1, 2/3, ...) and punctuation
The mapping table itself is the key — different seeds, different tables

Here's the core of it in JavaScript:

const DEFAULT_MASTER = {
  a:'α', b:'α', c:'β', d:'β', e:'γ', f:'γ', g:'δ', h:'δ', i:'ε', j:'ε',
  k:'ζ', l:'ζ', m:'η', n:'η', o:'θ', p:'θ', q:'ι', r:'ι', s:'κ', t:'κ',
  u:'λ', v:'λ', w:'μ', x:'μ', y:'ν', z:'ν',
  '0':'①','1':'①','2':'②','3':'②','4':'③','5':'③','6':'④','7':'④','8':'⑤','9':'⑤',
};

function encode(text, table = DEFAULT_MASTER) {
  return [...text].map(ch => table[ch.toLowerCase()] ?? ch).join('');
}

// 'stonevast' → 'κκθηγλακκ'

And a seed-derived version so everyone can generate their own codebook:

function deriveMaster(seed) {
  // SHA-256(seed) drives a Fisher-Yates shuffle of the target symbols
  // Same structure, different contents — the table IS the key
}

I also added an anti-phishing tail: 4 base-13 characters appended to every ciphertext, computed as SHA-256(cipher | masterKey) mod 28561. If someone tampers with a single character — or tries to forge a message without knowing the master — verification fails and the message is rejected:

correct key  → stonevast ✓
tampered     → ⚠ anti-phishing check failed, rejected
wrong key    → ⚠ anti-phishing check failed, rejected

Is this cryptographically serious? No. It's a learning toy — a Caesar-cipher-grade curiosity. Real systems should use AES/ChaCha20 and established libraries. But building it taught me more about key derivation, checksums, and why real ciphers are hard, than reading three articles would have.

What I'm actually building

The mining rig is the mascot; the real product is a mining revenue workbench: transparent dashboards, daily snapshots, risk calculators — every number checkable by anyone.

Live site: https://haolinsong32-lang.github.io/stonevast/
Node program (free tier, genesis perks): same link
Pool dashboard: search my worker sg4070 on HeroMiners

If you're into honest data, tiny ciphers, or laptop mining experiments — say hi in the comments. Tell me what you'd verify first.

All figures from public pool APIs and exchange tickers. Mining has electricity-cost risk; do your own math. This is a data journal, not investment advice.

I mined crypto for 40+ hours and made less than a dollar. I'm publishing every number anyway.

StoneVast — Sun, 18 Oct 2026 16:56:48 +0000

Subtitle: No lambo. No "passive income" pitch. Just a laptop GPU, a power meter, and three independent sources you can check yourself.

Most crypto mining content falls into two buckets: screenshots of impossible profits, or vague "it's not worth it anymore" takes. Both are useless if you actually want to decide whether to plug in your own GPU.

So I'm running a different kind of experiment: one RTX 4070 laptop, one Iron Fish miner, and a public dashboard where every number — down to the last $0.01 — is cross-verified against sources you can open in your own browser.

Here's everything I've learned, with receipts.

The honest numbers

Hardware: RTX 4070 Laptop (8GB), nothing fancy
Coin: Iron Fish (IRON), via the FishHash algorithm
Miner: lolMiner 1.98a on Windows
Pool: HeroMiners
Time mining so far: ~40+ hours (with interruptions I'll explain below)
Total paid out: ~2.5 IRON ≈ $0.19 at current prices
Net after electricity: roughly ¥0.5–0.8/day (~$0.07–0.11)

Yes. Less than a dollar. That's the real number, and it's the whole point: small is fine, as long as it's true and verifiable.

Every figure above can be checked against three independent mirrors:

The pool's own dashboard — HeroMiners shows my hashrate, shares (2,189 accepted, 0 rejected), and every payout to my address.
The Iron Fish block explorer — my address b88f8b...df80 is on-chain. The payments are real transactions anyone can look up.
The exchange price feed — IRON/USDT on Gate.io, the same live price my dashboard uses.

Pool ledger ↔ on-chain ledger ↔ exchange price. If any one of them disagrees, you'll see it.

The one tweak that gained me 25%: clock locking

Gaming laptops are terrible miners out of the box — not because of the GPU, but because of thermals. Boost clocks spike, the fans scream, the chassis heats up, and then everything throttles.

The fix was counterintuitive: lock the core clock lower.

Core clock	Hashrate	Power	Temp	Net/day
Unlocked (boosting)	~20 MH/s peak, unstable	50–60W	60–68°C	+¥0.58
Locked 1600 MHz	~15–18 MH/s	~46W	~55°C	+¥0.45
Locked 1200 MHz	~14.5 MH/s	~40W	~48°C	+¥0.31

I settled on 1600 MHz. On Windows with an NVIDIA card:

nvidia-smi -lgc 1600

That one command took me from 13.33 → 16.67 MH/s sustained (+25%), while dropping temps and fan noise. The machine now sits next to me, barely audible, mining through the night.

Note: lolMiner's own --cclk flag did nothing on this laptop GPU. nvidia-smi -lgc (needs admin) is what actually worked.

The breakeven math, published openly

Here's the part most mining content hides: the exact point where this stops making sense.

At my locked setting (14.5 MH/s @ 40W, ¥0.55/kWh electricity):

Breakeven coin price: IRON at $0.047 (currently ~$0.075 → 37% safety margin)
Breakeven electricity price: ¥0.88/kWh (~$0.12)

If IRON drops below $0.047, my watchdog script parks the miner automatically. Mining at a loss is a choice, not an accident.

If your electricity costs more than ~¥1.20/kWh (~$0.17), my own site will tell you not to bother — a hard "DO NOT RUN" verdict. I'd rather lose a visitor than have someone lose money on my numbers.

What killed my miner for 5 hours (a cautionary tale)

Confession: I lost 5 hours of mining because I forgot I was running a local AI image generator (ComfyUI) on the same GPU. FishHash needs ~4.6GB of VRAM for its DAG; my 8GB card had 183MB free. The miner silently failed.

Lesson: on an 8GB card, local AI and mining don't coexist. Pick one, or script the switching.

What I'm actually building

The mining itself earns pocket change. The real project is the layer on top:

A public, self-verifying dashboard — hashrate, power, payouts, breakeven points, all updated daily from the pool API and exchange feeds. No login. No ads.
A free "am I profitable?" checker — pick your GPU (20 presets, including AMD and datacenter cards), set your electricity price, get an honest RUN / MARGINAL / DO NOT RUN verdict in 20 seconds. It will happily tell you not to mine.
A node program — first 100 "genesis nodes" lock in 30% of their direct referrals' paid revenue, forever. One level only, no entry fee, no promised returns, monthly settlement ≥20 USDT paid on-chain. Genesis nodes also get a free numbered NFT (#001–#100) as an identity credential — explicitly not an investment.

It's all live, in Chinese and English, here:

http://47.236.149.136.nip.io (a proper domain is on the roadmap — funded by mining revenue, because of course it is)

Why publish numbers this small?

Because the alternative is what this industry usually does: hide the losses, crop the screenshots, and sell the dream.

I'd rather build the boring version — where a ¥0.83 profit is shown as ¥0.83, where the breakeven price is printed next to the current price, where the answer to "should I mine?" is sometimes a flat no.

If that sounds refreshing instead of disappointing, come watch the number tick up, one satoshi-sized drip at a time. First 100 nodes get the genesis terms permanently.

Every claim in this article can be verified from the dashboard itself: pool API, block explorer, exchange feed. That's the whole idea.