Testnet4 Mining (Discussion)

[BayAreaCoins]

In general, I agree that giving large amounts is not right.

I think the best way to give larger amount of Testnet is with a deposit that is refunded upon return.

It could be fairly easily automated... perhaps that is less greasy and still fills the needs of people who need (or want) more than dust.

I give 0.005 TBTC every ten minutes on our faucet... which I think is more than enough.

It gives people dust for free without being greasy or battling dirty () testnet miners.

[BayAreaCoins]

I found a public v4 pool:  https://testnet4.btclab.dev/#/

10% fee, which I think is fair.

I'm probably going to play with this tomorrow. (I've not tested it)

[BayAreaCoins]

I found a public v4 pool:  https://testnet4.btclab.dev/#/

10% fee, which I think is fair.

I'm probably going to play with this tomorrow. (I've not tested it)

I've now used this pool for a few days... it does pretty good.

The CPU miners are still annoying, but this is a way to at least get coins without using faucet or trading.

I'm hitting the network with 1.5 PH and due to the CPU fuckery, I'm only hitting around 5% of produced blocks. 

[LoyceV]

(Great estimation!)

That's because the spambot copied it from mikeywith's post.

[BayAreaCoins]

(Great estimation!)

That's because the spambot copied it from mikeywith's post.

God, that would be a thing now days... my bad, I thought I got caught again

[stwenhao]

I'm only hitting around 5% of produced blocks.

Of course. If you mine zero blocks with CPU difficulty, then you will only raise the network difficulty. In general, only 1/6th of all blocks have ASIC difficulty. And it is still adjusted every 2016 blocks. Which means, that if you will produce one block per 10 minutes, then the presence of CPU blocks will raise the difficulty again.

Also note, that when you make any block, then you can put any timestamp you want, in a consensus-acceptable range. Which means, that if you put the current date and time, then CPU miners can immediately produce blocks, up to 2 hours in the future, which means, that you will instantly see 6 new blocks with CPU-difficulty. However, if you put your timestamp 1 hour 59 minutes and 59 seconds forward, then no new CPU-mined block will appear for the next 20 minutes.

Another thing is that by having an ASIC, you can just reorg all CPU-mined blocks, and decrease the difficulty in that way. As long as you can see any chain of CPU-only blocks, up to the chain tip, you can replace it with a single ASIC block, and all of them will be reorged.

Code changes to throw away CPU-mined blocks: https://github.com/bitcoin/bitcoin/pull/31117

[LoyceV]

Another thing is that by having an ASIC, you can just reorg all CPU-mined blocks, and decrease the difficulty in that way. As long as you can see any chain of CPU-only blocks, up to the chain tip, you can replace it with a single ASIC block, and all of them will be reorged.

Code changes to throw away CPU-mined blocks: https://github.com/bitcoin/bitcoin/pull/31117

Do you know what happened to the plan to create a pool for this, so anyone with an ASIC can easily wipe out CPU blocks?

[stwenhao]

Do you know what happened to the plan to create a pool for this

As you can see, the code on GitHub is in "draft" status, so you have to compile it yourself, if you want to use it. Or you can ask "fjahr", what is the plan (but I guess if some people from the mailing list are talking about hard-forking testnet4, and removing CPU mining from consensus rules, then they don't have any plans to develop testnet4 further).

[BayAreaCoins]

Do you know what happened to the plan to create a pool for this

As you can see, the code on GitHub is in "draft" status, so you have to compile it yourself, if you want to use it. Or you can ask "fjahr", what is the plan (but I guess if some people from the mailing list are talking about hard-forking testnet4, and removing CPU mining from consensus rules, then they don't have any plans to develop testnet4 further).

Why would they bother to remove the CPU mining if they don't have plans to develop Testnet4 further?  Why bother?

Also, your last email on the mailing list had some really good points.  Probably the best quality email I've seen regarding Testnet in 2025.

From: Saint Wenhao
May 6, 2025, 10:25:25 AM (yesterday)
to Greg Maxwell, Bitcoin Development Mailing List

(> = Greg Maxwell quotes)

> Some believed that resetting alone would discourage the conduct,

I wonder, why many people keep saying "reset", when in practice, the old network is just "abandoned", and nothing is really "resetted". We would have a true reset, if a new testnet would lead to the full chain reorganization of the old chain. And this thing never happened, all previous testnets were just abandoned, as they were, and there were just not enough people, willing to keep old networks alive.


> Hard fork in an ultramassive premine

Well, if testnet4 will be "fixed" by doing a hard-fork on top of existing chain, then it would be equivalent to releasing testnet5 with "premine" of all testnet4 blocks.


> so maybe an additional 21 million testnet btc?

Using doublings, instead of halvings, is future-proof. Because then, if someone will continue using such test network forever, then later, it would release 21 million coins per block (which would be roughly equivalent to having a superchain, where you can peg one altcoin per 10 minutes).


> if someone starts trading tnbct for money, whomever has the premine wallet should feel feel to sell into it what they can

In practice, trading will start, when the first version of the official wallet will be released. It was the case in testnet4: when it was possible to mine new coins, only when you created your own binary from the source code, then there were only devs, and some skilled users. However, when it was officially supported in a new Core version, then it was quickly listed on centralized exchanges, and a lot of people dumped all of their test coins, and walked away with real BTCs.


> the custodian of the premine will use it for the betterment of Bitcoin development,  but if it's found that they're not-- solution is just another reset

Which means, that Bitcoin developers will turn into altcoin developers, where they will release new altcoins called "testnets", only to dump all of them, and get BTCs out of it.

I know you've seen my thread regarding Bitcoin Testnet actually being a Altcoin.

I'm glad that you are at least open-eyed enough to grasp the concept.  Now you are seeing the true "Shitty premined make devs rich" side of the equation that has *always* been there... (I have a good view at who is dumping these coins into the market.  It's not folks that bought them on Coinbase!!!  It should be noted that I do not think this is a bad thing.)

The only thing I will say is that the part of "In practice" is not set in stone at all.  I am applying Gamesmanship to this network & our business plan in order to protect ourselves and our customers + what I think is best for Bitcoin. (Fuck testnet, but I firmly believe in its purpose and benefiting Bitcoin.)

We may take a totally different approach for this next round!  Part of testing is doing things a differently to see what works best.  I sit around all day and think about this.  V4 was the most boring and straight forward approach... I can think of several other approaches that likely would have worked better in hindsight... Futures?!?! Maybe we will delist v4 after people put work in on v5?! Maybe we list both forks of v4 or only one or maybe none! (Lol). Maybe we do everything the same?  Maybe we don't list it at all?  This maybe list can go on and on... but one thing is for certain, I will be here.

It keeps people on their toes!   Gotta make those Premine shitcoin devs at least work for it <3. I thought their premine in Testnet 4 was enough!

[stwenhao]

Why would they bother to remove the CPU mining if they don't have plans to develop Testnet4 further?

Because if you have any hard-fork, which is strictly incompatible with the previous version, then it is very similar to the situation, when you just make a completely different network.

So, if they will fix testnet4 in a soft-fork, then all new changes can stay in the same network, if hashrate majority will support them. But if they make backward-incompatible changes (like doublings, instead of halvings, or using the ASIC difficulty in the "bits" field from the block header, even if 20 minutes passed, and old nodes are enforcing the CPU-difficulty blocks), then you have new, incompatible rules, and some nodes will enforce the old rules, and others will not.

To sum up: soft-forks will keep you in the same network. Hard-forks will split it between old and new nodes.

You may not understand it, if you saw altcoins, which did hard-forks on a regular basis, but technically, each hard-fork is roughly equivalent to releasing a new altcoin, which is incompatible with the old version. So, even if changes to testnet4 will be made on the same chain, then if they will be incompatible with existing nodes, I will consider it as "testnet5", because already deployed nodes don't have to update their code. It is P2P world: you are never forced to upgrade.

[BayAreaCoins]

Why would they bother to remove the CPU mining if they don't have plans to develop Testnet4 further?

Because if you have any hard-fork, which is strictly incompatible with the previous version, then it is very similar to the situation, when you just make a completely different network.

So, if they will fix testnet4 in a soft-fork, then all new changes can stay in the same network, if hashrate majority will support them. But if they make backward-incompatible changes (like doublings, instead of halvings, or using the ASIC difficulty in the "bits" field from the block header, even if 20 minutes passed, and old nodes are enforcing the CPU-difficulty blocks), then you have new, incompatible rules, and some nodes will enforce the old rules, and others will not.

To sum up: soft-forks will keep you in the same network. Hard-forks will split it between old and new nodes.

You may not understand it, if you saw altcoins, which did hard-forks on a regular basis, but technically, each hard-fork is roughly equivalent to releasing a new altcoin, which is incompatible with the old version. So, even if changes to testnet4 will be made on the same chain, then if they will be incompatible with existing nodes, I will consider it as "testnet5", because already deployed nodes don't have to update their code. It is P2P world: you are never forced to upgrade.

Our first version of "FreeBitcoins.com" was helping people harvest forks and CLAMS (a blockchain snapshot airdropped to all BTC, DOGE, LTC people, my current partner created CLAMS).  Helping people dump their forks without downloading clients for 10%. https://medium.com/freebitcoins-com/freebitcoins-com-2951cd5fa0a8

I would definitely consider the new fork Testnet 5 as well.

you are never forced to upgrade.

People and exchanges can always run both.  I'm mining Bitcoin Cash when my mining power isn't rented.  (Dump the blocks and then do my thing in Bitcoin)

I'm really excited to see normal POW in Testnet. It's a great update for Testnet IMO.  It's going to help prepare a lot of people & products for mainnet in a good way.

If you have Testnet coins on AltQuick... be damn sure to get them off before the fork. 

Does AltQuick.com credit forks?
We do not credit forks to users that hold a balance with us. Please withdraw your balance so that you hold your own keys to claim forks.

https://altquick.com/exchange/faq

Proof of Work puzzle in testnet4: https://mempool.space/testnet4/tx/cc159432ffb7a166abeccc79800e9616a09ea9ac6937080c2ca37b38671970e5

I noticed the puzzle you are promoting are coins that were bought from my exchange.  Even the left over coins in the puzzle transaction are sent back to the exchange directly in that transaction. 

That makes me really happy.

I don't see how someone can think what I'm helping facilitate is a bad thing.

Unsurprisingly, it's one of the leading commentators on the Bitcoin Mailing list and they are day trading testnet coins... lol *facepalms*

Garlo Nicon

[moderator's note: consecutive posts merged]

[stwenhao]

I noticed the puzzle you are promoting

Note that I don't know the solution upfront. It is a Proof-of-Work-based puzzle, where you have to mine the solution, by producing smaller and smaller signatures. Which means, that I cannot just sweep all coins, by importing some kind of wallet: if I would want to take any coins out of it, I would have to prepare some transaction, and then run a miner, which will grind signatures, until it will find some match.

Garlo Nicon

Of course. I got these coins from him, and I sent them back to him. Maybe he gave me some address from your exchange?

I don't see how someone can think what I'm helping facilitate is a bad thing.

Because test coins were supposed to be worthless, like for example regtest coins are. And your exchange just raised their value, and pushed some people to sell their test coins.

Imagine that you play a monopoly game with a group of friends. If some stranger will suddenly join, and pay you real money for your monopoly money, wouldn't you sell them?

Some people may consider it "a bad thing", just because when they started playing monopoly, they assumed, that no real funds are involved. Everyone assumed, that it is just a game, where you can just play, and have some fun. And when fake coins were made real, then it simply introduced value into something, which was considered "worthless", which defeated the purpose, why some people joined the game in the first place.

And now, developers are trying to make a new monopoly game, where coins will be really worthless, and no stranger will suddenly join, and start trading monopoly coins for real dollars. When you allow trading everything they release, then they don't want to play anymore, because then it is harder to test some real attacks, when funds are not worthless. The original purpose of test networks, was to make something, where you can check some attacks, without harming the main network. By raising value of test coins, you just make it harder, because then, there is no longer any good, decentralized chain, which can be safely attacked, and tested more seriously.

Edit:

Please do these tests on the test network.  That's what it's for.  Thanks.

See? You can read the whole topic, to have a context, how test networks were used in the past. The purpose was quite simple: you read the original code, and thought about some kind of attack. And then, when you wanted to check, if it really works or not, you didn't want to break the original, main network, with real coins. You needed some kind of environment, where you can break something, and nobody will blame you for doing that.

And now, when you bring value into test coins, you simply break that kind of use case. And currently, there is no good alternative network, which could be used instead, when testnet3 and testnet4 coins are traded.

[BayAreaCoins]

Imagine that you play a monopoly game with a group of friends. If some stranger will suddenly join, and pay you real money for your monopoly money, wouldn't you sell them?

Some people may consider it "a bad thing", just because when they started playing monopoly, they assumed, that no real funds are involved. Everyone assumed, that it is just a game, where you can just play, and have some fun. And when fake coins were made real, then it simply introduced value into something, which was considered "worthless", which defeated the purpose, why some people joined the game in the first place.

And now, developers are trying to make a new monopoly game, where coins will be really worthless, and no stranger will suddenly join, and start trading monopoly coins for real dollars. When you allow trading everything they release, then they don't want to play anymore, because then it is harder to test some real attacks, when funds are not worthless. The original purpose of test networks, was to make something, where you can check some attacks, without harming the main network. By raising value of test coins, you just make it harder, because then, there is no longer any good, decentralized chain, which can be safely attacked, and tested more seriously.

https://www.ccn.com/gambit-shuts-bitcoin-board-gaming/

It's more easy for me to imagine myself playing Monopoly for money and mopping people so hard that the website has to shut it down.  lol!

I'm pretty sure I hold a world record for playing a 10-15 BTC game of Rock Paper Scissors (best of 5?) and I won.  Playing whole Bitcoin Monopoly games was pretty normal for a group of us (we totally played each other, that site owner was a pussy.)

"Sharks" are always going to eat folks lunch.  For some reason, news articles just call me a shark and don't bother to list my name... it happened here too:  https://www.dailydot.com/debug/dogecoin-mining-dogehouse-ddos/

So I guess it just depends on what kind of game you're playing... if it's a decentralized game that you promote as permissionless borderless money, your playing with fake money, and someone show up with real cash to play against your fake bills... I'd be pretty god damn happy!  I would not be trying to kick the retard out with real money... he's buying property and playing by the rules otherwise except his bills are worth something after the game.

Even during a game of normal Monopoly.  Monopoly money still have "value" during the game.  After the game it becomes worthless.  This is yet another reason that Testnet needs a firm end date, so people know when the game is "done".


Also, regarding your post from Satoshi, the post right under it from Theymos highlights the importance of having a robust testing network.  

Please do these tests on the test network.  That's what it's for.  Thanks.

You can't get real data from such a small network. Besides, a reliable network must be able to withstand any kind of testing.

Issues and bugs are more likely to be found or used in a network that is being used.  This has been shown to be true in both v3 and v4.  

It's also much easier to make fun puzzles like in your signature!

Bitcoin was *so* small back then.  Bitcoin was trading for around 0.01-0.1 cents USD at the time of that post.

The times have changed.  The big issue was annoying Bitcoin by testing it back then... Now it isn't annoying (atm  ), it is expensive, but being expensive has become a use case for Bitcoin... sooooo.....  now we are trying:

The goal (my goal, at least) is to have some working distributed monetary system, not to keep a (potentially) broken system online without interruption.

It seems to be working. V3 would have never been attacked by Lopp if they didn't need a fresh block reward for the company causing all this OP_return bullshit.  V4 would have been a broken mining system without us and now it's potentially getting upgraded to what it should be... like Bitcoin.  2 for 2 thus far my dude.


Bitcoin Testnet won't cross $100 a coin, IMO.  $100 is perfect for a fuck around network.

[BayAreaCoins]

Basically, if the code allows for any mining fuckery, someone is going to hit it.

So breaking transactions to prove a point?

Submitting blocks with zero transactions is faster than including transactions.

[Sledge0001]

Submitting blocks with 0 data / or only 1 transaction maybe faster.....

But ends up being essentially useless for real dev imho....

Here's doing it right...

https://mempool.space/testnet4/block/000000000000000401af6aef678e8a9b780bcdfc9b0722370c9cbde345d76a9c

https://mempool.space/testnet4/block/00000000000000044d644ebd60091e96670a3175fea617e21dd476479a60d252

https://mempool.space/testnet4/block/0000000000000004d23b1ef0525014d34d95b003677686716e7b11d36d338b86

https://mempool.space/testnet4/block/0000000000000003827b15101ba8321e7e9cc8f5e335a1b5f40e3b1bbeb64379

[BayAreaCoins]

Submitting blocks with 0 data / or only 1 transaction maybe faster.....

But ends up being essentially useless for real dev imho....

Here's doing it right...

https://mempool.space/testnet4/block/000000000000000401af6aef678e8a9b780bcdfc9b0722370c9cbde345d76a9c

https://mempool.space/testnet4/block/00000000000000044d644ebd60091e96670a3175fea617e21dd476479a60d252

https://mempool.space/testnet4/block/0000000000000004d23b1ef0525014d34d95b003677686716e7b11d36d338b86

https://mempool.space/testnet4/block/0000000000000003827b15101ba8321e7e9cc8f5e335a1b5f40e3b1bbeb64379

All is fair in love and crypto mining.

Early Jan 2026 (edit note: fixed 2025 typo) the 20 min reset should be removed and Testnet won't have this issue any further.  Full POW is clearly the way, Testnet has pool activity now too, which is far different than back in the day.  

I'll definitely mine with y'all if it is ever open to the public or if you need a tester for some blocks.

Right now, I'm on BTClab.dev, paying a 10% fee.

My blocks include transactions as well:  https://mempool.space/testnet4/block/0000000000000002cedaf2784715288a9b3765e71d2c6c6c4b9c040b092a1e5d

[Sledge0001]

Submitting blocks with 0 data / or only 1 transaction maybe faster.....

But ends up being essentially useless for real dev imho....

Here's doing it right...

https://mempool.space/testnet4/block/000000000000000401af6aef678e8a9b780bcdfc9b0722370c9cbde345d76a9c

https://mempool.space/testnet4/block/00000000000000044d644ebd60091e96670a3175fea617e21dd476479a60d252

https://mempool.space/testnet4/block/0000000000000004d23b1ef0525014d34d95b003677686716e7b11d36d338b86

https://mempool.space/testnet4/block/0000000000000003827b15101ba8321e7e9cc8f5e335a1b5f40e3b1bbeb64379

All is fair in love and crypto mining.

Early Jan 2025 the 20 min reset should be removed and Testnet won't have this issue any further.  Full POW is clearly the way, Testnet has pool activity now too, which is far different than back in the day.  

I'll definitely mine with y'all if it is ever open to the public or if you need a tester for some blocks.

Right now, I'm on BTClab.dev, paying a 10% fee.

My blocks include transactions as well:  https://mempool.space/testnet4/block/0000000000000002cedaf2784715288a9b3765e71d2c6c6c4b9c040b092a1e5d

Soon. We are testing all aspects of block creation with the most recent version of core on testnets 3 and 4 and regnet.

When we launch (not too far away in public testing) we will have a fully unique coded Mainnet, Testnet 3 and Testnet 4 pool.

[LoyceV]

Early Jan 2025 the 20 min reset should be removed

That didn't happen. Unless you meant 2026.

Right now, I'm on BTClab.dev, paying a 10% fee.

My blocks include transactions as well:  https://mempool.space/testnet4/block/0000000000000002cedaf2784715288a9b3765e71d2c6c6c4b9c040b092a1e5d

If you're using an ASIC, why don't you wipe out those 8 CPU blocks before your block? Get BTClab to up their game a bit!

[BayAreaCoins]

Early Jan 2025 the 20 min reset should be removed

That didn't happen. Unless you meant 2026.

My bad, thanks Loyce, 2026*.

If you're using an ASIC, why don't you wipe out those 8 CPU blocks before your block? Get BTClab to up their game a bit!

Haha I don't think many pools want to do that.  If one does, it's going to be a game changing lol.

Soon. We are testing all aspects of block creation with the most recent version of core on testnets 3 and 4 and regnet.

When we launch (not too far away in public testing) we will have a fully unique coded Mainnet, Testnet 3 and Testnet 4 pool.

That's awesome.  Looking forward to that.