BTC/NMC merged mining available for testing

[nodemaster]

If you are brave and want to test merged mining on the most alpha status pool you have ever mined on have a look a this post: http://dot-bit.org/forum/viewtopic.php?f=5&t=217&p=1284#p1284  We need as much people testing and confirming the proper function of different miners as possible.

[1715156631]

1715156631

[1715156631]

1715156631

[1715156631]

1715156631

[TeraPool]

I am working on creating a bitcoin pool right now but would like to implement this feature, or at least get started testing it.

So there would be just 1 instance of pushpoold running? And 1 instance of namecoind running? Along with a proxy patch of some sort?

How would pushpool tell the mysql databases that it received a valid bitcoin solution or namecoin solution?

As it stands, pushpool simply reports "Y", "Y" for it's "upstream_result" and "our_result" mysql columns. So how would I know that I have mined a namecoin?

Perhaps from the "version" field of the valid solution?

[nodemaster]

So there would be just 1 instance of pushpoold running? And 1 instance of namecoind running? Along with a proxy patch of some sort?
 

You have a patched bitcoind (and its blockchain), a namecoind (and its blockchain) the merged-mining-proxy that know how to connect to both daemons and the pushpoold that connects to merged-mine-proxy.

 

How would pushpool tell the mysql databases that it received a valid bitcoin solution or namecoin solution?

As it stands, pushpool simply reports "Y", "Y" for it's "upstream_result" and "our_result" mysql columns. So how would I know that I have mined a namecoin?

Perhaps from the "version" field of the valid solution?

You don't have to rely on any information from the proxy. You could for example let pushpoold write the shares to the database and monitor the blockchains for generated blocks. Calculate how much valid shares each miner delivers between two blocks for each blockchain. Or you could patch both blockchain  daemons to deliver those information. Both is common amongst current pools

[Zibbo]

This would be akin to everyone being able to print their own money and trying to get others to accept it.  With a few select currencies, it's easy for them to succeed.  With thousands to choose from, people aren't going to accept every type of payment someone thinks up.

This is already true (and has always been), as anybody can start up their own blockchain. The challenge is still to try to get other people use it (not just mine it). The only difference with merged mining is that any blockchain can now be just as secure as bitcoin, but that doesn't make then as valuable.

It's like we are building this huge proof of work machine, and any future application that wants to use blockchains (and there will be many, not just currencies), can just tap in to existing hashing power. They don't even have to provide much (or at all) financial incentive for miners, as for them there is no downside in adding another blockchain to hash. It only needs to be for something a big part of miners would support for one reason or another, and you get your hashing power.

I see this as strengthening the whole idea of virtual currencies, as it kind of levels the playing field for everyone. Sure, it might eventually devalue BTC (or not), but only if some other virtual currency takes its place because it's thought to be better than BTC for some reason. I see no problem with that.

[JohnDoe]

They don't even have to provide much (or at all) financial incentive for miners, as for them there is no downside in adding another blockchain to hash. It only needs to be for something a big part of miners would support for one reason or another, and you get your hashing power.

The downside would be the disk space and bandwidth required to maintain new chains. If what one is able to mine isn't valuable enough to pay for those costs then most people would not support that chain with hashing power.

[enmaku]

Why pick an arbitrary block so far in the future to change namecoind? Why not work until development is through then just give a fair amount of notice that clients will need to be upgraded before date x or blocks might get rejected? Better yet just pick a block much closer to actual release date. 24k is a long way off...

[TeraPool]

Why pick an arbitrary block so far in the future to change namecoind? Why not work until development is through then just give a fair amount of notice that clients will need to be upgraded before date x or blocks might get rejected? Better yet just pick a block much closer to actual release date. 24k is a long way off...

Because if you upgrade the clients too soon and not everybody upgrades... then you will risk splitting the blockchain iirc.

And don't forget that NMC will be up to block 18,000 or so within the week I would guess once the difficulty drops to 22k.

[TeraPool]

So there would be just 1 instance of pushpoold running? And 1 instance of namecoind running? Along with a proxy patch of some sort?
 

You have a patched bitcoind (and its blockchain), a namecoind (and its blockchain) the merged-mining-proxy that know how to connect to both daemons and the pushpoold that connects to merged-mine-proxy.

 

How would pushpool tell the mysql databases that it received a valid bitcoin solution or namecoin solution?

As it stands, pushpool simply reports "Y", "Y" for it's "upstream_result" and "our_result" mysql columns. So how would I know that I have mined a namecoin?

Perhaps from the "version" field of the valid solution?

You don't have to rely on any information from the proxy. You could for example let pushpoold write the shares to the database and monitor the blockchains for generated blocks. Calculate how much valid shares each miner delivers between two blocks for each blockchain. Or you could patch both blockchain  daemons to deliver those information. Both is common amongst current pools

Thanks for that.

So what exactly must I do to bitcoind in order to start merged mining on my pool?

Are there any instructions to install this? https://github.com/vinced/namecoin/tree/namecoind-mergedmine

[OgNasty]

Very interesting project.  I'm not sure if I should be scared or excited.

[paraipan]

...They don't even have to provide much (or at all) financial incentive for miners, as for them there is no downside in adding another blockchain to hash. It only needs to be for something a big part of miners would support for one reason or another, and you get your hashing power.see no problem with that....

meeeh... wrong answer . No.hashing.power if incentives <= 0
everyone is free to have it's own block-chain. I would definitely not mine 10% or less of my total hashing power for another block-chain if that percentage it's not equal in price with the main block-chain. Actually I'm free to mine namecoins if I want to support distributed DNS, which I do, but my incentives are low for the moment. Namecoin will have to make it's way like bitcoin did, slowly, gaining it's users trust and no free hash power. It has the same security as bitcoin, don't need to worry bout it being fragile. Just my opinion.

[Xenland]

+1 this must be done. I will help test with my Mining Farm server

[coblee]

I think this is a very bad idea. (Assuming I understand this correctly)

First, you are polluting the bitcoin blockchain with useless namecoin data and vice versa. So you are increasing the size of the blockchain that everyone has to store for this purpose. This has a non-trivial cost on the whole network.

Second, you are tying the generation of namecoins to the generation of bitcoins. This leads to the value of namecoin being tied to the value of bitcoins. In the end, you are just effectively doubling the number of bitcoins from 21 million to 42 million.

The only benefit is a short term benefit of being able to mine more at the current hashrate. As soon as this is fully implemented, the value of bitcoin/namecoin will just decrease until the total value mined is the same. Because this is a zero sum game. You can't just generate value out of nowhere.

There is a perceived benefit that this will make the namecoin blockchain stronger. That's true, but effectively, because namecoins will be so tied together with bitcoins, it will no longer have its own identity and worth because it will be overruled by the value of bitcoins. Namecoin defines the value of a NMC by domain names. Right now, it costs 50 NMC to get a domain name. If you tie the value of NMC to BTC and BTC has a real value, then it would make the value of domain names either too expensive or too cheap. Let say in the future, 1 BTC is worth $1000 and 1 BTC is pegged at 10 NMC due to this merged mining. This makes 1 NMC worth $100. Is that the right price for a .bit domain name? We don't know. But the value of NMC cannot be changed by market forces because it's pegged to BTC. Assuming bitcoin is still the dominating crypto currency, the usefulness of namecoin's original purpose is gone.

If we go through with this merged mining, we are tying 2 crypto currencies with 2 different goals (finance vs. DNS naming) into 1 fate. This will effectively devalue both currency and likely destroy the usefulness of the auxiliary currency and prevent it from succeeding by itself. And it will add a non-trivial cost to the main currency by dirtying its blockchain.

This may seem like a win-win situation on the surface, but in reality, I believe it's a lose-lose situation for bitcoin and namecoin.

[TeraPool]

you are polluting the bitcoin blockchain with useless namecoin data and vice versa. So you are increasing the size of the blockchain that everyone has to store for this purpose. This has a non-trivial cost on the whole network.

Anybody can still run bitcoind and be none the wiser about namecoin. This will not affect bitcoin's blockchain at all as far as I know.

Second, you are tying the generation of namecoins to the generation of bitcoins. This leads to the value of namecoin being tied to the value of bitcoins. In the end, you are just effectively doubling the number of bitcoins from 21 million to 42 million.

That is for the market to decide. The current price of NMC to BTC is roughly that of it's generation difficulty now because both currencies are virtually worthless in terms of spending power.

Once namecoins can be (well) used for domain name creation, and bitcoin for buying goods, the market will create it's own prices for both blockchains. I see no reason why difficulty should play a hand in value, it doesn't for bitcoins at least.

[coblee]

Maybe an image helps.
Yeah, jumping on bitcoin's back to climb the mountain seems like a win/win solution. But in the end, the destination is different. Bitcoin does a lot more work to climb Mt BTC. Namecoin will eventually realize that he's going up the wrong mountain.

[coblee]

Anybody can still run bitcoind and be none the wiser about namecoin. This will not affect bitcoin's blockchain at all as far as I know.

If miners are working on the same work, wouldn't bitcoin's block also contain namecoin's transactions some how? Otherwise, how could the same share be used for both networks? Someone who understands this better, please correct me.

That is for the market to decide. The current price of NMC to BTC is roughly that of it's generation difficulty now because both currencies are virtually worthless in terms of spending power.

Once namecoins can be (well) used for domain name creation, and bitcoin for buying goods, the market will create it's own prices for both blockchains. I see no reason why difficulty should play a hand in value, it doesn't for bitcoins at least.

I think the value and difficulty of a crypto currency is tied closely together. Although I can't prove it, I think it's not just higher value causes higher difficult, but higher difficulty causes higher value also because the blockchain is more secure. Because of that, I don't think both currency can succeed at the same time if their generation/mining/difficulty is tied together.

[drawoc]

Saying that the value of NMC is linked to BTC isn't a sane argument as to why this shouldn't be implemented.

The only reason that the difficulty is linked to the exchange rate right now is because I have to choose between mining NMC and BTC.

As an example, let's say the difficulty of NMC is 1/2 the difficulty of BTC right now.
If the exchange rate is more than 2 NMC per BTC, and I want Namecoins, I'd be better off mining BTC and exchanging them to NMC (Thus lowering the exchange rate).
If the exchange rate is less than 2 NMC per BTC, and I want Bitcoins, I'd be better off mining NMC and exchanging them to BTC (Thus increasing the exchange rate).
And so because I can choose between mining NMC and BTC, and I will always do what's the best deal for me, the exchange rate is directly linked to the difficulty.

When I can mine both BTC and NMC at once, the linkedness falls apart completely. Suddenly, if I want BTC, it's best I mine both and exchane the NMC. If I want NMC, it's best I mine both and exchange the BTC.

So now everyone's mining both, so the difficulty becomes equal, and since the miners no longer look at the difficulty and exchange rate of both, and they are no longer linked by difficulty, only by how many people want BTC and how many want NMC.

Since BTC has caught on much more than NMC, I predict the exchange rate for NMC will drop a ton because all the Bitcoin miners will suddenly have Namecoins and want to get Bitcoins (and while the same thing will happen for namecoin miners, there are much fewer of them).

[Zibbo]

...They don't even have to provide much (or at all) financial incentive for miners, as for them there is no downside in adding another blockchain to hash. It only needs to be for something a big part of miners would support for one reason or another, and you get your hashing power.see no problem with that....

meeeh... wrong answer . No.hashing.power if incentives <= 0
everyone is free to have it's own block-chain. I would definitely not mine 10% or less of my total hashing power for another block-chain if that percentage it's not equal in price with the main block-chain. Actually I'm free to mine namecoins if I want to support distributed DNS, which I do, but my incentives are low for the moment. Namecoin will have to make it's way like bitcoin did, slowly, gaining it's users trust and no free hash power. It has the same security as bitcoin, don't need to worry bout it being fragile. Just my opinion.

But you would not need to sacrifice your bitcoin hashing to do namecoin (or whatever blockchain) hashing. If you have 1GHash/s of hashing power, with merged mining you could mine bitcoins with 1Ghash and namecoins with 1Ghash. If you refuse to give some project a free ride "just because", even though there is no loss to you, that's your decision, and you are free to do it. I suspect you would be in a minority though.

[Mike Hearn]

TeraPool is correct. Merged mining does not pollute the Bitcoin block chain with Namecoin data. The only addition to the chain is a single hash in the coinbase transaction - ie, an additional 33 bytes per block. It isn't significant. This is the whole point of having split chains that share work.

[TeraPool]

TeraPool is correct. Merged mining does not pollute the Bitcoin block chain with Namecoin data. The only addition to the chain is a single hash in the coinbase transaction - ie, an additional 33 bytes per block. It isn't significant. This is the whole point of having split chains that share work.

And just to re-iterate.

That "addition to the chain" is only in the namecoin blockchain.

Bitcoiners will be none the wiser unless they switch to a pool that is helping them mine namecoins as well. In which case the bitcoin blockchain is still completely unaffected.

[coblee]

TeraPool is correct. Merged mining does not pollute the Bitcoin block chain with Namecoin data. The only addition to the chain is a single hash in the coinbase transaction - ie, an additional 33 bytes per block. It isn't significant. This is the whole point of having split chains that share work.

And just to re-iterate.

That "addition to the chain" is only in the namecoin blockchain.

Bitcoiners will be none the wiser unless they switch to a pool that is helping them mine namecoins as well. In which case the bitcoin blockchain is still completely unaffected.

That's wrong. 33 bytes are added to the bitcoin blockchain. As Mike Hearn said, it's not significant. So I agree, it's not really polluting the bitcoin blockchain. But I'm still uneasy about what the consequences to namecoin would be if you tie the generation of namecoin and bitcoin together.