Bitcoin Forum

The three pools known with “merged mining”:

1. masterpool.eu  did 20 NMC-Blocks since merged mining
2. nmcbit.com did 6 NMC-Blocks since merged mining
3. simplecoin.us did 10 NMC-Blocks since merged mining

36 blocks have been “officially-merged-mining” found, hashrate complete: ~100 Gh/s

130 have been found

94 by the “rest” of NMC-Miners, it must have a hash-rate ~350 Gh/s to do so

backdoor-merged-mining is a thinkable reason, with cheating miners, because there is no post I have seen

I think this will be bad for namecoin protocol as well as the miners being cheated.

What does that mean, in technical terms?

EDIT: Do you mean some pools are doing merged mining but not passing back the coins to the miners?

what do you think, when there is a necessary power of 350 Gh/s with no infos where it might comes from?

I think one/some pools do merged mining without informing their miners. Backdoor in their own wallet.

Is it impossible to detect? Because the way it works is that the pool pushes to the miners basically a hash of the block transactions that need to be signed, and the miners try to generate a valid hash on the hashed data that they received? But maybe you can detect that the nonce range is fishy, because the nonce has to contain the hash of the namecoin block? If it's impossible to detect then the only way for a miner to avoid being cheated is to use one of the merged-mining pools?

i dont know, how to detect them, but maybe there are some coders who can do this.

the pool(s) who do so, must live with that big abuse of confidence - yet and in the future

This is VERY LIKELY to be happening. Bring down the scammers !

OK, I asked in #bitcoin-dev on irc:
http://bitcoinstats.com/irc/bitcoin-dev/logs/2011/10/10/3
Backdoor merged-mining can be detected after the block was generated, but not while generating it.
See e.g. http://digbtc.com/ regarding which pools found which blocks.

Where can i see, who found NMC-blocks?

Another question is: Could it be seen in the BTC-block if there was some (backdoor) merged-mining?

This is most likely due to the operator of a large pool using all the hashing power for merged mining but only paying out the Bitcoins.  Any miners with a standard Bitcoin pool are expecting a certain payout in Bitcoin for their hashing power.  Starting merged mining at the pool does absolutely nothing to change this (aside from possible a little server downtime) and so is a good opportunity for profit.  If miners actually want the namecoins then there is a selection of pools which payout namecoins too.

Does anyone have any evidence to support the idea of there being a way of actually cheating with merged mining?

Quite possibly.

We have not implemented merged mining yet, busy getting our new server up and setting up a US server, then a couple more things in our priority list.
If/when we decide to MM we will email our members and announce in this forum - as we do with all major changes.

We do recognise the tough times miners are coping with currently so are paying 55BTC per block as a way to offer extra value to miners :)

This are the arguments everyone likes very much. Your opinion in other words:

"We can cheat everyone, who doesn't care about it!"

I hope every pool operator puts his cards on the table, to see who is not informing his miners about what is done with the miners shares on the miners payed electricity.

Well, i don't like to be scammed!

Here's one reason the rate is high: https://bitcointalk.org/index.php?topic=42667.msg566146#msg566146  who knows if the other pools are doing it too.

I dont understand why do even mine in btc-only pools, when u earn much more on merged mining pool..

If they wanna give away their nmc coins to pool operator whos back mining merged, thats fine, serve them right..

cheers

Don't put words in my mouth!

I would guess that most people mining at a standard Bitcoin mining pool are expecting a "fair share" of Bitcoins for their hashing power.  By using the hashes to generate Namecoins I would argue that a pool operator is doing nothing to change this.  I fail to see how any cheating or scamming is going on.  No rules are being broken and no fraud is taking place.  If a miner doesn't care about namecoins then this is a non issue.  If a miner does care then they might consider using a pool which pays out the namecoins too; it's a free market.

If a miner actually has a moral problem with Namecoins (but not with Bitcoins) then there could be something to this but I honestly feel that most of the comments here are driven by personal greed and envy.


There are several reasons.  Low variance, reputation, reliability, lower risk.  Honestly though, with BTC income close to electricity cost for most and with merged mining offering a greater than 40% income bonus I would expect most people to try merged mining.

Here is the solution:

slush has tested merged mining with his pool...

rising up of the "missing" 100 blocks...

he will spread NMC between his miners...

There is nothing "fishy" about the nonce range.  It is still 2^32.

The namecoin data isn't in the nonce range it is in the block transactions itself.

There may be other ways to detect it but it won't be by looking at the nonce.

The very simple solution to this is for people to move away from pools that do not officially support merged mining. This will put pressure on the pools that do not (and perhaps are "cheating") to officially support it.

Problem solved.

Also, could this discrepancy not be explained by merged solo miners?

Nobody has to.  You believe pool operator secretly mining namecoins uses miner's hashing power and keeping the profits to be SUPER DUPER OK!

99% of people would consider that cheating.

If a pool want to keep the NMC that is fine but they should announce it.  "We pay you 100% of the BTC and keep 100% of the NMC".   To do it secretly is stealing.  They are stealing the hashing power of pool users.


Yeah and greed has nothing to do with merge mining, not telling your pool participants and keeping the namecoins.  At current rate that works out to a 28.5% pool fee.  Miner's aren't being told they are giving 28.5% of total revenue to the pool operator.

Some of it likely is.  But I doubt there are 350GH/s worth of solo miners who also implemented merged mining this early.    Some of it is likely some adventurous solo miner but I doubt all of it is.

The only way I can think you can detect it is after the parent chain has made a block... if the aux chain has a lower difficulty there should be a corresponding block.  Both block contain the data needed to confirm they linked.  However if a pool op wanted to hide it all they'd have to do is when they won a bitcoin block don't send the solution to the aux chain.  Then there's no link... They lose 1 block on the aux chain but they get all the aux chain blocks that meet aux difficulty but not bitcoin difficulty, which is most of the.

+1

Nonsense. Everytime pool is doing MM on BTC network, his coinbase is different (much longer) than usual. So you can detect if pool is doin MM, lets watch "raw data" of his blocks. There's no way ho to hide this information.

Example of coinbase while merge mining: http://blockexplorer.com/rawblock/00000000000008371089fddbdee1f7cf580d98e64c276c649102d2ab95fd9844

Slightly OT: how do mining pools prevent miners from taking successes out of the pool and keeping them for themselves?  Do they track the nonces used in recently minted bitcoins?

What if a pool operator decides to use the merged mining software for reasons of compatability and efficiency but sends all of the generated NMC (the public keys to spending them) to /dev/null?  Would it be ethical then?

What if the pool operator runs the merged mining software but patches it so as not to broadcast winning namecoin hashes to the network?  Is this still stealing hashing power from miners?

What if a pool operator uses the latest bitcoind (no merged mining at all) and just pays out the Bitcoins to the users.  Are such pool operaters guilty of theft?  Do you accuse deepbit and BTC Guild of suddenly charging a 30% fee because they didn't immediately and rashly start using merged mining?


What about a pool operator which does merged mining with bitcoin and foocoin (worth nothing) but only pays out the bitcoin and doesn't even mention the foocoin?  Are they required to inform users of how they are using the shares then?

What if an pool operator finds a way of using the winning hashes in a clever way to add a small amount of processing power to Folding@Home or similar?  Are they required to state exactly what they are doing?  Would they be required to pay the miners extra for this?

What if they print off all of the winning hashes and hang them on a wall as a piece of art for their own personal enjoyment?  Are they required to let the miners know of this?

Merged mining takes nothing from the Bitcoin miners at all.  There is no stealing because nothing is being taken?  There is no fraud because no-one is being mislead.  Bitcoin miners receive just as they would have before merged mining became available.  All that is different is that the pool operator is secretly being more efficient.

All that a miner and a pool operator have is an agreement.  There may not be anything in writing but this is where ethics comes in.  The pool operator specifies the reward system and fees, and then the miner supplies shares to the pool operator.  If evidence can be found that the operator is not living up to the agreement then this is cheating, plain and simple.  If the pool operator advertises the pool as "giving you a fair share for your work" or "making your hashes work as hard as possible" then they are arguably guilty of fraud or theft if they use the proportional reward system or don't use merged mining respectively but few make such claims and expect miners to be intelligent enough to assess the proposed deal on its own merits.  I find it laughable that you would suggest that I'm in a 1% minority of people on this.

This is responded thousand times on this forum already, but let's say that for 1001 ;).

a) Miner don't have complete block sources, only block header for finding corresponding hash. He don't know what to push to bitcoin network.
b) There's special (coinbase) transaction inside block source for 50 BTC adressed to pool's wallet. If somebody somehow got block sources for that block he found and he'll submit this block to bitcoin network directly, 50BTC reward still come to pool.

A miner can't keep the success for himself because the block solved includes a coinbase transaction going to the pool's wallet.  The hash the miner finds only meets target requirement for that block.  A different coinbase address would have a difference hash and thus require  a different nonce to meet difficulty target.

A miner could simply destroy the "winning hash" and not report it but they can't directly steal the winning hash however that is another issue. So far pool poisoning seems to be a non-issue.  If it became an issue there are countermeasures which could be used.

Gah! beat me to it.

Does this mean you only have 303 posts on this forum which are not about this? ;)

Well that wouldn't be stealing as they are benefiting from other's hashing power.  Still it would be nice for pool operator to be open with what they are doing with hashing power.

Well that is kinda stupid as you have reduced the efficiency of pool server but likewise not stealing as the pool operators aren't gaining anything.


More nonsense.  NO because they aren't skimming 30%+ off the top and hiding that from users. However if they DO keep NMC rewards then they are skimming 30% off the top and if they don't inform the users then it is both lying and stealing.

Make up all the nonsense comparisons you want but it doesn't change the fact.  A pool can do four things.

a) not merge mine = fine (we will see how market forces affect those pools)
b) merge mine, notify users they are keeping NMC as pool cost (as many pools do w/ transaction fees now)
c) merge mine, notify users and split NMC (or sell NMC and split BTC)
d) steal from users by using their hashing power to enrich the pool operators

It is stealing and lying.  Lying by omission is still lying.  Most pools indicate how the pool is compensated and how rewards are split.  Taking the NMC rewards without notifying the pool is stealing and lying.  Period.


Then laugh on but you are in the minority and your attempt to obfuscate by including nonesense examples.

Here is deepbit homepage:
Your part would imply rewards are split by number of shares submitted.  ith merged mining a block reward is higher.  It isn't merely 50BTC.  It is somtimes (most of the time) 50BTC + 50NMC.  If one only splits the 50BTC and keeps the 50NMC then it isn't being split equally.

Most pools have similar language indicating the "reward" is split equally by shares (or via some other system).  With merged mining that reward now sometimes includes 50 NMC.

Let us consider your excerpt from deepbit.


In the case of the Pay per share, the offer is for a certain amount of BTC per every submitted share.  It matters not what the pool operator does with those shares so long as the appropriate amount of BTC is paid.  If the pool operator is able to sell the shares on honestly and fairly for 1000 dollars a pop then more power to them!  If the pool operator is using the shares to do something ethically questionable then that is another matter altogether but I personally don't put "mining namecoins" in this category.

In the case of the proportional system I'm right with you.  They say that you receive your part of every solved block less 3% fee.  To apply this only to the solved bitcoin blocks but to use the hashing power to solve namecoin blocks in secret on the side is dishonest and misleading (namecoin blocks are blocks too after all).  If they went on to clarify that a block was specifically a Bitcoin block then they are above board again.

Reviewing your answers to my questions it seems your objection is not with the mining of Namecoins in secret but purely with the fact that the pool operator is profiting.  You are happier for namecoins to be mined and destroyed in secret than to be mined and pocketed in secret.  Denying a pool operator this would not be unlike requiring a FGPA miner to sell his bitcoins much more cheaply simply because he can.  If a pool operator started a PPS pool offering just 0.00001 BTC per every submitted share (no other details given at all) I would not accuse them of theft (I will not be so devious as to claim that 99% of people would agree but just suggest that more than 1% would).  Their ethics come into question only when they fail to deliver the promised BTC.  This is such a basic tenant of the free market that I expect we are using different definitions for some term and I cannot for the life of me think what it could be.

By "This" I am referring to an offer like "we pay 0.00002663771002209 BTC per every submitted share".  You instead seem to be referring to something like "we pass all the proceeds of our mining activities onto the miners".  For the second case, secretly keeping namecoins would be unjust whereas simply not mining them in the first place is sound.  The way I see it, most pools operate by specifying an explicit reward system so the former case applies.


+1

This is a job for the free market.  Instead of declaring certain reward systems immoral pool operators should be free to offer whichever reward systems they like.

I don't mean to appear a troll so I will stop posting here now.  I feel there has simply been a breakdown of communication.  I believe that saying you will do one thing and then proceeding to do something entirely different is dishonourable.  I'm arguing about buying shares for bitcoins and I don't think that the buyer is required to detail what they intend to do with those shares.  What is important is that they are honest about the actual trade and provide the promised bitcoins in exchange for the shares.

Doesn't anyone think that slushs "testing" implementation is also cheating?

Even if thier miner get payed afterwards, nobody will notice if a fraction of the nmc not get payed out, and most importantly he is stealing from the official merged mining pools by increasing the overall hashrate.
So the way I put it: Slush gtfo of our merge  >:(

No.

"Even if thier miner get payed afterwards, nobody will notice if a fraction of the nmc not get payed out,"

If the miners are paid in full them it isn't stealing.  How about you wait and see first.

Your logic seems to be that because he could cheat his miners that means he already has cheated them.  

Future crime anyone?

lol, why do you think I'm "cheating"?


Well, my pool is also "official merged mining pool", I just didn't finished GUI&other stuff. Don't forget that I have already big load on pool, so releasing such bleeding edge software is much harder than for 20GHash pools...


Of course I can cheat! Any pool operator can cheat. I can not publish some merged mined blocks for namecoin and nobody will notice, as any other merged mining pool can do. So why do you have problem with me? I think that I already prove that I'm not a scammer, by running pool for the longest time ever.

Yeah, The same could be said for normal BTC mining, pool operators could keep every 10th block found for themselves and no one would ever know.  It all comes down to being able to trust the operator.

My issue is: That hashrate has no point in being there. If people want merged mining they go where it is offered. Unofficial testing is just a convenient way of risk reduction (for you).
The thing was announced way in advance so there was plenty of time to test on testnet.

Stealing may not be the right term, but we (the people mining at official merged pools) would get more nmc if it weren't for you, even if you pay out the earning to your miners.

Who gives a flying crap?  If you can't compete then die.  Sorry this is capitalism at its finest.

I don't use slush pool but your claims he took "your namecoins" or that "his hashes don't belong there" are beyond pathetic.  If your hashing power sucks so bad you can't compete then close up shop. The network doesn't need you (as in you the individual), it just need aggregate power.   You aren't entitled to any monopoly of namecoins.  

I I were a miner @ slushes I might be annoyed that he risked the stability of the pool for testing but that would be the extent of my beef.  Still the idea that "his hashes don't belong there" just makes me laugh (not with you but at you).

As far as using test-net. Most of the issues w/ merge mining deal w/ large pool integration.  It is kinda hard to find 100GH/s worth of hashing power willing to spend couple hours on test-net simultaneously for free.  

Grow a pair and stop whining.  If you think slush's test was bad just wait I predict within a month most major pool will be merge mining (because miners will flow to those who are from those who aren't) so you can expect 5TH+ more competition.

Wait - what? You're jealous that I'm mining namecoins for my users? :)


Or they just don't need to do anything because they know that will earn namecoins for mining on my pool.


It is "unofficial" because there were silly problems before and I didn't finish GUI yet. Sorry, but it looks like you never developed something what must be together 24/7 up and on bleeding edge (which MM definitely is).


Haha, the best thing on this sentence is that I have prepared merged mining weeks ago. But it started crashing on production network. Yes, sometimes things are crazy.


I don't know what to say to this, except - sorry,  world isn't fair. It's the same as I can say "damn, I might earn much more if deepbit never come".

I'm not whining, but anyone resorting to those "not nice" tactics should grow some and accept being responded with "not nice" words.  :P

I admit it I am, so what?
It feels like some daddy getting in line for free popsicles for the kids.

What is the timeline for getting the gui online?
I think it's good for NMC to have at least one of the big boys on board.

Fair call...  A pool op would still have plausible deniability.  You can insert any old junk into the coinbase.   Correct me if I'm wrong but the MM specific stuff in the coinbase is the merkle root + tree size + nonce so with a random nonce the only real signature is the tree size and length of the data.

However, plausible deniability is probably largely academic since any pool up with unexplained data in their coinbases will no doubt come under suspicion now.

BTW this looks wrong to me: coinbase":"045dee091a03ac8701522cfabe6d6defb6519b6439e5852d2cdba258fe21438204f3eb6f69001fc d74e884ee87482b0100000000000000"

it indicates a tree size of 0?

lol not going to kill off the exchange, It will dip low as all the new hashing power mines a bunch of blocks quickly and the Difficulty stabilizes back to a normal 6 blocks an hour.  This will just give people the ability to buy into NMC cheaply if people are willing to sell off there NMC cheaply.