Wonder who this solominer is? 88.6.216.9

[Stardust]

TX fee increase will not solve the problem, because probably profit is not his goal. Were that the case, it would not deter similar attacks in the future.

A party interested in destroying Bitcoin, would be first interested in people losing interest in Bitcoin.  And what better way to accomplish that than to slow down transactions?

[1714070950]

1714070950

[1714070950]

1714070950

[1714070950]

1714070950

[pieppiep]

Without including tx, profit can still be his goal.
Assuming it's a botnet, he don't want people to notice the trojan/virus/whatever, and the little program must be as small as possible so doesn't download the complete blockchain to check the tx for validity. If he would download the blockchain, he would a very little more profit each block he finds, but he will find less blocks because more people would notice the trojan.

[DeathAndTaxes]

TX fee increase will not solve the problem, because probably profit is not his goal. Were that the case, it would not deter similar attacks in the future.

A party interested in destroying Bitcoin, would be first interested in people losing interest in Bitcoin.  And what better way to accomplish that than to slow down transactions?

It seems the stupidest and highest cost possible way to kill Bitcoin.  This is what the 20% attack which costs the attacker a huge sum of money and slows down avg 6 confirm time by a whole 3%.   Oh noes Bitcoin is doomed because my 6 confirm time went from 60 minutes to 63 minutes.

If this thread didn't exist nobody would even notice.  The 95% confidence interval of 6 confirm time is 60 minutes +/- 18 minutes.  The slowdown isn't even statistically valid.

[Stardust]

It seems the stupidest and highest cost possible way to kill Bitcoin.  This is what the 20% attack which costs the attacker a huge sum of money and slows down avg 6 confirm time by a whole 3%.   Oh noes Bitcoin is doomed because my 6 confirm time went from 60 minutes to 63 minutes.

Yes because a botnet is costly, and Windows users are security savy, and botnets never grow.

If this thread didn't exist nobody would even notice.  The 95% confidence interval of 6 confirm time is 60 minutes +/- 18 minutes.  The slowdown isn't even statistically valid.

Indeed because we are stupid.

[DeathAndTaxes]

A botnet is costly.  It is called OPERTUNITY COST.

If I have a botnet that can be used to generate $10,000 per day doing X and I instead use to "attack" Bitcoin where my attack slows the network down to less than statistical variance in normal confirmations that is stupid.

Simplest explanation is likely the most accurate:
botnet = lots of computing power = easy to amortize mining blocks.
tx fees = nearly worthless = not worth the cost & complexity to add them.

[CA Coins]

If it is a botnet, what is the difference between:

1.  hacking into 100k computers, and "mining", without consent or processing any transactions, roughly 700 BTC a day (est. 1 Th/s)

and

2.  hacking into Linode and stealing 43k+ BTC?

1. Hacking into 100k computers is a crime against the owners of those computers. It's not an offense against the Bitcoin network, which does not require processing of transactions and certainly doesn't require "consent".

2. Stealing 43k BTC from Linode is a crime against the owners of those BTC.

Since you asked, that's the difference.

Also, there is an upside (of sorts). If mystery miner is getting 25% of the blocks, there's less risk of DeepBit getting 51%.

I guess we differ in opinion.  Hacking into other people's computers and mining without incorporating transactions is IMHO stealing from the bitcoin network.  Those are blocks and rewards that you and any other miner could be solving.  This MM is essentially stealing 0.1 BTC (figuratively speaking, whatever your block reward is) from each miner every time MM solves a block.  But I guess there are no rules saying you have to mine with computers that you have permission to use, so I see your perspective.  As to the potential upside, what happens if/when MM grows to 1 million computers (less than 1% of PCs shipped last year) and achieves 10Th?  I agree that it might not be profitable as D&T pointed out, but just for the sake of argument.

[Jan]

A botnet is costly.  It is called OPERTUNITY COST.

If I have a botnet that can be used to generate $10,000 per day doing X and I instead use to "attack" Bitcoin where my attack slows the network down to less than statistical variance in normal confirmations that is stupid.

Simplest explanation is likely the most accurate:
botnet = lots of computing power = easy to amortize mining blocks.
tx fees = nearly worthless = not worth the cost & complexity to add them.

+1

[Isokivi]

Does the MM vote on bip16 ?

[DeathAndTaxes]

No.  There is no BIP 16 in the coinbase flag.   Still with 70%+ support he will have to upgrade or risk generating orphaned blocks.

[guruvan]

A party interested in destroying Bitcoin, would be first interested in people losing interest in Bitcoin.  And what better way to accomplish that than to slow down transactions?

It seems the stupidest and highest cost possible way to kill Bitcoin.  This is what the 20% attack which costs the attacker a huge sum of money and slows down avg 6 confirm time by a whole 3%.   Oh noes Bitcoin is doomed because my 6 confirm time went from 60 minutes to 63 minutes.

Quite right. Irrelevant to anyone with enough reason & capital to kill bitcoin.

If you're a government, or a bank, trying to kill bitcoin, the way to kill interest in bitcoin is simple - kill the value of the coin. At $41M market cap, it wouldn't be hard to kill it with money, no need whatsoever for hash power. That's way too much work. No need to slow transactions, too much work again.
Acquire coins, dump 'em, tank the market. At the level of trillions vs. millions, it doesn't take much effort to do this.  Do it a few times, and no one wants to hold that asset. Oh, and maybe stop allowing transactions related to it, but that's probably secondary.

Only a smaller competitor to bitcoin would likely want to try to defeat bitcoin by slowing the network - and this is only due to the attacker's lack of capital. Honestly, if it's this, it seems more like s script kiddie style attack, and not much to be concerned about in the long run.

Will the protocol allow for a free market solution to this? Could one pool offer to process transactions for X fee, and another for Y? While the client doesn't seem to do so today, could the client then be altered to allow the user to direct the transaction at a given pool ("clearinghouse") If it can work this way, then there is no reason to care about the no TX miner - people would direct TXs to the fastest/cheapest processors), and miners not processing TXs would have no effect.

Honestly, it seems to me that not finding a way for the miners to compete on fees is the real longer term problem. The only way for a miner/pool to make more money is to control more of the network.

[DeepBit]

No.  There is no BIP 16 in the coinbase flag.   Still with 70%+ support he will have to upgrade or risk generating orphaned blocks.

Don't spoil the fun please.

[DeathAndTaxes]

Will the protocol allow for a free market solution to this? Could one pool offer to process transactions for X fee, and another for Y? While the client doesn't seem to do so today, could the client then be altered to allow the user to direct the transaction at a given pool ("clearinghouse") If it can work this way, then there is no reason to care about the no TX miner - people would direct TXs to the fastest/cheapest processors), and miners not processing TXs would have no effect.

It already does.  Miners are free to choose which transactions to include.  To date most pools & miners simply include all valid transactions.  The subsidy is very high and that is the default option of bitcoind.

All that is necessary is for the client to be aware of the fee policy of all/most miners and the current pending transactions and it could provide a recommended fee (include 0.01 for for 95% chance of inclusion in next 3 blocks, include a 0.0005 fet for 95% chance of inclusion in next 18 blocks).

Of course it is a chicken & egg scenario.  If all miners include all tx (fee or not) then there is no real reason for the client to be upgraded. If users have no way of knowing what fee pools/miners require then there is no reason for pools to exclude tx with low fee.

Personally I think two simple thing can happen to get the ball rolling.  1 some % of miners simply agree to not include tx w/ fee below X BTC.  If enough do then there will be a catalyst for making the client "network aware".  The second event is simply the subsidy cut.  If not this one (50->25 then certainly the next one 25->12.5).  As subsidies decline the free ride miners have been granting will come to an end.

I am thinking of offering a bounty for a patch to bitcoind which allow for a command line argument (--min-fee x) which excludes work without sufficient fee from getworks.  That would allow even p2pool users to set a fee policy.

Honestly, it seems to me that not finding a way for the miners to compete on fees is the real longer term problem. The only way for a miner/pool to make more money is to control more of the network.

Which is why I think a minimum fee other than 1 satoshi will be necessary but we have plenty of time to think about it.

[Polvos]

I am thinking of offering a bounty for a patch to bitcoind which allow for a command line argument (--min-fee x) which excludes work without sufficient fee from getworks.  That would allow even p2pool users to set a fee policy.

Is it possible for that kind of patch to include the option of mining your own fake high fee transactions that you don't broadcast to the network? (meltmining)

If the answer is yes add me to the bounty (I'm poor, but half of my coins will go there)

[molecular]

If you're a government, or a bank, trying to kill bitcoin, the way to kill interest in bitcoin is simple - kill the value of the coin. At $41M market cap, it wouldn't be hard to kill it with money, no need whatsoever for hash power. That's way too much work. No need to slow transactions, too much work again.
Acquire coins, dump 'em, tank the market. At the level of trillions vs. millions, it doesn't take much effort to do this.  Do it a few times, and no one wants to hold that asset. Oh, and maybe stop allowing transactions related to it, but that's probably secondary.

And how exactly would you go about acquiring the necessary sizeable pile of coins?

[ribuck]

Here's another way to think about this:

Satoshi was quite clear that it was reasonable for each miner to set a fee threshold for transactions to be included in their blocks, and that a market would develop amongst miners.

So mystery miner's actions are no different from a miner who sets a fee threshold of 21 million bitcoins to include a transaction in the mined block.

[guruvan]

If you're a government, or a bank, trying to kill bitcoin, the way to kill interest in bitcoin is simple - kill the value of the coin. At $41M market cap, it wouldn't be hard to kill it with money, no need whatsoever for hash power. That's way too much work. No need to slow transactions, too much work again.
Acquire coins, dump 'em, tank the market. At the level of trillions vs. millions, it doesn't take much effort to do this.  Do it a few times, and no one wants to hold that asset. Oh, and maybe stop allowing transactions related to it, but that's probably secondary.

And how exactly would you go about acquiring the necessary sizeable pile of coins?

Well, how about stealing them? seriously, If I'm a huge $XXbillion business I can buy coins all day and then sell them all off at a huge loss, and not care financially. A $41M loss to JP Morgan is a bad trade on a bad day. Not really a big loss. What do they care how they acquire the coins? All they have to do is make sure the acquisition doesn't drive the price higher than they could drop it.

To do it, I certainly don't have to be obvious as I start my plan, only at the crucial execution in the market (when I dump) - penny stocks have been famous in the US for this type of manipulation.

Here's another way to think about this:

Satoshi was quite clear that it was reasonable for each miner to set a fee threshold for transactions to be included in their blocks, and that a market would develop amongst miners.

So mystery miner's actions are no different from a miner who sets a fee threshold of 21 million bitcoins to include a transaction in the mined block.

Right! The free market at work.

[conspirosphere.tk]

It would be interesting to see what is he doing with the BTC stolen from legit miners.
Is he selling already?

[DeepBit]

It would be interesting to see what is he doing with the BTC stolen from legit miners.
Is he selling already?

Nothing was stolen from legit miners.

[conspirosphere.tk]

It would be interesting to see what is he doing with the BTC stolen from legit miners.
Is he selling already?

Nothing was stolen from legit miners.

A lot was stolen from miners in 2 ways:
1) seizing a big percentage of the BTC created (which otherwise would have be gone to legit miners)
2) Increasing the difficulty, so reducing the rate at which legit miners gain

Now can you see the light?

[Littleshop]

It would be interesting to see what is he doing with the BTC stolen from legit miners.
Is he selling already?

Nothing was stolen from legit miners.

Correct.  If it was a botnet, the computer access and power are being stolen from those owners.  If it is a ASIC, nothing is being stolen at all.