Hidden Resiliency of Network

[cypherdoc]

there have been no shortages of theories of how an adversary to btc could attack the network.  most revolve around taking over 50% of the network to destroy the validity of the block chain.  i was thinking how i as an individual have stopped mining long ago b/c of my reduced ability to compete in generating coins with the pooled miners.  however, i do have 3 rather new and speedy CPU computers that can generate a not too insignificant hash rate but are off the btc network.  i also have developed a significant financial interest in btc and stock, not to mention a philosophical love of the concept.  if an attack by an outsider were to occur i for one would have no problem turning on my 3 computers to add to the resiliency of the network.  i'm wondering if anyone has worked out any models trying to estimate what the hidden resiliency of network from others like me might be?  i'm willing to bet that well over 60% of btc users might be in my category.  in fact, one of the administrators should try to outline a plan to mobilize as many btc users with dormant clients in event of an attack like this.  the way i see it, this could be a good way to snuff out and severely damage an adversary.

[unk]

it's a good question, but it doesn't have a good answer. we could compare various complicated calculations, but it's all just guesswork. my own conclusion (admittedly a guess) is that reserve hashing power won't be a meaningful way (given its size) to secure the network in the event of a threat that would otherwise overpower the active miners.

indeed, unused hashing power is itself a potential threat to the value of bitcoins in the currently prominent block chain. for example, what will happen to all the gpus that people are currently using if gpu mining becomes unprofitable in that chain? if i had been lured into buying a fancy gpu and then using it toward the current block chain were made unprofitable through the use of fgpas, or just through sheer competition, why wouldn't i apply my gpu to an alternative block chain as an attempt to capture some early value in that alternative chain?  (i think i would, in that situation, if i were rational and selfish.)

i'm actually surprised that there aren't already competitive financial block chains to prey upon that hope. surely if the first one has succeeded in doing that, others can as well. the marketing is easy: 'be first here, and you can be like satoshi / artforz / the very early adopters. why give them your money rather than being wealthy yourself?' whether it attracts attention or not is, of course, a matter of incentives and psychology. given your personal investment in the primary block chain, it's unlikely you'd want to support others. but a kid who owns 20 btc in the current chain and has just bought a $1300 mining rig that's soon going to be unprofitable in that main chain may seek value elsewhere, and why shouldn't he?

it will be interesting to see if members of the community attempt to stifle competitive block chains in the way that they've accused oligopolistic bankers of stifling competition in the traditional financial sector. for example, will bitcoinwatch.com display an exchange rate if i set up a new block chain called 'bitnote' that uses bitcoin or a bitcoin-like technology? will mt gox allow trades in it and conversion to btc?

[MoonShadow]

I would also be one of those non-mining users willing to mine at a loss for the good of the system, should it ever come under attack.

A plan isn't really necessary.  A 'watchdog' process needs to be developed to notify users should something strange start to occur.  For example, a watchdog process could warn the client owner of the possibility of a blockchain split in progress by watching for a relatively sudden drop in effective network hashing power.  This could be done by watching the interval between accepted blocks, and if the time between any two blocks more than doubles this average, flag a warning.  If the next one is over 180% of the noted prior average time, confidence in a problem increases.  If a third is over 150% of average time, a network split has almost certainly occurred, and users should be aware of a possible double spending attempt. 

Likewise, a watchdog process could be set up to look for signs that the blockchain is under attack, such as several consecutive blockchain reorganizations occurring in a relatively short time frame. 

[rezin777]

it will be interesting to see if members of the community attempt to stifle competitive block chains in the way that they've accused oligopolistic bankers of stifling competition in the traditional financial sector. for example, will bitcoinwatch.com display an exchange rate if i set up a new block chain called 'bitnote' that uses bitcoin or a bitcoin-like technology? will mt gox allow trades in it and conversion to btc?

Competition is good. I will gladly switch to a better system. I will not switch to a system that is exactly the same just to be an early adopter, others may.

[rezin777]

Likewise, a watchdog process could be set up to look for signs that the blockchain is under attack, such as several consecutive blockchain reorganizations occurring in a relatively short time frame. 

There is a bounty for this that arose from concerns about the growing size of Deepbit.

http://forum.bitcoin.org/index.php?topic=7622.0

[cypherdoc]

I would also be one of those non-mining users willing to mine at a loss for the good of the system, should it ever come under attack.

A plan isn't really necessary.  A 'watchdog' process needs to be developed to notify users should something strange start to occur.  For example, a watchdog process could warn the client owner of the possibility of a blockchain split in progress by watching for a relatively sudden drop in effective network hashing power.  This could be done by watching the interval between accepted blocks, and if the time between any two blocks more than doubles this average, flag a warning.  If the next one is over 180% of the noted prior average time, confidence in a problem increases.  If a third is over 150% of average time, a network split has almost certainly occurred, and users should be aware of a possible double spending attempt. 

Likewise, a watchdog process could be set up to look for signs that the blockchain is under attack, such as several consecutive blockchain reorganizations occurring in a relatively short time frame. 

couldn't this be automated with a generated alert to the Forum?

[MoonShadow]

I would also be one of those non-mining users willing to mine at a loss for the good of the system, should it ever come under attack.

A plan isn't really necessary.  A 'watchdog' process needs to be developed to notify users should something strange start to occur.  For example, a watchdog process could warn the client owner of the possibility of a blockchain split in progress by watching for a relatively sudden drop in effective network hashing power.  This could be done by watching the interval between accepted blocks, and if the time between any two blocks more than doubles this average, flag a warning.  If the next one is over 180% of the noted prior average time, confidence in a problem increases.  If a third is over 150% of average time, a network split has almost certainly occurred, and users should be aware of a possible double spending attempt. 

Likewise, a watchdog process could be set up to look for signs that the blockchain is under attack, such as several consecutive blockchain reorganizations occurring in a relatively short time frame. 

couldn't this be automated with a generated alert to the Forum?

sure, but not all users pay any attention to the forum today, and most certainly will not in the future.

[asdf]

I would think that if a pool operator started playing up, the members would migrate to another pool. It's sort of democratic like that. A pools members are a asset which it has incentive to protect.

[rezin777]

I would think that if a pool operator started playing up, the members would migrate to another pool. It's sort of democratic like that. A pools members are a asset which it has incentive to protect.

Yes of course, but if an attack did occur, would the migration be quick enough to stop the initial damage? The initial damage being peoples' trust in the block chain.