Bitcoin Forum

MODS- I'm posting this hear because it's about bitcoin

Right now, TRC has ground to a halt because it's difficulty was taken sky-high thanks to some ASICS that hopped on the network and then hopped off, leaving the difficulty incredibly high, and it taking everyone else forever to solve the block required to bring it down. It's brought the network to a standstill (although there is a fix being released for this).

Although of course TRC difficulties are incredibly low, so it makes this sort of attack easy to do- it does make one wonder how the bitcoin network would react. As we grow and grow, eventually we will get to a size large enough  that we are threatening for various other interest groups (if we aren't already). Although ASICS have made it much more expensive to attack the network, they haven't yet become widespread enough to claim we are really 'safe'.

If a State actor (for example China) were to put some time and money into it, they as well could develop a line of ASICS (indeed the only functioning ASICS are out of china- the government need really only seize their already completed designs and pay the factory to seriously ramp up production). With a large ASIC farm one could take the Bitcoin difficulty so high that when they leave the network the remaining bitcoin miners are stuck for hours trying to solve one block. As TRC shows you need only to slow down block production by a couple hours to induce widescale panic.

Thoughts?

I have been wondering the same thing recently.

https://bitcointalk.org/index.php?topic=169299.msg1761526#msg1761526

Indeed, what if for example the domestic ASIC market of China gets very large (Lets say Avalon does brisk business there), and some sort of political change happens and China blocks the internet for one reason or another. (It would be a tough call for them, but they could do it).

Then what for the rest of the world? Are we a) stuck with really long block confirms or b) stuck with really long block confirms while China's now isolated BTC network of mostly ASICS unintentionally creates a 51% situation?

Actually, lets ignore the conspiracy  theories (CHINA attacks bitcoin) or far fetched (Earthquake kills all the ASICS), and lets consider a very plausible senario- lets say there is a manufacturing fault in the ASIC machines, one that will take awhile to notice. What if the machines all fail after x number of hours- and what if an entire batch were to fail more or less, at the same time? That's a more plausible situation. Someone should be able to compute the effect this would have on the Block creation rate.

Does PPC suffer the same Problem....?

Could difficulty be set by inverse of transaction volume level being sort x hash rate

so the more transaction you have the dificulty goes down

ir decoupling difficulty to generate new block and the block chain from the transaction difficulty, thus avoiding the point your asic to mine, they could still do that but not affect the transactions side of the CC

Difficulty is set in the block and there is a good reason it's set the way it is currently. Its just that now with ASICS even bitcoin (although even more dramatically so for Terracoin) hash power is disproportionately concentrated. Imagine a perfect storm of Manufacture defect so ASICS failing at the same time, plus rising value of a coin like LITECOIN that for price reasons sucks the GPU miners onto the LTC chain.

With Diff on Bitcoin ASIC high, and price Litecoin higher then usual, you would have an unnaturally low amount of hash power on the BTC network while simultaneously having an ASIC high difficulty. And you would be stuck there for 2 weeks.

[/quote]

Difficulty is set in the block and there is a good reason it's set the way it is currently.
[/quote]

Elaborate::

WHY

why can't this be teased out?

This is actually an interesting attack vector.

The hardware requirements would probably be the same or a little less than a 51% attack - but while a 51% attack would only be good for a few seconds, this could potentially last for days.

Has it ever come up before?  I only ask because we get so bombarded by noobs about all the 'flaws' and 'mistakes' they discover in bitcoin on a daily basis, that I don't usually pay too much attention.

+1

I concur

I think ASIC jumping could be a problem for any coin.  Even Bitcoin, theoretically, if another coin got large enough to threaten it.

So it might be good to have a good solution for difficulty to quickly but securely adjust downward.  Bitcoin Testnet already has one potential solution: after ten minutes, difficulty goes back down to the minimum.  Perhaps this strategy could be adjusted: after a set long period goes with no block being found, difficulty might be reduced by a factor, then possibly reduced more for further intervals.  I think 10 minutes would be a good interval for coins like Terracoin; for coins like Bytecoin I think you'd want an interval more on the order of an hour.

It will be interesting in the near future to see if Bytecoin goes through the same thing that Terracoin is going through right now.

that sounds like it might be a decent idea walter



hopefully terracoin paves the way for all the alternate coins in ways of figuring out this tricky problem set.  if terracoin succeeds I'm sure bytecoin will look to terracoins solutions and implement them,

so hold on terracoin, there are some jagged bumps in the road, but your helping all the new generations!

p.s. I think we should set a goal to finding the solution by 4/15 when the new exchange opens, they will be offering TRC their platform and it could go over very well

The difficulty adjustment system is symmetrical, and this is for a reason.  Any asymmetry introduced would become a tool to manipulate difficulty.

Seriously guys, you need to take a deep breath and try really hard not to panic about everything.  Bitcoin is still small, it still has several problems related to smallness.  They will work themselves out over time as we grow.  Don't go off half cocked, proposing permanent bad ideas to counter temporary threats.  Losing a big chunk of hashing power is possible today, but won't be tomorrow.

The Fed is turning on their supercomputers next week. The plan is to pay off the national debt by mining bitcoins until the difficulty is 0.

Right now there are 750 petaflops of total computing power on the network.

The worlds fastest known supercomputer is a $100 million monster with 17.59 petaflops.

So we can say $1bn will buy you 750 petaflops.

Is it worth $1bn to attack bitcoin? To what end?

And when ASICS come into play in the next few years it's going to require ever more resources to throw at it.

Also the source of such an attack would surely be leaked? That is a massive project....

This issue was brought up back when Solidcoin was first created. If I'm remembering correctly, it was the first to address the issue of chain-hoppers pumping up difficulty and then leaving.

Changing Bitcoin's rules to address this kind of attack would be of marginal benefit at best. If the attacker has enough hash power to pump the difficulty up to problematic levels, they also have enough to just perform a 51% attack. If they don't have enough for a 51% attack, then at worst we'd just have to deal with 20-minute blocks for four weeks.

1. if it takes a block 2 weeks to be solved due to asic jumping, wouldnt that drive the price of btc up, since no coins being mined for two weeks?

2. i was under the impression asic mining isnt possible with litecoin yet.

I have been saying this since TRC launched. That ASICs would point to TRC and break it by running the difficulty into the sky.

The algorithm for TRC is identical to Bitcoin's and this will continue to be a problem for the foreseeable future.

Namecoin suffered this but at the hands of GPU miners flocking to NMC then run away.

NMC had a difficulty that did not adjust for 3 months.

TRC devs put such little effort into their "new coin" that they did not even manage to remove one of the biggest issues there is when creating an alt-coin, changing the mining algorithm.

If no coins are being mined, then no transactions are being included in blocks - how do you intend to send the bitcoins to be sold?

Namecoin is an utter failure, there isn't even a single person who developed a Chrome extension yet these people seem to think that if they sit on their namecoins for long enough someone will do the work for them.

What stops someone from copying Namecoin, purchasing all the names, then investing tens of thousands of dollars to make it widespread.

I think if a big enough batch of ASICs would crash someone would use the opportunity to replace them and get a profit in the meantime, especially with an economy which is relatively big like BTC.

I think the idea was more like an undersea cable was cut or some other disaster isolated hashing power in geographic areas cutting it off from the rest of the network. So imaging rather then the difficulty first increasing, it instead precipitously falls.