Deepbit Approaching 50% Once Again

[gsan]

But yes, bigger pools have lower variability on proportional or score based schemes as well. So best way to solve the current problem would be taking away the control of the list of included transactions from the pools.

To do that you would have to modify the protocol. Go ahead and write a patch to do that. It won't be accepted into the official bitcoin client and I won't download any client that has that in it.

I'm at a loss here. What's the connection with the protocol? And why would you object? What do you say to this proposal?

Everyone gets all information from the network. You would have to modify the bitcoin protocol to limit someone from getting all information. It is a reactive response and against the ideals of the system.

The thread does not solve any problem, it proposes an alternative and has nothing tangible to show yet. It was written when there were only a few other pools than deepbit. There are more alternatives now, there are 6-7 other pools and growing every week. It does not solve the fact that there are other pools and people aren't using them.

I don't know what you mean by limiting someone from getting all information, or how it would be applied in this case. Maybe we aren't talking about the same problem. I haven't seen a proposal involving a protocol change in this manner.

The basic idea is, if the pool operator doesn't have deciding power on the transaction list that gets in the block, a "public" pool, or any colluding group of them wouldn't be considered an attacker with high computing power. I'm not suggesting that there is an easy way to do that, but merely, if it could be done, it would solve the problem at hand. I don't have a clue how to do this with a protocol change, it's best if we approach this from the angle of, where the pool gets work from.

The basic idea in the mentioned thread does solve this specific problem, if it can be feasibly implemented, which is something yet to be proven. Either way, I think it's a step in the right direction. Of course it doesn't solve the fact that there are other pools and people aren't using them, that's an unrelated matter. The idea is applicable whether there is only one pool or a multitude of them.

[1713275250]

1713275250

[1713275250]

1713275250

[1713275250]

1713275250

[Findeton]

Right now deepbit is like 46% of the network 

[kiwiasian]

I'd like to see deepbit take over more than half, so that people will stop complaining once they realize that Tycho is not going to do anything evil.

Sigh, people need to stop freaking out.

[tehcodez]

I'd like to see deepbit take over more than half, so that people will stop complaining once they realize that Tycho is not going to do anything evil.

Sigh, people need to stop freaking out.

And...I'd like to see a giant comet hit your house, then see if you complain, or exist.

It's not about Tycho and his intentions...its about the fundamentals plain and simple.

Man...moment...machine. Let's hope some intervention happens...

[kiwiasian]

I'd like to see deepbit take over more than half, so that people will stop complaining once they realize that Tycho is not going to do anything evil.

Sigh, people need to stop freaking out.

And...I'd like to see a giant comet hit your house, then see if you complain, or exist.

It's not about Tycho and his intentions...its about the fundamentals plain and simple.

Man...moment...machine. Let's hope some intervention happens...

Nothing is going to happen once it reaches 50%.
And once it reaches that point we will all see.

[tehcodez]

We will see...someone pop the box and play make-a-chain.

Or, hopefully, a few will change those miner args and give it a break for a sec.

[lizthegrey]

It does need to be pointed out that the magical threshold is not 50%+1. It is theoretically possible for an adversary corrupting 30% or 40% of the computation to eventually succeed trick the rest of the network into forking the block chain within several hours to days according to a study performed by a group of MIT students for the 6.857 Computer and Network Security class. I don't have their graphs/slides, but I can ask them for their data (and suspect they already haunt these forums).

[Findeton]

Nothing is going to happen once it reaches 50%.
And once it reaches that point we will all see.

Perhaps nothing will happen the first day. But IT IS A security RISK and that's a fact.

You could play to try shooting yourself with one revolver and one bullet. And maybe the first time you won't die, but that doesn't mean it's not a risk.

It does need to be pointed out that the magical threshold is not 50%+1. It is theoretically possible for an adversary corrupting 30% or 40% of the computation to eventually succeed trick the rest of the network into forking the block chain within several hours to days according to a study performed by a group of MIT students for the 6.857 Computer and Network Security class. I don't have their graphs/slides, but I can ask them for their data (and suspect they already haunt these forums).

That means it already is a security risk.

[kiwiasian]

Nothing is going to happen once it reaches 50%.
And once it reaches that point we will all see.

Perhaps nothing will happen the first day. But IT IS A security RISK and that's a fact.

You could play to try shooting yourself with one revolver and one bullet. And maybe the first time you won't die, but that doesn't mean it's not a risk.

It does need to be pointed out that the magical threshold is not 50%+1. It is theoretically possible for an adversary corrupting 30% or 40% of the computation to eventually succeed trick the rest of the network into forking the block chain within several hours to days according to a study performed by a group of MIT students for the 6.857 Computer and Network Security class. I don't have their graphs/slides, but I can ask them for their data (and suspect they already haunt these forums).

That means it already is a security risk.

Oh, wow! A security risk!

I swear, some people.

[tehcodez]

Nothing is going to happen once it reaches 50%.
And once it reaches that point we will all see.

Perhaps nothing will happen the first day. But IT IS A security RISK and that's a fact.

You could play to try shooting yourself with one revolver and one bullet. And maybe the first time you won't die, but that doesn't mean it's not a risk.

It does need to be pointed out that the magical threshold is not 50%+1. It is theoretically possible for an adversary corrupting 30% or 40% of the computation to eventually succeed trick the rest of the network into forking the block chain within several hours to days according to a study performed by a group of MIT students for the 6.857 Computer and Network Security class. I don't have their graphs/slides, but I can ask them for their data (and suspect they already haunt these forums).

That means it already is a security risk.

Oh, wow! A security risk.

I swear, some people.

Some people = You.

[coined]

hmm, I thought we were all on the same page regarding a single entity gaining 50% hashing power being bad, I'm very surprised that some people claim not to see the problem. apart from the double spend threat from the owner himself. somebody could take over deepbit and have instant control, not a hack, literally kicking his door down, for somebody looking to attack bitcoin, like the government or somebody else, the problem goes from how do possibly gain 50% of this?, 20 new supercomputers? or somehow get worldwide synchronised take downs with cooperating governments? hmm how will we get the budget for that? oh wait, there's deepbit... 1 SIMPLE DOOR! 50% ta daa!  or it could be targeted via severe denial of service to be taken down briefly to weaken the total network for a fast attack, if any of those happen the confidence in bitcoin may never bounce back,

I remember somebody saying we could check to see if a double spend ever happened, like that's the end of that problem, but that is completely missing the point, its not the single double spend transactions he would get away with that's the problem, the fact that we know it happened after the fact would be the CAUSE of the massive loss in confidence in bitcoin, the news stories would be EVERYWHERE, panic selling would start immediately causing more panic selling it could potentially kill off bitcoin right now before it gets going,

it IS a big deal.

[bronte]

It´s because everyone has the $ $ sings glowing in his eyes. Fast and easy money!11! independent currency? who cares about that idea?
Tycho, if you are long-term interested in this project then disable registrations for some time ...

[Findeton]

Right now:



This is bad news for bitcoin.

[KnuttyD]

Holy lap dancing jesus.


If we start to see a breakdown in Bitcoin security, then the value of bitcoin diminishes greatly. People will sell. The coin itself goes all great-depression on our asses.
Honestly, it is in not just OUR best interest, but DEEPBITS best interest to shave off some of that power.

Or DDOS. I'm not saying we should as a first resort, but DDOS is always an option when this magnitude of cash is at stake.

[BitCheddar]

I started with bitcoinpool.  I got tired of idle miners.
I played around with some of the other pools, slush, etc.  Didn't seem to provide enough info/data for me to know how I was doing.
Went to deepbit.   No idling miners (so far), lots of data available to judge progress, etc.

People want minimum drama and results for their efforts.  The security threat people are talking about is below 50% anyways, so it just sounds like jealousy to me at the worst, anti-free choice at best.

[bcpokey]

Although I've said it before I'll say it again, while I fully support the idea that deepbit getting too many people is a bad thing I will have to interject some things here.

Most miners are not on these forums, and so these warnings will go unheeded because most people just found out about bitcoin from their friend, or some news article that mentioned bitcoin and the largest pool deepbit.

These same people are the ones driving the price of bitcoin through the roof and making it all kinds of profitable. So... we all enjoy enhanced profitability and hate the threat to bitcoins security. But for now one comes with the other. You'll have to get the word out through other channels than this thread though if you want to have any chance of making a difference.

I started with bitcoinpool.  I got tired of idle miners.
I played around with some of the other pools, slush, etc.  Didn't seem to provide enough info/data for me to know how I was doing.
Went to deepbit.   No idling miners (so far), lots of data available to judge progress, etc.

People want minimum drama and results for their efforts.  The security threat people are talking about is below 50% anyways, so it just sounds like jealousy to me at the worst, anti-free choice at best.

If I wasn't already sure that you knew nothing from your avatar (it could be ironic) the above confirms it. Plenty of pools provide plenty of information/return comparable or superior to deepbit. There is nothing special about the pool except for two things:

1) Highest hashing rate -- translates into lowest variance (although when there is variance, the low payout makes it very very very unprofitable).

2) Pay Per Share -- The best option for casual miners. This to my knowledge only exists on deepbit. If you are a serious miner though this is a horrible option (10% off the top).

Payouts, information, all that other stuff is just bunk though, even steven.

[BitCheddar]

If I wasn't already sure that you knew nothing from your avatar (it could be ironic) the above confirms it. Plenty of pools provide plenty of information/return comparable or superior to deepbit. There is nothing special about the pool except for two things:

Judging people by their avatars.  You must be new to the internet.  Real mature.  Plenty of larger pools? No.  bitcoinpool was one of the best in this aspect, but given I had to restart my miners 2-3 times a day, and they were idle most of the evening while I was sleeping and most of the day while I was at work, who cares?  I can't speak for the plethora of other smaller pools because I haven't tried them and I am not interested in wannabes.

1) Highest hashing rate -- translates into lowest variance (although when there is variance, the low payout makes it very very very unprofitable).

Haven't been there long enough to confirm or deny, ill take your word for it.  Ill see how it goes for a while and adjust if necessary.

2) Pay Per Share -- The best option for casual miners. This to my knowledge only exists on deepbit. If you are a serious miner though this is a horrible option (10% off the top).

Payouts, information, all that other stuff is just bunk though, even steven.

People need to see data to feel comfortable.  Or at least I do.  It helps drive the illusion of "can I trust this pool operator."

[TheSeven]

Its getting worse!


Jesus fucking christ you morons! MOVE AWAY FROM DEEPBIT!

The "50% of the hashing power" problem is no hard limit. It's an average case calculation. It is very well possible for deepbit to fork the block chain *right now*, and it would not require much luck. If they had 55% of the hashing power, it would require slightly less luck than it would require now.

The good news is that such an attempt would not go unnoticed...

[inh]

Until slush gets DDOS'd...

[bcpokey]

If I wasn't already sure that you knew nothing from your avatar (it could be ironic) the above confirms it. Plenty of pools provide plenty of information/return comparable or superior to deepbit. There is nothing special about the pool except for two things:

Judging people by their avatars.  You must be new to the internet.  Real mature.  Plenty of larger pools? No.  bitcoinpool was one of the best in this aspect, but given I had to restart my miners 2-3 times a day, and they were idle most of the evening while I was sleeping and most of the day while I was at work, who cares?  I can't speak for the plethora of other smaller pools because I haven't tried them and I am not interested in wannabes.

1) Highest hashing rate -- translates into lowest variance (although when there is variance, the low payout makes it very very very unprofitable).

Haven't been there long enough to confirm or deny, ill take your word for it.  Ill see how it goes for a while and adjust if necessary.

2) Pay Per Share -- The best option for casual miners. This to my knowledge only exists on deepbit. If you are a serious miner though this is a horrible option (10% off the top).

Payouts, information, all that other stuff is just bunk though, even steven.

People need to see data to feel comfortable.  Or at least I do.  It helps drive the illusion of "can I trust this pool operator."

I didn't say anything the first time I saw your avatar, only in conjunction with your post. This negates your response.

Anywho, just look at the graph that's been posted a number of times, bitcoinpool is relatively small. Large pool alternatives -- Slush, btcguild btcmine. All larger than bitcoinpool. I've used slush and am not a fan, I've used btcguild and it's basically a smaller deepbit with idle notification (but no PPS) and a lower (or 0) fee. I can't speak to btcmine, but I've heard good things about it.

The point though isn't just that there are alternatives to deepbit that are just as good as it (which there are), the point is that if you are concerned that your magical money fountain is going to dry up, then mining on deepbit makes as little sense as staying on a pool that requires your miners to restart every night. Maybe less sense.