Topic: .. (Read 3016 times)

I don't deny it is a problem I just don't see a good solution. All of the ideas that have been proposed seem like they might do more harm than good, or perhaps do nothing despite large investment to implement and deploy.

It may require a Satoshi-level breakthrough.

The main fundamental problem with 2P-PoW and all of these non-outsourcable solutions is that they just lead to huge mining farms, which is even more centralized. A lot of the ghash.io hash rate is already in house, as are other huge farms. (KnC etc.) Together those might already be 51%+, and are certainly large enough for some other forms of misbehavior. This just increases the incentives to build such huge farms.

There are other problems with that particular proposal but until the fundamental problem (not enforcing actual decentralization, only decentralization away from large pools, but toward what?) is solved somehow, those don't really matter

On the surface 2P-PoW seems like a reasonable solution, as does P2Pool. The problem isn't that solutions exist it's that no one in charge of the code is willing to change anything. I suppose I should amend my first sentence above to read, "Because the devs don't know what to do, don't like any solution for fear of breaking Bitcoin so they are unwilling to do anything". The problem with Bitcoin is it was written by a bad coder.  The code doesn't make it easy to determine overall architecture at a glance, explicit block structure is missing in places, goto statements and logic flow issues. It's a mess that makes it difficult to change anything without breaking something. This is essentially keeping anyone from changing it.

This is really well done.  What do you think of the solution proposed by Hacking, Distributed the two phase proof of work?

I would love to see your explanation on how bitcoin is going to become 80 billion market cap.
The thing is that even current 8 billion cap. is pure BS, no chance in hell you could even pull out 1 bilion (cashout).
The main reason big fishes are not playing with bitcoin is because its a one sided when ure talking about huge amounts, 1 way in , no way out

For a chain of 1 blocks, yes. The Original Bitcoin whitepaper lays out the probability that 10% of the hash-power will win chains of block races (z= number of blocks-1).
So 10% of the hash-power has a 20% chance of getting two blocks in a row, 5% chance of getting 3, 1.3% chance of getting 4. and a 0.3% chance of getting 5 (for food labeling they would just call it "0%").

10% attack can happen every 1 in 10 times.  so just sit back and watch. business as usual.

Because the devs don't know what to do. Bitcoin is proving to not be decentralized or trustless. Mining pools are necessary because no one will invest money into mining for the benefit of Bitcoin without the reward. Even without any single pool gaining a majority share of the network (51% is not required) damage can be done. It can be fixed (forked) but the cost in consumer confidence will be devastating. This is really the issue and it's being ignored.

Other trust issues revolve around an evil actor (govt, competing business) that could make multiple pool operators an offer that's too good to refuse. What would happen if I bought the the three largest pools away from the current operators? What would my share of the network security system be? There are multiple ways that Bitcoin has developed into a trust system just like the banking and finance industry. Mining is only one of them. Convenient thin clients are concentrating the full nodes into fewer hands every day. Eventually electricity and equipment costs will keep all but the seriously motivated from burning a 24/7 node. Cloud nodes are a solution but who is really in control of the network then? Bitcoin developers are an elite group of highly educated coders that are maintaining a C++ language (how you can tell Satoshi was not in finance, probably should have been Cobol) that few coders even bother learning. Just like most coders don't know how to code PostScript most wouldn't know the first thing about Bitcoin. It's the specialized knowledge base that makes the small group of developers another point of trust centralization. It's kind of funny that Bitcoin has developed into exactly the thing that it was invented to overcome.

Ha, but you don't actually have to have 8 billion to move bitcoin. That's just the market cap. It doesn't actually mean that 8 billion have been 'put' into bitcoin! Even a measly 2 million could easily move the market a lot. Just take a look at the depths graphs!

I think it's just like Congress in the US.  No one will do anything about it until it becomes a major problem, and by then it would be too late.  Some of the ideas people have just won't happen while there is a chance of a 51% attack.  For instance, all these people saying, oh BTC could become a world currency reserve...well no government is going to do that when there is a chance it could be destroyed for a relatively small amount of money.  There's a reason why Fort Knox is so heavily guarded.

At 8 Bil market cap it's unlikely any 'billionaire' would devote that much dough to attacking the nodes. The mega-billionaires aren't dumb, now some billionaire aires are dumb as wood, but their fortunes aren't that big compared to the new mega-billionaires today.

The next value growth bubble will put btc into the 80 bil market cap, and at that point its beyond one nuts reach no matter how rich, he would be better off trying to just urinate in the ocean and see if his volume has any effect.

It will choose the fork that uses the original protocol unless it is upgraded to new code that chooses the fork with the updated protocol.


That depends on which fork you want to be on, but in general if there is an intentional fork in the bitcoin protocol to fix some issue, and you want to be operating on the new fixed protocol, you'll need to update.


You agree by choosing software that implements the protocol that you agree with.  99% of the altcoins are all essentially bitcoin forks that fork off at the genesis block of the blockchain.  If you agreed with the decisions made by the developers of those forks, you'd choose to run the software that was written for the fork you liked (litecoin? dogecoin? etc). By choosing software that is compatible with the current bitcoin protocol, you have chosen to enforce the current rules of the bitcoin protocol.

If the bitcoin protocol were to fork, and some people chose to continue running wallets that supported the old protocol rules, then there would be an "old bitcoin", and a "new bitcoin".  Since the blockchain would fork later than the genesis block, any bitcoins generated prior to the fork would be spendable on both forks.  Bitcoins generated after the fork would only show up on the fork where they were generated.  If a single transaction spent pre-fork bitcoins and post fork bitcoins together, then all of those bitcoins would only be spendable from the side of the fork that the post-fork bitcoins came from, from then on.


The reason that your multibit wallet only sees bitcoin transactions, and not any other altcoins, is because it is designed to implement the current bitcoin protocol.  If there is a fork in the bitcoin protocol to fix some issue, and you want to continue to operate on that new fork, you would need to wait for the MultiBit developers to update their software, and then you'd need to upgrade.

If a pool starts behaving badly with its miner's hashing power, miners will leave that pool. Simple. A rich kid wrecking Bitcoin? Maybe DDOS the network and stop it in its tracks for a few hours. And what would he achieve?

Think the other way around: Why would it happen and stay 51% for a long time? Miners are not going to destroy Bitcoin - It's their investment at stake. The mining power is not controlled by large mining pools, merely "administered".

Im curious what will happen with my multibit client when a fork happens? Does this mean that multibit also needs to be updated aswell as Bitcoin core ?

When we say everyone agrees, how do I agree? What do I need to do, I don't run a full node nor do I use Bitcoin Core wallet.

Why do you seem to think nothing is being done about it...

BItcoin is of strategic importance. That means that the US, Great Britain, France, Russia and China must all agree that Bitcoin is bad before directly attacking it: otherwise, they may spark a new arms race.

Absolutely! 51% Attack can't make you earn anything and make you lose $50 million. Then the  fork chain 1 and original chain are existing at the same time. The majority of miners will know which chain is created by evil people and choose the right chain.   

Interesting... I had been thinking lately about 2/3 vote... seems to be some kind of universality to it (2/3 vote in congress, etc).

Obviously this is incompatible with Satoshi's scheme in its current incarnation because
Bitcoin not only requires distributed consensus, but ALSO is intolerant to any
stagnation or inactivity -- we MUST have a decision roughly 10 minutes or
else blocks won't get published.

The reason nothing is being done about the 51% attack is that nothing can be done about it other than scrounging as much hash-power as possible to point at the network (making the attack more difficult).

If you believe in Bitcoin, you believe that people are generally good. If the "good" hash-power outnumbers the "bad" hash-power, Bitcoin works just fine. If "bad" hash-power wins out over good, Bitcoin fails: at least until evil gets bored.

There are no "technical" fixes to this: only social ones. The reason is that trusting the majority of the hash-power is how Satoshi solved the distributed consensus problem. Somewhat relevant: Byzantine fault tolerance. According to that page, for most algorithms, you must trust 67% rather than 51% of the actors.