Topic: Solution to a 51% attack? - page 3. (Read 5482 times)
Nice to see some constructive criticism. All this will be usefull to improve new update.
There is none, high network hashrate would reduce the chances
Who ever wrote this, seriously needs a basic tutoring in exactly what a 51% attack is....The whole thing is so flawed I don't even know where to begin.
A 51% attack in no way has to involve a Double Spend on an exchange. I can kill your block chain with a 51% attack and never spend a single coin. It seems to me the author doesn't understand what a 51% attack is or the difference between a 51% and a Double Spend attack. They are entirely two different exploits and neither one will be limited by a 5 blocks in 10 minutes limitation....
Also let's not forget about either one of the Time Warp Exploits, the ArtForz version from 2011 or the new and improved BCX version that no one yet has seemed to mitigate at all.
I like Goldcoin and I like Microguy but if you guys keep posting basic stupid crap like this, someone will come along and shred your coin for you just for the hell of it and no it won't be me. I can think of about 5 or 6 right off the bat that would do it just for the challenge.
~BCX~
Where can I read more on this? I understand that double spend attempts don't even necessarily require 51% of total network. It's just that with 51% you are statistically guaranteed success. Right? But how does a 51% attack kill the blockchain? A 51% attack in no way has to involve a Double Spend on an exchange. I can kill your block chain with a 51% attack and never spend a single coin. It seems to me the author doesn't understand what a 51% attack is or the difference between a 51% and a Double Spend attack. They are entirely two different exploits and neither one will be limited by a 5 blocks in 10 minutes limitation....
Also let's not forget about either one of the Time Warp Exploits, the ArtForz version from 2011 or the new and improved BCX version that no one yet has seemed to mitigate at all.
I like Goldcoin and I like Microguy but if you guys keep posting basic stupid crap like this, someone will come along and shred your coin for you just for the hell of it and no it won't be me. I can think of about 5 or 6 right off the bat that would do it just for the challenge.
~BCX~
The solution to a 51% attack is never let one entity amass that amount.
It doesn't sound legitimate at all. It sounds like the writing of someone who doesn't even have a basic understanding of the problem space.
It ignores the existence of sybils, assuming that an attacker is confined to appear to be a single peer to the whole network. This is crazy, If not for sybils there would be no need for mining at all you could just have nodes produce blocks round-robbin or just directly sense a majority of nodes for consensus.
Even the false assumption that sybils don't exist isn't enough for this scheme, as it requires (see the last page) a centrally announced "checkpoint" to make the protection reliable as nodes enter and leave the network. Moreover, it doesn't describe how it's mechanism wouldn't falsely partition the network in the event that some kind of topology disruption resulted in the network having a small min-cut (e.g. fiber cuts between continents), this wasn't adequate explained but many proposals along these lines in the past have resulted in schemes which would never re-converge again after a modest partitioning: from one currency two result, and all coins can be spent twice (one in each side).
It's funny because not only would this proposed system not even work, the goons who proposed it don't even have the technical skills to put it into place to begin with. It ignores the existence of sybils, assuming that an attacker is confined to appear to be a single peer to the whole network. This is crazy, If not for sybils there would be no need for mining at all you could just have nodes produce blocks round-robbin or just directly sense a majority of nodes for consensus.
Even the false assumption that sybils don't exist isn't enough for this scheme, as it requires (see the last page) a centrally announced "checkpoint" to make the protection reliable as nodes enter and leave the network. Moreover, it doesn't describe how it's mechanism wouldn't falsely partition the network in the event that some kind of topology disruption resulted in the network having a small min-cut (e.g. fiber cuts between continents), this wasn't adequate explained but many proposals along these lines in the past have resulted in schemes which would never re-converge again after a modest partitioning: from one currency two result, and all coins can be spent twice (one in each side).
What stops me from doing this:
1. Control a large number of nodes.
2. On all except one, broadcast a chain that is always 1 block behind the main chain.
3. Suddenly from the last one release 4 blocks in quick succession. Now you win by 3 blocks (and don't trigger the 51% defense).
1. Control a large number of nodes.
2. On all except one, broadcast a chain that is always 1 block behind the main chain.
3. Suddenly from the last one release 4 blocks in quick succession. Now you win by 3 blocks (and don't trigger the 51% defense).
It doesn't sound legitimate at all. It sounds like the writing of someone who doesn't even have a basic understanding of the problem space.
It ignores the existence of sybils, assuming that an attacker is confined to appear to be a single peer to the whole network. This is crazy, If not for sybils there would be no need for mining at all you could just have nodes produce blocks round-robbin or just directly sense a majority of nodes for consensus.
Even the false assumption that sybils don't exist isn't enough for this scheme, as it requires (see the last page) a centrally announced "checkpoint" to make the protection reliable as nodes enter and leave the network. Moreover, it doesn't describe how it's mechanism wouldn't falsely partition the network in the event that some kind of topology disruption resulted in the network having a small min-cut (e.g. fiber cuts between continents), this wasn't adequate explained but many proposals along these lines in the past have resulted in schemes which would never re-converge again after a modest partitioning: from one currency two result, and all coins can be spent twice (one in each side).
It ignores the existence of sybils, assuming that an attacker is confined to appear to be a single peer to the whole network. This is crazy, If not for sybils there would be no need for mining at all you could just have nodes produce blocks round-robbin or just directly sense a majority of nodes for consensus.
Even the false assumption that sybils don't exist isn't enough for this scheme, as it requires (see the last page) a centrally announced "checkpoint" to make the protection reliable as nodes enter and leave the network. Moreover, it doesn't describe how it's mechanism wouldn't falsely partition the network in the event that some kind of topology disruption resulted in the network having a small min-cut (e.g. fiber cuts between continents), this wasn't adequate explained but many proposals along these lines in the past have resulted in schemes which would never re-converge again after a modest partitioning: from one currency two result, and all coins can be spent twice (one in each side).
Seems someone came up with a solution to a 51% attack with goldcoin.
http://gldcoin.com/documents/GoldCoin_0.7_51percent_defense_october_11_2013.pdf
I'm not an expert in cryptography at all but does it sounds legit and if so, is it applicable to bitcoin?
http://gldcoin.com/documents/GoldCoin_0.7_51percent_defense_october_11_2013.pdf
I'm not an expert in cryptography at all but does it sounds legit and if so, is it applicable to bitcoin?
Jump to: