Why the developers of bitcoin don’t want to fix a bug Time Warp
Blockstream co-founder mark Friedenbach during his recent studies under the name of Forward Blocks found that a long-known bug of bitcoin, which potentially allows to implement a Time Warp attack (attack time warp, or TW— attack), can be used to scale the network the main cryptocurrency. The author of the concept says that if it is approved the developers will attract more users and process transactions much faster.
However, the proposal of Friedenbach caused a new debate around the TW attack. A malicious miner can implement it by mining the last block before the allocation of complexity of mining: an unscrupulous player assigns the unit about the time stamp, thereby reducing complexity. This vector was intensively discussed in the developer community of bitcoin back in 2012, but due to the complexity of the implementation (due to recalculation every two weeks with a great herate the probability of success is negligibly small), this attack is considered to be impracticable.
Greg Maxwell, Blockstream co-founder and one of the most famous developers of bitcoin, in a recent letter to my colleagues suggested to fix this bug. He has not commented on the proposal Friedebach, although a letter was sent out on the background of the discussion surrounding his research.
Friedenbach, describes the concept of increasing the bitcoin block without the need for software updates network members (some believe this idea a breakthrough). Along with this research has drawn even more attention to the possibility of eliminating bug of a Time Warp.
First you need to understand why this threat still exists.
Some participants (miners) report the time when the network has occurred any event (a committed transaction or create a new block). There is a small possibility of manipulation of time even without breaking the rules of the bitcoin code, compliance with which is constantly checking the nodes of the network. Thus, miners can report a block with the wrong timestamp.
It is worth mentioning that bitcoin does meet periodically blocks c timestamps not in chronological order. To find out why this is happening, analysts from blockchain companies Chainalysis prepared a report that analyzed how changing the frequency of such anomalies.
The report shows a reduction in the frequency error time stamp that reflects the evolution of mining in the network. Philip Gradwell, chief economist Chainalysis and one of the authors of the report, States that, according to collected data, the error of timestamps in bitcoin look like “spikes” during technological shifts in the mining industry. For example, when miners started to band together into pools in early 2012, the timestamp error began to occur 8% more often. Gradwell says: these data indicate that the errors are random, not created by criminals; the miners just need to “get used” to the new equipment.
TW— attack requires very precise manipulation by the miners, who hope to earn way more money. Attack is possible when the miners unite and misleading incorrect timestamps for the purpose of manipulating difficulty of mining. Maxwell says:
For me and many others this issue was not a priority, because such an attack requires Hasrat most of the miners and in the case of can be easily stopped.
Even if one group of miners will collect most of Hasrat, TW— attack will be the least threat to bitcoin. According to Gradwell, then there is other problems like the risk of attack 51%.
Looking at a TW attack as a problem, which if necessary can be easily fixed. In his letter to the developers of Maxwell mentioned that he has an idea that he checked in testnet bitcoin a few years ago, but he won’t dwell on it:
Before I blow away the dust with his old ideas and may be premature to focus on a specific approach, I would like to know about other possible backwards compatible solutions to eliminate TW attack.
The term “backward compatibility” here is critical. The essence of this characteristic is that the update should not assume the possibility of partitioning the network.
In response to the letter from Maxwell several people have offered their solutions. Bitcoin Core developer Johnson Lau described a few ideas (good and bad) to find a compromise approach. According to him, the “naive” solution is to prohibit new block to show an earlier timestamp, but it can lead to the fact that the software of bitcoin will be divided into two versions. He says:
The goal is to find a short time period to prevent TW-attack sufficient to prevent separation of the network.
He added that this approach can be implemented in a weaker version of the “naive” approach.
The idea of Lau caused a bit of discussion about softforce — backward compatible way of amending the code of bitcoin. BitTorrent Creator Bram Cohen says that, in General, softforce represent the best solution because they are not “thrown overboard” miners who chose not to upgrade. He was supported by Lau, but his proposal for a three-hour window called controversial. Another developer, Scott Robert, proposed an alternative solution, but it is, in his own words, was wrong on bitcoin. Robert agrees with the opinion of Cohen, though he believes that the three-hour interval is “too narrow”, and adds:
I don’t know what decision will eventually be made, but I think that the fix should be as simple as limiting time stamping the plus-minus 24 hours from the previous label. Another idea
On the other hand, the elimination of the possibility of TW attack completely destroys the concept of Forward Blocks. Friedenbach says:
Elimination of TW’s attacks will make it impossible to implement the concept of scaling the Forward Blocks. Perhaps we should deploy the update Proof-of-Work, or to increase centrosoyus with sharding, but the main advantage of scaling of bitcoin in this case, it will disappear.
Thinking about it, Friedenbach made another proposal that maintains the concept of Forward Blocks, but excludes the “worst exploits” TW— attack. He stated that this proposal should deploy as early as possible to prevent the use of the Time Warp bug.
However, many developers of bitcoin agree that the concept of Forward Blocks it is not necessary to implement in its pure form. CEO of Blockstream’s Adam Back says that, considering this study to be interesting, he is not sure that it will support community:
I think it would be useful to study the technical possibilities that mark offered, but the main question is whether a compromise between decentralization and centrosoyus for brute force scaling of the first level
And, although the proposal Forwawd Blocks interesting in itself, because it increases the throughput of bitcoin without the need for hard forks, the type of change being proposed maintains a high probability of partitioning the network. Since the changes that could reduce the decentralization of the network, always cause in the community particularly fierce debate, Buck argues, such a proposal will not be accepted. He adds that “there are probably more straightforward and less hacking approaches” to scaling bitcoin than the concept of Friedenbach. He Friedenbach continues to assert that the concept Forwad Blocks deserve to keep it as a tool:
Dangerous consequences of the bug Time Warp can be prevented, and does not correct the error completely and thus not writing off from the accounts of Forward Blocks and other solutions for scaling bitcoin.
According to the materials CoinDesk