Jan 12, 2019
Proposal to shorten the block time of BCH
By Prof. Liu Changyong
The fourth edition
February 31, 2018
Since Aug 2017, BCH have reached a consensus on the development path of evolution to adapt to the market at a great cost of two splits. This allows us to concentrate on discussing and implementing some important updates.
Based on the market demands and community opinions, this proposal analyzes whether the block time should be shortened. The recommendation is to shorten the block time to 1 or 2 minutes as much as possible in 2019.
1 Objective/Purpose
As a peer-to-peer electronic cash, the core of BCH’s user experience is to enable users to realize BCH transferssafely, conveniently and economically. The main purpose of shortening the block time is to improve the convenience of transfer.
BCH and BTC block time is 10 minutes. Taking into account of the contingency, the guaranteed time for one confirmation is approximately 30 minutes. This is one of the defects of long-term criticism since the birth of Bitcoin. In the early days, most of the altcoins relied on solving this defect and obtained the first users. Both Litecoin’s 2.5 minutes and Doge’s 1 minute have achieved long-term stable operation, and the On-chain transaction volume has surpassed BCH’s.
Shortening the block time will improve the user experience for the three main applications of BCH:
(1) Transfer
When the demand for transfer in the market surges, especially in the bull market where users are flooding in, the demand for transfer mainly comes from the exchanges. The faster confirmation time is more attractive to the users of the exchanges, and the exchange is willing to reduce the number of confirmations in order to compete. At this time, the shorter block time has obviously more advantages. In the case of Bitcoin congestionin March 2017, the number of users, transaction size and Legal currency prices of ETH and LTC increased rapidly.
BCH must quickly increase its market share when the next buill market of cryptocurrency is coming (probably in 2020), otherwise BCH is likely to lose the chance of winning the cryptocurrency war.
(2) Blockchain certification
Using OP_Return to record witness information on the BCH chain is an important application. The fast confirmation time allows users to see the information stored in the chain faster and improve the user experience. Due to the limited space of OP_Return, some information needs to be split into multiple transactions. At present, the memory pool only accepts ancestor transactions within 25. When the information is split more than 25 copies, it must be stored in two or more blocks. Shortening the block time is more valuable for Blockchain certification.
(3) Token
The market development in 2017 has proved that the sub-currency or token issued on the public chain has a wide market demand, which can bring rapid growth of users. In particular, the emergence of stablecoin has solved the defects of the current Cryptocurrency price fluctuations affecting the payment function, and is an important transition from the legal currency economy to the cryptocurrency economy. Shortening the time of the block can greatly facilitate the issuance and circulation of tokens and stablecoins.
Since BCH has established a iteration to improve the user experience, the market has given an important evolutionary window in 2019. Shortening the block time is an upgrade in priority.
Of course, we need to consider the problems that may be faced in shortening the time of the block carefully.
2 Problems
Whether or not should BCH shorten the block time has been fully discussed for a long time, especially in Chinese community. The main problems that people reject or worry about are as follows:
(1) If shortening the block time affects the total amount of 21 million?
Worry: Shortening the time of the block will cause the coin to be mined faster and generate more than 21 million coins.
Answer: This worry is unnecessary. If it is shortened to 2 minutes, the amount of each new block reward will be reduced to 2.5. If it is shortened to 1 minute, it will be reduced to 1.25. It does not affect the total amount, nor does it affect the issuance speed and miners’ income, and does not make anyone’s benefits gain or lose.
(2) Is there any essential difference between 2 minutes and 10 minutes?
Worry: Regardless of 10 minutes or 2 minutes, the user needs to wait for confirmation. There is no essential difference. Also Considering the contingency of the block, it usually takes 10 minutes to actually wait for one confirmation if shortening to 2 minutes.
Answer:
1) Anyone can easily experience a significant difference between waiting for 10 minutes and 2 minutes. Please close your eyes and count to 120, then to 480.
2) Considering the contingency, the actual block out of the 2-minute block may take 10 minutes, but in the case of 10 minutes, it often takes 30 minutes. In fact, Bitcoin’s 10 minutes is more volatile than the Litecoin’s 2.5 minutes and the Doge’s 1 minute. See figure below:
3) More confirmations are more stable during the same time period. Statistically large samples converge more to the mean value than small samples. Therefore, the total time of five 2-minute confirmations must be closer to 10 minutes than the actual time spent on a 10- minute confirmation, which can significantly improve the user experience.
(3) Does the exchange increase the number of confirmations to offset the advantage of shortening the time of the block?
Worry:Assuming an exchange now requires 6 confirmations, if the block time is reduced to 2 minutes, the exchange will require 30 confirmations. So, the shortening is completely meaningless.
Answer: The number of confirmations required by the exchange will not increase multiplied by a fixed ratio.
. They will not strictly determine the number of confirmations based on the total time. The exchange is concerned with multiple factors such as double-spend attack, user experience, influence of the coin, and short-term risks (such as upgrades). Under the premise of security, the exchange is willing to reduce the number of confirmations as much as possible to improve the user experience, and respond to the fierce market competition.
For example,the Huobi.com requires 1 confirmation to confirm BTC deposit, according to the principle of equal time, ETH should be 40 confirmations, but the actual number of confirmations of ETH is only 15. The exchanges only require 1–6 confirmations before BCH split. During the war, the exchange requires 30 or even 144 confirmations. When BCH applied Replay attack protection, the exchange only requires 10 confirmations against double-spend attack.
With the end of the battle, the risk of 51% attack is declining. The exchanges will not require 5 times the number of confirmations when BCH is shortened to 2 minutes, but will be less.
After all, they are businessmen who are making money, not students taking math exams.
(4) Will shortening the block time make BCH’confirmation experience surpass the experience of ETH and EOS?
Worry: Even if we shorten the block time to 2 minutes, or even 1 minute, it is completely impossible to compete with ETH’s 15 seconds or EOS’s 3 seconds.
Answer: The user experience and market competitiveness are comprehensive and it is impossible to exceed all competitors in all indicators. The competition of cryptocurrency should compete in many aspects such as security, stability, convenience, and scalability. These aspects are often difficult to achieve. We can only weight and make improvement to get greater multi-faceted improvements at a small cost.
The advantages of BTC and BCH are mainly in terms of safety and stability, and there is a loss in convenience and scalability. The shortening of block time and the wormhole-like Layer 2 protocol is sacrificing a small price in security and stability, in exchange for greater convenience and scalability.Our goal is not to exceed ETH and EOS in confirming time, but to improve our user experience and market competitiveness.
(5) Does shortening the block time increase the orphan block rate?
Worry: Shortening the block time will cause more orphan blocks and reduce the stability of the system when the block broadcast and confirm time stay the same.
Answer:
1) Under the same conditions, the block time will be shortened, and the orphan block rate will increase, but how much it would be increased? Does it affect security? Is the price worth it? These are the decisions that need attention.
2) Improvements in network and hardware conditions. In the 10 years since 2009, the global network and storage technology has been greatly improved, which will inevitably allow the shortening from 10 minutes to lower under the premise of the orphan block rate.
3) Mining specialization. In recent years, mining technology upgrading and mining competition have promoted the specialization of mining. The network and storage hardware conditions and software technology of the mining node have been greatly improved compared with the previous ones, and the block time can be shortened under the premise of the orphan block rate is constant.
4) Other technologies are reducing the orphan block rate. The main purpose of CTOR, Avalanche, and graphene technology is to improve node synchronization efficiency and reduce orphan block rate. With the upgrade of these technologies, there will be more space for shortening the time of the block.
5) Shortening the time will reduce the size and number of transactions per block, reducing the time required for propagation and verification without causing a proportional increase in the orphan block rate.
6) The orphan block rate of the POW based coinsthat have shorter block time such as LTC and Doge is not high.
Taking all of the above factors into consideration, BCH does not shorten the time to the second level sharply, but considers 2 minutes or 1 minute. The actual risk and cost of increasing the orphan block rate is not high. LTC and Doge have made successful 5years test to show us that a pow system with 2 or 1 minute block time is safe.
(6) Is it possible to strengthen the mining centralization with shorter block time?
Worry:Shortening the block time leads to an increase in the broadcast time. The large pools can save broadcast time because they can generate blocks continuously,, and gain mining advantages, which will eliminate small mines in competition, and leading to centralized mining.
Answer:
1) The above worries do give a logic for large pools to gain advantage after shortening block time, but also to consider weighing other logics.
2) The main significance of mining pool is to smooth out the luck of mining. Shortening the block time will make the block denser in unit time, reducing the risk of bankruptcy of small mine pools caused by fluctuations in the luck , which is conducive to the entry and survival of small mine pools.
3) Mining centralization is a long-standing problem of POW. The key problem is not the block time, but the trend of market competition and specialization. Dynamically, the long-term monopoly mining centralization is not stable. From Ghash.io, F2Pool, to AntPool, the centralized mining pool is replaced in competition and cannot obtain long-term absolute control. Enterprises that temporarily hold the advantages of centralization are not willing to launch 51% attack for their own interests.
4) Shortening the block time will enable the mining node to distribute the block verification task more evenly. For example, assigning a 10-minute verification task to five 2-minute ones can improve mining efficiency. According to this, the optimized mining pool will gain a competitive advantage. This advantage of improving overall efficiency is worth encouraging.
Shortening the block time has advantages and disadvantages for mining centralization, but it is not the decisive factor of mining centralization, and it should not be the main consideration factor for shortening the time of the block.
(7) Does shortening block time make the DAA more divergent?
Worry: The BCH adopts Difficulty Adjustment Algorithm (DAA). Shortening block time would affect the current DAA model, resulting in increasing fluctuations in the time of the block?
Answer:DAA is designed to allow the block time to be quickly adjusted back to the target time by the difficulty adjustment when hash power swarm in and out.There are two possible options for DAA after shortening block time:
1) Still refer to the previous 144 blocks to adjust difficulty. The current DAA is that each block is adjusted according to the previous 144 blocks (blocks in one day). After shortening to 2 minutes, if you still refer to 144 blocks, it will refer to the blocks in 5 hours. When faced with hash power swarming, it can adjust back to normal difficulty faster than before. Since the number of referenced blocks does not change, the degree of smoothness of the difficulty adjustment is the same as that of 10 minutes, and does not cause an increase in the fluctuation of the real time of block.
2) Still refer to the blocks in a previous day. This requires the DAA to modify the referring amout of blocks from 144 blocks to 720 blocks. In this way, the convergence effect of hash power swarming is equivalent to that of 10 minutes, but due to the reference of the number of blocks increased 5 times, the fluctuation of the block will be significantly reduced.
For the above reasons, regardless of whether the DAA follows the shortening of block time and adjust the number of reference blocks, shortening block time can make the block more stable. However, we may not consider other factors, so developers need to measure and test the impact on the DAA based on specific adjustment formulas.
(8) Does shortening the block time affect the Time-lock transaction?
Worry:Shortening the block time will cause the transaction with Time lock to be unlocked in advance.
Answer: There are two ways to lock the time. One is to lock the transaction according to the absolute time. This situation is completely unaffected by the shortening of block time. The other way is to lock the transaction according to the block height. Shortening block time will lead to the remaining locktime reduced to one-fifth. This only affects individuals who lock in transactions and does not affect the overall interests. Should it be considered as a reason to refuse to shorten block time, it needs to look at the current number and value of the transaction locked by the block height?
(9) Is it necessary to shorten the blocking time if there is zero confirmation transaction
Worry: The zero-confirmation transaction is to reduce the time for people to accept BCH payment to the second level, and many technologies are also improving the security of zero confirmation transaction. With zero confirmation of the second level, it is unnecessary to shorten block time.
Answer:
1) The significance of POW is that the competitive accounting guarantees the security of payment. If there is technology that can improve the zero-confirmation security to the level of the usual block confirmation, then POW loses its meaning.
2) 0 and 1 are fundamentally different. Although some small or certain trust-assisted transactions can accept zero confirmation, most exchanges and merchants consider 1 confirmation as the minimum standard for “transaction completion”.
3) For a long time, we should not try to technically convince each merchant and user to confirm that zero confirmation is safe enough. This is difficult and unfriendly. We should comply with their custom and provide a quicker 1 confirmation.
4) Zero confirmation and shortened block time are not contradictory, and each is applied to an appropriate scenario, which can improve the user experience of BCH together.
(10) Does shortening the time of the block affect Avalanche and future block size increasing?
Worry: Avalanche helps to improve zero confirmation security and combined with other technologies, improves block synchronization efficiency, allowing for larger-size blocks to be disseminated, increasing block size, and shorten block time that can slow down this process.
Answer:
1) Shortening the time of the block is not inconsistent with the Avalanche which can improve synchronization efficiency and help to shorten block time. Avalanche requires several rounds of voting between the mining nodes to eliminate double-spend transactions. In order to avoid risks, the mining pool will choose not to pack uncompleted voted double-spend transactions. Therefore, shortening block time does not affect Avalanche.
2) Shorten block time could also increase the block size. If the time is shortened to 2 minutes and the block size is kept constant, the actual capacity is expanded to 5 times, and the capacity per 10 minutes reaches 32*5=160M. If you are concerned about the significant increase in the orphan-block rate, you can reduce the size of each block to 8M or 16M. At this time, the size will still reach 40M or 80M every 10 minutes, achieving size expansion.
3) It is more meaningful to shorten block time than to increase the block size. The actual utilization of BCH’s 32M block size is currently less than 1%. Releasing block size limits does not give users any experience boost. Shortening block time can make a significant improvement in each transfer and improve the experience of each user.
4) The conflict relies mainly on the time and energy of the developer. At present, developers have invested more time in Avalanche and increasing the block size, and shortening the block time requires developers to change their existing plans and invest time and energy. We are demonstrating that this adjustment is worthwhile.
(11) Is it against Satoshi Nakamoto’s intention to shorten block time?
Worry:The 10-minute block time was determined when Nakamoto created Bitcoin, and we should not change it.
Answer:
1) This is mysterious, and we need to take a rational look at Nakamoto’s decision at the beginning of Bitcoin. 10-minutes block is based on the network, computing and storage hardware conditions in 2009 to ensure the stable operation of the peer-to-peer system, and rounded up the results. Ten years have passed and the conditions have changed dramatically. This important parameter to the user experience should be adjusted.
2) Nakamoto should not be sanctified. Recently, BSV successfully split the community in the name of returning to Nakamoto, indicating that there is a tendency to sanctify Nakamoto Satoshi: It is believed that every detail set by Nakamoto is perfect and cannot be changed. Any problem that seems have to be adjusted has the deep meaning that our intelligence has not yet reached. This deviates from the scope of science and is not falsifiable. One of the achievements of this community split is the separation of this mysticism. We should treat the development of BCH more rationally and scientifically, and demonstrate how to do it instead of seeking the enlightenment and blessing of Nakamoto.
(12) Is it against community’s intention?
Worry: The 10-minute block has been implemented for 10 years, making it difficult for the community to accept shorter block times.
Answer: Community consensus is indeed an important obstacle, because changing people’s habits is very difficult.
1) We cannot give up the market in order to maintain the so-called “orthodox”. The important reason that some people are opposed to shortening block time is that if it is shortened, it will no longer be orthodox, and it will become an altcoin. Experienced the market in 2017–2018, especially the decline of BTC market share and the rise of ETH, LTC, XRP, etc., we must give up arrogance and conceit, seriously face the market, study hard, catch up, and surpass all kinds of competition coins. Nakamoto Satoshi and our goal is to achieve decentralized global free circulation of currency, the specific implementation must be constantly adjusted according to changes in the new situation.
2) The split war has made us see the danger of blindly pursuing “orthodox”.
This community split was due to that CSW splits two thoughts of“maintaining orthodoxy” and “adapting to the market”, which exist in the community and even anyone’s mind. CSW becomes the leader of “maintaining orthodoxy”, which has being a true dogma. Although CSW split the community, it also enabled BCH to get rid of the dogma. It is true that Nakamoto’s decentralized world currency ideal needs action, not sticking and waiting, and is more willing to accept market promotion and continuous improvement.
A survey of the BCH Chinese core community after the split showed that 83.8% supported shortening block time, and only 2.7% opposed it. In another Chinese community that retained some of the BSV supporters, the opponents reached 13.7%. The Chinese core community members who oppose shortening block time have basically separated to the BSV community.
3) The community can gradually understand and accept the necessary improvements. The Chinese community is also gradually accepting the suggestion of shortening block time. On November 30, 2017, Jihan Wu suggested in the Chinese community to shorten block time, and immediately met with opposition from most people. But after a year, most people turned to support him. The survey of Chinese core communities shows that more than 61% of the supporters of shortening block time were initially opposed and turned to the supporters afterwards. The opposition of the non-Chinese community to shortening block time may be relatively large. I believe that after sufficient discussion and given a certain amount of time and patience, most people can gradually understand the need to shorten the time of the block.
3 Suggestions
Based on the above discussion, the following suggestions are given:
