|
On the morning of September 10th, Beijing time, the Filecoin team once again launched an AMA event for the community. The mainnet launch window will end on September 28, 2020. The following is the full transcript: Q1: When will the mainnet be launched? A1: You can check Filecoin's latest roadmap for August 2020: ( https://filecoin.io/blog/roadmap-update-augut-2020/ ), The latest timeline can also be found in the public project route Gantt chart: ( https://app.instagantt.com/shared/s/1152992274307505/latest ) The tentative mainnet launch window will end on September 28, 2020, of course, this time will be revised. Q2: When transferring funds from miner to worker, I get this error message: "Get key address: Address t010372 is not for an account executor?" lotus-daemon import –snapshot acquisition error: The gap between the information random number and the expected random number is too large (Nonce: 675457, nextNonce: 631868): Unimplemented random number gap?" A2: This is not an error you need to worry about and has nothing to do with importing from snapshots. While syncing, your message pool cannot determine if the incoming message is valid, so some message pool errors may occur. You can safely ignore this error. Q3: Will the algorithm change after the main network is launched? A3: After the mainnet is launched, we will upgrade the network to introduce new proof mechanisms, user-defined contracts, node specification improvements, and more. Q4: Under Filecoin’s economic model, how do small and medium-sized miners survive? A4: Same with big miners! Economies of scale aren’t actually that significant in Filecoin — its reliance on commodity hardware means that large players don’t have the same scale advantages as some other networks. There’s no doubt that mining Filecoin right now requires active technical management, which may favor better resourced teams, but this is improving rapidly and will continue to improve in the future. If you have ideas for improving the miner UX and helping small miners with this developer incentive program, please contact us! Q5: Protocol Labs team, thank you for your support during the competition. A5: Thank you and your team for the hard work you do to help build a secure, resilient, and strong network. Q6: How to transfer funds from miner to worker? A6: The "lotus-miner actor withdraw" CLI command will withdraw all available funds to your owner address. You can then transfer these funds to your worker or control address with a simple send. Q7: As the third week draws to a close, are there any celebrations after the Space Race? A7: Yes, we will hold a closing ceremony next Tuesday, September 15 at 7:00 am Beijing time. More details will be released here soon: https://spacerace.filecoin.io/events Q8: We would like to know the fees/costs on the mainnet? How much FIL a miner needs to have on hand to start working? A8: Assuming you mean gas fees and the cost of the main network, the base fee will be determined by the market, that is, the balance between the fees processed by the network and the fees participants are willing to pay. On top of this there is a gas surcharge, which is the priority fee that the message sender is willing to pay and is specified by the sender. We are also exploring ways to improve the allocation of chain bandwidth resources. Q9: Will the Protocol Labs team come to Shanghai before the mainnet launch? A9: The Filecoin team is excited to meet with the Filecoin community around the world once travel restrictions are lifted and it is safe to do so. In the meantime, we will host and participate in virtual events such as Shanghai Blockchain Week in October. Q10: What are the advantages of official hard disks in mainnet mining? Especially in the beginning, there may be only a few real users compared to miners? A10: There are no official hard drives. You may be referring to DumboDrop which uses hard drives as a cheap and effective way to transfer data offline, but hard drives themselves do not give miners any additional advantage. DumboDrop is one of many certified clients, and transactions from certified clients will give miners greater effective hashrate. Verified client transactions will go live once details are released. Q11：How many sectors can be processed simultaneously with 128gb or 256gb of RAM on the same mining machine in precommit #1? We have a lot of swappable stuff. A11: Currently, each PreCommit1 (for 32G sectors) requires 56GiB of available RAM, 64GiB of available RAM + Swap. This can be adjusted in extern/sector-storage/resources.go. Q12: Which parts/stages of sector encapsulation can be done in parallel on a single machine? A12: All except PreCommit2 and Commit2 are OK, but we are working on making them run in parallel with other things. Q13: Can the mainnet be launched in October? Did this space race achieve the expected results? A13: The competition has greatly exceeded our expectations. Note that we are testing the Space Race network very robustly – there have been network upgrades, outages, and a very large number of transactions (nearly 100,000 so far!). Despite this, the network continues to grow at twice the expected rate, and many miners are seeing close to 100% transaction success rates even before the scoring adjustments. We have also learned about many bugs and made UX improvements to make it easier for clients and smaller miners. Q14: The hardware requirements are very high now. Have you seen any major improvements after the testnet? A14: We are actively researching more efficient PoRep algorithms. For example, the path upgrade of SDR-NSE will greatly speed up the packaging speed and efficiency, which in turn will allow miners of all sizes to add more storage space, and the network's overall time to exceed the baseline will be faster. There are other benefits, such as better retrieval time. Q15：Do you have enough funds to support the ecosystem? Because if the profit cannot cover the cost, miners will not stay. A15: As some miners leave, the remaining miners will gain greater profits, so that a new balance can be reached. In the long run, participants should not regard protocol subsidies as their main income. The goal of all participants should be to provide high-quality storage-related goods and services to attract storage demand and transaction payments. Q16: Asia seems to be very active in Filecoin mining, yet the rewards are the same across all continents. Is this really fair? Will there be users in Africa? A16: In fact, the reward pools are balanced to deal with this situation - the global reward pool is currently the largest and treats all participants equally, while smaller regional reward pools provide moderate incentives to promote mining around the world. Q17: Your official miners have higher luck than other miners. Why is that? Is there anything you can share with miners? A17: We do not have official miners. Having a reliable network connection, establishing on-chain miner information and control addresses, and ensuring that PoSt is submitted are key steps. Q18: Given the current shortage of FIL, is there an updated configuration guide? This is the current default value: MaxWindowPoStGasFee = "50 FIL" A18: Chain bandwidth is a limited resource. Message senders allocate resources through GasFeeCap bidding. Given that participants have a large number of test coins, they tend to raise gas fees, so MaxWindow-PoStGasFee is introduced as the default upper limit to avoid excessive distortion of the fee market. Sometimes, when the chain is clogged, it's a good idea to let the network wait for messages to be processed. With the gas fee update, it looks like miners have managed to increase MaxWindow-PoStGasFee by 3-5 times. We are looking into ways to improve the allocation of chain bandwidth resources. Q19: When mining, some people say that AMD CPU is better than Intel CPU. Is it true? A19: For now, yes. Filecoin's current Proof of Replication (PoRep) prefers to run on AMD processors. More precisely, it runs much slower on Intel CPUs. It runs extremely fast on some ARM processors, such as the ones on new Samsung phones, but these processors do not have enough RAM to seal larger sectors. The main reason we see an advantage on AMD processors is because they can implement SHA hardware instructions. Q20: Is it profitable to join the network as a miner now? Will there be a difference in where the rewards will be during the mainnet period? A20: Miners must answer questions about profitability themselves, but the protocol is designed to continuously and sustainably incentivize miners to participate in the network. There will be verified clients on the mainnet with different storage requirements, and many clients may prefer storage located in different geographical locations with different retrievability guarantees. Q21: Are there any plans to integrate Ethereum domain name service to make it easier to make immutable web domain names? A21: We are working on a number of integrations with partners in the Web 3.0 space and are excited to announce more soon. Q22: How long does lotus daemon import-snapshot take? A22: Note that the command is “lotus daemon –import-snapshot”. The command does not take long to run, but you will first need to download a very large file. This may take several hours, depending on your network connection. Q23: Are you satisfied with the participation of global miners in the space race? Do you think the proportion of Chinese miners is too large? A23: We certainly hope to have more miners participating in every region to develop a healthy, distributed global network for distributed storage. That said, we are seeing amazing activity in many parts of the world. We have some plans in the works to encourage further decentralization, but are very grateful to Filecoin's many enthusiastic users so far. Q24: Our miner has been offline for more than 24 hours and is no longer on the dashboard. Our problematic sectors are now recovering. How long will it take for them to reappear on the dashboard? A24: Usually you will recover the problematic sectors in the next successful PoSt. The dashboard should re-display your node soon. Q25: Do you have any official recommendations for mining hardware now? It was not clear last time. But you also participated in the test, any news? A25: Miners have had success on a wide variety of hardware. Watch the Miner Show & Tell to learn more: https://youtu.be/ jO5hJiwKAjA & https://youtu.be/pHi85ZNs3AY We strongly recommend using an AMD processor instead of an Intel one. Click here to learn more: https://docs.filecoin.io/downloadtion/faq/#why -does-filecoin-mining-work-best-on-amd Q26: What do you think about PoSts greater than 1x per sector per day? We don’t want customers to lose data access links within 24 hours if it happens to close the PoSt window for a few minutes. A26: Data access is actually independent of PoSt submission - PoSt is how the chain knows that data is being stored, but retrieval requests occur directly between clients and miners. ——End—— |
