Do you still remember the last time Juan led the team in AMA? That was March 6th from 10:00-11:00 am.

The stars moved, and two months passed by accidentally.

Today, when Interstellar Vision hit Juan's third AMA again in the same time period, the second test has been successfully launched for 4 days. Looking back a little bit, it is inevitable that there will be some small emotions and sighs, and everyone who understands the feeling will understand it.

The official time given this time is still one hour, but judging from the number of miners’ questions, it is obviously much less than last time. A brief look at everyone's Ask, although some very core issues have not been seen, but the miners seem to have matured a lot.

When will Retrieval Miner appear? The miners think that the verified data is very concentrated and does not express the spirit of the blockchain. What do the officials think?

The official is still very "official" as expected, but we can still read some key information from the back of each set of answers:

1. The premise of the decision to launch the Filecoin main network is the security and stability of the network;

2. Filecoin project developedhttps://github.com/filecoin-project/sector-storagemodule, which is used by both Lotus and Go-filecoin. Miners are free to customize or re-implement the tool as they see fit;

3. Verification is not to verify the data itself, but to verify the client and trust the client for a certain amount of data. Solving the problem of "cheating" must be solved by incentivizing the network;

4. Regarding the allocation of 1 million FIL bonuses in the second phase of the testnet, the official recommendation is to set up mining operations to compete for storage and retrieval transactions from real users.

Q1: What is the current interopnet status of two interoperable filecoin implementations, filecoin and lotus?

Aayush: Go-filecoin and Lotus are currently interoperating on Testnet Phase 2. We found issues and patched them, but they are fully interoperable for now.

Q2: When will the detailed economic model parameters be determined?

A2: Detailed parameters will be released only after all core functions are completed and thoroughly tested on testnet. The current parameters in testnet are temporary placeholders.

Q3: When can I join the testnet? When are you actually launching testnet/2?

Aayush: You can join testnet phase 2 now! It started last Thursday, and you can join anytime.

Q4: What are the incentives for Testnet miners? The required/recommended hardware (2TB, 8-core CPU, 128GiB ram) is very expensive.

A4: Yes, there will be a testnet competition, and miners will have a chance to earn a lot of Filecoin. In order to participate, miners should be prepared to store and retrieve real files for clients and perform sector lifecycle upgrades.

Q5: If I store files on the Filecoin network and no longer wish to store them, how do I delete them?

A5: There is currently no planned mechanism to request or require a storage provider to delete data.

Q6: Is customer data stored directly in nodes or distributed after slicing?

A6: Customer data is directly sealed into the sectors stored by the mining nodes. Customers may, at their discretion, split the data into parts and/or store multiple copies with different providers. We anticipate the emergence of Filecoin client software that automates redundant long-term storage.

Q7: Most of the hard drive orders purchased by Chinese miners have been returned. Is this an official action? Will you continue to sell hard drives in the future?

A7: For questions about filecoin-discover, you can consult the official, or ask through the official Slack.

Q8: If the client wants to store 1GB of data, will it independently generate a 32GB sector to store it, or will it be pieced together with other users' data?

A8: It can be decided by storage miners at their own discretion. There is no explicit mechanism for clients to request dedicated sectors.

Q9: If user data is pieced together, metadata must be generated to mark the sectors and locations where the user data resides. So where does that metadata exist?

A9: The sector storage module used by Lotus and Go-filecoin stores metadata about sectors and conducts transactions in a key-value data store, separate from the sectors themselves

Q10: Will ASIC chips be officially supported? If supported, when will it start? Will the current need for the GPU go away during the sealing process?

A10: (by Juan) No one we know is developing ASICs. We specifically chose the PoRep algorithm to work with commodity hardware (rather than ASICs).

The only ASIC we are considering developing in the future is the VDF ASIC with Ethereum 2.

Note that PoRep and consensus algorithms may change and evolve after Mainnet, so ASICs may no longer work and go to waste - we target commodity hardware

Q11: How to define valid data, if a third party is used to view my data, there is a possibility of human manipulation, how to solve the problem of cheating?

A11: (By Juan ) This is an issue related to "verified customers" - which we haven't explained very well. First, look for a talk I'll be giving in the next few days explaining these mechanisms in more detail.

To answer the question in short, validation is not about validating the data itself, but validating the client and trusting the client for a certain amount of data. This is a pragmatic solution to the very difficult problem of distinguishing malicious fake transactions from real customer transactions and being able to incentivize the network to function correctly.

First, note that data from unverified clients is still good and profitable, and accepted by the network. Data obtained from verified customers may not be as lucrative as mined from it

Cheating is addressed by incentivizing the network to:

(a) There must be many validators,

(b) the verifier trusts the real client with at most some reasonable data based on the client's application (not infinite),

(c) The verifier itself is checking to do accurate work. If not, they will lose the ability to verify.

Q12: Is it confirmed that the mainnet will be launched between July and August, or may it be postponed again for the security and stability of the network?

A12: The security of the Filecoin protocol, implementation and network is our top priority. We will not launch Filecoin until we are confident it is safe.

You can check out our latest roadmap update for more information on delivery estimates. NOTE: These are estimates, launch has not yet been determined. We'll announce it on the blog when the final release date comes.

Q13: If the miner's hard drive has previously stored the data he generated himself, should the original data be overwritten after the verified user gets the transaction?

A13: The storage marketplace separates sectors that have "committed capacity" (no customer transactions) from sectors that store customer data. Committed capacity sectors will be upgradeable to replace zero data with client data (this mechanism is not yet fully implemented). However, the sector of any transaction (whether the client is the miner itself) cannot be replaced with a "verified transaction" - only blank committed capacity sectors can be upgraded.

Q14: In the current testnet, we can see that some code-related work still needs to be done. When will the testnet be reset? Will the mainnet launch go according to plan?

A14: (By Aayush) Yes, there is still some work to be done before the mainnet protocol is frozen. We're heading towards the final sprint of performance and UX issues and protocol work.

You can check out our latest roadmap update for more information on delivery estimates. NOTE: These are estimates and have not yet been scheduled. We'll announce it on the blog when the final release date comes.

Q15: Will the Filecoin network design data redundancy and encryption in the future?

A15: Data encryption and redundancy is up to the client and there are currently no plans to build it into the core protocol. We expect standards and conventions to emerge in the storage market and work with client software designers and implementers.

Q16: Which industry scales are supported by the official release? When will hardware standards be published?

A16: Shortly after the final decision is made between NSE and SDR, department size and hardware standards will be determined - depends on that.

Q17: For the second phase, the testnet seems to be very memory intensive. Does this have anything to do with algorithms?

A17: (By Aayush) I'm not sure if you mean memory has increased due to basic node operations or mining (sealing). I don't believe there hasn't been a noticeable increase since Phase 1.

Q18: When will we switch to NSE?

A18: (By Juan) NSE is in progress and we are still evaluating. We hope to have an evaluation as soon as possible (before testnet2 incentives). We don't want anyone wasting any money, so be careful not to buy a lot of hardware one way or another.

Q19: How many full-time R&D personnel are there in the Lotus and go-filecoin projects?

A19: The core team of the Filecoin project has a total of about 50 engineers and researchers responsible for various components of the implementation, which does not include those working on supporting protocols such as libp2p and IPFS. In addition to the core team, dozens of people in external groups work on various parts of the protocol — starting with building the developer tools protocol, security audits, performance improvements, and more. Finally, there are hundreds of external contributors in the open source community, making meaningful contributions to the project.

Q20: After changing the current SDR algorithm to NSE algorithm, does AMD server sealing still have an advantage? Will Intel servers work?

A20: It is difficult to benchmark an algorithm until it has been fully developed and its final parameters determined. We will make an announcement in Slack once NSE is ready for benchmarking. That said, the current construction does involve SHA hashing at some stages, so CPUs with SHA extensions will benefit in those stages.

Q21: When will the details of the encrypted economic model be released?

A21: We have published a blog post describing the overall cryptoeconomic structure. Many of these mechanisms are already on the testnet, and we encourage you to try them out! We hope to describe each mechanism in detail later. Parameters can only be finalized after all core mechanics have been completed and tested.

Q22: When can Filecoin store hot data?

A22: Filecoin can already store hot data through Filecoin payment channels and development tools such as Powergate. Warm data is served quickly, and we expect it to work well for Web3 applications.

- If NSE finally solves the problem, it will make the storage market data warmer, although it is still not warm enough.

- We hope that future PoRep constructions will be faster and faster, as suggested by filecoin researchers. It just didn't make it to mainnet in time.

Q23: What is the biggest technical problem of Filecoin at present?

A23: The Filecoin protocol is complex, with many interacting parts. The biggest difficulty is making sure the many parts work together to achieve a secure and useful protocol.

Q24: How to deal with data corruption on one or more nodes?

A24: (By Aayush) Miners should keep backups if they are concerned. If you lose data due to an in-progress transaction, or miss a window to post a challenge, then you will be penalized.

There's no way to avoid this, because proving you've stored data is the whole point of Filecoin, and there's no way to tell the difference between "my data is corrupt" and "I never stored it like I said I would."

Q25: How does the Filecoin network handle content deletion requests? (e.g. child pornography) If I am a storage service provider, what are my responsibilities here?

A25: Filecoin is a protocol that helps coordinate file storage across a decentralized network of storage providers and customers. The network cannot delete files from participants' computers. However, we are developing a powerful set of tools to help users control the type of data they are willing to store and serve from their computers.

Q26: If the valid user data stored in the hard disk has expired (the storage period in the storage order has expired), can miners delete the data?

A26: Yes, when all transactions in a sector have expired and reached the end of their stated lifetime, the sector may be terminated and the storage area will be used to seal the new sector.

Q27: How to distribute the 1M FIL bonus for the second phase of the testnet?

A27: We will be announcing the final structure of the testnet rewards soon, if you are interested in participating we strongly recommend setting up a mining operation to compete for storage and retrieval transactions from real users.

Q28: The data verification design can motivate miners to store valid data. For example, miners with orders can get a coefficient reward of 1.1, which is a very meaningful weakly dominant strategy. However, using the hard disk sales method obviously cannot guide miners from the perspective of technological improvement, and it is easy to lead to short-term behavior. Have you considered how to deal with the possible negative effects of the 10 times coefficient, such as: contrary to decentralization, too high a coefficient leads to the concentration of block rewards so that it becomes a playground for capital rather than a paradise for technological progress?

A28: Discover will only operate at a very small scale and will have a large, decentralized set of validators and many validated clients.

Q29: The NetworkBaseLine setting is very important. Which standard is being considered? For example, how long do you expect the network to reach the defined goals?

A29: The official is doing a lot of work, and the network parameters will be determined later. We're excited to see the baseline running on the testnet, and block rewards have increased 7x since the testnet launch!

Q30: Will the official design and develop mining software?

A30: The filecoin project has developedhttps://github.com/filecoin-project/sector-storagemodule, which is used by both Lotus and Go-filecoin. Miners are free to customize or re-implement the tool as they see fit.

Q31: Regarding the data hard disk, how much data is there at present? How much did the first batch sell for? How much do you plan to sell?

A31: Filecoin Discover has been well-received, and the orders we have already received represent data that would have taken years to ship directly over the Internet. We are pleased to support such an important data preservation effort.

It must be emphasized that Discover is just one of many verified clients. Over the next few days we will be announcing some small changes to Discover to emphasize that participation is not mandatory and should not be the main focus of mining operations.

Q32: How does Filecoin storage use the ipfs system.

A32: (By Aayush) Filecoin storage won't "leverage" IPFS, but you can easily store IPFS data on Filecoin, and you can easily store Filecoin data on IPFS.

Q33: Will this test be restarted next?

A33: We will restart when we are ready to start more functional tests. Not sure when yet.

Q34: How to understand the exact meaning of committed capacity sector? And how do I avoid my ability to commit to the mistake of just getting paid?

A34: Good question! When there is no demand for storage on the network, miners can choose to attest to "committed capacity" sectors, which can then be upgraded to sectors with actual transactions to earn transaction payments (no lockup). For miners with no storage needs, it makes more sense to seal committed capacity sectors than to trade on their own.

Q35: There are Fil rewards in the late stage of the second stage, but it is not recommended to buy hardware in the early stage. How should we choose?

A35: When we announce the bounty program, it will be accompanied by a description of the intended algorithm we will use for the launch. We'll make sure miners have a few weeks between any incentive announcements and the start of the competition (and we'll do what we can).

Q36: In addition to the Filecoin discover plan, are there any other important pre-launch plans for the official launch of the main network?

A36: We see a lot of ecosystem activity from groups looking to build on top of Filecoin. These groups help expand and expand each of the following major Filecoin stakeholders: developers, customers, token holders, partners, and miners. Some exciting examples include:

- Textile's Powergate platform (here's a recent demo: https://www.youtube.com/watch?v=aiOTSkz_6aY&feature=youtu.be )

Glif web wallet: https://filecoin.io/blog/community-jonathan-schwartz-owl/ - Starling storage: an application that facilitates verified storage of archival datasets on Filecoin: https:// /starlingstorage.io/

Q37: Has the epidemic in the United States affected our progress? When will the official come to Shanghai, China?

A37: (By Juan) The epidemic has slowed us down a little bit, but not by much. We are prepared for this and working hard to deal with it. We hope to go to China when it is safe to travel, but unfortunately this has not been possible in recent months. I hope to be able to go in 2021 or sooner.

Q38: When will the 5PB large miner test start?

A38: We will announce as soon as we have developed the program to this point.

The above is the first-hand arrangement of the 38 questions in this AMA. I hope that the mainnet will be launched in the next AMA.