LEGAL DISCLAIMER: The following questions and answers are provided for general informational purposes only, and may contain forward looking statements subject to risks and uncertainties. The actual properties and behavior of the protocol may differ materially from the information provided here. Statements in this document are not investment or legal advice.
- 中文（中国） (zh-cn): Filecoin常见问题解答
1. Filecoin 101
Filecoin is a decentralized storage network that turns cloud storage into an algorithmic market. Miners earn the native protocol token (also called “filecoin”) by providing data storage and/or retrieval. Conversely, clients pay miners to store or distribute data and to retrieve it. “Filecoin” can refer to a) the network, b) the protocol, c) the token powering the network, and d) the project.
For more details, please see the Glossary below and https://filecoin.io/.
You can check out the Filecoin codebase at https://github.com/filecoin-project.
Filecoin and IPFS are complementary protocols, both created by Protocol Labs. IPFS allows peers to store, request, and transfer verifiable data with each other. IPFS is open-source and free to download and use, and is already in use by numerous groups. With IPFS, individual nodes store data that they consider important; there’s no simple way to incentivize others to join the network or store specific data.
Filecoin is designed to solve this key problem by providing a system of persistent data storage. Under Filecoin’s incentive structure, clients pay to store data at specific levels of redundancy and availability, and miners earn payments and rewards by continuously storing data and cryptographically proving it.
In short: IPFS addresses and moves content; Filecoin is the missing incentive layer.
Filecoin also employs many components of IPFS. For example:
- Filecoin uses IPLD for blockchain data structures
- Filecoin nodes use libp2p to establish secure connections to each other
- Messaging between nodes and Filecoin block propogation utilizes libp2p pubsub
Moreover, the Filecoin core team includes some members of the IPFS core team. Compatibility is intended to be as seamless as possible between IPFS and Filecoin. Even after Filecoin launches, we expect the IPFS and Filecoin open-source communities to continue to collaborate and improve compatibility. You can view the draft spec for IPFS<>Filecoin Interoperability and our ideas for future improvements to learn more.
First, it’s worth repeating that Filecoin and IPFS are intended to complement each other and have significant cross compatibility. We are working to make it easy to migrate between voluntary IPFS storage and paid Filecoin storage.
With IPFS, you are responsible for your own storage nodes by directly provisioning hardware or purchasing storage from a third party. On IPFS, individual nodes store content that they consider important; there’s no simple way to incentivize others to guarantee to store your data on their systems. Filecoin provides the missing incentive structure.
If you prefer to maintain your own storage nodes, or have external arrangements to store data cooperatively with peers, IPFS may be your preferred option. If you prefer to pay competitive prices and have storage managed for you at specific levels of redundancy and availability, Filecoin may be your preferred option.
When you type a URL into your web browser to access information on the internet today, the URL resolves to an IP address that identifies one server storing the information you are looking for. Nearly every publisher, vendor, and service on the internet stores information on servers located in specific data centers under their control, which makes the internet as we know it today “centralized”.
Filecoin is a decentralized storage system that uses cryptographic hashes to identify and distribute data stored on the network. The network protocol itself works to ensure your files are stored. Novel cryptographic proofs create a useful and valuable storage service for clients as a byproduct of the Filecoin mining process. The miners are incentivized to monetize open hard drive space on Filecoin’s verifiable marketplace for storage, and they are awarded proportionally: the more they store, the more filecoin tokens they earn.
We believe this additional supply of verifiable, decentralized storage will both reduce costs and improve performance for clients. As a decentralized protocol, the data stored in the network and access to it cannot be controlled or exploited by a central entity, which improves robustness. At scale, information distributed across Filecoin miners will be stored nearer to users than is currently possible with monolithic centralized servers and large content distribution networks, making information retrieval much faster. Indexed data via cryptographic hashes on Filecoin enables users to efficiently orchestrate and update massive datasets. Lastly, as an open-source project, the software itself can be examined, verified, and improved publicly, unlike most cloud storage and distribution platforms today. As Filecoin evolves and new capabilities are added, we hope the network can provide a platform for most, if not all, information storage and delivery needs across the internet.
Once a quarter or so, the Filecoin team leads take the time to coordinate on this and make a good estimate as of that date.
The most recent update can be found on the Filecoin Network roadmap.
Unfortunately, we can’t provide interim updates.
2.1 Setting up to mine
Please see our blog post for more information about getting started with Filecoin. We will continue to share more information and issue guidance as development progresses. In short, you will be able to run Filecoin mining node software in a way that is similar to networks like Bitcoin and Ethereum.
Please be sure to register on the Miner Testing Form to be notified when features are ready for testing by prospective Filecoin miners.
We have not yet released exact hardware specifications for Filecoin mining.
Storage miners earn filecoin by storing data for clients, and computing cryptographic proofs to verify storage across time. The probability of earning the block reward and transaction fees is proportional to the amount of storage the miner contributes to the Filecoin network, and not hashing power.
Retrieval miners earn filecoin by winning bids and mining fees for a particular file that are determined by the market value of that file. A retrieval miner’s bandwidth and bid/initial response time for deals (i.e., latency and proximity to clients) will determine its ability to close retrieval deals on the network. The maximum bandwidth of a retrieval miner will set the total quantity of deals it can make.
Therefore, mining hardware should include ample storage space for storage mining, and/or a fast internet connection for retrieval mining. We expect that mining hardware will contain configurations with many hard drives, likely similar to a network-attached storage (NAS) device. Exact specifications for CPU and RAM have not yet been published. We also note that early storage and retrieval miners will be required to run full nodes of the Filecoin network.
We don’t have specific recommendations for mining hardware. You can view the draft specs for storage mining and retrieval mining to get a sense of what they involve. Please note that the specs are works in progress and are likely to change before mainnet launch, so we don’t recommend purchasing hardware based on the specifications of the devnet or testnet.
We’re doing our best to design the protocol so that “specialized Filecoin mining hardware” is most likely a highly optimized hard drive. Because Filecoin’s consensus mechanism is based on Proof-of-Replication and Proof-of-Spacetime, there is no hash-intensive Proof-of-Work. While the Filecoin proofs do require frequent storage lookups that could be facilitated by hardware, it would be expensive and likely impractical to design ASICs that support such random access to large amounts of memory. Realistically, we expect commodity hard-drives to generate and verify proofs at competitive speeds.
We have not started testing the Filecoin Network with any miners. All miners are invited to explore the Filecoin codebase as of February 2019.
Interested miners are encouraged to check out the code and participate in our community discussion by joining our Slack, Matrix, and Filecoin Forum. Miners should not invest in mining hardware based on any current Filecoin specs or implementations, as they are certainly going to change. Those who seek to mine Filecoin will want to wait for a future release (likely, the testnet milestone).
If you’d like to help us prepare for the Filecoin main network launch, please sign up on the Miner Testing Form to be notified when features are ready for testing by prospective Filecoin miners.
2.2 Running your node
The Filecoin whitepaper states that storage miners will store two types of data: (1) sealed copies of the data they have agreed to store and (2) blockchain data. We expect miners will act as full nodes to better support the protocol, storing and verifying the complete blockchain. Since stored files will not be included in the blockchain itself, total storage required for storing the entire blockchain will be much lower than the sealed data that miners will store for deals.
Yes, retrieval miners can store data without being storage miners. However, storing data does not make a retrieval miner a storage miner. All storage miners are required to engage in the Filecoin Mining Protocol. That means: pledge storage, commit collateral, and prove storage continuously through Proofs-of-Spacetime.
We anticipate retrieval miners will also be storage miners (in which case they get the data directly from the client) or will retrieve a file from other storage miners if they realize the file is in high demand. The retrieval miner is not restricted to obtaining the data through Filecoin. It is possible that the data was temporarily available on IPFS or a server for free.
While the storage market takes place on-chain, the retrieval market will operate entirely off-chain. We also expect that retrieval miners serving popular data will be globally distributed and efficient. You can watch the Filecoin Storage Market Basics video for more information.
2.3 Rewards and profitability
Individual miners can participate in both the storage and retrieval markets to earn filecoin.
In the storage market, miners are paid filecoin over time as they continue to prove they are storing a particular file. In addition, storage market miners have some likelihood of winning the next block, and earning the block reward and fees from transactions included in the block (similar to block rewards in other cryptocurrencies).
In the retrieval market, miners are paid in filecoin via off-chain payment channels for quickly providing files to clients who request them.
No. Similar to other networks, Filecoin will include a devnet and testnet to allow developers to test and experiment without putting actual tokens at risk. These early networks will be unstable and much less secure than the main network, and you should assume devnet and testnet tokens are worthless. The filecoin main network token will not be released until the main network has launched.
3.1 Usage and costs
Filecoin uses a native currency to engineer an incentive structure for the participation of storage and retrieval miners, or the supply side, of our network. Filecoin was designed with the goal of ensuring that the most profitable choice of every participant (including clients, miners, investors, and developers) is to act to improve the quality of service for the network.
Filecoin is built around a novel blockchain consensus protocol based on Proof-of-Storage. Clients who want to store files on the Filecoin network hire miners to store multiple copies of their files across the network. Each miner must submit publicly verifiable cryptographic proofs that they are continuing to store the data in order to receive payment and block rewards in filecoin. Payment and rewards in filecoin power the incentive structure to guarantee a fair, permissionless, robust, and decentralized storage network.
In some cases yes, and in some cases no. Someone needs to pay for the bandwidth costs incurred by miners serving the data to whoever is retrieving it. The simplest structure is for whoever is retrieving the data to pay for the costs of their own request.
Of course, that simple structure is not enough to satisfy the wide variety of data storage and distribution arrangements in the internet today. It is anticipated that Filecoin will offer constructions that allow one party to pay for another party to retrieve data (for example, via the use of vouchers or tickets). That will make many other kinds of models work, including the main content distribution model of web 2.0 websites. In that model, website owners pay for infrastructure services to serve data to their users for free, and then monetize the content in some other way. We aim for Filecoin to support structures like these. But we also aim to enable users and content creators to explore all kinds of new content distribution and economic models.
IPFS is and will remain free to download, run, and use, and will run independently from Filecoin. Once Filecoin launches, IPFS nodes may also offer retrieval of their files on the Filecoin retrieval market at no cost or for a profit.
In addition, we believe that data sets with significant scientific or cultural value that are currently hosted on IPFS may have their Filecoin costs sponsored by nonprofits and governments.
3.2 Usage and guarantees
The Filecoin test network and main network are not yet live, but you can sign up for our early user program and follow the Filecoin blog for updates. You can also see the Filecoin Project codebase on Github.
Filecoin is designed for highly redundant, highly available, and highly geo-replicated storage. Nobody can cover against all data loss scenarios, but we can drastically increase the probability of success by encouraging high replication, repairing damage proactively, and making it expensive for miners to fail.
Filecoin stores data redundantly across multiple storage miners. If one of those miners loses the file, they then lose their committed collateral (as a fee for failing to honor the deal), and a new miner will be automatically hired to store another replica of the file. In this way, the Filecoin network is expected to be self-healing, and repair damage over time. This is similar in some ways to how professional cloud storage providers work today, but with an added repair that works across multiple storage providers.
On the Filecoin network, clients can adjust the number of times their data is replicated, also known as the replication factor (or erasure coding), to achieve higher levels of redundancy and safety, though at a higher cost. Replication defaults will be carefully chosen to optimize safe storage and reduce the risk of data loss. Clients can adjust these defaults to achieve even higher (or lower) levels of redundancy and safety, at different costs.
In addition, we expect reputation systems to emerge. We expect that client nodes will then be able to run software that learns the past history of storage miners and aims to hire more reliable miners.
It’s difficult for a protocol to distinguish a miner who has lost connectivity from a miner who has maliciously deleted the data they’ve promised to store. As a result, miners are required to lock up collateral, which they lose if they fail to prove that they’re storing the files. If a single miner is offline for a short period of time, and continues to generate proofs, they will be subject to little or no penalization. Eventually, if the miner is offline for too long, the miner is declared to have failed. The miner loses its collateral, and the network triggers the repair process, hiring a new miner to store the data.
Like IPFS, Filecoin uses content addressing. We use cryptographic hash functions to generate concise content identifiers (CIDs). This acts as a unique identifier or “fingerprint” for each file, regardless of name.
We believe the probability of encountering a collision with modern, secure, collision resistant cryptographic hash functions is astronomically low. Further, the likelihood of finding a second file that hashes to a specific value (second preimage attack) is even more unlikely. Cryptosystems around the world rely on these values being secure, and thus we believe it is safe to do so in Filecoin.
That said, secure hash functions have broken in the past. The collisions we have found in secure hash functions have generally been due to the hash function breaking under cryptanalysis (such as md5 and sha1). This is always a possibility, and most other cryptosystems – such as Bitcoin, Ethereum, DNSSEC, Certificate Authorities, HTTPS, and more – are equally vulnerable. Nevertheless, we have planned for this unlikely event! Our content identifiers (CIDs) use the Multihash format, allowing safe upgrades to a new hash function.
If you want to remember that QmXoypizjW3WknFiJnKLwHCnL72vedxjQkDDP1mXWo6uco gets you a Wikipedia mirror, you have our respect. But users should not be expected to remember content identifiers (CIDs) in this format. Over time, we and other developers building on Filecoin plan to make it easy for users to seamlessly and conveniently find files using human-readable naming systems.
4. Filecoin 201
For those who want to dig deeper…
The Filecoin network is decentralized because, anybody, without requiring permission from any central party, can:
- Participate by providing data storage or retrieval
- Use the network to store and retrieve data
- Audit and verify the codebase
- Improve the protocol by proposing and implementing improvements
- Build applications, features, or functionality on top of the network without asking for permission
- Voice dissent and propose other options or directions for the network
- Fork the codebase
- Fork the blockchain
- Hold or transact with the filecoin token
Filecoin will be an open-source protocol to which anyone can contribute or choose to run; moreover, the protocol, not a particular company, will mediate interactions between the participants. Filecoin will rely on support from its communities of developers, storage providers, and storage users building and using the protocol, rather than on the efforts of any particular company.
Clients who want to store files on this network hire miners to store multiple copies of their files across the network. Each of these miners must submit proofs that they are continuing to store the correct number of copies of the data in order to receive payment and rewards. The publicly verifiable proofs allow the Filecoin protocol to permissionlessly admit any aspiring participant (without verification or centralized approval) and still provide a system for guarantees based on the engineered incentives: Filecoin was designed with the goal that the most profitable choice of every participant (including clients, miners, investors, and developers) is to act to improve the quality of service for the network.
Most blockchains use a Proof-of-Work consensus protocol to incentivize mining and provide network security. These proofs are generally energy intensive as they require computers running nodes to calculate many computations of a hash function in order to verify the transactions on the network. While some novel projects are working toward a Proof-of-Stake consensus mechanism, the Filecoin protocol will use Proof-of-Replication (PoRep) and Proof-of-Spacetime (PoSt) to demonstrate that files have been allocated their own unique space in storage memory and are being stored over time. A Proof-of-Replication is a type of Proof-of-Storage with specific properties, which are outlined in detail in our Technical Report on Proofs-of-Replication.
Proof-of-Replication was developed to solve the specific problems of a verifiable, decentralized storage network that could incentivize and reward file storage. One of the biggest draws for developing a blockchain consensus algorithm based on Proof-of-Replication was, in fact, our dismay for the wasted energy from hash-based Proof-of-Work schemes. Proof-of-Replication not only requires much less energy than Proof-of-Work, but the work required is a demonstration that miners are contributing a valuable resource to the network, namely additional storage capacity. Furthermore, the block reward serves to incentivize miners to contribute as much storage as possible, driving down the price of storage on the Filecoin network.
We find the power consumption of wasteful proofs-of-work incredibly alarming. Given the quality and number of proposals for useful Proof-of-Work and Proof-of-Stake consensus protocols under development, we think that wasteful proofs-of-work should and will slowly wane as various networks adapt, or rise and fall. We hope and expect the Filecoin network to grow and adapt in a sustainable manner. We aim for the energy expenditure of the Filecoin network to be a negligible fraction of the energy spent maintaining Bitcoin, while providing the same level of security, and the utility of file storage as a side effect. We also hope to establish an optimization structure that may one day yield more power-efficient cloud storage networks. To that end, we’re looking into estimating the environmental impact of all aspects of Filecoin.
Filecoin uses a Proof-of-Storage consensus protocol because it is simple, aligns incentives, and drives miners to provide real, new storage at competitive prices. In developing a new type of Proof-of-Storage that we call Proof-of-Replication, we will enable miners’ storage devices to contain useful data, making them a valuable resource for the network. Proof-of-Stake is an interesting area of research that we closely follow.
We intend for Filecoin tokens to be distributed when the main network is launched (please see the development roadmap for the latest estimated timeframe). Shortly before then, we’ll send instructions to the email address associated with your CoinList account.
No K-1s will be issued for investments in Filecoin SAFTs. To review your final Filecoin SAFTs, follow the steps in this guide we have published on our website. Unfortunately we’re unable to give tax advice, so we suggest consulting with your tax advisor.
The filecoin token vesting period begins at main network launch. Shortly before then, we’ll send instructions to the email address associated with your CoinList account. Please see development roadmap for the latest estimated timeframe.
No. The Filecoin token sale was closed in September 2017, and we’re not able to complete any new investments at this time. Please be aware that any offers by third parties to buy, sell, or trade Filecoin are fraudulent.
There is currently no way to purchase or sell Filecoin, and therefore no price. Protocol Labs is focused on the engineering and community challenges involved with building a more efficient, resilient, and decentralized web. We believe this is the best way to create a network that’s useful to diverse groups of people for a very long time.
Investor caution: During and after the Filecoin token sale, we have seen parties claiming to sell Filecoin SAFTs or tokens, as well as exchanges listing filecoin and filecoin derivatives. Remember that the filecoin token will not be live until network launch, and any exchange that lists Filecoin or IPFS tokens right now is likely fraudulent. We will announce on the Filecoin blog when the Filecoin network and token go live. Until then, we strongly recommend staying away from any exchange or entity who claims to be buying, selling, or trading Filecoin or IPFS SAFTS, tokens or derivatives.
We ask that you kindly review our Code of Conduct prior to participating in the Filecoin community channels.
Please look also for official updates on the Filecoin blog.
Protocol Labs is excited to see that IPFS is being embraced and promoted by many communities around the world. Some of those communities and their members have been in contact with us, and have offered to help us with things like translations. In some cases, we may accept open-source help, but that does not establish any sort of official or formal relationship. At this time, Protocol Labs has no official partnership with any community or individual, including Chinese communities or individuals. Any updates about official community partners will be announced on our website, or on the Filecoin blog.
The code repositories for Filecoin project can be found here. We open-sourced several Filecoin project repos on 2/13/19.
This is an exciting time to get involved! Everybody interested can come learn about Filecoin, help build parts of it, and start building on top of it.
You may also refer to the Contributing Section of the Filecoin discussion forum for more information about contributing prior to getting started.
We love to see volunteers developing communities around Filecoin. Please email firstname.lastname@example.org with your information (name, background, experience, and location) and we’ll try to help however we can!
Ambitious projects like Filecoin succeed or fail depending on their teams. If you might be excited to help us build the Filecoin network, and share our vision, mission, and values, please apply for full time roles or check out our RFP program if you’d like to specifically help research our open problems. You can also make open-source contributions to the Filecoin project on Github.
Filecoin is a decentralized data storage network powered by a blockchain and a native token. Filecoin’s blockchain runs on a novel proof, called Proof-of-Spacetime, where blocks are created by miners that are storing data. The Filecoin network refers to the network of independent data storage and retrieval miners (or providers) and the clients that can hire them. Clients can hire miners via two decentralized, verifiable markets: the Storage Market and the Retrieval Market. The Filecoin network handles storage and retrieval requests on its verifiable markets without relying on centralized coordination from a single company. Protocol Labs initially proposed the Filecoin network in July 2014 (link), and the design was substantially improved in early 2017 (link).
The Filecoin protocol refers to the communication standard that independent participants (miners and clients) adopt as part of our network. This standard is implemented in the Filecoin codebase and run by all participants on the Filecoin network. The protocol enables Filecoin’s data storage and retrieval service without relying on a single coordinator. The Filecoin protocol builds on the decentralized storage network, novel Proofs-of-Storage, verifiable markets and useful Proof-of-Work.
The filecoin token is the native crypto token of the Filecoin network, similar to bitcoin and ether. Filecoin is used to participate and transact in the Filecoin network (i.e., pay for data storage and retrieval).
The Filecoin project includes the Filecoin protocol, Filecoin codebase, filecoin token, supporting communities (participants, developers, investors and token holders), project communications, and a team of individuals. The Filecoin project team is responsible for developing the Filecoin network’s codebase, launching the network, and cultivating the ecosystems that support the miner, user, developer, and investor communities of the network.
Access is provided to anybody; no approval or formal verification from a single entity or authoritative third party is required to participate.