Search content
Sort by

Showing 20 of 26 results by yingyush55
Post
Topic
Board Development & Technical Discussion
Topic OP
Next generation decentralized application platform
by
yingyush55
on 24/04/2023, 13:55:41 UTC
User Uniqueness White paper
Next-generation user unique decentralized application platform
A user-unique decentralized blockchain technology where each user can only register one account for computing data to verify the validity and integrity of the blockchain data. A blockchain with a fully fledged Turing-complete programming language built in would require only a few lines of code to write the desired logic.
Bitcoin's workload proves that the PoW mechanism is a breakthrough because it solves two problems at once. First, it provides a simple and relatively efficient consensus algorithm, allowing nodes in the network to all agree on a set of canonical updates to the Bitcoin ledger's state. Second, it provides a mechanism for allowing free access to the consensus process, solving the political problem of determining who influences consensus while preventing witch attacks. In Ethereum's proof-of-interest PoS, the election of verification nodes is based on the number of pledged assets held. A node with more Ether has a higher chance of being selected as a verifier, so it doesn't consume power but PoS is an equity mechanism which defeats the purpose of de-centring. Therefore, we will not use the PoW and PoS consensus mechanism, and use the advantage of user uniqueness to a new Proof of Randomness (PoR) consensus mechanism, which selects the bookkeeper by random number, so as to realize random bookkeeping. Unlike other consensus mechanisms, PoR's bookkeepers do not need to solve any mathematical problems or pledge any tokens, making PoR more efficient and centralised. We can even record and verify transactions across the network on mobile devices such as mobile phones to get rewards.
One important scalable feature of Bitcoin is that its blocks are stored in a multi-level data structure. The hash of a block is really just the hash of the block header, which is a piece of data of about 200 bytes containing a timestamp, random numbers, and the hash of the previous block.
A full node in the Bitcoin network -- the node that stores and processes all the data in all blocks -- took up 15GB of disk space in April 2014, and is growing by more than 1GB every month. At the moment, this is acceptable for desktop computers, but mobile phones can no longer carry such huge amounts of data, and in the future only businesses and hobbyists will act as full nodes. The Simplified Payment Confirmation Protocol (SPV) allows another type of node, called a "light node," to download the block header, use the block header to confirm proof of work, and then download only the Merkle tree branches associated with its transaction. This allows light nodes to securely determine the status of any one bitcoin transaction and the current balance of an account by downloading a small portion of the entire blockchain.
User uniqueness
The purpose of user uniqueness is to create a decentralized application of web3.0 unique identity. We believe that only user uniqueness can solve financial crimes brought by the current blockchain anonymity and various crimes brought by the future web3.0 anonymity. PoR consensus mechanism can solve the waste of power resources and computing power resources caused by the current proof of work. And the blockchain, with its Turing complete programming language, allows anyone to write smart contracts and decentralised applications in which they can freely define ownership rules, transactions and state-transition functions. This is far more powerful than the smart contracts that Ethereum and Bitcoin can offer.


User unique account
In user uniqueness, each user generates a unique code through face recognition, which binds the public key and keeps the private key by himself, so that he has his own unique account. Although the user's account is unique but does not mean that the user's personal information is public, because a string of code on the network and can not know the user's identity, unless the account has criminal behavior, the country is able to obtain the code through the face recognition company portrait. Here we should note that we are introducing the face recognition company interface and face company is centralized, can add more family face recognition company interface, at the same time through the generation of unique code to solve the problem of face recognition company centralization.
Blockchain and computing
The user-unique blockchain is completely different from the Ethereum bitcoin blockchain,
Bitcoin is extremely inefficient because it needs to store the entire state of each block. The reason is that the state is stored in the tree structure and only a small portion of the tree needs to be changed after each block is added. So in general, most of the tree should be the same between two adjacent blocks, so data can be stored once and referenced twice with Pointers (that is, hashes of subtrees). A special type of tree called a "Patricia tree" is used for this purpose, which includes modifications to the Merkle tree concept, allowing efficient insertion and removal of nodes rather than just changes. In addition, since all state information exists within the last block, there is no need to store the entire blockchain history, and calculations using this strategy can save 5-20 times the space if applied to bitcoin. However, at the beginning of the program operation, I hope all nodes will be heavy nodes to record the history of the whole blockchain, because the data at the beginning is not large, and each node can easily record the data of the whole network.
application
In general, there are three types of applications based on user uniqueness. The first type is the identity application of web3.0, and the second type is the application of data conversion between users. Finally, there are applications such as online voting and decentralized governance.
Identity and reputation system
Based on the premise of user uniqueness, users will be responsible for their own web3.0 identity, behavior, reputation, web3.0 era will come sooner or later, before the arrival of web3.0, we can no longer use the anonymous wallet address like Bitcoin to bring a variety of money laundering crimes. If Bitcoin had been designed to be unique to its users, no one would be using it to break the law.
Go to the center for client updates
Whether it's Bitcoin or Ethereum, they're just de-centralizing the transaction data, and the updates on the client side are still dictated by the development team, which is not really de-centralizing. We're going to hand over the right to update the content of the client, not the right to develop, but the final decision to implement the update to the client requires a majority vote of the user. The development team still has the right to initiate version updates, but whether the version can be updated depends on the votes of all users. For example, when Ethereum updates pos, it does not need the consent of ordinary miners. It can change from pow to pos today and update tomorrow in a favorable direction for the development team.
Decentralized file storage
Over the past few years, a number of popular online file-storage startups have sprung up, most notably Dropbox. Dropbox wants to allow users to upload hard drive backups, provide backup storage and allow users to access backups for a monthly fee. At this point, however, the file storage market is sometimes relatively inefficient. A cursory look at the various available solutions reveals that the monthly price of mainstream file storage is higher than the cost of an entire hard drive, especially in the 20-200 GB category known as the "Uncanny Valley," where there are neither free credits nor enterprise discounts. User-only contracts have enabled the development of a decentralized file storage ecosystem, where individual users can rent out their hard drives for a small profit, while unused space can be used to further reduce the cost of file storage.
Decentralized autonomous organization
Generally speaking, a "decentralized autonomous organization" is a virtual entity with a certain number of members or shareholders, who own a majority stake of approximately 67%. This essentially replicates the legal means of a traditional corporation or non-profit organization, but is only enhanced by crypto blockchain technology. So far, much of the discussion about decentralized autonomous organizations has revolved around the "capitalist" model of decentralized autonomous corporations, with dividend shareholders and tradable shares; As an alternative, what might be called a "decentralized autonomous community" would give all members equal power in decision-making and require a majority of 67 percent of existing members to approve any increase or decrease in membership. Since each person can only have one membership, groups are needed to perform collectively.
Another model is a decentralized company, where any account can own zero or more shares and decisions require a two-thirds majority. The complete framework would include asset management capabilities, the ability to bid to buy or sell shares and the ability to accept offers (preferably with order matching in the contract). Entrustment also provides a democratic form of appointment system and popularizes the concept of "board of directors".
More applications
Cloud computing
User-unique virtual machine technology can also be used to create a verifiable computing environment, because each computer only records and verifies the data of the entire network without taking up computing power. This allows the entire network to call on a user's computing power, allowing the user to ask others to perform a calculation, and then selectively ask for proof that the calculation was done correctly at some randomly selected checkpoint. This could create a cloud computing marketplace where any user can participate from their desktop, laptop, or dedicated server, and call on the power of the entire network when they need it.


Centralization of mining
The Bitcoin mining algorithm works by having miners perform millions of SHA256 calculations on a slightly modified version of the block header, again and again, until eventually a node produces a version with a hash smaller than the target value (currently about 2192). However, this mining algorithm is vulnerable to two forms of centralized attack. First, the mining ecosystem is already dominated by ASics (application-specific integrated circuits), computer chips designed specifically for specific Bitcoin mining tasks and thus thousands of times more efficient. This means that bitcoin mining is no longer a highly decentralized and egalitarian enterprise that requires huge amounts of capital to participate effectively. Second, most bitcoin miners don't actually do block verification locally; Instead, they rely on centralized pools to provide block heads. The problem is arguably worse: as of this writing, the top three mining pools indirectly control about 50% of the processing power in the Bitcoin network, although the problem is mitigated by the fact that miners can switch to other pools when a mining pool or alliance attempts a 51% attack.
Ethereum is now intended to use a mining algorithm that requires miners to take random data from the state, calculate some randomly selected transactions from the last N blocks of the blockchain, and return a hash of the result. This has two important benefits. First, Ethereum contracts can include any type of computing, so Ethereum ASics are essentially ASics for general computing, i.e., better cpus. Second, mining requires access to the entire blockchain, which forces miners to store the entire blockchain and at least be able to verify every transaction. This eliminates the need for a centralized pool; While a pool can still legitimately balance the randomness of reward distribution, a peer-to-peer pool without centralized control works just as well.
Whether Bitcoin or Ethereum huge capital can have more computing power, when the computing power reaches 51% can launch attacks. Ethereum does not have all nodes that record the whole network but only the heavy nodes that record the whole network which makes Ethereum not decentralized enough. On the other hand, with the user-unique PoR consensus mechanism, everyone who wants to get the block billing reward needs to keep the computer or mobile phone on to record the data of the whole network before they can get the billing reward. Under such a premise, it is almost impossible to control the computing power of the whole network, which can only control more than 51% users.
scalability
Like Bitcoin, Ethereum has the drawback that every node in the network needs to process every transaction. With Bitcoin, the current size of the blockchain is about 15 GB, growing by about 1 MB per hour. If the Bitcoin network were to process 2,000 transactions per second like Visa, it would grow by 1 MB every three seconds (1 GB per hour, 8 terabytes per year). This is helped by the fact that user-unique integrity nodes only need to store state rather than a complete blockchain history.
The problem with large blockchains is the centralization risk. If blockchain size is increased to 100 terabytes, it is likely that only a few large enterprises will be able to run full nodes, while all ordinary users will use light SPV nodes. In this case, there may be concerns that full node partners are fraudulently profiting (such as changing block rewards, giving themselves bitcoins, etc.). The light node cannot detect this immediately. Of course, there may be at least one honest full node, and a few hours later information about the scam will leak through a channel like Reddit, but by then it's too late: The massive, and probably impractical, collaboration of ordinary users to blacklist designated blocks is on the scale of a successful 51% attack.
conclusion
User uniqueness protocol is an upgraded version of blockchain technology, which makes users responsible for their own behaviors through user uniqueness and solves the energy consumption problem of proof of work through PoR consensus mechanism. Highly general-purpose programming languages offer advanced capabilities, and the existence of a Turing complete programming language means that, in theory, arbitrary contracts can be created. Around decentralized file storage, decentralized computing, and many other such concepts, the computing industry has the potential to become much more efficient. Finally, there are plenty of apps that have nothing to do with money at all.
Whether it is the current Bitcoin or Ethereum, they only decentralize the recorded data, while the content of version update is still decided by the development team, which is not the true sense of decentralization. User uniqueness is open in design, which not only decenters the data, but also enables the version update of the client to achieve decentralization. We believe it will be a good fit as a foundation layer for Web 3.0 for years to come.
Post
Topic
Board Bitcoin Discussion
Topic OP
Revealing the Essence of Virtual Currency
by
yingyush55
on 04/04/2023, 01:16:37 UTC
Today, let's reveal the secret of virtual currency. Is the huge development of virtual currency based on the fact that Bitcoin data cannot be tampered with? In fact, anonymity of Bitcoin is not the demand point of most of us because we do not have criminal demand in this regard. Our demand point is that data cannot be tampered with, and the operating code of Bitcoin is the consensus mechanism for users who hold Bitcoin, We can imagine this mechanism as if I signed a contract with Bitcoin, and we will always operate according to this contract.
Okay, now the focus is on whether there is a project leader and investment institution for any coin except Bitcoin. They hold a large number of tokens and their underlying code project can be changed. Taking Ethereum as an example, can so many miners agree to convert POW to POS? Finally, the development team can also change the operating mechanism of the underlying code to pos. This way, the development direction of Ethereum must be in the direction of benefiting developers and investors. In fact, there is no absolute right or wrong about this thing. We can also imagine the operating mechanism of Ethereum as a contract.
Let's take an example now. One is Bitcoin, who is relatively honest. The contract we signed at the beginning will never change, so we will continue to cooperate according to the rules we agreed upon.
Now there is another partner, Ethereum, who is quite intelligent and has a wide range of social resources and connections. However, he can change the contract I signed with him at the beginning. The changed contract may be in the direction of our company's interests, but he has the power to change this contract.
These two partners each have their own advantages and disadvantages, and there is no absolute right or wrong. However, if your original intention in entering virtual currency is to prevent data tampering, which partner should you choose? Don't go against your original intention for those prospects for the future.
Post
Topic
Board Development & Technical Discussion
Re: A more decentralized blockchain technology than Bitcoin
by
yingyush55
on 02/04/2023, 14:43:47 UTC
Bitcoin was originally designed to allow everyone to participate in mining, but now ordinary people in Bitcoin cannot participate in mining. The computational power is concentrated in several large mines.
Both Bitcoin and Ethereum can theoretically be centralized through huge amounts of money.
Can a new blockchain technology be designed to optimize Bitcoin blockchain technology from the bottom.
First of all, this blockchain technology must be that one person can only own one node. Otherwise, if a person like Ethereum has enough funds, it means generating enough nodes. That's how to go to the center. Bitcoin also means that a person can use huge funds to generate huge computing power, and that's how to go to the center.
The blockchain technology that I will design next is truly decentralized.
First, copy the underlying code of Bitcoin
Through face recognition technology, each person can only generate one node, a private key, and a public key.
By setting an upper limit on the computing power of each miner, it is possible to mine on a mobile phone without consuming too much computing power.
If we are interested, we can design such blockchain technology to run on mobile phones together.
By the way I don't know it's answer properly but I will say that currently Polygon Chain and also Ethereum Layer 2 projects could offer more defi solution and also lower transactions with less fee and faster transactions and that could lead to the adoption of more and more decentralized solutions and also a lot of Defi applications to run on these blockchains. So In future we might see full decentralized finance Blockchain that could revolutionize Bitcoin and also crypto.
Are you wrong? Do you seriously think that there are so many miners in Ethereum who will agree to switch from pow to pos? The development team can change the underlying code of Ethereum itself, and in the future, Ethereum will only develop in a direction that is beneficial to developers. Don't trust any currency except Bitcoin now.
Post
Topic
Board Development & Technical Discussion
Re: Feasibility of limiting the computational power of each node
by
yingyush55
on 02/04/2023, 11:29:27 UTC
If you add a computational power limit to each node in the Bitcoin base code, is this limit easy to crack? What I mean by cracking is not to increase the computational power by increasing the number of nodes, but can the set computational power be exceeded on a single node?
while it is possible to crack a computational power limit that is added to the bitcoin base code, the difficulty of doing so would depend on several factors. if the limit is properly designed and enforced, it should be difficult for the attackers to exceed the limit without being detected. However, it is still important for developers to carefully consider the potential risks and vulnerabilities of any changes to the bitcoin codebase.
If we can solve the energy consumption problem and fairness issue between users of Bitcoin, all efforts will be worthwhile.
Post
Topic
Board Development & Technical Discussion
Re: Feasibility of limiting the computational power of each node
by
yingyush55
on 02/04/2023, 11:13:10 UTC
--snip--
Mining pools are a large node.

I disagree. I expect pools have few nodes across different region for various reasons such as ensuring faster block propagation.

Will such centralized computing power affect the security of Bitcoin?

There are some concern about centralization of mining hashrate. But almost all them are not economically feasible.

Can several large mining pools jointly launch 51% computing power attacks

It's theoretically possible. But their pool would be abandoned by miners and face possibility of legal sue due after attempting 51% attack.

Quote
Limiting computational power can also filter non mining nodes. Is there a problem?
Yes, because then only miners will be able to verify data. If you ban all nodes that are not mining, then you will have only miners in the network, and then the whole P2P concept is gone, because all other users will then be entirely banned, and would need to connect to the network by using some miner as a proxy.
Suppose I created a fake coin using Bitcoin running code, but I added two restrictions to this fake coin. Each user can only have one account, and the computing power of each account is limited to 100M. This fake coin has 1000 users. If someone tries to attack with 1T computing power, is the network still safe?

Bitcoin network would be safe / not affected since your modification break isn't compatible with current Bitcoin protocol/consensus. Other node would just put you into ban.
If we use face recognition technology on this basis so that everyone can only generate a unique account, will it be fairer, safer, and more energy efficient than current Bitcoin.
Post
Topic
Board Development & Technical Discussion
Re: Feasibility of limiting the computational power of each node
by
yingyush55
on 02/04/2023, 11:01:52 UTC
--snip--
Mining pools are a large node.

I disagree. I expect pools have few nodes across different region for various reasons such as ensuring faster block propagation.

Will such centralized computing power affect the security of Bitcoin?

There are some concern about centralization of mining hashrate. But almost all them are not economically feasible.

Can several large mining pools jointly launch 51% computing power attacks

It's theoretically possible. But their pool would be abandoned by miners and face possibility of legal sue due after attempting 51% attack.

Quote
Limiting computational power can also filter non mining nodes. Is there a problem?
Yes, because then only miners will be able to verify data. If you ban all nodes that are not mining, then you will have only miners in the network, and then the whole P2P concept is gone, because all other users will then be entirely banned, and would need to connect to the network by using some miner as a proxy.
Suppose I created a fake coin using Bitcoin running code, but I added two restrictions to this fake coin. Each user can only have one account, and the computing power of each account is limited to 100M. This fake coin has 1000 users. If someone tries to attack with 1T computing power, is the network still safe?

Bitcoin network would be safe / not affected since your modification break isn't compatible with current Bitcoin protocol/consensus. Other node would just put you into ban.
If we limit the number of registered users, will this system be fairer and more energy efficient than the current Bitcoin system.
Post
Topic
Board Development & Technical Discussion
Re: Feasibility of limiting the computational power of each node
by
yingyush55
on 01/04/2023, 21:34:51 UTC
Quote
Limiting computational power can also filter non mining nodes. Is there a problem?
Yes, because then only miners will be able to verify data. If you ban all nodes that are not mining, then you will have only miners in the network, and then the whole P2P concept is gone, because all other users will then be entirely banned, and would need to connect to the network by using some miner as a proxy.
Suppose I created a fake coin using Bitcoin running code, but I added two restrictions to this fake coin. Each user can only have one account, and the computing power of each account is limited to 100M. This fake coin has 1000 users. If someone tries to attack with 1T computing power, is the network still safe?
Post
Topic
Board Development & Technical Discussion
Re: Feasibility of limiting the computational power of each node
by
yingyush55
on 01/04/2023, 19:15:02 UTC
Quote
this would involve verifying a LOT of shares
Note that shares are artificial concept, present in mining pools. In the Bitcoin network, you have valid blocks, or invalid blocks. Any block that does not meet the target is considered invalid, even if all included transactions are valid. Miners are rewarded only by mining pools for those invalid blocks, because they meet some lower difficulty, set by that mining pool.

Quote
a desire for everyone to have equal power, unlike the power currently owned by wealthy people
Note that there are also non-mining nodes, and using Proof of Work to filter connections will eliminate them. The question is: do you want to also eliminate for example block explorers, if their operators will not start mining?
"I don't understand English. I communicate with you using a translator. I'm a bit confused about what you said later. Perhaps you should use a simpler and more straightforward expression.". Limiting computational power can also filter non mining nodes. Is there a problem?
Post
Topic
Board Development & Technical Discussion
Re: Feasibility of limiting the computational power of each node
by
yingyush55
on 01/04/2023, 19:04:57 UTC
You can only create a consensus flag that would verify a proof that the work was only so difficult, this would involve verifying a LOT of shares so it is not extremely likely to be implemented naively if at all.
"If a consensus flag is created to limit computational power, can users bypass or crack computational power constraints?"
Post
Topic
Board Development & Technical Discussion
Re: Feasibility of limiting the computational power of each node
by
yingyush55
on 01/04/2023, 18:55:34 UTC
Quote
If you can limit the number of registered nodes per person, do you think such a system will become very fair?
No, it will increase pooled mining. Currently, you can mine coins without running a full node. If you restrict the number of nodes, then there will be even less miners with their own node. You will only harm people that run nodes without mining, which is probably not what you want.

Quote
The Pow mechanism is very complete and mature, but its node power is concentrated in the hands of wealthy people.
Not really, because anyone can run a non-mining node. And it seems to be useful in some situations, for example here: https://bitcointalk.org/index.php?topic=5447129.0

Without full node running 24/7, getting block 00000000000000000002ec935e245f8ae70fc68cc828f05bf4cfa002668599e4 would be very hard, because block explorers won't show you that, so you won't see, why this block was not accepted.

In theory, nothing stops non-mining nodes from charging fees for their services. In practice, most of them are not rewarded in any way, the only reward is that the node owner can be 100% sure what is going on in the network, if that node is online 24/7.

Quote
Mining pools are a large node. Will such centralized computing power affect the security of Bitcoin?
Why you want to solve "large node" problem by limiting the number of nodes? You should do it in exactly opposite way: encourage people to run nodes, so there will be a lot of smaller ones.
Limiting the number of nodes owned by each person and limiting the upper limit of the computing power of each node is not a solution to the problem of large nodes, but rather a desire for everyone to have equal power, unlike the power currently owned by wealthy people
Post
Topic
Board Development & Technical Discussion
Re: Feasibility of limiting the computational power of each node
by
yingyush55
on 01/04/2023, 18:19:05 UTC
Assuming you're talking about computation power to perform Bitcoin mining, the answer is no. Few reasons why it's impossible,
1. Most miner these days don't run full node. They only connect to mining pool where the pool give them some computational task.
2. Since most miner connect to pool, that means most Bitcoin block is mined and broadcasted by few pools.
3. There's no way to check how many miners involved to mine a block without trusting the pool.
Mining pools are a large node. Will such centralized computing power affect the security of Bitcoin? Can several large mining pools jointly launch 51% computing power attacks
Post
Topic
Board Development & Technical Discussion
Re: Feasibility of limiting the computational power of each node
by
yingyush55
on 01/04/2023, 18:11:06 UTC
So, for example, one computer can only output 1 GH/s mining hash rate, and if they want to increase it legally, they have to add more computers. Is this what you're trying to say?

If this is the case, I can see a lot of people using other clients or just straight up forking the network if they think it is worth the cost. Whether it is hackable or not is just speculative at this point since there is no code to scrutinize. Even if it is unhackable though, what I mentioned previously is still likely to happen. What is the purpose of this limitation btw? To prevent centralization or something else? I don't see how it works if the former is the purpose. CMIIW.
Mining pools are a large node. Will such centralized computing power affect the security of Bitcoin? Can several large mining pools jointly launch 51% computing power attacks
Post
Topic
Board Development & Technical Discussion
Re: Feasibility of limiting the computational power of each node
by
yingyush55
on 01/04/2023, 18:03:39 UTC
So, for example, one computer can only output 1 GH/s mining hash rate, and if they want to increase it legally, they have to add more computers. Is this what you're trying to say?

If this is the case, I can see a lot of people using other clients or just straight up forking the network if they think it is worth the cost. Whether it is hackable or not is just speculative at this point since there is no code to scrutinize. Even if it is unhackable though, what I mentioned previously is still likely to happen. What is the purpose of this limitation btw? To prevent centralization or something else? I don't see how it works if the former is the purpose. CMIIW.
If you can limit the number of registered nodes per person, do you think such a system will become very fair? The Pow mechanism is very complete and mature, but its node power is concentrated in the hands of wealthy people.
Post
Topic
Board Development & Technical Discussion
Topic OP
Feasibility of limiting the computational power of each node
by
yingyush55
on 01/04/2023, 11:28:53 UTC
If you add a computational power limit to each node in the Bitcoin base code, is this limit easy to crack? What I mean by cracking is not to increase the computational power by increasing the number of nodes, but can the set computational power be exceeded on a single node?
Post
Topic
Board Development & Technical Discussion
Re: A more decentralized blockchain technology than Bitcoin
by
yingyush55
on 01/04/2023, 00:31:32 UTC

Through face recognition technology, each person can only generate one node, a private key, and a public key.
By setting an upper limit on the computing power of each miner, it is possible to mine on a mobile phone without consuming too much computing power.
If we are interested, we can design such blockchain technology to run on mobile phones together.

I can find your face by just looking up your social media account, mean while can you discover the scalar that leads to these coordinates when multiplied by the secp256k1 curve Generator? :

x: 83260022149314607220299798889850163602723925708954433956645823190147746092229
y: 32501403673747823353314855720543420818188361784787882868346931239726724853713

If not I believe bitcoin is much more sound than your facial recognition scheme, do you want someone trying to cut your face off to access your funds? Is that better than protecting against data breaches really?
Stealing my facial images or videos cannot be obtained through face recognition, just as you cannot use photos or videos to recognize faces on your phone. A face can generate a corresponding string of codes, but this code cannot infer what a person looks like.

Not to mention that a government authority would then have the REASON to strap you down to a chair, peel your eyes open, and make you stand still while they scan your face to extract blockchain data. This is a terrifying idea.
The government has tied me to a chair and left me standing with my eyes open, unable to scan my face and extract my blockchain data. What is the difference between tying my body directly and asking me for my bank card password? If this is a security risk, then you know that your bank card password also has the same problem.

You are now just being delusional, if the government does that very thing to you they WOULD be able to access all your blockchain secrets immediately (assuming people actually used your insane face id chain) it would be trivial as THEY NOW HAVE COMPLETE ACCESS TO YOUR ENTIRE REAL FACE AND HEAD.

As for how is this a bigger security risk than a bank? It is just as insecure as a bank, BITCOIN is way more secure than this scheme because private keys can be stored OFF OF YOUR BODY.
The government can force me to perform the facial actions required for face recognition, and similarly, it can force me to say the bank card number and password, or the private key of Bitcoin. In terms of absolute power, no encryption technology will work. Bitcoin is not stored outside of my body, it is also stored in my brain.

I dont believe you or anyone truly remembers their private key if its any good, the real security of a private key comes from the discrete log problem and opting in to remembering it is your way of attaching a biological security vulnerability to it. There is nothing in bitcoin core that forces you to memorize your private key, its 100% optional.
Although I cannot remember the number of the private key, I can remember the specific location where the private key password is stored
Post
Topic
Board Development & Technical Discussion
Re: A more decentralized blockchain technology than Bitcoin
by
yingyush55
on 01/04/2023, 00:08:31 UTC
Yes of course as you say it can be done through identities. But then you have something very different than Satoshi's idea. You have a system where every single person on the network must be identified. Of course a system can be designed that puts limits based on identity (distribute mining rewards evenly, only allow one address per person), but then you have no privacy. Also this relies on the idea that no one can spoof someone's identity through whatever identity recognition system is used.


The reason Satoshi used PoW is because you can't spoof processing power. And of course he wanted to avoid the hassles and issues involved with identities.
Regarding face recognition, there is no privacy for the unique node. This may be another development direction of blockchain, which can prevent legal elements from using token transactions. The most important thing is that everyone can use mobile phones to mine, so that blockchain data can be distributed on everyone's mobile phones. Is it safer and more accessible for ordinary people to participate


If you destroy the whole concept just to make it more accessible you haven't done anything good.

Sure you can do one-id-one-vote mining but in order to do that you are going to have to rely on trusted third parties, and then you no longer have decentralization. Facial recognition wouldn't work well because you'd have people stealing other people's faces and generating fake faces and whatnot.

Also at this point you wouldn't be "mining". There would be no point in PoW mining if this is your setup because the idea you've set forth is to make the minting process available to anyone so you're not using competition for resources any more. You're using a identification system, where it is decentralized and secured through IDs. So in your setup there is no mining, there's just running your ID node from your phone and you are participating in the transaction verification network.

Anyway, since the essence of your system is based on identification what you really need is to involve trusted third parties that can verify IDs. There must be a whole legal apparatus put in place in case someone steals someone else's ID on the network. At which point instead of creating a more decentralized Bitcoin you've created a centralized trust-required crypto-banking industry with universal basic income. That's a very different thing.
Mobile phone transfer face recognition, mobile phone lock screen face recognition, and identity verification face recognition in your life are all based on trusted third parties. Why can they use a trusted third party when I just want to generate unique code using face recognition instead of using a trusted third party? Does anyone in your current life steal your face to unlock your phone, or transfer money through the stolen face?
"I just want a person to have only one node, and face recognition is the most feasible method I can think of currently.". Why must a person have only one node? Because if it weren't for this, the power of the node would be concentrated in the hands of wealthy people, like Bitcoin and Ethereum today.
Post
Topic
Board Development & Technical Discussion
Re: A more decentralized blockchain technology than Bitcoin
by
yingyush55
on 31/03/2023, 16:42:52 UTC

Through face recognition technology, each person can only generate one node, a private key, and a public key.
By setting an upper limit on the computing power of each miner, it is possible to mine on a mobile phone without consuming too much computing power.
If we are interested, we can design such blockchain technology to run on mobile phones together.

I can find your face by just looking up your social media account, mean while can you discover the scalar that leads to these coordinates when multiplied by the secp256k1 curve Generator? :

x: 83260022149314607220299798889850163602723925708954433956645823190147746092229
y: 32501403673747823353314855720543420818188361784787882868346931239726724853713

If not I believe bitcoin is much more sound than your facial recognition scheme, do you want someone trying to cut your face off to access your funds? Is that better than protecting against data breaches really?
Stealing my facial images or videos cannot be obtained through face recognition, just as you cannot use photos or videos to recognize faces on your phone. A face can generate a corresponding string of codes, but this code cannot infer what a person looks like.

Not to mention that a government authority would then have the REASON to strap you down to a chair, peel your eyes open, and make you stand still while they scan your face to extract blockchain data. This is a terrifying idea.
The government has tied me to a chair and left me standing with my eyes open, unable to scan my face and extract my blockchain data. What is the difference between tying my body directly and asking me for my bank card password? If this is a security risk, then you know that your bank card password also has the same problem.

You are now just being delusional, if the government does that very thing to you they WOULD be able to access all your blockchain secrets immediately (assuming people actually used your insane face id chain) it would be trivial as THEY NOW HAVE COMPLETE ACCESS TO YOUR ENTIRE REAL FACE AND HEAD.

As for how is this a bigger security risk than a bank? It is just as insecure as a bank, BITCOIN is way more secure than this scheme because private keys can be stored OFF OF YOUR BODY.
The government can force me to perform the facial actions required for face recognition, and similarly, it can force me to say the bank card number and password, or the private key of Bitcoin. In terms of absolute power, no encryption technology will work. Bitcoin is not stored outside of my body, it is also stored in my brain.
Post
Topic
Board Development & Technical Discussion
Re: A more decentralized blockchain technology than Bitcoin
by
yingyush55
on 31/03/2023, 15:32:48 UTC
Ethereum is a decentralized blockchain platform that allows developers to create and deploy decentralized applications (dApps). Ethereum uses a consensus algorithm called Proof of Stake (POS), which is designed to be more energy- efficient and decentralized than Bitcoin's Proof of Work (PoW) algorithm.
Is ETH really decentralized? Can you think about it carefully and see if he changed the underlying code from pow to pos after all the nodes voted for it? Will these miners agree to convert to pos? Their development team can change the underlying code themselves, and this power is concentrated in a few hands
Post
Topic
Board Development & Technical Discussion
Re: A more decentralized blockchain technology than Bitcoin
by
yingyush55
on 31/03/2023, 12:41:36 UTC
By setting an upper limit on the computing power of each miner, it is possible to mine on a mobile phone without consuming too much computing power.
Sounds terrible. First of all, why would you want to intervene in miners' belongings? What's the problem with one individual owning more than 100 others do? Secondly, why do you find this more decentralized? To forbid someone from using multiple faces, KYC must be implemented; and even that way, you can't be sure the mechanism won't be violated. Thirdly, how will the minting work?

Yet, another well intended idea which would have opposite results.
First of all, if I don't interfere with the miners' computing power, it will become more and more concentrated in the hands of wealthy people like Bitcoin. "A person who owns 100 other people's facial codes is stealing other people's information. If it is approved by others, why not?". Each of us is a node of the blockchain, and we can also give up the power that the blockchain gives us. The operation of coinage is that everyone is mining with their mobile phones on
Post
Topic
Board Development & Technical Discussion
Re: A more decentralized blockchain technology than Bitcoin
by
yingyush55
on 31/03/2023, 07:46:15 UTC

Through face recognition technology, each person can only generate one node, a private key, and a public key.
By setting an upper limit on the computing power of each miner, it is possible to mine on a mobile phone without consuming too much computing power.
If we are interested, we can design such blockchain technology to run on mobile phones together.

I can find your face by just looking up your social media account, mean while can you discover the scalar that leads to these coordinates when multiplied by the secp256k1 curve Generator? :

x: 83260022149314607220299798889850163602723925708954433956645823190147746092229
y: 32501403673747823353314855720543420818188361784787882868346931239726724853713

If not I believe bitcoin is much more sound than your facial recognition scheme, do you want someone trying to cut your face off to access your funds? Is that better than protecting against data breaches really?
Stealing my facial images or videos cannot be obtained through face recognition, just as you cannot use photos or videos to recognize faces on your phone. A face can generate a corresponding string of codes, but this code cannot infer what a person looks like.

Not to mention that a government authority would then have the REASON to strap you down to a chair, peel your eyes open, and make you stand still while they scan your face to extract blockchain data. This is a terrifying idea.
The government has tied me to a chair and left me standing with my eyes open, unable to scan my face and extract my blockchain data. What is the difference between tying my body directly and asking me for my bank card password? If this is a security risk, then you know that your bank card password also has the same problem.