Analysis of Vitalik's 'A Guide to 99% Fault Tolerant Consensus'

aliashraf

legendary

Activity: 1456

Merit: 1177

Always remember the cause!

Quote from: buwaytress on August 21, 2018, 05:10:16 AM

Quote from: aliashraf on August 20, 2018, 03:58:19 PM

Quote from: buwaytress on August 20, 2018, 08:33:44 AM

Quote from: aliashraf on August 19, 2018, 04:44:08 PM

Firstly, Vitalik is not a prominent figure in cryptocurrency,
~shnip

Not a fan of Buterin by any means, but let's be fair, he is a "prominent figure in cryptocurrency". Whatever his beliefs, goals, purposes, however mature they seem to be growing as he ages... and by which ever means he has used to reach this status, it's arguing semantics. He is prominent, and he is influential.

I don't agree. He just pretends to mature and the primary sign of being matured is avoiding such a hypocritical act, pretending to be matured. His semantic arguments are ridiculously superficial.
And even if he knew things, I don't think it would be enough to be a prominent figure in crypto community, one should be committed to the cause which is decentralization primarily.

Buterin is permanently insisting that decentralization is in contradiction with performance and security (implying that it should be put away) this is not what a prominent crypto advocate is supposed to say.

Our job is not to announce crypto agenda dead, we are here to make it happen.

I actually agree with you that his views are off track. I think that he and others like him stand for corporate interests, rather than for decentralisation. I also believe there is a lot of hypocrisy in developers but as well as with us the community. But at the end of the day these conflicts of ideology and profit are easy to see, and we can't really fault the directions it has taken. Bitcoin as a commodity, and a highly valuable one at that, has resulted in this "market" psychology that's seeped in.

The true advocate for decentralised tech are quietly sticking to their work, they know all of this, and they don't need to be prominent nor influential.

Piece. Kudos.

buwaytress

legendary

Activity: 3010

Merit: 3724

Join the world-leading crypto sportsbook NOW!

Quote from: aliashraf on August 20, 2018, 03:58:19 PM

Quote from: buwaytress on August 20, 2018, 08:33:44 AM

Quote from: aliashraf on August 19, 2018, 04:44:08 PM

Firstly, Vitalik is not a prominent figure in cryptocurrency,
~shnip

Not a fan of Buterin by any means, but let's be fair, he is a "prominent figure in cryptocurrency". Whatever his beliefs, goals, purposes, however mature they seem to be growing as he ages... and by which ever means he has used to reach this status, it's arguing semantics. He is prominent, and he is influential.

I don't agree. He just pretends to mature and the primary sign of being matured is avoiding such a hypocritical act, pretending to be matured. His semantic arguments are ridiculously superficial.
And even if he knew things, I don't think it would be enough to be a prominent figure in crypto community, one should be committed to the cause which is decentralization primarily.

Buterin is permanently insisting that decentralization is in contradiction with performance and security (implying that it should be put away) this is not what a prominent crypto advocate is supposed to say.

Our job is not to announce crypto agenda dead, we are here to make it happen.

I actually agree with you that his views are off track. I think that he and others like him stand for corporate interests, rather than for decentralisation. I also believe there is a lot of hypocrisy in developers but as well as with us the community. But at the end of the day these conflicts of ideology and profit are easy to see, and we can't really fault the directions it has taken. Bitcoin as a commodity, and a highly valuable one at that, has resulted in this "market" psychology that's seeped in.

The true advocate for decentralised tech are quietly sticking to their work, they know all of this, and they don't need to be prominent nor influential.

aliashraf

legendary

Activity: 1456

Merit: 1177

Always remember the cause!

Quote from: buwaytress on August 20, 2018, 08:33:44 AM

Quote from: aliashraf on August 19, 2018, 04:44:08 PM

Firstly, Vitalik is not a prominent figure in cryptocurrency,
~shnip

Not a fan of Buterin by any means, but let's be fair, he is a "prominent figure in cryptocurrency". Whatever his beliefs, goals, purposes, however mature they seem to be growing as he ages... and by which ever means he has used to reach this status, it's arguing semantics. He is prominent, and he is influential.

I don't agree. He just pretends to mature and the primary sign of being matured is avoiding such a hypocritical act, pretending to be matured. His semantic arguments are ridiculously superficial.
And even if he knew things, I don't think it would be enough to be a prominent figure in crypto community, one should be committed to the cause which is decentralization primarily.

Buterin is permanently insisting that decentralization is in contradiction with performance and security (implying that it should be put away) this is not what a prominent crypto advocate is supposed to say.

Our job is not to announce crypto agenda dead, we are here to make it happen.

Quote

I will say though that the opening of the article is presumptious. I don't think more than a handful actually think Buterin's given birth to another consensus algorithm, much less a "black tech" one.

Well, it IS "black tech", because Buterin has monopolized Ethereum and this is not just an accidental grave digging by a hobbyist , it is purposefully done by the god of Ethereum to introduce a trusted/famous class of entities and more importantly he is announcing it as a consensus algorithm!

buwaytress

legendary

Activity: 3010

Merit: 3724

Join the world-leading crypto sportsbook NOW!

Quote from: aliashraf on August 19, 2018, 04:44:08 PM

Firstly, Vitalik is not a prominent figure in cryptocurrency,
~shnip

Not a fan of Buterin by any means, but let's be fair, he is a "prominent figure in cryptocurrency". Whatever his beliefs, goals, purposes, however mature they seem to be growing as he ages... and by which ever means he has used to reach this status, it's arguing semantics. He is prominent, and he is influential.

I will say though that the opening of the article is presumptious. I don't think more than a handful actually think Buterin's given birth to another consensus algorithm, much less a "black tech" one.

Quote from: aliashraf on August 19, 2018, 04:44:08 PM

~schnippp
the resurrected "signed messaging" proposal for "finalizing" blockchain is actually the final nail on the coffin of Ethereum as a decentralized cryptocurrency.

Also, wasn't the final nail on that coffin the DAO hack? Not sure if I'm trolling, even.

aliashraf

legendary

Activity: 1456

Merit: 1177

Always remember the cause!

Quote from: KingScorpio on August 19, 2018, 03:10:59 PM

Quote from: aliashraf on August 17, 2018, 03:24:00 AM

Quote from: KingScorpio on August 16, 2018, 08:39:40 PM

Quote from: aliashraf on August 16, 2018, 10:53:18 AM

The only shortcoming in your article is its comparative approach, imo. I admit that you have pointed out some limitations of both "original" and "implementation" versions (your terminology) but the most important point is missing and it is the key to understanding why Buterin published such a copycat version of the original Lampart work.

Historically, Satoshi chose "oral messaging" against "encrypted messaging" for bitcoin because of the latter's centralization consequences. In classical Byzantine Generals, analogically speaking, having famous generals with known signatures is a prerequisite for applying "encrypted messaging" schemas and establishing such a fame system is vulnerable to centralization.

Obviously you are aware of this, we all are, even Buterin (well, probably ) is aware of such a centralization consequence, the thing is he exactly wants it, a second layer "finalization supervisory" or something, with his org signature as one of the most important ones in the chain of command, I suppose.

Anyway, your comparative study was great. Thank you

buterin used his prominence to spread that concept. without him it wouldnt have been known by any.and forgett

What concept? There is a reason that a concept has not received enough attention through like half a century. Right?

So what is Buterin doing here? Kinda digging graves and resurrecting dead ideas? Who is he? Dr. Frankenstein?

whats the problem if he is a prominent person in the business and he puts old ideas out of the libary and puts them into the community to inspire them to find solutions to the current problems of the blockchain technology, then this is just buterins work, there is nothing wrong about it. its good that someone does that maybe we have luck and one day we will get a 99% secure algorithm

You are absolutely wrong, unbelievable.

Firstly, Vitalik is not a prominent figure in cryptocurrency, he is a trojan with zero belief in the most basic idea of this technology, decentralization. Right now he has put himself and his org in charge of everything in Ethereum and had gone so far to impose his old agenda for getting rid of PoW by his ridiculous "weak subjectivity" and Casper shit.

Plus, this specific "resurrecting" attempt is not just a simple act of digging graves, it is intentionally done to help him and his org to inject a "circle of trust" concept to the ecosystem by adding a "finalization" layer of "reputable" nodes that recognise each other's signatures. It is nothing less than a take-over strategy.

Ethereum is in danger of transferring power to big stakeholders (that Vitalik happens to be one of the biggest ones, thanks to the premined scam) by Casper and this one, the resurrected "signed messaging" proposal for "finalizing" blockchain is actually the final nail on the coffin of Ethereum as a decentralized cryptocurrency.

KingScorpio

sr. member

Activity: 1470

Merit: 325

Quote from: aliashraf on August 17, 2018, 03:24:00 AM

Quote from: KingScorpio on August 16, 2018, 08:39:40 PM

Quote from: aliashraf on August 16, 2018, 10:53:18 AM

The only shortcoming in your article is its comparative approach, imo. I admit that you have pointed out some limitations of both "original" and "implementation" versions (your terminology) but the most important point is missing and it is the key to understanding why Buterin published such a copycat version of the original Lampart work.

Historically, Satoshi chose "oral messaging" against "encrypted messaging" for bitcoin because of the latter's centralization consequences. In classical Byzantine Generals, analogically speaking, having famous generals with known signatures is a prerequisite for applying "encrypted messaging" schemas and establishing such a fame system is vulnerable to centralization.

Obviously you are aware of this, we all are, even Buterin (well, probably ) is aware of such a centralization consequence, the thing is he exactly wants it, a second layer "finalization supervisory" or something, with his org signature as one of the most important ones in the chain of command, I suppose.

Anyway, your comparative study was great. Thank you

buterin used his prominence to spread that concept. without him it wouldnt have been known by any.and forgett

What concept? There is a reason that a concept has not received enough attention through like half a century. Right?

So what is Buterin doing here? Kinda digging graves and resurrecting dead ideas? Who is he? Dr. Frankenstein?

whats the problem if he is a prominent person in the business and he puts old ideas out of the libary and puts them into the community to inspire them to find solutions to the current problems of the blockchain technology, then this is just buterins work, there is nothing wrong about it. its good that someone does that maybe we have luck and one day we will get a 99% secure algorithm

aliashraf

legendary

Activity: 1456

Merit: 1177

Always remember the cause!

Quote from: KingScorpio on August 16, 2018, 08:39:40 PM

Quote from: aliashraf on August 16, 2018, 10:53:18 AM

The only shortcoming in your article is its comparative approach, imo. I admit that you have pointed out some limitations of both "original" and "implementation" versions (your terminology) but the most important point is missing and it is the key to understanding why Buterin published such a copycat version of the original Lampart work.

Historically, Satoshi chose "oral messaging" against "encrypted messaging" for bitcoin because of the latter's centralization consequences. In classical Byzantine Generals, analogically speaking, having famous generals with known signatures is a prerequisite for applying "encrypted messaging" schemas and establishing such a fame system is vulnerable to centralization.

Obviously you are aware of this, we all are, even Buterin (well, probably ) is aware of such a centralization consequence, the thing is he exactly wants it, a second layer "finalization supervisory" or something, with his org signature as one of the most important ones in the chain of command, I suppose.

Anyway, your comparative study was great. Thank you

buterin used his prominence to spread that concept. without him it wouldnt have been known by any.and forgett

What concept? There is a reason that a concept has not received enough attention through like half a century. Right?

So what is Buterin doing here? Kinda digging graves and resurrecting dead ideas? Who is he? Dr. Frankenstein?

KingScorpio

sr. member

Activity: 1470

Merit: 325

Quote from: aliashraf on August 16, 2018, 10:53:18 AM

The only shortcoming in your article is its comparative approach, imo. I admit that you have pointed out some limitations of both "original" and "implementation" versions (your terminology) but the most important point is missing and it is the key to understanding why Buterin published such a copycat version of the original Lampart work.

Historically, Satoshi chose "oral messaging" against "encrypted messaging" for bitcoin because of the latter's centralization consequences. In classical Byzantine Generals, analogically speaking, having famous generals with known signatures is a prerequisite for applying "encrypted messaging" schemas and establishing such a fame system is vulnerable to centralization.

Obviously you are aware of this, we all are, even Buterin (well, probably ) is aware of such a centralization consequence, the thing is he exactly wants it, a second layer "finalization supervisory" or something, with his org signature as one of the most important ones in the chain of command, I suppose.

Anyway, your comparative study was great. Thank you

buterin used his prominence to spread that concept. without him it wouldnt have been known by any.

buterin has a great live now he is rich and he increasingly can use others thinking and inventing for him.

aliashraf

legendary

Activity: 1456

Merit: 1177

Always remember the cause!

The only shortcoming in your article is its comparative approach, imo. I admit that you have pointed out some limitations of both "original" and "implementation" versions (your terminology) but the most important point is missing and it is the key to understanding why Buterin published such a copycat version of the original Lampart work.

Historically, Satoshi chose "oral messaging" against "encrypted messaging" for bitcoin because of the latter's centralization consequences. In classical Byzantine Generals, analogically speaking, having famous generals with known signatures is a prerequisite for applying "encrypted messaging" schemas and establishing such a fame system is vulnerable to centralization.

Obviously you are aware of this, we all are, even Buterin (well, probably ) is aware of such a centralization consequence, the thing is he exactly wants it, a second layer "finalization supervisory" or something, with his org signature as one of the most important ones in the chain of command, I suppose.

Anyway, your comparative study was great. Thank you

Fireflyangle

newbie

Activity: 1

Merit: 4

A recent blog post by Vitalik titled has led many to think that a new consensus algorithm like black tech has been born. However, as Vitalik himself said, this consensus algorithm is still the classic Byzantine general problem algorithm. Through analysis, we can see that the research and innovation of consensus algorithm still need to follow the proven theories such as CAP. On this basis, it is possible to apply all kinds of classic distributed algorithms and encryption algorithms to the blockchain field, and get good results.
1. Distributed system theory
In fact, before the birth of the blockchain, computer science had already done a lot of research on the consistency problem, forming a strictly proven distributed system theory. The more classical theories include FLP and CAP.
The principle of FLP impossibility is: "in a minimum asynchronous model system with network reliability and node failure (even if there is only one), there is no deterministic algorithm that can solve the consistency problem. "That is, the theoretical lower limit of the consistency problem is unsolved. There is no common algorithm that can be implemented in any scenario in an asynchronous distributed system.
The CAP principle is called the CAP impossible triangle, that is, consistency, availability and partition tolerance cannot be met simultaneously, and a feature needs to be weakened to design a distributed system. Therefore, under the premise of FLP impossibility principle, CAP principle provides theoretical guidance for engineering practice.
More generally than CAP, there are definitions of network environment characteristics in distributed systems theory: safety, liveness, communication unreliable. Distributed systems typically balance safety against availability.
2, The classic solution to the Byzantine general problem
This article will not elaborate on the description of BFT problem itself. In addition to the problems of the Byzantine general, Lamport et al. also provided two solutions in their classic paper.
The first is the OM(m) protocol for "oral messages", which means that no encryption algorithm is allowed except for the encrypted security on the link. The protocol requires a large number of messages to be delivered recursively between two parties, so the message complexity is high, is exponential, and is not very practical. However, this algorithm still has a high value. Firstly, it lays a foundation for the birth of the polynomial level complexity agreement, namely, Practical Byzantine Fault Tolerance. In addition, 1/3 of the fault-tolerant nodes are proved to be the theoretical upper limit of the algorithm.
The second is the SM(m) protocol for "encrypted messages." This algorithm is different from the first one in that it USES the signature algorithm. Each node can generate a non-forgerable signature that can be validated by other nodes. When a message is received, the node determines and verifies that the message has been received by signing. When the message is finally no longer received, the message consensus ends.
This paper has proved that the second algorithm can be fault-tolerant to any number of nodes (of course, the network should include at least two normal nodes, otherwise meaningless). The specific process can refer to the original paper.
However, this algorithm also has its limitations. Different from the hypothesis of many Byzantine algorithms in an asynchronous or semi-synchronous network environment, it is assumed to be carried out in a "synchronous" network, ignoring the communication delay between network nodes. In addition, signature identity system information needs to be determined before network operation, which is difficult to realize extension. Thus, according to the CAP theory, the approach is to achieve A high degree of consistency (C) and availability (A) without regard to tolerance for situations such as network partitioning (P).
3. "99% fault-tolerant consensus algorithm" and comparative analysis
In contrast, let's move on to what Vitalik refers to as the consensus algorithm invented by Lamport to describe and simplify the implementation (hereinafter referred to as the "implementation version"). The version in the Lamport paper becomes the "original").The implementation version still retains the original digital signature system, that is, each node can generate a non-forgery signature and can be verified by other nodes.
With the original version, in order to achieve the messaging between nodes, realize the timeout specifies the message version of the algorithm, namely node receives a message, check for news in addition to see if the signature has been received and in the collection, but also check the time should not be later than the signature of the received message corresponding to the time node.

After the specified time (calculated according to the number of rounds), the node will stop listening and select a value from the checked valid message as the consensus result according to some determined rule.

By comparison, we can find that there is no essential difference between the two algorithms. The essence of the two algorithms is based on the signature system. And Vitalik implementation version of the consensus algorithm increased the delay time requirements. This design implementation is frequently encountered when writing implementations of consensus protocols to determine the propagation cycle of messages and to ensure that message propagation can end within a specified time.

In addition, the implementation version also discusses the Observer as an independent role to pass messages across the network. The observer, acting as a passive viewer in the network, can receive, examine, and forward (unsigned) messages directly to other nodes. This needs to introduce a different delay time for the observer to solve the problem of malicious nodes intentionally sending messages to the observer late to make the normal messages time out.
4. Application and inspiration
As discussed above, the main problem of this kind of consensus method is the high demand for synchronization of network and poor scalability. Another drawback of the implementation version is that the amount of messages is also large (the process of sending messages from N nodes in round n-1 to other N nodes, that is, the message complexity is), so this kind of consensus method is more difficult to apply directly in the actual scenario.
In order to be more applicable to the blockchain field, Vitalik also mentioned in his article that this method can be combined with other current consensus algorithms (such as PBFT, PoS, etc.). For example, the algorithm can be run at intervals for some specific time. However, if the assumptions related to the two consensus algorithms cannot be satisfied, the consensus algorithm will also fail, that is, the improved optimization cannot violate the original theoretical system.
However, we can still get a lot of inspiration: fully exploring the classical theory of distributed systems and transforming it into a consensus algorithm applicable to the blockchain domain can achieve unexpected results. For example, PBFT combined with nakamoto consensus, and SM(m) algorithm in BFT paper proposed by Vitalik combined with existing blockchain consensus.
In addition, trying to apply a variety of encryption algorithms in the security field may also achieve good results. For example, for example, the use of encryption algorithms such as aggregated signatures to optimize the consensus mechanism is also being attempted by the likes of ByzCoin, which can greatly reduce the communication complexity.
my facebook：https://www.facebook.com/bay.yu.988373， more question Contact me.

Topic: Analysis of Vitalik's 'A Guide to 99% Fault Tolerant Consensus' (Read 349 times)