Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Snowflake to Avalanche: A Novel Metastable Consensus Protocol Family for Crypto

ipfs.io

Post reply on HN
Print view View original

31–40 of 53 posts

Re: Snowflake to Avalanche: A Novel Metastable Consensus Protocol Family for Crypto

#31

Post by zodiac »

gtrubetskoy wrote: post #26
apo wrote: post #12

Initial reactions: 1. Sybil-resistance (faking strong consensus by deploying cheap replica nodes you control) in a protocol like this is crucial. All I could find is this: To prevent Sybil attacks, it uses a mechanism like proof-of-stake that assigns weights to participants in committee selection based on the money in their accounts. 2. Every non-proof-of-work protocol I've seen, including Ripple Consensus Process an…

Add to that - I do not see any mention of the source of timing or "timestamp server" as the Nakamoto paper refers to it. The details of why it is important are in my blog post here https://grisha.org/blog/2018/01/23/explaining-proof-of-work/ (which was once at the top of HN).

I don't think your post shows that any public ledge must provide a timestamp server (which I think is your claim?). For instance PBFT (with a fixed validator set) works perfectly fine without providing a timestamp server since it's asynchronously safe - you can take as long as you want to complete every PBFT round.

Basically, PoW provides consensus by providing an absolute timestamp (we know that at the difficulty adjustment equilibrium, a certain block header must have taken 10 minutes to produce), but this does not imply that a consensus algorithm that all public ledger consensus algorithms must produce an absolute timestamp (the later claim is stronger).

Re: Snowflake to Avalanche: A Novel Metastable Consensus Protocol Family for Crypto

#33

Post by wildbunny »

apo wrote: post #12

Initial reactions: 1. Sybil-resistance (faking strong consensus by deploying cheap replica nodes you control) in a protocol like this is crucial. All I could find is this: To prevent Sybil attacks, it uses a mechanism like proof-of-stake that assigns weights to participants in committee selection based on the money in their accounts. 2. Every non-proof-of-work protocol I've seen, including Ripple Consensus Process an…

Agree completely. Nice paper, utterly useless in practice due to sybil attack.

edit: not only that, but it will be vulnerable to all kinds of history attacks for bootstrapping nodes

Re: Snowflake to Avalanche: A Novel Metastable Consensus Protocol Family for Crypto

#34

Post by gtrubetskoy »

zodiac wrote: post #31
gtrubetskoy wrote: post #26

Earlier quoted context omitted.

Add to that - I do not see any mention of the source of timing or "timestamp server" as the Nakamoto paper refers to it. The details of why it is important are in my blog post here https://grisha.org/blog/2018/01/23/explaining-proof-of-work/ (which was once at the top of HN).

I don't think your post shows that any public ledge must provide a timestamp server (which I think is your claim?). For instance PBFT (with a fixed validator set) works perfectly fine without providing a timestamp server since it's asynchronously safe - you can take as long as you want to complete every PBFT round. Basically, PoW provides consensus by providing an absolute timestamp (we know that at the difficulty ad…

My post merely shows that PoW is the mechanism via which this absolute timestamp (I like this terminology, BTW) is provided and seems very important, but I think the real question (to which I do not know the answer) is: can it be proven that an absolute timestamp is essential or non-essential, because it would answer the question of whether a distributed ledger without PoW (timelock puzzle) is fundamentally possible.

Re: Snowflake to Avalanche: A Novel Metastable Consensus Protocol Family for Crypto

#36

Post by stri8ed »

simias wrote: post #30
mkirklions wrote: post #8

At this point, are these papers simply to reduce the cost of verification? It seems thats the only problem in the crypto world, but I dont know if verification will ever be scalable.

What do you mean about "cost of verification"? Verification is not too difficult, it's reaching a consensus among trust-less nodes that is. PoW solves the problem by making it so that any node which receives two valid but conflicting versions of blockchain has an objective metric to decide which one is the "right" one. The answer being whichever has the most work put into it. Since you can't fake work you can't arbit…

Verification becomes difficult, when the system gains adoption and processes thousands of TX per second. Then the issue becomes about the trade-off between node resource requirements, and decentralization. Fundamentally, every node having to process/verify every single transaction in the network,is not scalable.

Re: Snowflake to Avalanche: A Novel Metastable Consensus Protocol Family for Crypto

#37

Post by KirinDave »

woah wrote: post #29
apo wrote: post #12

Initial reactions: 1. Sybil-resistance (faking strong consensus by deploying cheap replica nodes you control) in a protocol like this is crucial. All I could find is this: To prevent Sybil attacks, it uses a mechanism like proof-of-stake that assigns weights to participants in committee selection based on the money in their accounts. 2. Every non-proof-of-work protocol I've seen, including Ripple Consensus Process an…

Is PoW really a “clean distribution mechanism”? Economically, it’s like proof of stake, but all stake is divided up among the small set of people who have the electrical connections and equipment to mine profitably.

This statement is not always true.

May deployed PoW implementations are provably worse than most modern proposals for Proof of Stake. PoW mining opens selfish mining strategies, whereas Proof of Stake fixes the set of actors opens to scrutiny the mechanism for "who gets to mine the next block."

This doesn't mean that proof-of-stake is magical, but it's certainly less prone to issues than Proof of Work. It's also less inundated by religious zeal; PoS proposals face healthy skepticism and more vetting BEFORE they tend to be deployed. PoW is the axiomatic and beloved sacred "nakamoto consensus" in (incorrectly, but to many in the space) a platonic form.

Re: Snowflake to Avalanche: A Novel Metastable Consensus Protocol Family for Crypto

#38

Post by zodiac »

gtrubetskoy wrote: post #34
zodiac wrote: post #31

Earlier quoted context omitted.

I don't think your post shows that any public ledge must provide a timestamp server (which I think is your claim?). For instance PBFT (with a fixed validator set) works perfectly fine without providing a timestamp server since it's asynchronously safe - you can take as long as you want to complete every PBFT round. Basically, PoW provides consensus by providing an absolute timestamp (we know that at the difficulty ad…

My post merely shows that PoW is the mechanism via which this absolute timestamp (I like this terminology, BTW) is provided and seems very important, but I think the real question (to which I do not know the answer) is: can it be proven that an absolute timestamp is essential or non-essential, because it would answer the question of whether a distributed ledger without PoW (timelock puzzle) is fundamentally possible.

Yeah, I'll have to think a lot harder about full PoS to maybe answer that question. One thing to note us that there are certain hidden timing assumptions (eg in casper-ffg, there is the 4 month unbonding period, and the timescale over which an inactivity leak occurs) that are necessary for security (they seem to correspond to the time scales needed for weak subjectivity social consensus and hard forks respectively)

I think the most precise name might be something like "affine timestamp"; you don't actually need timestamp wrt the big bang but being able to measure the rate-of-time is stronger than merely being able to order events

Re: Snowflake to Avalanche: A Novel Metastable Consensus Protocol Family for Crypto

#39

Post by snissn »

wildbunny wrote: post #33
apo wrote: post #12

Initial reactions: 1. Sybil-resistance (faking strong consensus by deploying cheap replica nodes you control) in a protocol like this is crucial. All I could find is this: To prevent Sybil attacks, it uses a mechanism like proof-of-stake that assigns weights to participants in committee selection based on the money in their accounts. 2. Every non-proof-of-work protocol I've seen, including Ripple Consensus Process an…

Agree completely. Nice paper, utterly useless in practice due to sybil attack. edit: not only that, but it will be vulnerable to all kinds of history attacks for bootstrapping nodes

And it's very odd that the paper doesn't acknowledge or anticipate those concerns.

Re: Snowflake to Avalanche: A Novel Metastable Consensus Protocol Family for Crypto

#40

Post by SkyMarshal »

stri8ed wrote: post #36
simias wrote: post #30

Earlier quoted context omitted.

What do you mean about "cost of verification"? Verification is not too difficult, it's reaching a consensus among trust-less nodes that is. PoW solves the problem by making it so that any node which receives two valid but conflicting versions of blockchain has an objective metric to decide which one is the "right" one. The answer being whichever has the most work put into it. Since you can't fake work you can't arbit…

Verification becomes difficult, when the system gains adoption and processes thousands of TX per second. Then the issue becomes about the trade-off between node resource requirements, and decentralization. Fundamentally, every node having to process/verify every single transaction in the network,is not scalable.

Hence the ongoing debate around whether verification should be less computationally costly than computation (BTC) or whether a system can be architected to successfully scale even when verification == computation (ETH).
Post reply on HN