I apologize if I was rude or dismissive but I've found it rare that new users from other communities post here and stay long enough to keep an open dialog but you seem to be an exception so welcome! Could you elaborate on your point regarding offline nodes? If I'm understanding correctly, I don't believe this to be much of an issue because if a node (delegate) goes offline and misses a block then it just goes to the next node... if the offline node consistently misses blocks for a long duration they can be voted out relatively quickly.
No worries, I understand where you're coming from
Let's me start by saying I'm not nearly as well-versed in DPoS as I am in PoS.
The idea is that the rolling checkpoints allow nodes that have been continuously aware of the state of the blockchain to refuse forks that started at a height smaller than current_height - X (whether the criteria is height or time doesn't change much).
However, when bootstrapping a new node, the client cannot independently determine which version of history is the main chain. Indeed, although an attacker might have been able to create a longer branch (through a grinding attack for example) which will have been rejected by the online nodes thanks to the rolling checkpoints, the new node has no way of knowing that.
In Bitcoin, the client just needs to look at which chain has the most total trust. With rolling checkpoints receiving the genesis block isn't enough to guess which branch is the mainchain.
The question is "should the consensus mechanism be built around an absolute rule that cannot be superseded ("main branch is the one with most total trust") or is there an alternative"?
As I said, I'm not sure that this is a fundamental problem but some people in the cryptocommunity seem to think this is unacceptable.
I'm reasoning by analogy with PoS here so I might wrong about how this affects bitshares.
Moreover, I wish I had more time to study Bitshares
but I don't have enough time
If anyone knows how bitshares handle the issue of the new nodes & the nodes come back online I'm very curious
Great job on trying to convince the PoW maximalists that PoS ain't all that bad. We need people spending some effort on outlining the issue properly without bias. It seems like most guys here, and I guess NXTers as well have just given up on trying to convince the PoW camp. Lately it seems Ethereum guys, Vitalik in particular, has made a real effort to understand the pros and cons of both consensus methods. I know Daniel had some discussions with him as well.
What do you think of Vitaliks' recent style of argumentation?
Thanks a lot
Like pretty much anyone else in the community I have a great deal of respect for Vitalik and have been following his blog posts for quite some time.
I think he's been doing a lot for PoS by vouching for it and writing about it given the fact that he's certainly one of the most respected personalities of the scene.
His posts are extremely technical so I highly doubt that many people have the time and/or the knowledge to fully understand them (me included
) They seem to really find an echo within a limited scene (which is not necessarily a bad thing, he's definitely one of the main progressive forces research-wise in the field).
What's remains extremely difficult is to coordinate our efforts. Proof-of-stake algorithms are so complicated that I don't know anybody that has a profound understanding of all of them (I know for a fact that Vitalik isn't very familiar with Peercoin's implementation).
For all these reasons, we decided to make our paper as accessible as possible. I think that understanding the pros and cons of each implementation would definitely be of great help to the community.
NeuCoin's goal - as far as the white paper is concerned - is to reach to the community (especially pro-PoW people on reddit for example or btctalk) to try to spark a discussion. It might be utopic but it's worth the try
An excellent paper. Thank you for sharing this.