Re: Proof-of-stake can never scale without blowing up, because PoS isn't trustless
Charles has agreed that PoS requires centralized checkpointing:
https://www.youtube.com/watch?v=6zP4Chk8g7A#t=1632
https://www.youtube.com/watch?v=6zP4Chk8g7A#t=1632
You will need some hard checkpoints. In my opinion, a widely disbursed and actively staked coin does not need a checkpoint server. So the centralization would come from the source code, just like the protocol rules do for every coin.
you either have checkpoints or you dont. having just some hardcoded checkpoints makes no sense?
Here is the difference,
a hard coded checkpoint gives you the protection of a checkpoint, but does not give you a single point of failure for your coin like a checkpoint server would.
Because if a Hacker can control your Checkpoint server , he can control your coin.
A Hacker can do nothing against a hard coded checkpoint , which is why PoS & PoW Devs used hard coded checkpoints.
Kito explained it best.
So much anger. I have read previous posts debating this topic. One thing that comes up is POS needs checkpoints but POW doesn't. I believe POW and POS needs to have checkpoints. Here are my thoughts. Actually these ideas are not mine just what I have read previously on the same topic. But for new users this might help.
There is a perception that a long POW chain has a lot computation work behind it. This is not always true. POW difficulty changes to meet the target block generation period. There is no guarantee that a longer block chain has more work than a shorter chain.
On POW chain it is implied that work is being done and a longer POW chain has more work on it. This is because on honest chains which have equal difficulty targets are for most part true. But on this topic we are not worried about honest nodes trying to reach consensus with long running nodes which already have the honest chains.
In this case we are talking "checkpoint POS vs POW". Or as I like to say, "what happens when a new node comes online in a hostile hacker zone- how does a client trust one long chain vs another."
A hacker / attacker could basically fake blocks /w fake creation times for POS and POW chains. For POW, the attacker would generate chains with lowest allowed difficulty level and then keep it this way while generating a longer block chain. In this way an attacker can create a valid long POW block chain with low work / energy while following all the rules. This is sometimes overlooked because people look at bitcoin chain and see tons of energy expended in the mining races. The difficulty target in POW is to maintain block generation time not to insure there is a provable amount of work being done on the chain.
Enter the trusty checkpoint. Without the checkpoints a POS or POW would fail in this scenario. A checkpoint hash however cements the block chain up to a given date. With checkpoints an attacker has to mount the attack after this point which is more difficult to do with both POW or POS coins. For POW, the attacker has to reduce the difficulty target using energy. For POS, he would have to purchase a stake to reduce difficulty to generate a new fake chain.
So I accept checkpoints. They are cheap and make it highly difficult for attackers to workaround it.
There is a perception that a long POW chain has a lot computation work behind it. This is not always true. POW difficulty changes to meet the target block generation period. There is no guarantee that a longer block chain has more work than a shorter chain.
On POW chain it is implied that work is being done and a longer POW chain has more work on it. This is because on honest chains which have equal difficulty targets are for most part true. But on this topic we are not worried about honest nodes trying to reach consensus with long running nodes which already have the honest chains.
In this case we are talking "checkpoint POS vs POW". Or as I like to say, "what happens when a new node comes online in a hostile hacker zone- how does a client trust one long chain vs another."
A hacker / attacker could basically fake blocks /w fake creation times for POS and POW chains. For POW, the attacker would generate chains with lowest allowed difficulty level and then keep it this way while generating a longer block chain. In this way an attacker can create a valid long POW block chain with low work / energy while following all the rules. This is sometimes overlooked because people look at bitcoin chain and see tons of energy expended in the mining races. The difficulty target in POW is to maintain block generation time not to insure there is a provable amount of work being done on the chain.
Enter the trusty checkpoint. Without the checkpoints a POS or POW would fail in this scenario. A checkpoint hash however cements the block chain up to a given date. With checkpoints an attacker has to mount the attack after this point which is more difficult to do with both POW or POS coins. For POW, the attacker has to reduce the difficulty target using energy. For POS, he would have to purchase a stake to reduce difficulty to generate a new fake chain.
So I accept checkpoints. They are cheap and make it highly difficult for attackers to workaround it.
