Hacker News Reader

IMO i think there are only a couple of long term solutions for ethereum with regard to scaling. Sharding is one. I think setting agreed markers and signatures at points in time is what they're going to be forced to do though. POS is a big big risk and i haven't heard a solution to those issues yet. Not even close.

I actually hope they figure it out. But they do have to do it quickly. For now we have PoV. Proof-of-Vitalik. That'll just have to do.

View on HN

I am curious to know what types of "big big risk" POS presents. As far as I can tell, there are several POS cryptocurrencies today and they seem to work fine. The Ethereum team claims to modify their POS approach to avoid known issues like the 'nothing at stake' problem. But it would be interesting to know what other possible risks you're talking about.

They work fine if there's trust.

Non-proof-of-work consensus always ends up depending either on trusted third parties or majority vote. Anyone can create a valid-looking non-PoW chain in a short time, and there's no way to separate valid chains from each other. In Bitcoin, when you want to bootstrap, you connect to n nodes and if just a single node is honest you will get the right chain. For non-PoW, if there's disagreement, you have to trust either a majority on what the right chain is, or add trusted nodes that overrule a majority vote. For proof-of-work there will always be a chain with most work done, provided >50% of miners are honest (hard dependency).

I have written a bit longer on the subject here: https://runeksvendsen.github.io/blog/posts/2017-06-14-bitcoi...

Reply on news.ycombinator.com