4 ms·
It’s good to see more research and development funds being invested in alternative consensus mechanisms beyond PoW/PoS. Solana looks interesting given it’s hig
by mattdesl 5y ago
It’s good to see more research and development funds being invested in alternative consensus mechanisms beyond PoW/PoS.
Solana looks interesting given it’s high transaction throughput and novel PoH algorithm - assuming it can scale in a decentralized way beyond the support of its foundation.
Though, unfortunately it still seems quite challenging to run a Solana validator (which ultimately defines the health & decentralization of the chain). The current specs and docs are pretty daunting last I checked; hopefully that can be improved in time.
Edit:
I found this article to be a good introduction to the idea of PoH. It appears Solana is a mix of PoH (the novel high throughput validation mechanism) and PoS (used to incentivize validation and penalize negative behaviour).
https://lsmod.medium.com/what-makes-solana-the-fastest-public-blockchain-cb8ffde1a7f4 https://lsmod.medium.com/what-makes-solana-the-fastest-publi...
- Vespasian 5y agoI'm quite interested in this topic. Does PoH encourages the "excessive" use of hardware and energy like PoW or PoTaS (chia) does? I'd be very interested in any solution which scales its resource usage with usability/throughput rather than token price. (I'm probably missing the correct terms)
- SheinhardtWigCo 5y agoYes, it is PoW, plus brilliant marketing.
- mattdesl 5y agoI haven’t looked too deeply at it yet - but I don’t think it would lead to more excessive energy costs than PoW. It’s using PoS in addition to their PoH validation algorithm; and you have to “stake” or lock some amount of tokens up in order to validate and earn rewards. So you wouldn’t simply be able to spin up hundreds of validators (GPUs) because you’d quickly have no funds left to stake.
- dlubarov 5y agoI wouldn't expect PoH to lead to much energy use, since nodes are evaluating a single hash chain. Since the computation can't be parallelized, computational throughput isn't a factor; only latency is.
- mech422 5y agoI can't remember which coin it was, but we were looking at one the other day that required about $20K of hardware for a full node and 2 workers nodes to go with it... Just insane. Won't be playing with a lot of the new gen. coins with beer money :-P
- remolacha 5y agoDefinitely agree. Also, one minor point: PoW/PoS are technically Sybil control mechanisms, not consensus mechanisms. PoW/PoS rate-limit block producers, but actual consensus is achieved by something like Practical Byzantine Fault Tolerance. There are only a few blockchains that actually use a different mechanism for consensus, such as Avalanche.
- baby 5y agoThey are also a consensus mechanism: it’s called leader election. You need a system for participants to agree on who will be the next person that get to choose the next block of transaction.
- btown 5y agohttps://docs.solana.com/running-validator/validator-reqs https://docs.solana.com/running-validator/validator-reqs describe a high-end dedicated machine, perhaps out of range for most individual hobbyists. But it seems perfectly sane for any reasonable-scale business building nontrivial features on top of Solana to run its own validator. And https://docs.solana.com/clusters https://docs.solana.com/clusters describes exactly how to run validators that connect to the test and mainline-beta networks. IMHO that's more than enough to provide robust decentralization. I can't speak to Solana's other claims but there seems to have been a ton of thought put into the scalability of the system.