Hacker News Reader

Comment by duskwuff | Hacker News Reader

And what's worse, the mining process has some rather absurd system requirements: 128+ GB of RAM, 1 TB of SSD cache storage (separate from the data being stored!), and a high-end GPU is "recommended". The sense I get is that it's impossible to casually offer up some spare storage from a typical home system, or even from a NAS -- the system requirements effectively limit mining to people willing to spend a couple thousand dollars on hardware.

https://docs.filecoin.io/mine/hardware-requirements/

View on HN

If you take a look at the top storage providers, there's dozens who are providing 10 PB+. These are folks that have invested millions to earn FileCoin.

https://stats.filecoin.io/d/z6FtI92Zz/filecoin-chain-stats?o...

Moreover, it's greatly disingenuous to claim that this system lets casual users let their HD space given the system requirements are so high.

Overall bandwidth in the system is 125TB/week. Filling up the storage provided by "dozens who are providing 10 PB+" would require about 20 years at this rate. And another 20 years to download it back

This raises a giant red flag for me. Storage is storage. This is not like hash mining, which is based on probabilities. You either have the data or you don't. If you have the data, even partially, this should be useful to the system. I can conceive of no reason why a reasonably designed distributed storage system would have such insane requirements.

Security, mostly - if the proofs fail, it’s not just the data that could be corrupted - bad actors could attack the consensus itself and steal unlimited funds. The params for the proofs could be loosened, at the cost of extra risk of an attack. They’ve gone the safe route and picked challenging parameters that would be uneconomic to brute force attack.

If you want to use spare space to "mine" crypto, look at Chia from the creator of BitTorrent.

The space isn't used for anything useful like IPFS though.

I wish Wuala was still around, damn Lacie for buying - then removing the most interesting part of project (exchanging local storage for cloud storage)... then killing the entire thing. https://en.wikipedia.org/wiki/Wuala#Trading

Their cryptree [0] design for access control was also pretty stunning. We've adopted and improved it in Peergos.

[0] https://raw.githubusercontent.com/peergos/peergos/master/pap...

Yeah seems like a weird combination of both PoS and PoW which gets the best of neither. Would've been cool to see a novel PoW here that actually translates into actual "work" for the network.

That seems crazy, especially considering Sia and others have functioning alternatives with no such insane requirements. Maybe they're trying to corner some certain performance oriented market, but that guarantees nothing about network performance which I suspect will be a much larger impact than any of it.

Besides the GPU, that's a pretty basic server config. Shouldn't cost more than 200/mo

or "borrow" the hardware and power from their school/employer either without any authorization or for the ostensible purpose of "testing/burn-in"

Reply on news.ycombinator.com