46 ms·
There's a bunch of projects trying to solve the distributed mass storage problem. I am not convinced this problem exists. (1) Hard drives are dirt f'ing cheap.
by api 9y ago
There's a bunch of projects trying to solve the distributed mass storage problem. I am not convinced this problem exists.
(1) Hard drives are dirt f'ing cheap. Buying more than one and putting them at different locations (home and office etc.) is not hard. Storage is still experiencing a Moore's Law type geometric increase in density so this is going to get easier not harder in the future.
(2) Cloud storage is cheap too. S3 is cheap. S3 reduced redundancy is really cheap. Backblaze B2 is stupid cheap. There's no privacy problem here if you encrypt your data on your end. (1) and (2) are not mutually exclusive. They go together nicely. Keep your data local and then back it up encrypted to cheap cloud storage. Many very affordable NAS devices will do this for you.
(3) Enterprise users categorically don't want a system where they don't know who's running their infrastructure. In some cases such a thing would be prohibited by data residency regulations. You will never ever ever convince a CIO to use this. In some cases it might be illegal for them to use this.
(4) The real problem for end users is data handling UX, not actual storage. (1) and (2) are adequate for virtually all users.
(5) The web does have an ephemerality problem but I fail to see how this will solve it.
(6) These systems are mostly too heavy for mobile and don't cope well enough with node ephemerality for laptops. They only work well on desktops and servers and those already have mega-cheap storage and backup options in the form of (1) and (2). Look into how much disk you can cram in a desktop these days for under $500.
(7) The anonymous/uncensorable niche is already filled by Tor, I2P, and Freenet.
This really looks like a solution to a non-existent problem. Am I missing something?
Not trying to hate, just questioning.
- stale2002 9y agoThink about all the extra space on all the harddrives in the world that isn't being used. For example, when you buy a new computer, with a 500 GB drive, 400GB of it isn't going to be used for a while. Hard drives may be cheap, s3 may be cheap, but that extra 400 GB of hard disk space? That space is literally FREE. Hard disk space is weird, because there is so much completely unused space out there, that would be literally free to use. Filecoin will allow this free space to be put on the market.
- lordCarbonFiber 9y agoI think the OP's point is the cost of infrastructure means that this space on the market will never be less than the cost of S3 or local storage. So why buy the "free space", which leads to the bigger question. why sell it. Because operating a node for File Coin is certainly not free in terms of network costs and requires the opportunity cost of giving up your local storage for the duration.
- okabat 9y agoYou pay the operational burden of keeping your computer networked 24/7, and likely receive cents in exchange.
- int_19h 9y agoHow many people already keep their computer on and working 24/7 (or close enough) in practice, though?
- dfinch 9y agoI buy your arguments. However, I am curious (primarily as a thought experiment): what would happen if Filecoin was able to solve (3) - (8)? By opening up the barriers to entry and creating a perfectly competitive marketplace, economics says that the price of storage would then equal the marginal cost of providing that storage (i.e., electricity, networking, and amortized hardware cost). Presumably that's lower than what Amazon and others charge. Would love to see a theoretical analysis of Filecoin's impact on the storage market, presuming that it can build the foundations for competing in that market.