Warning I as AnonyMint proposed this in mid-2013 and referred to it as Proof-of-Storage. I also discovered it was fundementally flawed. If you continue, you will be wasting your time. Eventually I will come back and explain to you and Andrew Miller PhD why this is flawed.
Details are definitely needed here. I know about Proofs-of-Space and Proofs of Space-Time(
http://eprint.iacr.org/2016/035.pdf ), Miller's Proof-of-Retrievability(Permacoin), and White's Proof-of-Storage in Qeditas. What's Anonymint's proposal about?
We found some possible drawbacks and attack vectors in Permacoin, but no fundamental flaws.
In my original analysis in 2013, I went down the same rat hole of flaws as in section "4.2 Local-POR Lottery" of the white paper. They assume so many things (including for example that Amazon bandwidth is 10 - 100X more expensive than dedicated host), and when you work through all the analysis, then the scheme does not work to prevent centralization of the mining, and thus the permissionless (uncensored) and robustness/durability of file storage attribute will not be sustained.
However, revisiting this has given me a new idea that I want to think about regarding using
hard-disk storage space in a PoW hash. Hmmm.
P.S. I also read these:
https://bitcointalk.org/index.php?topic=186601.0 (this was an offshoot I think of my idea which was in
my March 2013 thread)
https://bitcointalk.org/index.php?topic=555375.msg6536798#msg6536798