πfs - revolutionary file system without data storage

What is it?

πfs is a revolutionary file system that, instead of wasting space on your hard drive, stores all the data in π. You will never have to worry about free space again! Have you been told that 100% compression is not possible? Yes, here it is!

How to assemble?

πfs is going to be elementary:

./configure
make

And using it is not more difficult:

πfs -o mdd=<metadatadirectory><mountpoint>

Where metadata directory is the directory with metadata (file names, offset in π), and mountpoint is the mount directory.

What does π do with my data?

π is one of the most important constants of mathematics, and it has a bunch of interesting properties (which can be found in the Wikipedia article )
One of the properties of the number π is supposedly normal, which means that all its numbers are distributed evenly, provided that it is a disjunctive sequence, i.e. all finite numerical sequences are inside it. If we consider the base π number 16 (HEX), then this assumption is true. The first record of this was in 2001 .
Well, if so, then why do we need to store all these exabytes of data on hard drives, if we can find them in π?

And what, right every-every file can be found in π?

And THAT! Any file that you or someone else created, or have not even created yet! Are you being tried for copyright infringement? Yes, to hell with two - you just got the sequence from a constant, it has been there all its life!

But how do I find my data in π?

If you know the file offset and its length in π, you can get them using the Bailey – Borwein – Plouffe formula . Similarly, you can use this formula to find out the offset of your file.
Since finding the whole desired sequence will take too much time, we need to split the file into several small sequences to increase performance.
In this implementation, we split the file into separate bytes, and we look for these bytes in π.

Well, I found the file offset in π, what should I do with it?

Well, obviously, you need to write it somewhere, on a piece of paper, for example, but now we have a bunch of free space on the hard drive, why not save it there?

Well, if I lose my bias?

It's okay - your data from this from π will not be lost! Find someday.

Why is this shit so slow? It took me 5 minutes to save 400 lines of the test!

Well, this is a prototype, and besides, we have Moore’s law!

GitHub source