YouTube has no limit on amount of video that you can upload. This means that it is effectively infinite cloud storage if you were able to embed files into video with some kind of tool. ISG is the tool.
This has been quite heavily inspired by suckerpinch's [Harder Drive](https://www.youtube.com/watch?v=JcJSW7Rprio) video and [discord as a filesystem](https://github.com/pixelomer/discord-fs). Unfortunately no filesystem functionality as of right now
I doubt there is any part of the TOS saying that you can't upload videos containing files, but I also did not want to shovel through all the legalese. I still don't condone using this tool for anything serious/large. YouTube might understandably get mad.
If you want to or already have went through the hassle of installing Rust, you can ```git clone``` this repository, then ```cargo build --release``` and run the program. Probably better in case I forget to update the executable.
The principle behind this is pretty simple. All files are made of bytes and bytes can be interpreted as numbers ranging from 1-255. This number can be represented with pixels using one of two modes: RGB or binary.
The cooler mode. Every byte perfectly fits inside one of the colors of an rgb pixel. One rgb pixel can contain 3 bytes at a time. You just keep adding pixels like this until you run out of data. It is leagues more efficient and quick than binary.
Binary:
Born from YouTube compression being absolutely brutal. RGB mode is very sensitive to compression as a change in even one point of one of the colors of one of the pixels dooms the file to corruption. Black and white pixels are a lot harder to mess up. Every pixel is either bright representing a 1 or dark representing a 0. We string these bits together to get bytes and continue until we run out of data.
Both of these modes can be corrupted by compression, so we need to increase the size of the pixels to make it less compressable. 2x2 blocks of pixels seem to be good enough in binary mode.
To make it easier on the user, we also include all the relevant settings used to create the video on the first frame of the video. This allows the program to know what mode the video is in and what size to use in order to avoid making the user remember.
