'The Dropbox for IPFS' // 'Hide in plain sight'

Why do I need horcrux?

horcrux is a client which can hopefully resolve some of the drawbacks IPFS has.

• Data on IPFS is permanent! So you better be careful of what you put on there
• Your data is on someone else's computer, how do you secure it?
• IPFS can get slow and if you're uploading a huge file, it might take a while
• A teeny tiny change to a huge file? You'll have to upload the entire file again!

What does horcrux do?

• horcrux uses deduplication to find similar patterns in a file and breaks down the file into those chunks.
• It then encrypts those chunks and pushes them to IPFS
• It keeps track of which encrypted chunk belongs to which file.
• It can then recreate the file based on this info.

In this way:

• Only chunks with updates will be pushed to IPFS, saving time and bandwidth
• Everything that's pushed is encrypted

Visual representation

This project is currently under development!

TODO:

• Find a more elegant way to generate and download blk and enc files without changing pyrabin
• Refactor code to be more elegant, refactor variable names to be consistent
• Introduce a config file which can have the encryption password and other details
• Reintroduce UI and hook it up with the horcrux via Flask
• Come up with unit tests

Dependencies and credits

horcrux is written in such a way that all the dependencies it uses can be swapped out easily. However the prototype was made with the help of these fine repos:

• python3
• ipfsapi (https://ipfs.io/docs/getting-started/)
• simple-crypt (https://github.com/andrewcooke/simple-crypt/)
• pyrabin (https://github.com/aitjcize/pyrabin)
• DragAndDrop UI (https://github.com/pcote/DragDropProject)