This project is awesome! I've always wanted to implement a userspace content addressable filesystem and one fine day I found this which does it in just a few lines of js code : )
Instead of content addressing at the file level, how about splitting files into chunks using rabin fingerprints and content addressing those. That would improve deduplication, make snapshotting more space-efficient and also speed up synchronization.
Thoughts about this? Would it be worth the overhead?
I'm interested in implementing this functionality.
This project is awesome! I've always wanted to implement a userspace content addressable filesystem and one fine day I found this which does it in just a few lines of js code : )
Instead of content addressing at the file level, how about splitting files into chunks using rabin fingerprints and content addressing those. That would improve deduplication, make snapshotting more space-efficient and also speed up synchronization.
Thoughts about this? Would it be worth the overhead?
I'm interested in implementing this functionality.