2 ms·
I'm really, really curious to how it actually works under the covers. Needless to say, I'm sceptical, primarily because "near plaintext speed" fundamentally isn
by hn_throwaway_99 4y ago
I'm really, really curious to how it actually works under the covers. Needless to say, I'm sceptical, primarily because "near plaintext speed" fundamentally isn't currently possible with true fully homomorphic encryption, at least in my understanding.
For example, they give an example of running queries against data that is never decrypted. I'm very curious as to how they do this. I've used blind indexes [1] to solve the "encrypted data searchability problem" in the past, but with blind indexes you're still left with the fact that you can only do exact matches - you can't sort the data or use less than/greater than queries.
With true FHE you should be able to sort results, but my understanding is that it's several orders of magnitude slower than plaintext searching, so I'm very curious as to what Vaultree is actually doing.
1. https://medium.com/@joshuakelly/blind-indexes-in-3-minutes-making-encrypted-personal-data-searchable-b26bce99ce7c https://medium.com/@joshuakelly/blind-indexes-in-3-minutes-m...
- bawolff 4y ago> "near plaintext speed" fundamentally isn't currently possible with true fully homomorphic encryption, at least in my understanding. If you skip the security requirements, applying rot13 twice is a fully homomorphic scheme that achieves plaintext speeds ;)
- fragmede 4y agoIt's up to them to define "near", but I'd consider https://spiralwiki.com/ https://spiralwiki.com/ (Samir Menon and Prof. David Wu at UT Austin) "nearly" usable compared to regular wikipedia.