SEGRETO

Devlog #2:
It’s been a long time since my last devlog (during which I learned RUST (thanks to Freecodecamp and BekBrace). In the end I was finally able to put my Rust crypto library to the test against PyNaCl to see if writing raw bindings was actually worth the effort. I wrote a benchmark script to measure execution times for encryption, signatures, key exchange, and hashing across both backends, but only after running a cross check to guarantee that both libraries produce bit for bit identical results. The results were amazing: Rust won ten out of the thirteen test cases with a mean of 2x gain respect PyNaCl and a 2x lost in 3 test cases.
In the end I opted for RUST for it’s velocity and possibility to control the raw code.

I’ve started this project this afternoon and I’m already surprised by the complexity I’ve found in it by exploring different peer to peer (P2P) libraries. My first idea was to build S.E.G.R.E.T.O. as a web app using HTML, CSS, and JavaScript. I investigated PeerJS because it makes web connections easier, but I misunderstood the role of its server and thought it stored messages. So as I didn’t want to depend on a server or on several pcs I researched a local P2P architecture where every device acts as both client and server. So after an excruciating research I’ve finally opted for a web app using HTML for graphic + python to host server and I’ve found out that I can create my own python wheels with rust (so wish me luck and I’ll dive right into this mess of learning practically everything).
The screenshot is teh beginning of the REPO