Ambraven 
I'm sticking with nalgebra for now because it turns out I also need QR decomposition and I haven't found it in the other crates proposed. I figured out a few things that makes it work...
Well, I'm sure there is a better and more optimized way to write it, but I got it working.
Still it was interesting to try to do the same (supposedly) basic stuff with several crates.
Spoiler, it still a pain, no matter the crate.
I don't think I can post my code since I'm basically trying to port *bad* C++/Eigen/assimp code from work. Maybe a very small part, so someone can tell me how bad my code is and offer a correction.