Rodrigo Setti

yes, unfortunately, C++ will beat racket no matter how much we optimize. Things to try: * Use [Racket Profile](https://docs.racket-lang.org/profile/) to identify the bottleneck * Using more of typed racket improves...

Yes, that would be the obvious place. I guess the sub-problems we need to solve to optimize the performance of your benchmark case are: 1. an efficient test to decide...

My bad, I misunderstood you! I thought you're referring to the `path-record%` class (path-record.rkt). The class intent of the class is simply to be a place where: 1) you can...

Nice! makes a lot of sense. I didn't know Racket Real's didn't use machine floats. Good learning.

@2mol the attributed is for a different implementation: https://github.com/rodrigosetti/messagepack

Thanks Gautam! would like to send a PR?

Thanks for trying to compile in Ubuntu. I don't have a Linux installation readily available right now... can you please remove the "-stdlib=libc++" part from [CMakeLists.txt line 49](https://github.com/rodrigosetti/azteroids/blob/master/CMakeLists.txt#L49) and try...

@ryankennedyio I don't get that... but I get a message saying that the notebook is old and it needs updating. I committed the updated version and now it's running without...

:+1: sure