In the linked article, the author also included some videos of himself implementing some of the presented parts e.g. exclusive Fibers [0], including writing the necessary C code using the Ruby C API.

Recording yourself writing code and then publishing it takes quite a lot of courage, even moreso if it's a complex topic such as the one presented. Additionally, it's such a valuable resource: You can follow the whole implementation step by step.

That's just awesome.

[0] https://www.youtube.com/watch?v=cNaqbeagqUw

That's even super awesome. I didn't watch the whole thing yet, but he seems to comment extensively, as opposed to staying quiet on the inner thoughts during the whole thought process – which is of course the most crucial part.

Thanks for sharing.

This is really cool. Do you have any recs for similar vid of someone hacking on c python API? I want to gain a better understanding of how these langs interface w their implementation, but Ruby isn't really in my wheelhouse.

The internals aren't THAT different, so while I can't really answer your question (I'm not working on CPython), you might still find it interesting from the POV as to how interpreters work.

Thanks for the link. This is great!

It's fascinating how many event-driven IO libraries and associated web servers have come in and out of fashion in the Ruby community over the last decade. I hope this or another effort finally sticks and becomes part of the stdlib, so we can have a real lasting ecosystem built around it.

Why would one want to replace nginx as a reverse proxy?

nginx is one of the most battle tested code on earth and I don't see a benefit in replacing it with a ruby implementation.

Also why are there a new http server coming out for ruby like every 3 years?

Is it not suitable to just put nginx at front and the app server in the back?