What I want is to gain the Fiber's high efficiency on context switching and yield CPU time during I/O, but when the workload is enough to feed more CPUs I would like to gain the performance from the fact of do things in parallel.
I read this great article about Threads and Fibers https://octarineparrot.com/article/view/getting-more-fiber-in-your-diet
And read this question regarding of comparison between fibers and threads. Fibers over Threads in D
But I want to get any light about a good way to mix them agnostic and gracefully. Of course, code samples are welcome.