In the current version of FASTER C++, on Linux, we use libaio for async IO handling. It is a known issue that libaio is not very efficient. Recently, io_uring is released with Linux kernel 5.1, which advertises to be a high performance aysnc IO library. It would be useful to try it and see if we can improve disk performance on Linux by replacing libaio with it.

More details for io_uring

CE have released their own take on tracing, with following notable changes compared to original ZIO Tracing:

• Instead of parsing lambdas, CE throws exceptions in IO constructors and cleans up stacktraces. That also means there is no regional control of tracing and AFAIK no way to avoid slight overhead even when tracing is completely disabled – but OTOH it may probably work on Scala.js
• Trace

Let tofu-optics users do this:

import tofu.optics.syntax.tupleN._

val lense = Tuple3._2[Int, String, Long]   // : Contains[(Int, String, Long), String]

tupleN object should contain all possible implicit extensions to Tuple{N} companion objects. Implementations should be macro generated.