Carl first demonstrated a serial port interface from Rebol in Montreal, as a simple project he undertook for controlling a water heater:
@kealist then took this on to bring to R3-Alpha, and Ren-C.
In the intervening time, I learned more about the device model and ports (mostly that there wasn't much of a design there). Experiences with trying to factor this code contributed to that knowledge.
It was an early aggressive user of the libRebol API, as non-essential code was untangled from the low-level core APIs. So it has had an impact.
But to try and rein in the project's sprawling scope, I think it's time to say that the serial code is something that we can put aside. Especially since we don't have a great way to test it in CI.
I won't even try to lobby @kealist to get a serial device out of the closet to spin more cycles on it...and instead point to the more interesting efforts (like UPARSE, or Blazor interop...)
Once a port model is implemented that can accommodate websockets and other interesting things, I'm sure the serial port code could be wired up by a sufficiently-motivated-individual.
Archived repository at: https://github.com/metaeducation/rebol-serial