I have made control programs for an emulated Wang 2200T and a real Arduino Uno R3, fiddling around with massive memory constraints (32kb RAM in case one (The language is T-BASIC) and 2KB of RAM in case two (However 31.5 KB of storage; The variant is Arduino C)), and actually getting functional programs going that can do things (mostly). In case one, I read over some manuals on how to write a program to a floppy to copy the system there, but alas, no dice. The interface of it, though, is as top notch as it can be. Case two, my current development platform, is the one that actually has some abilities (I/O, though only output ready as is (apart from the console, which is the only input available), I plan on implementing networking somehow later), one of them being able to run up to three regular Arduino programs in the same flash (Escape key not yet implemented; This also allows programs to read variables from others, thus allowing piping, and for more advanced cases, use the escape mechanism to actually switch programs). I figured it also needed delay to allow commands to function properly (The timer is stupid inaccurate on this thing (0.5% off circa) and actually take arguments. It can change this delay in runtime. It can also change its baud rate in runtime (However, only useful on the Leonardo to which it has been ported, since reinitializing the serial console on the computer at a higher baud resets an Uno due to there being a separate USB microcontroller which temporarily gets disconnected, which somewhy restarts the main Atmel controller). And, for when I tested it on my friend's Leonardo, I did write a soft reset for these cases where we'd actually need to null out only variables we need to null, instead of a complete hard reset. I also used relatively few ifs (Undesired behavior) and a lot of gotos (No other statement really took care of getting me out of a ten-level nest to root in a single statement) which returns the user to the prompt when the command has done executing. Also, considering this is serial (Raw, no less) and the board's limitations, I can chain short scripts of up to sixty five characters into the board's RAM which it then executes in correct sequence.

If source code is wanted, it is happily provided.

The reason for "proof of concept" and all that? I am going to found a company, and the system architectures I've used so far will be the CLI base of a system I'll be making later.

coding sounds fun i have a freiend who is into coding i'll reference this info to him as i'm learning as well