Discussion
Your definition is more useful and practical.
Mine is more definitional and categorical.
A script is interpreted by an interpreter, then machine code instructions execute. It's a run time affair, turning bytes in the script file into instructions for the CPU.
A program is compiled by a compiler from source into machine code and then stored in a file for later execution.
The program is the compiled code file.
Python scripts are all "scripts" run by the interpreter "program."
@pseudonym @KatS @zwol I used to draw the line this way, but after doing a lot requiring knowledge of linkers and loaders, I've come to realize how much extra work has to be done to your average ELF binary before it can be executed. Indeed, the ELF loader is often called an interpreter because it's ultimately taking the ELF file and turning it into an actual runnable binary in memory.
@pseudonym That's one of the distinctions I considered and dismissed, because it breaks down quickly.
Off the top of my head: Java is compiled into bytecode and dynamically compiled into machine code at runtime, and Bitbucket was (and presumably still is) implemented in Python - that's a pretty hefty set of "scripts."
Meantime, my test scripts in Common Lisp are compiled-then-executed by the "interpreter", as are my exploratory bits of introspection on code and objects when I'm debugging.
@zwol
This is a bonfire demo instance for testing purposes
