Threads Basics

A threading.Thread is an OS-backed thread managed by the Python runtime. In CPython (the runtime almost everyone uses), the Global Interpreter Lock (GIL) lets only one thread execute Python bytecode at a time, so threads don't run Python bytecode in true parallel — but the GIL is released during I/O and by some C-level operations, so threads still overlap I/O and blocking work. This course uses threads to practice coordinating concurrent tasks — the coordination patterns are the point, not raw CPU speedup.

import threading

def worker():
    print("hello from a thread")

threading.Thread(target=worker).start()

Thread.start() begins running target in a new thread and returns immediately — the calling thread keeps going without waiting. Python's default (non-daemon) threads do keep the process alive until they finish, even if you never call .join(), but that isn't the same as your own code knowing when a thread is done. Call .join() on a thread to block until it completes, which is how you safely use a value a thread produced — the sum_nums challenge below needs both halves' sums before it can add them, so it must join both threads first. (A thread started with daemon=True is the exception: the process kills it outright on exit, unjoined.)