Exercise 2: A Bytes-Type Contract Violation — Possible Solution
====================================================================

THE APPLICATION
------------------------------
    def string_body_app(environ, start_response):
        start_response('200 OK', [('Content-type', 'text/plain')])
        return ["this is a plain str, not bytes"]  # violates the real WSGI contract

start_response is called correctly here - the violation is specifically
in the returned iterable's own element type: PEP 3333 requires "an
iterable yielding zero or more bytestrings," and a plain Python str is
not a bytestring.

THE REAL, VERIFIED RESULT
------------------------------
Run against a real wsgiref server and hit with a real urllib request:

    Server-side traceback:
        AssertionError: write() argument must be a bytes instance

    Client-side result:
        urllib.error.HTTPError: HTTP Error 500: Internal Server Error

WHY THIS WORKS AS AN ANSWER
------------------------------
It isolates exactly one contract violation (the body's element type)
while keeping start_response's own call correct, then reports both the
real server-side AssertionError and the real client-side HTTPError -
showing wsgiref actively enforces the bytes-only body requirement at
runtime rather than silently accepting a str, and that the failure
surfaces to a real client as a genuine HTTP 500, not a hang or a
malformed response.