Exercise 2: Why the Capstone Chose Separate Directories Over Workspaces — Possible Solution
====================================================================

Chapter 8's own conclusion, reached via Exercise 2's own comparison,
was that a team managing a genuinely high-stakes production environment
should prefer separate directories over workspaces, specifically
because workspaces only isolate STATE -- the configuration, backend,
and (critically) provider credentials are all still shared across
every workspace. A mistaken workspace selection can still execute a
real `apply` against real production infrastructure using real
production credentials, since nothing about workspace selection itself
changes which credentials are available in that shell session.

This capstone's own environment is exactly the kind of case that
reasoning applies to: it includes a production database with
`prevent_destroy = true` protection (Step 5) precisely because an
accidental destroy would be severe -- the same category of "genuinely
high-stakes" resource Chapter 8's own Exercise 2 discussion was built
around. Using workspaces here would mean the dev and prod copies of
this environment share the same credentials and the same
configuration, with only state kept separate -- so a team member
intending to test something in a dev workspace, but accidentally still
selected into prod, could still run a real, credentialed `apply`
against the production database and the rest of the production
environment, with nothing structurally preventing it.

Separate `environments/dev/` and `environments/prod/` directories, each
with their own backend key and credentials, remove that failure mode
by construction: someone working in the dev directory simply doesn't
have production credentials available in that context at all, so even
a serious mistake can't reach the production resources this capstone
is specifically trying to protect with Step 5's `lifecycle` block.

WHY THIS WORKS AS AN ANSWER
------------------------------
This ties the capstone's specific choice directly back to Chapter 8's
own stated reasoning (shared credentials as the real risk with
workspaces) and connects it concretely to this capstone's own
production database and its Step 5 protection, rather than restating
Chapter 8's general conclusion without applying it to this specific
scenario.