Exercise 1: Declarative vs. Imperative Provisioning, With Examples — Possible Solution
====================================================================

DECLARATIVE: you describe the DESIRED END STATE, and the tool figures
out what changes are needed to get there.

Example: a Terraform config file states "I want 3 web server
instances and 1 load balancer." If only 1 web server currently exists,
the tool computes on its own that it needs to create 2 more, without
being told the individual steps -- and if 5 instances already exist
instead of 3, running the same config again would compute that 2
need to be REMOVED to match the desired state. The config always
describes the END GOAL, not the path to it.

IMPERATIVE: you write a specific SEQUENCE OF STEPS to be executed, in
order.

Example: a shell script that runs "create server 1," then "create
server 2," then "create server 3," one command after another. If this
script is run a second time without modification, it doesn't know
that 3 servers already exist from the first run -- it would simply
try to create 3 MORE servers, resulting in 6 total, because it has no
concept of a "current state" to compare against; it just executes its
fixed sequence of steps every time it's run.

WHY THIS WORKS AS AN ANSWER
------------------------------
The examples specifically illustrate the chapter's own distinguishing
property -- a declarative tool reconciles CURRENT state against
DESIRED state and computes the difference, while an imperative script
just re-executes its fixed steps with no awareness of what already
exists -- which is exactly why the chapter notes imperative scripting
"doesn't inherently know how to safely reconcile drift" the way a
declarative plan/apply cycle does.