CHAPTER 33 VISUAL LABS — boards 33a and 33b

These are original fictional teaching examples. Work locally.
Read the named boards first. Predict before running, and explain any
difference before changing a second thing.

HOW TO REPRODUCE
Open tracking.html. Predict the message after two clicks, then click twice.
Expect Local events: 2. In developer tools Console inspect
window.dataLayer
It contains two objects, both named practice_check with training_form.
This page includes no Google scripts and makes no network requests.

CHANGE / PREDICT / CHECK
Reload to reset the local example and click five times. Expect five local
practice records and zero real stored enquiries. These are occurrences,
not distinct people. Use a second run with one click to check that the
result follows the action rather than always displaying five.

BOARD 33b — DESIGN AND TEST THE BUSINESS CONDITION
Inspect production_test_plan.csv. It is a hypothetical test plan, not an
implemented receiving service. Explain why a server rejection must create
zero confirmed-lead events and repeated confirmation of the same stored
record should leave one event under this definition. Merely counting
clicks cannot meet that requirement.
To investigate doubled counts, inspect distinct handlers/installations.
Registering the exact same function twice for the same event and options
does not itself double that listener. Repeating this exact script can
instead create duplicate-const errors. Read the actual console evidence.
