Exercise 1: A MovingAverageTracker Observer — Possible Solution
====================================================================

THE NEW OBSERVER
------------------------------
    class MovingAverageTracker:
        def __init__(self):
            self.prices = []
        def update(self, symbol, price):
            self.prices.append(price)
        def average(self):
            return sum(self.prices) / len(self.prices)

Follows this chapter's own observer shape exactly - a single
update(symbol, price) method matching PriceLogger and AlertService's
own signature, so Stock's _notify() loop can call it identically to
every other attached observer without any special-casing.

VERIFYING IT ALONGSIDE THE EXISTING OBSERVERS
------------------------------
Attaching logger, alert, AND tracker to the same Stock('ACME', 100),
then sending three price updates (120, 160, 140):

    tracker prices: [120, 160, 140]
    tracker average: 140.0
    manual average: (120+160+140)/3 = 140.0

All three observers received all three notifications - tracker's own
average is confirmed to match a manual calculation exactly, computed
independently of whatever logger and alert were doing with the same
three notifications.

WHY THIS DEMONSTRATES OBSERVER'S OWN VALUE FURTHER
------------------------------
Stock's own set_price()/_notify() code needed zero changes to support
a third, differently-behaved observer - exactly the same finding this
chapter already established with two observers, now confirmed to hold
for three. Each observer keeps its own private state (a log list, an
alerts list, a running price list) completely independently; Stock
never needs to know any of that exists.

WHY THIS WORKS AS AN ANSWER
------------------------------
The new observer matches the established update(symbol, price)
interface exactly, is verified running alongside the two existing
observers rather than in isolation, and its own computed result is
cross-checked against a manual calculation.