Exercise 3: Why LC-3 Postpones Real-Hardware Quirks — Possible Solution
====================================================================

Two specific quirks this course's use of LC-3 lets you avoid until
cpu8bit1:

1. 6502 -- its famously tiny register set. Per the chapter's own
"Why Not Just Learn a Real Chip First?" section, the 6502 was built
with so few registers largely because MOS Technology needed it to
sell for $25 in 1975 -- fewer transistors meant a cheaper, more
competitive chip. That's a manufacturing-cost decision, not a
teaching decision.

2. Z80 -- its 8080-compatibility baggage. A meaningful chunk of the
Z80's instruction set exists purely so it could run 8080 software,
not because those instructions represent some ideal, necessary set
of operations a CPU should have.

WHY POSTPONING THESE HELPS RATHER THAN JUST DELAYS THE LEARNING
------------------------------
If a learner met the 6502's tiny register set before ever learning
what a register even is or why a CPU has them at all, the two
questions -- "what is a register, generally" and "why does THIS chip
specifically have so few of them" -- would get tangled together. It
would be easy to mistake a 1975 manufacturing constraint for some
universal truth about how registers have to work.

By learning registers first on LC-3 -- a chip designed with a clean,
deliberate register file (8 general-purpose registers, R0-R7, chosen
for pedagogical clarity) -- the concept of "a register" gets
established on its own terms first. Then, when cpu8bit1 introduces
the 6502's much smaller register set, it reads as a genuine, nameable
trade-off ("this chip sacrificed register count for manufacturing
cost") instead of unexplained noise. The same applies to the Z80's
8080-compatible instructions: once "why does a chip have the
instructions it has" is already a familiar question from LC-3's own
clean design, the Z80's answer ("backward compatibility") lands as a
real, understandable trade-off rather than one more unexplained fact
to memorize.

WHY THIS WORKS AS AN ANSWER
------------------------------
It picks one quirk per chip exactly as asked, ties each directly back
to the chapter's own stated reasons (manufacturing cost for the 6502,
backward compatibility for the Z80), and explains the postponement's
actual pedagogical benefit -- concepts learned cleanly first turn
later quirks into recognizable trade-offs -- rather than just
asserting that "it's easier this way."