The article’s overall point (break hard-looking things down into manageable chunks) is good, and it presents a nice, easily-grasped example of what a simple GC might look like.
However, like an OS, a filesystem, a programming language interpreter/compiler, or any number of other things, it’s also an example of something where a straightforward brute-force proof-of-concept implementation is simple enough for one person to knock together over a few hours/days, but building one that offers enough performance/features/generality for people to actually want to use it balloons the complexity by an order of magnitude or two (or twelve).
The article’s overall point (break hard-looking things down into manageable chunks) is good, and it presents a nice, easily-grasped example of what a simple GC might look like.
However, like an OS, a filesystem, a programming language interpreter/compiler, or any number of other things, it’s also an example of something where a straightforward brute-force proof-of-concept implementation is simple enough for one person to knock together over a few hours/days, but building one that offers enough performance/features/generality for people to actually want to use it balloons the complexity by an order of magnitude or two (or twelve).