Use built-in chips: Your HDL programs will most likely include chip parts that you've built in project 1. As a rule, though, we recommend using the built-in versions of these chips instead. The use of built-in chips ensures correct, efficient, and predictable simulation. There is a simple way to accomplish this convention: make sure that your project directory includes only the .hdl files of the chips developed in the current project.

​

Implementation order: We recommend building the chips in the order in which they appear in Chapter 2. However, since the simulator features built-in versions of these chips, you can use chip-parts without first building them: the simulator will automatically use their built-in implementations.

​

The Hack ALU produces two kinds of outputs: a "main" 16-bit output resulting from operating on the two 16-bit inputs, and two 1-bit "status outputs" named 'zr' and 'ng'. We recommend building this functionality in two stages. In stage one, implement an ALU that computes and outputs the 16-bit output only, ignoring the 'zr' and 'ng' status outputs. Once you get this implementation right (that is, once your ALU.hdl code passes the ALU-nostat test), extend your code to handle the two status outputs as well. This way, any problems detected by ALU.tst can be attributed to the incremental code that you've added in stage two. We thank Mark Armbrust for proposing this staged implementation plan, and for supplying the test files to support it.