MADR CLI interface: Febuary 3, 2026

Authors

Joe Starr, Ph.D.

I specialize in computational knot theory. I’m also a professional embedded software engineer.

Context and Problem Statement

The CLI needs a collection of command line arguments to be useful. We need to decide on what the shape of those arguments/options should be.

Decision Drivers

• Simplicity, the interface should favor simplicity over features.
• The interface has to be fast and flexible.
• The interface has to be obvious.

Decision Outcome

Multiple sets of options satisfy our needs. We will choose the following:

1. -c: Number of words to generate, default 10
2. -s: A stack size to allocate, default 100
3. input on standard in
4. output on standard out

Considered Options

1. -c: Number of words to generate, default 10
2. -l: A path to a file
3. -o: A path to an output file
4. -s: A stack size to allocate, default 100
5. input on standard in
6. output on standard out

Pros and Cons of the Options

-c

• Good, simple control of repetitions
• Good, "matches" DGL v1

-l

• Good, portable
• Good, doesn't need special termination
• Good, user friendly
• Bad, path manipulation can be annoying
• Bad, redundant with cat <> | pdgl

-o

• Good, portable
• Good, user friendly
• Bad, redundant with >

-s

• Good, simple control of repetitions
• Good, allows for system/run specific experimentation

stdin

• Good, simple
• Good, portable
• Bad, requires a special EOF terminal char

stdout

• Good, simple
• Good, portable
• Good, no special memory management
• Bad, path handling can be challenging

More Information

N/A

---