The Moby Dick Performance Benchmark (2022)

Posted on 12 Apr 2022

Is this the most urgent question in existence? Probably not, but I frequently find myself wondering: Exactly how many items should a todo list be capable of managing?

The Philosophy of Scaling

I believe that productivity software should be able to scale effectively to any "user-generated" content. While I don't expect a standard application to effortlessly process multi-gigabyte system logs, it should be able to handle anything a human can realistically write.

"If an app works well with 'Moby Dick,' then that’s a good indication that it will handle my needs also."

I recognize that Melville's prose is longer and more complex than anything I will ever produce. However, using it as a baseline provides a reliable stress test.

The "Workout" Process

To put an app through its paces, follow these steps:

• Paste the text of Moby Dick into the app.
• Repeat this process $2 \to 5$ times (or until you lose interest).
• Open the macOS Activity Monitor.
• Verify if the Memory usage remains within an acceptable range.

The Logic Flow

Technical Considerations

Modern computing power is immense; compared to the speed of a CPU, the text of Moby Dick is essentially $\text{tiny}$.

Mathematically, the test looks like this: $\text{Stress Level} = \text{Length of Moby Dick} \times \text{Repetitions}$

If an application fails this basic test, it suggests there may be underlying architectural problems. However, remember that this is merely a benchmark:

• A "good" app for your specific workflow might fail this.
• A "bad" app for your needs might pass it.

Test Resources

Below are the files I use to benchmark my own software, such as the Bike Outliner.

Example Data Structure

If you were to look at the raw data being injected, it would look something like this:

# Moby Dick
Call me Ishmael. Some years ago—never mind how long precisely—having little or no money in my purse...
[...thousands of lines of text...]

Note: Ensure you are monitoring the Memory tab specifically to catch memory leaks during the repetition phase.