This question involves architecting an organizational system for optimizing the learning of technical material via different types of highlights. The intent of this question is to define, with exacting specificity, what is "important enough" to highlight, with the goals of highlighting as little as possible, increasing signal, and decreasing noise.<p>Let's say that you were studying a new programming language book using Apple's iBooks app. iBooks affords six different highlight types: green, yellow, red, blue, purple, and underline.<p>1. How would you use these different highlight types to optimize your learning of the material?<p>2. What would you consider to be the principal benefit of your particular approach?<p>For example, hypothetically, one might reserve, say:<p>* Underline for "novelties", i.e., concepts that are entirely new.<p>* Green for benefits/pros of the language.<p>* Red for drawbacks/cons of the language.<p>This approach might be said to have the principal benefit, during review of the highlights, of reinforcing exposure to new concepts, and allowing one to know when best to use (or not use) this particular language.<p>Some other possibilities may include: best practices, anti-patterns, idioms, caveats, things that confuse you, and things that are contrary to your expectation.<p>It may be best that one's highlight types are mutually exclusive such that one highlighted passage can't fall within more than one highlight type, and it further may be best that each highlight type is "easy to judge" such that cognitive load isn't increased in trying to deliberate whether or not a given passage fits a given highlight type.<p>Imagine also that you'd be able to -- should you so optionally desire -- export the highlighted passages (with their types) for external use.