Effective OOP

Carl

Why Use Objects?

QA with students; I’d say the gist of it is that thinking with objects lets one do software engineering. Goal of this session would be to have the students experience some ways in which objects enable best practices.

Dividing Work

interfaces / contracts as a way to divy up a job

describes which side does what
having them lets you focus on one piece at a time
or if working in a team, dividing up work between developers
disciplined use keeps you from entangling ideas too much
when used with good API thinking, also a powerful way to communicate about a data structure

Use some of the Python OOP syntax to have take something bad and make it good?

Do Not Repeat Yourself

Using abstract classes to provide facade. Using generics (aka, typing / interfaces) to writing algorithms once (e.g., sorting)

Thinking with Tests

Revisit mock objects.

If you’ve translated requirements into tests, and then digested those requirements into a series of tests, also corresponding with tests, can think of combination of interface + tests as being the specification for objects.

Self Documenting Code

Objects are a means of saying these things go together.

Also, a way to make other things not be concerned with the details. Let’s the code speak more to what it’s doing.

OPP Danger Zone

Too Much Magic & YAGNI

When things are too abstracted to be able to figure out what’s going on where.

Hard to debug problems – often because interpreter reports errors in strange places.

End up doing unnecessary work - unnecessary work means more opportunity for real error.

Inheritence vs Composition