Simple, Easy, Risk and Change

Simple, Easy, Risk and Change

I want to share some of what I’ve been thinking about along those lines. What follows is a collection of some of my evolving thoughts relating to change and complexity. Let me know your thoughts in the comments.

4

Simple vs Easy

A few years ago, the creator of Clojure- did a talk called “Simple made Easy”. It’s an excellent talk, and I highly encourage you to watch it. In the talk, he talks about the difference between simple and easy. One is a measure of complexity, the other of effort.

One of the really interesting takeaways, is that we need to measure complexity separately from the effort that it takes to produce.

Change and Risk

All change introduces risk. Period. The amount of risk that is introduced does vary wildly with each change. Are you changing well understood code? Are you changing simple code? Are you changing well tested code? Are the changes well peer-reviewed?

Change is so risky, that decades ago we started writing software in a manner that reduced the amount of change we would have to do. We built plugin systems. We built configuration files. We built hooks and event systems to try to isolate change to minimize risk.

Optimizing for Simple

I think the reason is that we’ve trained ourselves to optimize for Easy. If you look at frameworks and libraries, they try as hard as possible to make themselves easy to use. They use the term “easy” as a tag-line. And they do so for good reason (marketing).

When you optimize for “easy” it becomes trivial to say “well, I don’t need a test here, the code is too hard to test”. When you optimize for “easy” you can fall into the trap of saying “this is just glue code, it doesn’t need a test”.


function makeFoo(): Foo {
return new Foo();
}


function makeBar(Foo $foo = null): Bar {
return new Bar($foo ?: makeFoo());
}


About the author

admin administrator

Leave a Reply