I see a lot of developers over-optimize or prematurely optimize their code.

Adding linting and testing and type-checking and transpilation and continuous integration and feature flags and an edge network can be powerful engineering tools. And when a project is just kicking off, they can also make you feel like you’re getting things done when you’re not at all.

I’d argue that, early in a project, it’s a lot more important to get something functional stood up so you can start experimenting with it, find the rough edges, and identify your engineering needs.

Prioritize working code over perfect code.

Similarly, I see developers try to abstract away their code bases into ever more complex spiderwebs of functions that import functions that import functions. I see complex function structures to account for variations, without considering how it makes the code harder to read and maintain.

Early on, it’s probably better and more effective to copy/paste/modify than the focus on abstraction.

Because at that phase of development, the implementation details change too much. You’ll probably just have to refactor your code again anyways.

But the flip-side is also true.

It’s easy to slip into this habit of moving fast and coding up the next feature, and never actually go back and clean things up.

You code base becomes a different kind of tangled mess, with lots of copy/pasted directories that do nearly the same exact thing, with just one or two small variations.

A lot of the code could be abstracted and combined, making the whole thing easier to manage and maintain. But doing it after a certain amount of code hoarding has taken place is a lot harder.

How do you resolve this tension?

I like to get working code stood up, and iterate and refine what it does first.

Once I’ve got it really dialed in, I’ll go back and clean up how it works under-the-hood as an implementation detail.

Doing that requires slowing down for a minute, though, and not everyone allows themselves that time.

Figuring out what that looks like is my specialty. I love helping people and organizations figure out how to structure their code base, how to refactor code for easier reuse and maintenance, and what the right mix of tooling is for the project.

If you’re struggling with this stuff, get in touch!

Like this? A Lean Web Club membership is the best way to support my work and help me create more free content.