Smaller Is Faster

“Smaller is faster” is a hardware design principle as you might already know it. Generally speaking, smaller pieces of hardware will be faster than larger pieces. I have always thought this principle in terms of computer hardware. Just a while ago, I started to think this principle is also applicable to software development. Throughout this post, I’ll try to give some insights and supporting examples for “smaller is faster” in software development.

The amount of data needs to be transferred has been important in software development. Obviously, the smaller data, the faster it’s transferred. Transferring fast is essential when it comes to web page loading. We want pages to load as fast as they can. As a result, we minify our JavaScript files. We also compress almost anything we can to make transfers as fast as possible. Thus, smaller data favors faster transfers.

Understanding a piece of code is an everyday activity for every programmer. It’s hard to focus when the code gets big. We have been trying to overcome code complexity by single responsibility principle and many others. Typically, we are trying to minimize or in other words make the code smaller so that we can understand it faster. A function or method with 100 lines of code would be scary to anyone. Hence, the smaller a piece of code, the faster you can understand it.

The team size can be one of the key factors for a successful software project. As team size grows, human communication complexities become a burden to the project. Adding more programmers to a project can make a project even worse than it was. In my experience, I have always experienced that smaller teams produce faster. I think it isn’t just communication. As the team gets bigger, you also need to deal with politics. In a team of three people, you can’t really have politics. Consequently, the smaller the team size is, the faster you develop.

In contemporary software development, microservices architecture has been widely adopted. Previously, we have been developing monolith applications where a single code base that contains all of an application’s logic. As the project grows bigger, it’s almost impossible to reason about the code base or manage each piece of software. So, decomposing project into smaller parts became the natural solution. Therefore, the smaller service you have, the faster it is to develop, test and deploy it.

All in all, “smaller is faster” is applicable to various aspects of software development. The smaller the software is, the faster it becomes to understand, develop, deploy, and optimize it.