Video version of this article: https://youtu.be/4Oo04kyM1Yg.
In this post I want to talk about why semantic naming is important. It's important in programming in general, but particularly I want to discuss its importance in CSS.
Basically, if you have a class
.color-red, or even
.color-primary, that is not semantically named.
It describes the style or look of the element, rather than its purpose.
A more semantic name may be
main-sidebar. Something that describes the purpose of the element.
It brings two important benefits:
- Labels / notes for the programmer
So with your
.color-red class, obviously the
h1 you apply it to is going to be red.
But the designer comes back tomorrow and tells you to make it blue.
Now you have to go to both your HTML and CSS to make the change.
In the HTML you need to change the class, because it's going to be the ultimate confusion if it has the class
.color-red and it looks blue. You also need to add a new class in your CSS as well:
.color-blue this time.
It works, but it's on cue to become a maintenance nightmare. Code should be optimised for maintainability.
Having to change two things (HTML and CSS) instead of one is more error-prone.
If it was just named
.page-heading, then you only have to change the color in one place (in the CSS).
This has tons of benefits:
- Changes are only done in one place, so the process is simpler (read: harder to make a mistake).
- It opens up the path for white-labelling (having a website where you can change the styling purely from the CSS).
- Maintenance is much easier. The element has the class
.page-headingtoday, and it will probably have the same class a year from now.
Back to the HTML with the class
<div class="color-red"> doesn't tell you anything about what that component is.
Is it the sidebar? The page heading? The comment box? It's extremely difficult to tell.
Compare this with
<div class="page-heading"> or
You immediately know the purpose of the component. You can find it on the website (even with a simple search on the dev tools), and you can find it in the codebase.
And remember, you don't have 5 components, you have 100 components for a large website.
I'm sure you've heard the phrase "There are two difficult things in computer science: Naming and cache invalidation".
Yes naming is difficult, and it's not uncommon for programmers to spend a good amount of time thinking of a good name for something.
The reason this is necessary is because code must be optimised for maintainability. You will read code and modify it a huge number of times. Even if it takes 3x to write the code upfront, it will pay dividends in efficiency if the code is easily understandable. And it will not take 3x longer to write it, only 1.2x.
Use semantic naming that describes the purpose of an element, rather than its look.
It provides lots of benefits:
- Changes only happen in one place, so less error-prone.
- Changes only happen in CSS (+ opens up white labelling).
- Labels for easily identifying elements.
And remember the project tomorrow will have 100s of components, so any problems you have today due to non-semantic names will greatly amplified tomorrow.