Here’s a hot take from me.

At the higher level, I will put on 2 sides: “Static” and “Dynamic”, as in Render the whole page again versus fetch and re-render HTML elements on the fly, respectively. At the developer’s level, we can think and categorize instead into websites that do not use APIs as oppose of those which do use APIs on the client-side.

The exception is a static website with don’t have moving data i.e. blogs. They can be just .html files, so APIs doesn’t matter. More suitable example is a web application, i.e. Facebook.

The no APIs approach#

TL;DR: HTML is heavy. You can’t make a mobile app.

First of all, you are making a Facebook with only HTMLs. Think, John.

  • Data is in the form of HTML. Hence, you can’t create a mobile app. Please don’t make apps with a webpage embedded in, for god’s sake.
  • In the case of Facebook, you can’t see new posts without downloading a whole new HTML. Your server only outputs HTMLs.
  • I challenge you to get the comments for a post without refreshing Facebook, without APIs.

That means, if you are making an application, you can forget what you learn about web at school.

My professor wanted me to use “AJAX” to make changes in a tabbed Edit page so that it doesn’t refresh the page without teaching APIs. Can you guess why?

AJAX without API

The APIs approach#

TL;DR: Hot-swap data. Can make mobile app. Can be used as 3rd party service.

Data is in the form of JSON/XML (usually). Now you can delete HTMLs and render new ones from JSON with JavaScript. Better yet, use Front-end frameworks. Even better, you’ll be able to send out immediately the static parts of the page and later on fetch and render other parts (because you request data on-demand via APIs).

Your Facebook is more lively than ever, all thanks to React. Now imagine Java’s JSP. And C#’s Razor Pages.

Now that you have APIs, you can create a mobile app. Lewlew?