What can and can't be polyfilled in vanilla JS?

In response to my article on Promises in vanilla JS the other day, one of my students told me that if they knew it could be polyfilled, they would have started using Promises a lot sooner.

This led to an interesting question: what can and can’t be polyfilled in vanilla JS.

Here’s the gist:

1. Any native method, property, or browser API can be polyfilled.
2. Operators and expressions cannot.

Why can one group of items be polyfilled while the other cannot?

For methods, properties, and browser APIs, the JavaScript language provides a mechanism for extending features and functionality. Everything is an object in JavaScript, so everything can be given properties and have functions added to it’s prototype.

The Array.prototype.map() method can be polyfilled by adding a new map() method to the Array.prototype object. The fetch() method can be polyfilled by creating a new global fetch() function that uses XHR under the hood.

Operators and expressions—things like the spread operator ([...someVar]), template literals, and const and let—cannot be polyfilled, because they don’t get attached to any sort of object.

They’re underpinnings of the language, baked into a browser’s JS parsing engine. In older browsers, using let to define a variable has no meaning, and things break. Unfortunately, there’s no way around that since we can’t replace the rendering engine itself in the old browser.

Hope that clears things up!