Reactive Programming For Girls

For those of you who are new to programming, you usually say, username = “alice”; followed by; refreshUserInterface().

The program will then use the name alice, in all the right places.

But as you add more to your program, manually refreshing the user interface, especially under some component architecture gets difficult, becomes unclear.

In reactive programming, you first subscribe to username, think of username as a radio station.

And then in your subscription, which is just a function, that is called, when username changes.

You put in the code that updates the UI, and your username is now bound to the User Interface.

You can just stop worrying about refreshing the UI, because you know any time you change the reactive variables

The UI will update, automatically.

And you do just do one UI binding at a time, if you want to update few things, in a singe change.

You just subscribe to the thing multiple times, a subscription is the end of the UI story, you set it and forget it.

To assign alice to the username, you now say, username .value = “alice”;

So username, is no longer just a primitive variable, it is an object with a .value getter and setter.

And then a .subscribe method, that accepts your function, and keeps track of all its subscribers.

When you first subscribe, and there is a value in there already, for example bob, your subscriber will instantly get called with the string bob.

But if the value is null, or undefined, nothing will happen, your UI component will show a default state.

Nothing also happens, if you change the value to what it already was, the subscribers will not be triggered.

That is reactive programming, everything else is up to you, you may move to functional reactive programming.

Or create something that acts as a dependency tree for nested signals, or just rely on your own application architecture.

Or just use reactive variables alone, with the Document Object Model, because it is just .value and .subscribe everything is clear.

Finally, I recommend you go reactive as soon as you can, it makes all of your programs simple, and self updatable.

It may not be possible to to take a non-reactive program, convert it to use reactivity, and benefit from that reactive simplicity.

I also worry that non-reactive one person projects, needlessly suffer, from necessary complexity, that may prevent shipping.

So reactive programming, is perfect for learning,

Because it adapts your mind to seeing a program for what it is: an interactive application.

And it is perfect for one person project and side projects, that it would be fun to complete.

Make your own reactive variable object, write from scratch or extent EventEmitter.

Don’t make it complicated, don’t add an addDependency method, or nested reactive variable detection via context proxies, or something.

Reactive programming is a big deal, and it is very friendly, it is just one little fancy thing.