Down With JavaScript!

training@instil.co
September 2018
© Instil Software 2018
Down With JavaScript!

Where Things Were Better…
Did you know that the original
4GL’s were designed to be an
ideal coding environment for
GUI’s? Where no one would
suffer, and everyone could
leave the office at 5pm?

Imagine a Developer Was Frozen in the 1990’s

The Myth Of The Full Stack Developer…

React 101
<Ticker name=“Fred”/>

class Ticker extends React.Component {
constructor(props) {
super(props);
this.state = {secondsElapsed: 0};
}
componentDidMount() {
this.interval = setInterval(() => this.tick(), 1000);
}
componentWillUnmount() {
clearInterval(this.interval);
}
tick() {
this.setState({secondsElapsed: this.state.secondsElapsed + 1});
}
render() {
return (
<div>
<h1>Hello, {this.props.name}!</h1>
<p>Page open {this.state.secondsElapsed} seconds.</p>
</div>
);
}
}

class Ticker extends React.Component {
constructor(props) {
super(props);
this.state = {secondsElapsed: 0};
}
componentDidMount() {
this.interval = setInterval(() => this.tick(), 1000);
}
componentWillUnmount() {
clearInterval(this.interval);
}

tick() {
this.setState({secondsElapsed: this.state.secondsElapsed + 1});
}
render() {
return (
<div>
<h1>Hello, {this.props.name}!</h1>
<p>Page open {this.state.secondsElapsed} seconds.</p>
</div>
);
}
}
Initial State
Lifecycle
Methods
Changing state marks component as dirty
Rendering uses embedded JSX templating language

We Still Need Services and a Data Source
Having defined the problem, the first step
towards a solution is the acquisition of data.

•  Using Kotlin.js
Demo No1

➔ Scala code compiled to JS
➔ Full Scala support (apart from reflection)
➔ Fast dev feedback loop
➔ Compact production JS code generated
➔ Performance within 1-3x as optimised JS
➔ Type safety across multiple platforms
What is Scala JS

Scala for JS
Devs
Ref: scala-js.org

JS Comparison: Variables
// mutable variable
let x = 5;
// immutable variable
const y = "Constant";
// mutable variable
var x = 5
// immutable variable
val y = "Constant"

JS Comparison: Functions
function mult(x, y) {
return x * y;
}
def mult(x: Double,
y: Double): Double = x * y

JS Comparison: Anonymous functions
const f = (x, y) => x + y;
const p = ["Fox", "jumped",
"over", "me"];
const l = p.map(s => s.length)
.reduce((a, b) => a + b, 0);
// == 15
val f = (x: Double,
y: Double) => x + y
val p = Array("Fox", "jumped",
"over", "me")
val l = p.map(s => s.length)
.foldLeft(0)((a,b) => a + b)
// == 15

JS Comparison: Higher order functions
function minmaxBy(arr, f) {
return arr.reduce(
([min, max], e) => {
const v = f(e);
return [Math.min(min, v),
Math.max(max, v)]
},
[Number.MAX_VALUE,Number.MIN_VALUE]
)
}
const [youngest, oldest] =
minmaxBy(persons, e => e.age);
def minmaxBy[T](seq: Seq[T],
f: T => Int): (Int, Int) = {
seq.foldLeft(( Int.MaxValue,
Int.MinValue)) {
case ((min, max), e) =>
val v = f(e)
(math.min(min, v),
math.max(max, v))
}
}
val (youngest, oldest) =
minmaxBy[Person](persons, _.age)

JS Comparison: Futures
function onLoadPromise(img) {
if (img.complete) {
return Promise.resolve(img.src);
} else {
const p = new Promise((success)=> {
img.onload = (e) => {
success(img.src);
};
});
return p;
}
}
def onLoadFuture(img:
HTMLImageElement) = {
if (img.complete) {
Future.successful(img.src)
} else {
val p = Promise[String]()
img.onload = { (e: Event) =>
p.success(img.src)
}
p.future
}
}

Other features
● Traits (Mixins)
● Singleton Objects
● Immutable Collections
● Case Classes
● Tuples
● Pattern Matching / destructuring
● Default parameter values
● Lazy Initialization
● Reified Generics
● TypeClass Support
● Higher Kinded Types Support
● Macros
…….

JavaScript Interop : Facades
@js.native
trait Window extends js.Object {
val document: HTMLDocument = js.native
var location: String = js.native
def innerWidth: Int = js.native
def innerHeight: Int = js.native
…
} Typescript definitions to Scala.js Binding transformer available
https://github.com/sjrd/scala-js-ts-importer

JavaScript Interop : Exposing Scala.js
package example
import scala.scalajs.js.annotation._
@JSExportTopLevel("HelloWorld")
object HelloWorld {
@JSExport
def sayHello(): Unit = {
println("Hello world!")
}
}
HelloWorld.sayHello();
Also @JSExportTopLevel("..")
to export a class

Slinky
➔ Scala JS wrapper for React
➔ Stays close as possible to React
conventions
➔ Excellent integration with existing react
components

React Stateless Component expanded

Slinky: getting started
● Dependent on Java 8, SBT, NPM

Diode
● Immutable application state
● Controlled data flow
● Similar to react
● React components integration

Slinky React Native
● Dependent on SBT and Java 8, NPM, xcode

training@instil.co
September 2018
F# and Fable
Functional Client Side Development
Eamonn.Boyle@instil.co
@BoyleEamonn

F# and Fable
Functional Client Side Development

.NET strongly & statically typed multi-paradigm language
• Imperative
• Object Oriented
• Functional First
Open source, cross platform
• Mono, .NET Framework & .NET Core
Stable – currently on version 4.5 (August 2018)
• Version 1.0 released in 2005
• Passionate Community
A lot of nice C# features originated in F#
F#

Originally a .NET implementation of OCaml so the syntax is similar
Heavy use of type inference
• Expressions, function parameters and return types etc all inferred from usage
• Can reduce the amount of typing (pun intended)
• Can lead to very strange error messages
Focuses on pushing runtime errors to compile time
• Very strict type system
Easy interop with .NET (C#) libraries
F# Syntax

Type bindings and helpers for easy interop with JavaScript
• Including Type Safe helpers for React
Supports most of the F# core library
There’s a tool to convert TypeScript bindings to Fable bindings
• https://github.com/fable-compiler/ts2fable
Fable
Fable is an F# to JavaScript compiler powered by
Babel, designed to produce readable and standard
code.

Elm-like abstractions for F#
Facilitates creating Model-View-Update applications
• Similar to using Redux with something like React
• Does not include the View part, instead you use the Fable React helpers
The main building blocks are
• Model
• Messages
• Commands
• Initialisation
• Update
• Views
• Program
Elmish

The Bad Stuff

“This was the hardest development
experience, period!!”

Type inference can be very confusing
• I enjoy type inference in C#, Kotlin etc
• This is at another level
• Inference in one part of the program can be dictated by usage far away
• This can lead to confusing error messages
• Explicit types can help here
• But then the syntax can be longer than other languages
File order matters – this caught me out a few times
Tooling is not perfect
• I found myself moving between VS Code and JetBrains Rider
Runtime errors can still occur and can be difficult to debug
• Some of the error messages are minimal
The Bad Stuff (very subjective)

I found the React bindings a pain compared to using HTML
• It’s better because it’s type safe but…
• Tooling around HTML + CSS is pretty mature while this isn’t
• More difficult if working with other tools, designers and existing skills
• I find standard JSX with React much easier
White space significance is a pain
I found it difficult to structure the app
• Some blocks of code were getting very large
• I would need to build much larger apps to see how this scales
The Bad Stuff (very subjective)

The Good Stuff

The language is pretty cool - succinct functional syntax
• Pipes
• Function Composition
• Type Definition and Domain Modeling
• Operator Overloading
• etc
Strict Type System
• Catches a lot of errors
• Exhaustive pattern matching
• Units of measure
• Etc
Easy interop and drop down to OO or imperative style if required
The Good Stuff

This ability to use OO & imperative is essential when targeting JavaScript platform
• Good bindings to access dynamic types
Tooling is very good
• Very quick to create a project from templates
• Visual Studio Code + Ionide is pretty good and is completely free
• HTML to Elmish converter - https://mangelmaxime.github.io/html-to-elmish/
• Hot Module Reloading while retaining state
Elmish MVU pattern is very good and is well implemented
• Very succinct to build up types, state and views
The Good Stuff

Conclusion

It is very impressive what these projects have accomplished
Some aspects of the development are succinct, safe and neat
If you are a functional fan & especially if doing F#, then this is pretty awesome
My Conclusion
But…

For my money, TypeScript is a better solution
Better balance (compromise) on safety, succinctness & productivity
• Gives me type safety (even if less expressive)
• Much easier interop with JavaScript (it’s a superset)
• Easier consumption of existing JavaScript libraries
• Better tooling
I would stick with standard HTML/JSX for my view definitions
• Not because it is necessarily better for defining views
• Simply because the tooling, documentation and onboarding is easier
Angular has excellent TypeScript APIs
My Conclusion

Down With JavaScript!

More Related Content

What's hot (16)

Similar to Down With JavaScript! (20)

More from Garth Gilmour (20)

Recently uploaded (20)

Down With JavaScript!