CPINFO20 Web Development

1 Java vs JavaScript

	java	javascript
Execution	bytecode executed in JVM	script interpreted
Compilation	source code compiled to byte code (type checking)	sometimes transformed and/or compressed (minimized), but compilation not necessary
Types	statically typed	dynamically typed
Numeric types	byte, short, int, long, float, double	Number, BigInt
Portability	“write once, run anywhere” (still depends on JVM version and implementation)	Was horribly fragmented, today much better thanks to ECMA
OOP	Everything is OO	possible with prototypes
Inheritance	OO Class	prototypal
Concurrency	Thread model	Async callbacks with event loop
First class functions	“kind of” since java 8	Yes
Function overloading	Yes	“kind of” with … rest operator in args
XML + JSON	XML is a language. XML parsing supported out of the box. JSON OK	JSON is a format (not a language). JSON is a subset of JS (is JS). XML possible via npm.
Exception Handling (try, catch, finally)	Yes	Yes
Deployment / Scaling	typically big server, scaled horizontally (add CPU+mem, i.e. add floors to building)	typically containerized, scaled vertically (add instances, i.e. add building to neighborhood)

2 What is node.js

• History
  • 1995: Brendan Eicht 👉 Netscape
  • “LiveScript” -> “JavaScript” (marketing)
  • 1997: ECMAScript standard 1
    • es3, es4, es5, es2016, …, es2019
  • JavaScript is ECMAScript
  • Google Chrome v8
  • 2009: Ryan Dahl 👉 node.js
  • Containerization
  • 2013: Docker
  • 2014/2015: Kubernetes
  • Serverless / FaaS
  • 2010: PiCloud (python)
  • 2014: Amazon AWS
  • 2016: MS Azure Functions (GA)
  • 2018: Google Cloud Functions (GA)
• Under Attack! 💣 🔫 ⚔️ Competition
  • Microsoft: JScript, silverlight, asp.net, typescript
  • Adobe: flash
• Node.js is a JavaScript engine (Google’s v8) with a collection of modules that allow file system access, networking, and other functionality required to create applications.
• No relation to Java (marketing), but certainly inspired by Java
• Managed by Linux Foundation (OpenJS Foundation)
• Allows same language on client and server

3 Types in JavaScript

📢 JavaScript is dynamically typed 🚨

Theoretic types:

• Number
  • IEEE 754 Floating Point
  • ⚠️ Careful with rounding, i.e. money!!!

    ((0.1 + 0.2) + 0.3) === (0.1 + (0.2 + 0.3)) // false
    

• String
  • use “backquotes” ‘`’ for string interpolation:

    const name = 'cody';
    const location = 'annecy';
    const message = `${name} is in `${location}`;
    

• Boolean
• Object object literal:

  const user = {
      name: 'cody',
      address: {
          city: 'annecy'
          code: 74000
      },
  };
  

• Array
• null & undefined
  • both represent “no value”
  • prefer undefined because it is built into the langauge

Inheritance:

• JavaScript uses “prototypal inheritance” (subtle difference from class based inheritance)
• The class keyword makes JavaScript appear like Java inheritance (even though it uses prototypes)

4 What is npm

• npm registry is a public registry of node.js packages
• npm cli tool is a tool to install/manage packages from the registry
• yarn is an alternative cli tool

5 SQL vs NoSQL 🤜💥🤛

	SQL	NoSQL
storage	tables	collections
object encapsulation	rows 👉 object may be split into multiple rows and across multiple tables (requiring ORM) 👉 normalized to remove duplication 👉 a single row may or may not have meaning to the application 👉 query with joins 👉 ORM to help read / write objects across tables	documents 👉 "semi-structured data" 👉 an entire object is typically encoded into a single document 👉 de-normalized to improve query perf & API requests (makes updates and deletions complex 👎😿)
identification	primary key 👉1 row may not be an entire object	document id 👉 key value pair 👉 entire document can be fetched with id
organization	table names	tree 👉 similar to filesystem (directories and files)
encoding	table schema 👉 column types	encoded into standard format 👉 JSON, XML, etc.
indexes	yes ✔️	yes ✔️
queries	SQL query language 👉 powerful and complicated	Query API or language 👉 generally simpler but limited
schema	defined at table creation before adding data 👉 can be modified, but every row must match schema	no schema 👉 documents still need common structure to facilitate queries, but this is not enforced by the DB
schema migrations	data must always be valid 👉 add new columns then complete existing rows 👉 often done with service "offline"	data is never validated 👉 modify all existing data to new structure (may require going "offline") 👉 migrate data at application time when documents are accessed (requires code that supports multiple versions of structure)
scalability	vertical 👉 add memory, CPU, SSD, etc. 👉 adding floors to a building	horizontal 👉 sharding (replication over multiple servers/sites) 👉 adding buildings to a neighborhood 👉 ideal for large DBs and geographic distribution
examples	PostgreSQL, MySQL, Oracle, Microsoft SQL server	MongoDB, Redis, Cassandra, CouchDB, Firestore