API CREATION :: Creating RESTful APIs with NodeJS and MongoDB Tutorial (Part 2)

RESTful API using Node.js (Express.js) and MongoDB (mongoose)! We are going to learn how to install and use each component individually and then proceed to create a RESTful API.

MEAN Stack tutorial series:

1. AngularJS tutorial for beginners (Part I)
2. Creating RESTful APIs with NodeJS and MongoDB Tutorial (Part II) 👈 you are here
3. MEAN Stack Tutorial: MongoDB, ExpressJS, AngularJS and NodeJS (Part III)

What is a RESTful API?

REST stands for Representational State Transfer. It is an architecture that allows client-server communication through a uniform interface. REST is stateless, cachable and has property called idempotence. It means that the side effect of identical requests have the same side-effect as a single request.

HTTP RESTful API’s are compose of:

• HTTP methods, e.g. GET, PUT, DELETE, PATCH, POST, …
• Base URI, e.g. http://adrianmejia.com
• URL path, e.g. /blog/2014/10/01/creating-a-restful-api-tutorial-with-nodejs-and-mongodb/
• Media type, e.g. html, JSON, XML, Microformats, Atom, Images…

Here is a summary what we want to implement:

Resource (URI)	POST (create)	GET (read)	PUT (update)	DELETE (destroy)
/todos	create new task	list tasks	N/A (update all)	N/A (destroy all)
/todos/1	error	show task ID 1	update task ID 1	destroy task ID 1

NOTE for this tutorial:

• Format will be JSON.
• Bulk updates and bulk destroys are not safe, so we will not be implementing those.
• CRUD functionality: POST == CREATE, GET == READ, PUT == UPDATE, DELETE == DELETE.

Installing the MEAN Stack Backend

In this section, we are going to install the backend components of the MEAN stack: MongoDB, NodeJS and ExpressJS. If you already are familiar with them, then jump to wiring the stack. Otherwise, enjoy the ride!

Installing MongoDB

MongoDB is a document-oriented NoSQL database (Big Data ready). It stores data in JSON-like format and allows users to perform SQL-like queries against it.

You can install MongoDB following the instructions here.

If you have a Mac and brew it’s just:

1	brew install mongodb && mongod

In Ubuntu:

1	sudo apt-get -y install mongodb

After you have them installed, check version as follows:

1 2 3 4 5 6 7 8 9

# Mac mongod --version # => db version v2.6.4 # => 2014-10-01T19:07:26.649-0400 git version: nogitversion # Ubuntu mongod --version # => db version v2.0.4, pdfile version 4.5 # => Wed Oct 1 23:06:54 git version: nogitversion

Installing NodeJS

The Node official definition is:

Node.js® is a JavaScript runtime built on Chrome’s V8 JavaScript engine. Node.js uses an event-driven, non-blocking I/O model that makes it lightweight and efficient. Node.js’ package ecosystem, npm, is the largest ecosystem of open source libraries in the world.

Node.js Websitenodejs.org

In short, NodeJS allows you to run Javascript outside the browser, in this case, on the web server. NPM allows you to install/publish node packages with ease.

To install it, you can go to the NodeJS Website.

Since Node versions changes very often. You can use the NVM (Node Version Manager) on Ubuntu and Mac with:

1 2 3 4 5 6 7 8 9

# download NPM curl -o- https://raw.githubusercontent.com/creationix/nvm/v0.31.4/install.sh | bash # load NPM export NVM_DIR="$HOME/.nvm" [ -s "$NVM_DIR/nvm.sh" ] && . "$NVM_DIR/nvm.sh" # This loads nvm # Install latest stable version nvm install stable

Check out https://github.com/creationix/nvm for more details.

Also, on Mac and brew you can do:

1	brew install nodejs

After you got it installed, check node version and npm (node package manager) version:

1 2 3 4 5

node -v # => v6.2.2 npm -v # => 3.9.5

Installing ExpressJS

ExpressJS is a web application framework that runs on NodeJS. It allows you to build web applications and API endpoints. (more details on this later).

We are going to create a project folder first, and then add express as a dependency. Let’s use NPM init command to get us started.

1 2 3 4 5 6 7 8 9

# create project folder mkdir todo-app # move to the folder and initialize the project cd todo-app npm init . # press enter multiple times to accept all defaults # install express v4.14 and save it as dependency npm install express@4.14 --save

Notice that after the last command, express should be added to package.json with the version 4.14.x.

Using MongoDB with Mongoose

Mongoose is an NPM package that allows you to interact with MongoDB. You can install it as follows:

1	npm install mongoose@4.5.8 --save

If you followed the previous steps, you should have all you need to complete this tutorial. We are going to build an API that allow users to CRUD (Create-Read-Update-Delete) Todo tasks from database.

Mongoose CRUD

CRUD == Create-Read-Update-Delete

We are going to create, read, update and delete data from MongoDB using Mongoose/Node. First, you need to have mongodb up and running:

1 2	# run mongo daemon mongod

Keep mongo running in a terminal window and while in the folder todoApp type node to enter the node CLI. Then:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

// Load mongoose package var mongoose = require('mongoose'); // Connect to MongoDB and create/use database called todoAppTest mongoose.connect('mongodb://localhost/todoAppTest'); // Create a schema var TodoSchema = new mongoose.Schema({ name: String, completed: Boolean, note: String, updated_at: { type: Date, default: Date.now }, }); // Create a model based on the schema var Todo = mongoose.model('Todo', TodoSchema);

Great! Now, let’s test that we can save and edit data.

Mongoose Create

1 2 3 4 5 6 7 8 9 10 11

// Create a todo in memory var todo = new Todo({name: 'Master NodeJS', completed: false, note: 'Getting there...'}); // Save it to database todo.save(function(err){ if(err) console.log(err); else console.log(todo); });

If you take a look to Mongo you will notice that we just created an entry. You can easily visualize data using Robomongo:

You can also build the object and save it in one step using create:

1 2 3 4

Todo.create({name: 'Create something with Mongoose', completed: true, note: 'this is one'}, function(err, todo){ if(err) console.log(err); else console.log(todo); });

Mongoose Read and Query

So far we have been able to save data, now we are going explore how to query the information. There are multiple options for reading/querying data:

• Model.find(conditions, [fields], [options], [callback])
• Model.findById(id, [fields], [options], [callback])
• Model.findOne(conditions, [fields], [options], [callback])

Some examples:

Find all

1 2 3 4 5

// Find all data in the Todo collection Todo.find(function (err, todos) { if (err) return console.error(err); console.log(todos) });

The result is something like this:

results

1 2 3 4 5 6 7 8 9 10 11 12

[ { _id: 57a6116427f107adef36c2f2, name: 'Master NodeJS', completed: false, note: 'Getting there...', __v: 0, updated_at: 2016-08-06T16:33:40.606Z }, { _id: 57a6142127f107adef36c2f3, name: 'Create something with Mongoose', completed: true, note: 'this is one', __v: 0, updated_at: 2016-08-06T16:45:21.143Z } ]

You can also add queries

Find with queries

1 2 3 4 5 6 7 8 9 10 11

// callback function to avoid duplicating it all over var callback = function (err, data) { if (err) { return console.error(err); } else { console.log(data); } } // Get ONLY completed tasks Todo.find({completed: true }, callback); // Get all tasks ending with `JS` Todo.find({name: /JS$/ }, callback);

You can chain multiple queries, e.g.:

Chaining queries

1 2 3 4 5 6 7 8

var oneYearAgo = new Date(); oneYearAgo.setYear(oneYearAgo.getFullYear() - 1); // Get all tasks staring with `Master`, completed Todo.find({name: /^Master/, completed: true }, callback); // Get all tasks staring with `Master`, not completed and created from year ago to now... Todo.find({name: /^Master/, completed: false }).where('updated_at').gt(oneYearAgo).exec(callback);

MongoDB query language is very powerful. We can combine regular expressions, date comparison and more!

Mongoose Update

Moving on, we are now going to explore how to update data.

Each model has an update method which accepts multiple updates (for batch updates, because it doesn’t return an array with data).

• Model.update(conditions, update, [options], [callback])
• Model.findByIdAndUpdate(id, [update], [options], [callback])
• Model.findOneAndUpdate([conditions], [update], [options], [callback])

Alternatively, the method findOneAndUpdate could be used to update just one and return an object.

Todo.update and Todo.findOneAndUpdate

1 2 3 4 5 6 7 8

// Model.update(conditions, update, [options], [callback]) // update `multi`ple tasks from complete false to true Todo.update({ name: /master/i }, { completed: true }, { multi: true }, callback); //Model.findOneAndUpdate([conditions], [update], [options], [callback]) Todo.findOneAndUpdate({name: /JS$/ }, {completed: false}, callback);

As you might noticed the batch updates (multi: true) doesn’t show the data, rather shows the number of fields that were modified.

1	{ ok: 1, nModified: 1, n: 1 }

Here is what they mean:

• n means the number of records that matches the query
• nModified represents the number of documents that were modified with update query.

Mongoose Delete

update and remove mongoose API are identical, the only difference it is that no elements are returned. Try it on your own ;)

• Model.remove(conditions, [callback])
• Model.findByIdAndRemove(id, [options], [callback])
• Model.findOneAndRemove(conditions, [options], [callback])

ExpressJS and Middlewares

ExpressJS is a complete web framework solution. It has HTML template solutions (jade, ejs, handlebars, hogan.js) and CSS precompilers (less, stylus, compass). Through middlewares layers, it handles: cookies, sessions, caching, CSRF, compression and many more.

Middlewares are pluggable processors that runs on each request made to the server. You can have any number of middlewares that will process the request one by one in a serial fashion. Some middlewares might alter the request input. Others, might create log outputs, add data and pass it to the next() middleware in the chain.

We can use the middlewares using app.use. That will apply for all request. If you want to be more specific, you can use app.verb. For instance: app.get, app.delete, app.post, app.update, …

Let’s give some examples of middlewares to drive the point home.

Say you want to log the IP of the client on each request:

Log the client IP on every request

1 2 3 4 5

app.use(function (req, res, next) { var ip = req.headers['x-forwarded-for'] || req.connection.remoteAddress; console.log('Client IP:', ip); next(); });

Notice that each middleware has 3 parameters:

• req: contain all the requests objects like URLs, path, …
• res: is the response object where we can send the reply back to the client.
• next: continue with the next middleware in the chain.

You can also specify a path that you want the middleware to activate on.

Middleware mounted on "/todos/:id" and log the request method

1 2 3 4

app.use('/todos/:id', function (req, res, next) { console.log('Request Type:', req.method); next(); });

And finally you can use app.get to catch GET requests with matching routes, reply the request with a response.send and end the middleware chain. Let’s use what we learned on mongoose read to reply with the user’s data that matches the id.

Middleware mounted on "/todos/:id" and returns

1 2 3 4 5 6

app.get('/todos/:id', function (req, res, next) { Todo.findById(req.params.id, function(err, todo){ if(err) res.send(err); res.json(todo); }); });

Notice that all previous middlewares called next() except this last one, because it sends a response (in JSON) to the client with the requested todo data.

Hopefully, you don’t have to develop a bunch of middlewares besides routes, since ExpressJS has a bunch of middlewares available.

Default Express 4.0 middlewares

• morgan: logger
• body-parser: parse the body so you can access parameters in requests in req.body. e.g. req.body.name.
• cookie-parser: parse the cookies so you can access parameters in cookies req.cookies. e.g. req.cookies.name.
• serve-favicon: exactly that, serve favicon from route /favicon.ico. Should be call on the top before any other routing/middleware takes place to avoids unnecessary parsing.

Other ExpressJS Middlewares

The following middlewares are not added by default, but it’s nice to know they exist at least:

• compression: compress all request. e.g. app.use(compression())
• session: create sessions. e.g. app.use(session({secret: 'Secr3t'}))
• method-override: app.use(methodOverride('_method')) Override methods to the one specified on the _method param. e.g. GET /resource/1?_method=DELETE will become DELETE /resource/1.
• response-time: app.use(responseTime()) adds X-Response-Time header to responses.
• errorhandler: Aid development, by sending full error stack traces to the client when an error occurs. app.use(errorhandler()). It is good practice to surround it with an if statement to check process.env.NODE_ENV === 'development'.
• vhost: Allows you to use different stack of middlewares depending on the request hostname. e.g. app.use(vhost('*.user.local', userapp)) and app.use(vhost('assets-*.example.com', staticapp)) where userapp and staticapp are different express instances with different middlewares.
• csurf: Adds a Cross-site request forgery (CSRF) protection by adding a token to responds either via session or cookie-parser middleware. app.use(csrf());
• timeout: halt execution if it takes more that a given time. e.g. app.use(timeout('5s'));. However you need to check by yourself under every request with a middleware that checks if (!req.timedout) next();.

Wiring up the MEAN Stack

In the next sections, we are going to put together everything that we learn from and build an API. They can be consume by browsers, mobile apps and even other servers.

Bootstrapping ExpressJS

After a detour in the land of Node, MongoDB, Mongoose, and middlewares, we are back to our express todoApp. This time to create the routes and finalize our RESTful API.

Express has a separate package called express-generator, which can help us to get started with out API.

Install and run "express-generator"

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

# install it globally using -g npm install express-generator -g # create todo-app API with EJS views (instead the default Jade) express todo-api -e # create : todo-api # create : todo-api/package.json # create : todo-api/app.js # create : todo-api/public # create : todo-api/public/javascripts # create : todo-api/routes # create : todo-api/routes/index.js # create : todo-api/routes/users.js # create : todo-api/public/stylesheets # create : todo-api/public/stylesheets/style.css # create : todo-api/views # create : todo-api/views/index.ejs # create : todo-api/views/layout.ejs # create : todo-api/views/error.ejs # create : todo-api/public/images # create : todo-api/bin # create : todo-api/bin/www # # install dependencies: # $ cd todo-api && npm install # # run the app on Linux/Mac: # $ DEBUG=todo-app:* npm start # # run the app on Windows: # $ SET DEBUG=todo-api:* & npm start

This will create a new folder called todo-api. Let’s go ahead and install the dependencies and run the app:

1 2 3 4 5 6 7 8

# install dependencies cd todo-api && npm install # run the app on Linux/Mac PORT=4000 npm start # run the app on Windows SET PORT=4000 & npm start

Use your browser to go to http://0.0.0.0:4000, and you should see a message “Welcome to Express”

Connect ExpressJS to MongoDB

In this section we are going to access MongoDB using our newly created express app. Hopefully, you have installed MongoDB in the setup section, and you can start it by typing (if you haven’t yet):

1

mongod

Install the MongoDB driver for NodeJS called mongoose:

1	npm install mongoose --save

Notice --save. It will add it to the todo-api/package.json

Next, you need to require mongoose in the todo-api/app.js

Add to app.js

1 2 3 4 5 6 7 8 9 10

// load mongoose package var mongoose = require('mongoose'); // Use native Node promises mongoose.Promise = global.Promise; // connect to MongoDB mongoose.connect('mongodb://localhost/todo-api') .then(() => console.log('connection succesful')) .catch((err) => console.error(err));

Now, When you run npm start or ./bin/www, you will notice the message connection successful. Great!

You can find the repository here and the diff code at this point: diff

Creating the Todo model with Mongoose

It’s show time! All the above was setup and preparation for this moment. Let bring the API to life.

Create a models directory and a Todo.js model:

1 2	mkdir models touch models/Todo.js

In the models/Todo.js:

1 2 3 4 5 6 7 8 9 10

var mongoose = require('mongoose'); var TodoSchema = new mongoose.Schema({ name: String, completed: Boolean, note: String, updated_at: { type: Date, default: Date.now }, }); module.exports = mongoose.model('Todo', TodoSchema);

diff

What’s going on up there? Isn’t MongoDB suppose to be schemaless? Well, it is schemaless and very flexible indeed. However, very often we want bring sanity to our API/WebApp through validations and enforcing a schema to keep a consistent structure. Mongoose does that for us, which is nice.

You can use the following types:

• String
• Boolean
• Date
• Array
• Number
• ObjectId
• Mixed
• Buffer

API clients (Browser, Postman and curl)

I know you have not created any route yet. However, in the next sections you will. These are just three ways to retrieve, change and delete data from your future API.

Curl

Create tasks

1 2 3 4 5

# Create task curl -XPOST http://localhost:3000/todos -d 'name=Master%20Routes&completed=false&note=soon...' # List tasks curl -XGET http://localhost:3000/todos

Browser and Postman

If you open your browser and type localhost:3000/todos you will see all the tasks (when you implement it). However, you cannot do post commands by default. For further testing let’s use a Chrome plugin called Postman. It allows you to use all the HTTP VERBS easily and check x-www-form-urlencoded for adding parameters.

Don’t forget to check x-www-form-urlencoded or it won’t work ;)

Websites and Mobile Apps

Probably these are the main consumers of APIs. You can interact with RESTful APIs using jQuery’s $ajax and its wrappers, BackboneJS’s Collections/models, AngularJS’s $http or $resource, among many other libraries/frameworks and mobile clients.

In the end, we are going to explain how to use AngularJS to interact with this API.

ExpressJS Routes

To sum up we want to achieve the following:

Resource (URI)	POST (create)	GET (read)	PUT (update)	DELETE (destroy)
/todos	create new task	list tasks	error	error
/todos/:id	error	show task :id	update task :id	destroy task ID 1

Let’s setup the routes

Create a new route called `todos.js` in the `routes` folder or rename `users.js`

1	mv routes/users.js routes/todos.js

In app.js add new todos route, or just replace ./routes/users for ./routes/todos

Adding todos routes

1 2	var todos = require('./routes/todos'); app.use('/todos', todos);

All set! Now, let’s go back and edit our routes/todos.js.

List: GET /todos

Remember mongoose query api? Here’s how to use it in this context:

routes/todos.js

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

var express = require('express'); var router = express.Router(); var mongoose = require('mongoose'); var Todo = require('../models/Todo.js'); /* GET /todos listing. */ router.get('/', function(req, res, next) { Todo.find(function (err, todos) { if (err) return next(err); res.json(todos); }); }); module.exports = router;

Harvest time! We don’t have any task in database but at least we verify it is working:

Testing all together

1 2 3 4 5 6 7 8 9

# Start database mongod # Start Webserver (in other terminal tab) npm start # Test API (in other terminal tab) curl localhost:3000/todos # => []%

diff

If it returns an empty array [] you are all set. If you get errors, try going back and making sure you didn’t forget anything, or you can comment at the end of the post for help.

Create: POST /todos

Back in routes/todos.js, we are going to add the ability to create using mongoose create. Can you make it work before looking at the next example?

routes/todos.js (showing just create route)

1 2 3 4 5 6 7 8

/* POST /todos */ router.post('/', function(req, res, next) { Todo.create(req.body, function (err, post) { if (err) return next(err); res.json(post); }); });

diff

A few things:

• We are using the router.post instead of router.get.
• You have to stop and run the server again: npm start.

Everytime you change a file you have to stop and start the web server. Let’s fix that using nodemon to refresh automatically:

Nodemon

1 2 3 4 5

# install nodemon globally npm install nodemon -g # Run web server with nodemon nodemon

Show: GET /todos/:id

This is a snap with Todo.findById and req.params. Notice that params matches the placeholder name we set while defining the route. :id in this case.

routes/todos.js (showing just show route)

1 2 3 4 5 6 7

/* GET /todos/id */ router.get('/:id', function(req, res, next) { Todo.findById(req.params.id, function (err, post) { if (err) return next(err); res.json(post); }); });

diff

Let’s test what we have so far!

Testing the API with Curl

1 2 3 4 5 6 7 8 9 10 11 12 13 14

# Start Web Server on port 4000 (default is 3000) PORT=4000 nodemon # Create a todo using the API curl -XPOST http://localhost:4000/todos -d 'name=Master%20Routes&completed=false&note=soon...' # => {"__v":0,"name":"Master Routes","completed":false,"note":"soon...","_id":"57a655997d2241695585ecf8"}% # Get todo by ID (use the _id from the previous request, in my case "57a655997d2241695585ecf8") curl -XGET http://localhost:4000/todos/57a655997d2241695585ecf8 {"_id":"57a655997d2241695585ecf8","name":"Master Routes","completed":false,"note":"soon...","__v":0}% # Get all elements (notice it is an array) curl -XGET http://localhost:4000/todos [{"_id":"57a655997d2241695585ecf8","name":"Master Routes","completed":false,"note":"soon...","__v":0}]%

Update: PUT /todos/:id

Back in routes/todos.js, we are going to update tasks. This one you can do without looking at the example below, review findByIdAndUpdate and give it a try!

routes/todos.js (showing just update route)

1 2 3 4 5 6 7

/* PUT /todos/:id */ router.put('/:id', function(req, res, next) { Todo.findByIdAndUpdate(req.params.id, req.body, function (err, post) { if (err) return next(err); res.json(post); }); });

diff

curl update

1 2 3

# Use the ID from the todo, in my case 57a655997d2241695585ecf8 curl -XPUT http://localhost:4000/todos/57a655997d2241695585ecf8 -d "note=hola" # => {"_id":"57a655997d2241695585ecf8","name":"Master Routes","completed":true,"note":"hola","__v":0}%

Destroy: DELETE /todos/:id

Finally, the last one! Almost identical to update, use findByIdAndRemove.

routes/todos.js (showing just update route)

1 2 3 4 5 6 7

/* DELETE /todos/:id */ router.delete('/:id', function(req, res, next) { Todo.findByIdAndRemove(req.params.id, req.body, function (err, post) { if (err) return next(err); res.json(post); }); });

diff

Is it working? Cool, you are done then! Is NOT working? take a look at the full repository.

What’s next?

Connecting AngularJS with this endpoint. Check out the third tutorial in this series.

The Software Economy: Why Software Jobs Are Taking Over

We have entered what I like to think of as "The Software Economy." Software is everywhere, and the market for software and computer science careers has exploded.
Look at the products and services around us. New cell phones (Look at the Moto X, HTC, Samsung, ...) are largely differentiated by software and user interface design. The MotoX is getting great reviews because you can turn on the camera by rotating it quickly, for example.
Think about automobiles - they are becoming highly programmable devices filled with smartphone and other software tools. The Chevy Volt, for example, has 10 million lines of code, 2 million more than the code in the F-35 fighter jet. The Tesla, which is an engineering marvel in many respects, has among the most advanced software interfaces in the market (shown below).

Fig 1:  Software-based Control Panel in Tesla
Wearable computing is largely software based: the Nike FuelBand, Jawbone UP, and Fitbit are exciting devices which create your own personal "BigData" for exercise and wellness. And of course Google Glass will enable an even bigger market for wearable computing.

Fig 2:  Wearable Computing Devices
Think about how mobile apps have transformed business service-delivery. Apple AAPL -0.18% tells us that 50 billion apps have been downloaded, making "developing an app" one of the most important projects in nearly every company.
When I entered the technology markets in 1978 the software industry was kind of a niche. Software engineers had to work in big companies who could invest in mainframe computers. The discipline existed, but it was a small branch of math or computer science.

Fig 3:  Software Development Environment in 1970s and 1980s
Since then, of course, we've seen the explosion of PC's, the internet, laptops, mobile devices, and now wearable and embedded computers. All these programmable devices need software.
Let me give you some insights on this fast-growing economy for jobs and careers:

1. The Software Industry is Expanding Faster than Other Occupations
Burning Glass Technologies, one of the leading providers of BigData for employment and job dynamics, recently did some analysis for us and found that the from 2007 to 2012 the number of software jobs grew by 31%, three times faster than US jobs overall. IDC estimates that the market for various software products is over $200 billion, and Comptia, the industry trade association, sizes the total IT industry at $3.6 Trillion, 17% of which is software (which would translate to over $400 billion).

Comptia believes there are around 9.2 million technical and managerial jobs in software, and millions more in peripheral industries. If we look only at core IT positions (ie. those in IT-related companies, not general manufacturing and other service industries), the market for software positions is growing at nearly twice the rate of other jobs in the US economy.

Fig 4:  Core IT positions
The unemployment rate in software and IT positions is particularly low, by the way, around 3.8% in the last BLS survey.

2. Software Skills Have Expanded Well Beyond IT Industries
As I mentioned above, software is everywhere.
Getting back to autos: Ford, GM, Chrysler and other automakers are now competing with Silicon Valley for software engineers. In the article "Calling all Codaholics", automaker Ford talks about how most of their most critical positions are now in software and systems engineering, creating a new war for technical talent in Detroit.
Burning Glass data shows that over the last five years software jobs in manufacturing grew at 31%, software jobs in healthcare grew by 72%, software jobs in financial services grew by 40%, and software jobs in retail industries grew by 98%. Retailers need software engineers to develop e-commerce systems, develop apps, and analyze BigData.

3. Software Jobs are No Longer Concentrated in High-Tech Havens
Software engineering used to be focused in Boston, San Francisco, Atlanta, and a few key geographies where software companies were born. Today software skills are needed everywhere and we can hire and manage people from around the world.
The Burning Glass research shows that Detroit, Baltimore, Austin, and Phoenix are all exploding with software positions (significantly higher growth rates than even Silicon Valley and Boston). And cities like Vancouver, Toronto, and eastern European locations are now becoming huge havens for software teams.

4. Software Skills are Easier Than Ever to Obtain
While software continues to be a professional career, more and more data shows that it is turning into a "craft."
You can learn to program online very easily (MOOC-enabled courses in computer science are available from Stanford, Harvard, MIT, and many other top educatinoal institutions) and if you're so inclined you can start writing code at home with no more investment than a laptop.
Back in the 1980s when I entered the tech industry you needed an employer with a huge mainframe and lots of tools. Today all these tools are free, so people with technical minds and enterprising hearts can learn to program themselves.
One of the best developers we have in our company has a history degree. He is just amazingly smart and loves to solve puzzles and learn new things. Many of the most creative new apps are built by people with very little formal software training. So this marketplace will become broader and more dynamic each year.

5. Software Engineers Trained Outside the US are Rapidly Growing in Number
While this is clearly a renaissance for job opportunities in the US, the global competition is fierce. In China, 41% of college grads have STEM degrees (science, technology, engineering, math) - in India it's 26%. In the US it's still only 13%, which is alarmingly low. (STEM research from Accenture.)
Companies are looking everywhere to find these skills, and this is why new marketplaces for technical skills have opened up. Look at Kaggle, which lets you put your data science projects online for competitive bidding. Companies like Odesk, Elance, Freelancer, and dozens more startup marketplaces are helping connect programmers to buyers.

 6. The Technologies and Tools for Software are Expanding
Software, like all "expanding economies," is a fragmented and disruptive space. Today there are dozens of types and disciplines within software (Burning Glass data tells us that Hadoop "BigData" and security are the hottest skills at the moment), so if you are entering this field it's important to think of yourself as a "continuous learner."
Twenty years ago software engineers learned about client/server technologies. Ten years ago we were focused on Java and the internet. Today the new areas are BigData, mobile, user experience design, search, and mobility applications. Artificial intelligence, machine learning, and gamification are coming next. The range of opportunities in software is broader than ever.

7.  There is a New War for Software Talent, Upping the Arms Race
For those of us who need to find and hire great software engineers, the demand for new hiring and sourcing tools is exploding. Innovative recruitment tools for software talent like Gild, TalentBin, RemarkableHire, and Entelo are transforming the way we recruit.  These new tools use the concepts of BigData in recruiting to search and find great engineers through their social footprint. (Read my article on the 9 Hottest Trends in Recruiting for more.)
The arms race for software talent has changed the way we pay and manage people. Think about the new perks like unlimited vacation, free food, and flexible work environments. Where did they start? Largely among Silicon Valley software companies who had to compete vigorously for engineers.
I'm an engineer and so are many of my friends. Engineers look for companies that give them great projects and an exciting, creative work environment. Companies have re-designed their workplace to attract great technical teams - and much of this work is paying off for the rest of us. (Read Deloitte's Center of the Edge new research on the work environment to learn more.)
The competition to find "great" engineers remains very high. Studies have shown that a top software developer can product 10-20X the output (and quality) of a junior or lower skilled developer. New tools like Agile and Pair development are improving productivity rapidly.
And age is not always an issue. Many senior software programmers who grew up using Cobol (which was object-oriented), for example, have turned into great developers in modern object-oriented programming environments.

You Need to Understand Software to Succeed

We really have entered a new "software economy."  Software has permeated every business function.
Can your VP of Marketing do their job without understanding the role of CRM, SEO, social networks, analytics, and customer segmentation? Can your product management team design great products without understanding information architecture, search, and mobile computing? Can manufacturers design supply chains without understanding how ERP and other forms of software manage data?
Software is now one of the most powerful ways to add value to your product, your marketing, and your business.

What's y if 9y=9÷9×9÷9×9?

What is the answer of 6:2(1+2)?

• What is the answer of 6:2(1+2)?