lunis suppe

  

The Web has become increasingly interactive over the years. This trend is set to continue with the next generation of applications driven by the real-time Web. Adding real-time functionality to an application can result in a more interactive and engaging user experience. However, setting up and maintaining the server-side real-time components can be an unwanted distraction. But don’t worry, there is a solution.

Cloud hosted Web services and APIs have come to the rescue of many a developer over the past few years, and real-time functionality is no different. The focus at Pusher, for example, is to let you concentrate on building your real-time Web applications by offering a hosted API which makes it quick and easy to add scalable real-time functionality to Web and mobile apps. In this tutorial, I’ll show how to convert a basic blog commenting system into a real-time engaging experience where you’ll see a comment made in one browser window "magically" appear in a second window.

Why Should We Care About The Real-Time Web?

Although the Real-Time Web is a relatively recent mainstream phrase, real-time Web technologies have been around for over 10 years. They were mainly used by companies building software targeted at the financial services sector or in Web chat applications. These initial solutions were classed as "hacks". In 2006 these solutions were given an umbrella term called Comet, but even with a defined name the solutions were still considered hacks. So, what’s changed?

In my opinion there are a number of factors that have moved real-time Web technologies to the forefront of Web application development.

Social Media Data

Social media, and specifically Twitter, has meant that more and more data is becoming instantly available. Gone are the days where we have to wait an eternity for Google to find our data (blog posts, status updates, images). There are now platforms that not only make our data instantly discoverable but also instantly deliver it to those who have declared an interest. This idea of Publish/Subscribe is core to the real-time Web, especially when building Web applications.

Increased User Expectations

As more users moved to using applications such as Twitter and Facebook, and the user experiences that they deliver, their perception of what can be expected from a Web application changed. Although applications had become more dynamic through the use of JavaScript, the experiences were seldom truly interactive. Facebook, with it’s real-time wall (and later other realtime features) and Twitter with it’s activity stream centric user interface, and focus on conversation, demonstrated how Web applications could be highly engaging.

WebSockets

Earlier on I stated that previous solutions to let servers instantly push data to Web browsers were considered "hacks". But this didn’t remove the fact that there was a requirement to be able to do this in a cross-browser and standardised way. Our prayers have finally been answered with HTML5 WebSockets. WebSockets represent a stardardized API and protocol that allows realtime server and client (web browser) two way communication over a single connection. Older solutions could only achieve two-way communication using two connections so the fact the WebSockets use a single connection is actually a big deal. It can be a massive resource saving to the server and client, with the latter being particularly important for mobile devices where battery power is extremely valuable.

How are Real-Time Technologies being used?

Real-time Web technologies are making it possible to build all sorts of engaging functionality, leading to improved user experiences. Here are a handful of common use cases:

• Realtime Stats – The technology was first used in finance to show stock exchange information so it’s no surprise that the technology is now used more than ever. It’s also highly beneficial to sports, betting and analytics.
• Notifications – when something a user is interested in becomes available or changes.
• Activity Streams – streams of friend or project activity. This is commonly seen in apps like Twitter, Facebook, Trello, Quora and many more.
• Chat – the 101 or real-time Web technology but still a very valuable function. It helps delivery instant interaction between friends, work colleagues, customers and customer service etc.
• Collaboration – Google docs offers this kind of functionality within its docs, spreadsheets and drawing applications and we’re going to see similar use cases in many more applications.
• Multiplayer Games – The ability to instantly deliver player moves, game state changes and score updates is clearly important to multiplayer gaming.

In the rest of this tutorial I’ll cover building a basic blog commenting system, how to progressively enhance it using jQuery and finally I’ll also progressively enhance it using the real-time Web service I work for, Pusher, which will demonstrate not just how easy it can be to use real-time Web technology, but also the value and increased engagement that a real-time factor can introduce.

Creating Generic Blog Commenting System

Start from a Template

I want to focus on adding real-time commenting to a blog post so let’s start from a template.

This template re-uses the HTML5 layout defined in the post on Coding An HTML 5 Layout From Scratch and the file structure we’ll start with is as follows (with some additions that we don’t need to worry about at the moment):

• css (dir)
  • global-forms.css
  • main.css
  • reset.css
• images (dir)
• index.php

HTML

The template HTML, found in index.php, has been changed from the HTML5 Layout article to focus on the content being a blog post with comments. You can view the HTML source here.

The main elements to be aware of are:

• <section id="content"> – the blog post content
• <section id="comments"> – where the comments are to appear. This is where the majority of our work will be done

Comments

Now that we’ve got the HTML in place for our blog post and for displaying the comments we also need a way for our readers to submit comments, so let’s add a <form> element to collect and submit the comment details to post_comment.php. We’ll add this at the end of the <section id="comments"> section wrapped in a <div id="respond">.

<div id="respond">

	<h3>Leave a Comment</h3>

	<form action="post_comment.php" method="post" id="commentform">

		<label for="comment_author" class="required">Your name</label>
		<input type="text" name="comment_author" id="comment_author" value="" tabindex="1" required="required">

		<label for="email" class="required">Your email;</label>
		<input type="email" name="email" id="email" value="" tabindex="2" required="required">

		<label for="comment" class="required">Your message</label>
		<textarea name="comment" id="comment" rows="10" tabindex="4"	 required="required"></textarea>

		<-- comment_post_ID value hard-coded as 1 -->
		<input type="hidden" name="comment_post_ID" value="1" id="comment_post_ID" />
		<input name="submit" type="submit" value="Submit comment" />

	</form>

</div>

Comment Form CSS

Let’s apply some CSS to make things look a bit nicer by adding the following to main.css:

#respond {
	margin-top: 40px;
}

#respond input[type='text'],
#respond input[type='email'],
#respond textarea {
	margin-bottom: 10px;
	display: block;
	width: 100%;

	border: 1px solid rgba(0, 0, 0, 0.1);
	-webkit-border-radius: 5px;
	-moz-border-radius: 5px;
	-o-border-radius: 5px;
	-ms-border-radius: 5px;
	-khtml-border-radius: 5px;
	border-radius: 5px;

	line-height: 1.4em;
}

Once the HTML structure, the comment form and the CSS are in place our blogging system has started to look a bit more presentable.

Handling Comment Submission

The next step is to write the PHP form submission handler which accepts the request and stores the comment, post_comment.php. You should create this file in the root of your application.

As I said earlier I’m keen to focus on the real-time functionality so a class exists within the template that you’ve downloaded which encapsulate some of the standard data checking and persistence functionality. This class is defined in Persistence.php (you can view the source here), is in no way production quality, and handles:

• Basic validation
• Basic data sanitization
• Very simple persistence using a user $_SESSION. This means that a comment saved by one user will not be available to another user.

This also means that we don’t need to spend time setting up a database and all that goes with it and makes post_comment.php very simple an clean. If you wanted to use this functionality in a production environment you would need to re-write the contents of Persistence.php. Here’s the code for post_comment.php.

<?php
require('Persistence.php');

$db = new Persistence();
if( $db->add_comment($_POST) ) {
	header( 'Location: index.php' );
}
else {
	header( 'Location: index.php?error=Your comment was not posted due to errors in your form submission' );
}
?>

The above code does the following:

1. Includes the Persistence.php class which handles saving and fetching comments.
2. Creates a new instances of the Persistence object and assigns it to the variable $db.
3. Tries to add the comment to the $db. If the comment is successfully added it redirects back to the blog post. If it fails the redirection also occurs but some error text is appended to an error query parameter.

Displaying the Comments with the Blog Post

The final thing we need to do to have our Generic Blog Commenting System up and running is to update the blog page, index.php, to fetch and display the comments from the Persistence object.

• Since this isn’t a real blogging system we’ll hard code the $comment_post_ID value to be 1.
• Include the Persistence.php class and fetch the comments from it. Comments are associated with a blog post using a unique $comment_post_ID.

<?php
require('Persistence.php');
$comment_post_ID = 1;
$db = new Persistence();
$comments = $db->get_comments($comment_post_ID);
$has_comments = (count($comments) > 0);
?>

Since we now have the $comment_post_ID accessible via PHP we should update the HTML for the comment form to use this value.

<input type="hidden" name="comment_post_ID" value="<?php echo($comment_post_ID); ?>" id="comment_post_ID" />

We now have all the comments related to the blog post referenced by the $comments variable we need to display them on the page. To do this we need to update the PHP in index.php to iterate through them and create the required HTML.

<ol id="posts-list" class="hfeed<?php echo($has_comments?' has-comments':''); ?>">
	<li class="no-comments">Be the first to add a comment.</li>
	<?php
		foreach ($comments as $comment) {
			?>
			<li><article id="comment_<?php echo($comment['id']); ?>" class="hentry">
				<footer class="post-info">
					<abbr class="published" title="<?php echo($comment['date']); ?>">
						<?php echo( date('d F Y', strtotime($comment['date']) ) ); ?>
					</abbr>

					<address class="vcard author">
						By <a class="url fn" href="#"><?php echo($comment['comment_author']); ?></a>
					</address>
				</footer>

				<div class="entry-content">
					<p><?php echo($comment['comment']); ?></p>
				</div>
			</article></li>
			<?php
		}
	?>
</ol>

You’ll notice that if the value of $has_comments is true an additional CSS class is applied to the ordered list called has-comments. This is so we can hide the list item with the "Be the first to add a comment" message when comments are being displayed using CSS:

#posts-list.has-comments li.no-comments {
	display: none;
}

Now that all this is in place we have a functional blog commenting system. If you would like to start writing your code from this basic functioning blog commenting system you can also download the code completed up to here.

Progressive Enhancement With jQuery

The first step in making our blog commenting system feel less like a Web page and more like an application is to stop page reloads when a user submits a comment. We can do this by submitting the comments to the server using an AJAX request. Since jQuery is probably the defacto standard for cross browser JavaScript functionality we’ll use it here. Although I’m using jQuery here, I’d also like to highlight that it’s a good idea to not always use jQuery. Instead, analyze your scenario and make a considered decision because there are some cases where you are best not to.

In an attempt to try and keep as much scripting (PHP and JavaScript) from the index.php file we’ll create a new folder for our JavaScript and in there a file for our application JavaScript. The path to this fill should be js/app.js. This file should be included after the jQuery include.

<script src="http://code.jquery.com/jquery-1.7.1.min.js"></script>
<script src="js/app.js"></script>

Capture the Comment Form Submission

When the page is ready bind to the submit event of the form.

$(function() {
	$('#commentform').submit(handleSubmit);
});

When the form is submitted and the handleSubmit function is called the comment data we want to send to the server is extracted from the form. There are more elegant ways of fetching the data from the form but this approach clearly shows where we’re getting the values from and the data object we are creating.

function handleSubmit() {
	var form = $(this);
	var data = {
		"comment_author": form.find('#comment_author').val(),
		"email": form.find('#email').val(),
		"comment": form.find('#comment').val(),
		"comment_post_ID": form.find('#comment_post_ID').val()
	};

	postComment(data);

	return false;
}

function postComment(data) {
	// send the data to the server
}

POST the Comment with AJAX

Within the postComment function make a POST request to the server passing the data that we’ve retrieved from the form. When the request is made an additional HTTP header will be sent to identify the request as being an AJAX request. We want to do this so that we can return a JSON response if it is an AJAX request and maintain our basic functionality if it isn’t (for more information on this see Detected AJAX events on the Server). We also define two handlers; postSuccess for handling the comment being successfully stored and postError to handle any failures.

function postComment(data) {
	$.ajax({
		type: 'POST',
		url: 'post_comment.php',
		data: data,
		headers: {
			'X-Requested-With': 'XMLHttpRequest'
		},
		success: postSuccess,
		error: postError
	});
}

function postSuccess(data, textStatus, jqXHR) {
	// add comment to UI
}

function postError(jqXHR, textStatus, errorThrown) {
	// display error
}

Dynamically Updating the User Interface with the Comment

At this point the comment data is being sent to the server and saved, but the AJAX response isn’t providing any meaningful response. Also, the comments section isn’t being updated to show the newly submitted comment so the user would have to refresh the page to see it. Let’s start by writing the code to update the UI and test that functionality.

If you are thinking "Hang on a minute! We don’t have all the data we need from the Web server to display the comment" then you are correct. However, this doesn’t stop us writing the code to update the UI and also testing that it works. Here’s how:

function postSuccess(data, textStatus, jqXHR) {
	$('#commentform').get(0).reset();
	displayComment(data);
}

function displayComment(data) {
	var commentHtml = createComment(data);
	var commentEl = $(commentHtml);
	commentEl.hide();
	var postsList = $('#posts-list');
	postsList.addClass('has-comments');
	postsList.prepend(commentEl);
	commentEl.slideDown();
}

function createComment(data) {
	var html = '' +
	'<li><article id="' + data.id + '" class="hentry">' +
		'<footer class="post-info">' +
			'<abbr class="published" title="' + data.date + '">' +
				parseDisplayDate(data.date) +
			'</abbr>' +
			'<address class="vcard author">' +
				'By <a class="url fn" href="#">' + data.comment_author + '</a>' +
			'</address>' +
		'</footer>' +
		'<div class="entry-content">' +
			'<p>' + data.comment + '</p>' +
		'</div>' +
	'</article></li>';

	return html;
}

function parseDisplayDate(date) {
	date = (date instanceof Date? date : new Date( Date.parse(date) ) );
	var display = date.getDate() + ' ' +
								['January', 'February', 'March',
								 'April', 'May', 'June', 'July',
								 'August', 'September', 'October',
								 'November', 'December'][date.getMonth()] + ' ' +
								date.getFullYear();
	return display;
}

The code above does the following:

• Within the postSuccess function we clear the form values and call displayComment.
• displayComment first calls the createComment function to create the list item (<li>) HTML as a String.
• We then convert the HTML to a jQuery object using $(commentHtml) and hide the element.
• The comment list item is then prepended to the comments ordered list (<ol>). The list also has a class called has-comments added to it so we can hide the first list item which contains the "Be the first to comment" statement.
• Finally, we call commentEl.slideDown() so that the comment is shown in what is becoming the standard "here’s a new item" way.

The functionality is now implemented but we want to test it out. This can be achieved in two ways:

• The displayComment is a global function so we can call it directly using the JavaScript console of the browser.
• We can bind to an event on the page that triggers a fake update which calls the displayComment function

Let’s go with the latter and bind to the "u" key being released by binding to the keyup event. When it is, we’ll create a fake data object containing all the information required to create a new comment and pass it to the displayComment function. That comment will then be displayed in the UI.

Hit the "u" key a few times and see the comments appear.

$(function() {

	$(document).keyup(function(e) {
		e = e || window.event;
		if(e.keyCode === 85){
			displayComment({
				"id": "comment_1",
				"comment_post_ID": 1,
				"date":"Tue, 21 Feb 2012 18:33:03 +0000",
				"comment": "The realtime Web rocks!",
				"comment_author": "Phil Leggetter"
			});
		}
	});

});

Great! We now know that our displayComment function works exactly as we expect it to. Remember to remove the test function before you go live or you’ll really confuse your user every time they press "u".

Detect and Responding to the AJAX request

All that’s left to do is update the post_comment.php file to detect the AJAX call and return information about the newly created comment.

Detecting the AJAX request is done by checking for the X-Requested-With header:

$ajax = ($_SERVER[ 'HTTP_X_REQUESTED_WITH' ] === 'XMLHttpRequest');

Once we know the request is an AJAX request we can update the code to respond with an appropriate status code and the data representing the comment. We also need to ensure that the original functionality is maintained. The post_comment.php code now looks as follows:

<?php
require('Persistence.php');

$ajax = ($_SERVER[ 'HTTP_X_REQUESTED_WITH' ] === 'XMLHttpRequest');

$db = new Persistence();
$added = $db->add_comment($_POST);

if($ajax) {
	sendAjaxResponse($added);
}
else {
	sendStandardResponse($added);
}

function sendAjaxResponse($added) {
	header("Content-Type: application/x-javascript");
	if($added) {
		header( 'Status: 201' );
		echo( json_encode($added) );
	}
	else {
		header( 'Status: 400' );
	}
}

function sendStandardResponse($added) {
	if($added) {
		header( 'Location: index.php' );
	}
	else {
		header( 'Location: index.php?error=Your comment was not posted due to errors in your form submission' );
	}
}
?>

Key points from the above code are:

• The $db->add_comment($_POST) call returns the data from the added comment which is assigned to the $added variable.
• By setting the response Content-Type to “application/json” we tell jQuery to convert the returned string into a JavaScript object. For more information on calling JSON Web services see A Beginner’s Guide To jQuery-Based JSON API Clients.
• The 201 status code indicates a successful call and also that a resource (the comment) was created by the call.

The blog commenting system now works in a much more dynamic way, instantly showing the user that their comment has been posted without refreshing the page. In addition, the way the we’ve added the JavaScript based functionality to the page means that if JavaScript is disabled or a JavaScript file fails to load that the system will fallback to the standard functionality we first implemented.

Getting Real-Time—Finally!

As with any "from scratch" tutorial it can take a bit of time to get to the really interesting part, but we’re finally here. However, all the work we’ve up in has been worth it. Because we’ve built our commenting system up in a progressively enhanced way, plugging Pusher into it is going to be really easy

What is Pusher?

At the start of the tutorial we said that we would use Pusher to add the realtime functionality to the application. But what is Pusher?

Pusher is a hosted service for quickly and easily adding realtime features into Web and mobile applications. It offers a RESTful API that makes it really easy to publish events from any application that can make a HTTP request and a WebSocket API for realtime bi-directional communication. You don’t even need to use the APIs directly as there are server (PHP, Ruby, node.js, ASP.NET, Python and more) and client (JavaScript, iOS, Android, .NET, ActionScript, Arduino and more) libraries available in a host of technologies which means you can add realtime functionality to an app within minutes ‐ I’m confident you’ll be surprised just how easy!

Sign up for Pusher and get your API Credentials

In order to add Pusher-powered real-time functionality to a Web application you first need to sign up for a free Sandbox account. After you have signed up you’ll be taken to the Pusher dashboard where you’ll see that a "Main" application has been created for you. You’ll also see you are in the "API Access" section for that application where you can grab your API credentials.

For ease of access create a pusher_config.php file and define the credentials in there so we can refer to them later:

<?php
define('APP_ID', 'YOUR_APP_ID');
define('APP_KEY', 'YOUR_APP_KEY');
define('APP_SECRET', 'YOUR_APP_SECRET');
?>

In your version of pusher_config.php be sure to replace the values which being ‘YOUR_ with your actual credentials.

You should also require this in your index.php file. We should also make the APP_KEY available to the JavaScript runtime as we are going to need it to connect to Pusher.

<?php
require('pusher_config.php);
?>

<script>
var APP_KEY = '<?php echo(APP_KEY); ?>';
</script>

Real-time JavaScript

The first thing you need to do when adding Pusher to a Web application is include the Pusher JavaScript library and connect to Pusher. To connect you’ll need to use the key which you grabbed from the Pusher dashboard. Below you can see all that is required to hook up the front-end application to Pusher.

Include the Pusher JavaScript library after the app.js include:

<script src="http://code.jquery.com/jquery-1.7.1.min.js"></script>
<script src="http://js.pusher.com/1.11/pusher.min.js"></script>
<script src="js/app.js"></script>

Add the Pusher functionality to app.js:

var pusher = new Pusher(APP_KEY);
var channel = pusher.subscribe('comments-' + $('#comment_post_ID').val());
channel.bind('new_comment', displayComment);

This probably looks too easy to be true, so here are the details about what the above code does:

• var pusher = new Pusher(APP_KEY);
  Creates a new instance of a Pusher object and in doing so connects you to Pusher. The application to use is defined by the APP_KEY value that you pass in and that we set up earlier.
• var channel = pusher.subscribe('comments-' + $('#comment_post_ID').val());
  Channels provide a great way of organizing streams of real-time data. Here we are subscribing to comments for the current blog post, uniquely identified by the value of the comment_post_ID hidden form input element.
• channel.bind('new_comment', displayComment);
  Events are used to further filter data and are ideal for linking updates to changes in the UI. In this case we want to bind to an event which is triggered whenever a new comment is added and display it. Because we’ve already created the displayComment function we can just pass in a reference to the call to bind.

Sending Real-Time Events using the Event Creator

We can also test out this functionality without writing any server-side code by using the Event Creator for your app which can also be found in the Pusher dashboard. The Event Creator lets you publish events on a channel through a simple user interface. From the code above we can see that we want to publish an event named "new_comment" on the "comments-1" channel. From the earlier test function we also have an example of the test data we can publish.

Real-time PHP

Again, we’ve proven that our client-side functionality works without having to write any server-side code. Now lets add the PHP code we need to trigger the new comment event as soon as a comment is posted in our comment system.

Pusher offers a number of server-side libraries which make it easy to publish events in addition to helping with functionality such as private channel authentication and providing user information for presence channels. We just want to use the basic event triggering functionality in the post_comment.php file so we need to download the Pusher PHP library (direct zip file download).

Once you’ve downloaded this zip file, unzip it into the directory along with your other files. Your file structure will now look something like this:

• index.php
• css (dir)
• images (dir)
• post_comment.php
• pusher_config.php
• Persistence.php
• squeeks-Pusher-PHP (dir)
  • lib (dir)
    • Pusher.php

An event can be triggering in just a few lines of code:

<?php
require('squeeks-Pusher-PHP/lib/Pusher.php');
require('pusher_config.php');

$pusher = new Pusher(APP_KEY, APP_SECRET, APP_ID);
$pusher->trigger('comments-1', 'new_comment', array(
	"comment_post_ID" => 1,
	"date" => "Tue, 21 Feb 2012 18:33:03 +0000",
	"comment" => "The realtime Web rocks!",
	"comment_author" => "Phil Leggetter"
));
?>

However, we need to apply a some additional logic before we trigger the event:

• Check that the comment was added.
• Extract the unique comment identifier from the $added array.
• Build the text to identify a channel name for the submitted comment.
• Trigger a new_comment event on the channel with the $added data. Note: The library automatically converts the $added array variable to JSON to be sent through Pusher.

Therefore the full post_comment.php file ends up looking as follows:

<?php
require('Persistence.php');
require('squeeks-Pusher-PHP/lib/Pusher.php');
require('pusher_config.php');

$ajax = ($_SERVER[ 'HTTP_X_REQUESTED_WITH' ] === 'XMLHttpRequest');

$db = new Persistence();
$added = $db->add_comment($_POST);

if($added) {
	$channel_name = 'comments-' . $added['comment_post_ID'];
	$event_name = 'new_comment';

	$pusher = new Pusher(APP_KEY, APP_SECRET, APP_ID);
	$pusher->trigger($channel_name, $event_name, $added);
}

if($ajax) {
	sendAjaxResponse($added);
}
else {
	sendStandardResponse($added);
}

function sendAjaxResponse($added) {
	header("Content-Type: application/json");
	if($added) {
		header( 'Status: 201' );
		echo( json_encode($added) );
	}
	else {
		header( 'Status: 400' );
	}
}

function sendStandardResponse($added) {
	if($added) {
		header( 'Location: index.php' );
	}
	else {
		header( 'Location: index.php?error=Your comment was not posted due to errors in your form submission' );
	}
}
?>

If you run the app now in two different browser windows you’ll see that as soon as you submit a comment in one window that comment will instantly ("magically") appear in the second window. We now have a real-time commenting system!

But…, we’re not done quite yet. You’ll also see that the comment is shown twice in the window of the user who submitted it. This is because the comment has been added by the AJAX callback, and by the Pusher event. Because this is a very common scenario, especially if you’ve built an application in a progressively enhanced way as we have, the Pusher server libraries expose a way of excluding a connection/user from receiving the event via Pusher.

In order to do this you need to send a unique connection identifier called a socket_id from the client to the server. This identifier can then be used to define who will be excluded.

function handleSubmit() {
	var form = $(this);
	var data = {
		"comment_author": form.find('#comment_author').val(),
		"email": form.find('#email').val(),
		"comment": form.find('#comment').val(),
		"comment_post_ID": form.find('#comment_post_ID').val()
	};

	var socketId = getSocketId();
	if(socketId !== null) {
		data.socket_id = socketId;
	}

	postComment(data);

	return false;
}

function getSocketId() {
	if(pusher && pusher.connection.state === 'connected') {
		return pusher.connection.socket_id;
	}
	return null;
}

The changes we’ve made are:

• A new function called getSocketId has been added to get the socket_id. It wraps a check to ensure that the pusher variable has been set and also that the client is connected to Pusher.
• The handleSubmit has been updated to check to see if a socket_id is available. If it is, this information is posted to the server along with the comment data.

On the server we need to use the socket_id parameter if it is present and therefore exclude the connection and user who submitted the comment, or pass in null if it’s not:

$channel_name = 'comments-' . $added['comment_post_ID'];
$event_name = 'new_comment';
$socket_id = (isset($_POST['socket_id'])?$_POST['socket_id']:null);

$pusher = new Pusher(APP_KEY, APP_SECRET, APP_ID);
$pusher->trigger($channel_name, $event_name, $added, $socket_id);

And as simple as that we have a fully realtime enabled blog commenting system and we only send messages to users who really need to see them. As with the AJAX functionality the realtime functionality has been added in a progressively enhancing way, to ensure it doesn’t impact on any other functionality. You can find the a demo running here and the completed solution in the realtime commenting repository in github.

Good Real-Time App Development Practices

Real-time application development shares common good development practices with general Web development. However, I thought I would share a couple of tips that can come in particularly handy.

Use Browser Developer Tools

When you start doing a lot of JavaScript development the browser developer tools becomes your best friend. It’s the same when adding realtime functionality to your Web app, not only because you are using JavaScript, but also because the JavaScript library you are using is likely to be doing some reasonably complex things internally. So, the best way of understanding what is going on and if your code is using it as expect is to enable logging which usually goes to the developer console. All major browser vendors now offer good developer tools which include a console:

• Firebug addon for Firefox
• Google Chrome Developer Tools
• Internet Explorer F12 developer tools
• Opera Dragonfly
• Safari Developer Tools

The Pusher JavaScript library provides a way to hook into the logging functionality. All you need to do is assign a function to the Pusher.log static property. This function will then receive all log messages. You can do what you like within the function but best practice is to log the messages to the developer console. You can do this as follow, ensuring the code it executed after the Pusher JavaScript library include:

Pusher.log = function(msg) {
	if(window.console && window.console.log) {
		window.console.log(new Date().getTime() + ': ' + msg);
	}
};

The code above checks to make sure the console and log function is available – it’s not in older browsers – and then logs the messages along with a timestamp to the JavaScript console. This can be incredibly handy in tracking down problems.

Check Connectivity

Any good real-time technology will maintain a persistent connection between the client (web browser) and the Web server or service. Sometimes the client will lose connectivity and when the client isn’t connected to the Internet the real-time functionality won’t work. This can happen a lot with applications running on mobile devices which rely on mobile networks. So, if your application relies on that connectivity and functionality then it’s important to deal with scenarios where the client isn’t connected. This might be by displaying a message to tell the user they are offline or you might want to implement some alternative functionality.

The Pusher JavaScript library exposes connectivity state via the pusher.connection object, which we briefly saw earlier when fetching the socket_id. Binding to state changes and implementing your required functionality is quite simple as it follows the same mechanism as binding to events on channels:

var pusher = new Pusher(APP_KEY);

pusher.connection.bind('state_change', function(states) {
	Pusher.log('Connection state changed from: ' + states.previous + ' to ' + states.current);
});

Conclusion

We’re seeing real-time functionality appearing in a large number of high profile applications: some have it at the core of what they offer whilst others use it sparingly. No matter how it is used the end goal is generally the same; to enhance the user experience and keep users engaged with an application. By converting a basic blog commenting system into a more engaging communication platform I hope I’ve demonstrated that the functionality and experience can easily be layered on existing application.

The ease of access to this technology is a relatively new thing and we’ve only just touched the potential uses for the technology. Real-time stats, instant news updates, activity streams, social interaction, collaboration and gaming are just a few common uses but as more developers become aware of, and comfortable with, the technology I’m confident that we’re going to see some truly amazing innovation. An "Internet of Real-Time Things?"?

© Phil Leggetter for Smashing Magazine, 2012.

  

Welcome to the first in a new series of interviews called “How I Work”. These interviews revolve around how thinkers and creators in the Web world design, code, and create. The goal is not to get into the specific nuances of their craft (as that information already exists online), but rather step back and learn a bit about their habits, philosophies, and workflow for producing great work.

Meet Doug Crockford

First up is Douglas Crockford who believes JavaScript might just be the most elegant language ever. Learn why he thinks you should study the history of computer science, the value of reading your code in front of other people, and that jQuery really is a good thing.

Douglas Crockford is known as The JavaScript Guy. He’s famous not only for his O’Reilly book JavaScript: The Good Parts but even more so as the visionary behind the JSON data format as well as the JSLint tool. He was featured in the book Coders at Work for his contributions and philosophies on what JavaScript got right, and what it didn’t.

As a native of Southern California, Doug has the build of a surfer; lean and tall with white hair and a beard. A veteran of Silicon Valley, he’s worked at Atari Labs, founded and worked at numerous software start-ups, was head of technology at Lucas Films and now has the enviable job of being a JavaScript evangelist at Yahoo!.

Image credits go to Eric Miraglia.

Self-taught (as many of the greats are), he says his goal is simply to get more people coding in JavaScript, or any language for that matter. While his day job may be as a JavaScript evangelist, speaking with Doug you get the sense he really is an evangelist for programming in general.

Below is a conversation that took place in Bozeman, Montana prior to a talk at Montana State University. Doug freely shared his thoughts on great programmers, user empathy, and how JSON restored his faith in humanity.

Why do you feel programmers should study the history of Computer Science?

Well, first semester of physics is a history class. You study Galileo and Newton and all their contributions to the field and that gives us the overall view of physics. It’s a really nice place to start.

I wish CS would do that. It doesn’t seem to have enough value in its history and it’s a really amazing history that’s completely neglected. It’s rarely that the best idea won. So, we’ve taken different paths over the years and maybe haven’t realized why.

Ironically, despite the rate of change in technology, we see in the story of software that it takes a generation to retire or die off before we have a critical mass of bright young minds to embrace new ideas.

I think if people were more aware of their history, they could see these patterns more easily.

What were the traits of the weak programmers you’ve seen over your career?

That’s an easy one—lack of curiosity. They were so satisfied with the work that they were doing was good enough (without an understanding of what ‘good’ was) that they didn’t push themselves.

I’m much more impressed with people that are always learning. The brilliant programmers I’ve been around are always learning.

You see so many people get into one language and spend their entire career in that language, and as a result aren’t that great as programmers.

Do you feel that the pain a programmer goes through in learning a language contributes to this unhealthy attachment to using only one language?

My advice to programmers to avoid this trap is to learn lots of different languages. We’re in sort of a language renaissance right now and there are a ton of brilliant languages to learn from.

To learn new languages takes nights and weekends outside of work, and that’s a commitment. The great programmers are the people that are constantly picking a project and diving into it, learning a language that way.

In Coders at Work, you stress the importance of doing code readings with teams. Why do you feel it’s important to present your code in front of other people?

Well, over the years I noticed that there are some terrific programmers out there that are completely content to sit in their cave all day long writing brilliant code. But they don’t interact much with their team, which means it’s a lost opportunity for mentoring other members.

As you know, a lot of coders aren’t the most socially adept animals either.

So, my idea with code reading sessions is to provide a forum where people can come together and read for each other to get them out of their caves. The masters read for the beginners, and vice versa, as a team-building exercise.

The trick for success is to set up rules ahead of time so that nobody is going to get spanked and everyone is respectful in their feedback. It has to be a good learning experience for everyone. You have to be careful with a dysfunctional team, because it can quickly tear apart the group. But I always call the game before it gets that far.

The rules are that it’s about improving the quality of the code that we’re all responsible for, improving the quality of our team, and improving our individual capabilities.

Some people see this as a terrible time sink. Yet, I’ve found by doing this exercise, bugs are caught way ahead of time and you can prevent a team member from going off the tracks. But again, that’s not the goal, it’s about team building.

Over time the masters help pull up the beginners and the overall output from the team gets better.

Are programmers getting better at user empathy?

The best experience I had with empathy was working in marketing support. There were times I would go out into the field and hold hands with the customers and see the consequences firsthand of some of the crap we were delivering to them.

I was shocked when I moved into systems programming and how the programmers actually held the customer in contempt.

I think every programmer should work in customer support for the product they’re delivering.

It’s another case of over-specialization. “I just write the code,” is the response you get and the programmers don’t see it as a chance to improve peoples’ lives.

How much of a language do you need to know?

Virtually every programming language is too big. Language standards have difficulty removing unnecessary features but as users we can choose not to use it.

I would say you can do 100% with knowing 50% of the language.

The language that taught me that lesson the most was JavaScript, because it has more bad parts than good parts. It gave me a very strong motivation for figuring out what are the good parts and what are the bad parts, and what the criteria is for deciding what’s in or out.

And the good parts are just so good. Be sure to watch Doug’s Google Tech talk titled “JavaScript: The Good Parts.”

What approaches would you say a master has versus a beginner?

When I was a journeyman, I was a maximilist. I tried to use the whole language. While I don’t know if I would call myself a master now, I’m certainly a minimalist. I’ve tried to get good at using as little of the language as possible.

I place a lot of value in simplicity and minimalism.

What are your thoughts on jQuery? Some JS enthusiasts feel like it’s letting people off the hook from truly learning JS.

There is some really clever stuff in jQuery and I think John Resig did some very good work there.

I do have a problem with anybody doing anything without understanding what they’re doing. I’m not going to fault jQuery for attracting those sorts of people.

But I do think there are some other AJAX libraries that maybe doing a better job that aren’t quite as accessible. However, I think there is a place for all of these things.

When you were developing JSON was it tough to pull back and not put too much into it?

My design criteria were three things: minimal, contextual, and a subset of JavaScript.

The last constraint was to keep us from going off the rails and inventing new stuff. We had to only use stuff that was in JavaScript, which meant that our unicode handling wasn’t quite right because JS isn’t quite right, which was disappointing. We don’t have proper support for dates because JS didn’t have it. But we can work around both of these.

But it also meant that when somebody proposed, “Hey we should do this crazy thing” we could be like “Nope”. So, we had a really easy criteria for stopping extra features from being added.

One interesting story about leaving things out: as we got closer to releasing JSON I decided to take out the ability to do comments. When translating JSON into other languages, often times the commenting piece was the most complicated part. By taking the commenting out we reduced the complexity of the parsers by half—everything else was just too simple.

One of the best features of JSON is that it’s stable. If your program works now, it will work forever, and that is an attractive thing.

I still get notes from people saying they’ve got great ideas for the next version. But there isn’t going to be a next version. I always say you’re free to invent a new standard and promote it as much as you like.

How did JSON get adopted?

You know, the adoption of JSON sort of restored my faith in humanity because it was a good idea that won out, only because it was a good idea.

It was a case where there were no slick marketing campaigns. In 2001, I started working on it as a way to tie the browsers to the server. At the time, everyone thought XML had to be used or they’d say “that’s a great idea but JSON isn’t a standard”. So, I bought json.org, made a logo, threw up a Web page and it sat out on the Web for three years.

In the meantime, AJAX happened and when it became the way for writing applications JSON was there. There was counter promotion from the XML community, of course.

But when I arrived at Yahoo! some kids at the company started thinking it was okay to start shipping JSON API’s through Web services. And developers found the apps got faster and were easier to write.

It sort of took off from there—no slick campaigns. So a good idea based on simplicity won out for once.

Watch Doug Crockford At Google Speaking On “JavaScript: The Good Parts”

In this presentation from Google Tech Talks, Doug goes over the ideas behind his landmark book, JavaScript: The Good Parts, and dives into the areas of what JavaScript got right and what it didn’t. Learn about the history and common roadblocks programmers run into when developing with this language.

Learn About The JSON Saga

In this video, Doug tells the interesting tale of how JSON was discovered, and sheds some light on how it became a major standard for describing data in an interesting turn of events.

(jvb) (il)

© Jacob Cook for Smashing Magazine, 2012.

  

Thanks to CSS3, we can create effects and animations without using JavaScript, which will facilitate the work of many designers.

But we must be careful to avoid abusing CSS3, not only because old browsers do not support all of its properties. In any case, we all see the potential of CSS3, and in this article we’ll discuss how to create an infinitely looping slider of images using only CSS3 animation.

Sections of This Article

To get a solid sense of the process from beginning to end, below is an outline of the article. Please note that this effect will only work properly in modern browsers that support the CSS3 properties being used.

1. Introduction
   Learn basic concepts related to CSS3 transitions and keyframe animation.
2. HTML markup
   Create the HTML markup for the sliding images.
3. CSS styles
   Create the style sheet to display the slider.
4. CSS3 keyframe animation
   Add animation to the slider (we’ll explain the various processes happening here).
5. Progress bar
   Add a progress bar for our slider.
6. Tooltip
   Add a tooltip to display the title of the image.
7. CSS3 transitions
   Animate the tooltip using CSS3 transitions to make it appear to hover over the image.
8. Pause and restart
   Pause the slider and restart it on hover.
9. Demo
   Check out the slider in action.
10. Conclusion
    Final considerations.

Screenshot of the pure CSS3 cycling slideshow.

1. Introduction

To follow this tutorial, having a basic understanding of CSS, especially CSS3 transitions and keyframe animation, is important. Using this simple concept, we will see how to make a functional image slider.

Basic Concepts of CSS3 Transitions

Normally when you change a CSS value, the change is instant. Now, thanks to the transition property, we can easily animate from the old to new state.

We can use four transition properties:

1. transition-property
   Defines the name(s) of the CSS properties to which the transitions should be applied.
2. transition-duration
   Defines the duration over which the transitions should occur.
3. transition-timing-function
   Determines how intermediate values of the transition are calculated. The effects from the timing function are commonly called easing functions.
4. transition-delay
   Defines when the transition starts.

At the moment, CSS3 transitions are supported in Safari 3.2+, Chrome, Firefox 4+, Opera 10.5+ and IE 10. Because the technology is still relatively new, prefixes for browsers are required. So far, the syntax is exactly the same for each browser, with only a prefix change required. We will omit them in the snippets in this article, but please remember to include the prefixes in your code.

Let’s see how to apply a simple transition to a link:

a {
   color: #000;
   transition-property: color;
   transition-duration: 0.7s;
   transition-timing-function: ease-in;
   transition-delay: 0.3s;
}

a:hover {
   color: #fff;
}

When assigning an animation to an element, you can also use the shorthand:

a  {
   color: #000;
   transition: color 0.7s ease-in 0.3s;
}

a:hover {
   color: #fff;
}

The W3C has a list of all “Animatable Properties.”

Basic Concepts of CSS3 Animations

CSS animation enables us to create animations without JavaScript by using a set of keyframes.

Unlike CSS transitions, keyframe animations are currently supported only in Webkit browsers and Firefox and soon in IE 10. Unsupported browsers will simply ignore your animation code.

The animation property has eight subproperties:

1. animation-delay
   Defines when the animation starts.
2. animation-direction
   Sets the animation to play in reverse on alternate cycles.
3. animation-duration
   Defines the length of time an animation takes to complete one cycle.
4. animation-iteration-count
   Defines the number of times an animation cycle should play before stopping.
5. animation-name
   Specifies the name of the @keyframes rule.
6. animation-play-state
   Determines whether an animation is running or paused.
7. animation-timing-function
   Describes how an animation progresses over one cycle.
8. animation-fill-mode
   Specifies how a CSS animation should apply styles to its target before and after executing.

Let’s see how to apply a simple animation to a div.

/* This is the element we are applying the animation to. */

div {
   animation-name: move;
   animation-duration: 1s;
   animation-timing-function: ease-in-out;
   animation-delay: 0.5s;
   animation-iteration-count: 2;
   animation-direction: alternate;

   -moz-animation-name: move;
   -moz-animation-duration: 1s;
   -moz-animation-timing-function: ease-in-out;
   -moz-animation-delay: 0.5s;
   -moz-animation-iteration-count: 2;
   -moz-animation-direction: alternate;

   -webkit-animation-name: move;
   -webkit-animation-duration: 1s;
   -webkit-animation-timing-function: ease-in-out;
   -webkit-animation-delay: 0.5s;
   -webkit-animation-iteration-count: 2;
   -webkit-animation-direction: alternate;
}

/* This is the animation code. */

@keyframes move {
   from {
      transform: translateX(0);
   }
   to {
      transform: translateX(100px);
   }
}

@-moz-keyframes move {
   from {
      -moz-transform: translateX(0);
   }
   to {
      -moz-transform: translateX(100px);
   }
}

@-webkit-keyframes move {
   from {
      -webkit-transform: translateX(0);
   }
   to {
      -webkit-transform: translateX(100px);
   }
}

But we can use the shorthand property to conveniently set all of the animation properties at once.

div {
   animation: move 1s ease-in-out 0.5s 2 alternate;
   -moz-animation: move 1s ease-in-out 0.5s 2 alternate;
   -webkit-animation: move 1s ease-in-out 0.5s 2 alternate;
}

Keyframes

Each keyframe describes how an animated element should render at a given point in the animation sequence. The keyframes take a percentage value to specify time: 0% is the start of the animation, while 100% is the end. You can optionally add keyframes for intermediate animations.

/* Animation from 0% to 100% */

@keyframes move {
   0% { transform: translateX(0); }
   100% { transform: translateX(100px); }
} 

/* Animation with intermediate keyframes */

@keyframes move {
   0% { transform: translateX(0); }
   50% { transform: translateX(20px); }
   100% { transform: translateX(100px); }
}

The W3C has a lot of useful and detailed information on “CSS3 Animations.”

Basic Structure of Our Slider

Now that we know how transitions and animation work, let’s see how to create our slider using only CSS3. This sketch shows how the animation should work:

How the animation slider functions

As you can see, the slider will be a container inside of which the images will be displayed.

The animation is very simple: the image follow a predefined path, animating the top property and changing the z-index and opacity properties when the image returns to its initial position.

Let’s dive right into the HTML markup to create the slider.

2. HTML Markup

The HTML markup is very simple; it’s all organized and SEO-friendly. Let’s see the full code first and then figure out in detail how everything works.

<div class="container">
   <div id="content-slider">
      <div id="slider">  <!-- Slider container -->
         <div id="mask">  <!-- Mask -->

         <ul>
         <li id="first" class="firstanimation">  <!-- ID for tooltip and class for animation -->
         <a href="#"> <img src="images/img_1.jpg" alt="Cougar"/> </a>
         <div class="tooltip"> <h1>Cougar</h1> </div>
         </li>

         <li id="second" class="secondanimation">
         <a href="#"> <img src="images/img_2.jpg" alt="Lions"/> </a>
         <div class="tooltip"> <h1>Lions</h1> </div>
         </li>

         <li id="third" class="thirdanimation">
         <a href="#"> <img src="images/img_3.jpg" alt="Snowalker"/> </a>
         <div class="tooltip"> <h1>Snowalker</h1> </div>
         </li>

         <li id="fourth" class="fourthanimation">
         <a href="#"> <img src="images/img_4.jpg" alt="Howling"/> </a>
         <div class="tooltip"> <h1>Howling</h1> </div>
         </li>

         <li id="fifth" class="fifthanimation">
         <a href="#"> <img src="images/img_5.jpg" alt="Sunbathing"/> </a>
         <div class="tooltip"> <h1>Sunbathing</h1> </div>
         </li>
         </ul>

         </div>  <!-- End Mask -->
         <div class="progress-bar"></div>  <!-- Progress Bar -->
      </div>  <!-- End Slider Container -->
   </div>
</div>

1. div id="slider"
   This is the main container of the slider. It does not have a particular function, but we will need it to pause the animation.
2. div id="mask"
   We will use this to hide everything that happens outside of the slider. In addition to hiding the content, the mask allows us to display the contents of the slider.
3. li id="first" class="firstanimation"
   Every list item has an ID and a class. The ID displays the tooltip, and the class is tied to the animation that has to occur.
4. div class="tooltip"
   This simply displays the title of the image. You can modify it to your needs; for example, by making it clickable and adding a short description.
5. div class="progress-bar"
   This contains the function that shows the progress of the animation.

Now it’s time for the CSS file.

3. CSS Style

Let’s create the basic structure of the slider. It will have the same image size. The border property will be useful to create a frame around the image.

/* SLIDER STRUCTURE */

#slider {
   background: #000;
   border: 5px solid #eaeaea;
   box-shadow: 1px 1px 5px rgba(0,0,0,0.7);
   height: 320px;
   width: 680px;
   margin: 40px auto 0;
   overflow: visible;
   position: relative;
}

The mask class will hide all of the elements that lie outside of the slider; its height must be equal to the height of the slider.

/* HIDE ALL OUTSIDE OF THE SLIDER */

#mask {
   overflow: hidden;
   height: 320px;
}

Finally, to sort the list of images, we’ll have position: absolute and top: -325px so that all of the images are positioned outside of the slider.

/* IMAGE LIST */

#slider ul {
   margin: 0;
   padding: 0;
   position: relative;
}

#slider li {
   width: 680px;  /* Width Image */
   height: 320px; /* Height Image */
   position: absolute;
   top: -325px;	/* Original Position - Outside of the Slider */
   list-style: none;
}

With these few lines of code, we have created our slider. Now we just need to add the animation.

4. CSS3 Keyframes Animation

Image animation for the slider

Before we begin with the animation, we have to specify some parameters in order to get the right view of the animation.

As we know, the total duration of the animation will be 25 seconds, but we have to know how many keyframes equals 1 second.

So, let’s work out a series of operations that gives us the exact number of keyframes based on the images we have and the total duration of the animation. Here are the calculations:

1. Define the total number of images to use in the slider
   5
2. Define the length of animation for each image
   5 seconds
3. Define the total duration of the animation
   Multiply the total number of images by the duration of each image:
   5 images × 5 seconds = 25 seconds
4. Calculate how many keyframes equals one second
   Divide the total number of keyframes by the total duration of the animation.
   100 keyframes / 25 seconds = 4 keyframes
   4 keyframes = 1 second

With all of this math, we can now apply the CSS animation to the slider. We will be able to put the animation on infinite loop because each image will follow its own animation that activates once it comes up in the slider.

#slider li.firstanimation {
   animation: cycle 25s linear infinite;
}

#slider li.secondanimation {
   animation: cycletwo 25s linear infinite;
}

#slider li.thirdanimation {
   animation: cyclethree 25s linear infinite;
}

#slider li.fourthanimation {
   animation: cyclefour 25s linear infinite;
}

#slider li.fifthanimation {
   animation: cyclefive 25s linear infinite;
}

Once the properties of the animation have been assigned, we need to use keyframes to set the animation in motion.

Following this principle, we can connect the animations to each other even though they are separate, which will give us an infinite loop.

I’ve added the opacity and z-index properties to make the transition from one image to the next more attractive.

As you can see in the code, the first animation has more keyframes than the rest. The reason for this is that when the gallery is started, the first image is positioned to make way for the second image; but when the last image has finished its animation, the first image has to have additional keyframes in order for the user not to see a break between animation cycles.

Here is all of the code for the animations:

/* ANIMATION */

@keyframes cycle {
   0%  { top: 0px; } /* When you start the slide, the first image is already visible */
   4%  { top: 0px; } /* Original Position */
   16% { top: 0px; opacity:1; z-index:0; } /* From 4% to 16 % = for 3 seconds the image is visible */
   20% { top: 325px; opacity: 0; z-index: 0; } /* From 16% to 20% = for 1 second exit image */
   21% { top: -325px; opacity: 0; z-index: -1; } /* Return to Original Position */
   92% { top: -325px; opacity: 0; z-index: 0; }
   96% { top: -325px; opacity: 0; } /* From 96% to 100% = for 1 second enter image*/
   100%{ top: 0px; opacity: 1; }
}

@keyframes cycletwo {
   0%  { top: -325px; opacity: 0; } /* Original Position */
   16% { top: -325px; opacity: 0; }/* Starts moving after 16% to this position */
   20% { top: 0px; opacity: 1; }
   24% { top: 0px; opacity: 1; }  /* From 20% to 24% = for 1 second enter image*/
   36% { top: 0px; opacity: 1; z-index: 0; }   /* From 24% to 36 % = for 3 seconds the image is visible */
   40% { top: 325px; opacity: 0; z-index: 0; } /* From 36% to 40% = for 1 second exit image */
   41% { top: -325px; opacity: 0; z-index: -1; }   /* Return to Original Position */
   100%{ top: -325px; opacity: 0; z-index: -1; }
}

@keyframes cyclethree {
   0%  { top: -325px; opacity: 0; }
   36% { top: -325px; opacity: 0; }
   40% { top: 0px; opacity: 1; }
   44% { top: 0px; opacity: 1; }
   56% { top: 0px; opacity: 1; }
   60% { top: 325px; opacity: 0; z-index: 0; }
   61% { top: -325px; opacity: 0; z-index: -1; }
   100%{ top: -325px; opacity: 0; z-index: -1; }
}

@keyframes cyclefour {
   0%  { top: -325px; opacity: 0; }
   56% { top: -325px; opacity: 0; }
   60% { top: 0px; opacity: 1; }
   64% { top: 0px; opacity: 1; }
   76% { top: 0px; opacity: 1; z-index: 0; }
   80% { top: 325px; opacity: 0; z-index: 0; }
   81% { top: -325px; opacity: 0; z-index: -1; }
   100%{ top: -325px; opacity: 0; z-index: -1; }
}
@keyframes cyclefive {
   0%  { top: -325px; opacity: 0; }
   76% { top: -325px; opacity: 0; }
   80% { top: 0px; opacity: 1; }
   84% { top: 0px; opacity: 1; }
   96% { top: 0px; opacity: 1; z-index: 0; }
   100%{ top: 325px; opacity: 0; z-index: 0; }
}

Having created the animations, we have to add a progress bar to display the duration of each animation.

5. Progress Bar

The progress bar for each animation

The process of animating the progress bar is the same as it was for the slider. First, we create the progress bar itself:

/* PROGRESS BAR */

.progress-bar {
   position: relative;
   top: -5px;
   width: 680px;
   height: 5px;
   background: #000;
   animation: fullexpand 25s ease-out infinite;
}

Don’t be afraid of the syntax here. It has the same function as from to; you can see that the keyframes set the appearance and disappearance of each image.

/* ANIMATION BAR */

@keyframes fullexpand {
   /* In these keyframes, the progress-bar is stationary */
   0%, 20%, 40%, 60%, 80%, 100% { width: 0%; opacity: 0; }

   /* In these keyframes, the progress-bar starts to come alive */
   4%, 24%, 44%, 64%, 84% { width: 0%; opacity: 0.3; }

   /* In these keyframes, the progress-bar moves forward for 3 seconds */
   16%, 36%, 56%, 76%, 96% { width: 100%; opacity: 0.7; }

   /* In these keyframes, the progress-bar has finished his path */
   17%, 37%, 57%, 77%, 97% { width: 100%; opacity: 0.3; }

   /* In these keyframes, the progress-bar will disappear and then resume the cycle */
   18%, 38%, 58%, 78%, 98% { width: 100%; opacity: 0; }
}

6. Tooltip

The slider is more or less complete, but let’s add a few details to make it more functional. We’ll insert tooltips for the image titles that will be visible on hover.

Simple tooltip

Here is the CSS for the tooltips:

   #slider .tooltip {
   background: rgba(0,0,0,0.7);
   width: 300px;
   height: 60px;
   position: relative;
   bottom: 75px;
   left: -320px;
}

#slider .tooltip h1 {
   color: #fff;
   font-size: 24px;
   font-weight: 300;
   line-height: 60px;
   padding: 0 0 0 10px;
}

Here we’ve made only the image titles visible, but you can do the same to custom text, links or whatever you like.

7. CSS3 Transitions

Animate the tooltip on hover

We have seen how to apply CSS3 transitions to elements; now let’s do it to the tooltips.

If you remember, we added an ID to each list (first, second, etc.) to have only the tooltip associated with an image appear on hover, rather than all of the tooltips appear together.

#slider .tooltip {
…
   transition: all 0.3s ease-in-out;
}

#slider li#first: hover .tooltip,
#slider li#second: hover .tooltip,
#slider li#third: hover .tooltip,
#slider li#fourth: hover .tooltip,
#slider li#fifth: hover .tooltip {
   left: 0px;
}

8. Pause And Restart

Stop slider on mouse hover

To allow users to pause to read content or look at an image, we should stop the animation when they hover over an image. (We’ll also have to stop the animation of the progress bar.)

#slider: hover li,
#slider: hover .progress-bar {
   animation-play-state: paused;
}

9. Demo

Finally, we’ve reached the end of the tutorial. The slider is now 100% complete!

Pure CSS3 cycling slider demo

Check out the demo. It works in Firefox 5+, Safari 4+ and Google Chrome, as well as the iPhone and iPad. You can also download the ZIP file.

Thanks to Massimo Righi for his images.

10. Conclusion

The effect is impressive, but admittedly, this slider is not very versatile. For instance, to add more images, you have to edit all of keyframes. CSS3 has great potential, but it does have limits, and sometimes JavaScript is preferable, depending on your target users.

This slider has some interesting features, such as pausing on hover and uniques link for the images, which allow users to interact with the slider. If you want full support among browsers, this is not possible, so JavaScript is recommended. Unfortunately, CSS3 animation has many limitations; its lack of flexibility in particular will prevent its widespread use for now. Hopefully this will spur you on to further study of CSS3.

Feel free to share your thoughts in the comments section below!

(al) (il)

© Alessio Atzeni for Smashing Magazine, 2012.

  

For years, the Web standards community has talked about the separation of concerns. Separate your CSS from your JavaScript from your HTML. We all do that, right? CSS goes into its own file; JavaScript goes in another; HTML is left by itself, nice and clean.

CSS Zen Garden proved that we can alter a design into a myriad of permutations simply by changing the CSS. However, we’ve rarely seen the flip side of this — the side that is more likely to occur in a project: the HTML changes. We modify the HTML and then have to go back and revise any CSS that goes with it.

In this way, we haven’t really separated the two, have we? We have to make our changes in two places.

Exploring Approaches

Over the course of my career, I’ve had the pleasure and privilege to work on hundreds of different websites and Web applications. For the vast majority of these projects, I was the sole developer building out the HTML and CSS. I developed a way of coding websites that worked well for me.

Most recently, I spent two years at Yahoo working on Mail, Messenger, Calendar and other projects. Working on a much larger project with a much larger team was a great experience. A small team of prototypers worked with a larger team of designers to build out all of the HTML and CSS for multiple teams of engineers.

It was the largest-scale project I had worked on in many aspects:

• Yahoo’s user base is massive. Mail alone has about 300 million users.
• Hundreds of people spread across multiple teams were working with the HTML and CSS.
• We were developing a system of components to work across multiple projects.

It was during my time at Yahoo that I began to really examine how I and the team at Yahoo build websites. What pain points did we keep running into, and how could we avoid them?

I looked to see what everyone else was doing. I looked at Nicole Sullivan’s Object-Oriented CSS, Jina Bolton’s presentation on “CSS Workflow” and Natalie Downe’s “Practical, Maintainable CSS,” to name just a few.

I ended up writing my thoughts as a long-form style guide named “Scalable and Modular Architecture for CSS.” That sounds wordy, so you can just call it SMACSS (pronounced “smacks”) for short. It’s a guide that continues to evolve as I refine and expand on ways to approach CSS development.

As a result of this exploration, I’ve noticed that designers (including me) traditionally write CSS that is deeply tied to the HTML that it is designed to style. How do we begin to decouple the two for more flexible development with less refactoring?

In other words, how do we avoid throwing !important at everything or falling into selector hell?

Reusing Styles

In the old days, we wrapped font tags and applied background attributes to every HTML element that needed styling. This was, of course, very impractical, and thus CSS was born. CSS enabled us to reuse styles from one part of the page on another.

For example, a navigation menu has a list of items that all look the same. Repeating inline styles on each item wouldn’t be practical. As a result, we begin to see CSS like this:

#nav {
   margin: 0;
   padding: 0;
   list-style: none;
}

#nav li {
   float: left;
}

#nav li a {
   display: block;
   padding: 5px 10px;
   background-color: blue;
}

Sure beats adding float:left to every list item and display:block; padding:5px 10px; to every link. Efficiency, for the win! Just looking at this, you can see the HTML structure that is expected:

<ul id="nav">
   <li><a href="/">Home</a></li>
   <li><a href="/products">Products</a></li>
   <li><a href="/contact">Contact Us</a></li>
</ul>

Now, the client comes back and says, “I want a drop-down menu to appear when the user clicks on ‘Products.’ Give them easy access to each of the pages!” As a result, our HTML changes.

<ul id="nav">
   <li><a href="/">Home</a></li>
   <li><a href="/products">Products</a>
      <ul>
         <li><a href="/products/shoes">Shoes</a></li>
         <li><a href="/products/jackets">Jackets</a></li>
      </ul>
   </li>
   <li><a href="/contact">Contact Us</a></li>
</ul>

We now have a list item within a list item, and links within it. Our menu has a horizontal navigation when the client wants a vertical list, so we add some rules to style the inner list to match what the client wants.

#nav ul {
   margin: 0;
   padding:0;
   list-style:none;
}

#nav li li {
   float: none;
}

#nav li li a {
   padding: 2px;
   background-color: red;
}

Problem solved! Sort of.

Reducing The Depth Of Applicability

Probably the most common problem with CSS is managing specificity. Multiple CSS rules compete in styling a particular element on the page. With our menu, our initial rules were styling the list items and the links in the navigation and the menu. Not good.

By adding more element selectors, we were able to increase specificity and have our menu styles win out over the navigation styles.

However, this can become a game of cat and mouse as a project increases in complexity. Instead, we should be limiting the impact of CSS. Navigation styles should apply to and affect only the elements that pertain to it, and menu styles should apply to and affect only the elements that pertain to it.

I refer to this impact in SMACSS as the “depth of applicability.” It’s the depth at which a particular rule set impacts the elements around it. For example, a style like #nav li a, when given an HTML structure that includes our menus, has a depth of 5: from the ul to the li to the ul to the li to the a.

The deeper the level of applicability, the more impact the styles can have on the HTML and the more tightly coupled the HTML is to the CSS.

The goal of more manageable CSS — especially in larger projects — is to limit the depth of applicability. In other words, write CSS to affect only the elements that we want them to affect.

Child Selectors

One tool for limiting the scope of CSS is the child selector (>). If you no longer have to worry about Internet Explorer 6 (and, thankfully, many of us don’t), then the child selector should be a regular part of your CSS diet.

Child selectors limit the scope of selectors. Going back to our navigation example, we can use the child selector to limit the scope of the navigation so that it does not affect the menu.

#nav {
   margin: 0;
   padding: 0;
   list-style: none;
}

#nav > li {
   float: left;
}

#nav > li > a {
   display: block;
   padding: 5px 10px;
   background-color: blue;
}

For the menu, let’s add a class name to it. This will make it more descriptive and provide a base for the rest of our styles.

.menu {
   margin: 0;
   padding: 0;
   list-style: none
}

.menu > li > a {
   display: block;
   padding: 2px;
   background-color: red;
}

What we’ve done is limited the scope of our CSS and isolated two visual patterns into separate blocks of CSS: one for our navigation list and one for our menu list. We’ve taken a small step towards modularizing our code and a step towards decoupling the HTML from the CSS.

Classifying Code

Limiting the depth of applicability helps to minimize the impact that a style might have on a set of elements much deeper in the HTML. However, the other problem is that as soon as we use an element selector in our CSS, we are depending on that HTML structure never to change. In the case of our navigation and menu, it’s always a list with a bunch of list items, with a link inside each of those. There’s no flexibility to these modules.

Let’s look at an example of something else we see in many designs: a box with a heading and a block of content after it.

<div class="box">
   <h2>Sites I Like</h2>
   <ul>
      <li><a href="http://smashingmagazine.com/">Smashing Magazine</a></li>
      <li><a href="http://smacss.com/">SMACSS</a></li>
   </ul>
</div>

Let’s give it some styles.

.box {
   border: 1px solid #333;
}

.box h2 {
   margin: 0;
   padding: 5px 10px;
   border-bottom: 1px solid #333;
   background-color: #CCC;
}

.box ul {
   margin: 10px;
}

The client comes back and says, “That box is great, but can you add another one with a little blurb about the website?”

<div class="box">
   <h2>About the Site</h2>
   <p>This is my blog where I talk about only the bestest things in the whole wide world.</p>
</div>

In the previous example, a margin was given to the list to make it line up with the heading above it. With the new code example, we need to give it similar styling.

.box ul, .box p {
   margin: 10px;
}

That’ll do, assuming that the content never changes. However, the point of this exercise is to recognize that websites can and do change. Therefore, we have to be proactive and recognize that there are alternative elements we might want to use.

For greater flexibility, let’s use classes to define the different parts of this module:

.box .hd { }  /* this is our box heading */
.box .bd { }  /* this is our box body */

When applied to the HTML, it looks like this:

<div class="box">
   <h2 class="hd">About the Site</h2>
   <p class="bd">This is my blog where I talk about only the bestest things in the whole wide world.</p>
</div>

Clarifying Intent

Different elements on the page could have a heading and a body. They’re “protected” in that they’re a child selector of box. But this isn’t always as evident when we’re looking at the HTML. We should clarify that these particular hd and bd classes are for the box module.

.box .box-hd {}
.box .box-bd {}

With this improved naming convention, we don’t need to combine the selectors anymore in an attempt to namespace our CSS. Our final CSS looks like this:

.box {
border: 1px solid #333;
}

.box-hd {
margin: 0;
padding: 5px 10px;
border-bottom: 1px solid #333;
background-color: #CCC;
}

.box-bd {
   margin: 10px;
}

The bonus of this is that each of these rules affects only a single element: the element that the class is applied to. Our CSS is easier to read and easier to debug, and it’s clear what belongs to what and what it does.

It’s All Coming Undone

We’ve just seen two ways to decouple HTML from CSS:

1. Using child selectors,
2. Using class selectors.

In addition, we’ve seen how naming conventions allow for clearer, faster, simpler and more understandable code.

These are just a couple of the concepts that I cover in “Scalable and Modular Architecture for CSS,” and I invite you to read more.

Postscript

In addition to the resources linked to above, you may wish to look into BEM, an alternative approach to and framework for building maintainable CSS. Mark Otto has also been documenting the development of Twitter Bootstrap, including the recent article “Stop the Cascade,” which similarly discusses the need to limit the scope of styles.

(al) (il)

© Jonathan Snook for Smashing Magazine, 2012.