Advertisements

User Controled Color Theme

Works on Desktop Safari, desktop Google Chrome, desktop Firefox 3.6-4, iPhone, iPod Touch, iPad.

So, in the last blog post I showed how to make RGB slides with HTML, CSS and some JavaScript. I thought about it and, while interesting, it doesn’t have a whole lot of practical application. Sure you could take that and hook up any other type of value to get whatever result you might need for your interface. Well that got me to thinking, so I threw together an implementation of the RGB sliders that allow a user to change the color scheme of a Web app. No need to spend time creating different color themes. Let the user do it.

OK, before you think I’m crazy, especially you folks from the design community, let me explain. I came up with a basic theme technique. I call it chromaeleon &mdash because the app’s chrome can change colors like a chamaeleon. The way this works is, instead of solid color gradients, you create gradients with transparent values of black and white. Behind this you have a background color which shows through the transparent gradients. This way, when the user drags the sliders, the background colors update and the look of the interface changes. Now in the real world you’d want to provide a way for the user to save their color choice. You could save the choice to localStorage. Then when the app loads, it checks to see it the user saved a color choice, if not, it goes to the default. Sorry, I didn’t do all of that. Just the part to update the background colors. Here’s what it will look like:

iPhone Chromaeleon Interface

The structure we’re going to use is pretty must standard as we’ve used elsewhere, a header, a section, some buttons.

<body>
	<header>
		<a href="http://css3wizardry.com" class="button back"><span class="pointer"></span><span>Back</span></a>
		<h1>Chromaeleon Theme</h1>
		<span class="button">Click Here</span>
	</header>
	<section>
		<h2>Use the sliders to adjust the colors of the theme.</h2>
		<div class="colorRow">
			<div id="redSlider" class="slider">
				<div class="thumb"></div>
			</div>
			<div id="redColor" class="colorOutput"></div>
			<span> Red</span>
		</div>
		<div class="colorRow">
			<div id="greenSlider" class="slider">
				<div class="thumb"></div>
			</div>
			<div id="greenColor" class="colorOutput"></div>
			<span> Green</span>
		</div>
		<div class="colorRow">
			<div id="blueSlider" class="slider">
				<div class="thumb"></div>
			</div>
			<div id="blueColor" class="colorOutput"></div>
			<span> Blue</span>
		</div>
		<div class="colorRow finalResult">
			<span>Final Color: </span>
			<div id="rgbColor" class="colorOutput"></div>
			<br />
			<span>RGB: </span><span id="rgbResult">0, 0, 0</span>
			<br />
			<span>HEX: </span><span id="hexResult">#000000</span>
		</div>
	</section>
</body>

So, for the header and the buttons, we need to change their default gradients, as I mentioned above, to have RGBA transparency values. This is my basic gradient:

background-image: 
	-moz-linear-gradient(top, 
		rgba(255,255,255,.5), 
		rgba(30,30,30,.65) 50%, 
		rgba(0,0,0,.7) 50%, 
		rgba(0,0,0,.8)); 
background-image: 
	-webkit-gradient(linear, left top, left bottom, 
		from(rgba(255,255,255,.5)), 
		color-stop(0.5,rgba(30,30,30,.65)), 
		color-stop(0.5, rgba(0,0,0,.7)), 
		to(rgba(0,0,0,.8)));

And for the hover state of the button, we use this gradient:

background-image: 
	-webkit-gradient(linear, left top, left bottom, 
		from(rgba(0,0,0,.1)), 
		color-stop(0.5,rgba(0,0,0,.5)), 
		color-stop(0.5, rgba(0,0,0,.6)), 
		to(rgba(255,255,255,.2)));
background-image: 
	-moz-linear-gradient(top,
		rgba(0,0,0,.1), 
		rgba(0,0,0,.5) 50%, 
		rgba(0,0,0,.6) 50%, 
		rgba(255,255,255,.2));

Now to change the color, all we need to do is introduce a new method to our existing code:

/**
*
* Method to update chrome colors according to the RGB value of the sliders.
*
*/
$.updateInterfaceColors = function() {
	$("header").css("background-color: rgb(" + $.rgbColor[0] + "," + $.rgbColor[1] + "," + $.rgbColor[2] + ")");
	$$(".button").forEach(function(button) {
		button.css("background-color: rgb(" + $.rgbColor[0] + "," + $.rgbColor[1] + "," + $.rgbColor[2] + ")");
	});
	$("section").css("background-color: rgb(" + $.rgbColor[0] + "," + $.rgbColor[1] + "," + $.rgbColor[2] + ")");
	$(".pointer").css("background-color: rgb(" + $.rgbColor[0] + "," + $.rgbColor[1] + "," + $.rgbColor[2] + ")");
};

To execute this method, we invoke it in the slide mouse event handlers for each slider, and for touch-based mobile devices we invoke it in the updateSliderTouch method:

/**
*
* This is for the red slider's mouse interaction, you'd do the same for the green and blue sliders' setup scripts as well.
*/
// Set up three sliders for Red, Green and Blue:
$.slider("#redSlider", { 
	onDrag : function() {
		$("#redSlider").setColorFromSlider("red");
		$.updateInterfaceColors();
	},
	// onDragEnd function necessary to remove hover state off of slider thumb when drag ends.
	onDragEnd : function() {},
	top : -6
});
/**
*
* This is for touch-enabled devices. You invoke the $.updateInterfaceColors() method just once inside the updateSliderTouch method's definition, at the very end.
*/
Element.prototype.updateSliderTouch = function( color ) {
	this.style.left =  curX + 'px'; 
	if (color === "red") {
		$("#" + color + "Color").css("background-color: rgb(" + curX +",0,0)");
		$.rgbColor[0] = curX;
	}
	if (color === "green") {
		$("#" + color + "Color").css("background-color: rgb(0," + curX +",0)");
		$.rgbColor[1] = curX;
	}
	if (color === "blue") {
		$("#" + color + "Color").css("background-color: rgb(0,0," + curX +")");
		$.rgbColor[2] = curX;
	}
	$("#" + color + "Slider").css("-webkit-background-size:" + (curX + 1) + "px 9px, 100% 9px");
	$("#" + color + "Slider").css("background-size:" + (curX + 1) + "px 9px, 100% 9px");
	$("#rgbColor").css("background-color: rgb(" + $.rgbColor[0] + "," + $.rgbColor[1] + "," + $.rgbColor[2] + ")");
	$("#rgbResult").fill($.rgbColor[0] + ", " + $.rgbColor[1] + ", " + $.rgbColor[2]);
	$("#hexResult").fill("#" + $.rgb2hex($.rgbColor[0]) + $.rgb2hex($.rgbColor[1]) + $.rgb2hex($.rgbColor[2]));
	$.updateInterfaceColors();
};

This works great on desktop Safari, Chrome and even Firefox (Yay!), and fine on the iPad. For iPod Touch or iPhone you need to load it in portrait mode. it’s a bit cramped due to the size of the sliders. I needed them to be at least 255px long for the RGB values, and then borders, box shadows and the extra space for the thumbs made them barely fit in the iPhone’s and iPod Touch’s default width. Try hitting the plus icon at the bottom of the browser and save it to you device’s desktop. Then run it from there, you’ll have more vertical space. You can try this out online or download the source code. Enjoy!

Advertisements

Range Slider with CSS and JavaScript

Works on desktop Safari, desktop Chrome, desktop Firefox 3.5 – 4, iPhone, iPod Touch, iPad.

In this post I’m going to show how to make a range slider using HTML, CSS and JavaScript that works with both a mouse and a finger. The mouse-enabled version required a small drag-and-drop JavaScript framework. Fortunately I had already put that together several years back. After spending some time playing around with touch events on mobile Webkit, I was able to come up with a way to implement horizontal dragging for the range slider.
range slider

I’m not going to go into details about my mouse-enable drag-and-drop framework. You can popup it open and read the copious comments in the example. However, I will explain how I implemented the touch-enabled drag for the slider.

The structure for a slider is fairly straightforward. You need a track and a thumb:

<div id="redSlider" class="slider">
	<div class="thumb"></div>
</div>

Since the structure is so simple, you might be wondering how we give it the look. The thumb gets border radius to make it round, along with a box shadow and a background-gradient, including a blue background gradient for hover. The slider track is styled with two background gradients, the bottom-most gradient is the default grey which swans the width of the slider. Layered on top of the same track is a second, bluish gradient. By using CSS3’s background sizing property, we will dynamically resize it as the slider’s thumb is dragged.

Of course, just dragging a slider thumb back in forth is not suck a big deal. I therefore created three sliders implemented as RGB pickers. By dragging each thumb, you add or subtract from a red, green or blue value. Down below you’ll see the final RGB and Hex values.
RGB Slider

Here are the basic styles for the slider:

.slider {
	display: inline-block;
	-webkit-box-sizing: border-box;
	-moz-box-sizing: padding-box;
	box-sizing: padding-box;
	-webkit-box-shadow: 2px 2px 4px #666;
	-moz-box-shadow: 2px 2px 4px #666;
	box-shadow: 2px 2px 4px #666;
	height: 9px;
	width: 277px; 
	padding: 1px;
	-webkit-border-radius: 4px;
	-moz-border-radius: 4px;
	border-radius: 4px;
	background-image: 
		-webkit-gradient(linear, left top ,left bottom,
		   from(#0a3a86),
		   color-stop(.5, #4c8de7),
		   color-stop(.95, #6babf5),
		   to(#0a3a86)),
		-webkit-gradient(linear, left top ,left bottom,
			from(#919191),
			color-stop(.5, #f0f0f0),
			color-stop(.5, #fff),
			color-stop(.95, #fff),
			to(#919191));
	background-image: 
		-moz-linear-gradient(top,
		   #0a3a86,
		   #4c8de7 50%,
		   #6babf5 95%,
		   #0a3a86),
		-moz-linear-gradient(top,
			#919191,
			#f0f0f0 50%,
			#fff 50%,
			#fff 95%,
			#919191);
	background-repeat: no-repeat, repeat-x;
}
.thumb {
	position:relative;
	-webkit-box-shadow: 2px 2px 3px #666;
	-moz-box-shadow: 2px 2px 3px #666;
	box-shadow: 2px 2px 3px #666;
	height:20px;
	width:20px;
	left: 0px; 
	top: -6px;
	-webkit-border-radius: 10px;
	-moz-border-radius: 10px;
	border-radius: 10px;
	background-image: 
	   -webkit-gradient(linear, left top, left bottom,
		   from(#aaa),
		   color-stop(.5, #ddd),
		   to(#ccc));
	background-image: 
	   -moz-linear-gradient(top,
		   #aaa,
		   #ddd 50%,
		   #ccc);
	cursor: move;
	-webkit-tap-highlight-color: transparent;
}
.thumb:hover, .thumb.hover {
	background-image: 
	   -webkit-gradient(linear, left top, left bottom,
		   from(#6297f2),
		   color-stop(.5, #0251ae),
		   to(#6297f2));
	background-image: 
	   -moz-linear-gradient(top,
		   #6297f2,
		   #0251ae 50%,
		   #6297f2);
}

Notice the slider’s background gradient style. The first background gradient will be the top-most. The last will be the bottom-most. But the top-most is going to be the blue part of the track that appears to the left of the thumb as it is dragged away from the left start of the range slider.

We also need some styles to set the initial states of the three thumbs. Notice that I’ve used background sizing to control the two background gradients. The first is for the blue top-most gradient, the second is for the full width grey gradient.

#redSlider .thumb {
	left: 121px;
}
#redSlider {
	-webkit-background-size: 123px 9px, 100% 9px;
	-moz-background-size: 123px 9px, 100% 9px;
	background-size: 123px 9px, 100% 9px;
}
#greenSlider .thumb {
	left: 156px;
}
#greenSlider {
	-webkit-background-size: 158px 9px, 100% 9px;
	-moz-background-size: 158px 9px, 100% 9px;
	background-size: 158px 9px, 100% 9px;
}
#blueSlider .thumb {
	left: 230px;
}
#blueSlider {
	-webkit-background-size: 232px 9px, 100% 9px;
	-moz-background-size: 232px 9px, 100% 9px;
	background-size: 232px 9px, 100% 9px;
}

So, I’ve defined two gradients with different background repeats: background-repeat: no-repeat, repeat-x; and background sizing with values such as: 123px 9px, 100% 9px. 9px is the height of the slider track. The bottom-most gradient has a width of 100%, and the top-most bluish one gets a width of 123px. By using these values, with very little markup, we can create visually and functionally complex structures.

/**
* Touch enabled support:
*/
/**
*
* Method to set the colors of color swatches and width of the slider progress track when the slider thumb is dragged.
*/
Element.prototype.setupSliderTouch = function( event ) {
	event.preventDefault();
	var el = event.target;
	var touch = event.touches[0];
	curX = touch.pageX - this.parentNode.offsetLeft;
	if (curX <= 0) { 
		curX = 0;
	}
	if (curX > 255) {
		curX = 255;
	}
};
Element.prototype.updateSliderTouch = function( color ) {
	this.style.left =  curX + 'px'; 
	if (color === "red") {
		$("#" + color + "Color").css("background-color: rgb(" + curX +",0,0)");
		$.rgbColor[0] = curX;
	}
	if (color === "green") {
		$("#" + color + "Color").css("background-color: rgb(0," + curX +",0)");
		$.rgbColor[1] = curX;
	}
	if (color === "blue") {
		$("#" + color + "Color").css("background-color: rgb(0,0," + curX +")");
		$.rgbColor[2] = curX;
	}
	
	$("#" + color + "Slider").css("-webkit-background-size:" + (curX + 1) + "px 9px, 100% 9px");
	$("#" + color + "Slider").css("background-size:" + (curX + 1) + "px 9px, 100% 9px");
	
	$("#rgbColor").css("background-color: rgb(" + $.rgbColor[0] + "," + $.rgbColor[1] + "," + $.rgbColor[2] + ")");
	$("#rgbResult").fill($.rgbColor[0] + ", " + $.rgbColor[1] + ", " + $.rgbColor[2]);
	$("#hexResult").fill("#" + $.rgb2hex($.rgbColor[0]) + $.rgb2hex($.rgbColor[1]) + $.rgb2hex($.rgbColor[2]));
};

$("#redSlider > .thumb").bind('touchmove', function(event) {
	this.setupSliderTouch(event);
	this.updateSliderTouch("red");
});
$("#greenSlider > .thumb").bind('touchmove', function(event) {
	this.setupSliderTouch(event);
	this.updateSliderTouch("green");
});
$("#blueSlider > .thumb").bind('touchmove', function(event) {
	this.setupSliderTouch(event);
	this.updateSliderTouch("blue");
});

Basically, I attach a touchmove event to the slider thumbs. The event listener passes the event to the setupSliderTouch method. The first thing the setupSliderTouch method does is to prevent the default interaction from taking place, such as page scrolling. We want the user to be able to move the thumb without scrolling the page. From the event passed in to setupSliderTouch we get the touch event and calculate its x coordinate on the screen. To calculate the touch’s position in relation to the slider, we subtract the left offset of the slider from the pageX of the touch. This gives us the left-most edge of the slider’s thumb. We store this as curX. We check the value of curX. If it is less than zero, we set it back to zero. We do this because this value will be used to set the position of the thumb and one of the RGB values. We don’t want either the thumb being dragged off of the left edge of the slider, nor a value less than zero, since RGB values start at zero. We do the same thing when the curX value is greater than 255 for the same reasons.

The updateSliderTouch method uses the value of the slider’s thumb to calculate and update RGB and Hex values, giving the user visual feedback as the thumb is dragged. Then we use the value of curX to update the background size of the blues background gradient on the slider track:

$("#" + color + "Slider").css("-webkit-background-size:" + (curX + 1) + "px 9px, 100% 9px");
$("#" + color + "Slider").css("background-size:" + (curX + 1) + "px 9px, 100% 9px");

That’s all there is to it. You can try this out online using desktop Safari, Chrome, or Firefox for the mouse version, or on an iPhone, iPod Touch or iPad for the touch version. Or you can download the source code, which I recommend, so you can dig into the CSS and JavaScirpt.

Update: September 15, 2010
If you’re trying this out on an iPhone, I noticed that there seems to be a very slight delay before an initial touch is registered on the screen. This means that in order to slide the thumb, you need to press and hold for a very brief moment before sliding, otherwise no touch gets registered and nothing happens. This doesn’t seem to happen when performing the same action on the iPad. Touches seem more responsive.

Activity Indicator with CSS3

Works on desktop Safari, desktop Google Chrome, Android, iPhone, iPod Touch and iPad.

You’ve seen them. The things on the screen that mean you’ve got to wait for some activity to complete. They’re called activity indicators. They’re usually several pieces arranged in a circle. They spin around. And if things go really bad, you could be watching them spin for quite some time. Hopefully not.
activity indicator

Usually people make these with images. There are even Web sites dedicated to making this spinning baubles for you in animated gif format. Except that gif stink. Pngs look better. I’m going to show you two ways to implement an activity indicator. One starts off with an image that has been converted into 64 bit data. We’re talking about dataurls here. Dataurls eliminate the need for separate images. It’s the same image but reduced to data which you can paste directly in your document. Heck, you can paste it directly into your CSS file so you never have to worry about losing an image or rewriting the image path when you move the CSS. Cool, huh? And if you start with a large image, you can scale it down for small screens and low resolution devices, and scale it up for hi res devices, like the iPhone and iPod Touch retina display. Cooler still, you can make the activity indicator completely out of HTML and CSS, no images, no dataurl. And by using a scale transform, you never have to worry about resolution. It will always render smoothly.

So, let’s look at dataurls. The concept is simple. Instead of having a resource reside as an external file, you convert it into a 64 bit data sample which you can include in your document. In the case of an image, it’s basically embedding it in your HTML/CSS. You can encode images as dataurls in many ways, using PHP, Python, etc., or you can simply upload a file to a Web site that will conveniently do it for you. A dataurl looks something like this (Note: I trimmed off the data to fit here):

data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAATEAAAEwCA+

For an image, you would do something like this:

<img src="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAATEAAAEwCA+...">

When you create your first dataurl you’ll probably be shocked at how big the resulting code is. Don’t worry. Somehow this winds up being more efficient for download and rendering than a traditional image. Besides being the source for an inline image, you could also embed the dataurl into your CSS as a background image. As I mentioned, my original image was large to allow for scaling up or down without loss of quality. I therefore add in some resizing to the CSS for the size of the element and the size of the background image:

.activityIndicator {
	/*
	The original height is:
	height: 304px;
	width: 305px;
	*/
	height: 40px;
	width: 40px;
	-webkit-background-size: 40px 40px;
	margin: 0px auto;
	background-image: url("data:image/png;base64,iVBORw0KGgoAAAANUb...");

This is what I used to produce the image above. Now we need to animate it. That’s actually quite easy, just a couple of lines of CSS:

.activityIndicator {
	/*
	The original height is:
	height: 304px;
	width: 305px;
	*/
	height: 40px;
	width: 40px;
	-webkit-background-size: 40px 40px;
	margin: 0px auto;
	background-image: url("data:image/png;base64,iVBORw0KGgoAAAANUb...");
	-webkit-animation-duration: 1s;
	-webkit-animation-iteration-count: infinite;
	-webkit-animation-timing-function: linear;
	-webkit-animation-name: spinnerAnim;
}
@-webkit-keyframes spinnerAnim {
	0% { -webkit-transform: rotate(0deg); }
	100% { -webkit-transform: rotate(360deg); }
}

By using the keyframe animation, we can implement the continuously spinning activity indicator without using JavaScript. But, as I mentioned, dataurls tend to be verbose and are rather unsightly. I therefore came up with a reproduction of the activity indicator using only HTML and CSS3. At the end of this post are links to the working file. Download or view source of the online document to see what the dataurl looks like. It’s not pretty, but it is efficient. However, by reproducing the same image using markup and CSS, we can achieve the same result with even less code. The markup isn’t terribly complex, two parent contains and twelve blades:

<div id="activityIndicator">
	<div>
		<div class="blade"></div>
		<div class="blade"></div>
		<div class="blade"></div>
		<div class="blade"></div>
		<div class="blade"></div>
		<div class="blade"></div>
		<div class="blade"></div>
		<div class="blade"></div>
		<div class="blade"></div>
		<div class="blade"></div>
		<div class="blade"></div>
		<div class="blade"></div>
	</div>
</div>

That’s really all we need, the rest is done with CSS3 properties. You’ll notice that all the blades have the same class “blade.” We don’t need to give each one anything special because we can use CSS3 selectors to indicate each blade individually. To do this we’ll use the :nth-child selector. We first give all the blades some generic styling. Then we start from the second blade using the CSS3 nth-child selector: .blade:nth-child(2), then .blade:nth-child(3), etc. Since there are twelve blades, we need to rotate each one 30 degrees more than the previous. We also need to change the color by approximately 21.25% to go from and rgb value of 0 to 255. Since rgb values expect positive integers between 0 and 255, we need to round the float off to the nearest whole number. As you can see, CSS3 selectors allow us to create complex structures with minimal markup and CSS.

#activityIndicator {
	position: relative;
	width: 130px;
	height: 130px;
	margin: 0 auto;
	-webkit-perspective: 500;
	-webkit-animation-duration: 1s;
	-webkit-animation-iteration-count: infinite;
	-webkit-animation-timing-function: linear;
	-webkit-animation-name: spinnerAnim2;
}
@-webkit-keyframes spinnerAnim2 {
	0% { -webkit-transform: rotate(0deg) scale(.29); }
	100% { -webkit-transform: rotate(360deg) scale(.29); }
}
#activityIndicator > div:first-of-type {
	margin-left: 58px;
	margin-top 0;
	width: 50%;
	height: 50%;
}
#activityIndicator .blade {
	position: absolute;
	height: 40px;
	width: 14px;
	background-color: rgba(234,234,234, .30);
	-webkit-border-radius: 10px;
	-webkit-transform-origin-x: 50%;
	-webkit-transform-origin-y: 165%;
}

#activityIndicator .blade:nth-child(2) {
	-webkit-transform: rotate(30deg);
	background-color: rgba(212,212,212, .70);
}
#activityIndicator .blade:nth-child(3) {
	-webkit-transform: rotate(60deg);
	background-color: rgba(191,191,191, .70);
}
#activityIndicator .blade:nth-child(4) {
	-webkit-transform: rotate(90deg);
	background-color: rgba(170,170,170, .70);
}
#activityIndicator .blade:nth-child(5) {
	-webkit-transform: rotate(120deg);
	background-color: rgba(149,149,149, .70);
}
#activityIndicator .blade:nth-child(6) {
	-webkit-transform: rotate(150deg);
	background-color: rgba(128,128,128, .70);
}
#activityIndicator .blade:nth-child(7) {
	-webkit-transform: rotate(180deg);
	background-color: rgba(106,106,106, .70);
}
#activityIndicator .blade:nth-child(8) {
	-webkit-transform: rotate(210deg);
	background-color: rgba(85,85,85, .70);
}
#activityIndicator .blade:nth-child(9) {
	-webkit-transform: rotate(240deg);
	background-color: rgba(64,64,64, .70);
}
#activityIndicator .blade:nth-child(10) {
	-webkit-transform: rotate(270deg);
	background-color: rgba(42,42,42, .70);
}
#activityIndicator .blade:nth-child(11) {
	-webkit-transform: rotate(300deg);
	background-color: rgba(21,21,21, .70);
}
#activityIndicator .blade:nth-child(12) {
	-webkit-transform: rotate(330deg);
	background-color: rgba(0,0,0, .70);
}

You can try this out online or download the source code.

Update: September 14th, 2010
I forgot to mention one thing. If you look at the styles on #activityIndicator .blade you’ll notice the last two property definitions:

	-webkit-transform-origin-x: 50%;
	-webkit-transform-origin-y: 165%;

By setting the transform origin x value to 50% we fix the horizontal rotation to the blade’s center. By setting the transform origin vertical value to 165% we define the turning point at that distance from the start of the blade. Together these values cause the blades to rotate around leaving and empty circular space in the center, thus reproducing the appearance of the png image.

iPhone Modal Popup with HTML5, CSS3 & JavaScript

Works on Desktop Safari, Desktop Google Chrome, iPhone, iPod Touch, iPad. Note that I’ve included some styling for Firefox, even though it has no presence to speak of in the mobile space. In particular, Firefox 4 beta still lacks support for CSS3 keyframe animation, although that will make it into a later update.

If you’ve used an iPhone, iPod Touch or iPad, then you’re familiar with the modal popup dialog boxes that the native system uses. Here’s a typical iPhone popup:
Native iPhone modal popup

Notice the white radial gradient behind the popup. I was able to replicate this, but when the user was on a long document and scrolled down to do something that would trigger a popup, I could find no way to center that radial gradient based on the vertical page scroll. I therefore went with a whitesh blur around the popup itself using a CSS3 box shadow. Here’s what my HTML5/CSS3 version looks like:

Originally I thought I would use just one popup per app, re-assigning values to the popup’s part each time the popup was invoked. However I ran into the problem of events from different and I failed to find an elegant way to resolve this. I therefore came up with a scheme where you initialize a popup at the view level, allowing each view to have a custom popup. The initializing script creates the popup and injects it as the last child of the view. The setup script creates the markup for the popup and populates it with values passed as an argument to the initializing script. The setup script also adds basic functionality to the buttons so that clicking either of them will close the popup. The setup script also creates a screen cover which traps events to prevent user interaction with what is behind the popup until it is closed.

The setup script accepts a single argument—an object literal containing key/values pairs to populate the popup. In order for the setup script to create a popup, you must at least pass a value for a valid view in your Web app. This would be like selector: "#Popup". If no other values are passed, the script will produce a basic popup that looks like this:
Basic popup

I used the ChocolateChip mobile JavaScript library to add the interactive functionality to the popup. Here’s the JavaScript that creates the markup and functionality for the popup:

/** 
* 
* A method to initialize a modal popup. By passing a valid selector for a view, this method creates a view based popup with the properties supplied by the options argument. It automatically binds events to both popup buttons to close the popup when the user clicks either. If a callback is passed as part of the opts argument, it gets bound to the "Continue" button automatically.
*
* @method
* 
* ### setupPopup
*
* syntax:
*
*  $.setupPopup({selector: "#News", title: "Subscribe", cancel: });
*
* arguments:
* 
*  - string: string A valid selector for the parent of the tab control. By default the an object literal.
*  - string: string An object literal which can have the following properties:
	title: a string defining the title in the popup.
	message: a string defining the popup message.
	cancelButton: a string defining an alternate name for the cancel button.
	continueButton: a string defining an alternate name for the confirm button.
	callback: a function to run when the user touches the confirm button.
	If no title is supplied, it defaults to "Alert!".
	If no cancelButton value is supplied, it defaults to "Cancel".
	If no continueButton value is supplied, it defaults to "Continue".
* example:
*
*  $.setupPopup({selector: "#buyerOptions"});
*  $.setupPopup({
		selector: "#Popup",
		title: 'Attention Viewers!', 
		message: 'This is a message from the sponsors. Please be seated while we are getting ready. Thank you for your patience.', 
		cancelButton: 'Skip', 
		continueButton: 'Stay for it', 
		callback: function() {
			$('#popupMessageTarget').fill('Thanks for staying with us a bit longer.');
			$('#popupMessageTarget').removeClass("animatePopupMessage");
			$('#popupMessageTarget').addClass("animatePopupMessage");
		}
	});
*
*/
$.setupPopup = function( opts ) {
	if (opts.selector) {
		var selector = opts.selector;
	} else {
		return false;
	}
	var title = "Alert!";
	if (opts.title) {
		var title = opts.title;
	}
	var message = "";
	if (opts.message) {
		var message = opts.message;
	}
	var cancelButton = "Cancel";
	if (opts.cancelButton) {
		cancelButton = opts.cancelButton;
	}
	var continueButton = "Continue";
	if (opts.continueButton) {
		continueButton = opts.continueButton;
	}
	var popup = '<div class="screenCover hidden"></div>';
	popup += '<section class="popup hidden"><div>';
	popup += '<header><h1>' + title + '</h1></header>';
	popup += '<p>' + message +'</p><footer>';
	popup += '<div class="button cancel">' + cancelButton + '</div>';
	popup += '<div class="button continue">' + continueButton + '</div></footer></div></section>';
	$(selector).insertAdjacentHTML("beforeEnd", popup);
	// Bind event to close popup when either button is clicked.
	$$(selector + " .button").forEach(function(button) {
		button.bind("click", function() {
			$(selector + " .screenCover").addClass("hidden");
			$(selector + " .popup").addClass("hidden");
		});
	});
	
	if (opts.callback) {
		var callbackSelector = selector + " .popup .continue";
		$(callbackSelector).bind("click", function() {
			opts.callback();
		});
	}
	
};

And here is an initialization of a popup:

$.setupPopup(
	{
		selector: "#Popup",
		title: 'Attention Viewers!', 
		message: 'This is a message from the sponsors. Please be seated while we are getting ready. Thank you for your patience.', 
		cancelButton: 'Skip', 
		continueButton: 'Stay for it', 
		callback: function() {
			$('#popupMessageTarget').fill('Thanks for staying with us a bit longer.');
                        // Remove this class in case the popup was opened previously.
			$('#popupMessageTarget').removeClass("animatePopupMessage");
                        // Then add the class to trigger an animation of the message being displayed.
			$('#popupMessageTarget').addClass("animatePopupMessage");
		}
	}
);

Now that a popup has been created and populated with the desired values, we need a way to show it. Before actually showing the popup, the $.showPopup method display a screen cover which captures user interaction and thereby prevents the interface behind the popup from being accessed until the popup is dispelled. The showPopup method accepts one argument, a selector indicating a uniquely identifiable node that contains the popup as a descendant.

$.showPopup = function( selector ) {
	var screenCover = $(selector + " .screenCover");
        // Make the screen cover extend the entire width of the document, even if it extends beyond the viewport.
	screenCover.css("height:" + (window.innerHeight + window.pageYOffset) + "px");
	var popup = $(selector + " .popup");
	$(selector + " .popup").style.top = ((window.innerHeight /2) + window.pageYOffset) - (popup.clientHeight /2) + "px";
	$(selector + " .popup").style.left = (window.innerWidth / 2) - (popup.clientWidth / 2) + "px";
	$(selector + " .screenCover").removeClass("hidden");
	$(selector + " .popup").removeClass("hidden");
};

With this method defined we can now show the popup as need. Here’s a script that attaches an event handler to a button with a class of “openPopup” for a popup somewhere among the descendant nodes of a node with an id of “Tabs”:

$("#Tabs .openPopup").bind("click", function() {
	$.showPopup("#Tabs");
});

OK, so we have the markup and functionality for the popup, but we don’t have the look. We’ll take care of that next. In order to create the unique look of the iPhone popup, I use several layers for encasing borders and composited transparent background gradients. Originally I had two gradients, the dark blue linear gradient and the whitish radial gradient, layered on top of each other as multiple backgrounds. But Google Chrome had a problem rendering the underlying linear gradient, ignoring its transparent alpha values and rending the colors as opaque. I was therefore forced to break them out into separate elements. The end result is the same. When the popup is created by the setup script, it is given a class of “hidden.” This defines its scale as 0% and its opacity as 0%. When we execute the showPopup method, it removes that “hidden” class. Because the popup has basic transitions properties defined on it, its scale and opacity transition from zero to full, making it appear to popup out of no where. The scripts also always make sure that the popup is centered in the viewport, regardless of where it was displayed when scrolling down a long document.

For their modal popups, Apple always indicates the default button, what would be equivalent to a submit or OK button, with slightly lighter colors so that it stands out from the other button, which is the equivalent of a cancel/close button. I have the buttons located in a footer and I use CSS3’s flexible box model styles to make the buttons position and size them selves according to available space.

/* Modal Popup Styles */
section.popup {
	width: 75%;
	max-width: 300px;
	border: solid 1px #72767b;
	-webkit-box-shadow: 0px 4px 6px #666, 0 0 50px rgba(255,255,255,1);
	-moz-box-shadow: 0px 0px 1px #72767b,  0px 4px 6px #666;
	box-shadow: 0px 0px 1px #72767b, 0px 4px 6px #666;
	-webkit-border-radius: 10px;
	-moz-border-radius: 10px;
	border-radius: 10px;
	padding: 0px;
	opacity: 1;
	-webkit-transform: scale(1);
	-webkit-transition: all 0.25s  ease-in-out;
	position: absolute;
	z-index: 1001;
	margin-left: auto;
	margin-right: auto;
	background-image: 
		-webkit-gradient(linear, left top, left bottom,
			from(rgba(0,15,70,0.5)),
			to(rgba(0,0,70,0.5)));
}
section.popup.hidden {
	opacity: 0;
	-webkit-transform: scale(0);
	top: 50%;
	left: 50%;
	margin: 0px auto;
}
section.popup > div {
	border: solid 2px #e6e7ed;
	-webkit-border-radius: 10px;
	-moz-border-radius: 10px;
	border-radius: 10px;
	padding: 10px;
	background-image: 
	   -webkit-gradient(radial, 50% -1180, 150, 50% -280, 1400,
		   color-stop(0, rgba(143,150,171, 1)),
		   color-stop(0.48, rgba(143,150,171, 1)),
		   color-stop(0.499, rgba(75,88,120, .9)),
		   color-stop(0.5, rgba(75,88,120,0)));
	color: #fff;
	text-shadow: 0px -1px 1px #000;
}
section.popup header {
	background: none;
	-webkit-border-top-left-radius: 10px;
	-webkit-border-top-right-radius: 10px;
	-moz-border-radius-topleft: 10px;
	-moz-border-radius-topright: 10px;
	border-top-left-radius: 10px;
	border-top-right-radius: 10px;
	border: none;
	color: #fff;
	text-shadow: 0px -2px 1px #000;
}
section.popup header > h1 {
	letter-spacing: 1px;
}
section.popup footer
{
	display: -webkit-box;
	-webkit-box-orient: horizontal;
	-webkit-box-pack:justify;
	-webkit-box-sizing: border-box;
	display: -moz-box;
	-moz-box-orient: horizontal;
	-moz-box-pack:justify;
	-moz-box-sizing: border-box;
}
section.popup footer > .button {
	-webkit-box-flex: 2;
	-moz-box-flex: 1;
	display: block;
	text-align: center;
	-webkit-box-shadow: none;
	-moz-box-shadow: none;
	box-shadow: none;
	margin: 10px 5px;
	height: 32px;
	font-size: 18px;
	line-height: 32px;
	-webkit-border-radius: 8px;
}
section.popup footer > .button.cancel {
	background-image: 
		-webkit-gradient(linear, left top, left bottom, 
			from(#828ba3), 
			color-stop(0.5, #4c5a7c), 
			color-stop(0.5, #27375f), 
			to(#2e3d64));
}
section.popup footer > .button.continue {
	background-image: 
		-webkit-gradient(linear, left top, left bottom, 
			from(#b0b6c4), 
			color-stop(0.5, #7a839b), 
			color-stop(0.5, #515d7c), 
			to(#636e8a));
}
section.popup footer > .button:hover, .popup footer > .button.hover {
	background-image: 
		-webkit-gradient(linear, left top, left bottom, 
			from(#70747f), 
			color-stop(0.5, #424857), 
			color-stop(0.5, #171e30), 
			to(#222839));
}
.screenCover {
	width: 100%;
	height: 100%;
	display: block;
	background-color: rgba(0,0,0,0.5);
	position: absolute;
	z-index: 1000;
	top: 0px;
	left: 0px;
}
.screenCover.hidden {
	display: none;
}

You can try this out online or download the source code to play around with it.

iPhone Style Radios Buttons with HTML, CSS & JavaScript


Works on desktop Safari, desktop Google Chrome, Android, iPhone, iPod Touch and iPad.

Have you ever surfed to a Web page on the iPhone or iPod Touch’s Safari browser and come across a form with standard radio buttons? It’s a pretty miserable experience trying to hit them with your finger. You have to zoom in to do so, maybe zoom in a lot. When Apple was designing the interface for the iPhone, they put a lot of thought into how to make conventional interface elements easier to use in a touch environment. If you think about it, what is a group of radio buttons but a list of items to select from. And only one item can be selected at a time. To cover this requirement Apple came up with the radio table control. A radio table is just a list of items, same as a radio button group. Only one list item can be chosen, and this is indicated by a checkmark on that particular list item which correlates to the single radio button being selected out of a group.

The radio button list looks like this:
Radio Button List

One of the things that I really can’t understand is why people think a mobile touch interface needs radio buttons like on the desktop browser. All the other mobile frameworks are providing ways to implement the standard tiny, round radio buttons. They don’t work for touch interfaces. Get over it. The radio list works better for touch. Embrace it and love it and it will love you. Wait, I didn’t really mean that, but you get the picture. I have a hard enough time hitting normal sized controls designed for the iPhone. Heck, sometimes I can’t even find my iPhone, but that’s another issue.

To make this more like the Web equivalent of radio buttons I added real radio buttons to my solution. Here’s the markup to implement them (note that you still need to great the grouping of the radio buttons by giving each radio button in the list the same name):

<ul id="activityChoices" class="radioList">	
	<li>
		<span>Go eat something</span> 
		<span class="check">&#x2713</span>
		<input type="radio" name="activity" value="Go eat something" />
	</li>
	<li>
		<span>Take a nap</span> 
		<span class="check">&#x2713</span>
		<input type="radio" name="activity" value="Take a nap" />
	</li>
	<li>
		<span>Get some work done</span> 
		<span class="check">&#x2713</span>
		<input type="radio" name="activity" value="Get some work done" />
	</li>
	<li>
		<span>Play a game</span> 
		<span class="check">&#x2713</span>
		<input type="radio" name="activity" value="Play a game" />
	</li>
</ul>

Notice that the last item in each list item is the radio input. Please leave this as such, since it makes it easy for us to target the actual radio button as the last child of the list items child nodes. If you have a need to add other things into the list, insert them elsewhere in the list items collection of child nodes.

We’ll use CSS to hide the radio buttons and when a user touches a list item, we’ll use JavaScript to set the checked state of that list item’s radio button to true. After that, what you do with the user interaction is up to you. In many cases that initial choice can immediately trigger a corresponding action, or you may wait until the user takes a decisive final action that triggers a submit or post of all the selected inputs. Notice the span with the class “check.” It contains a hex value of “&#x2713” which is an HTML entity for a standard check mark. We’ll use CSS to position and hide or show it depending on the user’s interaction.

Here’s the CSS needed to make our list look like the iPhone one. Since the radio button group is based on the list control type, it shares some styles with standard lists:

.list, .radioList {
	-webkit-box-shadow: 2px 2px 4px #666;
	-webkit-border-radius: 12px;
	-moz-box-shadow: 2px 2px 4px #666;
	-moz-border-radius: 12px;
	box-shadow: 2px 2px 4px #666;
	border-radius: 12px;
}
.list li, .radioList li {
	cursor: pointer;
	padding: 8px;
	border-left:  1px solid #acacac;
	border-right: 1px solid #acacac;
	border-bottom: 1px solid #acacac;
	background-color: #fff;
	font-weight: bold;
	-webkit-tap-highlight-color: transparent;
}
.list li:hover, .radioList li:hover {
	background-image: 
		-webkit-gradient(linear, left top, left bottom, 
			from(#4286f5), 
			to(#194fdb));
	background-image: 
		-moz-linear-gradient(top, 
			#4286f5, 
			#194fdb);
	color: #fff;
}
.list li:hover:after, .radioList li:hover:after {
	color: #fff;
}
.list li:first-of-type, .radioList li:first-of-type {
	border-top: 1px solid #acacac;
	-webkit-border-top-right-radius: 10px;
	-webkit-border-top-left-radius: 10px;
	-moz-border-radius-topright: 10px;
	-moz-border-radius-topleft: 10px;
	border-top-right-radius: 10px;
	border-top-left-radius: 10px;
}
/** 
	Styles for single choice lists.
	These are the same in functionality as a 
	radio button group.
*/
.radioList li > .check {
	float: right;
	-webkit-transition: all 0.125s  ease-in-out;
	-moz-transition: all 0.125s  ease-in-out;
	transition: all 0.125s  ease-in-out;
	opacity: 0;
}
.radioList li.selected > .check {
	opacity: 1;
	color: #496691;
}
.radioList li > .check, .radioList li.selected:hover > .check {
	color: #fff;
}
.radioList li > input[type="radio"] {
	display: none;
}

The selector radioList li > .check defines the check mark. We set it’s initial opacity to 0 so that it is completely transparent. When the user selects a list item by clicking/touching, we add a “selected” class to the list item. The selector .radioList li.selected > .check then sets the check mark’s opacity to 100%.

To make all the behavior work, we need to write some JavaScript for a reusable control. We’ll use the light, mobile JavaScript framework ChococlateChip.

/** 
* 
* A method to initialize a list of radios buttons to present the user with a group of single choice options. It takes as the main argument, a unique selector identifying the view or section where the radio list resides.
*
* @method
* 
* ### RadioButtons
*
* syntax:
*
*  $.RadioButtons(selector);
*
* arguments:
* 
*  - string: string A valid selector for the parent of the tab control. By default the selector will target a class, id or tag of the radio list itself, so if you want to pass in a selector for a parent tag, such as an article, section or div tag, you'll need to make sure to put a trailing space on the end of the selector string.
*  - function: function A valid function as a callback. This is optional. The callback gets passed a reference to the clicked item, so you can access it in your callback function.
* 
* example:
*
*  $.RadioButtons("#buyerOptions");
*  $.RadioButtons("#buyerOptions", function(choice) {
	   // Output the value of the radio button that was selected.
	   // Since the actual radio button is the last item in a radio
	   // button list, we can use the last() method to get its value.
	   console.log(choice.last().value);
   };
*
*/
$.RadioButtons = function( viewSelector, callback ) {
	var items = viewSelector + ".radioList li";
	var radioButtons = $$(items);
	radioButtons.forEach(function(item) {
		item.bind("click", function() {
			radioButtons.forEach(function(check) {
				check.removeClass("selected");
			});
			this.addClass("selected");
			this.last().checked = true; 
			if (callback) {
				callback(item);
			}
		});
	});
};	

Because the radio input is the last child of the list’s node collection, we can set its checked value to true when the user clicks or touches a list item. We do this with the line: this.last().checked = true; We also manage toggling of the selected state of a list item by adding and removing a “selected” class. This hides or shows the check mark. We also have a conditional block to check for a callback. If one was passed as an argument, we invoke it. We pass in a reference to the list item that was clicked using the term “item.” This allows us to reference the clicked item in our callback. We can initialize a radio button list as follows:

// Radio button initialization:
$.RadioButtons("#activityChoices", function(item){
	$("#RadioButtons .response").fill(item.last().value);
});

In the above code we’re passing in a reference to the clicked list item in an anonymous function which fills a span with a class of “response” with the value of the list item’s radio button.

The radio button list has an id of “#activityChoices,” so I just pass that in. If “#activityChoices” were a parent node, I would have had to written the selector thus (notice the trailing space at the end before the closing parenthesis): $.RadioButtons(“#activityChoices “); This differentiation is necessary because the selector passed in gets concatenated with “.radioList li” to set up the radio button list’s functionality. There is no way for the control to know when you are targeting the list itself or a parent node. If the selector is the list itself, “.radioList li” gets appended to that, but if the selector is a parent node, you need to indicate that with a trailing space so that the “.radioList li” gets appended with the space separating. Otherwise the resulting complete selector will not identity the radio button list properly and nothing will get initialized. I hope this is clear. Yeah, I could have written the control to check to see if the selector was the list or a parent node, but that would have resulted in a performance hit as the code would have had to do quite a bit of evaluation to determine what the selector relation was.

Remember that the callback is optional. That means if you are construction a form with a submit process, then all you need to do for the radio button list is pass in a correct selector to initialize its behavior. You’ll then get the user’s choice during the submit process.

You can try out an example online or download the source code.

iPhone Style Tab Control with HTML, CSS and JavaScript


Works with Desktop Safari, Desktop Google Chrome, Desktop Firefox 4, Android, iPhone, iPod Touch and iPad.
Please not that this was originally designed for Android/iPhone/iPad use. It can be used on modern desktop browsers. Just note that the attached example will only work with Firefox 4 or later because it uses CSS3 transitions for navigation.

If you’ve used an iPhone or iPad, you’ve seen them. Actually, Apple uses them on the operating system. They even use them on their Website and online assets. As is usual for Apple, they take a very different approach to implementing tabs. They use segmented buttons to indicate tabs. This works especially well in the mobile space where you want the tab to stand out and be an easy target for a fat finger. It also eliminates all the futzing around to try to get all the border of tabs and tab panels to fit together properly. The panels have no borders, just their content. The segmented buttons capture your attentions as the device that toggles the content.

The interaction is simple and immediate. As soon as a user clicks or touches a segmented button, it reveals the new content below. On the iPhone, there are not even any special effects, not even a fade. It just directly switches out the data.

Let’s go over the markup for the tab control. Note that the class names it uses are important. By using standardized markup and classes, we can reuse the tab wherever we need it in our app, and initialize it with one line of JavaScript. The tab controls has two parts. The first is a div with the class “Tabs segmentedControlBase.” This is where the segmented buttons that hide and show the tab panels reside. Following this there is another div with the class “TabPanels,” with obviously holds the tab panels. Here’s the markup:

<div class="Tabs segmentedControlBase">
	<div class="button segmentedControl leftEnd tab"><span>Search</span></div>
	<div class="button segmentedControl tab"><span>Directions</span></div>
	<div class="button segmentedControl rightEnd tab"><span>Options</span></div>
</div>
<div class="TabPanels">
	<div class="tabPanel">
		<p class="rounded">Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.</p>
	</div>
	<div class="tabPanel">
		<p class="rounded">Sed ut perspiciatis unde omnis iste natus error sit voluptatem accusantium doloremque laudantium, totam rem aperiam, eaque ipsa quae ab illo inventore veritatis et quasi architecto beatae vitae dicta sunt explicabo. Nemo enim ipsam voluptatem quia voluptas sit aspernatur aut odit aut fugit, sed quia consequuntur magni dolores eos qui ratione voluptatem sequi nesciunt.</p>
	</div>
	<div class="tabPanel">
		<p class="rounded">Neque porro quisquam est, qui dolorem ipsum quia dolor sit amet, consectetur, adipisci velit, sed quia non numquam eius modi tempora incidunt ut labore et dolore magnam aliquam quaerat voluptatem. Ut enim ad minima veniam, quis nostrum exercitationem ullam corporis suscipit laboriosam, nisi ut aliquid ex ea commodi consequatur?</p>
	</div>
</div>

As you can see, the markup is very straightforward and we can easily reuse the pattern through an application as necessary. Of course, this needs some styling so that it looks like something. The most visual part of the tab control is the set of segmented buttons which constitute the tabs. If you’re going to use these on a mobile phone, be aware the you can probably at maximum fit in four tabs. Two or three is optimal. If your tabs are cramped, hard to read or even hard to touch because there so many, you should probably rethink how you’re present your data. Maybe a drilldown menu would serve your purposes better. Realistically, the tab control is for a small number of data sets which you wish to present to the user.

Here’s the markup. Remember, the left and right ends of a segmented button set are normally rounded. They don’t have to be, but they definitely look more elegant that way. The button or buttons between the two ends are squared off.

.segmentedControlBase {
	display: -webkit-box;
	-webkit-box-orient: horizontal;
	-webkit-box-pack:justify;
	-webkit-box-sizing: border-box;
	display: -moz-box;
	-moz-box-orient: horizontal;
	-moz-box-pack:justify;
	-moz-box-sizing: border-box;
}
.button.segmentedControl {
	display: block;
	-webkit-box-flex: 1;
	-moz-box-flex: 1;
	border: solid 1px #9a9a99;
	border-left: none;
	-webkit-border-radius: 0px;
	-moz-border-radius: 0px;
	border-radius: 0px;
	text-align: center;
	background-image: 
		-webkit-gradient(linear, left top, left bottom, 
			from(#fbfbfb), 
			to(#bdbdbd));
	background-image: 
		-moz-linear-gradient(top, 
			#fbfbfb, 
			#bdbdbd);
	color: #6b6b6b;
	font-size: 16px;
	padding: 10px;
}
.button.segmentedControl:hover, .button.segmentedControl.hover, .button.segmentedControl.selected {
	background-image: 
		-webkit-gradient(linear, left top, left bottom, 
			from(#2a55b1), 
			to(#6297f2));
	background-image: 
		-moz-linear-gradient(top, 
			#2a55b1, 
			#6297f2);
	color: #fff;
	border-color: #193a7f;
}
.button.segmentedControl.leftEnd {
	-webkit-border-top-left-radius: 10px;
	-webkit-border-bottom-left-radius: 10px;
	-moz-border-radius-topleft: 10px;
	-moz-border-radius-bottomleft: 10px;
	border-top-left-radius: 10px;
	border-bottom-left-radius: 10px;
	border-left: solid 1px #9a9a99;
}
.button.segmentedControl.rightEnd {
	-webkit-border-top-right-radius: 10px;
	-webkit-border-bottom-right-radius: 10px;
	-webkit-border-top-right-radius: 10px;
	-webkit-border-bottom-right-radius: 10px;
	border-top-right-radius: 10px;
	border-bottom-right-radius: 10px;
}
/* Styles for Tab states */
.TabPanels .tabPanel {
	display: none;
}
.TabPanels .tabPanel.selected {
	display: block;
}
/* End Styles for Tab states */

This will give us a tab control that looks like this:

tabs

Now we’re going to create the JavaScript to make the tabs function. I’m going to use the ChocolateChip JavaScript framework because it’s tiny, just 12k, and does everything I need it to do in the mobile space. For those of you who would rather use jQuery, you should be able to replicate my control’s functionality fairly easily. If you don’t know jQuery good enough to do that, then maybe you want to give ChocolateChip a look. If you compare the two, you’ll see that ChocolateChip and jQuery code look very similar. The difference is that with ChocolateChip you’re always dealing with real JavaScript. ChocolateChip gives you some simple helpers to make it easier for you to write clean, maintainable and legible code. Yet everything ChocolateChip returns is normal JavaScript. With jQuery, you’re always dealing with an object wrapper and therefore have to do things the jQuery way. You cannot use normal JavaScript with what the jQuery returns.

Anyway, here’s the code that creates the tab:

/** 
* 
* A method to initialize a set of tabs using lozenge or segmented buttons to toggle data sets in a view. It takes one argument, a unique selector identifying the view or section where the tabs reside.
*
* @method
* 
* ### setupTabs
*
* syntax:
*
*  $.setupTabs(tabsSelector);
*
* arguments:
* 
*  - string: string A valid selector for the parent of the tab control.
* 
* example:
*
*  $.setupTabs("#buyerOptions");
*
*/
$.setupTabs = function( viewSelector ) {
	var tabsSelector = viewSelector + " .Tabs .tab";
	var panelsSelector = viewSelector + " .TabPanels .tabPanel";
	var tabs = $$(tabsSelector);
	var panels = $$(panelsSelector);
	
	tabs.forEach(function(tab) {
// Set the first tab and tab panel to the selected state automatically.
		tabs[0].addClass("selected");
		panels[0].addClass("selected");
		tab.bind("click", function() {
			var i = 0, len = tabs.length;
			var panelToHide = null;
			while(i < len) {
				tabs[i].removeClass("selected");
				panels[i].removeClass("selected");
				if (this == tabs[i]) {
					panelToHide = i;
				}
				i++
			}
			this.addClass("selected");
			panels[panelToHide].addClass("selected");
		});
		
		tab.bind("touchstart", function() {
			var i = 0, len = tabs.length;
			var panelToHide = null;
			while(i < len) {
				tabs[i].removeClass("selected");
				panels[i].removeClass("selected");
				if (this == tabs[i]) {
					panelToHide = i;
				}
				i++
			}
			this.addClass("selected");
			panels[panelToHide].addClass("selected");
		});
	});
};

If you look at the argument passed in, setupTabs expects a selector indicating the parent element that the tab control abides in. This can be the article tag, the section tag, or whatever you choose. I wouldn’t expect there to be more than one set of tabs in a particular view. If no selector is supplied, the script searches the entire document for the first occurrence of the tab control’s classes and initializes it. So, if you do have more than one tab control in your mobile app, you’ll need to pass in a selector indicating where it resides.

Since by default the tab control’s CSS sets the segmented buttons to unselected and the tab panels to hidden, the script does some initialization to select the first tab and show it’s tab panel. You don’t have to tell the script how many tabs there are or how many tab panels there are. It searches the markup and figures that out itself.

Now, its time to discuss the data in the tabs. How do you intend to get it in them? The simplest way is to do it on page load. If you have many tab controls, and your app is large, you might want to avoid that. In that case you code do something creative, such as binding an event listener to the transition effect of the view the first time the user navigates to it. You could use the webkitAnimationStart event to load your data with an Ajax call. You’d only want to do this the first time the user comes to the view, so you might want to use either a custom object or a value registered in the sessionStorage object to indicate that the user has visited the section already so you don’t need to reload the data.

You can see this tab control live, or download the code to pick it apart.

Update 9/1/10
What was I thinking? My bad. The two classes “leftEnd” and “rightEnd” are totally unnecessary. Using CSS3 selectors we can always target the first and last tabs using :first-of-type and :last-of-type. That way, you never have to worry about naming the first and last. Just put the tabs and the CSS will create the first and last. So the line:

.button.segmentedControl.leftEnd

should be:

.button.segmentedControl:first-of-type

and the line:

.button.segmentedControl.rightEnd

should be:

.button.segmentedControl:last-of-type

The online sample and downloadable code has been updated for this.

CSS3 Gradient Image Masks


Works on Desktop Safari, iPhone, iPod Touch, iPad.

Besides using CSS3 gradients as background images, you can also use them as image masks. An image mask will reveal whatever is behind the element whose background it is applied to while simultaneously masking whatever is in its parent element. This affects both any background properties, colors, images or gradients, as well as the foreground content, whatever it may be. This allows for some interesting layered effects, but you do need to take care that your content gets masked by the mask. I’ll say that again, don’t let your content get masked by the mask. It can happen, so use masks sparingly and with care.

The notation for an image mask is fairly straightforward. First off, you can use an image, either a png or and svg graphics as a mask. The notation is like this:

-webkit-mask-box-image: url(image-mask.png) 75 stretch;

In this post we’re going to look at how to use CSS3 gradients to mask. So my little trick is to nest the element with the mask inside another element which has a background color, background image or background CSS3 gradient which the mask will reveal. In my first example I’m going to give the parent element a background color of #000 (black). The mask will therefore reveal that black to varying degrees added a patterned image burn effect to an otherwise flat foreground. So, here’s the styles for the parent:

article {
	width: 600px;
	height: 500px;
	margin: 40px auto;
	border: solid 6px gold;
	background: #000;
	border-radius: 60px;
	-moz-border-radius: 60px;
	-webkit-border-radius: 60px;
	-webkit-box-sizing: border-box;
	overflow: hidden;
	-webkit-background-clip: padding;
	box-shadow: 
		-2px -2px 2px rgba(255,255,255,.75), 
		2px 2px 10px #000;
	-webkit-box-shadow: 
		-2px -2px 2px rgba(255,255,255,.75), 
		2px 2px 10px #000;
	-moz-box-shadow: 
		-2px -2px 2px rgba(255,255,255,.75), 
		2px 2px 10px #000;
}

And its child element will get this style:

section { 
	-webkit-box-sizing: border-box;
	-moz-box-sizing: border-box;
	-o-box-sizing: border-box;
	-webkit-background-clip: padding;
	height: 100%;
	padding: 20px;
	border-radius: 54px;
	-moz-border-radius: 54px;
	-webkit-border-radius: 54px;
	background-color: #800023;
	background-image: 
		-webkit-gradient(linear, 100% 0%, 0% 100%, 
			color-stop(0.15, #8a3726),  
			color-stop(0.15, #800023), 
			color-stop(0.50, #800023), 
			color-stop(0.50, #8a3726), 
			color-stop(0.65, #8a3726),
			color-stop(0.65, #800023));
	-webkit-background-size: 6px 6px, 100% 20px;
	background-image: 
		-moz-linear-gradient(right top, 
			#8a3726,
			#8a3726 15%, 
			#800023 15%, 
			#800023 50%, 
			#8a3726 50%, 
			#8a3726 65%, 
			#800023 65%, 
			#800023);
	background-size: 6px 6px;
	-webkit-mask-image:
		-webkit-gradient(linear, left top, left bottom,
			color-stop(0.123, rgba(0,0,0,1)),
			color-stop(0.18, rgba(0,0,0,0.6)),
			color-stop(0.18, rgba(0,0,0,1)),
			color-stop(0.8, rgba(0,0,0,.5)),
			color-stop(0.8, rgba(0,0,0,0.8)),
			color-stop(0.85, rgba(0,0,0,.9)),
			color-stop(1, rgba(0,0,0,.5))); 
}

Notice in the code above the line beginning with “-webkit-mask-image.” What follows that property is a normal looking CSS3 gradient. Except that we’re using it as an image mask. Because we gave it no size or position, it will automatically stretch over the entire surface of the element. This is what the styles will produce:
CSS3 gradient image mask

As you can see, the image mask reveals the black background color in the shape of the gradient image mask. Now if we want to get creative, we can put something more interesting in the parent element’s background, and image or a CSS gradient pattern. Here’s an example of a CSS3 background gradient pattern being revealed by a mask:
CSS3 gradient image mask

You can see the document live or download the file as you wish.