By Louis Lazaris on April 27th, 2009
Categories: Scripting, Web Design Articles | 22 Comments
Although the days of long and tedious code branches to target specific browsers in JavaScript are over, once in a while it’s still necessary to do some simple code branching and object detection to ensure that a certain piece of code is working properly on a user’s machine.
In this article, I’ll outine 7 areas where Internet Explorer and Firefox differ in JavaScript syntax.
The basic syntax for accessing a specific css property for any given object is object.style.property, using camel casing to replace a hyphenated property. For example, to access the background-color property of a <div> whose ID is “header”, we would use the following syntax:
document.getElementById("header").style.backgroundColor= "#ccc";
But since the word “float” is already reserved for use in JavaScript, we cannot access the “float” property using object.style.float. Here is how we do it in the two browsers:
document.getElementById("header").style.styleFloat = "left";
document.getElementById("header").style.cssFloat = "left";
JavaScript can easily access and modify CSS styles that have been set on objects using the object.style.property syntax outlined above. But the limitation of that syntax is that it can only retrieve styles that have been set inline in the HTML or styles that have been set directly by JavaScript. The style object does not access styles set using an external stylesheet. In order to access an object’s “computed style”, we use the following code:
var myObject = document.getElementById("header");
var myStyle = myObject.currentStyle.backgroundColor;
var myObject = document.getElementById("header");
var myComputedStyle = document.defaultView.getComputedStyle(myObject, null);
var myStyle = myComputedStyle.backgroundColor;
As is the case with the “float” property, our two major browsers use different syntax to access this attribute via the getAttribute method in JavaScript.
var myObject = document.getElementById("header");
var myAttribute = myObject.getAttribute("className");
var myObject = document.getElementById("header");
var myAttribute = myObject.getAttribute("class");
This syntax would also apply using the setAttribute method.
See a demo page that demonstrates these two different syntaxes
Similar to number 3, we have different syntax to access a <label> tag’s “for” attribute in JavaScript.
var myObject = document.getElementById("myLabel");
var myAttribute = myObject.getAttribute("htmlFor");
var myObject = document.getElementById("myLabel");
var myAttribute = myObject.getAttribute("for");
It would be rare that you would want to find the cursor position of an element, but if for some reason you need to, the syntax is different in IE and Firefox. The code samples here are fairly basic, and normally would be part of a much more complex event handler, but they serve to illustrate the difference. Also, it should be noted that the result in IE will be different than that of Firefox, so this method is buggy. Usually, the difference can be compensated for by getting the “scrolling position” — but that’s a subject for another post!
var myCursorPosition = [0, 0]; myCursorPosition[0] = event.clientX; myCursorPosition[1] = event.clientY;
var myCursorPosition = [0, 0]; myCursorPosition[0] = event.pageX; myCursorPosition[1] = event.pageY;
Sometimes it’s necessary to find out the size of the browser’s available window space, usually called the “viewport”.
var myBrowserSize = [0, 0]; myBrowserSize[0] = document.documentElement.clientWidth; myBrowserSize[1] = document.documentElement.clientHeight;
var myBrowserSize = [0, 0]; myBrowserSize[0] = window.innerWidth; myBrowserSize[1] = window.innerHeight;
Okay, this is not a JavaScript syntax issue — alpha transparency is set via CSS. But when objects fade in and out via JavaScript, this is done by accessing CSS alpha settings, usually inside of a loop. The CSS code that needs to be altered via JavaScript is as follows:
#myElement {
filter: alpha(opacity=50);
}
#myElement {
opacity: 0.5;
}
So, to access those values via JavaScript, we would use the style object:
var myObject = document.getElementById("myElement");
myObject.style.filter = "alpha(opacity=80)";
var myObject = document.getElementById("myElement");
myObject.style.opacity = "0.5";
Of course, as mentioned, normally the opacity/alpha would be changed in the midst of a loop, to create the animating effect, but this simple example clearly illustrates how it’s done.
These are some of the differences I’ve personally come across either in development or research, but I’m sure there are others. Leave a comment if you think I’ve left out an important one, and I’ll try to update this list, or else do a “Part 2″ post.
NEW COMMENT RULES: I no longer have any tolerance for people that use blog comments for link bait and branding. Although comments are pre-approved, comment rules are as follows:
If any of the above rules are broken, one of the following will happen:
I apologize for the inconvenience, but I'm no longer interested in comments that detract from the discussion. Thank you for your cooperation.
Nice find!
I personally use jQuery and don’t run into these issues.
A good reference none the less.
IE / FF
Position:
offset / layer (offsetX / layerX ….)
page / ??? (PageX, PageY)
Event..
??? / currentTarget
returnValue / preventDefault
srcElement / target
fromElement, toElement / relatedTarget
cancelBubble / stopPropagation
has more …
All the more reasons to use a JS library like jQuery which normalizes all this.
Good, informative article. As a nitpick I would have to argue that these aren’t really JavaScript language differences, but differences in the DOM API implementations of different browsers.
And of course, using a framework (Prototype, jQuery, Dojo, YUI, etc.) will help you not have to worry about these differences and not waste time worrying about these annoying differences.
For the always-comprehensive list of differences: http://quirksmode.org/compatibility.html
#3 and #4 are misleading. Your examples simply point out a problem with IE’s getAttribute() method, which internally returns a DOM property rather than an actual attribute. In Firefox, you also need to use element.className and element.htmlFor if you want to access the DOM property value, but since its getAttribute() is not broken, you can pass it the correct attribute name.
#6 is also misleading because it depends on whether IE is in standards mode or not. In quirks mode, your IE solution will not work. And in FF, your IE solution works just fine.
You should also be careful to mention browser versions. IE6 != IE7 != IE8.
Um, I think you pasted the wrong line in example 1:
to access the background-color property of a whose ID is .header., we would use the following syntax:
document.getElementById(“header”).style.borderBottom= “1px solid #ccc”;
Anyway, nice article!
Awesome list, i’m so used to frameworks, that i forgot a lot of basic stuffs like this ones. Bookmarked, thx.
Do you know if the “style.filter” property works in IE 8?
I know Microsoft changed the CSS “filter” to “-ms-filter” and altered the syntax, but I’ve not been able to find if this affects scripting.
@tz:
Thanks for catching that. I had originally used the border property while writing the examples, but changed it later.
@Matt Kruse:
Thank you for clarifying the issue with getAttribute. My list was very basic though, so I suppose, yes, it may be a little misleading.
document.getElementById(), when used on form elements in IE returns the name, not the id [affects 6 ]
so if yo have
getElementById(“stuff”) will return the form element not the div
You left out a big one that’s had me caught up at least once. That would be IE’s ability to grab an element by id, without using getElementById(). This issue hung me up when I had a div with an id of “totalPrice”, and used the same “totalPrice” as a variable name in the js on the page. IE thought I was trying to access the element when I called the variable.
Sounds like another seven reason for not gong near Firefox to me.
I guess it is too much to hope for that all the volunteers get together and sort out Firefox so that it works
in the same way as IE… Still you can always hope :-)
Two things that have bitten me because firefox is a little bit more lax on syntax, and that cause errors in IE (and should probably in firefox)
Objects with trailing commas
In firefox, it won’t complain if you do this:
var foo = {
field1: 0,
field2: 1, // notice the comma
};
Indexing a character in a string with a subscript
var foo = “abc”;
var bar = foo[1]; // should be charAt()
There is an issue with the class property… in IE you have to access it with the .className notation
document.getElementBYId(“myid”).className = “my_class”;
Working of javascript code according to the different browsers like IE,Firefox…
is given in your website, is very helpful for me to enhance my knowledge in javascript. I really got high benefit from this.
I also highly expect more new programming tips better than this in your website in future days.
keep it up
By
Ramesh.S,
Software Engineer,
Chennai,
TamilNadu
India.
Good Post
was really helpful :P
Sorry, I’m missing a subtlety here. On #3, getting (or in my case setting) the class, what exactly is the diff between FF and IE? If someone could put just a few more words to it, it would help a lot! Thanks.
@Ken Blackman:
The difference is that if you use
getAttributeto access the “class” attribute of any given element, you use different syntax. In IE, you use the syntax “className”, whereas in FF you use just plain “class”. I’ve set up a demo page that might make it more clear:getAttribute Demo.
I will introduce the more 2 differences between FireFox and IE for which I bumped into it when I worked with my projects:
1. Strict Syntax Validation
FireFox, Safari, Google Chrome validates syntax very strict. The following example doesn’t work correctly without “px”:
myDiv.style.width = (myDiv.offsetWidth – 50) + ‘px’;
Whereas, IE and Opera automatically add “px” if developer overlooks this string.
2. Look up text value of parent node. IE and FireFox treats it in different ways.
if(navigator.appName == “Microsoft Internet Explorer”) //IE
{
selectedValue = document.getElementById(myID).parentNode.innerText;
}
else if(navigator.appName == “Netscape”) //Firefox
{
selectedValue = document.getElementById(myID).parentNode.childNodes.item(1).innerHTML;
}
Thank you. This post was incredibly useful.
I really wish all browsers shared the same javascript. I have wasted years of my life fixing scripts for IE!
thanks for the article…. not everyone cares to use frameworks (as more power can be achieved in direct js most of the time), but it helps to remember these key points ;)
Thank you g, I totally agree that tools like JQuery are useful, but they can never be as powerful as building JS from the ground up. I prefer to have my own library of scripts for firefox/ie compatibility (and I add to it all the time). I feel as though tools like JQuery dull a programmer’s ability to think with logical creativity when confronted with a problem that the framework cannot handle.