Namespacing your JavaScript
Perhaps a very uncommon approach to developing web applications that require JavaScript (but should be more common) is namespacing your scripts. This can be done very simple-like in a manner that is painless and nice-looking. And I know “looking-nice” should generally not be a determining factor on how your code should be written, but this indeed has some benefits as well.
Coming up with a Name
First things first, we need a name to space. If I were me (sometimes I’m not), but if I were, I would go with my own name Dustin Diaz
. But of course, I’d want to abbreviate that to just DD. However since I don’t want this to be confused with Drag n’ Drop,
I’ll add my middle initial which will make it DED. Now from this point on, all JavaScript for DustinDiaz.com or developed by me for others, it will live under the DED Object.
Extending the namespace
Perhaps my most favorite way of setting up an application object is to do the following steps:
- Create the
DEDObject as a function literal - self-invoke the function
- return my methods
This is demonstrated below:
Namespace Object example
var DED = function() {
return {
method_1 : function() {
// do stuff here
},
method_2 : function() {
// do stuff here
}
};
}();The benefit of this structure is that due to the fact that we have a function literal, we can add private properties and methods into the function before returning an object. This can be done as follows:
adding private properties and methods
var DED = function() {
var private_var;
function private_method() {
// do stuff here
}
return {
method_1 : function() {
// do stuff here
},
method_2 : function() {
// do stuff here
}
};
}();In this case if we were to set our private_method to return… oh I dunno, “Hello World” - then we can call that within our public methods, alert it, and indeed receive a nice little message. However one thing to note, I would argue whether or not our public methods are… public. Of course, in this simple example, they are nothing but public. But there are also “privledged methods” which have access from the outside (public) and also have access to private methods. They are indeed possible to delete and replace, but you can never get at the information to which they were trying to keep secret.
Douglas Crockford also explains a bit on this in relation to Objects and its members, and specifically using the this keyword in his article labeled private memebers in JavaScript. Hopefully the examples on this page will get you thinking of how better to organize your code, avoid conflicts with external application code, playing nice with the “other JavaScript kids,” and indeed the benefits of closures (inner functions). Have fun!
April 25th, 2006 at 12:59 am
I never felt the need to do this and I can’t agree that it’s nice-looking. I understand why you’d want this though.
April 25th, 2006 at 3:06 am
Good article. After thinking alot about it I came to the conclusion that this is just about the best way to namespace in JavaScript too. One other advantage to that way of laying out code is that you can easily include initialisation code in the function body that get evaluated as soon as the script is loaded. It’s a good place to attach your onload/DOMContentReady handlers:
var Thing = {
function initialise() {
// do some stuff on load.
}
return {
my_method : function() {}
}
window.onload = initialise;
}
April 25th, 2006 at 5:37 am
Nice thinking Dustin. I do something similar as far as name-spacing goes using the SI object but never delved into true private functions. After seeing your example, I liked the idea but not the lack of symmetry. Here’s what I propose (some reasoning first for Fry). Say I have an object that encapsulates XMLHttpRequest, it has two public methods, get() and form() that use the private request() method to handle all the object detection and correctly setting the request method (as in GET or POST). I don’t want people to use the request() method directly.
SI.Request = function()
{
// private functions and properties
var _private =
{
request : function(method, url, data)
{
// handle XHR
}
}
var _public =
{
get : function(url)
{
return _private.request(’GET’, url, null);
},
form : function(form)
{
// parse form
return _private.request(form.method, url, data);
}
};
return _public;
}();
I like it. I just did some testing and I thought there might be issues with _public functions calling _private functions that rely on other _public functions but everything seems copacetic.
April 25th, 2006 at 6:32 am
Using all caps for your namespace doesn’t really sit well with me, but that’s a pretty trivial matter.
What exactly makes your var ‘private_var’ private? Seems like I can do this no problem:
DED.private_var = “You’re not very private”;
April 25th, 2006 at 7:01 am
Justin, that will create a new property named private_var on the DED object which isn’t the same as a private variable. Try it out.
April 25th, 2006 at 7:30 am
Looks very interesting, though I must have forgotten some of the Javascript lessons since I can’t get any working samples out of any of the codes above.
I have recently started to move all my functions into the :
var myscript = {
varname : ’something’,
somefunction : function(){
alert(this.varname);
}
};
myscript.somefunction();
If someone would be nice enough to add some alert() to the code snippets above and some invocation examples, like my selv explanatory example, then I can skip class and dive into it :)
April 25th, 2006 at 7:33 am
By the way, your blog messes up the Javascript code, single quotes are replaced with those fancy webquotes which doesn’t exactly work in Javascript. Most people probably knows this, but the cut’n'paste people should know this as they will get a Javascript error.
April 25th, 2006 at 7:36 am
Hey Dustin!
Quick question… Would you recommend this only for classes you don’t wish to have instantiated (ie: static class)?
One thing I’ve learned recently is that when encapsulating any methods of a class within the constructor function, any object instances of that class will each have their own copy of the methods, rather than sharing a copy from the prototype object; thus using more memory.
I noticed the method you mentioned is being used in the YUI libraries (kickass, btw!)… Can you shed any light on whether or not this is true for the namespacing technique here?
Thanks!
Frank
April 25th, 2006 at 7:54 am
This technique creates a namespace rather than a ‘class’ that can have multiple instances. Be careful not to confuse them. You can put object constructors inside your namespaces though:
var animals = (function() {
function Dog(name) {
this.name = name;
}
Dog.prototype.bark = function() { alert(”Woof! ” + this.name) }
return {
Dog : Dog;
}
})();
Then you can make lots of dogs like this:
var fido = new animals.Dog(’fido’);
fido.bark() //=> alerts ‘Woof! fido’
var rover = new animals.Dog(’rover’);
rover.bark(); //=> alerts ‘Woof! rover’
April 25th, 2006 at 7:55 am
Shaun,
Sure enough :)
I did try it out before, but it didn’t occur to me that I was creating a new variable not setting the existing one. How wacky.
April 25th, 2006 at 8:10 am
Dan,
Ahh!! That makes sense. I totally confused myself with the namespace/class difference. I’ve been burying myself with as much OO-related JS content these past few weeks, that I seem to automatically swerve to that end of the curve. ;)
Your example makes total sense. Thank you!
One question about the example though… The anonymous function that you declared at the beginning of the “animals” namespace; what is the purpose of that? Is it called at any point, or does it serve some other purpose?
Another thing I noticed (not sure if it was just a typo on your part), was that having a semi-colon in the return:
return { Dog : Dog; }
seems to throw an error. Removing it works. Not sure why, but it does.
Thanks again for the clarification!
April 25th, 2006 at 8:14 am
That anonymous function encloses the entire namespace as in Dustin’s example. As for the ; that was meant for the line below (oops) but removing it is fine.
April 25th, 2006 at 8:49 am
I really need to come up with a better system for people to post code on this blog. I realize with as much as I talk about it, people are definitely going to post it.
As regards to Justin’s comment on “What makes them private” - they indeed cannot be called from outside the DED Object. For instance, in this case, you can’t call
DED.private_method(). However the other methods (in the return) can. That’s why I think it’s best to call them those ones ‘privledged’.April 25th, 2006 at 8:51 am
Dan, you also beat me to the true OO example. I was going to post something similar last night but got too sleepy. I may just have to single you out on a later entry.
April 25th, 2006 at 8:57 am
single me out? That sounds scary…
April 25th, 2006 at 9:54 am
Hehe. It will be completely in a good way. Basically you’ve demonstrated how to use function literals (or object literals) and make them reusable like classes.
April 25th, 2006 at 9:58 am
Another site detailing OOP in JS. It’s more digestible than Crockford’s for me at least.
April 25th, 2006 at 3:00 pm
I’ve seen another attempt at ‘namespacing’ in FlashObject on deconcept. In version 1.3 he even went as far as “com.deconcept.*”. If you really want to go all the way with namespacing you should probably use both of them combined. What if somebody wanted to add animals.Cat to Dan’s example?
As a way of cleanly adding private members I do agree this is a good techinque. But you have to remember that private methods in javascript are not really private since anybody can see them in the code and can change the script to make them public. It forces people to use the API instead of the core, but still…
April 25th, 2006 at 5:49 pm
Fry, you may be mistaken by what a private method is. Just because you can see it doesn’t mean it’s public. If we were to go by that definition, then PHP is an entirely hidden language just because we can’t see the source code. Read Crockfords article and you might have a better understanding on why they’re called private.
Also, the com.domainName is a technique used in Java - or so says Jonathan Snook who told me so.
Another point on Justin’s comment about all capitals on ‘DED’ - since JavaScript is case-sensitive when declaring variables, it is less likely that the ‘DED’ namespace will have a collision with another.
April 25th, 2006 at 5:53 pm
Just one thing to note though on Fry’s comment again, the deconcept method is indeed a good approach. But although at first pass it seems like a bullet-proof approach, it requires everyone to be on board with it. Declaring variables like ‘com’, ‘net’, or ‘eu’ seem very likely to be collided with. But then again, so can ‘DED’. Nevertheless, nothing really is full proof and requires attention to detail when you’re integrating code from multiple applications and developers.
April 25th, 2006 at 8:08 pm
> the com.domainName is a technique used in Java
I think this is more an API/SOAP/SOA thing, not exactly Java specific. .NET uses the same syntax to when creating namespaces for webservices.
> it is less likely that the ‘DED’ namespace will have a collision with another.
Yeah I was just saying that b/c it looks like a constant. Of course, an acronym isn’t a constant though. Nevermind, I’m just being a pedantic nut.
April 26th, 2006 at 5:28 am
Dustin, I know what you’re saying about private methods and what the difference is (I am a JavaScript and Java user/developer). My point was that when using a javascript library I can just copy/paste the code around to make a method public. With compiled languages you can’t…
I don’t think that there are many people who use private methods in JavaScript and I rarely see a private method that has a valid reason to be private (why shouldn’t a power user use the request method from Shaun’s example?).
April 26th, 2006 at 8:16 am
There are two great articles on this subject by Tom Trenka: Coding for Portability 1 and Coding for Portability 2.
May 29th, 2006 at 5:40 am
[...] Namespacing your Javascript. [...]
May 29th, 2006 at 10:26 am
Dustin - Have you checked out the Dojo toolkit? It has a great namespacing feature. You just do:
dojo.provide(’com.dustindiaz.tetris.ScorePanel’);
Then later you can import it using:
dojo.require(’com.dustindiaz.tetris.ScorePanel’);
May 31st, 2006 at 1:06 am
Dustin,
Have you seen Ajile [http://ajile.sf.net/]? It’s a very effective approach to JavaScript namespaces with most of the end-user complexity removed.
Ajile lets you do define a namespace as simply as:
Namespace (”com.dustindiaz”);
and then import from it as easy as:
Import (”com.dustindiaz.MyObject”);
or
ImportAs (”MyAlias”, “com.dustindiaz.MyObject”);
It’s really worth a look.
June 28th, 2006 at 11:23 pm
[...] Granted, I’ve written on this topic briefly when talking about JavaScript namespaces but didn’t get into full detail on how to achieve all three styles, and how it all actually works. [...]
July 17th, 2006 at 8:52 am
[...] This step ties in with unobtrusive JavaScript, so let’s not dwell on it. In essence it means that a maintainer could add your script to any page and don’t risk overwriting other code by doing so. You can keep scripts self-contained by namespacing them or wrapping them in an object via the object literal. You ensure you don’t hijack variables by keeping them in scope of your functions via the var keyword and you don’t hijack events by using addEvent()code> or its derivates. [...]
July 17th, 2006 at 10:00 pm
i suppose this is namespacing as javascript allows it, but i am not sure i would classify it as such with regards to traditional language semantics, particularly with regards to scoping rules and their verification (for classes as well as objects). this is simply the semantics of functions in javascript, and your ‘namespace’ is just a viable retval. why not then just refer to this as what it is - a function?
July 27th, 2006 at 1:20 am
I’m using your approach for creating namespaces but run into a problem.
I need a way to return an object.
I have the BLA namespace with some functions in it.
Now I need to add a cookie object to the BLA namespace but people need to call it like this:
new x = BLA.cookie();
x.set(”test”, “value”);
x.get(”test”);
Is there a way to create such an object using your approach?
BLA.cookie = function() {
return {
new Object();
}
}();
the above doesn’t work but would it be possible to to something like that?
August 1st, 2006 at 11:35 pm
Late to the party, but….
Desparately wanting namespacing to give some organization to my code, I tried the following and was generally pleased with the result, but did have a couple of questions.
I was able to use this for nested namespaces. I would create the namespace as an empty object in the parent namespace, then redefine it later to be a namespace created in the same form as you describe. Is there an easier way?
Finally, I had some difficulty referring to public members from private methods and referring to public members from other public members. It turns out I needed to fully qualify the names. Is there an easier way?
Thanks,
Brad
August 3rd, 2006 at 11:32 pm
Brad,
You might want to look at: http://ajile.iskitz.com/
Ajile is a JavaScript namespacing library with importing.
Good luck.
August 4th, 2006 at 5:50 am
[...] Namespacing in its essence is the art of clearly seeing your code as one being. Namespacing is the path to prevent clashing of variable and function names in the jungle of scripts, and is one of the points of the JavaScript-Fu Star of Unobtrusiveness.The way of the Namespace is simple and clear once you learn to understand its true meaning. The following mantra can help you better understand the path you have to follow to reach the Namespace nirvana through the use of Object Literal Namespacing: [...]
August 11th, 2006 at 1:45 am
It’s nice, i didn’t know that it’s possible to define a function in another function in javascript.
Anyway it’s a great method when you have 2 function with the same name in a page, just like what you do in common programming language like JAVA.
September 2nd, 2006 at 12:50 am
I don’t think that there are many people who use private methods in JavaScript and I rarely see a private method that has a valid reason to be private (why shouldn’t a power user use the request method from Shaun’s example?).
A power user can modify the script if it seems necessary. Encapsulation’s purpose is not to hide source or render it untouchable. It merely helps enforce integrity. By making members private, the developer is saying, “This is intended for internal use. If you mess with it, the data or functionality may become untrustworthy.” That’s why I use encapsulation in my own projects, even when I’m the only developer who will ever touch the code. It can keep you from shooting yourself in the foot.
September 3rd, 2006 at 8:01 pm
Incidentally, I was playing around with this concept and found a syntax for namespacing that I consider a little more readable:
var NAMESPACE = new function() {
// Private stuff
var private_variable = ‘foo’;
function private_function() { alert(’private’); }
// Public stuff
return {
public_variable: ‘bar’,
public_function: function() { alert(’public’); }
};
}
I like it because the “new function()” in the NAMESPACE declaration is an immediate indication that the contained code is a single instance of the object.
September 15th, 2006 at 5:44 am
Could I ask a beginner question…
I’m developping a suite of controls in ASP.Net and I’m trying to figure out how I could code the client side script (js) the best way… using JS inheritance for all my controls…
As example, let’s say I have following server-side controls (coded in c#):
MyTextBox -> provides some extra client side functionalities using js (like watermarking)
MyMaskedTextBox -> inherits from MyTextBox (and from its watermark functionality) but adds masking capability. In c#, no problem to inherit the MyTextBox class, it’s common practice. But in JS, how can this be done using the best practices… And also not using some special features (i need it to work on almost any common browser)…
And so I would create MyOtherTextBox, inheriting from MyTextBox too…
The problem here does nothing to do with C# coding, I just want to learn how I could achieve a better object oriented implementation for my components at client side.
So my questions are :
- What would you advice in order to create some clean javascript objects ?
- I first though it would be nice to wrap all my objects (MyTextBox, MyMaskedTextBox, …) in a single namespace, but does it really matter in Javascript ? It disturbs me because the syntax to define a Namespace I can read here seems to be quite the same as I would define an Object. For example :
var NAMESPACE = new function() {
// Private stuff
var private_variable = ‘foo’;
function private_function() { alert(’private’); }
// Public stuff
return {
public_variable: ‘bar’,
public_function: function() { alert(’public’); }
};
}
If i replace the word NAMESPACE by MYOBJECT, wouldn’t it be like a new object definition ? (I’m not familiar with JS OO syntax so if i’m wrong plz tell me ;o) )
Well, I have many questions to ask, I anybody could give me some answers, I really would appreciate. Actually, I could make it by myself, the way I think, but I really want to improve my practices in JS and JS OO so … I hope that you guys will make me a better JS coder :-)
Thanks by advance !
Mike.
September 15th, 2006 at 10:02 am
JavaScript inheritance… Here’s a quick example. SubClass inherits the properties of it’s prototype, so you just overwrite the default prototype (Object) with your SuperClass:
September 15th, 2006 at 10:06 am
In JavaScript2.0 there are already built in language approaches for javascript solving the name space problem.
September 15th, 2006 at 2:34 pm
I forgot to mention… the YUI Library also provides a mechanism for extending classes:
http://developer.yahoo.com/yui/yahoo/#extend
September 20th, 2006 at 4:27 am
Ok, thanks for the first answers.
So some quick questions:
- speaking about prototype, its a word used on every post speaking about Javascript. Sorry for asking that but is there any relation between Peter Foti example and the Prototype library ? I think there isn’t, but it’s just to make it sure. Is “prototype” a reserved word in STANDARD Javascript or does it need the prototype library to get working ?
What I need is to implement an independant (and cross browser) solution (at least, IE and FF) so I don’t want to use external functionalities (like prototype, moofx and others…)
Talking about Javascript 2.0, is it compatible in all browsers ? (even older ones ?)
Thank you for your help !
Mike.
September 20th, 2006 at 8:05 am
@Miketrix
Just to clarify, yes “prototype” (in the way I used it above) is built into JavaScript and should not be confused with the Prototype JavaScript Framework (which I personally think was a very stupid naming choice as it only causes confusion). You do NOT need the Prototype Framework stuff for my examples.
Why do you need to implement a totally independant solution? I highly recommend the YUI Library. It has a BSD license. It’s VERY well put together, with a growing user base and a great support mailing list. It works in A-Grade browsers (http://developer.yahoo.com/yui/articles/gbs/gbs_browser-chart.html), and makes development SOOOO much easier. Don’t reinvent the wheel.
September 20th, 2006 at 12:56 pm
Hi Peter and thank you so much for your answoer.
Now I won’t be confused anymore on prototype.
Actually, I’m developping some customized controls in C# for ASP.Net 1.1 (kind of enhanced input controls… Sure I’m not the first to do that). Those are developped for many reasons.
The first is that we need them in our financial software, and therefore we don’t want to use external libraries… Well it’s more that our chiefs don’t agree on using external stuff … I know it’s not always a good choice but … anyway. Furthermore, our software must run on older IE versions too. I’m not targetting only IE for my controls, but at least I need to try to make them the more compliant I can.
The second reason why I’m asking about Javascript Practices is that I want to improve my knowledge and abilities in JS development. Actually, I know like most developers how to code functions in Javascript, getting hand on the DOM in order to make some common features (collapse/expand, validation of input fields, text/image hover effects…) but what I’m interested in, is to achieve using Javascript as a “truly” OO language… Getting benefits of inheritance, encapsulation, …
The first controls I’m developping for the moment are quite simple :
- TextBox : providing Watermark functionality
- NumericBox (inherits from TextBox) : providing real-time numeric typing, including currency symbol, group and decimal separator, …
- MaskedBox (inherits from TextBox) : providing mask support (for ex: 999-999999-99)
- DatePicker (inherits from MaskedBox) : providing dynamic calendar popup + date input field.
As you can see, all components I have are in a same hierarchy. Of course, many logic and validations are done at server side (in C#) and all Javascript is included dynamically in the page also using Javascript.
But instead of declaring a bunch of Javascript Functions, passing tons of parameters to each function, I would prefer to declare a clean object for each control to handle. SO the object can handle itself, calling methods that get access to the properties of the object, without the need of passing them again and again… And also, some methods of TextBox control could be overrided for other controls inheriting from it.
Well, not much more to say, I just want to use OO as far as possible in common Javascript programming.
SO, any extra advices are still welcome.
Thanks again !
September 21st, 2006 at 9:48 am
@Miketrix
Your approach is a sound one. I, too, prefer an OO methodology. I like to think of it like this:
1. The Content should be accessible for the widest audience possible. That means not reliant on JavaScript or CSS.
2. Additional behaviors (via JavaScript) should be unobtrusive, and get “attached” to the content.
For example, I prefer to define my JavaScript objects and then “attach” whatever event handlers (vs. defining intrinsic event handlers using onclick attributes, etc.). Sometimes scoping issues can be a little tricky, but once you get the hang of it I’m sure you’ll find it’s a much “cleaner” way to work.
therefore we don’t want to use external libraries… Well it’s more that our chiefs don’t agree on using external stuff
Understandable I suppose. But one could argue that the YUI Library is more of a Framework with which to build your application. The methods it exposes (particularly in the Dom and Event Utilities) can greatly increase productivity.
Also, I think it will be difficult to create JavaScript that is 100% compatible with new and ancient browsers. You’ll probably want to find some good references that discuss exactly what versions of JavaScript are supported by each browser version, along with something that defines when a method was added to JavaScript (for example, the String.substr() method became available in JavaScript 1.2, which was around the time of Netscape 4 and IE 4). You might be better off to define a minimum browser requirement.
Good luck.
September 24th, 2006 at 12:20 pm
@Miketrix: Regarding your question about namespaces in JavaScript:
In the example I provided, NAMESPACE would be a single instance of an object. “
var NAMESPACE = new function() {}” creates an anonymous constructor and instantiates the object immediately. NAMESPACE is therefore the object itself. The constructor is no longer available. In that respect, its usage is comparable to the singleton design pattern.Just thought I’d clarify how it works to avoid confusion.
September 24th, 2006 at 12:32 pm
Oops… there was a bug in my previous example. The bar variable isn’t available to the show_bar() function. That’ll teach me to post code without running it in a browser first. Here’s a working version.
September 26th, 2006 at 9:09 am
Regarding the question about JS2.0, it is not in old browsers, it is not even in new browsers. The lastest new version of JS coming to us is 1.7 (with generators and destructuring) in Firefox 2.0. And even that is along way from being widely used enough for using in public web pages (it will surely be a while before IE implements it).
December 6th, 2006 at 11:21 pm
[...] Namespacing your JavaScript Create the DED Object as a function literal, self-invoke the function, return my methods. Now, we can add private properties and methods into the function before returning an object. [...]
January 26th, 2007 at 2:54 am
What if I want to create namespace of the variant “com.mycomp.section” instead of just DED. I guess it’s not this simple:
Any hints?
Also, we use YUI as our main library. But I guess using YAHOO.namespace() is mostly for extentions for the YUI lib. Or is it a good choise for all our functions?
What do you think?
January 26th, 2007 at 11:20 am
YAHOO.namespace() is good for all your library methods. But it still essentially extends everything from the YAHOO object… which I don’t see as all that necessarily bad. In the end, if for instance I had DED.util.Dom… You can’t directly extend all the way out to ‘Dom’ using var.. Eg:
In this case, ‘DED’ needs to be declared before it’s extended, and same with ‘util’. So here’s a few that would work:
The above example might not be preferred, but it works. I would do something more like the following:
There are of course other favorable ways too.
January 29th, 2007 at 7:45 pm
@Svein Are
You should take a look at Ajile (http://ajile.iskitz.com/). Lets you create the kind of custom namespace you’re talking about without it being nested under something like YAHOO.
Using it you can do:
Namespace(”com.mycomp.section”);
com.mycomp.section.Method = function(info) {
alert(info);
};
Pretty nifty I’d say…
April 20th, 2007 at 12:33 pm
[...] Nachtrag: Scripts schreiben, die mit andern Scripts gut können durch Namespacing: Namespacing your JavaScript, sprich Namen wo es geht privat lassen und global schreiben, ihnen sozusagen einen eigenen Raum geben. [...]
July 3rd, 2007 at 1:17 pm
Great site, I love the content. I hate to complain but any chance you could change the type or background color? It is a bit hard on my eyes to read. The black type is a bit strong on the white background. Try #333333 for the type. With the black background the white type is also too strong. Thanks for all the great coding tips.
August 21st, 2007 at 1:50 pm
OK, I’ve read this article about ten times since you published it. I think I’m finally getting it. I’m a real fast learner. I’m trying to apply this method to an existing class (Grid.js) that I use that manipulates some table (DOM) properties. I had it set up that the class was used on all pages that contained a certain table. I then had some prototype methods in a separate file that I only used on a specific few of those pages. I’m kinda stuck now because Grid.prototype.someMethod = function () {…} isn’t going to work. I want to be able to exclude the protoype methods on pages that don’t require them in order to keep file size down. Is it possible to have prototype methods outside the object literal?
Thanks
March 11th, 2008 at 7:02 am
I’ m also very interested in any answers could be given to Stuart’s problem
Prototyping of the closure is assumed as working in Dustin’s book (pag 36) but, as far as I can see, it doesn’t
March 18th, 2008 at 1:59 pm
If that code isnt sexy to you then you need some viagra. Im currently rewritting a library and I’ve been griping with whether or not I should. I finally think I will, might as well. i want to be a cool kid too!
June 27th, 2008 at 7:23 am
See JavaScript: The Definitive Guide 5th Edition by David Flanagan Chapter 10: “Modules & Namespaces” for an excellent (but advanced) discussion on the matter.
July 26th, 2008 at 11:17 am
Could someone please, please, please post a working example of Dustin’s routine? Preferably one where you pass something in, transform it and pass it out.
Thanks very much.
Ken
August 5th, 2008 at 1:51 am
In regards to using the Yahoo namespacing, as mentioned by Dustin Diaz and Svein Are; I’d prefer to keep my own code as fairly unrelated to the Yahoo methods as possible. Of course, its just a matter of personal taste really, but I like to think of my code as a separate entity to any of the libraries I employ.
August 14th, 2008 at 1:47 pm
[...] page and don’t risk overwriting other code by doing so. You can keep scripts self-contained by namespacing them or wrapping them in an object via the object literal. You ensure you don’t hijack variables [...]