Section 1: General JavaScript Info

Section 2: JavaScript Programming Basics

Section 3: Syntax

Section 4: Frames & Windows

Section 5: Forms

Section 6: JavaScript Security

Section 7: LiveConnect Techniques

1. What is JavaScript?
2. What is the goal of the JavaScript Newsgroup?
3. Aren't JavaScript and Java the same thing?
4. What browsers support JavaScript, and where can I find documentation?
5. What are the compatibility issues for the Netscape Developer?
6. Where is the current bug list for JavaScript?
7. How do I report a bug?
8. What differences exist between JavaScript versions?
9. What is the difference between Client-Side and Server-Side JavaScript?

JavaScript is a compact, object-based scripting language for developing client and server Internet applications. Netscape Navigator interprets Client-Side JavaScript statements embedded in an HTML page. Server-Side JavaScript is compiled into platform-independent bytecode and used by LiveWire (now part of Enterprise Server 3.0), enabling you to create server-based applications similar to Common Gateway Interface (CGI) programs. This FAQ describes the JavaScript language on the client, specifically its use in Navigator. For information on developing server-based JavaScript applications, see the LiveWire Developer's Guide.

The JavaScript Newsgroup is established as a forum for developers to discuss hot topics, work out mutual problems, and expand the overall knowledge and awareness of JavaScript. Discussions range from the beginner who wants to open a new window to the advanced scripter who wants to manipulate multiple frames in multiple windows. Compatibility issues are also very common within the newsgroup.

No, they are different! JavaScript and Java are similar in some ways but fundamentally different in others.

Two sites that give an excellent description on the differences between Java and JavaScript are:

1. JavaScript and Java
2. Java vs. JavaScript

Both Netscape and MSIE support JavaScript, with newer browsers supporting later versions of JavaScript:
Client side javascript:

• js1.3 reference documentation
• js1.3 guide
• DHTML guide

  Javascript 1.4 core language:
  

• js1.4 core reference documentation
• js1.4 guide

  Server side javascript:
  

• Reference documentation
• Guide

• Index of js documentation

Trying to make the universally-compatible script these days is almost pure folly. In order to support the broadest possible audience, you should author code that is either sensitive to the browser that's executing it, or code that uses only JavaScript 1.0, which is supported by most browsers currently in use.

 These sites contain information that can help make your JavaScript as "browser-friendly" as possible:

• Gordon McComb's JavaScript Compatibility Issues
• WebReference - "It works in Netscape 3.0x but not in MSIE3.0!"
• Danny Goodman's JavaScript Object Road Map and Compatibility Guide

Although Netscape does have an official Known Bugs document for JavaScript, for the most updated and accurate information, please try posting to the JavaScript newsgroup.

DevEdge bug reports are viewed directly by Netscape engineering teams and help identify problems in shipping products and betas. It is suggested you view the various DevEdge newsgroups prior to reporting a bug since identified problems (bugs) are often discussed within the relevant newsgroups. A bug report form is available for developers who wish to report new findings.

Netscape has written a document outlining the features added after JavaScript 1.0.

 A new document outlining What's new in JavaScript for Navigator 4.0 is also available.

Client-Side JavaScript and Server-Side JavaScript have been different, but now are unified as JS 1.2 in Communicator 4.0 and ES 3.0. Client-Side JavaScript is interpreted only within browsers that support it, and the code is visible to the user. Server-Side JavaScript is stored in a pre-compiled state on the server, so it is browser-independent, and only the results of the JavaScript programs are passed to the browser, so that code is never revealed. There are also fundamental differences in the available objects. An article written by Paul Dreyfus may be seen at View Source on the DevEdge Web site that describes JavaScript on the Server. Server-Side JavaScript is supported separately from client-side JavaScript -- use the snews://secnews.netscape.com/netscape.dev.livewire.programming and snews://secnews.netscape.com/netscape.dev.livewire.dbconfig newsgroups, or the LiveWire FAQ instead.

1. How can I use JavaScript to detect the browser type & version?
2. Is JavaScript's math inaccurate (0.119 * 100 = 11.899999)?
3. How do I do image "rollovers" (change the image when the mouse is over it)?
4. How do I use the different substring functions to extract part of a string?
5. How do I round numbers off to a certain number of decimal places?
6. Why do I have trouble with JavaScript within tables?
7. Where can I get information on Cookies?
8. Why do I sometimes see a URL beginning with wysiwyg?
9. Problems caching .js source files?
10. Can I detect the back button?
11. What's the difference between Grey and Gray?
12. Why doesn't document.lastModified work when I put my files on a web server?
13. Why do I have trouble adding values in forms?

By inspecting the navigator object's properties. You should consult:
• JavaScript Reference: Navigator Object
• Comprehensive list of user-agent strings (Courtesy I. Thum)
• Detecting a JavaScript Cleint (Danny Goodman / View Source)

It's not an issue of accuracy, but rather of the way that JavaScript does math internally. Generally, you can avoid any unexpected results by rounding the result of your calcuations to the precision of the most precise number in your equation. In the example above, you would get the correct answer if you rounded the output to three decimal places.

It's easier than you think. Just use the onMouseOver and onMouseOut events to change the .src property of the image. Note that images are named using the name= element of the <image> tag:

<a href="home.html"> onMouseOver = "document.homeImage.src='homeLitUp.gif'; return true;"> onMouseOut = "document.homeImage.src='homeNormal.gif'; return true;"> <img name="homeImage" src="homeNormal.gif"> </a>

 

There are several ways of selecting substrings in JavaScript, including substring(), substr(), slice(), and the regexp functions.

In JavaScript 1.0 and 1.1, substring() exists as the only simple way to select parts of a larger string. For example, to select the string press from Expression, use "Expression".substring(2,7). The first parameter to the function is the character index at which to start, while the second parameter is the character index at which to end (non-inclusive): substring(2,7) includes the indices 2, 3, 4, 5, and 6.

In JavaScript 1.2, substr(), slice(), and the regexp functions can also be used to dissect strings.

substr() behaves in the same fashion that Perl's substr does, where the first parameter indicates the character index at which to start, while the second parameter indicates the length of the substring. To perform the same task as in the previous example, "Expression".substr(2,5) can be used. Remember, 2 is the start point and 5 is the length of the resulting substring.

When used on strings, slice() behaves very similarly to the substring() function. It is, however, a much more powerful tool, being able to operate on any type of array, and not only on strings. slice() also makes use of negative offsets to refer to a position taken from the end of the string rather than the beginning. "Expression".slice(2,-3) will return the substring found between the second letter and the letter third from the end, returning press once more.

The final, and most versatile, method for dealing with substrings is through the regular expression functions in JavaScript 1.2. Once again, demonstrating with the same example, the substring "press" is obtained from the string "Expression ":

write("Expression".match(/press/));

While this example is trivial, the regexp documentation is quite complete, providing more complex examples of the RegExp object and its related functions.

Use Math.round() to multiply then number by ten times the number of decimal places you want to round off by, then divide by that number.

For example:

var pi = "3.1415926535"; // Round to tenths: pi = Math.round(pi*10)/10; // Round to hundredths pi = Math.round(pi*100)/100; // etc...

 

There are a few bugs related to JavaScript and tables. They all relate to the number of times a page is actually parsed out by the browser before it is displayed. The worst of these is the one that displays your code within the table, even though it executes properly. These bugs are not limited to Netscape 2.x; you could encounter them in Netscape 3.x as well.

Your best course of action against this is to generate your entire table with JavaScript.

Information on cookies can be found in a wide range of places. Here are some examples:
• Netscape's Cookie Spec
• Cookie Central. Includes an FAQ written by David Whalen, previous JS Champion.
• InterNIC - RFC2109

(What you see is what you get!)

While working with client-side JavaScript code, one might see URL's in the format: wysiwyg:/0/http://www.myserver.com/document.html. These are internal references that are displayed in the location field and are not troublesome in any way to either users or developers. The "0" in the example is an internal document id, not a controllable parameter.

When reloading code that sources data from .js files, Netscape clients frequently seem to become confused and display the .js file as plain text on the HTML page instead of interpreting the code contained within it.

Netscape clients prior to version 3.01 are not actually affected by this problem as they do not cache the .js source file at all, and so always assume .js files to contain JavaScript. Beyond 3.01, the file is cached, and the given mime-type, not the extension is later used in withdrawing the file from cache. If the mime-type is set incorrectly by the server, then the .js file will be displayed as the type text/plain.

To force the server to accompany .js files with the correct mime type, find the mime.types file (or equivalent) for the server and add this line: application/x-javascript exts=js.

Unfortunately, no! However, the event handlers onBack and onForward are planned in a later version of Communicator.

Somehow, it seems that "lightgrey" managed to squeeze through as the only British-spelled grey/gray colour. Odd. Here's the complete grey/gray colour list:

darkgray darkslategray dimgray gray lightgrey ** lightslategrey slategray

 

Navigator expects to receive the modification date of a file in the HTTP response header from the web server. In the case of files that are being read from a local disk, Navigator is able to get the modification date from the operating system.

 Not all web servers send the modification date along with the file. When viewing a file with a call to document.lastModified on such a server, you'll get a meaningless date.

Form values are TEXT not NUMBERS and therefore have to be converted to numbers first. Here are two ways to do this.

  
   varName = parseInt(document.formName.elementName.value, 10) 
   varName = document.formName.elementName.value - 0 

In both cases varName will contain the element value as a NUMBER.

1. Why do the <script> container tags need to be enclosed inside comments?
2. Is the LANGUAGE='JavaScript' attribute required?
3. Where can <SCRIPT> container tags be placed within an HTML document?
4. Are semicolons(;)required at the end of all JavaScript statements?
5. How can I use quotation marks within a JavaScript statement?
6. What "scope" does a newly created variable have?
7. What does var do?
8. Is there any way to control the way a page prints using JavaScripts, can page breaks be forced, etc.?

In a nutshell, for compatibility with non-scriptable Web browsers. Potentially, the JavaScript code will be written verbatim to the screen if comments are not used. Any Web browser that is termed as being JavaScript-aware will scan the program code contained within the comments.

<script language="JavaScript"> <!-- alert("Hello, this is The DevEdge JavaScript Newsgroup"); //--> </script>

 

The language attribute is always recommended. Omitting the language tag will force the browser to use its default scripting language, which may not be JavaScript. Furthermore, by specifying a language attribute, you can custom-tailor your code on a scripting language version-by-version basis.

In general, the <SCRIPT> container tags may appear anywhere within an HTML document. It is more viable to have the tags placed within the <HEAD> container because this occurs at the top of the document. This guarantees that this code will be available to any code in the <BODY> of the document by the time it has been loaded.

This becomes useful if one requires function or "global" variable definitions. (i.e.)

<html> <head> <script language=JavaScript> function showDate(){ now = new Date(); document.write("<title>Today is " + now +"</title>"); } </script> </head> <body text="#FFFFFF"> <script language="JavaScript"> showDate(); </script> </body> </html>

 

Semicolons are optional in JavaScript statements that exist on separate lines. However, if you use statements within event handling attributes (as part of a string) you will need to terminate each statement with a semicolon.

<A HREF="http://www.mirana.com/page1.html" target="_top" onMouseOver="window.status='Click here for page 1'; return true;" onMouseOut ="window.status=''; return true;">Click Here!</a>

The semicolon is required to separate the two JavaScript statements in the above example. Although semicolons are optional, they are recommended both for debugging and readabilitiy reasons and for consistency with other programming languages.

Preface the quotation mark with a backslash("\"). In fact, the backslash can be used before any character to tell JavaScript to print it literally instead of using it as part of a statement. For example:

<A HREF="http://www.mirana.com/pages/pages2.html" target="_top" onMouseOver="window.status=\"Click Here for more pages \"; return true;" onMouseOut ="window.status=\"\"; return true;">Click for more, here!</a>

 

A variable created in JavaScript is associated with a document loaded into the current browser window. The variable appears to any code executing within the same document object scope (e.g. within event handlers, functions, and code within the body). It's good to remember that a variable created by your code is nothing more than a property of the current document object. As such, it is not visible, to other frames that share the same frameset, or other windows -- even those opened by the current document.

One way to overcome this is to use a frameset document and define your variables within the 'parent' document. In doing so the parent will always be in memory while the other frames are used to load into them HTML documents, etc. Variables defined by other documents are NOT global. Hence, in order to access a variable defined in the another frame an appropriate expression must be used. The following expression is an example:

parent.frames[n].document.variableName;

This also applies when the window object has been used to open a window:

windowName.document.variableName

 

Simply, var is used within a function to create a variable that is local to that function.

 For example:

function myFn() { x = 5; // creates global variable x. var y = 10; // creates local variable y (independent of global variable y). var z = 20; // creates local variable z. } y = 1;&nbsp;&nbsp;&nbsp; // creates global variable y (independent of local variable y). myFn(); // at this point (globally): // x: 5; // y: 1; <-- not 10! // z: undefined

 

No. JavaScript does not expose any control over how or what Netscape prints.

1. Why does document.write() open a new window instead of writing to the current one?
2. How can I load a frame without additional files?
3. How do I reference functions in a different frame?
4. How do I change two frames simultaneously with one click?
5. How can code in one frame determine if code in the other frames in my frameset have finished loading?
6. Why does my call to window.open() not display the window type and size I specify?
7. How can I create a modal window (one that can't be minimized) using JavaScript?
8. Can I write JavaScript that will execute when the user closes a window?
9. How can I control the placement of a new window using window.open()?
10. How can I close a window without causing a confirm prompt to appear?
11. I have a pop-up window for a menu. How can my selections from this menu open documents in a full size window?
12. Can I reuse the same pop-up window on different pages?

This is, by far, the most popular misunderstanding among JavaScript programmers.

Once the current document has completed loading, you absolutely can not use document.write() to modify its contents. The default behaviour of document.write(), once the current document is done loading, is to open a new window of the type text/html, and deliver your output to that window.

The rule to remember is that the only opportunity you have to modify the display of the currently loaded document object is:

1. The contents of form elements
2. Document characteristics that are exposed via the document object model (like bgcolor)
3. Screen real estate in which applets or plugins are running, and,
4. Layers, which have their own document object, and can thus be re-generated independent of your script's document object.

How do I load an empty frame? This is accomplished by the use of a variable then using JavaScript: URL to load the frame then referring to this variable via the frame tag's src attribute.

<script language=JavaScript> var emptyFrame = "<html><body>some <B>HTML</B> text</body></html>"; // or // var emptyFrame = "<html></html>"; </script> ... <frame src="javascript:parent.emptyFrame ">

For more information on using JavaScript with multiple frames, take a look at:
1. Netscape Introduction to Frames
2. JavaScript and Netscape Frames
3. Bill Dortch's hIdaho Frameset

It is best to define functions within a single document, parent frameset-document, meaning the HTML document that defines the frameset(s). Then a function may be referenced by using the following

parent.frameName.function();

You can also execute functions in other frames by using the

windowReference.function()

syntax, but be aware of the frame synchronization issues you'll face, as described below.

The best way to do this is to have a function that will do the dirty work for you:

function newPages() { parent.frameA.document.location.href="http://....."; parent.frameB.document.location.href="http://....."; }

Then you call this function via any object that supports an onClick event:

<a href="JavaScript:newPages( )">Continue to the new pages</a>

Note: If one of the pages you wish to change is the one with this function in it, you should put a setTimeout() on the command that changes the local page:

function newPages() { parent.frameA.document.location.href="http://....."; setTimeout('self.location.href="http://....."', 250); }

This way all of your other calls are executed before you close this page. Naturally, your specific needs will differ, and this is only a suggestion.

Often, while working with multi-framed documents, it is necessary to access data in objects belonging to sibling frames. It is conceivable that during the initial build of a frameset, while the child frames are loading, frame A might request data from frame B only to find that frame B has not finished loading or executing its code. This synchronization problem often results in errors similar to "object has no property indexed by zero" or "parent.frameB.variableName is undefined".

To resolve this, one can implement onLoad handlers in the frameset so that any required cross-boundary code is executed only after the entire document set is prepared:

frameset.html:

<frameset rows="50%,50%" onLoad="alert('Done the frameset')"> <frame name="frameA" src="frameA.html"> <frame name="frameB" src="frameB.html"> </frameset>

frameA.html:

<body bgcolor="white" onLoad="alert('Done A')"> </body>

frameB.html:

<body bgcolor="white" onLoad="alert('Done B')"> </body>

If all goes well, you'll see the alerts from frame A, then B, then from the frameset, because the frameset onLoad event handler files after all frames are done loading.

OnLoad Notes: Some Netscape clients mistakenly call the frameset onLoad function before each child document's onLoad function. In all Netscape clients with version numbers less than 3.0, resizing the browser window generates an onLoad event -- not always a desired action!

It is important to note that the third parameter to the window.open() function can contain no whitespace:

myWindow = window.open("http://www.netscape.com", "myWindowName", "toolbar=yes,location=yes,directories=no,status=yes,menubar=yes,scrollbars=yes,resizable=yes,height=400,width=600");

 

Using Navigator 2.x and 3.x:

<BODY onBlur="self.focus();">

Using a Communicator 4.x signed script:

window.alwaysRaised = true; // Sets window to always on top window.alwaysLowered = true; // Anchors window to desktop

 

Not reliably. You can use the onUnLoad event handler of the window object, but there's no guarantee that your code will finish executing before Navigator 4 shuts down and cuts you off. As of this writing, this was listed in the JavaScript Known Bugs document.

You could, however, use the onUnLoad event handler to execute code in another window, that presumably isn't in the middle of closing, if you can be sure that the other window is still open. This will be unreliable, at best.

This issue will presumably be addressed in a future versions.

This is only possible using Communicator 4.x, and is documented in What's New in JavaScript 1.2 :

windowReference = window.open("document.html","windowName","screenX=20,screenY=30")

 

The "close window" alert will not appear if:
• The window you're trying to close programmically was opened programmically by your script or your script's document OR
• The window you're trying to close is the first document loaded in that window (for example, it was created by a link with a previously-unused target, or all documents have been loaded in the window using location.replace()).
Using Navigator version 4 or later, a signed script with Universal Browser Write privileges can close any window.

In the head of the initial HTML document put a window.name statement.

 
  
   window.name="a_Name" 

Then in the window open statements in the menu use the same name.

 
  
   varName = open("whatever.htm","a_Name",""); 

Yes, use the same name each time in the window open. The window size has to be the same.

 
  
   varName = open("whatever.htm","same_Name",""); 

When you want to close the pop-up put a window open just before the close.

 
  
   varName = open("whatever.htm","same_Name",""); 
   varName.close() 

1. How can I use a select box as a navigation menu?
2. How can I use an image for a submit button?
3. How can I pass data between forms on different pages using JavaScript?
4. Why doesn't document.formName.selectObject.value reflect the value of the selected item in the list?
5. How do I get the value of the currently selected radio button in a radio group or group of checkboxes?
6. How do I get a form to submit with the Enter key?
7. How can I disable a text input field?

Using some simple JavaScript, you can use a select box for site navigation. Consider this example:

<FORM name="navForm"> <select name="menu" onChange = "self.location = document.navForm.menu[document.navForm.menu.selectedIndex].value;"> <option value="home.html">Home</option> <option value="links.html">Links</option> <option value="contact.html">Contact Info</option> </select>

Simply store the URL's of various pages in the value of the option's. The onChange event handler will cause the location to be redirected when the value in the box is changed.

You could use this code to change the location of another frame by changing

self.location = to top.framename.location =

Often, the question arises: "How does one submit a form with an image, not a submit button?" The solution is as simple as this example shows:

<form> <input type="image" name="cancel" src="cancel.gif" alt="Cancel!" border=0> <input type="image" name="continue" src="continue.gif" alt="Click to Continue" border=0> </form>

When either button is clicked, the form is submitted, and the x/y coordinates of the mouse click on the button are loaded into the request object. For example, if the user had selected Cancel!, the request object would contain cancel.x and cancel.y variables. Similarly, a click on Continue, would have resulted in continue.x and continue.y variables.

What is important to note here is that the .x and .y elements are not variables on a continue or cancel object but are actually part of the name: "continue.x ". Consequently, one cannot use the typical method for determining the existence of one of the request object's children:

if (request.normalItem) act(); if (request.continue.x) // Noooo! Produces horrible error messages! doContinuingThings();

However, consider this:

if (request['normalItem']) act(); if (<B>request['continue.x']</B>) // Much better :) doContinuingThings();

Thanks to JavaScript's named arrays, it is possible to check which image submitted the form.

For some applications, it is useful to know the coordinates of the cursor on the image clicked. That information, of course, is contained in the image_element.x and image_element.y pairs. The following Server-Side JavaScript could be used to detect this::

x=0; y=0; if ((request['my_submit_image.x']) (request['my_submit_image.y'])) { x = parseInt(request['my_submit_image.x']); y = parseInt(request['my_submit_image.y']); }

This addresses the issue of maintaining state between various forms.

As a simple example, suppose you a 'home' page to ask a user for his name, then use that name to refer to the user on later pages. You can collect the name using a form, then use JavaScript to pass the username to the next page using a decorated URL. The subsequent page could then parse the username out of the URL using the information in document.search .

home.html

<html> <head> <script language="JavaScript"> function nextPage() { self.location = "next.html?name=" + escape(document.theForm.userName.value); // Use escape() any time there might be spaces or } // non-alpa characters </script> </head> <body> <form onSubmit = "nextPage();return false;"> Enter your name: <input type=text name=userName> <input type=submit> </form> </body> </html>

next.html

userName = document.search; userName = userName.substring(userName.indexOf("=")+1) // Get rid of 'name=' document.write("Greetings, " + userName + "<P>");

Because that's not the way the object works. The correct and complete syntax for accessing the VALUE of the currently SELECTED item in a list is:
document.form.selectObject[document.form.selectObject.selectedIndex].value

. There is a property created in the form object for each set of radio buttons or checkboxes with the same name. For example, the following HTML code:

<form name='theForm'> <input type=radio name="gender" value="Male">Male <input type=radio name="gender" value="Female">Female <input type=radio name="gender" value="Evasive">Not Specified </form>

results in the creation of a 3 element array referenced by document.theForm.gender. In order to determine the value of the presently selected button (or checkbox), you'll need to examine each elements checked property. For example:

function checkIt() { theGroup = document.theForm.gender; for (i=0; i<= theGroup.length; i++) { if (theGroup[i].checked) { alert("The value is " + theGroup[i].value); break; } } }

Netscape will submit a form if the enter key is pressed while the only text input element of the form has the focus. You can affect simliar behavior in a form with more than one element by breaking the form into a number of separate forms, so that each form has only one text element. Use the onSubmit event handler (or action="javascript:myFunction( );") to gather the data from the other forms in your page and submit it all at once.

5.7: How can I disable a text input field?

Use the onfocus handler to call the blur() function:

<INPUT TYPE="text" NAME="aTextField" ONFOCUS="this.blur()">

If you want to disable/enable dynamically use function skip (e)

{ this.blur(); }
  <A HREF="javascript:document.formName.aTextField.onfocus = skip; void
0">disable text field</A>
  <A HREF="javascript:document.formName.aTextField.onfocus = null; void
0">enable text field</A> 

1. What is a signed script?
2. Access disallowed from scripts at one.ingenia.com to documents at two.ingenia.com?
3. Why aren't I able to read the URL history using the history object?

A signed script is a script which has a digital signature associated with it, and which can therefore request additional privileges from the user. Each privilege has a distinct name, such as "UniversalBrowserWrite" -- allowing almost total control over the browser's behaviour -- or "UniversalBrowserRead" -- allowing the script to read data from other pages, preferences and so forth. There is information about using signed scripts at http://developer.netscape.comhttp://developer.netscape.com/docs/manuals/communicator/jssec/index.htm. The section on "codebase principals" may be of special interest to intranet developers and others who wish to allow privileges to be requested without the hassle of signatures. Danny Goodman's article at http://developer.netscape.com/news/viewsource/goodman_sscripts.html is also recommended reading.

This error message indicates that a section of JavaScript code is attempting to access properties that it does not own. For example, imagine a framed document containing two pages: "http://one.ingenia.com/page1.html " and "http://two.ingenia.com/page2.html ". If otherLocation = parent.page2.location.href was to be executed by the first page, the Access Disallowed error would result, as page1.html is trying to get information from two.ingenia.com to one.ingenia.com.

Please see the relevant section of the Netscape JavaScript Guide for more information on using data tainting for security .

Unless your script is signed, or operating with privileges granted by codebase principals, JavaScript only allows you to access URL history of entries within the current domain. Check the references in first section, above, for more information.

1. How can I change the user's preferences using JavaScript?
2. How can I make Socket calls in JS?
3. Can I use JavaScript to read the source of an HTML page?

You can use LiveConnect to communicate with Navigator:

netscape.security.PrivilegeManager.enablePrivilege('UniversalPreferencesRead'); var home = navigator.preference('browser.startup.homepage'); if (home != 'http://www.someserver.com/') { netscape.security.PrivilegeManager.enablePrivilege('UniversalPreferencesWrite'); navigator.preference('browser.startup.homepage','http://www.someserver.com/'); }

The Communicator Preferences document describes the names and functions of the different preferences.

Use this code as an example which writes a http get request and reads the http status returned:

 

function fetchHTTPStatus (url) { if ((location.host == '') || (url.indexOf(location.host) == -1)) netscape.security.PrivilegeManager.enablePrivilege('UniversalConnect'); var urlObj = new java.net.URL (url); var host = urlObj.getHost(); var port = (urlObj.getPort() > 0) ? urlObj.getPort() : 80; var fileName = urlObj.getFile(); var sock = new java.net.Socket (host, port); dock = new java.io.DataOutputStream(sock.getOutputStream()); dock.writeBytes('GET ' + fileName + ' HTTP/1.0\r\n'); // could use HEAD dock.writeBytes('\r\n'); var dis = new java.io.DataInputStream(sock.getInputStream()); line = dis.readLine(); // get just status message dis.close(); dock.close(); sock.close(); return line; } // example call alert(fetchHTTPStatus('http://www.someserver.com/someUrl.html')) alert(responseInt)

 

 

function fetchURL(url) { if ((location.host == '' && url.indexOf(location.protocol) == -1) || url.indexOf(location.host) == -1) { netscape.security.PrivilegeManager.enablePrivilege("UniversalConnect"); } var dest = new java.net.URL(url); var dis = new java.io.DataInputStream(dest.openStream()); var res = ""; while ((line = dis.readLine()) != null) { res += line; res += java.lang.System.getProperty("line.separator"); } dis.close(); return res; } //example call alert(fetchURL(location.href)); }