<?xml version="1.0" encoding="UTF-8"?>

<CONTENT TITLE="Windows and Frames" ID="windows">

<P>

The <ICODE>window</ICODE> object (also referred to by <ICODE>self</ICODE>) is "DOM Level 0".

No formal standard for it exists.

</P>

<CONTENT TITLE="How can I disable the back button in a web browser?" ID="disableBackButton" NUMID="4_2">

<P>

You can't. The browser's history cannot be modified. However, you

can use <ICODE>self.location.replace(url);</ICODE> in some browsers to replace

the current page in the history.

</P>

<MOREINFO>

<URL>http://msdn.microsoft.com/en-us/library/ms536712%28VS.85%29.aspx</URL>

<URL>http://docs.sun.com/source/816-6408-10/location.htm#1194240</URL>

</MOREINFO>

</CONTENT>

<CONTENT TITLE="How do I access a frame's content?"

ID="frameRef" NUMID="4_8">

<P>

To reference another frame on the <EM>same domain</EM>:

</P>

<P>

The <DFN>content window</DFN> of a <ICODE>FRAME</ICODE> or <ICODE>IFRAME</ICODE> can be

accessed by the <ICODE>frames</ICODE> collection.

</P>

<P>

Example:

</P>

<CODE>

var fwin;

fwin = self.frames[0]; // or:

fwin = self.frames["iframeName"];

</CODE>

<P>or, from the <ICODE>IFRAME</ICODE> or <ICODE>FRAME</ICODE> element:</P>

<CODE>

var iframeEl = document.getElementById("myFrame");

var fwin = iframeEl.contentWindow; // Nonstandard, but widely supported.

var fdoc = iframeEl.contentDocument; // DOM2 HTML Standard.

</CODE>

<P>

A global identifier <ICODE>moomin</ICODE> in the the iframe's <DFN>content window</DFN>

is accessed as <ICODE>fwin.moomin</ICODE>.

</P>

<P>

To communicate between frames on <EM>different</EM> domains:

</P>

<P>

Where supported, (IE8, Firefox 3, Opera 9, Safari 4), use

<ICODE>window.postMessage( message[, port], otherDomain);</ICODE>.

</P>

<P>

Example:

<URL>http://jibbering.com/faq/example/postMessage.html</URL>

</P>

<P>

Where <ICODE>window.postMessage</ICODE> is not supported, the <ICODE>window.name</ICODE> property

can be set on the other window, which can poll for updates to that

property using <ICODE>setInterval(checkWinName, 100);</ICODE> where <ICODE>checkWinName</ICODE>

is a function that polls to check the value of

<ICODE>self.name</ICODE>.

</P>

<MOREINFO>

<URL>http://en.wikipedia.org/wiki/Same_origin_policy</URL>

<URL>http://www-archive.mozilla.org/docs/dom/domref/dom_frame_ref5.html</URL>

<URL>https://developer.mozilla.org/en/DOM/window.postMessage</URL>

<URL>http://msdn.microsoft.com/en-us/library/cc197015(VS.85).aspx</URL>

</MOREINFO>

</CONTENT>

<CONTENT TITLE="How do I find the size of the window?" ID="getWindowSize"

NUMID="4_9">

<P>

Here is a detailed explanation of a cross-browser strategy to

find the dimensions of the viewport, excepting all chrome

(excludes scrollbars, etc).

</P>

<P>

We can consider various properties:

</P>

<CODE>

window.innerWidth

document.clientWidth

document.documentElement.clientWidth

document.body.clientWidth

</CODE>

<P>

Of the browsers that have an <ICODE>innerWidth</ICODE> property, most

include scrollbar dimensions. Some versions of KHTML browsers

(including Safari 2) do <EM>not</EM> include scrollbar width.

</P>

<P>

The <ICODE>window.inner*</ICODE> properties are unreliable and not

useful here. We don't want scrollbar dimensions included.

</P>

<CODE> document.clientWidth</CODE>

<P>

Certain versions of KHTML, including Safari 2, have

<ICODE>document.clientHeight</ICODE> and <ICODE>document.clientWidth</ICODE>

properties. Where supported, these rare properties accurately

return the height and width of the viewport, without including

scrollbar dimensions.

</P>

<CODE>

document.documentElement.clientWidth

document.body.clientWidth

</CODE>

<P>

MSHTML (Trident), Firefox (Gecko), Opera (Presto), and Safari

(Webkit) all support <ICODE>clientHeight</ICODE> on <ICODE>document.body</ICODE>

and <ICODE>document.documentElement</ICODE>. The difficulty is figuring out

which one is reliable. In other words which object to get the

<ICODE>clientHeight</ICODE> property from:<ICODE>documentElement</ICODE> or <ICODE>body</ICODE>?

</P>

<P>

What the number returned from either of these properties

represents depends on the environment. The environment includes

the browser, its version, and the rendering mode of the document.

In quirks mode, we'll mostly want to use <ICODE>body.clientHeight</ICODE>

(except for in Safari 2).

</P>

<CODE>document.body.clientHeight</CODE>

<P>

Some environments will return the viewport height. Others will

return <ICODE>0</ICODE>. Yet others will return the <ICODE>clientHeight</ICODE> of

the <ICODE>BODY</ICODE> element.

</P>

<CODE> document.documentElement.clientHeight</CODE>

<P>

This is the more "standard" property for getting the height of

the viewport. It usually "works" in modern browsers in

<DFN>standards mode</DFN>. Notable exceptions include Safari 2 and

Opera &lt;= 9.25, both of which return the <ICODE>clientHeight</ICODE>

of the <ICODE>html</ICODE> <EM>element</EM>. (Oddly, Opera &lt;= 9.25

in standards mode returns the width of the viewport for

<ICODE>documentElement.clientWidth</ICODE>).

</P>

<P>

With the exception of Safari 2, <ICODE>body.clientHeight</ICODE> is reliable

where <ICODE>documentElement.clientHeight</ICODE> is found to be unreliable.

For example, in Safari 3+, Opera, and Mozilla, all in quirks mode,

<ICODE>document.documentElement.clientHeight</ICODE> returns the <ICODE>clientHeight</ICODE>

of the <ICODE>html</ICODE> element (this may seem unsurprising but

it is not what we want).

</P>

<P>

Conversely, <ICODE>document.body.clientHeight</ICODE> will return

the height of the viewport in most cases where

<ICODE>document.documentElement.clientHeight</ICODE> does not. An exception

to that is Safari 2, where <ICODE>documentElement.clientHeight</ICODE>

and <ICODE>body.clientHeight</ICODE> both return the height of their

corresponding element (not what we want).

</P>

<P>

By using a combination of <DFN>Feature Testing</DFN> and <DFN>Capability Testing</DFN>,

the dimensions of the viewport can be strategically retrieved

from the property that works in the environment the script is

running in. The trick is determining which property will give us

the value we want.

</P>

<P>

Since <ICODE>document.clientHeight</ICODE> is reliable where

(rarely) supported, and since browsers that support this property

don't return the viewport dimensions from

<ICODE>document.body.clientHeight</ICODE> or

<ICODE>document.documentElement.clientHeight</ICODE>, this should be the

very first condition:

</P>

<CODE>

// Safari 2 uses document.clientWidth (default).

if(typeof document.clientWidth == "number") {

// use document.clientWidth.

}

</CODE>

<P>

The next strategy is to determine if

<ICODE>document.documentElement.clientHeight</ICODE> property is unreliable.

It is deemed "unreliable" when it is either <ICODE>0</ICODE> or taller

than the viewport.

</P>

<P>

Determining if <ICODE>documentElement.clientHeight</ICODE> is <ICODE>0</ICODE> is easy.

The result is stored in a variable <ICODE>IS_BODY_ACTING_ROOT</ICODE>.

</P>

<CODE>

var docEl = document.documentElement,

IS_BODY_ACTING_ROOT = docEl &amp;&amp; docEl.clientHeight === 0;

docEl = null;

</CODE>

<P>

To determine if <ICODE>documentElement.clientHeight</ICODE> returns

a value taller than the viewport, we need a <DFN>Capability Test.</DFN>

</P>

<P>

If we can force <ICODE>documentElement</ICODE> to be very tall

(taller than a normal viewport) we can then check to see if

<ICODE>documentElement.clientHeight</ICODE> returns that "very tall" number.

If it does, then it is unreliable.

</P>

<P>

We can force <ICODE>documentElement</ICODE> to be taller than the viewport

(or any "normal" viewport) by adding a <ICODE>div</ICODE> to the <ICODE>body</ICODE>,

give that <ICODE>div</ICODE> a height larger than any normal monitor,

and then check to see if <ICODE>documentElement.clientHeight</ICODE> is

that high (or "almost" that high, to account for <ICODE>documentElement</ICODE>

having a border).

</P>

<CODE>

// Used to feature test Opera returning wrong values

// for documentElement.clientHeight.

// The results of this function should be cached,

// so it does not need to be called more than once.

function isDocumentElementHeightOff(){

var d = document,

div = d.createElement('div');

div.style.height = "2500px";

d.body.insertBefore(div, d.body.firstChild);

var r = d.documentElement.clientHeight &gt; 2400;

d.body.removeChild(div);

return r;

}

</CODE>

<P>

We can use this function to see if we should use

<ICODE>body.clientHeight</ICODE>, instead. (but only after checking if

<ICODE>document.clientHeight</ICODE> is supported).

</P>

<CODE>

// Safari 2 uses document.clientWidth (default).

if(typeof document.clientWidth == "number") {

// use document.clientHeight/Width.

}

else if(IS_BODY_ACTING_ROOT || isDocumentElementHeightOff()) {

// use document.body.clientHeight/Width.

} else {

// use document.documentElement.clientHeight/Width.

}

</CODE>

<P>

The preceding strategy was developed by Garrett Smith with input

from John David Dalton. A complete and tested example can be found

in APE Library under <ICODE>APE.dom.getViewportDimensions</ICODE>.

Source code:

<URL>http://wayback.archive.org/web/20110814142717/http://dhtmlkitchen.com/ape/build/dom/viewport-f.js</URL>.

APE is publicly released under Academic Free License.

APE home: <URL>http://wayback.archive.org/web/20110628160010/http://www.dhtmlkitchen.com/ape/</URL>.

</P>

<P>

Note: The dimensions cannot be determined accurately until after

the document has finished loading.

</P>

<MOREINFO>

<URL>http://msdn.microsoft.com/en-us/library/ms533566%28VS.85%29.aspx</URL>

<URL>http://developer.mozilla.org/en/DOM/window.innerWidth</URL>

<URL>http://dev.opera.com/articles/view/using-capability-detection/</URL>

</MOREINFO>

</CONTENT>

<CONTENT TITLE="How do I check to see if a child window is open, before opening another?"

ID="isWindowOpen" NUMID="4_10">

<CODE>

var myWin;

function openWin(aURL) {

if (!myWin || myWin.closed ) {

myWin = window.open(aURL,'myWin');

} else {

myWin.location.href = aURL;

myWin.focus();

}

}

</CODE>

<P>

Popup windows cause usability problems and are generally best avoided.

</P>

<MOREINFO>

<URL>https://developer.mozilla.org/en/DOM:window.open</URL>

<URL>http://msdn.microsoft.com/en-us/library/ms533574%28VS.85%29.aspx</URL>

<URL>http://docs.sun.com/source/816-6408-10/window.htm#1201877</URL>

<URL>http://www.useit.com/alertbox/990530.html</URL>

</MOREINFO>

</CONTENT>

<CONTENT TITLE="Why does framename.print() not print the correct frame in IE?"

ID="printFrame" NUMID="4_11">

<P>

IE prints the frame that has focus when you call the print

method <ICODE>frameref.focus();frameref.print();</ICODE>

</P>

<MOREINFO>

<URL>http://msdn.microsoft.com/en-us/library/ms976105.aspx</URL>

</MOREINFO>

</CONTENT>

<CONTENT TITLE="How do I close a window and why does it not work on the first one?"

ID="windowClose" NUMID="4_14">

<P>

If a window was opened by javascript, then it can be closed

without confirmation by using <ICODE>windowRef.close()</ICODE>.

</P>

<P>

Before calling <ICODE>windowRef.close()</ICODE> (or other <ICODE>window</ICODE> methods), make

sure the window reference is not null and its <ICODE>closed</ICODE> property is <ICODE>false</ICODE>.

</P>

<P>

Popup windows cause usability problems and are generally best avoided.

</P>

<MOREINFO>

<URL>http://www.useit.com/alertbox/990530.html</URL>

<URL>#isWindowOpen</URL>

<URL>http://msdn.microsoft.com/en-us/library/ms536367%28VS.85%29.aspx</URL>

<URL>https://developer.mozilla.org/en/DOM/window.close#Description</URL>

<URL>http://docs.sun.com/source/816-6408-10/window.htm#1201822</URL>

</MOREINFO>

</CONTENT>

<CONTENT TITLE="Why do I get permission denied when accessing a frame/window?"

ID="permissionDenied" NUMID="4_19">

<P>

In the normal browser security model, a script may only access the

properties of documents served from the same domain or IP address,

protocol, and port.

</P>

<P>

Any attempt to access a property in such cases will result in a &quot;Permission

Denied&quot; error. Signed scripts or trusted ActiveX objects can

overcome this in limited situations.

</P>

<MOREINFO>

<URL>http://msdn.microsoft.com/en-us/library/ms533028%28VS.85%29.aspx</URL>

<URL>https://developer.mozilla.org/En/Same_origin_policy_for_JavaScript</URL>

</MOREINFO>

</CONTENT>

<CONTENT TITLE="How do I make a 10 second delay?" ID="setTimeout" NUMID="4_20">

<P>

There is no built-in way to pause execution in javascript such

as a sleep function, but hosts usually provide a method of some

form. Web browsers are designed for event driven programming and

only provide the <ICODE>setTimeout</ICODE> and <ICODE>setInterval</ICODE> functions

to facilitate timed delays. The delay before calling <ICODE>getSnork</ICODE> may

exceed the second parameter to <ICODE>setTimeout</ICODE> and <ICODE>setInterval</ICODE>

due to implementation differences among browsers.

</P>

<P>

To call the function <ICODE>getSnork</ICODE>, approximately 10 seconds

after the function <ICODE>getMoomin()</ICODE> completes, you would do this:

</P>

<CODE>

getMoomin();

setTimeout(getSnork, 10000);

</CODE>

<P>

Script execution is not stopped, and adding <ICODE>getSnufkin()</ICODE> after the

<ICODE>setTimeout</ICODE> line would immediately execute the function <ICODE>getSnufkin</ICODE>

before <ICODE>getSnork</ICODE>.

</P>

<P>

Achieving delays through running a loop of some sort for a pre-defined

period is a bad strategy, as that will inhibit whatever was supposed to

be happening during the delay, including blocking user interation.

</P>

<P>

Other (less event driven) hosts have different wait functions,

such as <ICODE>WScript.Sleep()</ICODE> in the Windows Script Host.

</P>

<MOREINFO>

<URL>http://msdn.microsoft.com/en-us/library/ms536753%28VS.85%29.aspx</URL>

<URL>http://docs.sun.com/source/816-6408-10/window.htm#1203758</URL>

<URL>http://en.wikipedia.org/wiki/Event-driven_programming</URL>

<URL>faq_notes/misc.html#mtSetTI</URL>

</MOREINFO>

</CONTENT>

<CONTENT TITLE="How do I change print settings for window.print()?"

ID="printSettings" NUMID="4_23">

<P>

In a normal security environment, you can't change anything.

</P>

<P>

Print Stylesheet rules provide options.

</P>

<P>

For IE, <ICODE>ActiveX</ICODE> or Plugin ScriptX and

Neptune from Meadroid to give you more control for Windows

versions of Internet Explorer, Netscape, and Opera.

</P>

<MOREINFO>

<URL>http://www.meadroid.com/scriptx/</URL>

<URL>http://msdn.microsoft.com/en-us/library/ms976105.aspx</URL>

</MOREINFO>

</CONTENT>

<CONTENT TITLE="How do I change the confirm box to say yes/no or default to cancel?"

ID="changeBrowserDialog" NUMID="4_28">

<P>

The buttons on a confirm box cannot be changed, nor can a default

button be specified.

</P>

<P>

Change the question to a statement so that "OK" is suitable as the

default response.

</P>

<P>

Example:

"Would you like us to charge your credit card?" (wrong)

"We will now charge your credit card." (right).

</P>

</CONTENT>

<CONTENT TITLE="How do I prompt a &quot;Save As&quot; dialog for an accepted mime type?"

ID="fileDownload" NUMID="4_33">

<P>

It is not possible with client-side javascript.

</P>

<P>

Some browsers accept the Content-Disposition header, but this

must be added by the server. Taking the form:-

<ICODE>Content-Disposition: attachment; filename=filename.ext</ICODE>

</P>

<MOREINFO>

<URL>http://classicasp.aspfaq.com/general/how-do-i-prompt-a-save-as-dialog-for-an-accepted-mime-type.html</URL>

<URL>http://support.microsoft.com/kb/q260519/</URL>

</MOREINFO>

</CONTENT>

<!-- Can we remove this entry?

http://groups.google.com/group/comp.lang.javascript/msg/c70b57d86300fa91

CONTENT TITLE="I have window.status=&quot;Moomin&quot;; why doesn't the statusbar change?"

ID="returnValue" NUMID="4_35">

<P>

When changing the status in an event (e.g. <ICODE>onmouseover</ICODE>) you

should return true from the event. Also a number of browsers

require a short delay before setting the status to overcome their

default behaviour with the statusbar.

<ICODE>onevent=&quot;setTimeout('window.status=\'Moomin\'',15);&quot;</ICODE>

</P>

<P>

Many browsers are configured, by default, to disallow scripts from setting

the status bar text. Some browsers don't have such a bar, or don't have it by default.

</P>

</CONTENT-->

<CONTENT TITLE="How do I modify the current browser window?" ID="modifyChrome" NUMID="4_36">

<P>

In a default security environment you are very limited in how much

you can modify the current browser window. You can use

<ICODE>window.resizeTo</ICODE> or <ICODE>window.moveTo</ICODE> to resize or move a

window respectively, but that is it. Normally you can only

suggest chrome changes in a <ICODE>window.open</ICODE>.

</P>

<MOREINFO>

<URL>http://msdn.microsoft.com/en-us/library/ms536651%28VS.85%29.aspx</URL>

<URL>https://developer.mozilla.org/en/DOM:window.open</URL>

</MOREINFO>

</CONTENT>

<CONTENT TITLE="How do I POST a form to a new window?" ID="target" NUMID="4_37">

<P>

Use the target attribute on the form, opening a window with

that name and your feature string in the onsubmit handler of the

FORM.

</P>

<CODE>

&lt;form action=&quot;&quot; method="post"

target=&quot;wndname&quot; onsubmit=&quot;window.open('',this.target);return true;&quot;&gt;

</CODE>

<MOREINFO>

<URL>http://www.htmlhelp.com/reference/html40/forms/form.html</URL>

</MOREINFO>

</CONTENT>

<CONTENT TITLE="How do I open a new window with javascript?" ID="openWindow" NUMID="4_42">

<P>

New windows can be opened on browsers that support the

<ICODE>window.open</ICODE> function and are not subject to the action of any

pop-up blocking mechanism with code such as:-

</P>

<CODE>

var wRef;

if(window.open){

wRef = window.open("http://example.com/page.html","windowName");

}

</CODE>

<MOREINFO>

<URL>https://developer.mozilla.org/en/DOM:window.open</URL>

<URL>http://www.infimum.dk/HTML/JSwindows.html</URL>

</MOREINFO>

</CONTENT>

</CONTENT>