Dave Raggett Getting started with HTML

Dave Raggett, revised 13th February 2002.

This is a short introduction to writing HTML. Many people still write HTML by hand using tools such as NotePad on Windows, or SimpleText on the Mac. This guide will get you up and running. Even if you don't intend to edit HTML directly and instead plan to use an HTML editor such as Netscape Composer, or W3C's Amaya, this guide will enable you to understand enough to make better use of such tools and how to make your HTML documents accessible on a wide range of browsers. Once you are comfortable with the basics of authoring HTML, you may want to learn how to add a touch of style using CSS, and to go on to try out features covered in my page on advanced HTML

A convenient way to automatically fix markup errors is to use the HTML Tidy utility. This also tidies the markup making it easier to read and easier to edit. I recommend you regularly run Tidy over any markup you are editing. Tidy is very effective at cleaning up markup created by authoring tools with sloppy habits.

p.s. a good way to learn is to look at how other people have coded their html pages. To do this, click on the "View" menu and then on "Source". Try it with this page to see how I have applied the ideas I explain below. You will find yourself developing a critical eye as many pages look rather a mess under the hood!

This page will teach you how to:

If you are looking for something else, try the advanced HTML page.

Start with a title

Every HTML document needs a title. Here is what you need to type:

<title>My first HTML document</title>

Change the text from "My first HTML document" to suit your own needs. The title text is preceded by the start tag <title> and ends with the matching end tag </title>. The title should be placed at the beginning of your document.

To try this out, type the above into a text editor and save the file as "test.html", then view the file in a web browser. If the file extension is ".html" or ".htm" then the browser will recognize it as HTML. Most browsers show the title in the window caption bar.

Add headings and paragraphs

If you have used Microsoft Word, you will be familiar with the built in styles for headings of differing importance. In HTML there are six levels of headings. H1 is the most important, H2 is slightly less important, and so on down to H6, the least important.

Here is how to add an important heading:

<h1>An important heading</h1>

and here is a slightly less important heading:

<h2>A slightly less important heading</h2>

Each paragraph you write should start with a <p> tag. The </p> is optional, unlike the end tags for elements like headings. For example:

<p>This is the first paragraph.</p>

<p>This is the second paragraph.</p>

Adding a bit of emphasis

You can emphasize one or more words with the <em> tag, for instance:

This is a really <em>interesting</em> topic!

Adding interest to your pages with images

Images can be used to make your Web pages distinctive and greatly help to get your message across. The simple way to add an image is using the <img> tag. Let's assume you have an image file called "peter.jpg" in the same folder/directory as your HTML file. It is 200 pixels wide by 150 pixels high.

<img src="peter.jpg" width="200" height="150">

The src attribute names the image file. The width and height aren't strictly necessary but help to speed the display of your Web page. Something is still missing! People who can't see the image need a description they can read in its absence. You can add a short description as follows:

<img src="peter.jpg" width="200" height="150"
alt="My friend Peter">

The alt attribute is used to give the short description, in this case "My friend Peter". For complex images, you may need to also give a longer description. Assuming this has been written in the file "peter.html", you can add one as follows using the longdesc attribute:

<img src="peter.jpg" width="200" height="150"
alt="My friend Peter" longdesc="peter.html">

You can create images in a number of ways, for instance with a digital camera, by scanning an image in, or creating one with a painting or drawing program. Most browsers understand GIF and JPEG image formats, newer browsers also understand the PNG image format. To avoid long delays while the image is downloaded over the network, you should avoid using large image files.

Generally speaking, JPEG is best for photographs and other smoothly varying images, while GIF and PNG are good for graphics art involving flat areas of color, lines and text. All three formats support options for progressive rendering where a crude version of the image is sent first and progressively refined.

Adding links to other pages

What makes the Web so effective is the ability to define links from one page to another, and to follow links at the click of a button. A single click can take you right across the world!

Links are defined with the <a> tag. Lets define a link to the page defined in the file "peter.html":

This a link to <a href="peter.html">Peter's page</a>.

The text between the <a> and the </a> is used as the caption for the link. It is common for the caption to be in blue underlined text.

To link to a page on another Web site you need to give the full Web address (commonly called a URL), for instance to link to www.w3.org you need to write:

This is a link to <a href="http://www.w3.org/">W3C</a>.

You can turn an image into a hypertext link, for example, the following allows you to click on the company logo to get to the home page:

<a href="/"><img src="logo.gif" alt="home page"></a>

Three kinds of lists

HTML supports three kinds of lists. The first kind is a bulletted list, often called an unordered list. It uses the <ul> and <li> tags, for instance:

<ul>
  <li>the first list item</li>

  <li>the second list item</li>

  <li>the third list item</li>
</ul>

Note that you always need to end the list with the </ul> end tag, but that the </li> is optional and can be left off. The second kind of list is a numbered list, often called an ordered list. It uses the <ol> and <li> tags. For instance:

<ol>
  <li>the first list item</li>

  <li>the second list item</li>

  <li>the third list item</li>
</ol>

Like bulletted lists, you always need to end the list with the </ol> end tag, but the </li> end tag is optional and can be left off.

The third and final kind of list is the definition list. This allows you to list terms and their definitions. This kind of list starts with a <dl> tag and ends with </dl> Each term starts with a <dt> tag and each definition starts with a <dd>. For instance:

<dl>
  <dt>the first term</dt>
  <dd>its definition</dd>

  <dt>the second term</dt>
  <dd>its definition</dd>

  <dt>the third term</dt>
  <dd>its definition</dd>
</dl>

The end tags </dt> and </dd> are optional and can be left off. Note that lists can be nested, one within another. For instance:

<ol>
  <li>the first list item</li>

  <li>
    the second list item
    <ul>
      <li>first nested item</li>
      <li>second nested item</li>
    </ul>
  </li>

  <li>the third list item</li>
</ol>

You can also make use of paragraphs and headings etc. for longer list items.

Getting Further Information

If you are ready to learn more, I have prepared some accompanying material on advanced HTML and adding a touch of style.

W3C's Recommendation for HTML 4.0 is the authorative specification for HTML. However, it is a technical specification. For a less technical source of information you may want to purchase one of the many books on HTML, for example "Raggett on HTML 4", published 1998 by Addison Wesley. See also "Beginning XHTML", published 2000 by Wrox Press, which introduces W3C's reformulation of HTML as an application of XML. XHTML 1.0 is now a W3C Recommendation.

More advanced features

Dave Raggett, 29th August 2000.

Having mastered the basics, it is time to move on to more advanced features. The following will teach you how to:

p.s. I recommend you use HTML Tidy to keep your markup clean and free of errors.

How to force line breaks

Just occasionally, you will want to force a line break. You do this using the br element, for example when you want to include a postal address:

<p>The Willows<br>
21 Runnymede Avenue<br>
Morton-in-the-marsh<br>
Oxfordshire OX27 3BQ</p>

The br element never needs an end-tag. In general, elements that don't take end-tags are known as empty elements.

How to introduce non-breaking spaces

Browsers automatically wrap text to fit within the margins. Line breaks can be introduced wherever space characters appear in the markup. Sometimes you will want to prevent the browser from wrapping text between two particular words. For instance between two words in a brand name such as "Coca Cola". The trick is to use &nbsp; in place of the space character, for example:

Sweetened carbonated beverages, such as Coca&nbsp;Cola,
have attained world-wide popularity.

It is bad practice to use repeated non-breaking spaces to indent text. Instead, you are advised to set the indent via style rules.

How to use entities for special characters

For copyright notices, or trademarks it is customary to include the appropriate signs:

Symbol Entity Example
Copyright sign &copy; Copyright © 1999 W3C
Registered trademark &reg; MagiCo ®
Trademark &#8482; Webfarer™

Note HTML 4.0 defines &trade; for the trademark sign but this is not yet as widely supported as &#8482;

There are a number of other entities you may find useful:

Symbol Entity Example
Less than &lt; <
Greater than &gt; >
Ampersand &amp; &
nonbreaking space &nbsp;  
em dash &#8212;
quotation mark &quot; "

And then, there are entities for accented characters and miscellaneous symbols in the Latin-1 character set:

  &nbsp; &#160; Ð &ETH; &#208;
¡ &iexcl; &#161; Ñ &Ntilde; &#209;
¢ &cent; &#162; Ò &Ograve; &#210;
£ &pound; &#163; Ó &Oacute; &#211;
¤ &curren; &#164; Ô &Ocirc; &#212;
¥ &yen; &#165; Õ &Otilde; &#213;
¦ &brvbar; &#166; Ö &Ouml; &#214;
§ &sect; &#167; × &times; &#215;
¨ &uml; &#168; Ø &Oslash; &#216;
© &copy; &#169; Ù &Ugrave; &#217;
ª &ordf; &#170; Ú &Uacute; &#218;
« &laquo; &#171; Û &Ucirc; &#219;
¬ &not; &#172; Ü &Uuml; &#220;
­ &shy; &#173; Ý &Yacute; &#221;
® &reg; &#174; Þ &THORN; &#222;
¯ &macr; &#175; ß &szlig; &#223;
° &deg; &#176; à &agrave; &#224;
± &plusmn; &#177; á &aacute; &#225;
² &sup2; &#178; â &acirc; &#226;
³ &sup3; &#179; ã &atilde; &#227;
´ &acute; &#180; ä &auml; &#228;
µ &micro; &#181; å &aring; &#229;
&para; &#182; æ &aelig; &#230;
· &middot; &#183; ç &ccedil; &#231;
¸ &cedil; &#184; è &egrave; &#232;
¹ &sup1; &#185; é &eacute; &#233;
º &ordm; &#186; ê &ecirc; &#234;
» &raquo; &#187; ë &euml; &#235;
¼ &frac14; &#188; ì &igrave; &#236;
½ &frac12; &#189; í &iacute; &#237;
¾ &frac34; &#190; î &icirc; &#238;
¿ &iquest; &#191; ï &iuml; &#239;
À &Agrave; &#192; ð &eth; &#240;
Á &Aacute; &#193; ñ &ntilde; &#241;
 &Acirc; &#194; ò &ograve; &#242;
à &Atilde; &#195; ó &oacute; &#243;
Ä &Auml; &#196; ô &ocirc; &#244;
Å &Aring; &#197; õ &otilde; &#245;
Æ &AElig; &#198; ö &ouml; &#246;
Ç &Ccedil; &#199; ÷ &divide; &#247;
È &Egrave; &#200; ø &oslash; &#248;
É &Eacute; &#201; ù &ugrave; &#249;
Ê &Ecirc; &#202; ú &uacute; &#250;
Ë &Euml; &#203; û &ucirc; &#251;
Ì &Igrave; &#204; ü &uuml; &#252;
Í &Iacute; &#205; ý &yacute; &#253;
Î &Icirc; &#206; þ &thorn; &#254;
Ï &Iuml; &#207; ÿ &yuml; &#255;

You can also use numeric character entities for the greek letters and mathematical symbols defined in Unicode. For more details, take a look at the list specified in the HTML 4 specification. Note that the entity names for these characters aren't recognized in Navigator 4, so you are recommended to stick to the numeric entities instead.

Linking into the middle of Web pages

Imagine you have written a long Web page with a table of contents near the start. How do you make the entries in the table contents into hypertext links to the corresponding sections?

Let's assume that each section starts with a heading, for instance:

<h2>Local Night Spots</h2>

You make the heading into a potential target for a hypertext link by enclosing its contents with <a name=identifier> .... </a>

<h2><a name="night-spots">Local Night Spots</a></h2>

The name attribute specifies the name you will use to identify the link target, in this case: "night-spots". The table of contents can now include a hypertext link using this name, for instance:

<ul>
  ...
  <li><a href="#night-spots">Local Night Spots</a></li>
  ...
</ul>

The # character is needed before the target name. If the target is in a different document, then you need to place the web address of that document before the # character. For example if the document is in "http://www.bath.co.uk/", then the link becomes:

<a href="http://www.bath.co.uk/#night-spots">Local Night Spots</a>

In the future, you will eventually be able to define link targets without the need for the <a> element. The new method is much easier, as all you need to do is to add an id attribute to the heading, for instance:

<h2 id="night-spots">Local Night Spots</h2>

This method doesn't work for 4th generation or earlier browsers, so it should be used with care while these browsers are still in use!

Preformatted Text

One of the advantage of the Web is that text is automatically wrapped into lines fitting within the current window size. Sometimes though, you will want to disable this behavior. For example when including samples of program code. You do this using the pre element. For instance:

<pre>
    void Node::Remove()
    {
        if (prev)
            prev->next = next;
        else if (parent)
            parent->SetContent(null);

        if (next)
            next->prev = prev;

        parent = null;
    }
</pre>

Which renders as:

    void Node::Remove()
    {
        if (prev)
            prev->next = next;
        else if (parent)
            parent->SetContent(null);

        if (next)
            next->prev = prev;

        parent = null;
    }

The text and background colors were set via the style sheet. Note that all line breaks and spaces are rendered exactly as they appear in the HTML. The exception is a newline immediately after the start tag <pre> and immediately before the end tag </pre>, which are discarded. This means that the following two examples are rendered identically:

<pre>preformatted text</pre>

<pre>
preformatted text
</pre>

Preformatted text is generally rendered using a monospaced font where each character has the same width. If you define a style rule for the pre element, some browsers forget to use the monospace font, necessitating the use of the font-family property. For instance if you want to render all pre elements in green you can define the style rule:

<style type="text/css">
  pre { color: green; background: white; font-family: monospace; }
</style>

When setting the foreground color for text, you are advised to also set the color for the background. This will prevent accidents where the background color is hard to distinguish from the foreground. Rather than setting the background color on the pre element, you may find it more convenient to set it on the body element, for instance:

<style type="text/css">
  body { color: black; background: white; }
  pre { color: green; font-family: monospace; }
</style>

Flowing text around images

With HTML, you can choose whether any given image is treated as part of the current text line or is floated to the left or right margins. You control this via the align attribute. If the align attribute is set to left the image floats to the left margin. If it is set to right the image floats to the right margin. For instance:

<p><img src="sun.jpg" alt="sunburst graphic"
width="32" height="21" align="left"> This text will be
flowed around the right side of the graphic.</p>

which renders as:

sunburst graphic This text will be flowed around the right side of the graphic.

The following uses align="right"

<p><img src="sun.jpg" alt="sunburst graphic"
width="32" height="21" align="right"> This text will be
flowed around the left side of the graphic.</p>

which renders as:

sunburst graphic This text will be flowed around the left side of the graphic.

To force rendering to continue below the floated image you can use the <br clear=all> element, for example:

<p><img src="sun.jpg" alt="sunburst graphic"
width="32" height="21" align="left"> This text will be
flowed around the right side of the graphic.<br clear="all">
This starts a new line below the floated image.</p>

which renders as:

sunburst graphic This text will be flowed around the right side of the graphic.
This starts a new line below the floated image.

Clickable regions within images

The following image acts as a map of a group of Web pages. You can click on the circles to go to the corresponding page.

site map Getting Started A Touch of Style Web Page Design Advanced HTML

The markup for this is as follows:

<p align="center">
   <img src="pages.gif" width="384" height="245"
      alt="site map" usemap="#sitemap" border="0">
   <map name="sitemap">
      <area shape="circle" coords="186,44,45" 
         href="Overview.html" alt="Getting Started">
      <area shape="circle" coords="42,171,45" 
         href="Style.html" alt="A Touch of Style">
      <area shape="circle" coords="186,171,45"
         alt="Web Page Design">
      <area shape="circle" coords="318,173,45"
         href="Advanced.html" alt="Advanced HTML">
   </map>
</p>

The src attribute on the img element specifies the image "pages.gif". The usemap attribute references a map element. It uses a Web address to do so, hence the # character. The border attribute is set to "0" to suppress the blue border around the image.

The map element specifies which regions in the image act as hypertext links. The name attribute matches usemap attribute on the img element and acts much like the name attribute on the <a> element. In practice, the map element needs to be in the same file as the img element.

The area element is used to define a region on the image and to bind it to a Web address. The shape attribute specifies "rect", "circle" or "poly". The coords attribute specifies the coordinates for the region depending on the shape.

The top left pixel is considered as the origin of the image with x and y both equal to zero, x increases rightwards across the image and y increases downwards. Most image manipulation tools allow you to find the pixel coordinates of any given point in the image.

If two or more defined regions overlap, the region-defining element that appears earliest in the document takes precedence (i.e., responds to user input). For a complex shape such as an anular ring, you can make part of a region inactive by overlaying it with another region using the nohref attribute, for example:

      <area shape="circle" coords="186,44,50" nohref>
      <area shape="circle" coords="186,44,100" 
         href="Overview.html" alt="Getting Started">

Where the first circle creates an inactive region within the larger circle created by the second area element. To have any effect, the inactive shape needs to be placed first as otherwise it will be hidden by the active shape.

Why you need to specify the alt attribute

The alt attribute on the area element is used to supply a text label for the link. Without it the image map is inaccessible to people who can't see the image.

Tables

Tables are used for information as well as for layout. You can stretch tables to fill the margins, specify a fixed width or leave it to the browser to automatically size the table to match the contents.

Tables consist of one or more rows of table cells. Here is a simple example:

Year Sales
2000 $18M
2001 $25M
2002 $36M

The markup for this is:

<table border="1">
<tr><th>Year</th><th>Sales</th></tr>
<tr><td>2000</td><td>$18M</td></tr>
<tr><td>2001</td><td>$25M</td></tr>
<tr><td>2002</td><td>$36M</td></tr>
</table>

The table element acts as the container for the table. The border attribute specifies the border width in pixels. The tr element acts as a container for each table row. The th and td elements act as containers for heading and data cells respectively.

Cell Padding

You can increase the amount of padding for all cells using the cellpadding attribute on the table element. For instance, to set the padding to 10 pixels:

<table border="1" cellpadding="10">

this has the effect:

Year Sales
2000 $18M
2001 $25M
2002 $36M

Cell Spacing

By contrast the cellspacing attribute sets the space between the cells. Setting the cell spacing to 10:

<table border="1" cellpadding="10" cellspacing="10">

has the effect:

Year Sales
2000 $18M
2001 $25M
2002 $36M

Table Width

You can set the width of the table using the width attribute. The value is either the width in pixels or a percentage value representing the percentage of the space available between the left and right margins. For instance to set the width to 80% of the margins:

<table border="1" cellpadding="10" width="80%">

which has the effect:

Year Sales
2000 $18M
2001 $25M
2002 $36M

Text Alignment within Cells

By default browsers center heading cells (th), and left align data cells (td). You can change alignment using the align attribute, which can be added to each cell or to the row (tr element). It is used with the values "left", "center" or "right":

<table border="1" cellpadding="10" width="80%">
<tr align="center"><th>Year</th><th>Sales</th></tr>
<tr align="center"><td>2000</td><td>$18M</td></tr>
<tr align="center"><td>2001</td><td>$25M</td></tr>
<tr align="center"><td>2002</td><td>$36M</td></tr>
</table>

with the following result:

Year Sales
2000 $18M
2001 $25M
2002 $36M

The valign attribute plays a similar role for the vertical alignment of cell content. It is used with the values "top", "middle" or "bottom", and can be added to each cell or row. By default, heading cells (th) position their content in the middle of the cells while data cells align their content at the top of each cell.

Empty Cells

One quirk is the way browsers deal with empty cells, compare:

Year Sales
2000 $18M
2001 $25M
2002 $36M
2003

with

Year Sales
2000 $18M
2001 $25M
2002  

The former occurs when a cell is empty:

<td></td>

To prevent this, include a non-breaking space:

<td>&nbsp;</td>

Cells that span more than one row or column

Let's extend the above example to break out sales by north and south sales regions:

Year Sales
North South Total
2000 $10M $8M $18M
2001 $14M $11M $25M

The heading "Year" now spans two rows, while the heading "Sales" spans three columns. This is done by setting the rowspan and colspan attributes respectively. The markup for the above is:

<table border="1" cellpadding="10" width="80%">
<tr align="center"><th rowspan="2">Year</th><th colspan="3">Sales</th></tr>
<tr align="center"><th>North</th><th>South</th><th>Total</th></tr>
<tr align="center"><td>2000</td><td>$10M</td><td>$8M</td><td>$18M</td></tr>
<tr align="center"><td>2001</td><td>$14M</td><td>$11M</td><td>$25M</td></tr>
</table>

You can simplify this by taking advantage of the fact that browsers don't need the end tags for table cells and rows:

<table border="1" cellpadding="10" width="80%">
<tr align="center"><th rowspan="2">Year<th colspan="3">Sales
<tr align="center"><th>North<th>South<th>Total
<tr align="center"><td>2000<td>$10M<td>$8M<td>$18M
<tr align="center"><td>2001<td>$14M<td>$11M<td>$25M
</table>

Notice that as the heading "Year" spans two rows, the first th element on the second row appears on the second rather than the first column.

Borderless tables

These are commonly used for laying out pages in a gridded fashion. All you need to do is to add border="0" and cellspacing="0" to the table element:

Year Sales
2000 $18M
2001 $25M
2002 $36M

This was produced using the following markup:

<table border="0" cellspacing="0" cellpadding="10">
<tr><th>Year</th><th>Sales</th></tr>
<tr><td>2000</td><td>$18M</td></tr>
<tr><td>2001</td><td>$25M</td></tr>
<tr><td>2002</td><td>$36M</td></tr>
</table>

If you leave out the cellspacing attribute you will get a thin gap between the cells, as shown below:

Year Sales
2000 $18M
2001 $25M
2002 $36M

Coloring your tables

This page uses a style sheet to set the background colors for tables, with a different color for heading and data cells. The style rules I used are as follows:

table {
  margin-left: -4%
  font-family: sans-serif;
  background: white;
  border-width: 2;
  border-color: white;
}
th { font-family: sans-serif; background: rgb(204, 204, 153) }
td { font-family: sans-serif; background: rgb(255, 255, 153) }

The last two lines above set the background color for th and td cells to given red/green/blue values. The numbers are in the range 0 to 255 (fully saturated).

Another way to set the background color is to use the bgcolor attribute. This works with nearly all browsers, and doesn't rely on style sheet support. The first step is to determine the hexadecimal values for the red, green and blue components of the color you wish to use. A convertor is included in the style page.

<table border="0" cellspacing="0" cellpadding="10">
  <tr>
    <th bgcolor="#CCCC99">Year</th>
    <th bgcolor="#CCCC99">Sales</th>
  </tr>
  <tr>
    <td bgcolor="#FFFF66">2000</td>
    <td bgcolor="#FFFF66">$18M</td>
  </tr>
  <tr>
    <td bgcolor="#FFFF66">2001</td>
    <td bgcolor="#FFFF66">$25M</td>
  </tr>
  <tr>
    <td bgcolor="#FFFF66">2002</td>
    <td bgcolor="#FFFF66">$36M</td>
  </tr>
</table>

Making your tables accessible

If you are unable to see the table it can be quite hard to understand what the table is about. The first step is to add information describing the purpose and structure of the table. The caption element allows you to provide a caption, and to position this above or below the table. The caption element should appear before the tr element for the first row.

Projected sales revenue by year
Year Sales
2000 $18M
2001 $25M

which was produced by the following markup:

<table border="1" cellpadding="10" width="80%">
<caption>Projected sales revenue by year</caption>
<tr align="center">
  <th>Year</th><th>Sales</th>
</tr>
<tr align="center"><td>2000</td><td>$18M</td></tr>
<tr align="center"><td>2001</td><td>$25M</td></tr>
</table>

Here is the same table with align="bottom" added to the caption element:

Projected sales revenue by year
Year Sales
2000 $18M
2001 $25M

The table element's summary attribute should be used to describe the structure of the table for people who can't see the table. For instance: "the first column gives the year and the second, the revenue for that year".

<table summary="the first column gives the year
and the second, the revenue for that year">

Specifying the relation between header and data cells

When a table is rendered to audio or to Braille, it is useful to be able to identify which headers describe each cell. For instance, an aural browser could allow you to move up and down or left and right from cell to cell, with the appropriate headers spoken before each cell.

To support this you need to annotate the header and/or data cells. The simplest approach is to add the scope attribute to header cells. It may be used with the following values:

Applying this to the example table gives:

<table border="1" cellpadding="10" width="80%">
<caption>Projected sales revenue by year</caption>
<tr align="center">
  <th scope="col">Year</th>
  <th scope="col">Sales</th>
</tr>
<tr align="center"><td>2000</td><td>$18M</td></tr>
<tr align="center"><td>2001</td><td>$25M</td></tr>
</table>

For more complex tables, you can use the headers attribute on individual data cells to provide a space separated list of identifiers for header cells. Each such header cell must have an id attribute with a matching identifier.

A final point is that you should consider using the abbr attribute to specify an abbreviation for long headers. This makes it tolerable to listen to lists of headers for each cell, for instance:

<th abbr="W3C">World Wide Web Consortium</th>

Roll-Overs and other tricks

A little JavaScript can go a long way to enliven your pages. You will be shown below how to create "rollovers" where the appearence of a link changes as you move the mouse over it. You will also learn how to create cycling banner ads which help to direct visitors to your sponsors' sites

Roll-Overs

In the most common form, a roll-over consists of an image serving as a hypertext link. While the mouse pointer is over the image, it changes appearence to attract attention to the link. For example, you could add a glow effect, a drop shadow or simply change the background color. Here is an example:

<script type="text/javascript">
if (document.images)
{
    image1 = new Image;
    image2 = new Image;
    image1.src = "enter1.gif";
    image2.src = "enter2.gif";
}

function chgImg(name, image)
{
    if (document.images)
    {
        document[name].src = eval(image+".src");
    }
}
</script>

...

<a href="/" onMouseOver='chgImg("enter", "image2")'
onMouseOut='chgImg("enter", "image1")'><img name="enter"
src="enter1.gif" border="0" alt="Enter if you dare!"></a>

and here is what it looks like ...

Enter if you dare!

I created these images using a freeware painting tool by adding a hot wax effect and then a drop shadow to the text. You can find lots of advice and royalty free clipart on the Web via most search engines.

Banner Ads

If your website has several sponsors, then you can use an image link that cycles through each of the sponsors in turn. The first step is to create an image for each of your sponsors. All the images should have the same size. The corresponding URLs for the images and for the websites are then placed into the arrays named adImages and adURLs defined at the start of the script. The img element for the link should be initialized to the first image in the array. The cycle is started off using the onload event on the body element.

<html>
<head>
<title>cycling banner ads</title>
<script type="text/javascript">
if (document.images)
{
    adImages = new Array("hosts/mit.gif",
                "hosts/inria.gif", "hosts/keio.gif");
    adURLs = new Array("www.lcs.mit.edu",
                "www.inria.fr", "www.keio.ac.jp");
    thisAd = 0;
}

function cycleAds()
{
    if (document.images)
    {
        if (document.adBanner.complete)
        {
            if (++thisAd == adImages.length)
                thisAd = 0;

            document.adBanner.src = adImages[thisAd];
        }
    }

    // change to next sponsor every 3 seconds
    setTimeout("cycleAds()", 3000);
}

function gotoAd()
{
    document.location.href = "http://" + adURLs[thisAd];
}
</script>
</head>
<body onload="cycleAds()">
...

<a href="javascript:gotoAd()"><img name="adBanner"
src="hosts/mit.gif" border="0" alt="Our sponsors"></a>

Our Sponsors: Our sponsors

Note: you are recommended to make sure that all of the images are the same width and height. An alternative is to add width and height attributes to the img element to ensure the images are all shown at the same size.

What about browsers that don't support scripting?

The content of a noscript element is only shown if the browser doesn't support scripting. It should be used when you want to give people access to information that would otherwise be inaccessible to people with browsers that don't support scripting. Let's assume that you want to make the links for your sponsors available as text:

<noscript>
Our sponsors: <a href="http://www.lcs.mit.edu/">MIT</a>,
<a href="http://www.inria.fr/">INRIA</a>, and
<a href="http://www.keio.ac.jp/">Keio University</a>.
</noscript>

There are many free sources of information about scripting, which can be easily found via most search engines.

Getting Further Information

W3C's Recommendation for HTML 4.0 is the authorative specification for HTML. However, it is a technical specification. For a less technical source of information you may want to purchase one of the many books on HTML, for example "Raggett on HTML 4", published 1998 by Addison Wesley. See also "Beginning XHTML", published 2000 by Wrox Press, which introduces W3C's reformulation of HTML as an application of XML. XHTML 1.0 is now a W3C Recommendation.

Best of luck and get writing!

Dave Raggett <dsr at w3.org>

Copyright  ©  2000 W3C ® (MIT, INRIA, Keio ), All Rights Reserved. W3C liability, trademark, document use and software licensing rules apply. Your interactions with this site are in accordance with our public and Member privacy statements.