Wilbur - HTML 3.2

Syntax rules

In this overview of tags, I have used some simple rules to make the syntax clear. It also allows me to provide the information in a short format.

To illustrate the syntax rules, here's the section on IMG:

The first section, Appearance, gives a common way to use this tag. As you can see here, the IMG tag does not have an ending tag. If the beginning or ending tag appears inside square brackets, it is optional and may be left off.

The next section describes the attributes for the IMG tag. If an attribute appears in bold, it is required, otherwise it may be omitted. In the above case, SRC is required, but the other attributes are not. Note that the attributes themselves are listed in all caps, and the possible values (if possible) in lower case. Note that an attribute value must be quoted if it contains more than just letters, digits, hyphens and periods.

The contents section describes which tags are permitted inside this tag. For IMG, there are none. And last, you can see which tags allow IMG inside them.

The attributes and their values are noted in a very compact format as well. The "|" character is used to separate mutually exclusive attributes or values. For example, A=foo|bar indicates that attribute "A" may get foo or bar as value, but not both, or anything else. A=string|B=string indicates that you may use either A or B, but not both.

If an attribute can take more possible values than can be given in a list, the following special symbols are used:

n

A number. It must be an integer, and not have a "-" or "+" sign prepended. Numbers do not have to be enclosed in quotes.

p%

A percentage. The percentage must also be an integer. Exactly what the percentage applies to depends on the tag. Percentages must be enclosed in quotes.

URL

An URL. This can be an absolute or a relative URL, depending on the situation. In most cases, both are permitted. It is recommended that URLs always be enclosed in quotes.

string

A string of characters. Any character is permitted, including entities. It is recommended that strings are always enclosed in quotes.

#RRGGBB

A color code, in hexadecimal notation. The color is constructed in the red-green-blue format. Each part gets a hexadecimal number between 00 and FF, and it should be given in two digits at all times. Note that a color code must have a # as the first character, and it must be enclosed in quotes.

ADDRESS - Address information

The ADDRESS tag should be used to enclose contact information, addresses and the likes. It is often rendered with a slightly indented left margin and italics.

Notes:

• If you include an address in here, be sure to use BR for explicit linebreaks after every line, otherwise the address won't come out right.

APPLET - Java applet

The APPLET tag is used to include Java applets. The CODE attribute indicates the location of the class of the applet itself. CODEBASE can be used to specify an absolute URL for the applet, similar to the BASE element for HTML documents. Other classes for this applet will be searched at the location indicated in CODEBASE. If this is not specified, the current URL will be used for the location. The NAME attribute gives the name of the applet.

Just like with IMG, WIDTH and HEIGHT are used to specify the width and height of the applet's window, and HSPACE and VSPACE control horizontal and vertical spacing around the applet. ALIGN sets the horizontal or vertical alignment for the applet.

Arguments to the applet can be specified with the PARAM tag, which goes inside the APPLET tag.

The ALT text may contain text which should be displayed if the applet cannot be run, but you should use the contents of APPLET instead. In here you may use markup, so you can provide a better alternative than with the ALT text.

Notes:

• Not all browsers support Java applets, and those that do often allow the user to disable it.
• Always provide alternative text, so your visitors get something when they can't see the applet.

AREA - Client-side imagemap hotspot

Inside the MAP tag, each "hotzone" in the client-side imagemap is defined with an AREA tag. The HREF attribute specifies the URL for the destination that should be chosen if this area was selected. If you specify NOHREF instead, this area won't do anything.

SHAPE and COORDS define the actual region. SHAPE can be a rectangle, circle, or polygon, and COORDS should contain a set of coordinates describing that shape. This is done with a comma separated list of numbers, enclosed in quotes. If SHAPE is set to DEFAULT, no coordinates need to be specified. The default area is what will be chosen if no others match. The syntax for COORDS depends on what shape you choose.

rect - rectangle

A rectangle has four coordinates. The first specifies the top left corner, and the second the bottom right corner of the rectangle. For example, <AREA SHAPE=rect COORDS="0,0,9,9"> would specify a rectangle of 10x10 pixels, starting in the top left corner of the image.

circle - circle

A circle is defined by its center and radius. The COORDS attribute first specifies the coordinates of the center, and then the radius of the circle, in pixels. For example, <AREA SHAPE=circle COORDS="10,10,5"> would specify a circle with radius 5 at location (10,10) in the image.

poly - polygon

A polygon is built up by a list of coordinates. They are all connected in the order you present, and the last coordinate pair is connected to the first. This way you can build arbitrary figures. For example, <AREA SHAPE=poly COORDS="10,50,15,20,20,50"> would specify a triangle, with edge locations (10,50), (15,20) and (20,50).

default - default

The default location doesn't have any coordinates, and it should be used only once in the map. It is used to indicate what should happen if the user selects one of the coordinates which are not defined in any of the other elements.

The ALT text is used by text browsers to present the URLs in the imagemap in a more readable fashion. If you leave those off, the browser can only display the "bare" URLs. The ALT text is required if you want your document to be valid.

Notes:

• Coordinates are specified in X,Y order: COORDS="1,0,10,19" means from X=1, Y=0 to X=10, Y=19. The top left corner is (0,0).
• If you have an area which should not do anything, not even go to the URL specified in the default area, use NOHREF.
• Just like with IMG, there can be no markup inside the ALT attribute.

A - Hyperlinks

The anchor tag is the "glue" for hypertext documents. The enclosed text and/or image(s) will be selectable by the user, and doing so will take the user to the location specified in the HREF attribute. The TITLE attribute can be used to provide a description of that location, which is displayed by some browsers when the mouse moves over the URL.

The NAME attribute is used to set up "named anchors." The enclosed text will be marked as a "target" to which a browser can jump directly. For example, if you have "<A NAME="toc">Table of Contents</A>" somewhere in the document, and the user selects the URL "#toc" he will be taken to that line.

REL and REV are not widely used, although these attributes were already present in the HTML 2.0 specs. They are used to mark up relationships between the current document and the resource in the link. REL="foo" in document A, in a link pointing to B, indicates that document A has a relationship of "foo" with document B. REV="foo" indicates B has that relationship with A. Since these attributes are not widely used, there is no standard list of values for REL and REV.

Notes:

• Be sure to close the quotes around the value in HREF. Older browsers were often quite forgiving about them, but as this caused other problems, it was fixed in newer releases. This means that a hyperlink with an unclosed quote may not work correctly.
• You may not nest anchors, not even if one uses the HREF and the other uses the NAME attribute.
• The A element used with the NAME attribute requires a closing tag and non-empty content.
• TITLE is most often used for mailto URLs, where it is used to set the subject of the message. Some browsers also use it if you bookmark the link.
• Since you cannot nest anchors, you can't hyperlink a named anchor. You can however combine the NAME and HREF anchors into one anchor; <A NAME="foo" HREF="bar">text</A> works just fine.

BASEFONT - Default font size

The BASEFONT tag is used to suggest a default font size. FONT tags using a relative font size will use the value set here. The SIZE attribute is an integer value ranging from 1 to 7. The base font size applies to the normal and preformatted text but not to headings, except where these are modified using the FONT element with a relative font size.

Notes:

• Do not use FONT or BASEFONT to manipulate the font appearance, use BIG and SMALL for local changes, and style sheets as a more general solution.

BASE - Document location

The BASE tag is used to indicate the correct location of the document. Normally, the browser already knows this location. But in cases such as a mirrored site, the URL used to get the document is not the one that should be used when resolving relative URLs. That's when you use the BASE tag. The required HREF attribute must contain a full URL which lists the real location of the document.

For example, in a document which contains <BASE HREF="http://www.htmlhelp.com/">, the relative URL in <IMG SRC="icon/wdglogo.gif"> corresponds with the full URL http://www.htmlhelp.com/icon/wdglogo.gif.

Notes:

• It is not necessary to include this tag; use it only if relative URLs won't work otherwise.

BIG - Larger text

The BIG tag (as well as SMALL) is new. A browser should draw the enclosed text in a larger font if available, and ignore the tag otherwise. Since this tag is new, support for it is not universal. The FONT tag can do the same, with SIZE="+1".

Notes:

• Nesting BIG tags may produce text in a larger font than with just one BIG tag, but this is not required by the specs (although it is recommended).
• It is legal to nest BIG and SMALL, although the results are undefined.
• BIG is not supported by all browsers, so you might want to use <FONT SIZE="+1"> instead.

BLOCKQUOTE - Large quotations

If you are quoting more than a few lines from a document, use a BLOCKQUOTE to indicate this. Block quotations are often rendered with indented margins, and possibly in italics, although a rendering with the standard quotation symbol for E-mail, "> ", is of course also possible.

Notes:

• If you quote from someone else's work, don't forget to include a credit and/or copyright notice.
• Do not use BLOCKQUOTE simply to create indented text. This is not the required rendering, so you will not achieve the effect you want on all browsers. It will also confuse page indexers and summarizers.

BODY - Document body

The BODY tag contains the actual contents of the document. That contents should consist of block elements only. You may put in plain text in the body, this is then assumed to be inside a P container.

The attributes contain the appearance of the document. The BACKGROUND attribute should point to the location of an image, which is used as the (tiled) background of the document. The other attributes set the colors for the background, text, links, visited links and active (currently being selected) links, using the order above.

The color is composed by specifying the red, green and blue components of the color in hexadecimal notation, with a # in front. For example, to specify white, the red, green and blue components are 255, 255, 255, so you would use "#FFFFFF". You can also use the following color names, although they are not as widely supported as the codes:

Black:	#000000	Green:	#008000	Silver:	#C0C0C0	Lime:	#00FF00
Gray:	#808080	Olive:	#808000	White:	#FFFFFF	Yellow:	#FFFF00
Maroon:	#800000	Navy:	#000080	Red:	#FF0000	Blue:	#0000FF
Purple:	#800080	Teal:	#008080	Fuchsia:	#FF00FF	Aqua:	#00FFFF

The BODY tag is optional; if you put all the HEAD elements first, the browser can immediately see where the actual document body begins.

Notes:

• If the background image cannot be displayed, the color specified in BGCOLOR will be used.
• If you set one of the attributes, set them all. Otherwise your specified color may conflict with a user's default. This could result in unreadable text. For example, imagine that you set your TEXT color to light gray, but forget to set the background. Then someone with a light gray background will not see anything at all.
• Do not set unvisited and visited links to the same color, it will confuse your readers.
• The names that you can use instead of the hexadecimal values are not as widely supported as the color codes.
• Netscape 1.1 produced a "fade" effect when more than one BODY tag was used in a document. It would render the BGCOLOR colors in sequence. This bug has been fixed in later versions. Do not expect that using multiple BODY tags will give the intended results.

BR - Forced line break

The BR tag is used to force line breaks within text. Normally, linebreaks are treated as a space by browsers (except inside the PRE tag). The optional CLEAR attribute is used when you have an IMG image in your text. If that image uses ALIGN=LEFT or ALIGN=RIGHT, the text will flow around it. If you have text you want below the image, you can do this with <BR CLEAR=LEFT> or CLEAR=RIGHT to force scrolling down to a clear left or right margin, respectively. Using CLEAR=ALL will scroll down until both margins are clear. CLEAR=NONE is the default, and does nothing.

Notes:

• Some people use multiple BR tags to force whitespace. This is not required by the specs, so it may not work in all browsers.

B - Bold

B is used to indicate that the enclosed text must be rendered in a bold typeface. It must be rendered distinct from I-italics text.

If you want to indicate strong emphasis, use the STRONG element instead. B should only be used when you want bold typeface for some other reason than to denote strong emphasis. While the two tags usually produce the same output, the B tag does not provide any reasons why the enclosed text is in boldface. This means an indexer or text-only browser cannot pick a good alternative. With STRONG this is possible.

Notes:

• Putting large blocks of text in boldface makes the text hard to read.

CAPTION - Table caption

The CAPTION tag is used to provide a caption for a TABLE. This caption can either appear above or below the table. This can be indicated with the ALIGN attribute. It is usually centered with respect to the table itself, and usually appears in bold or is emphasized in some other way.

The tag should appear directly below the TABLE tag, before the first TR.

Notes:

• Although you can use all text-level markup inside a CAPTION, it should be brief, so don't include images or large amounts of text in it.

CENTER - Centered division

The CENTER tag is one of the first Netscape extensions. It is used to indicate that large blocks of text should appear centered. In the Wilbur standard, it is defined as an alias for <DIV ALIGN=CENTER>.

The tag is more widely supported than the DIV method, as it was the first widely implemented Netscape extension to HTML 2.

Notes:

• The CENTER tag is not text-level markup, so do not use it to center single lines of text inside a paragraph or other block element. It will introduce a new paragraph.
• Older versions of Netscape treated CENTER as if it were text-level markup, so it was rendered without a paragraph break there.
• For better portability with browsers which do not support this tag, use ALIGN=CENTER on headers and paragraphs if possible.

CITE - Short citations

The CITE element marks up the title of a cited work. For example, a text discussing HTML could say <CITE>RFC 1866</CITE> says ....

Text in CITE is typically rendered in italics.

Notes:

• Do not use CITE for anything but titles of cited works. It will confuse indexers and automatic programs to extract information from your documents. Use EM for emphasis or I for text which must appear in italics.
• There is no element in HTML 3.2 to mark up short cited phrases. For longer texts, you can use BLOCKQUOTE.

CODE - Code fragments

CODE is used for snippets of code which appear inside a paragraph of text. It is usually rendered in a monospaced font. You can use this tag to mark up things like <CODE>for ( ; ; ) ;</CODE> is a nice way to make an endless loop in C.

For larger blocks of code, use PRE instead. If what you are marking up is what a user should type in, use KBD.

Notes:

• CODE will usually be rendered in a monospaced font, but multiple spaces are collapsed, unlike in PRE. This can screw up the spacing in your code if you want to provide more than one line.

DD - Definition

The DD tag is used inside a DL definition list to provide the definition of the text in the DT tag. It may contain block elements but also plain text and markup. The end tag is optional, as it's always clear from the context where the tag's contents ends.

A typical rendering is indented, one line below the DT, but this is not guaranteed.

Notes:

• Some people use this tag out of its proper context (DL only) to achieve an "indent" in text. Don't do this, it is not valid and not guaranteed to work.

DFN - Definition of a term

DFN is used to mark up terms which are used for the first time. These are often rendered in italics so the user can see this is where the term is used for the first time.

Notes:

• Not all browsers render this tag in an appropriate way.

DIR - Directory list

The DIR element is similar to the UL element. It represents a list of short items, typically up to 20 characters each. Items in a directory list may be arranged in columns, typically 24 characters wide.

Notes:

• It is not permitted to use a block element, list element or TABLE in a LI inside a MENU or DIR.

DIV - Logical division

The DIV tag is used to mark up divisions in a document. It can enclose paragraphs, headers and other block elements. Currently, you can only use it to set the default alignment for all enclosed block elements. Future standards will most likely include more options for DIV.

Just like with other block elements such as P or H1, you can specify left, right and centered alignment.

Notes:

• The align attribute on a block element inside DIV overrides the align value of the DIV element.
• Instead of <DIV ALIGN=CENTER>, use CENTER. This element is more widely supported at the moment, even though HTML 3.2 defines both as identical.

DL - Definition list

DL is used to provide a list of items with associated definitions. Every item should be put in a DT, and its definition goes in the DD directly following it. This list is typically rendered without bullets of any kind.

While it is legal to have a DL with only DD or DT tags, it doesn't make much sense (what good is a definition without a term?) and you shouldn't expect it to get rendered as a normal list.

Notes:

• DL may not contain plain text or any tag other than DT or DD.
• Do not use DL to create an indented section of text. This is not guaranteed to work and it is syntactically invalid HTML.

DT - Definition term

The DT tag is used inside DL. It marks up a term whose definition is provide by the next DD. The DT tag may only contain text-level markup.

Notes:

• Although it is legal, using a DT without a corresponding DD tag is quite pointless.

EM - Emphasized text

EM is used to indicate emphasized text. While it is often rendered identical to I, italics, using EM rather than I is preferred. It allows the browser to distinguish between emphasized text and other text which can be drawn in italics (for example citations, CITE).

EM text should be rendered distinct from STRONG text.

Notes:

• Use EM only for emphasized text. If you want to use an italic font for some other reason, use a more appropriate element like CITE, DFN or I.

FONT - Font appearance

The FONT tag can be used to change the appearance of the current block, in terms of size and color. The SIZE attribute can either take an absolute value, ranging from 1 (smallest) to 7 (largest), or a relative value. Using the latter will change the font relative to the font size as set in the most recent BASEFONT tag. For example, <FONT SIZE="+1"> will make the font size one step bigger.

The COLOR attribute takes a hex value, which is the RGB-notation of the desired color. You can also use a color name, although the names are less widely supported than the codes. See the section on BODY for a more detailed explanation on how to specify colors.

In both cases, the closing tag </FONT> restores the previous font size and/or color.

Notes:

• Avoid making extreme font changes if possible. They can make a document hard to read. If possible, use BIG instead of <FONT SIZE="+1"> and SMALL instead of <FONT SIZE="-1">.
• The color attribute is not very widely supported, so do not rely on it.
• None of the browsers which support FONT allow their users to disable it. It is possible to disable body colors, so if you use FONT COLOR to change font colors and the visitor has overriden your body colors, the text may wind up invisible.
• Do not use the FONT tag to emulate headers. Indexers rely on the six header elements to generate an overview of a document, and they will not be able to index your document if you use FONT instead. FONT should be used only as an enhancement of the presentation.

FORM - HTML forms

Forms allow a person to send data to the WWW server. You can use the INPUT, TEXTAREA and SELECT tags to add individual elements, such as checkboxes, input fields or "drop down" lists to your form. A form may contain all markup (both text and body level tags), but it may not have a nested form.

FORM has one required attribute, ACTION, specifying the URL of a CGI script which processes the form and sends back feedback. There are two methods to send form data to a server. GET, the default, will send the form input in an URL, whereas POST sends it in the body of the submission. The latter method means you can send larger amounts of data, and that the URL of the form results doesn't show the encoded form.

You can specify an encoding type with ENCTYPE, the default of "application/x-www-form-urlencoded" is most widely supported. An alternative is "text/plain", which is typically used in combination when the ACTION attribute points to a mailto: URL. If a browser supports both, the contents of the form is sent in plain text to the indicated recipient.

Notes:

• A form should always have at least one submit button. This can be done with <INPUT TYPE=submit NAME=submitit> or with an image: <INPUT TYPE=image NAME=submitit>.
• More than one submit button is legal. If each submit button has a unique NAME attribute, the name of the selected submit button is sent along with the rest of the form input. This allows the parsing script to determine which button was pressed.
• The URL specified in the ACTION attribute does not have to be a CGI script, although you can get pretty weird results if you try to feed data to a document which isn't a CGI script. A popular reason to do this is to get a "button" which when pressed takes you to a new page. This can be done with:

  <FORM ACTION="destination_url" METHOD=GET>
  <INPUT TYPE=submit NAME=foo VALUE="Go to destination">
  </FORM>
  

H1 - Level 1 heading

The level 1 heading is the most important header in the document. It should be rendered more prominently than any other header. It is usually used to indicate the title of the document. Often it has the same contents as the TITLE, although this is not required and not always a good idea. The title should be useful out of context (for example, in a bookmarks file) but the level 1 heading is only used inside the document.

The optional ALIGN attribute controls the horizontal alignment of the header. It can be LEFT, CENTER or RIGHT.

Notes:

• Headers should be used in hierarchical order.
• Do not assume that this header means "very large font size and bold." While this is a popular rendering, it can be anything the browser chooses.
• Search engines may give words appearing in headers more importance in their index. The headers are also often used to build an "outline" of the document, which appears in the search results.
• Right alignment is not as widely supported as centered.

H2 - Level 2 heading

The level 2 heading is the second most important header in the document. It should be rendered more prominently than a H3, but less prominently than a H1. It is often used to mark up chapters in a document.

The optional ALIGN attribute controls the horizontal alignment of the header. It can be LEFT, CENTER or RIGHT.

Notes:

• Headers should be used in hierarchical order.
• Do not assume that this header means "very large font size and bold." While this is a popular rendering, it can be anything the browser chooses.
• Search engines may give words appearing in headers more importance in their index. The headers are also often used to build an "outline" of the document, which appears in the search results.
• Right alignment is not as widely supported as centered.

H3 - Level 3 heading

The level 3 heading is the third most important header in the document. It should be rendered more prominently than a H4, but less prominently than a H2. It is often used to mark up sections inside a chapter in a document.

The optional ALIGN attribute controls the horizontal alignment of the header. It can be LEFT, CENTER or RIGHT.

Notes:

• Headers should be used in hierarchical order.
• Do not assume that this header means "very large font size and bold." While this is a popular rendering, it can be anything the browser chooses.
• Search engines may give words appearing in headers more importance in their index. The headers are also often used to build an "outline" of the document, which appears in the search results.
• Right alignment is not as widely supported as centered.

H4 - Level 4 heading

The level 4 heading should be rendered more prominently than a H5, but less prominently than a H3. It is often used to mark up subsections in a document.

The optional ALIGN attribute controls the horizontal alignment of the header. It can be LEFT, CENTER or RIGHT.

Notes:

• Headers should be used in hierarchical order.
• Do not assume that this header means "large font size and bold." While this is a popular rendering, it can be anything the browser chooses.
• Search engines may give words appearing in headers more importance in their index. The headers are also often used to build an "outline" of the document, which appears in the search results.
• Right alignment is not as widely supported as centered.

H5 - Level 5 heading

The level 5 heading is the second least important header in the document. It should be rendered more prominently than a H6, but less prominently than a H4. Because it is often rendered in a small font, it is not used very often. It should be used to divide sections inside a H4.

The optional ALIGN attribute controls the horizontal alignment of the header. It can be LEFT, CENTER or RIGHT.

Notes:

• Headers should be used in hierarchical order.
• Do not assume that this header means "small font size and bold." While this is a popular rendering, it can be anything the browser chooses.
• Search engines may give words appearing in headers more importance in their index. The headers are also often used to build an "outline" of the document, which appears in the search results.
• Right alignment is not as widely supported as centered.

H6 - Level 6 heading

The level 6 heading is the least important header in the document. It should be rendered less prominently than a H5, but more prominently than normal text. Because it is often rendered in a small font, it is not used very often. It should be used to divide sections inside a H5.

The optional ALIGN attribute controls the horizontal alignment of the header. It can be LEFT, CENTER or RIGHT.

Notes:

• Headers should be used in hierarchical order.
• Do not assume that this header means "very small font size and bold." While this is a popular rendering, it can be anything the browser chooses.
• Search engines may give words appearing in headers more importance in their index. The headers are also often used to build an "outline" of the document, which appears in the search results.
• Right alignment is not as widely supported as centered.

HEAD - Document head

The HEAD part of the document provides information about the document. It should not contain text or normal markup. If a browser encounters such markup, it will assume it has arrived in the BODY section of the document already.

Notes:

• You may only omit this tag if you group all the tags which may go in it at the top.

HR - Horizontal rule

HR is used to draw horizontal rules across the browser window. If the margins are currently smaller, for example because of images (IMG) which are placed against the margins, the rule will extend to these margins instead of the whole window. A horizontal rule is typically used to separate sections within a document.

In HTML 3.2, the appearance can be controlled more than in HTML 2. You can specify the thickness of the rule with the SIZE attribute, which takes an integer number of pixels. The width of the rule can be specified in number of pixels or as a percentage of the currently available window width, using the WIDTH attribute. Don't forget that percentage values must be quoted! The NOSHADE attribute is used to indicate that the rule should not get its usual shaded appearance, but instead should be drawn as a thick line.

Notes:

• None of the attributes for HR existed in HTML 2, so they may not be supported by all browsers. This can produce bizarre effects if you are using multiple HRs in a row to produce growing or shrinking "stripes".
• If you use too many rules on a document, the end result can be that the document looks like a "sandwich" because there is little text between each rule.
• Setting an absolute width is not recommended, since you have no way to know how wide the currently available window is. Use a percentage if you have to change the width.

HTML - HTML Document

The HTML tag is the outermost tag. It is not required and may safely be omitted. It indicates that the text is HTML (the version can be indicated with the optional VERSION attribute), but this information is almost never used by servers or browsers.

Notes:

• If used, the HTML tags should go around the entire document, but directly after the DOCTYPE declaration.

IMG - Inline images

The IMG tag is used to insert images within text. These are often called "inline" images. Note that the IMG tag is not a block tag by itself, so it must be used only within a block element. The location of the image file should be specified in the SRC attribute. It can be a relative or an absolute URL. When the image cannot be displayed for whatever reason, the browser should display the ALT text instead. The WIDTH and HEIGHT attributes should contain the image's dimensions. This allows a browser to lay out the page in advance, as it now knows where the text below the image should be drawn.

ALIGN controls the alignment of the image with respect to the text. Using a value of LEFT or RIGHT will make the image line up against the left or right margin, and text will flow around it. To force text below such an aligned image, use BR with the CLEAR attribute. The values TOP, MIDDLE and BOTTOM specify where any text following the image should be put. If more than one line follows after the image, it will be put below the image.

VSPACE and HSPACE get a numeric value, indicating the number of pixels that should be left free around the image. VSPACE covers vertical spacing and HSPACE the horizontal spacing.

The BORDER attribute is used when the image is a link. It indicates that the browser should draw a border of the indicated size around the image to show that it is a link. It's most often used as BORDER=0 to turn it off. This has the disadvantage that the image must make it very clear that it's a hyperlink.

ISMAP and USEMAP are used for imagemaps. The ISMAP attribute specifies that the link that this image is in goes to an imagemap program on the server, so the browser can send the coordinates of the selected location to the server. USEMAP is used for a so-called client-side imagemap. It specifies the URL of the imagemap information. Support for this is limited, especially if the URL points to a different document rather than an inline anchor. See the section on the MAP tag for more information about client-side imagemaps.

Notes:

• Most browsers only support GIF and JPEG file types for inline images.
• The ALT text may not contain markup, other than entities. Not all browsers support entities in ALT text, so be careful.
• Although the ALT attribute is not required, it is good practice to add it. It should replace the image's meaning, and not just provide a description of the image! If the image is purely decorational, use ALT="" or a decorative ALT text like "* " if possible.
• Using WIDTH and HEIGHT with incorrect values, or with percentage values is not valid. Some browsers (most notably Netscape) resize the image to the indicated size, but this often gives very poor results. Other browsers simply ignore the WIDTH and HEIGHT attributes in such a case.
• Browsers which do not draw boxes around hyperlinked images will not honor the BORDER attribute.

INPUT - Input field

The INPUT tag is probably the most useful tag inside forms. It can generate buttons, input fields and checkboxes. In all cases, the NAME attribute must be set.

TYPE=text

This generates a input field, where the user can enter up to MAXLENGTH characters. The SIZE attribute lists the length of the input field (if the user enters more characters, the text will scroll). The VALUE attribute specifies the initial value for the input field.

TYPE=password

Same as TYPE=text, but the text will be hidden by "*" or similar characters. It is still sent in the clear to the server, though.

TYPE=checkbox

Produces a checkbox. It has two states, on and off. When it is on when the form is submitted, it will be sent as "name=on", otherwise it is ignored altogether. If you use CHECKED, it will come up checked (selected) initially.

TYPE=radio

Produces a radio button. A radio button always exists in a group. All members of this group should have the same NAME attribute, and different VALUEs. The VALUE of the selected radio button will be sent to the server. You must specify CHECKED on exactly one radio button, which then will come up selected initially.

TYPE=submit

Produces a button, which when pressed sends the contents of the form to the server. You can have more than one submit button in the form. Each should have a different NAME. The name and value of the pressed button will be sent to the server as well. The value of the VALUE attribute is typically used as text on the submit button.

TYPE=reset

Also produces a button, which will restore the form to its original state if pressed. The value of the VALUE attribute is typically used as text on the reset button.

TYPE=file

Allows the user to upload a file. It is still very new, so it is not very widely supported. It is typically presented as an input box with a button to start browsing the local hard disk. This way, a user can specify one or more filename(s) to upload.

TYPE=hidden

Allows you to embed information in the form which you do not want changed. This can be useful if the document is generated by a script and you need to store state information. NAME and VALUE of this input field will be sent to the server without modifications.

TYPE=image

Functions similar to a submit button, but uses an image instead. The ALIGN attribute controls the alignment of the image. The coordinates of the selected region will also be sent to the server, in the form of "NAME.x=n&NAME.y=n". A text browser will treat it as identical to a normal submit button.

Notes:

• Do not expect that if you set MAXLENGTH, you will get no more than so many characters. Anyone can modify his local copy of your form to enter as many characters as he wants.

ISINDEX - Primitive search

The ISINDEX tag was used before FORMs became more popular. When inserted in a document, it will allow the user to enter keywords which are then sent to the server. The server then executes a search and returns the results. The PROMPT attribute can be used to override the default text in the dialog box ("Enter search keywords: ").

Notes:

• This tag should be inserted by the server if the document can be searched. Merely inserting this tag will not make the document searchable!
• For more flexibility, use a FORM instead.

I - Italics

I is used to indicate that the enclosed text must be rendered in a italic (slanted) typeface. It must be rendered distinct from B-bold text.

You should use EM or CITE instead of I if you can. While they usually produce the same output, the I tag does not provide any reasons why the enclosed text is in italics. This means an indexer or text-only browser cannot pick a good alternative. With EM and CITE this is possible. The browser can now distinguish between emphasized text and citations and choose different methods to display them.

Notes:

• The I tag should only be used if text is in italics by convention.

KBD - Keyboard input

KBD is used to indicate text which should be entered by the user. It is often drawn in a monospaced font, although this is not required. It differs from CODE in that CODE indicates code fragments and KBD indicates input.

Notes:

• Do not use KBD if your text requires a monospaced font; use TT instead.

LINK - Site structure

LINK is used to indicate relationships between documents. There are two possible relationships: REL indicates a normal relationship to the document specified in the URL. REV indicates a reverse relationship. In other words, the other document has the indicated relationship with this one. The TITLE attribute can be used to suggest a title for the referenced URL or relation.

Some possible values for REL and REV:

REV="made"

Indicates the creator of the document. Usually the URL is a mailto: URL with the creator's e-mail address. Advanced browsers will now let the reader comment on the page with just one button or keystroke.

REL="stylesheet"

This indicates the location of the appropriate style sheet for the current document.

REL="home"

Indicates the location of the homepage, or starting page in this site.

REL="toc"

Indicates the location of the table of contents, or overview of this site.

REL="index"

Indicates the location of the index for this site. This doesn't have to be the same as the table of contents. The index could be alphabetical, for example.

REL="glossary"

Indicates the location of a glossary of terms for this site.

REL="copyright"

Indicates the location of a page with copyright information for information and such on this site.

REL="up"

Indicates the location of the document which is logically directly above the current document.

REL="next"

Indicates the location of the next document in a series, relative to the current document.

REL="previous"

Indicates the location of the previous document in a series, relative to the current document.

REL="help"

Indicates the location of a help file for this site. This can be useful if the site is complex, or if the current document may require eplanations to be used correctly (for example, a large fill-in form).

Notes:

• Support for the LINK tag is currently very limited, although it may be useful for site maintenance.

LI - List item

The LI element is used to mark list items within a list. When the list used is OL, ordered list, the LI element will be rendered with a number. The appearance of that number can be controlled with the TYPE attribute. Similarly, inside an unordered list UL the type of bullet that is displayed can be controlled with TYPE. DIR and MENU can't be controlled this way, as they are not required to be bulleted or numbered. For ordered lists, you can also reset the sequence with the VALUE attribute.

The TYPEs for ordered lists should give the following appearance:

• 1 - Arabic numbers (default) (1, 2, 3, 4, ...)
• a - Alphanumeric, lowercase (a, b, c, d, ...)
• A - Alphanumeric, uppercase (A, B, C, D, ...)
• i - Roman numbers, lowercase (i, ii, iii, iv, ...)
• I - Roman numbers, uppercase (I, II, III, IV, ...)

Notes:

• The attributes on LI are new with HTML 3.2, and so not supported by all browsers yet.
• Using LI outside its proper context to get a bullet in the text is not guaranteed to work. A browser is free to ignore such an out-of-context item.
• When the LI element is used inside MENU or DIR, it is not permitted to include block elements, l