Errata for Dynamic HTML: The Definitive Reference

well as the descriptions. This would be much faster to scan for what the reader wants. AUTHOR'S REPLY: I wholeheartedly agree and will try to get production to accede to my continued request in the next edition.

p:first-line {font-style:small-caps} NOW READS: p:first-line {font-variant:small-caps}

This is stated in the reference section, but it should be mentioned here as well.

If this function finds a match in the recursion, it doesn't stop. This is extra work. To fix, add "if (theObj) break" right after the recursion call. Also, I would think that the depth first search it does is also extra work. It would be better to look at all the objects at the current level before trying to recurse. AUTHOR'S REPLY: I agree that adding the statement "if (theObj) break" below the first code line on p.90 would avoid extraneous processing.

body {font-family:Ariel, NOW READS: body {font-family:Arial,

remove "and Internet Explorer 3", change "support" to "supports", remove "." at the end and replace with "; Internet Explorer 4 supports both values."

last para. of section 9.3.3.4 should read: "This property is not supported by Netscape; Internet Explorer 4 incorrectly implements *small-caps* as all upper case."

support..." with "Netscape does not support this style property."

effects, Netscape".

"only supported by Internet Explorer 4".

*underline* and *line-through* values" with "Internet Explorer 4 does not support the *blink* value".

Explorer 3." Replace "While supported by" with "In".

The default value for autocomplete should be "none (set by user)". The autocomplete attribute is somehow missing from the <input> tag discussion starting on p.327. I see that the autocomplete property is also missing from the 'form' and 'input type="password"' and 'input type="text"' discussions starting on pp. 683, 744, and 759, respectively.

with "Netscape 4 supports all but the *sub*, *super*, and *baseline* values only when this property is applied to the *<img>* tag. Internet Explorer 4 supports only *sub* and *super* when applied to text elements."

Explorer or Netscape support the *dashed* or *dotted* values."

Internet Explorer support the *border* property, but only Internet Explorer supports the individual side properties."

with "This property is only supported by Internet Explorer for images. Netscape honors it for textual elements as well."

supported by Internet Explorer 4."

supported by Netscape 4; Internet Explorer 4 supports only the *none* value."

property is not supported by Internet Explorer; Netscape only supports the *pre* and *normal* values."

1997". In last sentence, remove everything, including the parenthetic "see 13.4", after "in their most recent versions."

"Currently." At end of para, remove sentence "Netscape promises complete support...to keep pace with Netscape." In last sentence, replace "first half of 1997" with "last half of 1997" and replace "end of the year" with "end of 1998".

"makeword(query_string, '&'.;" should be "makeword(query_string, '&');" Similarly, four lines later, "makeword(entries[num_entries].val, '='.;" should be "makeword(entries[num_entries].val, '=');"

"fmakeword(stdin, '&'. &cl);" should be "fmakeword(stdin, '&', &cl);" Four lines later, "makeword(entries[num_entries].val, '='.;" should be "makeword(entries[num_entries].val, '=');"

behavior..." with "The default behavior, represented by the value *all*, is to draw borders around all cells." Replace last sentence beginning with "Using *rows* or *cols*..." with "Using *rows* or *cols* places borders only between every row or column, respectively, while using *none* removes borders from every cell in the table."

the multiple attribute of the <select> tag is described as having no effect if size is set to 1, and yielding a pop-up menu. In IE 5 and 6, a one-line scrollable list is displayed if both multiple and size=1 are specified. This is useful, e.g. as a way to implement a spinbox control. The error is repeated on page 411 in the first paragraph, describing the size attribute. AUTHOR: The reader is correct for IE/Windows. IE for the Mac (and Safari betas) automatically expand the dimensions of the select element to a default height of four rows when you turn on the multiple attribute.

They both refer to <quote> when instead you meant <q>. And, contrary to your text, in Chapter 8 neither <blockquote> nor <q> have these properties listed. AUTHOR'S REPLY: You're right that <quote> element should be <q>, but only on p.486. On page 488, the dateTime property should apply only to the del and ins elements. In line with the text, the <blockquote> and <q> tag references _do_ show the cite attribute.

AUTHOR'S REPLY: Yes, the Value should read: A string consisting of one or more (in a comma-delimited list) of the following values: all | print | screen.

In Mozilla 1.7, it is an 'Object MediaList', and useless as far as I can tell. AUTHOR'S REPLY: I have this discussion on pp.925-6. The Mozilla version available at press time had not yet implemented the property for me to see whether it was done the way the W3C DOM says it should be (although it had included the MediaList object). Mozilla has since implemented it and correctly at that. IE6, however, still reports the value as a string (as does IE7 through B2).

Text says "Once the range has been narrowed to the target text, assign values to its htmlText and text properties...." You can only assign values to "text," not to "htmlText" which is only read-only. Use pasteHTML() to change HTML.

This paragraph hints that a URL in a CSS is relative to the URL of the current document. This is not what the CSS specs say. CSS 1: Partial URLs are interpreted relative to the source of the style sheet, not relative to the document: CSS 2.1: In order to create modular style sheets that are not dependent on the absolute location of a resource, authors may use relative URIs. Relative URIs (as defined in [RFC1808]) are resolved to full URIs using a base URI. RFC 1808, section 3, defines the normative algorithm for this process. For CSS style sheets, the base URI is that of the style sheet, not that of the source document. AUTHOR: I think it's worthy of a clarification. Of course when the CSS rules are in the source document (e.g., in a <!-- <DEFANGED_STYLE> tag set or assigned to the style attribute of an element), relative URLs will be relative to the current document; but for imported .css files, the relativity is to the location of the .css file, which may or may not be in the same directory as the HTML document.

you don't say what elements it applies to. AUTHOR'S REPLY: The Applies To section should reference Netscape 4 where it currently references Netscape 6.

An optional zero-based integer of the last item of the subset from the current array. should be: An optional zero-based integer of the item after the end of the subset from the current array.

You say the string must begin with numerals, but negative numbers begin with a minus sign. I don't know about beginning with a plus sign or a decimal point. AUTHOR'S REPLY: Decimals aren't allowed, but + and - are. The string parameter descriptions for both functions should read: Any string that begins with one or more numerals (optionally, preceded by + or -).

Math.floor(Math.random() * n) NOW READS: Math.floor(Math.random() * (n+1))

Without the /g option, element 0 contains the matched text. This is described in the Definitive Guide to JavaScript. AUTHOR'S REPLY: Yes, the discussion section and Returned Value needs to be fleshed out more. Without the /g option, the method returns not only a one-element array, but an object with index and input properties. I'll fix this in the next edition.