Name attribute not XHTML valid

On Fri, 08 Apr 2005 15:17:54 -0700, Del Ferguson
<delferguson (AT) charter (DOT) net> wrote:
[...]

Why is XHTML Strict the wrong way to go?

Because all variants of XHTML 1.0 are designed to ride on lots of old
browser bugs to get rendered, bugs that was already there long before
XHTML came along.

--
Rex

Andy Dingley wrote:

A site coded according to Strict is in no possible way
"better" than one coded according to Transitional. In fact, it may
actually be better to use the Transitional DTD -- there are many
possible sites that are valid Trasitional (validity is good) whilst
not being valid Strict.

The benefit is in being _valid_, not in the choice of aspirational
target.

I agree that validity is as you stated, but my concern is that some
coders use Transitional as a means to "bang" out a page and get it over
with.

There are also arguments in favour of writing semantically-relevant
well-structured code. But Strict adds little to aid this - you can
still nest <blockquote> to create an indent, or to chain <a> links
rather than coding an obvious list or menu structure.

I guess I'm not the "purist" that I should be. I'm pragmatic about what
works in a page. If my pages (all Strict XHTML) pass the several
validity checks I do and look OK with FF and IE, that should be enough.

"Clearing up errors from Tidy" is not in itself sufficient to remove
all the inconsistencies from a switch to Strict.

I didn't mean to give the idea that a coder would only use Tidy;
although, using the "correct" DOC TYPE goes a long way towards that end.

Jan Roland Eriksson wrote:

On Fri, 08 Apr 2005 15:17:54 -0700, Del Ferguson
delferguson (AT) charter (DOT) net> wrote:

Why is XHTML Strict the wrong way to go?

Because all variants of XHTML 1.0 are designed to ride on lots of old
browser bugs to get rendered, bugs that was already there long before
XHTML came along.

I agree and HTML might be an easier way to go; however, that is why I
use a VisiBone chart to find code most compatible with the greater
percentage of browers out there.

On Fri, 08 Apr 2005 19:44:45 -0700, Del Ferguson
<delferguson (AT) charter (DOT) net> wrote:

some coders use Transitional as a means to "bang" out a page
and get it over with.

What's wrong with that?

If you "bang it out", and as a result it validates, then you've hit
the target

A valid Transitional page is still valid. It's no less valid, and no
less "right" than a valid Strict page..

I'm pragmatic about what
works in a page. If my pages (all Strict XHTML) pass the several
validity checks I do and look OK with FF and IE, that should be enough.

These tests meet certain goals, but they fail others.

Take a look at csszengarden - you can only do that sort of
dynamically-switched CSS if the page structure beneath is sound. The
structure must not only be amenable to having one specific CSS applied
to it to give one specific look, but it should permit generalised CSS
to be applied, hooking in through a sensible class structure.

There are as yet few benefits to doing this. But there are identified
reasons, mainly in machine-processing of the page, where sensible
semantic structure of the HTML becomes important.

As yet the minimal benefits are in considering a linear CSS-stripped
version of the site, such as encountered by screen readers. 2d tables
etc. should linearise in the sensible order, not the "cross reading"
order. We recently had a page posted with a directory of shops at
Pigeon Forge. The HTML for that linearised as {name, name, name} {
phone, phone, phone} when clearly it would have been improved as
{name, phone}{name,phone}{name,phone}

Andy Dingley wrote:

On Fri, 08 Apr 2005 19:44:45 -0700, Del Ferguson
delferguson (AT) charter (DOT) net> wrote:

some coders use Transitional as a means to "bang" out a page
and get it over with.

What's wrong with that?

If you "bang it out", and as a result it validates, then you've hit
the target

A valid Transitional page is still valid. It's no less valid, and no
less "right" than a valid Strict page..

Other than the fact that Transitional doesn't completely meet the
Standards Mode, you are right.

I'm pragmatic about what
works in a page. If my pages (all Strict XHTML) pass the several
validity checks I do and look OK with FF and IE, that should be enough.

These tests meet certain goals, but they fail others.

That is the reason I do multiple tests such as those listed in Firefox's
Checky. I rely on multiple tests to keep me within the Standards Mode.

Take a look at csszengarden - you can only do that sort of
dynamically-switched CSS if the page structure beneath is sound.

If I could even come close to being on the level of a css Zen Garden
coder, I'd consider myself a real coder, not a hobbyist.

As yet the minimal benefits are in considering a linear CSS-stripped
version of the site, such as encountered by screen readers. 2d tables
etc. should linearise in the sensible order, not the "cross reading"
order.

I agree. We should all look at our pages stripped of style to see if
they make sense.

piraticman (AT) gmail (DOT) com wrote:

But I still haven't fixed the 'name' problem.

I know it has to be changed to ID and then adding the
document.forms[0], but I don't know exactly how to implement this.

<head>
<script type="text/javascript">
<!--
window.onload = function () {
// The form has an id but no name attribute
// under XHTML 1.1
// So we provide a handy global reference.
// Use Pascal case to:
// 1. Designate it is a Global Variable; and
// 2. So it doesn't conflict with the IE 6 global form variable eg
"frmMain".
FrmMain = document.forms["frmMain"];
//-->
</script>
</head>
<body>
<form id="frmMain" onsubmit="return false;" action="" method="post">
....

Mellow Crow wrote:

head
script type="text/javascript"
!--
window.onload = function () {
// The form has an id but no name attribute
// under XHTML 1.1

Is this supposed to be script in XHTML? Then get rid of the <!-- //-->
wrapping the script as that way the script is commented out
(<http://www.w3.org/TR/xhtml1/#C_4>).
You might need other wrappers like
//<![CDATA[
script code
//]]>
then.


--

Martin Honnen
http://JavaScript.FAQTs.com/

Mellow Crow wrote:
head
script type="text/javascript"
!--
window.onload = function () {
// The form has an id but no name attribute
// under XHTML 1.1

Martin Honnen wrote:

Is this supposed to be script in XHTML? Then get rid of the <!-- //--
wrapping the script as that way the script is commented out
(<http://www.w3.org/TR/xhtml1/#C_4>).
You might need other wrappers like
//<![CDATA[
script code
//]]
then.

Thanks Martin. An issue worthy of it's own thread. Researching this now. I
might post this as
"Which comment wrappers for JavaScript in XHTML?" in comp.lang.javascript
when I get the time (in 5 days??).