W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > April to June 1998

Re: URL for new Draft to Review

From: Liam Quinn <liam@htmlhelp.com>
Date: Sat, 11 Apr 1998 13:14:17 -0400
Message-Id: <>
To: w3c-wai-gl@w3.org

Section 1.1 requires that authors "Use elements that comply with the strict
HTML 4.0 Definition".  The wording is a bit odd here in that one could
interpret it to mean that HTML 4.0 Strict elements have to be used but
attributes have no such restriction.  I assume the intention is to require
valid, HTML 4.0 Strict documents, which I think is unrealistic.  Valid HTML
(as defined by <http://validator.w3.org/>) should be required, but most
authors would prefer HTML 4.0 Transitional or Frameset over HTML 4.0 Strict
since that allows them to use APPLET and various harmless attributes such
as ALIGN.  As a side note, the draft declares itself to be HTML 4.0 Strict
but uses deprecated ALIGN attributes that are only in HTML 4.0
Transitional.  (There are also some &nbsp;'s that look like leftovers from
a WYSIWYG editor or a conversion program, e.g., "the WAI author guidelines
can&nbsp; also".)

Section 1.3:  We should probably explain what is meant by "proper" and
"improper" nesting of headings.

Section 1.8:  Confusing terminology is used as ALIGN, BORDER, and
BACKGROUND are lumped in as "presentational elements" when they are
presentational attributes.  As well, BORDER is not deprecated and it does
no harm (that I can see) from an accessibility point of view.

Section 1.10 mentions the CITE attribute as being used to indent a
paragraph.  The CITE attribute has no such use; it's not even supported by
any browsers that I know of.

Section 2.1.1 is a nice improvement.  Section 2.1.2 still uses "XYZ Logo"
as alternate content.

The example in section 2.5 is good, but I don't see the point in using the
TITLE attribute specifically for images used as links.  If the content of
the link is abbreviated or not context-free, then the TITLE attribute
should be used on the A element.  But if the content of the link (including
the ALT text) is a sufficient description of the link's contents, then
using the TITLE attribute is unnecessary and redundant.

Section 3.1 requires the use of alternative text for each script.  The
dynamic interactivity provided by client-side scripts generally cannot be
replaced by static content.  The user is often better off not knowing about
the script.  A good example is client-side form validation, for which no
alternate content would be necessary or appropriate since a good author
would duplicate the validation with the server-side form handler.

Section 3.2 requires alternative text for each applet and suggests using
the ALT attribute.  There is no point to the ALT attribute on APPLET; it
should not be used since markup-rich alternate content can be provided as
the content of the APPLET element.  The example's ALT text would be better
suited as a TITLE attribute value.  As well, decorative applets and objects
may not require alternative content, just as some IMG elements should use

Minor corrections:  Java applets as OBJECT elements should use
CLASSID="java:Press.class" instead of CLASSID="Press.class" according to
the examples in the HTML 4.0 Recommendation [2].  And section 3.2.2 uses
"application/mpeg" in an example when the registered media type for MPEG
videos is "video/mpeg" [3].

Sections 3.2 and 3.3 seem to be redundant.

Section 3.6 mentions "the following example", but there is no example.

Section 6.1 uses "for the text version click here" as an example of good
link text.  I don't think we should advocate any example that uses "click
here" due to its device dependence.

Section 6.3 uses "Press C to go to XYZ page" for a link with "C" as an
access key.  This is worse than "Click here for XYZ page" since almost no
browsers support the ACCESSKEY attribute.  We have to depend on the user
agent to indicate the access key; to facilitate this the access key should
be present in the link text, e.g., <A accesskey="X" href="doc.html">XYZ

Section 10.14 includes a link to A Kinder, Gentler Validator, which is
out-of-date.  The W3C Validator should be sufficient.

[1] http://lists.w3.org/Archives/Public/w3c-wai-gl/1998AprJun/0020.html
[2] http://www.w3.org/TR/REC-html40/struct/objects.html#edef-OBJECT
[3] ftp://ftp.isi.edu/in-notes/iana/assignments/media-types/media-types

Liam Quinn
Web Design Group            Enhanced Designs, Web Site Development
http://www.htmlhelp.com/    http://enhanced-designs.com/
Received on Saturday, 11 April 1998 13:14:32 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 16 February 2017 16:48:18 UTC