W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > October to December 2001

Re: title inclusion

From: Al Gilman <asgilman@iamdigex.net>
Date: Tue, 04 Dec 2001 15:28:12 -0500
Message-Id: <Version.32.20011204135022.040f7970@pop.iamdigex.net>
To: Joe Clark <joeclark@joeclark.org>, WAI-GL <w3c-wai-gl@w3.org>
What he said.  TITLE is a resource; we need to use it well.

And furthermore, we need the behavior defined in XHTML 2.0.

In the past there have been occasional disasters such as a screen reader
passing over link text and only using the TITLE on hyperlinks when the

It's going to sound complicated until we get it all spelled out, but all the
conditions of UAAG Guideline 2 apply to how to profile what of the various
information assets show when.  The language should provide the author with a
default profile, the author may revise this but what the author provides is a
default profile with a basis in declarative properties.  User policies rewrite
the profile based on the declarative properties and the user can, if they find
they need to, reprofile what shows and what remains hidden by drilling down
locally and peeking at what remains hidden.

Quote "the ALT vs. TITLE mess" unquote is on the list of ugly little nitty
gritty areas that we need to be cleaning up.  Part of the answer belongs in
XHTML 2.0 specification.  How it works is UAAG 1.0 plus WCAG 2.0.


References:  the thread on "the problem with OBJECT" that Sean started.  The
XHTML 2.0 issues heap in the wai-tech-comments archives.  The current
thread on
wai-ig on how to sneak extra orientation text into a page so it is there in
aural rendering and not there in visual rendering (give or take).  All these
content control scenarios need to flow from one master policy framework.  See
also WAI comments on the Device Independence scenarios at

At 01:13 PM 2001-12-04 , Joe Clark wrote:
>Here is a recurring bugbear I mentioned the other day.
>When it comes to providing alternate text for images, current WCAG 
>only ever bothers to talk about alt (universally supported, but short 
>and limited) and longdesc (almost completely unsupported, and way too 
>big most of the time). Meanwhile, the tremendously useful title 
>attribute, supported by virtually everything save for Netscape 4 and 
>OutSpoken for Macintosh, is ignored completely.
>title sits halfway between alt and longdesc. You can load it up with 
>several sentences of text and it will still display even in monopoly 
>browsers like Windows IE. Very often, four or five words in an alt 
>and a couple of sentences in a title fully and completely sum up an 
>As one of the "common attributes," title is looked down upon as Not 
>Really Being Our Kind of Thing Here. Since you can slap a title on 
>essentially everything, it is seen as not access-specific enough. But 
>it's perfectly useful for accessibility. In fact, it is already used 
>for accessibility. I know lots of page authors who never write a 
>hyperlink or include an image without a title on each. And if the big 
>screen readers cannot gain access to title text, they need to be 
>upgraded; it's basic HTML.
>So what I am suggesting is that title be explicitly included in WCAG 
>2.0 as a form of text equivalent for a wide range of objects, 
>including images and multimedia. I assume this would not be a 
>difficult thing to include and would not require huge rewrites.
>   Joe Clark | joeclark@joeclark.org |
>   Accessibility articles, resources, and critiques ||
>       "I can't pretend to understand the mind of Joe Clark"
>       -- Larry Goldberg
Received on Tuesday, 4 December 2001 15:19:09 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:39 UTC