Re: IMG ALT attribute in HTML 4.0

Jukka Korpela (jkorpela@cc.hut.fi)
Wed, 10 Sep 1997 09:02:15 +0300 (EET DST)


Date: Wed, 10 Sep 1997 09:02:15 +0300 (EET DST)
From: Jukka Korpela <jkorpela@cc.hut.fi>
To: www-html@w3.org
In-Reply-To: <3.0.3.32.19970909160059.008d4c90@undergrad.math.uwaterloo.ca>
Message-ID: <Pine.OSF.3.96.970910084633.12107B-100000@tuuba.hut.fi>
Subject: Re: IMG ALT attribute in HTML 4.0

On Tue, 9 Sep 1997, Liam Quinn wrote:

> I was glancing over the recently updated Cougar DTD at 
> <http://www.w3.org/MarkUp/Cougar/Cougar.dtd> where I noticed that the ALT 
> attribute for IMG is now a required attribute (a good thing, IMO).

A good thing indeed, although it is somewhat odd to improve elements
which should actually be deprecated (as IMG should be, in favor of
OBJECT).

> Both of these comments imply that 
> the ALT attribute is used to give a *description* of the image, but 
> commonly suggested practice (as well as the HTML 4.0 draft [1]) says that 
> the ALT attribute should be used to give *alternate text*

I remember reading somewhere in the draft an explicit statement which
says that comments in the DTD have no normative value. I think they
should be compatible with the normative part, however.

> With this in mind, I would suggest changing the DTD's comment to 
> "alternate text when image is not rendered".

The essential thing is the wording of the specification itself. If it is
difficult to make the comments in the DTD correspond to that, such
comments should be simply removed.

However, the ALT controversy is not just a matter of wording. There is
a fundamental problem. There are two masters to serve, two needs: the
need for replacement (when image is not shown at all, for some reason)
and the need for description. The latter is nowadays more common, as
many people use graphic browsers with image autoloading off. But both
needs must be taken into account. I'm not sure exactly how this should
be done for the IMG element, but I think the essential point is to
prevent the controversy from ever arising for the OBJECT element.
For my preliminary thoughts on this, see
http://www.hut.fi/home/jkorpela/HTML4.0/comments.html#alt

Yucca, http://www.hut.fi/home/jkorpela/