W3C home > Mailing lists > Public > w3c-wai-ua@w3.org > July to September 2001

Re: alt tag repair:

From: Al Gilman <asgilman@iamdigex.net>
Date: Wed, 22 Aug 2001 21:33:42 -0400
Message-Id: <Version.32.20010822142734.0400ff00@pop.iamdigex.net>
To: "David Poehlman" <poehlman1@home.com>, "User Agent Working group list" <w3c-wai-ua@w3.org>
It is almost as you describe.  What the guidelines say about browser behavior
could be developed in terms of cases graded into good, better, best.

Good:  When there is a non-null ALT and when images are off, the browser
displays the ALT. When there is no ALT or the ALT is the null string, the
browser shows you nothing.

Better:  The above, except that when there is no ALT or the ALT is the null
string, the browser synthesizes a repair text string which is presented
much as
the ALT text would have been.

Best:  The above, except that when the ALT text is the null string it ignores
the object or generates a synthetic placeholder depending on the setting of a
mode you can pick.

The third case, which is just about what you describe, is the best behavior. 
But there are allowed to be double-A compliant browsers that don't quite get
there.  They don't distinguish the null ALT case and they generate repair
placeholders anyway.  This was based on the feeling that it is more urgent to
have access to some information for sure, even though it is more better to
have
the option of either mode.  If you only support one of the modes, it should be
the "show repair text anyway" flavor.  Not the "hide objects with properly
indicated null ALT" flavor.

Best authoring practice would make the 'hide' mode the appropriate behavior. 
But we can't count on a perfect world where everyone is using the absolutely
best authoring practice.

Hope this hasn't confused the situation too badly.

Al

At 12:29 PM 2001-08-22 , David Poehlman wrote:
>I've noticed recently that it is taken that an alt tag such as:
>alt=""
>is no alt tag.  It was my understanding that this was to represent a
>null hence supressed conditional view.  If however this is so than user
>agents complying with this should when asked for the information not
>repair it to a condition of no alt tag.
>
>If I quirie the element, I should be provided with the information that
>is available for that element such as in the case of an image, the name,
>type and what ever other information has been provided for that element
>and not a notice that it has a lack of conditional content?
>
>
>In summary, Is not an null alt tag still an alt tag?
>Should it not in all instances of conditional content view be given that
>status?  if I ask for a place holder, I should be presented with <image>
>or the above value representation.  If I ask for conditional view, I
>should be given none.  correct?  I have reviewed the checkpoints on this
>and it seems that this is so.
>
>I'd like to know or perhaps if not, we should be certain that we have
>stated our intent clearly.
>
>Hands-on Technolog(eye)s
>Touching The Internet
><http://members.home.com/poehlman1/>http://members.home.com/poehlman1/
><mailto:poehlman1@home.com>mailto:poehlman1@home.com
>voice: 301.949.7599
>  
Received on Wednesday, 22 August 2001 21:14:12 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 October 2009 06:50:58 GMT