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

RE: D-link and LONGDESC (GL type stuff)

From: Gregg Vanderheiden <po@Trace.wisc.edu>
Date: Thu, 23 Apr 1998 09:17:14 -0500
To: "GL - WAI Guidelines WG" <w3c-wai-gl@w3.org>
Message-ID: <000701bd6ec2$83917260$7b865c90@vanderheiden>
Jon wrote:
I am not sure adding the D-Link to an image with a LONGDESC will be
accepted by designers.  When this was brought up before it was shot
down as
being unacceptable to WWW authors who would not like to see D on or
their images.  I think that LONGDESCs should be more like ALT text.  A
setable option would turn long descriptions on and it would appear as
inserted HTML in the document.
Using the DOM the assistive technology could always automatically
the LONGDESC and do what ever the assistive technology vendor wants to
with it.

Gv:  Couple of thoughts on this

GV:  Remember that LONGDESC does not contain a long description - or
any text.  It contains a URL.  Thus it is a link to  information...
not information itself

GV:  The suggestion I made below was that the user have the option of
deciding whether a browser would place a small icon next to picture
that a person could "click on" to send for a description - or - not
have any icon present.  (which is what could be the default behavior
of the browser).  Thus no one would see the little "D's" unless they
set their browser to show them.

GV:  Having LONGDESC without a D-LINK (old kind) is not be of any
benefit until browsers support LONGDESC.  That may be awhile - so I'm
not sure if we want to put that forward as helping today.     If both
are put in then the D-LINK would work for old browsers and it would
disappear when the new browsers came on line (if the browsers were
done as suggested).  It would only be recommended though, so that
no-one who doesn't want to use D-LINKS would have to.  It would just
be a recommendation for those who wanted to take the extra effort to
do so.

GV:  One disadvantage of the approach I described is that the page
will lay out slightly differently if the graphic D's are show or not.
This may not be a big problem since people look at the pages with so
many different sized screens that it lays out differently for them
anyway...    But it is something to think about.

GV:  I don't think very many people would ever want LONGDESCs inserted
into the text of a page as a regular event.   LONGDESCs are usually
detailed information that a person occasionally wants to look at.. but
usually doesn't  want to read again.   For example, you might be
interested in the description of the logo for a company but would not
want the description to be inserted into the running text wherever it

GV: And all of these are just thoughts to keep us thinking an moving

Your thoughts?


At 07:12 PM 4/22/98 -0500, Gregg Vanderheiden wrote:
>1) I would posit that an ALT text that is 65536 characters in length
>should probably have been done as a  LONGDESC.   (grin)
>2) I concur with the train of though that ALT and LONGDESC will see
>little use until they are easily and VISIBLY supported in the tools
>used to create web pages.      Jutta and crew are busy on this one.
>We ALL need to encourage tool makers to do this.
>3) I think the best way to support LONGDESC in a browser is to
>- have the browser add a small pretty graphic icon with a D on it
>to any graphic with a LONGDESC.
>- The browser can also have a setting which allows a user to HIDE
>doesn't like them doesn't have to see them.  (they could always see
>there was a LONGDESC and jump to it by right clicking on an icon and
>looking at the pop up menu).
>- If a page has a LONGDESC and an old fashioned D-LINK, then the
>browser could see that they pointed to the same place and hide the
>fashioned D-LINK  (or the new one - no matter)
>- Screen readers could recognize the LONGDESC Graphic (it would
>be a graphic since it was put there by the browser, not the html
>and either pronounce it or skip it when reading, preferably at the
>users choice.  (remember the user can cause them to not be displayed
>if they like anyway)
>Just some thoughts as of today.
>> Not true.  I've had ALT text in Internet Explorer longer than 3K
>> characters.  SGML/HTML do not define a maximum length.
>Not true.  Although not usually enforced, HTML's SGML declaration
>*does* define a maximum length (see
>                  ATTSPLEN 65536
>This limits the total length of attribute names and values, as
>contained in the start-tag, to 2^16 characters.
>                  LITLEN   65536
>This limits the length of any one attribute literal to 2^16
Jon Gunderson, Ph.D., ATP
Coordinator of Assistive Communication and Information Technology
Division of Rehabilitation - Education Services
University of Illinois at Urbana/Champaign
1207 S. Oak Street
Champaign, IL 61820

Voice: 217-244-5870
Fax: 217-333-0248
E-mail: jongund@uiuc.edu
WWW:	http://www.staff.uiuc.edu/~jongund
Received on Thursday, 23 April 1998 10:27:08 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:27 UTC