W3C home > Mailing lists > Public > w3c-wai-er-ig@w3.org > October 1999

Re: BLINK repair mechanisms (calling all CSS gurus!)

From: Gregory J. Rosmaita <unagi69@concentric.net>
Date: Thu, 21 Oct 1999 16:05:05 -0400
Message-Id: <4.1.19991021150453.00aab6b0@pop3.concentric.net>
To: "Chris Ridpath" <chris.ridpath@utoronto.ca>
Cc: Evaluation & Repair Interest Group <w3c-wai-er-ig@w3.org>
aloha, chris!

one small comment upon
        http://aprompt.snow.utoronto.ca/tags/blink1.html

whilst the screen-shot is commendably ALT-texted and LONGDESCed, would it be
possible to employ the following strategy to expose the contents of a quote
real unquote LONGDESC, rather than simply repeat the ALT argument as the
LONGDESC argument as you do on the above-referenced page?

1. enclose the graphic in an <A HREF ...> </A> that points to the same URI is
used for the LONGDESC -- for example, blink1_screenshot_description.html

2. describe the screenshot in the body of the page
blink1_screenshot_description.html

3. use the ALT attribute for the screenshot image to point the user towards the
LONGDESC -- ALT="Description of the screen shot of the dialog used to validate
BLINK element"

this would allow someone with image rendering turned off (or for whom image
rendering is superfluous) to get an idea of what the screenshot contains...

the use of a hyperlink to point to the URI of a LONGDESC is an "Until user
agents..." technique for exposing LONGDESCs that i would like to "see" added to
the WCAG techniques document, as it is a strategy that is more palatable to
those who bankroll commercial sites than is the suggested "D-link" technique

please refer also to my comments on the topic archived at:
        http://lists.w3.org/Archives/Public/w3c-wai-er-ig/1999Aug/0023.html
and note that the examples of this technique that are implemented at Camera
Obscura (and referenced in the above-mentioned post) are encoded using a
"hidden" style declaration because 

a) "display : none" isn't very widely supported, and
b) i was illustrating the use of the technique for a developer who did not want
large screen shots that i re-encoded using this technique to be visually
demarcated with a border and for who objected to my setting the BORDER
attribute in the IMG element to zero, as there are still UAs in common use (he
alleged) that do not respect the BORDER="0" argument when used in an IMG
declaration

the genesis of using display : none (or the same-background-as-foreground
kludge) to pass information on to assistive technologies came from my
experience encoding HTML forms for commercial sites -- while i wanted to pass
on supplemental semantic information about the number of radio buttons in a
particular field, for example, to speech and refreshable braille users, the
person paying me to do the work didn't want the quote extraneous unquote
information to quote clutter unquote the form he had envisioned, so i used
stylesheets to hide the extra info...

gregory.

At 02:33 PM 10/21/99 -0400, you wrote:
>Gregory makes a good point for including the Hx as replacements for
>BLINK/MARQUEE. Another reason for suggesting Hx is that it might get the
>user thinking about using these in the future instead of BLINK/MARQUEE.
>
>Should we also allow the author to change the BLINK/MARQUEE to something
>else?
>
>To repair a BLINK/MARQUEE we could offer the author 3 choices:
>1) remove it
>2) change it to: [EM, STRONG, SPAN, (perhaps Hx)]
>3) change it to: [enter your own tag]
>
>For example, here's how the A-Prompt dialog might look:
>http://aprompt.snow.utoronto.ca/tags/blink1.html
>
>Chris

------------------------------------------------
Writing is easy; all you do is sit staring at a 
blank piece of paper until the drops of blood 
form on your forehead.            -- Gene Fowler
------------------------------------------------
Gregory J. Rosmaita <unagi69@concentric.net>
WebMonster and Minister of Propaganda, VICUG NYC
     <http://www.hicom.net/~oedipus/vicug/>
------------------------------------------------
Received on Thursday, 21 October 1999 15:59:57 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:01:29 UTC