W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > October to December 2006

RE: WCAG 1.0 or 2.0?

From: John Foliot <foliot@wats.ca>
Date: Thu, 19 Oct 2006 14:30:27 -0700
To: "'Christopher Hoffman'" <christopher.a.hoffman@gmail.com>, "'John Foliot'" <jfoliot@stanford.edu>, "'WAI Interest Group list'" <w3c-wai-ig@w3.org>
Message-ID: <010701c6f3c5$cf0424e0$428e40ab@Piglet>

Christopher Hoffman wrote:
> 
> For real-world example, consider a dynamically-loaded image whose
> subject isn't necessarily known ahead of time. For example, my
> personal site has a "photo badge" from Flickr with six images that
> all have the same alt attribute: "A photo on Flickr." I would argue
> that such an attribute is perfectly reasonable, since I have no way
> of knowing whether a given randomly selected photo will be of my dog,
> of me on vacation, or of anything else.      

Chris, if you feel that the following is "perfectly reasonable":

A photo on Flickr. 
A photo on Flickr. 
A photo on Flickr. 
A photo on Flickr. 
A photo on Flickr. 
A photo on Flickr.

...then I can't argue it with you, although do you not think that it is
"unreasonable"?  What actual value is there here?  Outside of repeating
myself repeating myself (which alone could be annoying) it adds no
contextual information for the end user, so why bother?  To be "compliant"?

I would even suggest that it might be more appropriate to go:

<div title="Six random photos from my Flickr account">
	<img src="./path/" alt="" />
	<img src="./path/" alt="" />
	<img src="./path/" alt="" />
	<img src="./path/" alt="" />
	<img src="./path/" alt="" />
	<img src="./path/" alt="" />
</div>
 
...at least here there is some useful data for those that can access it.
Yes I know that not all AT (or AT configurations) will supply the title
information, but at least you as content producer have tried, using "W3C
approved technologies". 

I'm not looking to pick on you specifically Chris (you just opened the door
for me to go on...), but it never ceases to amaze me how many creative
developers/designers will talk about "user experience", and yet not think
outside of the box about "alternative" experiences.

> 
> Of course, I (or Flickr) could feed the photos from a database
> pre-matched with appropriate alt texts... 

Now you're talking.  This kind of "service" would be fairly trivial (I would
suspect) for Flickr to implement, and why they haven't yet done so is a bit
of a mystery (given the other bells and whistles they *do* provide).  At the
end of the day, it would be incumbent of the photo poster to also provide
appropriate alt text as part of the upload, but adding a bit of text along
with a JPEG file to the database is child's play.  And if they were to
borrow a page from Google's Image Labeler
[http://images.google.com/imagelabeler/] I think that you would begin to see
some powerful "alt text labeling" starting to emerge. "A photo on Flickr."
indeed...

> 
> On further thought, certainly the question of whether the image
> conveys additional meaning that isn't conveyed in the surounding text
> is an important one.

Yes, pity that LONGDESC is not more widely used and supported, as your
scenario below illustrates a perfect opportunity/use case

> An image of a planned escape route out of a
> building in case of a fire that has an alt attribute of "a picture"
> would surely be bordering on negligent if not criminally
> irresponsible.     
> 

This is where "compliance" and usability part company.  But of course, you
*are* correct, it would be criminal, yet the way things are currently worded
(especially Section 508, which is a Standard, as opposed to WCAG which is a
guideline), "a picture" would satisfy the requirement.

JF
Received on Thursday, 19 October 2006 21:30:56 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 5 February 2014 07:13:31 UTC