Re: The alt="" attribute

On Tue, 26 Aug 2008, Matt Morgan-May wrote:
> > 
> > When a restaurant has a picture hanging on the wall, do you require 
> > the restaurant to describe the image to you, or is the image of no 
> > consequence to you?
> 
> As has been mentioned before, comparisons to the physical world are 
> quite often irrelevant. There has been a lot of work done over the 
> decades on talking signs, for example, using directional RF and 
> specialized receivers, but the infrastructure required to make it 
> universal is prohibitively expensive -- in the trillions of dollars, at 
> least. The cost of putting @alt on an img, by comparison, in 
> infinitesimal. There is always a cost-benefit analysis done on 
> accessibility work, and in this case, at least, it is absolutely 
> clear-cut. Adding @alt is one of the simplest and cheapest accessibility 
> remedies, ever.

The point of my example is that even if the accessibility community had 
unlimited funds and could install all manner of hardware in this 
hypothetical restaurant, they _still_ wouldn't wire up the paintings to 
announce themselves to blind patrons. Equivalently, they wouldn't wire up 
the radio to display the lyrics and sheet music of the songs being played 
for the benefit of deaf users. Both the painting and the background music 
are purely decorative and not in any way "content" that a user needs or 
even wants if he lacks the relevant sensory capabilities.


> > If people do want to do actual research here, I would love to see more 
> > usability study videos of blind users using the Web without guidance, 
> > to see how they actually interact with images. I think that that is 
> > the level of research we need to really make more informed decisions.
> 
> The plural of "anecdote" is not "data". I've run numerous usability 
> tests, and they can help an organization make good guesses about what 
> they should do next. But they're just guesses. I wouldn't use the 
> results of usability testing or focus group-style polling to settle a 
> bar bet, much less a specification debate.

Certainly if more quantitative data can be collected that would be great 
too. Right now, we have neither.

(In practice, it turns out that you get diminishing returns beyond about 
seven usability studies -- people aren't as different from each other as 
one might think. The purpose of usability studies here is to see how users 
interact, as opposed to their opinions. For example, it's obvious from 
having watched usability studies of visually impaired users browsing the 
Web that there is a ridiculous level of repetition involved in browsing 
the Web with a screen reader, but it is so pervasive that users have 
become numb to it and don't even notice it any more. In several cases, I 
noticed this repetition actively harming the user, yet in each case the 
user assumed the problem was with his or her own interaction rather than a 
bug in the software.)

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'

Received on Tuesday, 26 August 2008 22:30:16 UTC