W3C home > Mailing lists > Public > public-html@w3.org > April 2008

Re: several messages relating to the alt="" attribute

From: Dr. Olaf Hoffmann <Dr.O.Hoffmann@gmx.de>
Date: Fri, 11 Apr 2008 19:12:07 +0100
To: public-html@w3.org
Message-Id: <200804112012.07327.Dr.O.Hoffmann@gmx.de>

Ian Hickson:
> On Fri, 11 Apr 2008, Dr. Olaf Hoffmann wrote:
> > Ian Hickson:
> > >The current text in the spec requires alt="" in all cases
> > >except when the page is generated in a manner where alternative text is
> > >not available, or when there is no possible way to provide text that is
> > >in any way a replacement for the image.
> >
> > Ian Hickson:
> > > Things that are impossible just take longer.
> >
> > If authors know what the purpose of the image is, it should be never a
> > problem to add it. If not, there is no need to put the image on the web
> > page.
>
> How about when the page is generated by someone who doesn't know what the
> image is, e.g. on Flickr?

Well, I never used 'Flickr' - I prefer my own image gallery without any nasty
restrictions due to the abilities of the authors of such tools.
 
This looks pretty much like distributed responsibilities with no one
responsible at the end, if it does not work (we already had much worse
cases than this in history).
a) The producer of the digital camera can provide some possibility to put an
alternative text in the EXIF including maybe UTC and GPS data.
b) programs to manipulate digital photos can provide input of alternative text
c) if there exist specific programs to upload photos on a specific service, 
they can provide a possibility to add text
d) such services can try to provide an input as for the name of the file
itself in the same step as the upload occurs (and can try to extract EXIF data
from a) and b) if this is specified in a useful way)
e) if a service does not provide a possibility to add alternative text, people
can simply use a better service.

Therefore everyone has some responsibility somehow to help to get this
improved, but finally it is the last one in this chain with the main 
responsibility. If there is currently no way to create this automatically,
the author of the gallery script has to provide a possibility to add 
alternative text. If the user thinks, that the photo has no meaning,
and leaves this input empty, this is a clear indication, that the photo
itself has no meaning and is purly decorative, therefore alt="" is
the intention of such an image author.
Therefore there is never a situation, where no human has the
responsibility at the end, because in our time things do not really
happen as a miracle without any reason or cause. What really exists 
is laziness, indifference and ignorance, but this is no reason to kill 
an accessibility feature already in the recommendation.

And if this possibilities are missing, this is mainly an indication 
of a miserable script or service, nothing to care about in (X)HTML
recommendations.
On the other hand, a gallery with only raster images will not
be interesting for people without the abilities to see them, 
therefore it is not really a problem, if such galleries often
use alt="" if the image author does not provide an alternative 
text.


>
> > >Rorschach inkblot test.
> >
> > could be the value of the alt attribute and anyone can look
> > on the web, what kind of test this is.
>
>    <figure>
>     <img src="r14.jpeg" alt="Rorschach inkblot test">
>     <legend>Rorschach inkblot test #14</legend>
>    <figure>
>
> ...is not any more accessible, and is arguably less accessible, than what
> the spec suggests now:
>
>    <figure>
>     <img src="r14.jpeg">
>     <legend>Rorschach inkblot test #14</legend>
>    <figure>


I think, currently it is not required to use figure and legend around
img - I personally like it, but if those gallery scripts do not manage
to add required alt attributes, they will not manage this too.
If it is required in HTML5 to put any img element in a figure with
a required not empty legend element, this is better than this
old img with alt. As error processing rule it could be useful,
that the img is not displayed, if there is not content in
legend or no alt attribute inside img (maybe a joke ;o)
Already for accessibility object is much better than img as
another approach, but up to now authors often prefer the 
old img - the complete img could be replaced with something 
like object, but I think, this is out of the scope of the 'HTML5' 
design principles.

In my own art gallery similar things happen as with figure and
legend, the main part of alternative text is on the page accessible 
for everyone, not inside the alt-attribute or inside an object element. 
Therefore typically I do not need much of this alternative
or fallback because I provide both anyway - after some
discussion with accessibility experts on the problem of an
alternative text for abstract arts and looking on the problem,
if and how people understand abstract arts in general. 
But for the huge group of designers with limited abilities
of textual articulation this required alt attribute is a nice
help and accessibility feature to trigger the brain to start
to think, what they are currently doing - it is often only
decorative, therefore alt="", but if they really have something
to say - this is a chance to put some more glyphs in it.
Received on Friday, 11 April 2008 18:16:32 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:16:14 GMT