W3C home > Mailing lists > Public > wai-xtech@w3.org > April 2008

RE: Request for review of alt and alt value for authoring or publishing tools

From: Ian Hickson <ian@hixie.ch>
Date: Wed, 16 Apr 2008 01:18:38 +0000 (UTC)
To: Jim Jewett <jimjjewett@gmail.com>
Cc: foliot@wats.ca, wai-xtech@w3.org, wai-liaison@w3.org, public-html@w3.org, list@html4all.org
Message-ID: <Pine.LNX.4.62.0804160109330.29828@hixie.dreamhostps.com>

On Tue, 15 Apr 2008, Jim Jewett wrote:
> Ian's point was that people won't really supply an alt -- even John 
> Foliot seemed to have forgotten.
> John's point was that he didn't forget -- the tool didn't *allow* him to 
> enter an alt.  Taking out the alt requirement will remove any pressure 
> on flikr to fix this.  On the other hand,
> (1)  if flikr had to publish "alt=_notsupplied" to claim validity, that 
> might annoy the coders enough that they would create at least an 
> optional way to supply alt text.
> And this leads in to the main reasons why "alt=_magicvalue" is different 
> from an omitted alt.
> (2)  There are already tons of pages (including this one) which 
> (wrongly) use alt="", and tons of other pages that (wrongly, for now) 
> omit the alt.  There always will be, because these are good 
> programmer-language defaults.
> There are not many pages saying "alt=_decorative" or "alt=_notsupplied", 
> and those values aren't likely to be supplied by accident -- so they 
> have a chance of remaining honest.

This is the only good argument I have seen for why we should make the alt 
attribute a required attribute.

I did consider this, however -- it has been suggested several times in the 
past. The problem is that I am skeptical that magic keywords won't be 
abused as much as the simpler syntax we have now. If anything, the 
longdesc="" attribute should show us how much authors are likely to use 
nonsense values. longdesc="" is supposed to take a URI, but a huge 
fraction of longdesc="" attributes have strings that are not URIs at all. 
(The most common longdesc="" value is the empty string!)

If we had reason to believe that these new magic values wouldn't be abused 
as much as the current proposal, then I'd agree. However, I think that 
that would be naive.

(Also, the _decorative value, if we made it required instead of alt="", 
would also make a large amount of existing content non-conforming, since 
people use alt="" today. But if we allowed both, then we'd lose the 
benefit of _decorative.)

> (3)  There could be more than one (or even two) magic tokens.  When I'm 
> browsing without images, I often want to make a special request for the 
> critical images.  I wouldn't grab _decorative.  I wouldn't grab most 
> images with a good alt.  I wouldn't want to grab images that are just a 
> "better" representation of the summary/description/caption.
> I would want to grab the main content.  __notsupplied tells me that 
> (unless the CMS is really lousy) the image is something a user supplied 
> specific to this page.  In this case, that would be the main image and 
> the teaser from the slideshow -- and two is less than the number of 
> images without an alt, according to someone else's count.

Flickr currently abuses alt all over the place. I have no confidence that 
they would use these new values correctly.

Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'
Received on Wednesday, 16 April 2008 01:19:28 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:51:35 UTC