WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: address tag

for

From: Jukka K. Korpela
Date: Feb 20, 2007 3:00AM

On Tue, 20 Feb 2007, Alastair Campbell wrote:

> Because of the uptake, people are actually building user agents to use
> them:

Of course one can create user agents that utilize whatever fancy markup
might appear. You can do it even without any programming, just by writing
a user style sheet (especially if you use a browser that supports
generated content so that you can turn attribute values into visible
texts).

There's so much you can play. But this won't help accessibility. Rather,
it makes things more difficult, since it results in artificial and even
semantically wrong markup, thereby confusing people who use browsers that
utilize the defined meaning of markup (as defined in authoritative
specifications).

> Even now, if you add microformats to an events listing, you can then add
> a link to 'subscribe' (e.g. at the bottom of
> http://ukwindsurfing.com/events/).

I have no idea of what you mean by that, but the text at the bottom,
"Apple iCal and Thunderbird users can subscribe",
looks like discrimination, not accessibility.

> (NB: I'll be changing the <abbr> to <span>, due to the title issue. I
> don't think it's an issue now, but it could be.)

Of course it's an issue, and it will remain an issue even if you switch to
the meaningless (semantically empty) element name "span". The title
attribute still means "advisory title" by the book, but you are not using
it as a title but hidden container for data for some processing. The vast
majority of users will see the title attribute's value if they mouse over
the element.

If I see "20070711" when mousing over "Nov 17", I can guess what's going
on, but it really won't help me. To many people, it's just gibberish.
That is, it helps nobody but hurts many. (Changing "abbr" to "span"
does not change this, though it removes the disturbing and confusing
"dotted underline" on some browsers.)

> Personally, I would do whatever works best for the user.

Then you surely should not create obstacles to the vast majority, by
exploiting tricks that are supposed to work on some marginal applications
but are known to be processed quite differently (and more to the book) by
popular browsers.

--
Jukka "Yucca" Korpela, http://www.cs.tut.fi/~jkorpela/