[uf-discuss] ISO Dates and Durations using Style
Benjamin Hawkes-Lewis
bhawkeslewis at googlemail.com
Sun Sep 28 13:36:20 PDT 2008
Martin McEvoy wrote:
> One question what is actually wrong with using an extension eg:
>
> <span class="bday" style="-uf-content:'1968-01-04';">4th Jan, 1968</span>
>
> apart from the CSS validation issues, which when you compare it with the
> current <abbr> and @title accessibility issues doesn't seem so bad to me?
The only direct accessibility issue I can think of is that if you had a
tool that stripped out styling information before presentation to a user
who needed to apply their own formatting, the data would be stripped
along with the presentational suggestions.
There are a couple more indirect considerations:
1. Encouraging non-conforming code could produce more mistakes; without
a quality control process, mistakes are less likely to be noticed when
they only affect users with certain disabilities.
2. It would constitute a barrier of adoption to publishers aiming to
produce webpages that conform to accessibility guidelines that require
conforming or even just valid CSS and reduce the benefits of such
microformats reaching users with disabilities via sites designed to be
usable for them. For example, a page using this syntax probably could
not pass WCAG 1.0 Checkpoint 3.2 ("Create documents that validate to
published formal grammars.") and therefore could not pass at the
Priority 2 level.
These issues are by no means as significant as the accessibility
problems with abbr and title.
But in terms of what is wrong with it more generally, this solution is
every bit as non-conforming (though not quite as risky) as an HTML
custom attribute, with the additional ugliness of putting required data
into a layer intended for optional presentational suggestions. Given
that HTML custom attributes are a non-starter for microformats because
they build on existing standards, I can't see how this proposal is going
to gain any traction.
--
Benjamin Hawkes-Lewis
More information about the microformats-discuss
mailing list