[uf-discuss] proposed title-design-pattern is not backwards compatible, too big of a change

[Tantek Ç elik]

On 4/29/07 7:44 AM, "Brian Suda" <brian.suda at gmail.com> wrote:

> On 4/29/07, Jeremy Keith <jeremy at adactio.com> wrote:
>> If we were to find an existing HTML element that was semantically
>> suited to encoding datetime and/or geo information *and* didn't cause
>> problems with assistive technology, then I would jump all over it and
>> agree wholeheartedly that the title-design-pattern should be
>> restricted to that particular element. But I don't believe such an
>> element exists.
> 
> the whole discussion begs the question about what people with
> assistive technologies ACTUALLY think? A while ago there was a whole
> report about who screen readers fail with AJAX apps, then someone
> actually ASKED some blind folks if they could navigate the site...
> they managed to do so just fine.

Agreed.  This is definitely one of the reasons why we emphasize real world
examples and data.


> We are naively ASSUMING that people with assistive technologies NEED
> our help. I would prefer, before WE think we can hand the right
> soltion down from on high, that someone who uses a screen-reader as
> their main browser give their feedback.

>From talking with them at SXSW I do know that at least Derek Featherstone
and James Craig have done some real world research on what ABBR does, but
what we need to do now is to document their results in detail on the wiki in
order to be sure that we are addressing the specific problems found, rather
than potentially theoretical expansions/generalizations to problems.

James, could you start with documenting the specific assistive technologies
that you are testing (with version numbers, estimated # of users etc.)

 http://microformats.org/wiki/assistive-technology

The next step would be to create a wiki page that documents tests that have
been done with results, using specific assistive technologies, and URLs to
specific content in the wild that uses microformats with abbr today.
Perhaps something like:

 http://microformats.org/wiki/abbr-assistive-technology-tests



> We skirt the issue by moving data to the title attribute of
> alternative elements, how do we know screen-readers now or later won´t
> read out those as well?

We don't.  And frankly, at this point, rather than take another "best guess"
as I did with the abbr for datetime, and later extended to other data types
(numbers, enumerated type values), any proposal should be tested as well in
those same specific assistive technologies with test cases.


> we are coding around a problem by potentiall
> creating other ones and ignoring the semantics of the HTML spec in the
> process.

Agreed.  Adopting a proposal like the one that has been put forth without
sufficient research and documentation may not actually make the situation
any better in practice (while being semantically weaker).

Tantek