[uf-discuss] Use of <abbr> (also <object>) and Accessibility

[Tantek Ç elik]

On 9/21/06 3:20 PM, "Ben Ward" <lists at ben-ward.co.uk> wrote:

> On 21 Sep 2006, at 20:05, Andy Mabbett wrote:
>> There're some interesting views about the use of <Abbr> by
>> microformats,
>> on the Accessify forum:
>> 
>>         <ttp://www.accessifyforum.com/viewtopic.php?t=6167>
>> 
>> also an interesting take on the non-use of <object>.
> 
> I'm a little confused reading through that: The debate about whether
> the use of ABBR is fine and well, I follow that, but maybe I came too
> late the Microformats community to understand the references to OBJECT.

OBJECT has been problematic in Safari for quite some time, and still is
AFAIK.

In terms of bug-reporting, I'd suggest pointing the Safari team at the draft
HTML 4.01 test suite to *at least* pass all the test cases there.

 http://www.w3.org/MarkUp/Test/HTML401/current/


> I was under the impression that the OBJECT bugs in Safari were
> related to the first generation include pattern,

That was the second time the OBJECT bugs got in our way.


> prompting the  
> creation of rel-include. Can someone link me to a post in the
> archives that concerns the OBJECT and ABBR for datetimes issue the
> Accessify thread raises?

http://tantek.com/log/2005/01.html#d26t0100

Following the descriptions contained within that post, it is trivial to
construct perhaps a half dozen or so OBJECT test cases which Safari fails.


> In general, the _theoretical_ ABBR discussion is well based, datetime
> is a new use and perhaps it stretches the element too farŠ maybe not.

The only objectionable examples listed in that accessify thread represent
edge cases, rather than the common case.  It's (I hate to say this, but
typical) reasoning by edge case rather than reasoning by 80% case.

The example I gave (which was then misquoted in the accessify thread) was:

 <abbr title="20050125">January 25th</abbr>

"January 25th" *is* an abbreviation in that context for *2005* January 25th.

When used with times in typical use with hCalendar, the ISO8601 datetime
typically includes the timezone offset in addition to the information
included inline in the element, and thus again, it is proper semantic use of
the abbr element to markup an abbreviation.

It is a *more* specific use of the abbr element, but certainly fits within
the broader abbr semantic.


> The thing is though, people love to talk their interpretation of
> the semantics and expected behaviours but I'm yet to see anyone with
> access to assistive technology produce examples to demonstrate
> problems (or otherwise).

A very good point Ben.  So far the critics have only been chicken-littling
which we should all have very little patience for.

Tantek