Re: @title's relation to accessibility

2007-09-04 06:33:12 +0200 Sander Tekelenburg <st@isoc.nl>:
> At 14:19 -0500 UTC, on 2007-09-03, Jon Barnett wrote:
>> On 9/3/07, Sander Tekelenburg <st@isoc.nl> wrote:

[ … @title anything to do with universality or accessiblity? … ]

> important, because [1] authors mix up @alt and @title, and [2] the HTML5
> draft now says that UAs must present @alt and @title differently and [3] the
> HTML5 draft now says that @alt is no longer required, making authors wonder
> again/more about @title...
> 
> So there appears to be confusion, and we should try to ensure that the spec
> ends that confusion.
> 
> So far, IIRC, the only universality aspect of @title that I think actually
> exists is as an attribute to iframe.

[ An aside: in iCab the IMG and IFRAME settings share the same preference panel.]

Sighted users how @TITLE can be useful as an accessibility attribute by considering how TITLE is used in tabbed web browsing. TITLE and @TITLE are discussed in same context in HTML4 [1]. And with tabbed browsing, the TITLE _element_ got a new use - as tooltip: Hovering over a tab, reveals the page title,  in all "tabbing" browsers. (Exception: Opera shows a  page preview, with the TITLE element as heading,  iCab does too, but still wtih TITLE as tooltip). [*]  

[*] Tab preview could also be a model for what UAs could make _both_
    equal content/alt text **and** the image itself to users: Hovering
    over IMG, one could get to see the @title as well as the @ALT text
    - and a possible @longdesc link. With image display turned off, it
    could work the other way around.

The usefullness of @TITLE for accessibility matters, depends on how long the alternative content is - and how difficult it is to load or read (compared with @TITLE). HTML4 specifies @LONGDESC for the @TITLE of FRAME and IFRAME [2] - and probably for 2 reasons: AT UAs may have (had more) trouble with loading/reading FRAME and IFRAME pages. And, FRAME/IFRAME is also often quite long. 

For IMG, the @ALT text is often short and probably there are no problems loading it. However, the @ALT text can still possibly be long - at least a @LONGDESC can. And no matter how crucial an image is, one can very often just skip it and still get a meaningful text. Therefore, a short title can be useful for creating a short glimpse of the @ALT text if the @ALT is long. (For typical text elements, like P, I do not see that it has particular relevance as accessibility feature.)

To decrease confusion about how accessibility is linked to @TITLE, perhaps these rules could be worth considering: 

 - say that @TITLE «tells about», while @ALT «is the story itself»
 - in general, use @TITLE when it is beneficial for deciding about
   whether to read a longer text or opening another page/resource
 - require @TITLE  when @LONGDESC is used [†]
 - decide a @ALT text limit, wherupon @TITLE SHOULD be added?
   (AT UAs could think: If long @ALT, then read @TITLE first)?
 - @TITLE on A-links a MUST?

[†] Btw, I wonder if one can say something about how @ALT should look when @LONGDESC is used? E.g. can @ALT be emtpy when @LONGDISK is used? Or should @ALT never be empty when @LONGDESC is used, since the alt text will kind of be used as link text/hypertext? Should the @ALT then describe the @LONGDESC - or should it be a short @ALT text for the image, while @TITLE tells about the @LONGDESC? Or the user of @ALT vs @TITLE be related to whether the IMG is part of a link - which possibly have its own @TITLE - which could lead to another place than to wher ethe @LONGDESC leads? 

For instance, in photo album - like Flicker, which have 2-3-4 levels (different sizes) of «thumbnail» images - until the largest version: For AT users, there might not be any point in looking at all these different sizes. Thus each such image - in all its size versions - could have the same @LONGDESC. Thus, we have practical example, I think about where the the A-link itself can point one place (to a larger version of an image), while the @LONGDESC can point to a independent description. If the descriptions could be collected independently from the images, then it would also not be difficult to present them as a whole, on a page particulary aimed at giving a textual version. (The proposal to have an ALT element comes to mind ...)

[1] <http://www.w3.org/TR/html401/struct/global.html#h-7.4>
[2] <http://www.w3.org/TR/html401/present/frames.html#edef-FRAME>
-- 
leif halvard silli

Received on Tuesday, 4 September 2007 16:36:52 UTC