Re: ARIA 1.1: Deprecate @aria-grabbed and @aria-dropeffect

Restarting the thread since ARIA drag/drop hasn't been deprecated yet.

Replying to one question by Bryan:
> If these attributes were no longer supported, how would a non-sighted user know when an item was ‘grabbed’, and where within the tree a focusable element was enabled for being ‘droppable’?


I think you're asking the wrong question, Bryan. Even if a web author did add the ARIA drag/drop attributes and a screen reader user knew they were drag or drop targets, the drag/drop functionality would not magically start working with a screen reader. The only recommended authoring pattern conflates drag attributes with a defaultAction-based selection model (e.g. press-to-select or drop) that is totally unrelated to actual drag/drop functionality. As a inverse case in point, none of Jon Gunderson's ARIA drag/drop examples work by using mouse-based drag/drop.

This is why ARIA drag/drop is a non-starter and needs to be deprecated. 

James



> On Jun 21, 2015, at 4:09 PM, Bryan Garaventa <bryan.garaventa@ssbbartgroup.com> wrote:
> 
> I agree that the way that drag and drop is worded is problematic, but I would like to ask about the value of providing textual equivalent feedback using these attributes.
>  
> I have never seen the value of using the ARIA drag and drop attributes as indicated in the spec for implementing behavior, because that is confusingly worded.
>  
> However, when there is drag and drop functionality attached to a particular widget or element, these attributes are useful for conveying that this functionality exists for non-sighted screen reader users. This is more in the sense of a textual equivalent.
>  
> E.G Within GWT, there is a tree component. You can do many things with this, and one of the options is to drag and drop tree leaf nodes from one place to another.
>  
> If these attributes were no longer supported, how would a non-sighted user know when an item was ‘grabbed’, and where within the tree a focusable element was enabled for being ‘droppable’?
>  
> From: Richard Schwerdtfeger [mailto:schwer@us.ibm.com <mailto:schwer@us.ibm.com>] 
> Sent: Friday, June 19, 2015 11:03 AM
> To: lwatson@paciellogroup.com <mailto:lwatson@paciellogroup.com>
> Cc: 'James Craig'; 'WAI Protocols & Formats'
> Subject: RE: ARIA 1.1: Deprecate @aria-grabbed and @aria-dropeffect
>  
> Just to be clear HTML5 tried to get common drag and drop done and it bombed. Perhaps the new web applications group can do a better job. 
> 
> 
> Rich Schwerdtfeger
> 
> Léonie Watson ---06/19/2015 06:46:57 AM---> From: James Craig [mailto:jcraig@apple.com <mailto:jcraig@apple.com>] > Sent: 19 June 2015 10:43
> 
> From: Léonie Watson <lwatson@paciellogroup.com <mailto:lwatson@paciellogroup.com>>
> To: "'James Craig'" <jcraig@apple.com <mailto:jcraig@apple.com>>, "'WAI Protocols & Formats'" <public-pfwg@w3.org <mailto:public-pfwg@w3.org>>
> Date: 06/19/2015 06:46 AM
> Subject: RE: ARIA 1.1: Deprecate @aria-grabbed and @aria-dropeffect
> 
> 
> 
> 
> > From: James Craig [mailto:jcraig@apple.com <mailto:jcraig@apple.com>]
> > Sent: 19 June 2015 10:43
> > In an effort to reduce the author complexity of ARIA, I'd like to propose
> the
> > spec's first deprecations: @aria-grabbed and @aria-dropeffect.
> 
> +1
> 
> [...]
> 
> > Accessible drag & drop is a feature that may be better left to native
> > implementations. It could potentially be solved by some future version of
> > ARIA, but I do not believe @aria-grabbed and @aria-dropeffect do the job.
> > It's a bad API that should be culled from the 1.1 spec.
> 
> 
> Do you think it would be worth proposing an HTML5 extension for this?
> > 
> > In case there is any objection: I could be convinced to drop the call for
> > deprecation if anyone can point to a single real-world web application
> (not a
> > test case) that works well in any browser+screenreader combo. The example
> > should use @aria-grabbed and @aria-dropeffect accurately in conjunction
> > with native or scripted drag and drop behavior.
> 
> 
> Even in test cases using ARIA to spec, I haven't yet found an example that
> works reliably across all (or even most) browser/AT combinations.
> 
> 
> Léonie.
> 
> -- 
> Léonie Watson - Senior accessibility engineer
> @LeonieWatson @PacielloGroup PacielloGroup.com <http://paciellogroup.com/>