> On Jun 30, 2016, at 4:18 AM, James Craig <jcraig@apple.com> wrote: > > >> On Jun 29, 2016, at 2:18 PM, Richard Schwerdtfeger <richschwer@gmail.com <mailto:richschwer@gmail.com>> wrote: >> >> roledescription is new for ARIA 1.1 and so the ATs really have not used it yet. > > It landed in WebKit not long after the feature went into the 1.1 spec. It works at least two releases of both iOS and macOS. > > You can try it out on the "arrow" within this bitmap SVG. > http://cookiecrook.com/longdesc/svg_bitmap/ <http://cookiecrook.com/longdesc/svg_bitmap/> >> That said we can work with them to ensure that they don’t say anything. >> >> net net. That should be doable. It is a great suggestion by Jason and it would avoid having another role. > > I consider an empty @aria-roledescription more of a temporary workaround than a long term solution. It suffices to postpone the other role discussion, but it's be ill advised to promote that technique as a recommended authoring practice. > I agree it is a temporary solution. However, if we can’t get the group to agree on the new role it may be needed as a technique. At this point my issues with the new role have been addressed. > James > >> Rich >> >> >>> On Jun 29, 2016, at 3:54 PM, Amelia Bellamy-Royds <amelia.bellamy.royds@gmail.com <mailto:amelia.bellamy.royds@gmail.com>> wrote: >>> >>> [Jason] Couldn’t you do this in ARIA 1.1 with aria-roledescription=””, thereby suppressing presentation of the image role by the screen reader? >>> >>> >>> I'm okay with recommending that the img role continue to be used for text-as-graphics, but I'm curious about Jason's suggestion of an empty role description value. Has anyone tested whether that actually works and makes sense when exposed by a screen reader? >>> >>> This wouldn't affect anything in ARIA 1.1, but would affect authoring guidance. >>> >> >
Received on Thursday, 30 June 2016 16:12:34 UTC