RE: Flexbox Spec Wording



From: Richard Schwerdtfeger [mailto:schwer@us.ibm.com]
Sent: Sunday, April 10, 2016 4:57 PM

What they need to do is create a mapping spec. that exposes the information to the AT in an order that reflects the order provided. The text that should be exposed should be on the order of:

The order in which the user agent exposes the elements to assistive technologies MUST reflect the order defined by the Flexbox used.

One of the points made in the CSS Flexbox draft, as currently written, and emphasized at TPAC last year, is that the order as defined by the “order” property may not correspond to a logical reading order, which is one reason why they advise authors to ensure that the DOM tree order is logical/reasonable.

Advising assistive technologies (e.g., screen readers, synchronized highlighting tools) to follow the presentation order won’t solve this problem.

I work in an organization where authors are prepared to do whatever is necessary to create a usable and accessible experience. Such authors need to have the flexibility to define both the logical reading order and the visual presentation order independently of each other, and to take advantage of the features provided by Flexbox and other forthcoming CSS specifications to change the visual layout depending on the capabilities of the device on which the content is rendered.

I am also concerned by the differences among screen readers in the handling of flexbox content. The example developed at TPAC last year was telling: VoiceOver under OS X followed the visual presentation order, apparently by increasing y-coordinate of the bounding boxes of the elements, as I recall (this may be inaccurate). The results were disastrous from the standpoint of a logical reading order unless the user set the configuration option that required the DOM tree order to be followed.

To summarize:

I’m raising the question of whether following the “order” property, in ascending sequence, would in a significant range of cases turn out to be the wrong response from the standpoint of generating a rational reading order.

I’m also very supportive of making it possible to decouple visual presentation from the logical reading order, so that the author can define both independently of each other. The current draft achieves this, as I understand it, allowing the author to define their DOM tree order and yet to set up the desired visual layout.


________________________________

This e-mail and any files transmitted with it may contain privileged or confidential information. It is solely for use by the individual for whom it is intended, even if addressed incorrectly. If you received this e-mail in error, please notify the sender; do not disclose, copy, distribute, or take any action in reliance on the contents of this information; and delete it from your system. Any other use of this e-mail is prohibited.


Thank you for your compliance.

________________________________

Received on Monday, 11 April 2016 15:46:25 UTC