How to make arbitrary text selections in a screen reader render visually

[Harnum, Alan]

I don't have great suggestions at first glance but I found the two links helpful in reading about the issue
-  https://wiki.mozilla.org/Accessibility/Virtual_buffer_smash
- http://tink.uk/understanding-screen-reader-interaction-modes/

It seems to me that the practical answer lies somewhere in indicating with ARIA that a section of page text is interaction-driven, causing the screen reader to operate in application mode, and bypassing the virtual buffer behaviour entirely. I don't know what that might look like in practice, but to the extent I understand the virtual buffer it seems the solution would have to involve bypassing it entirely somehow


From: fluid-work <fluid-work-bounces at lists.idrc.ocad.ca<mailto:fluid-work-bounces at lists.idrc.ocad.ca>> on behalf of "Hung, Jonathan" <jhung at ocadu.ca<mailto:jhung at ocadu.ca>>
Date: Wednesday, February 1, 2017 at 11:22 AM
To: Fluid Work <fluid-work at fluidproject.org<mailto:fluid-work at fluidproject.org>>
Subject: How to make arbitrary text selections in a screen reader render visually

Hi everyone,

In many of our designs (i.e. Metadata authoring, My Lifelong Learning, UI Options, and OER Authoring Tool) we have a feature which allows a user to select any arbitrary text and “do something with it”. This can take many forms such as:

* Selecting text to make an annotation
* Selecting text to perform some formatting (i.e. make it bold, or insert a link)
* Mark a section that has a footnote or citation

A tricky implementation issue has been brought up by our friends at Hypothes.is<http://Hypothes.is>:

If a user is using a screen reader and keyboard to interact with the website content, how do you get the interactions with the AT’s screen buffer to render visually? The issue here is that screen readers that use Virtual Buffers (like NVDA and JAWS), the user interacts with the buffer and nothing in the browser is affected. This is unfortunate for a screen reader user who has vision as they do not get any visual cues.

How would one implement a selection feature such that a user of a screen reader (with a virtual buffer) can make selections, and for the selections in the virtual buffer to appear visually in the browser? Is this even a good approach, or is there another way?

Thoughts and suggestions appreciated.

- Jon.

---
Jonathan Hung, Inclusive Designer
Email: jhung at ocadu.ca<mailto:jhung at ocadu.ca>
OCAD University
Inclusive Design Research Centre
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.idrc.ocad.ca/pipermail/fluid-work/attachments/20170202/20a80062/attachment.html>