RE: tab based navigation not as informative as virtual cursor navigation

Yes, that's totally expected. I don't wish to be rude, but do your testers actually know how people use screen readers? In my experience, a lot of testers have had no training and don't have any idea how screen reader users navigate within and between pages. As a result, they report a lot of nonsense.

Firstly, screen readers don't all work the same way. Professional products like JAWS and NVDA have two modes of operation, typically known as browse mode (where you are using the virtual cursor to navigate within the virtual object model) and forms mode (where you are interacting directly with the DOM). There is also application mode, but it is very similar to forms mode.

In browse mode, you can navigate to every element. In forms mode, you can only navigate between focusable elements, although you will also hear some other content, such as ARIA live regions. When you tab through a page, you can only navigate between focusable elements, so the effect is the same as if you were in forms mode. As such, you would not expect to hear a lot of the content.

To confuse matters further, even when you land on an element, you usually hear different things depending on which mode you are in.

In my experience, screen reader users almost never tab through a page. They only use the tab key to navigate between form controls.

Some cheap or free screen readers don't have a browse mode, and they can only navigate to focusable elements. Microsoft's Narrator used to be like that until just a few years ago.

Steve Green
Managing Director
Test Partners Ltd


-----Original Message-----
From: bryan rasmussen <rasmussen.bryan@gmail.com> 
Sent: 17 August 2021 10:27
To: WAI IG <w3c-wai-ig@w3.org>
Subject: tab based navigation not as informative as virtual cursor navigation

Hey,

I just ran into a situation with NVDA on Chrome where I had put quite a lot of effort into making a table accessible with labels and other information and when testers navigate through it with the virtual cursor it works great but when they navigate using the tab it doesn't read any labels and just reads out stuff like table row, clickable.

Is this expected?

Are there significant differences in different screenreaders using their virtual cursor to navigate through things and the information the tab gives you when you navigate?

Mvh,
Bryan Rasmussen

Received on Tuesday, 17 August 2021 10:14:05 UTC