<scribe> scribe: mf

screen reader vocabulary

matt_king: I think I was combining two different topics here.
... one topic is how we write sr instructions
... when I was writing expectations, I was struggling with the vocabulary to use. words like 'read', 'operate', 'convey', 'navigate', 'understand', etc.
... I think we need a really limited vocabulary so that it is very clear what exactly the expectation is.
... I also wonder if I'm blending screen reader behavior and screen reader behavior. It might be good to keep these concepts separate.
... menu example: when using arrow keys to navigate within the menu, are you reading the menu, navigating the menu, or both? Do we need a different word when in reading mode vs interaction mode? These are really two different operations and can have different expectations associated with them.
... for example, boundaries would be conveyed in interaction mode, but perhaps not in reading mode.
... other way around
... in interaction mode, you will also often hear posinset information, but perhaps not in reading mode
... so maybe 'navigate' just means you are moving a cursor and we define each mode in addition to the word 'navigate'

Example: task - read navigation region; expectation: region role, name, and boundaries are announced.

matt_king: there are many different ways to get to a landmark region. You can get into one in either reading or forms mode. The expectations are different in both. In forms mode, you might be aware that you crossed a boundary but you might not be aware where that boundary is.
... in this version, I used the word 'read' whenever we were in reading mode. But that's confusing because you can read the current line/position/etc without changing the cursor position. You can also read by navigating.

Yohta: my initial thought is to describe the vocab in written text

matt_king: if you change the state of the checkbox, you are operating the checkbox.
... discussed different ways to 'operate' menu items, via screen reader shortcuts and non-screen reader commands
... do we need different verbs for those?

<Matt_King> mf: how detailed should the data collect be? there are so many possibilities for output.

<Matt_King> mf: If we are too detailed, it could slow everything down too much.

<Matt_King> mf: Perhaps not our responsibility to be that detailed.

<Matt_King> mf: How do we get the most value for effort? 80/20 rule that we should apply? 80% coverage with only 20% of the commands?

<Matt_King> jg: We need to capture the effect of the aria markup.

<Matt_King> jg: if some of the aria is not exposed that is important to user, we need to capture that

<Matt_King> jg: difference between modes is kind of fuzzy

<Matt_King> jg: Compare VO and JAWS

matt_king: if we have different meaning for 'read', 'navigate', and 'operate', I want to make sure we are all on the same page.
... combobox example - when number of suggestions changes, the screen reader user is 'informed' of the change.

jg: have you thought of the word 'action'?
... navigate is usually tab key or arrow keys, and, read is command to read current item/position, operate (or activate) is usually space, enter, or click.