Mobile Accessibility Task Force Teleconference -- 16 May 2014

<trackbot> Date: 16 May 2014

<Kathy> http://www.w3.org/WAI/GL/mobile-a11y-tf/wiki/New_WCAG_2.0_Techniques

<scribe> scribe: jeanne

<Kathy> https://www.w3.org/2002/09/wbs/66524/20140512_survey/results

<Kathy> http://www.w3.org/WAI/GL/mobile-a11y-tf/wiki/Technique_Development_Assignments

[intro to survey]

KW: We still need to identify which apply to mobile, mobile apps, and hybrid

<Kathy> https://www.w3.org/2002/09/wbs/66524/20140512_survey/results

KW: I reviewed the funka nu best practices and did a gap analysis which i added to the new techniques page

Survey Q1

KW: we could break this into multiple techniques and possibly a failure technique
... there are a lot of ways that someone can control their device: bluetooth keyboard, switch, etc.
... what would be a path for WCAG for keyboard access?

JS: Alan had a good point that if the device doesn't support bluetooth or any other connection, it seems overboard to require the app to do so. I think we need to word the technique around this.

JA: If you have an app on a phone that doesn't support addon devices, then WCAG cannot be supported.
... you could never meet that

JS: Isn't that Accessibility Supported Technology?

JA: That's a part of it.
... Example: if you have audio description for a multimedia presentation, but didn't have enough room to pause, and you made extended audio descriptions, you could meet WCAG AAA, but not AA.

KW: If you could do everything with the keyboard, but you were on a device that didn't have a keyboard, could you claim compliance?
... the default interaction with a desktop is keyboard, so you have to meet keyboard access to comply, but on mobile, the primary mobile input is touch and gestures, so when we are looking at keyboard access and defining what keyboard access is, there has to be some way to input.
... what do we say for what the minimum, or what should be required on a mobile device?

JA: We could take a functional approach, and say that you have to have an input method that doesn't require speech, vision, etc.
... we could require a method that serves the most people

<KimPatch> Jeanne: keyboard is the most useful interface – not physical keyboard, but the concept of the keyboard was the most universal because which could use a keyboard, Bluetooth keyboards could be attached, individual tapping of a virtual keyboard is very common, so keyboard was the most universal of the input types because people can always emulate a keyboard. The danger with the functional...

<KimPatch> ...approach is that it can easily be set up to exclude people, especially people with multiple disabilities. I would really encourage us to keep looking at this and finding a way that we could approach this that would not leave out people – particularly people with multiple disabilities

KW: I think that the keyboard is not the most universal device on mobile

KP: I htink it depends on what you are doing. you could spend the whole day on keyboard on a mobile device.

kW: If you have everything you can access with a bluetooth keyboard, does that mean that everything can be accessed with switch and keyboard.

JA: i don't know how to support the iOS problem, where it is very accessible, but it doesn't support keyboard access to everything.
... there is a product that can take advantage of that interface.
... propose: defining that an alternative would be exposed and it would be programmatically supported to that interaction.
... the product can take advantage of the input exposed, and the application could take advantage of the action that the item can performed programmatically in some way.

<jon_avila> sounds good

KW: i think we should call this out in the notes and take it to the different working groups for more feedback from WcAGWG and UAWG
... I think it would be a sufficient technique, and could be a failure technique if it doesn't support keyboard

JA: In a programmatic way, if it didn't support keyboard emulation
... it goes back to Accessibility Supported. It's a broad term. It would also have to work with switch control or keyboard. It also goes back to the device, if it didn't have the device support, it isn't Accessibility Supported.

KW: the failure would be that it didn't have an input method that was accessibility supported.

kathy: that would be a failure of the device platform and not of the app

JA: If it worked on a platform, and failed on a different platform because the platform doesn't support it, then they could make an argument for accessibility supported.

Kathy: we shouldn't penalize the developer because of a failure of the platform

KW: you might define your platform as X and Y and not support Z because of a lack of accessibility support.
... it is so much more complex on mobile because of all the variations of platform

JS: The app must support the programmatic access to the platform
... We should take it to UAWG, because they have worked on this a lot and have a lot of language already worked out.

KW: We have it as a sufficient technique and as a failure technique

Survey Q2

2. Provide visual/audio indication for all functions

KW: on the desktop, you have focus indicator, but how do you alert people to the functional areas of the screen on mobile?/

ack {G

ack [GV

JA: Two things - I'm not sure everything has to have an audio -- if you have the programmatic correct, it could be announced by a screenreader. Are we aiming it so it is available to assistive technology? i don't think we need to have it read aloud.

KW: Let's considr this a duplicate of #28, that seems like better wording.

3. Provide instructions for custom functions and gestures

JA: Is advising the user different from providing instructions
... we have to be more clear about where it needs to be. Are we advising the user or providing instructions
... 3.3.2 is about instructions for input. This appears to be more about keyboard access -- maybe the keyboard trap technique may be helpful. It says if it requires more than one keystroke, the user must be advised of the method for moving away.
... we could use that language, this is not a keyboard trap
... if you have a custom gesture, then you have to advise the user.
... I'm not sure it should be connected to 2.2.1 keyboard input, maybe it maps to 3.3.2

kW: I'm not sure how it maps to WCAG, but I think i agree

JS: i'm not sure it is an accessibility issue, it seems more like a usability issue.

KW: Example: there is a gesture to expand an area by swiping up and down that people are using in apps. You need to perform the gesture to perform that action.

JA: And there is no instructions for how it would be executed with a keyboard.

jS: So it seems that the alternative input instruction is more of an issue than the custom gesture instruction.
... provide instructions for alternative input for custom gestures and functions

JA: I would prefer advise rather than instructions

<Kathy> Provide advisement on custom gesture and functions for alternative input

WCAG: The user is advised of the behavior before using the component

<Kathy> The user has been advised on custom gestures and functions for alternative input

The user is advised of the alternative input of custom gestures and functions

JA: Qualify it with "when alternative is not available with standard input method"

The user is advised of the alternative input of custom gestures and functions when alternative is not available with standard input methods.

kW: This would be a sufficient and also a failure.

next meeting

no meeting next week because of holiday

<Kathy> http://www.w3.org/WAI/GL/mobile-a11y-tf/wiki/Technique_Development_Assignments

Go through the techniques and identify which are applicable to mobile web, mobile apps or both.

- DRAFT -

Mobile Accessibility Task Force Teleconference

16 May 2014

Attendees

Contents

Survey Q1

Survey Q2

3. Provide instructions for custom functions and gestures

next meeting

Summary of Action Items

Scribe.perl diagnostic output