RE: Request from APA-WG re ARIA in HTML Spec

Okay, now I’m really confused.

Does this mean that the ARIA 1.1 specification is not actually meant to be used to define normative ARIA usage criteria, but that this is supposed to be defined by a different document instead?

Where is this defined?

From: Bryan Garaventa <bryan.garaventa@whatsock.com>
Sent: Thursday, March 15, 2018 7:16 PM
To: 'w3c/html-aria' <reply+001c7568bc22899eb274b9cc4b088ccd916d25f1be7fc70492cf0000000116c2b9fa92a16>
Subject: RE: [w3c/html-aria] Issues from ARIA Working Group review of the ARIA HTML spec document (#102)

“You do understand that it’s impossible to implement validator support for ARIA in HTML based simply on the ARIA spec, right?”

Is this question for me?

I’m not talking about HTML validation here, but ARIA usage validation. These two are not the same as you state.

From what I gather from this discussion, this ARIA HTML spec appears to be stating that the normative rules for correct ARIA usage, are not contained within the ARIA 1.1 specification, and that a different spec document is responsible for determining this instead. However this determination directly contradicts the ARIA 1.1 specification.



From: Michael[tm] Smith <notifications@github.com<mailto:notifications@github.com>>
Sent: Thursday, March 15, 2018 3:47 PM
To: w3c/html-aria <html-aria@noreply.github.com<mailto:html-aria@noreply.github.com>>
Cc: Bryan Garaventa <bryan.garaventa@whatsock.com<mailto:bryan.garaventa@whatsock.com>>; Mention <mention@noreply.github.com<mailto:mention@noreply.github.com>>
Subject: Re: [w3c/html-aria] Issues from ARIA Working Group review of the ARIA HTML spec document (#102)


As a simple example of this, according to the ARIA HTML spec document, no ARIA roles are allowed to be used on the HTML label element.

To my knowledge the ARIA spec makes no claim such as this

Right. That’s because by design, the ARIA spec makes no normative claims about what roles are allowed HTML elements, nor on elements in any other host language — instead it says that’s something which is left up to host language specifications to define.

You do understand that it’s impossible to implement validator support for ARIA in HTML based simply on the ARIA spec, right?

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub<https://github.com/w3c/html-aria/issues/102#issuecomment-373547134>, or mute the thread<https://github.com/notifications/unsubscribe-auth/ABx1aBt7cIc7jlYdE2amyleqIpMMTtC-ks5teu96gaJpZM4Srq39>.

Bryan Garaventa
Accessibility Fellow
Level Access, Inc.
Bryan.Garaventa@LevelAccess.com
415.624.2709 (o)
www.LevelAccess.com

From: Bryan Garaventa <bryan.garaventa@levelaccess.com>
Sent: Thursday, March 15, 2018 2:11 PM
To: Steve Faulkner <faulkner.steve@gmail.com>
Cc: Janina Sajka <janina@rednote.net>; W3C WAI ARIA <public-aria@w3.org>; Joanmarie Diggs <jdiggs@igalia.com>
Subject: RE: Request from APA-WG re ARIA in HTML Spec

I’m sorry, but this issue really is a huge one, and this means that the HTML Validator cannot be trusted.

From: Bryan Garaventa <bryan.garaventa@whatsock.com<mailto:bryan.garaventa@whatsock.com>>
Sent: Thursday, March 15, 2018 2:08 PM
To: 'w3c/html-aria' <reply+001c7568b35812b610f1587aabadf887a76ad9e6b8eb94d792cf0000000116c298a892a16>
Subject: RE: [w3c/html-aria] Issues from ARIA Working Group review of the ARIA HTML spec document (#102)

Yes, because nothing in the ARIA spec says that developers cannot do so. All of the roles I mentioned are simply those that I know of personally, but there are countless others not mentioned.

E.G It is perfectly valid to include role=checkbox on a label element that includes a hidden checkbox element so that the simulated checkbox control is focusable and interactive and toggles aria-checked while the result of this is reflected within the hidden checkbox for form submission.

There are so many examples of this with various markup structures that, if the HTML Validator is going to be flagging all such instances as being invalid, then the W3C HTML Validator cannot be trusted to provide accurate ARIA related information.


From: stevefaulkner <notifications@github.com<mailto:notifications@github.com>>
Sent: Thursday, March 15, 2018 1:25 PM
To: w3c/html-aria <html-aria@noreply.github.com<mailto:html-aria@noreply.github.com>>
Cc: Bryan Garaventa <bryan.garaventa@whatsock.com<mailto:bryan.garaventa@whatsock.com>>; Mention <mention@noreply.github.com<mailto:mention@noreply.github.com>>
Subject: Re: [w3c/html-aria] Issues from ARIA Working Group review of the ARIA HTML spec document (#102)

>So, if I understand correctly, are you saying that the HTML Validator will
flag any ARIA role present on a label element as being invalid?

yes, the aim of the conformance requirements is to deter authors from
misusing HTML elements that have strong semantics or behaviour. Can you
provide a use case for adding roles to the label element, that could not be
easily fulfilled using a span?

--

Regards

SteveF
Current Standards Work @W3C
<http://www.paciellogroup.com/blog/2015/03/current-standards-work-at-w3c/>

On 15 March 2018 at 19:30, Bryan Garaventa <notifications@github.com<mailto:notifications@github.com>> wrote:

> So, if I understand correctly, are you saying that the HTML Validator will
> flag any ARIA role present on a label element as being invalid?
>
> —
> You are receiving this because you commented.
> Reply to this email directly, view it on GitHub
> <https://github.com/w3c/html-aria/issues/102#issuecomment-373495949>, or mute
> the thread
> <https://github.com/notifications/unsubscribe-auth/AAzBE5ICQqMyiqku_WVd4eWDXC-rHkSHks5tesFAgaJpZM4Srq39>
> .
>

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub<https://github.com/w3c/html-aria/issues/102#issuecomment-373511269>, or mute the thread<https://github.com/notifications/unsubscribe-auth/ABx1aEzeTkBgIE1ymvk1pBGRcxs0W0tmks5tes4ogaJpZM4Srq39>.



Bryan Garaventa
Accessibility Fellow
Level Access, Inc.
Bryan.Garaventa@LevelAccess.com<mailto:Bryan.Garaventa@LevelAccess.com>
415.624.2709 (o)
www.LevelAccess.com<http://www.LevelAccess.com>

From: Steve Faulkner <faulkner.steve@gmail.com<mailto:faulkner.steve@gmail.com>>
Sent: Thursday, March 15, 2018 11:41 AM
To: Bryan Garaventa <bryan.garaventa@levelaccess.com<mailto:bryan.garaventa@levelaccess.com>>
Cc: Janina Sajka <janina@rednote.net<mailto:janina@rednote.net>>; W3C WAI ARIA <public-aria@w3.org<mailto:public-aria@w3.org>>; Joanmarie Diggs <jdiggs@igalia.com<mailto:jdiggs@igalia.com>>
Subject: Re: Request from APA-WG re ARIA in HTML Spec

Hi Bryan,
each of your suggestions will be reviewed and a decision will be made based on use cases and data provided.

As I stated in the issue you filed<https://github.com/w3c/html-aria/issues/102#issuecomment-373480208>
you wrote:
To my knowledge the ARIA spec makes no claim such as this, so it is unclear where these rules are being deduced from.

The ARIA spec naturally does not make any claims about what is allowed usage in host languages, what it does say is

Host languages MAY document features that cannot be overridden with WAI-ARIA
https://www.w3.org/TR/wai-aria-1.1/#host_general_conflict


Which is what ARIA in HTML does for the HTML specification. The rules have been developed over the past 8 years.

--

Regards

SteveF
Current Standards Work @W3C<http://www.paciellogroup.com/blog/2015/03/current-standards-work-at-w3c/>

On 15 March 2018 at 18:23, Bryan Garaventa <bryan.garaventa@levelaccess.com<mailto:bryan.garaventa@levelaccess.com>> wrote:
<reply+001c7568acfc056d811d78e8ce4acfb1444fde85c25ce0b892cf0000000116c1e51b92a16<mailto:reply%2B001c7568acfc056d811d78e8ce4acfb1444fde85c25ce0b892cf0000000116c1e51b92a>>
Subject: RE: [w3c/html-aria] Issues from ARIA Working Group review of the ARIA HTML spec document (#102)

Thanks, I understand, however I don't have time to write a separate markup case for every instance of all of these roles. Speaking personally, our clients at Level Access have had cases for all of these combinations, and all were valid usages of ARIA.

I wasn't aware until last week that this document is being used as the guide for the W3C HTML Validator, and that missing roles will be flagged as being invalid usages of ARIA. If this is a misunderstanding on my part, please let me know since this is the primary reason for my objection.

As a simple example of this, according to the ARIA HTML spec document, no ARIA roles are allowed to be used on the HTML label element.

To my knowledge the ARIA spec makes no claim such as this, so it is unclear where these rules are being deduced from.

However, if there is no risk that valid usages of ARIA will be flagged as violations by the HTML Validator, then I have no objections to this.



Bryan Garaventa
Accessibility Fellow
Level Access, Inc.
Bryan.Garaventa@LevelAccess.com<mailto:Bryan.Garaventa@LevelAccess.com>
415.624.2709 (o)
www.LevelAccess.com<http://www.LevelAccess.com>

From: Steve Faulkner <faulkner.steve@gmail.com<mailto:faulkner.steve@gmail.com>>
Sent: Thursday, March 15, 2018 1:21 AM
To: Bryan Garaventa <bryan.garaventa@levelaccess.com<mailto:bryan.garaventa@levelaccess.com>>
Cc: Janina Sajka <janina@rednote.net<mailto:janina@rednote.net>>; W3C WAI ARIA <public-aria@w3.org<mailto:public-aria@w3.org>>; Joanmarie Diggs <jdiggs@igalia.com<mailto:jdiggs@igalia.com>>
Subject: Re: Request from APA-WG re ARIA in HTML Spec

>I don't recommend entering CR until these issues are addressed, because there are currently many roles that are missing from the >accepted role usage column, and if they are omitted then HTML validators will start flagging valid constructs with false reports as >being invalid usages of ARIA.

as I stated on the bug Bryan filed<https://github.com/w3c/html-aria/issues/102#issuecomment-373287694>

  *   The default implicit column mappings are defined in the HTML AAM<https://w3c.github.io/html-aam/>, if you have issue with these you will need to file issues there.
  *   it is not simply enough to state that "the accepted roles/states/properties column is missing". As a host language HTML can and does limit the allowed roles on elements with meaningful default semantics. If it is requested that a particular role be allowed, then a case needs to be made for why.
Furthermore,
and if they are omitted then HTML validators will start flagging valid constructs with false reports as being invalid usages of ARIA.

What is in the current ARIA in HTML spec is what is implemented in the HTML validator. The rules have been implemented in the HTML validator in some form or other since 2010 as they were defined in HTML5. <https://www.w3.org/TR/2014/REC-html5-20141028/dom.html#sec-strong-native-semantics>
If you review those rules it will be evident that in most cases they are less stringent now than they were when initially defined. The relaxation of the rules has been, in general on a case by case basis, based on use cases.

--

Regards

SteveF
Current Standards Work @W3C<http://www.paciellogroup.com/blog/2015/03/current-standards-work-at-w3c/>

On 15 March 2018 at 07:35, Bryan Garaventa <bryan.garaventa@levelaccess.com<mailto:bryan.garaventa@levelaccess.com>> wrote:
Hi,
I've completed my initial review of the spec document as I was asked to do last week, and have filed the issues I discovered here
https://github.com/w3c/html-aria/issues/102


I don't recommend entering CR until these issues are addressed, because there are currently many roles that are missing from the accepted role usage column, and if they are omitted then HTML validators will start flagging valid constructs with false reports as being invalid usages of ARIA.

The changes are relatively simple though by adding the missing roles, so an update likely wouldn't take long.

There are some areas where the mappings seem unclear to me, or where I've added suggestions regarding default role mappings. Areas beginning with FYI are simply informative in case the info is helpful.

All of the other content appears clear to me though, and I found no issues while scanning the remainder of the document.

Bryan Garaventa
Accessibility Fellow
Level Access, Inc.
Bryan.Garaventa@LevelAccess.com<mailto:Bryan.Garaventa@LevelAccess.com>
415.624.2709 (o)
www.LevelAccess.com<http://www.LevelAccess.com>

-----Original Message-----
From: Janina Sajka <janina@rednote.net<mailto:janina@rednote.net>>
Sent: Thursday, March 08, 2018 6:38 AM
To: W3C WAI ARIA <public-aria@w3.org<mailto:public-aria@w3.org>>; Joanmarie Diggs <jdiggs@igalia.com<mailto:jdiggs@igalia.com>>
Subject: Request from APA-WG re ARIA in HTML Spec
Hello ARIA:

Wearing my hat as APA Chair, I'd like to ask you to consider whether an ARIA review of the Web Platforms WG specification ARIA in HTML is needed. The Web Platforms WG has requested to take this specification to Candidate Recommendation (CR):

http://w3c.github.io/html-aria/publish/index-CR-20180215.html


There's considerable history behind this document which I won't attempt to summarize in this email. However, it would be very helpful for the ARIA-WG to take at least a quick look to determine whether there are any problems in this specification that should be cleared up before the document moves further.

Janina
--

Janina Sajka

Linux Foundation Fellow
Executive Chair, Accessibility Workgroup:       http://a11y.org


The World Wide Web Consortium (W3C), Web Accessibility Initiative (WAI)
Chair, Accessible Platform Architectures        http://www.w3.org/wai/apa

Received on Friday, 16 March 2018 02:51:56 UTC