search

w3c / wcag

Web Content Accessibility Guidelines
https://w3c.github.io/wcag/guidelines/22/
Other
1.09k stars 242 forks source link

Revisiting Transcripts and Audio Description (SC 1.2.3/5/8) #782

Open bruce-usab opened 5 years ago

bruce-usab commented 5 years ago

The purpose of this issue is to determine whether or not WCAG 2.1, success criterion 1.2.3 Audio Description or Media Alternative (Prerecorded) is still relevant to the accessibility of web content for people with disabilities.

The initial proposal was to delete/remove SC 1.2.3 from WCAG 2.2. There may be better alternatives, so really this issue is about revisiting 1.2.3, 1.2.5, and 1.2.8.

In my experience, having one (and only one) instance of a Double A SC wholly replacing a Single A SC has proven to be a surprisingly difficult stumbling block with teaching about WCAG.

In theory, the concept is easy: At Double A, SC 1.2.3 is superseded by SC 1.2.5. In actual practice, the instruction on this detail is not worth the cognitive load required to convey this particular nuance.

Straw Poll:

In 2008, when WCAG 2.0 was published, the working assumption was that some regulatory bodies would adopt the Single A level. I submit that this (1) has proven to not be the case; and (2) therefore, having an SC that is irrelevant at Double A is harmful and counter-productive.

The simplest course of action would be for 2.2 to not include SC 1.2.3.

We might also consider promoting Success Criterion 1.2.8 Media Alternative (Prerecorded) to Single A.

This top card of this issue will be updated as discussion on the issue progresses.

mbgower commented 5 years ago

I think there may be situations where one cannot meet the synchronized requirement, for whatever reason. Being able to offer a transcript is better than nothing. I think someone being able to indicate that the transcript is provided is a definsible way to gracefully fail the AA requirement while still providing some facilitation.

johnfoliot commented 5 years ago

I'll also note that David remarked on the call today (06/11) that in some territories they have an explicit, policy-driven "exemption" around 1.2.5 (even if all other things considered the requirement is for WCAG AA compliance).

While Section 508 has effectively made SC 1.2.3 redundant for US Federal sites (et. al.), that may not be the case everywhere. For that reason I would oppose removal of this SC.

JF

On Tue, Jun 11, 2019 at 9:54 AM bruce-usab notifications@github.com wrote:

This top card of this issue will be updated as discussion on the issue progresses.

The purpose of this issue is to determine whether or not WCAG 2.1, success criterion 1.2.3 Audio Description or Media Alternative (Prerecorded) http://www.w3.org/TR/WCAG21/#audio-description-or-media-alternative-prerecorded is still relevant to the accessibility of web content for people with disabilities.

The proposal is delete/remove SC 1.2.3 from WCAG 2.2.

Having one (and only one) instance of a Double A SC wholly replacing a Single A SC has proven to be a surprisingly difficult stumbling block with teaching about WCAG.

In theory, the concept is easy: At Double A, SC 1.2.3 is superseded by SC 1.2.5. In actual practice, the instruction on this detail is not worth the cognitive load required to convey this particular nuance.

In 2008, when WCAG 2.0 was published, the working assumption was that some regulatory bodies would adopt the Single A level. This has not been the case. Therefore, having an SC that is irrelevant at Double A is harmful and counter-productive.

The simplest course of action would be for 2.2 to not include SC 1.2.3.

We might also consider promoting Success Criterion 1.2.8 Media Alternative (Prerecorded) http://www.w3.org/TR/WCAG21/#media-alternative-prerecorded to Single A.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/w3c/wcag/issues/782?email_source=notifications&email_token=AAJL445SX2TIZ6PAX5TNNYTPZ64DDA5CNFSM4HW7GZKKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4GY2Y5MA, or mute the thread https://github.com/notifications/unsubscribe-auth/AAJL44Y6CI3KRCW7NTDFECDPZ64DDANCNFSM4HW7GZKA .

-- ​John Foliot | Principal Accessibility Strategist | W3C AC Representative Deque Systems - Accessibility for Good deque.com

mraccess77 commented 5 years ago

I agree with the concerns that in regions 1.2.5 is exempted and that some videos don't have sufficient pauses for AD. Moving 1.2.8 to Level A should solve that issue and also ensure that there is something available for users who are deafblind. Right now transcripts are not required at Level AA and getting transcripts at Level AA seems reasonable in today's environment.

bruce-usab commented 5 years ago

@mbgower wrote:

I think there may be situations where one cannot meet the synchronized requirement

Can you think of some examples?

I think someone being able to indicate that the transcript is provided is a defensible way to gracefully fail the AA requirement while still providing some facilitation.

This would be passing 1.2.3 but failing 1.2.5. But is this a good argument for keeping 1.2.5 status quo?

@mraccess77 wrote:

...some videos don't have sufficient pauses for AD.

That use case would not actually fail against 1.2.5 (since, by definition, AD narration is only dubbed over the default sound track). Also, it is pretty easy to find examples of decent AD for pretty intense theatrical uses (e.g., action movies), so I am always skeptical of assertions that an A/V production does not lend itself to AD.

There are times when it is hard to do a great job with AD, and a transcript might be better for those use cases. Also, many people prefer a transcript over AD. And, of course, some users do not benefit at all from AD, but would benefit from a transcript. But I think this all argues for promoting 1.2.8 to Single A.

Moving 1.2.8 to Level A should solve that issue and also ensure that there is something available for users who are deafblind. Right now transcripts are not required at Level AA and getting transcripts at Level AA seems reasonable in today's environment.

I agree, and would also like for us to have a conversation about that!

mraccess77 commented 5 years ago

I have seen many training videos where there is not sufficient pauses to cover what is shown in a screenshare of a system. If there isn't enough room for pauses then you can't fit in the AD and as most agree it's not a failure of 1.2.5 -- but the person who is blind doesn't get the missed information at all and the content passed WCAG A AA. If we had a transcript requirement then at least that information would be available somehow. Agree on your point that there are some types of videos like this were the video is walking through entering code where the transcript would be the preferred method of access by people who are blind. But for other situations like movies the transcript is a very separate alternative that excludes people from the experience. So it really depends on the genre of the video content.

bruce-usab commented 5 years ago

@mraccess77 wrote:

I have seen many training videos where there is not sufficient pauses to cover what is shown in a screenshare of a system.

That is a great and common example actually! I will even go further, to generalize to like 99% of PowerPoint style presentations where the speaker cannot be bothered to read their slides. So the presenter is talking the whole time, but someone who is listening but not watching is lost.

If a bad presenter like that is being webcast live, they do not fail WCAG at any level! We do not have an AD requirement for live web casting. Maybe we can take a look a that also for 2.2?

bruce-usab commented 5 years ago

@johnfoliot wrote:

I'll also note that David remarked on the call today (06/11) that in some territories they have an explicit, policy-driven “exemption” around 1.2.5 (even if all other things considered the requirement is for WCAG AA compliance).

@DavidMacDonald can you provide the cite for that? I would like to see exactly how they phrase that exception.

While Section 508 has effectively made SC 1.2.3 redundant for US Federal sites (et. al.), that may not be the case everywhere.

But it is not just 508. EN 301 549 has the same result, as does every legal settlement that I know of.

Does anyone have examples of WCAG2 being required at the Single A and not the AA Level?

Aside from the Canadian example with exempting 1.2.5, are there other examples where WCAG 2.0 Level AA has been cherry picked?

mraccess77 commented 5 years ago

Some electronic health records requirements called out WCAG A https://www.levelaccess.com/electronic-health-records-accessibility/ but these might have been overcome by ACAA.

bruce-usab commented 5 years ago

Some electronic health records...

Good catch! But of course, it would be pretty unusual to have multimedia in EHR.

Here is a shortish URL for 45 CFR Part 170: www.ecfr.gov/cgi-bin/text-idx?node=pt45.1.170

To find where WCAG 2.0 Level A is applicable to EHR, search that page for “170.204(a)(1)”. (Just two instances, but they are quite significant.)

The single cite to WCAG 2.0 Level AA is written as “170.204(a)(2)” and is not actually all that helpful. Here is that excerpt:

Such access must be consistent and in accordance with the standard adopted in §170.204(a)(1) and may alternatively be demonstrated in accordance with the standard specified in §170.204(a)(2).

So one can meet the Level A requirement by conforming to Level AA. Duh!

bruce-usab commented 5 years ago

I have added a straw poll to my original post, and would ask folks to use that to indicate if you find 1.2.3 particularly annoying to teach about (or not).

Please do not use the straw poll to vote if we should deprecate 1.2.3 (or not). That sort of action should wait for a survey, and after more conversation about the issue.

A simple explanation for my experience is that I am poor teacher and/or have dull students!

awkawk commented 5 years ago

@bruce-usab OP?

bruce-usab commented 5 years ago

OP == original post / sorry

johnfoliot commented 5 years ago

@Bruce:

Audio Description (prerecorded video) except where the video provides information related to health and safety of Canadians WCAG 2.0 Success Criterion 1.2.5 Audio Description

Exclusions for a 10 business day grace period:

Live Video Captions exclusions refer to closed captioning and transcription of live video

Audio Description (prerecorded Videos) are only required for videos that relate to health and safety of Canadians; for remaining prerecorded videos, audio description is not required.

(source: https://www.tbs-sct.gc.ca/pol/doc-eng.aspx?id=23601 - Expand the Section "Appendix B: List of Exclusions: Published August 1, 2011 - highlight mine - the exclusion/exemption of SC 1.2.5 here means that it falls back to SC 1.2.3.)

I fail to see why this is an issue: if you've met SC 1.2.5, then by default you've also met SC 1.2.3. Your presumption that every site out there is striving for AA full compliance is not based on real evidence, but rather (I will suggest) your vantage point from the US Federal Government perspective.

Additionally, while I support moving SC 1.2.8 Media Alternative (Prerecorded) (a.k.a. Provide a Transcript) 'higher' in severity, jumping from AAA to A is quite radical. We could move it to a AA SC, in alignment with other video-related SC (1.2.5 Audio Descriptions and 1.2.4 Captions (Live) ) - remembering that the prioritization of A, AA, and AAA was based not only on impact to the user, but also effort on the part of the content creator.

Bottom line here Bruce - I will strenuously oppose the deprecation of SC 1.2.3 in the context of any WCAG 2.x work, unless you can categorically prove that it is not relevant in the context of internationalization support.

JF

On Wed, Jun 12, 2019 at 10:33 AM Andrew Kirkpatrick < notifications@github.com> wrote:

@bruce-usab https://github.com/bruce-usab OP?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/w3c/wcag/issues/782?email_source=notifications&email_token=AAJL44467K6DCBCMDF2DHOTP2EJK7A5CNFSM4HW7GZKKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODXQ2R5I#issuecomment-501328117, or mute the thread https://github.com/notifications/unsubscribe-auth/AAJL4457N7LOMNY543OZFQLP2EJK7ANCNFSM4HW7GZKA .

-- ​John Foliot | Principal Accessibility Strategist | W3C AC Representative Deque Systems - Accessibility for Good deque.com

lauracarlson commented 5 years ago

Hi Bruce and all,

I voted thumbs up as my experience teaching 1.2.3 is similar to yours. At first people struggle with the concept of SC 1.2.5 superseding SC 1.2.3 at AA. Once it is explain then they understand. Maybe that could be explained better in the understanding docs? People I work with have read the understanding docs, and still didn't get it until I explained it.

But I agree with Mike, providing a transcript is a way to fail the AA requirement while still providing some accessibility. If we have evidence that 1.2.3 is used internationally (Canada etc), we shouldn't deprecate it.

jake-abma commented 5 years ago

just a question / thought: why go through all the possible trouble and explaining of deleting a SC while working on Silver? Is it worth the effort or is the expectation of Silver that is will take so much time that the pros win from the cons?

johnfoliot commented 5 years ago

Laura writes:

"At first people struggle with the concept of SC 1.2.3 superseding SC 1.2.5 at AA"

Respectfully Laura, that's upside down: SC 1.2.5 supersedes 1.2.3, and not vice versa. It is similar to how Success Criterion 1.4.6 Contrast (Enhanced) at AAA (aka 7:1 color contrast) supersedes Success Criterion 1.4.3 Contrast (Minimum) at AA (aka 4.5:1 color contrast) - where achieving the AAA is harder than achieving the AA

How I explain this: At the single A level (1.2.3), content creators have a choice: provide audio descriptions OR a transcript, so in the context of what is less onerous on the content creator, providing a transcript (all things being equal) is easier to produce than hiring voice talent and seeking gaps in the existing audio-stream to inject those descriptions. THAT is why it is a single A requirement: it suggests that at a minimum, while not truly meeting the needs of non-sighted users (at least not elegantly), the transcript has some utility there none-the-less, albeit less useful, especially in real-time.

The additional effort of specifically providing AD was deemed "harder" on the content creator, and thus the AA categorization. (That reasoning, for example, is also why the Government of Canada has removed that requirement

Today the issue is that most sites are either mandated to, or strive to, meet AA conformance, which is why it seems that 1.2.3 is redundant. However I assert that this assumption is NOT universal, and thus suggesting that SC 1.2.3 is redundant for ALL content creators is an unproven assertion which I reject.

In the context of teaching this (something I am very familiar with, as it is part of my role at Deque), I've actually found this particular topic a good place to inject the "story" about how the SCs arrived at their A, AA, or AAA levels, and to remind folks that it was based not just on the impact of the end user - that WCAG 2.0 also considered the impact on authors as well. (I also surface the contrast between SC 1.4.3 and 1.4.6 to illustrate the pattern elsewhere in the document.)

When I wrap the "story" in that context, I've not had any confusion from my students - they seem to get it.

JF

On Wed, Jun 12, 2019 at 11:45 AM Jake Abma notifications@github.com wrote:

just a question / thought: why go through all the possible trouble and explaining of deleting a SC while working on Silver? Is it worth the effort or is the expectation of Silver that is will take so much time that the pros win from the cons?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/w3c/wcag/issues/782?email_source=notifications&email_token=AAJL446Z4QEV4TYYLHHHLQDP2ER4NA5CNFSM4HW7GZKKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODXRCF2I#issuecomment-501359337, or mute the thread https://github.com/notifications/unsubscribe-auth/AAJL446Y2ZNXJMUMK6YYIZDP2ER4NANCNFSM4HW7GZKA .

-- ​John Foliot | Principal Accessibility Strategist | W3C AC Representative Deque Systems - Accessibility for Good deque.com

lauracarlson commented 5 years ago

Yes. John, typo. Thanks.

On 6/12/19, John Foliot notifications@github.com wrote:

Laura writes:

"At first people struggle with the concept of SC 1.2.3 superseding SC 1.2.5 at AA"

Respectfully Laura, that's upside down: SC 1.2.5 supersedes 1.2.3, and not vice versa.

Thanks.

Kind regards, Laura

-- Laura L. Carlson

johnfoliot commented 5 years ago

@Jake - I sort of agree: this seems like a better discussion inside of the Silver work, or at least in that context. I'm quite fearful of tinkering with anything already published as a WCAG 2.x SC - leave sleeping dogs lay and we can address this topic more holistically in the context of Silver.

JF

On Wed, Jun 12, 2019 at 11:45 AM Jake Abma notifications@github.com wrote:

just a question / thought: why go through all the possible trouble and explaining of deleting a SC while working on Silver? Is it worth the effort or is the expectation of Silver that is will take so much time that the pros win from the cons?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/w3c/wcag/issues/782?email_source=notifications&email_token=AAJL446Z4QEV4TYYLHHHLQDP2ER4NA5CNFSM4HW7GZKKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODXRCF2I#issuecomment-501359337, or mute the thread https://github.com/notifications/unsubscribe-auth/AAJL446Y2ZNXJMUMK6YYIZDP2ER4NANCNFSM4HW7GZKA .

-- ​John Foliot | Principal Accessibility Strategist | W3C AC Representative Deque Systems - Accessibility for Good deque.com

bruce-usab commented 5 years ago

@johnfoliot wrote:

source: https://www.tbs-sct.gc.ca/pol/doc-eng.aspx?id=23601 Expand the Section “Appendix B: List of Exclusions: Published August 1, 2011”

Thanks!

...the exclusion/exemption of SC 1.2.5 here means that it falls back to SC 1.2.3.

I agree. So if WCAG 2.0 did not have SC 1.2.3, with their language as-is, TBS would not have had a requirement for either AD or transcripts. This observation does temper my enthusiasm for just deleting SC 1.2.3. On the other hand, making 1.2.3 a requirement for transcripts is another possibility...

(FWIW, I actually like how they write up their “list of exclusions” in that it is narrative instead of something terse like “1.2.5 is excepted”. Also, they are citing to 2.0 specifically, so the W3C putting out 2.1 (or 2.2) does make requirements retroactive.)

I fail to see why this is an issue: if you've met SC 1.2.5, then by default you've also met SC 1.2.3.

That is something I am trying to a get sense for. It is an issue because this single instance one SC superseding another SC has been an (unexpected) impediment to my own teaching about WCAG. But my experience might not be representative of others’ experiences. At Level AAA there are a few SC that make Level A and AA irrelevant, but having one-and-only-one example at Level AA is confusing and not necessary.

Your presumption that every site out there is striving for AA full compliance is not based on real evidence, but rather (I will suggest) your vantage point from the US Federal Government perspective.

Yes, I am prejudiced by my vantage point. The WCAG conformance model is based on wholly meeting levels.

Additionally, while I support moving SC 1.2.8 Media Alternative (Prerecorded) (a.k.a. Provide a Transcript) 'higher' in severity, jumping from AAA to A is quite radical. We could move it to a AA SC, in alignment with other video-related SC (1.2.5 Audio Descriptions and 1.2.4 Captions (Live) )

I agree with promoting SC 1.2.8 for 2.2. That would also be a way to resolve this issue.

bruce-usab commented 5 years ago

@lauracarlson wrote:

Maybe that could be explained better in the understanding docs? People I work with have read the understanding docs, and still didn’t get it until I explained it.

I submit that the Understanding is as good as it could be on this particular point (and if it were written perfectly, it would not matter). I think it is a serious defect with WCAG 2.0. I think the problem comes from WCAG being read (and implemented) by people who mostly ignore the AAA SC, so the single instance of one SC superseding another is weird and strange.

If we have evidence that 1.2.3 is used internationally (Canada etc), we shouldn’t deprecate it.

How do you feel about making transcripts a requirement at AA or A?

lauracarlson commented 5 years ago

@bruce-usab wrote:

How do you feel about making transcripts a requirement at AA or A?

+1

Laura

johnfoliot commented 5 years ago

Hi Bruce,

That is something I am trying to a get sense for. It is an issue because this single instance one SC superseding another SC has been an impediment to my teaching about WCAG. But my experience might not be representative of others experiences. At Level AAA there are a few SC that make Level A and AA irrelevant, but having one-and-only-one example at Level AA is confusing and not necessary.

Actually, I truly think about this in terms of approach. At Single A, the content author has a choice to meet SC 1.2.3: provided Audio Description OR provide a transcript. But as we get more "demanding" (at the AA level), the Standard removes that choice and demands Audio Description.

I think part of the issue (and I further believe that Silver will solve much of this issue), is that for the majority of content creators in the US Federal sphere (where Section 508 = WCAG A & AA) there is little difference to the none-aware about the difference between A conformance and AA conformance, as for them in particular, the distinction has been flattened down to "here are the 38 Success Criteria you must meet". However, I have found that when I explain it as "at single A, you have a choice, at double A you don't" seems to resolve any questions from the classroom.

HTH

JF

On Fri, Jun 14, 2019 at 9:16 AM bruce-usab notifications@github.com wrote:

@johnfoliot https://github.com/johnfoliot wrote:

source: https://www.tbs-sct.gc.ca/pol/doc-eng.aspx?id=23601 Expand the Section “Appendix B: List of Exclusions: Published August 1, 2011”

Thanks!

...the exclusion/exemption of SC 1.2.5 here means that it falls back to SC 1.2.3.

I agree. So if WCAG 2.0 did not have SC 1.2.3, with their language as-is, TBS would not have had a requirement for either AD or transcripts. This observation does temper my enthusiasm for just deleting SC 1.2.3. On the other hand, making 1.2.3 a requirement for transcripts is another possibility...

(FWIW, I actually like how they write up their “list of exclusions” in that it is narrative instead of something terse like “1.2.5 is excepted”. Also, they are citing to 2.0 specifically, so the W3C putting out 2.1 (or 2.2) does make requirements retroactive.)

I fail to see why this is an issue: if you've met SC 1.2.5, then by default you've also met SC 1.2.3.

That is something I am trying to a get sense for. It is an issue because this single instance one SC superseding another SC has been an impediment to my teaching about WCAG. But my experience might not be representative of others experiences. At Level AAA there are a few SC that make Level A and AA irrelevant, but having one-and-only-one example at Level AA is confusing and not necessary.

Your presumption that every site out there is striving for AA full compliance is not based on real evidence, but rather (I will suggest) your vantage point from the US Federal Government perspective.

Yes, I am prejudiced by my vantage point. The WCAG conformance model is based on wholly meeting levels.

Additionally, while I support moving SC 1.2.8 Media Alternative (Prerecorded) (a.k.a. Provide a Transcript) 'higher' in severity, jumping from AAA to A is quite radical. We could move it to a AA SC, in alignment with other video-related SC (1.2.5 Audio Descriptions and 1.2.4 Captions (Live) )

I agree with promoting SC 1.2.8 for 2.2. That would also be a way to resolve this issue.

I will strenuously oppose the deprecation of SC 1.2.3 in the context of any WCAG 2.x work, unless you can categorically prove that it is not relevant in the context of internationalization support.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/w3c/wcag/issues/782?email_source=notifications&email_token=AAJL44YLQSAV2PNJUR5AVFTP2OR4JA5CNFSM4HW7GZKKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODXW5NOY#issuecomment-502126267, or mute the thread https://github.com/notifications/unsubscribe-auth/AAJL444A2IG7EGM6KPWL4PLP2OR4JANCNFSM4HW7GZKA .

-- ​John Foliot | Principal Accessibility Strategist | W3C AC Representative Deque Systems - Accessibility for Good deque.com

bruce-usab commented 5 years ago

@jake-abma wrote:

just a question / thought: why go through all the possible trouble and explaining of deleting a SC while working on Silver?

The answer to this question is mostly the same answer as to, “Why work on 2.2 at all?” But IMHO fixing old problems with 2.0 is just as important as adding new SC to 2.2.

Is it worth the effort

Yes, full stop.

...or is the expectation of Silver that is will take so much time that the pros win from the cons?

Yes, there is that too. Also, it seems possible (maybe even probable) that Silver 1.0 won’t have as many requirements as WCAG 2.2. So even after Silver 1.0 is out, 2.2 might be more attractive to certain regulatory agencies.

bruce-usab commented 5 years ago

@johnfoliot wrote:

@jake - I sort of agree: this seems like a better discussion inside of the Silver work, or at least in that context. I'm quite fearful of tinkering with anything already published as a WCAG 2.x SC - leave sleeping dogs lay and we can address this topic more holistically in the context of Silver.

I don’t disagree that revisiting 1.2.3/1.2.5/1.2.8 makes more sense in context of the Silver work. That does not negate that revisiting 1.2.3/1.2.5/1.2.8 makes some sense in the context of work on 2.2.

Those of us wanting to fix a few broken things about 2.0 gave up that fight so that 2.1 could meet its publication schedule. I submit that the interaction between 1.2.3/1.2.5/1.2.8 is at least as broken as 4.1.1 and the relative luminance formula, and that the change needed is pretty lightweight. If we are making corrective changes to 2.0 anyway, addressing this issue is better than choosing to ignore it.

johnfoliot commented 5 years ago

Hi Bruce,

Again, I will strenuously oppose making any changes to existing SC in the 2.x family, as it breaks the backward compatibility requirement, and again, not every territory or site is mandated to meet WCAG A & AA compliance. (This would include making 1.2.3 simply mandate a transcript - that too breaks the backward compatibility paradigm). There is ample evidence of sites meeting this SC long before Section 508 mandated WCAG 2.0

I can support updating the Understanding documents if the current documents remain unclear (Hello EO WG?), but I will oppose any material change to SC 1.2.3 simply on the basis that it's hard to understand or teach when you are mandated to meet both A and AA Success Criteria (as is the current case for Section 508). There is enough evidence of sites meeting this SC long before Section 508 mandated WCAG 2.0 that it disproves your theory that it needs fixing.

You initially wrote:

Having one (and only one) instance of a Double A SC wholly replacing a Single A SC has proven to be a surprisingly difficult stumbling block with teaching about WCAG.

This is a teaching problem then, and not a problem with the actual SC.

...the working assumption was that some regulatory bodies would adopt the Single A level. This has not been the case.

Not every territory has 'regulatory' pressures to contend with (see: https://www.w3.org/WAI/policies/).

One such territory is Japan, where our good friend and colleague @Makoto Ueki has spent the past decade-plus advancing web accessibility in his home country. I don't want to put words in his mouth, but, given that in Japan, "web accessibility" is "recommended" but not mandated, I'm sure that Makoto takes his wins where he finds them, and getting a site to even full Single A compliance would be a win in his camp.

If you can provide evidence of materiel harm because of the current wording or intent of SC 1.2.3, then I would then re-consider re-opening this now 11-year-old Success Criteria, but personally that would be the very high bar I would set. It's not the SC that needs fixing, it's (apparently) the educational material behind it, including understanding why certain SC are given an A, AA or AAA ranking, and what that means in an International context.

Respectfully,

JF

On Fri, Jun 14, 2019 at 9:55 AM bruce-usab notifications@github.com wrote:

@johnfoliot https://github.com/johnfoliot wrote:

@jake https://github.com/jake - I sort of agree: this seems like a better discussion inside of the Silver work, or at least in that context. I'm quite fearful of tinkering with anything already published as a WCAG 2.x SC - leave sleeping dogs lay and we can address this topic more holistically in the context of Silver.

I don’t disagree that revisiting 1.2.3/1.2.5/1.2.8 makes more sense in context of the Silver work. That does not imply, however, that revisiting 1.2.3/1.2.5/1.2.8 makes some sense in the context of work on 2.2.

Those of us wanting to fix a few broken things about 2.0 gave up that fight so that 2.1 could meet its publication schedule. I submit that the interaction between 1.2.3/1.2.5/1.2.8 is at least as broken as 4.1.1 and the relative luminance formula, and that the change needed is pretty lightweight. If we are making corrective changes to 2.0 anyway, addressing this issue is better than choosing to do nothing.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/w3c/wcag/issues/782?email_source=notifications&email_token=AAJL447HZSH53YTNCHZQLA3P2OWPHA5CNFSM4HW7GZKKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODXXBBNA#issuecomment-502141108, or mute the thread https://github.com/notifications/unsubscribe-auth/AAJL44ZALXTJTUTOILHIFVDP2OWPHANCNFSM4HW7GZKA .

-- ​John Foliot | Principal Accessibility Strategist | W3C AC Representative Deque Systems - Accessibility for Good deque.com

bruce-usab commented 5 years ago

@johnfoliot wrote:

I will strenuously oppose making any changes to existing SC in the 2.x family, as it breaks the backward compatibility requirement ... This would include making 1.2.3 simply mandate a transcript - that too breaks the backward compatibility paradigm.

I am not following how other changes we are considering for 2.2 do not “break backwards compatibility” but effectively promoting 1.2.8 to Single A would somehow break backwards compatibility.

This is a teaching problem then, and not a problem with the actual SC.

It is both a teaching problem and a structural problem with the actual SC.

That said, I like your approach to teaching. OTOH, maybe you mostly teach young fresh developers who are new to accessibility? I mostly teach old tired admin/policy HR folks who were anointed 508 responsibilities under “other duties as assigned” and they have been struggling to make sense of WCAG on their own for many months before I get a chance to help them.

If you can provide evidence of materiel harm because of the current wording or intent of SC 1.2.3

I regard my first hand experience as evidence of material harm. This is not theoretical for me!

It's not the SC that needs fixing, it's (apparently) the educational material behind it, including understanding why certain SC are given an A, AA or AAA ranking

Rather obviously, I respectfully disagree. But again, as part of this discussion, we should have a survey to figure out if the defect with 1.2.3/1.2.5/1.2.8 warrants being in or out for 2.2. You have made your opinion on the matter clear. Got it! Please thumbs down the OP / top card.

johnfoliot commented 5 years ago

@bruce-usab I think that part of the problem here is that you are proposing multiple actions in one thread. For clarity, I personally am at the following:

ACTION 1:

ACTION 2

I'd also add a third Action:

ACTION 3

Thoughts?

patrickhlauke commented 5 years ago

Coming in late, but for what it's worth, I'd suggest that for content authors, it's orders of magnitude more difficult to provide an alternative version of a video with AD (in terms of production, particularly if it isn't produced in-house) than it is to provide a transcript (and leaving aside the fact as well that it's often not straightforward to just provide a single video with multiple audio tracks that can be selected, and it usually ends up requiring a completely separate video file).

(as a separate topic, i'm also reminded of the weird imbalance in requiring captions for live audio_video at AA, but for live audio only at AAA, but i'll repost this as a separate issue)

bruce-usab commented 5 years ago

@johnfoliot:

I think that part of the problem here is that you are proposing multiple actions in one thread.

Yes, but that was by design, as I wanted first to get a sense if there was general agreement that some changes should be made. It seems to me that there is decent support for Revisiting Transcripts and Audio Description.

I naively proposed just deleting 1.2.3 as the simplest/quickest fix, but I now agree that is not the best approach. The actions you enumerate are a good start, but I think that there are few other options as well. For example, rather than than just promoting 1.2.8 to AA, I think it would be a cleaner end-product if 1.2.5 were re-written to require both captions and audio description (because then it could have parallel construction to 1.2.3). But that option presumes that transcripts should be AA and not A, and we have not had even one phone call (let alone a survey) about any of this.

bruce-usab commented 5 years ago

@patrickhlauke

Coming in late, but for what it's worth, I'd suggest that for content authors, it's orders of magnitude more difficult to provide an alternative version of a video with AD...

Agreed. Especially with the non-uncommon example of the constantly-talking talking-head talking over (but not reading) their slides!

Since transcripts are so much easier than AD, and better for certain use cases, does that not argue for 1.2.8 being promoted to Single A?

Or should we try and craft an exception for transcripts being permissible at AA under certain conditions (e.g., when extended AD is needed to provide comparable information, and regular AD is not good enough)?

I'm also reminded of the weird imbalance in requiring captions for live audio_video at AA, but for live audio only at AAA, but i'll repost this as a separate issue.

I think that is because requiring captions on an audio-only live broadcast is something of a fundamental alteration.

I think it is a bigger gap that we do not have a AAA requirement for AD on live multimedia.

johnfoliot commented 5 years ago

@bruce-usab

I think it be a cleaner end product if 1.2.5 were re-written

I'll oppose that too I'm afraid, not because I don't understand the concern or your issue, but because it breaks the backward compatibility requirement, by changing normative text after the fact. Our options are limited here, and it also sounds like your problem is not with the requirements, but with how it is explained and taught.

I personally do not think there is a lot of confusion here if you respect the A, AA and AAA model - the issue is that Section 508 flattened that model to make A and AA equivalent (or at least equally required), which WCAG clearly states they are not. ¯_(ツ)_/¯

bruce-usab commented 5 years ago

@johnfoliot

I don't understand the concern or your issue

I agree that you do not understand the concern or my issue.

it also sounds like your problem is not with the requirements

That is not correct. My problem is with the requirements. I would like to see the normative text improved.

patrickhlauke commented 5 years ago

I think it is a bigger gap that we do not have a AAA requirement for AD on live multimedia

pardon my ignorance, but...can you even DO audio description of live multimedia? you can't foresee where the gaps in audio/speech occur if you're doing it live...and again from a technical standpoint it's non-trivial, having to provide two separate live streams instead of one

bruce-usab commented 5 years ago

@patrickhlauke asked:

can you even DO audio description of live multimedia?

Yes, but it is tricky, and the live event almost certainly would need to be following a script. OTOH live radio of sporting events is essentially live AD.

I will try and dig up some examples (which would be recordings of live events, e.g. the U.S. Presidential Inauguration). One can also find live theatrical productions that offer live AD. The Kennedy Center does a nice job with this. The ACB ADP has a full listing of Performing Arts with Audio Description.

EDIT: I could not quickly find a recording of the AD version of the 2013 inauguration. Here is the LOC page which mentions that it was, as well as this Cool Blind Tech article.

patrickhlauke commented 5 years ago

adding though that i can see how/why this whole thing is a confusing mess that should be sorted out...

1.2.3 EITHER provide AD or transcript at level A 1.2.5 provide AD at level AA 1.2.8 provide transcript at level AAA

1.2.8 is easier (technically and logistically) to achieve than 1.2.5, so it's odd that it's AAA compared to 1.2.5 which is AA

and then there's the weird logical conundrum of "at A you have to either do AD - a AA requirement - or transcript - a AAA requirement" ... the levels and just overall logic of that is ... odd

johnfoliot commented 5 years ago

@bruce-usab

I would like to see the normative text improved.

Unfortunately, I do not believe that is an option open to us, and I will (as repeatedly stated) strenuously oppose any attempt to make changes to WCAG 2.x that break backwards compatibility, sorry. That requirement is sacrosanct.

This normative text has been in place for over a decade now, and it is only now, with the advent of Section 508 adopting WCAG 2.0 A & AA that this has become an issue? I'm sorry, but respectfully, I do not think the source of the issue is with the text(s), but rather in the way it is being taught. I've previously stated I would favor and encourage re-visiting the non-normative Understanding documents around these Success Criteria if that is the root of the misunderstanding, and I'll again suggest that is the better path forward.

I would also support making SC 1.2.8 a AA going forward in WCAG 2.2 (if that helps any)

bruce-usab commented 5 years ago

@johnfoliot wrote:

I do not believe that is an option open to us

We do have that option, so long as we do not break backwards compatibility. You were on the call Tuesday and last week when this was discussed.

I will (as repeatedly stated) strenuously oppose any attempt to make changes to WCAG 2.x that break backwards compatibility

I have not proposed anything that breaks backward compatibility.

I do not think the source of the issue is with the text(s).

I understand that you do not think so. Sorry, but you are mistaken about that! :-)

johnfoliot commented 5 years ago

@patrickhlauke

adding though that i can see how/why this whole thing is a confusing mess that should be sorted out...

Yep, which is why I'd be OK with moving SC 1.2.8 to a AA level, so that we end up with:

At level A:

But at AA you must provide BOTH:

(And respectfully, if that is hard to explain, then I am unsure what to say... NOT EVERY TERRITORY IS MANDATED TO MEET BOTH A & AA REQUIREMENTS.)

Prior to the Section 508 refresh, 508 demanded less than WCAG A & AA, and yet, at that time, the common thought was "something was better than nothing". Now Section 508 has to meet both the A & AA requirements, but that requirement only extends to US entities covered by that act.

Meanwhile, beyond the reach of Section 508, some entities will strive to at least meet WCAG A (even if not legally required to do so, and again operating under the principle that something is better than nothing) and in those use-cases, those entities are at least required that they provide something to impacted users.

At AA conformance, we increase the demand for both accommodations.

Why is this hard?

johnfoliot commented 5 years ago

@bruce-usab

You were on the call Tuesday and last week when this was discussed.

Indeed, and I stated then (and will re-state now) that, in the case of SC 4.1.1, "removing" that requirement is moot, because the problem it sought to address is now 'fixed' in the browsers - the "user-agents" remediate the code on the fly. However, I also stated clearly that I would oppose any change to normative text that still has an impact on the content authors, which is what you are proposing here.

Bruce, throughout this thread, you have stated things such as:

"In my experience... (it) has proven to be a surprisingly difficult stumbling block with teaching about WCAG."

"...indicate if you find 1.2.3 particularly annoying to teach about (or not)."

"A simple explanation for my experience is that I am poor teacher and/or have dull students!"

"It is an issue because this single instance one SC superseding another SC has been an (unexpected) impediment to my own teaching about WCAG."

"It is both a teaching problem and a structural problem with the actual SC."

"I mostly teach old tired admin/policy HR folks who were anointed 508 responsibilities under “other duties as assigned” and they have been struggling to make sense of WCAG on their own for many months before I get a chance to help them."

@lauracarlson also chimed in:

At first people struggle with the concept of SC 1.2.5 superseding SC 1.2.3 at AA. Once it is explain then they understand. Maybe that could be explained better in the understanding docs? People I work with have read the understanding docs, and still didn't get it until I explained it.

I cannot help but conclude then that the problem is in the teaching, not in the requirement(s) themselves.

As Laura states, "Maybe that could be explained better in the understanding docs", to which I wholeheartedly agree. I get that you feel something needs fixing, but re-writing an existing SC, or eliminating it because Section 508 coordinators don't understand is something I reject as a viable option - the impact of that decision will extend far beyond the boundaries of the United States.

You've effectively made the point this is hard to teach - so let's fix the teaching materials.

patrickhlauke commented 5 years ago

Why is this hard?

can we just chill for a minute here? i'm adding some of my observations, not arguing for one side or another at the moment.

bruce-usab commented 5 years ago

@johnfoliot

I also stated clearly that I would oppose any change to normative text that still has an impact on the content authors

That cannot be what you exactly mean, since new SC in 2.1 and 2.2 have impact on the content authors. I think I am hearing you say that you would be okay with promoting 1.2.8 but not with incorporating the text of 1.2.8 into 1.2.5 (even though the functional end result would be the same). Am I correct about that? I am agnostic about the mechanics of any change at this point.

Bruce, throughout this thread, you have stated things such as:

I was only trying to make it clear that I read what you wrote.

I cannot help but conclude then that the problem is in the teaching, not in the requirement(s) themselves.

The problem is that the normative text is confusing on its face. If my students were first being introduced to WCAG from my teaching, there would probably not be any difficulty, but that is not usually the case.

An inherit defect with the SC text cannot be resolved by patching Understanding.

You've effectively made the point this is hard to teach - so let's fix the teaching materials.

It is not so hard to teach. It is harder than it needs to be for people to pick up this sort of nuance on their own.

bruce-usab commented 5 years ago

I have made a request for a couple survey questions:

(1) Should transcripts be required for prerecorded synchronized media at either Level A or Level AA?

(2) Should we consider requiring AD for live events at AAA?

Please read “transcript” as shorthand for “textual alternative for time-based media”, per the WCAG definition.

I think it is too soon to survey for how to best address (1) and (2) above, as I think we first need to decide if there even should be a change.

alastc commented 5 years ago

...some videos don't have sufficient pauses for AD. ... That use case would not actually fail against 1.2.5 (since, by definition, AD narration is only dubbed over the default sound track).

Hmm, 1.2.5 was raised for our 'showreel' video during 2.1 testing, and I was not convinced that audio desc was useful for that, at least when there is a transcript as well, and it is basically an alternative way of understanding the entire website. (We removed the video from the homepage to sidestep that for 2.1 testing.)

When providing training, our clients are an odd mix of multimedia heavy (e.g. broadcasters), or multimedia light (e.g. companies with the odd video). I generally start with the premise of "what do people do if they can't see it? And how about if they can't hear it?" and work out what the best solutions are.

For the non-broadcasters, even pretty big organisations I've worked with generally provide transcripts and not audio-desc, even whilst aiming for AA in every other respect. The audio-desc recommendation is made, but generally left sitting in a backlog somewhere.

The gripe I have is that the current setup discourages 'supplemental' videos. E.g. you have a page of text content, and a video that is saying the same thing. Great for people (including with cognitive issues) that don't like reading lots of text. I'd like to be able to treat that as a translation rather than independent content... but that would be a bigger change.

If we were starting from scratch (e.g. Silver) I'd propose an equivalent to:

In the WCAG 2.2 context I'd be happy to move transcripts up, and/or rationalise the requriements so long as they match/exceed the previous requirements (which means I can't have my wish for exceptions to AD).

If my students were first being introduced to WCAG from my teaching, there would probably not be any difficulty, but that is not usually the case.

That might be a cause of different experiences, most people in my training have never read the guidelines.

bruce-usab commented 5 years ago

@alastc, I agree with (almost) all your points, so I am just nit-picking one thing here…

The gripe I have is that the current setup discourages 'supplemental' videos. E.g. you have a page of text content, and a video that is saying the same thing. Great for people (including with cognitive issues) that don't like reading lots of text. I'd like to be able to treat that as a translation rather than independent content... but that would be a bigger change.

I think the 2.0 phrasing explicitly allows for this use case! All of 1.2.1, 1.2.2, 1.2.3 include “except when the media is a media alternative for text and is clearly labeled as such”. Is that not your “video that is saying the same thing”?

One issue (that I have not submitted yet) is that I am pretty sure that it is essentially a typo that 1.2.4 and 1.2.5 (and 1.2.8 and maybe some of the other AAA in Guideline 1.2) do not include this exception language.

That might be a cause of different experiences, most people in my training have never read the guidelines.

Yeah, I have a little too much empathy for your situation...

mraccess77 commented 5 years ago

When videos are a media alternative to other content then they don’t have to meet the audio description requirements

bruce-usab commented 5 years ago

@mraccess77 wrote:

When videos are a media alternative to other content then they don’t have to meet the audio description requirements

I agree with you. This is explicitly true at Single A. Unfortunately, the concept is only implicit with AA, and @alastc mentions getting caught by 1.2.5. This is another example of less-than-perfect drafting on our part.

mraccess77 commented 5 years ago

@bruce-usab I wonder if the definition of synchronized which is normative covers the AA SC because it address media alternatives.

bruce-usab commented 5 years ago

@bruce-usab I wonder if the definition of synchronized which is normative covers the AA SC because it addresses media alternatives.

It does indeed! Unfortunately, that observation begs the question: Why then does the exception appears in 1.2.1, 1.2.2 and 1.2.3? Answer: less-than-perfect drafting! I submit that 2.2 is a good opportunity to fix this sort of thing.

I just added a new issue highlighting this defect.

Ryladog commented 5 years ago

Bruce,

I am for making transcripts required at A or AA. I think it will have to be a new SC (with explanations in the Understanding). I have long felt this has been needed.

To answer your questions that you would like submitted to survey....

  1. Should transcripts be required for prerecorded synchronized media at either Level A or Level AA? KHS: YES

    • No, but I can live with a change.
    • No, and I oppose any change for 2.2
    • Yes, and I am open to transcripts being Level A or AA. KHS: YES
    • Yes, but only if transcripts are Level A
    • Yes, but only if transcripts are Level AA

  2. Should we consider requiring AD for live events at AAA? KHS: YES

From my experience teaching WCAG the media requirements of 1.2.3 and 1.2.5 have always been a source of confusion. Additionally I find that in many cases teaching those in government often brings more WCAG-first-timers.

I would like to advise the chairs to please monitor this and other threads as I see disrespect and bullying raising its ugly head yet again (no matter how many times one tries to qualify with 'Respectfully'). Having to sit by and listen to someone requiring a member to 'prove' the truth of their statement and the practice of demeaning the experience of others is very unfortunate - and a stain on this WG.

Peace out!

katie

Katie Haritos-Shea

*Principal ICT Accessibility Architect, Vice President of Accessibility at EverFi, *Board Member and W3C Advisory Committee Rep for Knowbility

WCAG/Section 508/ADA/AODA/QA/FinServ/FinTech/Privacy, IAAP CPACC+WAS = CPWA http://www.accessibilityassociation.org/cpwacertificants

Cell: 703-371-5545 <703-371-5545> | ryladog@gmail.com ryladog@gmail.com | Oakton, VA | LinkedIn Profile http://www.linkedin.com/in/katieharitosshea/

People may forget exactly what it was that you said or did, but they will never forget how you made them feel.......

Our scars remind us of where we have been........they do not have to dictate where we are going.

On Thu, Jun 20, 2019 at 1:37 PM bruce-usab notifications@github.com wrote:

@bruce-usab https://github.com/bruce-usab I wonder if the definition of synchronized which is normative covers the AA SC because it addresses media alternatives.

It does indeed! Unfortunately, that observation begs the question as to why the exception appears in 1.2.1, 1.2.2 and 1.2.3. It is just less-than-perfect drafting, and 2.2 is a good opportunity to fix this sort of thing.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/w3c/wcag/issues/782?email_source=notifications&email_token=ABL6VSVBOLG7CGPB5L53DCLP3O54BA5CNFSM4HW7GZKKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODYGDJCI#issuecomment-504116361, or mute the thread https://github.com/notifications/unsubscribe-auth/ABL6VSSWVE2WF5NORHF7SW3P3O54BANCNFSM4HW7GZKA .

alastc commented 2 years ago

Re-labelling as this isn't a 2.2 criteria, and can't be dealt with as a 2.2 update.