Re: My Thoughts on User content, 3rd party content and conformance

A lot of this is one of the things that ATAG was designed to cover, back last century. I believe that the more recent revision, ATAG 2, still covers it, largely in https://www.w3.org/TR/ATAG/#principle_b2 - https://www.w3.org/TR/ATAG/#principle_b2


While it is impossible to ensure that this actually works, since in the end ATAG indeed allows for authors to deliberately avoid any of the accessibility support and actively make things inaccessible if they so choose, it would seem reasonable to at least require that systems that ingest user-generated content apply ATAG to the mechanism used to do so, and to tie the level of WCAG compliance to the level of ATAG compliance. The levels are reciprocal, part of the design since well before ATAG 1 was released.


It is slightly unfortunate that ATAG 2 relies on the now somewhat outdated WCAG2, and that there is no obvious plan to update it to at least refer to WCAG 2.2 should that become a Recommendation. Although 


1. that seems like something that might not be terribly difficult to fix, and
2. it's a lot better than nothing, which seems to be where this conversation might otherwise settle. :(


cheers


Chaals


On Wednesday, May 24, 2023 14:27:49 (+02:00), John Foliot wrote:


Hi Wendy,


> It’s OUR responsibility to help content creators create accessible books


I agree 100% with this, but with the caveat that it is NOT our responsibility if the content creators ignore our help. 


I truly believe this is where we MUST draw the line: we cannot (I suspect legally or technically) FORCE the content creators to accept and implement all of that offered help, no matter how much we try (especially at scale). In other words, we cannot shift from ensuring there is a robust support ecosystem that fosters accessible development, to forcing site owners to police every piece of third-party content published on the site. 


So I think one of the bigger questions is "how much help is enough?" For example, using an easy scenario, when a third-party creator uploads an image file, we could say that the site *MUST* provide a means for also providing the text alternative. But is it the site's responsibility to also evaluate the quality of the alt text? Imagine a site where alt text is also suggested: that is a strong support tool for the content creator, but if they choose NOT to accept the suggested text, and instead provides alt text of "lower" quality, then this should not come back onto the site owner: they did their best to aid accessible development, but the creator chose not to accept the help.

Can some of this be shored up via Terms of Conditions (etc.)? Likely yes, and ensuring policy statements to that effect (i.e. "It is a requirement of this site that when publishing user-supplied content that all images must have a textual alternative" - and then "graying out" the submit button until that condition is met), might be an area where the level of technical and policy support would be a contributing factor when it comes to scoring. So for example, Bronze: editable alt text field is present; Silver, editable alt text field is present and pre-populated with a suggested alt text; Gold, editable alt text field is present, pre-populated, and requires that the author verify and 'accept' the suggested alt text or furnish a different text, and the submit button is disabled until that verification is completed (check the box that says "I accept this alt text" or similar).

JF


On Tue, May 23, 2023 at 5:09 PM Reid, Wendy <wendy.reid@rakuten.com> wrote:


I’m with Alastair, Sheri, and Jon here (this is a topic I’m particularly interested in/passionate about).

 

This is a really thorny topic, as evidenced by the discussion we’ve had already, but I think it can be summarized into a few essential elements.

 

When we’re thinking about responsibility, sourcing, and conformance, there are a few things to consider:

Is this thing content, or a component? (content being an object, a component being software or an API that may deliver content)
Who is creating this content or component? (An individual user, a company – big or small, the component itself)
What phase is the content in? (Creation, transmission, display, consumption)
Who is consuming the content? (Individual users, companies, APIs, components)
Will this content be transformed in any way during any phase?
Who is transforming the content, and when?
Will those transformations impact the end user?
Who is the end user?
What is the impact on the end user?

 

There are probably more questions that aren’t coming to me right now. To illustrate what I’m getting at here, I think responsibility and ownership matter a lot to this conversation, and I do think we need to increase responsibility in areas we have previously let off the hook because of the vagueness of “user generated content”.

 

This comes up a lot with my employer which is why I’ve spent so much time thinking about this. Some examples:

For every ebook on our retail site we have metadata (title, synopsis, author, price, etc.), that information is provided to us in a standardized format by the publisher and/or author of the book.
It’s OUR (the site’s) responsibility to display that information in an accessible format
It’s OUR responsibility to ensure our publishers/authors know which information to provide and in what format to make it possible for us to display in an accessible way
It’s NOT our responsibility to ensure every book cover meets WCAG contrast requirements for text on a background (we don’t make the book covers)
There is a standard for developing accessible EPUB files, we do not produce the files on behalf of publishers for sale, however we do produce the platforms that display them.
It’s OUR responsibility to make the platforms accessible
It’s NOT our responsibility to make the ebooks accessible
We SHOULD use whatever tools at our disposal to check the accessibility of the books when they are delivered (probably unpopular but I’d even raise this to a must when we consider the power/responsibility imbalance often present between platform and content creator)
We have a self-publishing arm that allows content creators to publish their content on our site for sale, as part of that we offer conversion of documents to EPUB (our primary format), as well as easy-to-use forms to provide necessary metadata.
It’s OUR responsibility to make that platform accessible.
It’s OUR responsibility to help content creators create accessible books  this is the contentious bit, because I think traditionally ebook content falls into “third party” territory, but when we consider the questions, the ownership, and the user impact, responsibility shifts.

 

If we are considering conformance exclusively, there are ways for us to create conformance conditions for many “third party content”/”user content” situations that can look at how we distribute responsibility and ownership in fair ways. I think traditionally we have put a lot of the responsibility on site owners, but with how the tech landscape has evolved there are so many more places that we can distribute responsibility that can even out responsibility to a more sustainable level.

 

-Wendy

 

From: Alastair Campbell <acampbell@nomensa.com>
Date: Tuesday, May 23, 2023 at 6:17 AM
To: Gregg Vanderheiden RTF <gregg@raisingthefloor.org>
Cc: w3c-waI-gl@w3. org <w3c-wai-gl@w3.org>
Subject: Re: My Thoughts on User content, 3rd party content and conformance

[EXTERNAL] This message comes from an external organization.

Hi Gregg,

 

I was reading your original email as present tense, i.e. wcag2. My misunderstanding, if it’s all in WCAG 3 context that makes sense.

 

On the points:

 

GV: 2)  Third party provided parts of a page created by any contractors the author hired to are part of the page   and are covered by WCAG.   

 

I’m not sure anyone is arguing against this, anytime the ‘website owner’ is making a procurement choice (e.g. supplier, library of code, API) that should be part of the site claiming conformance.

I agree with Sheri that it may or may not be a paid thing, but it is where the site owner has made a direct a choice about what to include. (Rather than opening the site to end-users to post things.)

 

The nature of these things is that they are perceived as part of the site and not authored individually. (So that would include a free or paid display of maps from a service, but not include individual users creating maps to display on the site.)

 

I guess there might be difficult situations where there are no accessible choices, such as (my experience with) GIS software, but that’s a wider problem with that type of software.

 


AC: However, once saved and part of the page (for that user and other users), why isn’t that covered?

GV: Because, for one - that would mean that all email providers would have to read your email and make it accessible before delivering it to your recipient. 

 

Ok, I can see problems with personal communications, particularly where the content could be authored by any number of applications and the service must accept and display them.

 

However, for a social media site which controls the authoring interface and displays the posts to the public, I’m not sure why we would differentiate that?

 

That’s something we know how to do and is practical today, at least for most content types.

 

Obviously, someone determined to post inaccessible content will find a way, but when you control the authoring interface there should be a responsibility to encourage accessible content.

 

In that scenario I think we’d want to have an assertion that they have reviewed the authoring interface and ensured that users are encouraged to create accessible content.

 

Kind regards,

 

-Alastair

 

-- 

 

@alastc / www.nomensa.com

 

 





--

John Foliot | 
Senior Industry Specialist, Digital Accessibility |
W3C Accessibility Standards Contributor |


"I made this so long because I did not have time to make it shorter." - Pascal
"links go places, buttons do things"

-- 
Charles 'Chaals' Nevile
Lead Standards Architect, ConsenSys Inc