Re: Collections of web pages

Matt,

I want to focus on one specific thing you wrote below: “…I assume the goal here is to be as minimally disruptive of the current terminology as possible…”  I want to push back on that.

Why should that be the goal?

As you have articulated nicely below, eBooks are a bit of different animal than a web site – being at times both more like a web page and at others more like a collection of web pages (or, dare I suggest, an entire website unto themselves).  Given this, I think the right approach is to start with an evaluation from first principles:

  *   What do the four key principles mean in the context of an eBook – perceivable, operable, usable, and robust?
     *   Is there anything about these principles that don’t map onto an eBooks [I suspect not]?
     *   Are there any principles that are missing when think about eBooks [perhaps; I’m not sure, but again I suspect not]?
  *   For each Guideline, what do the Guidelines contained within that principles mean in the context of an eBook?
     *   Any missing?  Any that don’t apply?  [I suspect both things are the case – especially as we think about cognitive impairments where there is significant understanding around techniques like adjustable justification of rendering to make a book’s text right ragged]
  *   [and then finally], For each success criterion…

And then, for each that apply, be unconstrained in modifying the language of that criterion so that fits well – not just “substitute eBook for webpage” (or “some portion of an eBook for webpage”).

I suspect most of the people in the publishing world who are striving to create accessible eBooks will be far more fluent in what it means to be a book than to be a web page.  I see no value in pushing them to first understand web terminology and lore, and then to get comfortable with the places where we simply directly substitute “eBook” for “web  page” (or “set of web pages” or whatever).


Also, finally, I think we should be aware of what is going on in Silver/WCAG 3.  Given that they are completely comfortable re-writing success criterion for their work, making eBook accessibility language slavishly follow a structure we are quite prepared to move away from, makes even less sense.


Regards,

Peter
--
Peter Korn | Director, Accessibility | Amazon Lab126
pkorn@amazon.com

From: Matt Garrish <matt.garrish@gmail.com>
Date: Monday, April 13, 2020 at 4:41 AM
To: 'Avneesh Singh' <avneesh.sg@gmail.com>, 'John Foliot' <john.foliot@deque.com>, "Korn, Peter" <pkorn@lab126.com>
Cc: 'Bruce Bailey' <Bailey@access-board.gov>, 'Alastair Campbell' <acampbell@nomensa.com>, 'Andrew Kirkpatrick' <akirkpat@adobe.com>, "w3c-wai-gl@w3.org" <w3c-wai-gl@w3.org>, 'Shawn Lauriat' <lauriat@google.com>, '508' <508@access-board.gov>
Subject: RE: [EXTERNAL] Collections of web pages


CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you can confirm the sender and know the content is safe.


Hi all,

The problem I find for EPUB is that it is both a “web page” – available at a single URL when it is on the web or at a file URL – and also a set of web pages once you obtain the packaged resource and begin to consume it. The URLs of the packaged pages may just be opaque to all but the reading system.

Similarly, the forthcoming audiobooks specification and any future publication formats like it are designed to exist both as expanded sets of pages and also as packaged single resources, as Avneesh has pointed out. They can’t be different things, and aren’t, just because of the packaging.

Where things get even more weird is that digital publications can be a single web page of content within the container file, so you have a set that itself only has to be one page (good old singletons). The current definition of a “set of web pages” doesn’t invalidate this, but there’s an inference of two or more.

What we want with digital publications is for the set to always be evaluated as a single entity independent of it being packaged. It’s why, for example, we chose with the publication manifest and audiobook specifications to focus more on the bounded aspect of the resources – that the one or more pieces of content are all connected as a single logical unit with a beginning and end (linear or non-linear reading not mattering).

But to avoid rambling on too much, I assume the goal here is to be as minimally disruptive of the current terminology as possible. If so, I’d look at massaging the “set of web pages” definition as I think it can be made to bridge some of the current ambiguity we have with publications.

My best take so far on this is maybe a bit of a mouthful, but goes:

set of web pages
one or more web pages<https://www.w3.org/TR/WCAG21/#dfn-web-page-s> logically connected, and with at least one progression, such that whole would be incomplete or unusable if any were unavailable
To that I would add a note like the following about packaging:

Note: A set of web pages may be packaged and made available as a single web page (e.g., EPUB), but the packaging does not alter the nature of the set.

I realize “one or more” may introduce some confusion with the single web page definition, so that could always be left as “collection of” if digital publications are unique in sometimes only consisting of a single resource. The singular nature could be hinted at in an example instead.

I’m sure this group can improve on what I’ve offered up, but hopefully this helps frame the publication issue a bit.

Matt


From: Avneesh Singh <avneesh.sg@gmail.com>
Sent: April 10, 2020 06:50
To: John Foliot <john.foliot@deque.com>; Korn, Peter <pkorn@lab126.com>; Matt-Garrish <matt.garrish@gmail.com>
Cc: Bruce Bailey <Bailey@access-board.gov>; Alastair Campbell <acampbell@nomensa.com>; Andrew Kirkpatrick <akirkpat@adobe.com>; w3c-wai-gl@w3.org; Shawn Lauriat <lauriat@google.com>; 508 <508@access-board.gov>
Subject: Re: Collections of web pages

“Now, what if instead we had a different format eBook, which didn’t use zip or any other component-based compression scheme, but which appeared to the reader to be the same content, presented in the same way, as its ePub equivalent?  Would that non-ePub eBook also be a “set of web pages”?  “

The definition should not be based on zip or container. The audio books specifications developed in Publishing Working Group allows publishers to publish a set of files as well as files encapsulated form of light weight zip container.
The best person from digital publishing group to work on definition is Matt Garrish, the chief editor of EPUB 3 specifications. He is copied.

With regards
Avneesh
From: John Foliot
Sent: Friday, April 10, 2020 3:38
To: Korn, Peter
Cc: Avneesh Singh ; Bruce Bailey ; Alastair Campbell ; Andrew Kirkpatrick ; w3c-wai-gl@w3.org<mailto:w3c-wai-gl@w3.org> ; Shawn Lauriat ; 508
Subject: Re: Collections of web pages

Peter asks (some very good questions):

Now, what if instead we had a different format eBook, which didn’t use zip or any other component-based compression scheme, but which appeared to the reader to be the same content, presented in the same way, as its ePub equivalent?  Would that non-ePub eBook also be a “set of web pages”?

I think Peter that part of the problem is that we keep calling them "web pages": what if, instead we just called them "pages"? In that case, any collection of "pages" would (by definition) be a "book" (publication?) and in that case I am simply suggesting that higher-order "collection of..." requirements are applied to whatever the parent container is called: book, site, (and perhaps, in the future with XR, "scenes" inside an "instance" - I don't know the answer to that specific a question today, but you get (I hope) the idea). I think the critical bit is that alone, the content (page, screen, scene, whatever) is incomplete at best, and completely unusable at worst, but by design: it is *intended* to be part of a larger "collection".

I think too that while there has been significant standardization in digital publishing since the early days (.mobi anyone?), I'd none-the-less seek some guidance from the W3C Digital Publishing Group on an appropriate term and perhaps even definition (if they have such a beast). In that regard, I know Avneesh to be an active member of that group, and welcome his thoughts - or perhaps he could take it back to that WG for more input? Perhaps eBook is right, or perhaps there is a better more inclusive term that we should be using. (Open Question: does 'pagination' have a place here? i.e. content that relies explicitly - or perhaps even implicitly - on pagination as part of the delivery process is a "collection" of pages/screens/contentblocks - and for those collections, 1 of *foo* is sufficient: findable help, privacy or accessibility policy, etc..)


No matter what, I am personally unhappy with "web page" as a unit of measurement, and I believe that WCAG 3.0 / Silver is trying to avoid that narrowly defined container as well.

FWIW

JF

On Thu, Apr 9, 2020 at 2:06 PM Korn, Peter <pkorn@lab126.com<mailto:pkorn@lab126.com>> wrote:
John, Avneesh,

If an ePub file/book is, under the covers, a zip file; and if this ePub file is a “set of web pages”, then what constitutes an individual web page within that set?  Each of the individual files if we were to unzip it?  Each page (what is a page when screen sizes & font sizes vary so much)?  Each chapter?

Now, what if instead we had a different format eBook, which didn’t use zip or any other component-based compression scheme, but which appeared to the reader to be the same content, presented in the same way, as its ePub equivalent?  Would that non-ePub eBook also be a “set of web pages”?

I’m sorry that I’m coming into this analysis of ePub through the lens of WCAG rather late; I know I’ve missed many conversations about this, and I appreciate patience as I come up to speed.  Looking through the 3 WCAG 2.1 SCs that reference “set of Web pages”, I see the following text:

2.4.5 Multiple Ways: More than one way is available to locate a Web page<https://www.w3.org/TR/WCAG21/#dfn-web-page-s> within a set of Web pages<https://www.w3.org/TR/WCAG21/#dfn-set-of-web-pages> except where the Web Page is the result of, or a step in, a process<https://www.w3.org/TR/WCAG21/#dfn-processes>.
2.4.8 Location: Information about the user's location within a set of Web pages<https://www.w3.org/TR/WCAG21/#dfn-set-of-web-pages> is available.
3.2.4 Consistent Identification: Components that have the same functionality<https://www.w3.org/TR/WCAG21/#dfn-same-functionality> within a set of Web pages<https://www.w3.org/TR/WCAG21/#dfn-set-of-web-pages> are identified consistently.

Thinking about the user needs, and translating this into an eBooks, I’d think the right approach is something like:

2.4.5 Multiple Ways: More than one way is available to locate a [page? Chapter? Section?] of the eBook.
   (e.g. a TOC, page numbers, section titles in footer)
2.4.8 Location: Information about the user's location within the eBook is available.
   (e.g. a page number, %, or other counter)
3.2.4 Consistent Identification: Interactive elements that have the same functionality<https://www.w3.org/TR/WCAG21/#dfn-same-functionality> within the eBook identified consistently.

This language doesn’t so easily map onto simply re-defining what “web page” and “set of web pages” means in an eBook/ePub context, but feels to me to do a better job capturing the intent of these SCs and the needs of customers with various disabilities in using eBooks.


Peter
--
Peter Korn | Director, Accessibility | Amazon Lab126
pkorn@amazon.com<mailto:pkorn@amazon.com>

From: Avneesh Singh <avneesh.sg@gmail.com<mailto:avneesh.sg@gmail.com>>
Date: Wednesday, April 8, 2020 at 10:43 PM
To: John Foliot <john.foliot@deque.com<mailto:john.foliot@deque.com>>, Bruce Bailey <Bailey@access-board.gov<mailto:Bailey@access-board.gov>>
Cc: Alastair Campbell <acampbell@nomensa.com<mailto:acampbell@nomensa.com>>, Andrew Kirkpatrick <akirkpat@adobe.com<mailto:akirkpat@adobe.com>>, "w3c-wai-gl@w3.org<mailto:w3c-wai-gl@w3.org>" <w3c-wai-gl@w3.org<mailto:w3c-wai-gl@w3.org>>, Shawn Lauriat <lauriat@google.com<mailto:lauriat@google.com>>, 508 <508@access-board.gov<mailto:508@access-board.gov>>
Subject: RE: [EXTERNAL] Collections of web pages
Resent-From: <w3c-wai-gl@w3.org<mailto:w3c-wai-gl@w3.org>>
Resent-Date: Wednesday, April 8, 2020 at 10:42 PM


CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you can confirm the sender and know the content is safe.




From: John Foliot
Sent: Wednesday, April 8, 2020 23:09
To: Bruce Bailey
Cc: Alastair Campbell ; Andrew Kirkpatrick ; WCAG list (w3c-wai-gl@w3.org<mailto:w3c-wai-gl@w3.org>) ; Shawn Lauriat ; 508
Subject: Re: Collections of web pages

Bruce writes:

> I am arguing that we make the same common sense leap for ePub and WCAG 2.2.  A typical ePub, posted online as a zip file, is a set of web pages, full stop.

+1

Avneesh: +1
EPUB is set of HTML pages in form of files properly arranged in zip container.

An E-Pub (Electronic Publication) is a singular unit that comprises multiple screens or views, but is traditionally thought-of as a single and complete entity.
It traditionally also has a single table-of-contents, which I will argue also suggests to me that a single "findable help" would be (in context) appropriate. This is not to say that
content creators cannot *also* provide contextual help in the 'footer' of each e-pub document if desired, only that it would not be mandated to do so.

Thoughts?

JF

On Wed, Apr 8, 2020 at 12:06 PM Bruce Bailey <Bailey@access-board.gov<mailto:Bailey@access-board.gov>> wrote:
Thanks Alastair for kicking off this discussion.  CC’ing John Foliot since he has some strong opinions about this.  CC’ing Shawn Lauriat because he has articulated how our current definition of web page does not stand up to technical scrutiny.

Forgive me, but I will remind folks that in 2006 the WG though we needed a new term, “web unit”.  The good old bad old days!
www.w3.org/TR/2006/WD-WCAG20-20060427/appendixA.html#webunitdef<http://www.w3.org/TR/2006/WD-WCAG20-20060427/appendixA.html#webunitdef>

Can we agree that there is a certain amount of hand waving required with our current definition of web page?

I agree that a typical PDF file is a web page.
I agree that a PDF collection could be posted in a way that it is a set of web pages.  I pretty confident we can agree it is not typical.  For this discussion, I would really rather we not spend cycles talking about PDFs.

I disagree that posting a .zip file (or similar archive of a collection) has any meaningful implication to our discussion of web page or set of web pages.  Yes, files posted online have a URI.  Not every URI is a web page!

If one archives a set of web pages into a single zip file (and posts the zip online), it would be nonsensical to assert that the URI is now a web page and no longer a set of web pages.

I am arguing that we make the same common sense leap for ePub and WCAG 2.2.  A typical ePub, posted online as a zip file, is a set of web pages, full stop.

I admit that my argument is not in the shape of good formal logic.  I would ask that anyone who disagrees (than an ePub is a set of web pages) make a recommendation to how our definition of web page and set of web page might be tweaked (so that they would agree that an ePub meets their modified definition for set of web pages).


From: Alastair Campbell <acampbell@nomensa.com<mailto:acampbell@nomensa.com>>
Sent: Wednesday, April 8, 2020 11:42 AM
To: Andrew Kirkpatrick <akirkpat@adobe.com<mailto:akirkpat@adobe.com>>; Bruce Bailey <Bailey@Access-Board.gov<mailto:Bailey@Access-Board.gov>>
Cc: WCAG list (w3c-wai-gl@w3.org<mailto:w3c-wai-gl@w3.org>) <w3c-wai-gl@w3.org<mailto:w3c-wai-gl@w3.org>>
Subject: Collections of web pages

Hi Andrew, Bruce & everyone,

During the discussion of two criteria (at least), the concept of “set of web pages” came up as a key point.


•  Findable help: Including ‘set of web pages’ helps to scope-out the very simple one-page websites and PDFs that are less likely to have human contact details.

•  Fixed reference points: It says “a web page or set of web pages" so that it covers ePub and non-ePub files .

Andrew mentioned that long PDFs could be considered a ‘set of web pages’, and that some PDFs techniques mention that.

As far as I can tell from our definition for a web page<https://nam01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.w3.org%2FTR%2FWCAG21%2F%23dfn-web-page-s&data=02%7C01%7CBailey%40access-board.gov%7Cc6d89b46797f49213b0808d7dbd36f2d%7Cfc6093f5e55e4f93b2cf26d0822201c9%7C0%7C0%7C637219573475002609&sdata=gHYOWONUhzMRcA04Vv1eJhLF1DSlhV93bdnPX6QIfnA%3D&reserved=0> and set of web pages<https://nam01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.w3.org%2FTR%2FWCAG21%2F%23dfn-set-of-web-pages&data=02%7C01%7CBailey%40access-board.gov%7Cc6d89b46797f49213b0808d7dbd36f2d%7Cfc6093f5e55e4f93b2cf26d0822201c9%7C0%7C0%7C637219573475012600&sdata=yFqJ7fz4UUlIGWgKiP%2F3wSjXdWjAUavo%2F3lEdu%2B7xHI%3D&reserved=0>, all of these would be considered a ‘web page’ as they are located at a single URI:

•         A PDF;

•         An ePub document;

•         A ‘single page app’, unless it adjusts the URI & browser history to appear to have multiple pages.

I can’t see a reference to ‘set of web pages’ in the PDF techniques<https://nam01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.w3.org%2FTR%2FWCAG20-TECHS%2Fpdf%23PDF2&data=02%7C01%7CBailey%40access-board.gov%7Cc6d89b46797f49213b0808d7dbd36f2d%7Cfc6093f5e55e4f93b2cf26d0822201c9%7C0%7C0%7C637219573475022599&sdata=l675wkH%2FdAOI%2BCs1gXjVqCzW%2FzcR%2FBdfolGwFJ96iNs%3D&reserved=0>, the closest is PDF2 but that doesn’t seem to reference the definition directly.

Can anyone see an issue with the uses of “set of web pages” in these two SCs?

Kind regards,

-Alastair

--

www.nomensa.com<https://nam01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.nomensa.com%2F&data=02%7C01%7CBailey%40access-board.gov%7Cc6d89b46797f49213b0808d7dbd36f2d%7Cfc6093f5e55e4f93b2cf26d0822201c9%7C0%7C0%7C637219573475032595&sdata=DG90CpyZ1cu8b9ZXKK5ZbRRWtRJ4U5d%2FHjqqhuwLuVo%3D&reserved=0> / @alastc


--
​John Foliot | Principal Accessibility Strategist | W3C AC Representative
Deque Systems - Accessibility for Good
deque.com<http://deque.com/>




--
​John Foliot | Principal Accessibility Strategist | W3C AC Representative
Deque Systems - Accessibility for Good
deque.com<http://deque.com/>