Digital Publishing Interest Group Teleconference

<mgylling> https://www.w3.org/2016/04/18-DPUB-minutes.html

Markus: "Minutes from last week: Any comments?"

Publishing the accessibility note

<mgylling> http://w3c.github.io/dpub-accessibility/

Markus: "Charles and Deborah - the accessability note - which has been worked on for quite some time, and people should review & comment - hopefully today we can give the green-thumbs up with consensus to publish as a W3C note. There was some activity in the issue tracker. Charles - quick status update?"

Charles: "We are ready to publish unless someone has an issue. Issues have been resolved as of this morning, there was some weekend activity. George Kersher added a section for PDF UA, which was some comments from Leonard. We had a great meeting with Leonard and George on Friday. That got resolved. A lot of little cleanup issues from Ivan. It was great to have his feedback, so we're good

to go unless someone here has issues."

<clapierre> http://w3c.github.io/dpub-accessibility/

David McDonald: "In the MathML - what it's saying here is that there is not a solution here, but there is not a braille and screen-readers?"

Charles: "Yes, pretty much. The problem is that not everyone has implemented it. That's why peter is leading the charge here and trying to get a consesus... Publisher are actually commenting out MathML and providing an image. They have to do that because the reading systems will crash. Publishers want to sell books. There is no real good solution. We're trying to get together. We get

asked what to put in our books - and we're not quite at the point where we can tell people to put in MathML in... We're not there yet - so we're pointing to the new community group."

David: "So people are aware on what's going on with the Math communities, right?"

Charles: "yes, we have had the top experts in the world together, things like NVDA - and yes want to see that move forward."

Markus: " Before we go to Peter, I pinged Shane - with some final language. He'll be here soon. Peter..."

Peter: "I wanted to add to what charles said - there are 2 problems. Accessibility of MathML which is complicated. The 2nd problem is that there is no way to solve accessibility rendering without MathML - which isn't really supported by browsers... So that's one aspect worth stressing. Things like JAWS, mathplayer, etc... They all only work, and are exclusively tied to MathML - we're stuck

in a world where we cannot render mathematics and be accessible."

Markus: "Ivan - do we do a call for concensus or do we do an email?"

Ivan: "here is fine. Then we have to set all the practicalities. "
... "We must agree on the short-name 'DPUB Accessibility'?"

<pkra> https://github.com/w3c/dpub-accessibility/pull/35/

<pkra> proposed a fix

<mgylling> +1

<pkra> +1

<HeatherF> +1

<dkaplan3> +1

<TimCole> +1

<clapierre> +1

<ivan> +1

Markus: "If you have any objections on moving forward with the shortname 'DPUB-Accessibility' then speak up now. In Favor do a +1"

+1

<nickbarreto> +1

<astearns> +1

<Bill_Kasdorf> +1

David: "Probably better to speak for WCAG conformance..."
... "Conformance"

<ShaneM> +1

Markus: "Ok - so that's a quick/easy fix. With that change, we have agreement to move on."

Ivan: "Practically, we can ask for authorization of the short name. Charles/Deborah - do you have any experience with publishing to W3C?"

Charles: "No."

<ivan> https://www.w3.org/2005/07/pubrules

<ShaneM> Note that you need to check a static version - not a raw ReSpec source

Ivan: "You will have to - or I can - you need to go to a pub rules checker, which I will sign. You can use that to check your HTML, etc. Link above. This will give you the first set of checks - you fix those issues, but it'll look at dangling links, namespaces, etc. Before you do that, you'll need to create a static version from the "ReSpec" button which will give you a save-as and you'll

save it in HTML. That is what you have to check. That is what we publish. Not the respec, but the fixed/static HTML."
...: Publication dates are Tuesdays and Thursdays. We don't need to rush, so lets set the publication date as May 3rd. When these steps are done, you ping me and we take care of the rest. What would be good is that if I have text that you're happy to give out by friday/Monday. The webmaster will publish it Tuesday. "

<ShaneM> feel free to ping me too - timezones

<ivan> http://w3c.github.io/dpub-accessibility/index.html?specStatus=IG-NOTE
...: "Nothing else, as you dont' have images or other files, just one single HTML file. You can even see how it will look if you see the link pasted above."

Markus: "Charles - do you know what to do next?"

Leonard: "I'll email charles - I noticed a minor editorial - section 3 is incorrectly headered."

Markus: "Before we move on - Charles/Deborah, do you know what you're doing next?"

<HeatherF> Yay!

Charles: "We have some use-cases that we're going to be starting to collate, and we're going to work on that, we have a lot of use cases, so heather get ready. We'll see based on our note some of the things - and we'll see what makes sense after that."

Markus: "No more questions? Lets move on. Back to original agenda item - HTML5 and notes. Shane/David, please introduce yourselves."

Notes in HTML

Shane: "I'm Shane and I've been involved win W3C for 20 years... I've been most recently involved with payments & accessibilities, which are new. I've been doing alot of work with DPUB and working with Tzviya and Markus. Through that we developed requirements around notes. I'm with SpecOps and we do standards, so that's what I'm doing. I believe I've always been a member."

David: "I'm david, and I've been on WCAG group since 2002, so I've been on for a while. I work on accessibility helping companies."

Markus: "Through the DPUB work you started working on notes..."

<mgylling> http://spec-ops.github.io/html-note/index.html

Shane: "The DPUB accessibility taskforce had been talking about various roles and how to make those roles and semantics available to eletronic publications. Those roles include things about footnotes and endnotes - there was a lot of discussion around those roles and making them well supported. That discussions migrated into the HTML accessibility taskforce - which is the overlap with HTML5 taskforce. We've talked about how there is bad support for notes and the navigability of notes within accessibility. David has been the champion for this project, so that all started to jell and I came to the DPUB accessibility group and we gathered requirements. The accessibility taskforce (HTML) brought it up quite a few times, so I revived it and reached out to David/Markus/Ivan and lets think about it more and develop a spec."
...: "I've come here because the spec doesn't really have a home. An IG doesn't really do specs, but you are the most affected by such a spec, so I wanted to see if there was interest from the group, to get some extended custom elements where we could get testing and such."

David: "The big thing is that you... Every academic institution that makes items uses word and adds footnotes - it's ubiquitious within papers. Look online and there are no standard way to do footnotes, and it's done differently by publishers. It hasn't really been standardized. As a blind person, you cannot easily navigate. Why could word do it? It was a dedicated element - the person could just click on it and it worked. If we could do this with browsers/languages, then footnotes will be used. This is the whole plan. There are complications in footnotes (multiple references) and we have many of the complications addressed in the standard we're writing as well as ways to address them all. We want anyone who uses a screen reader to be able to access these footnoes."

Shane: "I agree that what we're trying to do allows people with assistive technology be able to access - but i hope that's an awesome side-affect for people who are using it, but there are limits beyond that. "

Markus: "For example, maybe this is for Bill Kasdorf - it's great for publishing workflows to have better workflow for this."

<ShaneM> Notes - it's not just for scholarly literature any more! (tm)

Bill: "It's not just for scholarly publishing. I like the analogy with Word's functionality. In Word - there's a distinction between footnotes/endnotes, and the fact that foot is in the word, doesn't matter."

<Bill_Kasdorf> Just clarifying that I'm making a distinction that DEPARTS from Word's functionality, that there is NO distinction in this case between footnotes and endnotes

tim: "General - What are the use cases here, especially in terms of Annotations. This is really about authors/publishers being able to add in an easy/accessible way notes. Is that the case here? 1) is that right? 2) I was a little surprised by the detail level - the fact that the note-ref in common use-cases will be empty - which raises the issue - what is the note attached to? In your samples, you show two that have ref content, and one that doesn't - are they attached to a place in the document, or a place in the document?"

<ShaneM> http://spec-ops.github.io/html-note/index.html#examples

Shane: "Your analysis is correct - that is the direction we're trying to go. With regards to the second question - the thought is that an empty note... What content is rendered - will be supplied by the browser. So if I have an empty note ref, the user-agent would supply a "1" or whatever - a link to that that could be navigated to. I'm letting the system provide the footnote number. When it says in common use-cases, the note-ref would be empty. The system is providing the annotation to the note."

Tim: "So it's from a standpoint of the user - in classic traditional publishing, you need to figure out what the '1' is referring to...'

Shane: "We're allowing for specific referencing - as someone noted in footnotes/endnotes. We do have a 'type' of notes - so you could have multiple threads of notes embedded in your document. I've seen 3 different types of notes, with different numbering. all in the same doc. There was original footnotes, then there were annotations - which was a different type of note. "

Bill: "Yes - typing is the right solution to that."

<ShaneM> @group identifies the group in which a note is collected

Ivan: "I did not go into all the details - but I am curious about - what part of the browser behavior is specified and what is not. For example, you set - somebody said - that you would have a spec-reference from a note, to whatever it referred to. Is it part of the spec that user-agents have to provide something? Is that a specific implementation that has to do that in Javascript? what's the borderline between purely HTML elements and the behaviour?"

Shane: "I have an answer - but not sure I like it. It all depends on where we go with it. If we try to define this as an HTML5 extension, then there is a clear way to define requirements on user-agents. We can do that. If thats the course we take, I would do that. I would define the requirements for rendering/behaviour. On the other hand - if we implement as a stand alone spec as custom elements - which we can do with HTML - then I would think it is part of the implementation. Everyone uses the same one, etc..."

Ivan: "I'm official confused - what are the two ways we can go exactly? I don't always follows the HTML working groups."

Shane: "We could - as a group - I'm don't have an ask yet. Just telling you we're doing this and see if you want to help. One course is to propose an extension of HTML5 - you need to get the browsers to provide feedback - and they'll have to implement. Similar to <longdesc> - but it becomes and HTML5 tag..."
...: "Another option is to create an extension spec - so we provide an implementation as custom elements. The way you do custom elements is that you put in a link at the top of the document and it will load in the implementaiton."

Ivan: "When you do the 2nd is it an intermediary for doing the first?"

<ShaneM> https://www.w3.org/TR/2016/WD-custom-elements-20160226/

Shane: "Possibly, but not required."

David: "If we want something easy to get-to and do, we hope it will be an element they can pop into their code. We hope it's an element that gets added."

Shane: "With extensions - there's always a 'who is going to use that, and where is it going to be used."

Markus: "When it comes to option 1, have you had any response yet? Do you have a feeling for the viability of that route?"

Shane: "My preference would be to attempt route 1. If the feedback is that we need more evidence - it's heavyweight thing to implement - and they want a demonstration of demand - then having option 2 might be a good start. No, we don't have negative feedback yet. We have positive feedback from Yandex (sp) but they just use webkit or chrome and change a few things."

Ivan: "The other question I have is that obviously, when used, this is bound to use with CSS - or all kinds of things. Have you checked whether what you expect to do with notes - is that CSS can do? One thing this group has fought with - is having proper pagination done in CSS. One thing with footnotes, is that you have a paginated document where the footnotes are in the feet. But you need to have the right CSS statement. Is this something you have talked to the CSS group about?"

Shane: "We have not yet talked to them. The idea was that nothing would require changes to CSS. We talked about pagination, but we wanted not to require CSS changes. We were talking about how you could use a CSS selector for print... If you're not using the media-selector, I could think of ways to have footnotes appear on the screen, but it would not be the default presentation/implementation."

<ShaneM> Yandex Browser is based upon Chromium

Ivan: "Pagination is something that will be used generally once it's working..."

Shane: "Can we explore this deeper? Lets say that you wanted 'footnotes' on a screen-based view of something. If you've defined a region of your screen/viewport. That is the place that footnotes should display. As the user scrolling through the document, is it your view that a region at the bottom of the page could show relevant footnotes and grow/shrink as necessary... Could Flexbox be a solution for this?"

Shane: "If you look at the spec, there are a number of open issues, and I expect there to be more. Questions like yours are perfect to add here."

Leonard: "I was just going to say that these need to be called out as specific items for discussion. While the user-experience that a given author may want, there are different ways that content can be displayed. We may indeed find that there are missing pieces."

Shane: "Agreed. My inclination is to have as little of basic user-experience. The basic is the simplest case, and anything more complex is up to the author."

Ivan: "I try to imagine the reaction of HTML people - Why do you need the 'notes' element, isn't <aside> enough?"

Shane: "We've had that discussion. The answer is - generally - aside has different semantics. 2) note-ref isn't the interesting part, it's note and note-ref, so you need a way to reference those notes or a way to reference. Aside cannot do that. It doesn't let you have multiple reference from the same note, or to group them. "

Ivan: "I was looking at Aside being the same as the note..."

Shane: "I can specify my notes as an aside, but i cannot reference them and get the behaviour I was looking for."

<Bill_Kasdorf> +1 on the behavioral distinction between <aside> and <note> is critical

<ShaneM> also note with its "group" semantic isn't achieveable with aside.

David: "Is anyone doing that? Aside is not being used at all for footnotes. We're looking at - footnotes is such a common thing - and there is no uniform thing that it has been done. We're trying to rally around something. It's in nearly every academic document. Trying to repurpose something that wasn't especially created for notes - we need something specific for footnotes."

Leonard: "How do you see note related to references? Bibliographic, normative references... While the semantics are similar, how do you see that. Could note be used?"

David: "I'll speak for myself - if I see a little 1 and it's superscript, that is what note would be used for. If you're using that for bibliographic information, it would be used for it."

Shane: "We absolutely COULD do what you suggested. In that case, I would have those references handled differently - so notes in that group would be rendered differently. It would be the bibliographic group and things would handle differently. We have type and group - in this can I mean group - not type. "
...: "It's the same as 'type' on an ordered list. It specifies the 'type' in the numbers. Group attribute would be 'footnote' vs 'endnote'"

<ShaneM> group — Name of the note group

<ShaneM> value — Ordinal value of the note

<ShaneM> type — The type of the note markery

David: " Goal is to make it simple, but very rich. If you're a publisher, you can use the attributes to dig deep and get what you need."

Bill: "Thank you for the distinction between group and type - there are many co-existing types of these things in a document. A bibliographic reference, an end-note, and footnote, etc. But mainly when addressing the issue of asides. If they are all asides, you have a complex job of determining what is an aside and what is a note."

<ShaneM> <note group="bibliography" id="bib-1" title="DPUB-A11Y">The Dpub A11Y reference</note>

Bill: " There can be a paper with 100s of these things with different types"

Markus: " You were saying this doesn't have a home right now, and you wanted to gauge interest on the publishing side... I think it's safe to say nearly all publishers would want to see declaritive items for notes. One question is priority - it's something we don't really have clear. Especially other things we have listed in the IG. So that is something we need to think about as well in the

IG here, as well as other venues that come into play. There are lots of people who would love to work on this. What then would they not be working on...."

markus: "Lets say we should - if I understood you right - Shane - are you unclear about the next steps? Do you want feedback?"

Shane: "I welcome any feedback - but I don't think we know where to go next. I don't have a sense of priority."

Markus: "The document now lives on SpecOps. Would it make sense to try to bring it in under a W3C umbrella? Organizationally are we good or would you like to see an IG/CG/WG take it on?"

Shane: "It might be too early. We don't care where it lives... There isn't a WG that exists that could take it on. if it is going to move in, I'd put it into the web-platform group. The incubator group. That's where things are supposed to come into the W3C these days..."

<ShaneM> https://www.w3.org/community/wicg/

Markus: "Those are questions we don't need to answer just yet, but we need to see if the publishing industry will use this. Your requirements/markup looks good. We can take up with the IG to have tons of people look at this and comment with our publishing industry hat on. It will not just include rendering - lots of publishers look at HTML as an authoring platform. Which may or may not be valid use-cases for HTML. We can take that action to make sure that we spread the word around and dedicate time to that - to make sure it gets proper analysis and readthrough."

Shane: "Please review and provide feedback. In terms of other things, I'm talking with other people in the space - to get people who are interested - so we can put it into the web incubator. I want a heavy hitter on the list of editors (someone whos' from a browser) if we're going to go that route."

Ivan: "My feeling is that - having seen from far away - how these things work. Having an implementation as custom elements would make things easier. Because if you have that, you have proof of existence that it can be done, implemented, etc. You did it as a JS without any issues. That can then be used to make it a proper HTML extension.

Markus: "Ok, we're over time. David & Shane, thank you. For the rest of us, we will be talking about Annotations. And we missed some agenda items, but none critical. Lets make sure to do those next week instead. Thank you."