Digital Publishing Interest Group Teleconference

<trackbot> Date: 08 June 2015

<tzviya> agenda: https://lists.w3.org/Archives/Public/public-digipub-ig/2015Jun/0030.html

Tzviya: "First, lets approve the minutes. Any comments?"

<tzviya> http://www.w3.org/2015/06/01-dpub-minutes.html
...: "We did a recap of the face-to-face and talked about functional requirements for Packaging. Minutes approved"

Requirement for pagination

<tzviya> https://www.w3.org/dpub/IG/wiki/Functional_Requirements_for_Pagination
...: "Next item: Dave has put together a draft of the functional requirements for packaging. Based on Latin-req and pagination."

Dave: "Calling it a draft is a bit much, but it's a start. I tried to broadly categorize things so we had a list without 50 members on it. Basics: control margins, orientation, page display. How we can do rendering, pages, and some typography issues. I know even some very large corporations that should be members are interested in these sorts of enhancements for their ebooks. A section on interacting with pages - copy/pasting, Brady's requirements. I'm sure there are things that are missing. I would love other people's input. How much detail do we need? Relationship with Project Houdini..."

...: "There are like 10 links in the chain - requirements, the new APIs that browsers would have to expose... I'm technically unqualified to get the whole way there. "

Tzviya: "The point of the document is to get our ideas down, so that we can have that conversation. So we can sit down with project Houdini and say: 'these are our requirements, please tell us what's missing."
... "We need to know what is missing - so what about pagination would you like to see?"

Alan: "One thing I don't see on the list (not sure it should be there) is some greater control over line-breaks for a particular representation. We have pagination requirements about fitting things on a particular page. Dave - you use HTML as part of your workflow - and you'"

<tzviya> +1

<mgylling> +1
..: "You're printing to a particular medium/width, do you have the ability to set linebreaks, or is that not something we should be worrying about."

Dave: "This is currently a major issue with our print workflow."

Alan: "Until we have everything automated, we will need greater control over how things are rendered."

Dave: "I'd love input on what those features might be to control. Such as if you're in indesign and trying to set a line-break. your paragraph will continue to wrap apropriately."

Markus: "One thing I don't see as well is the notion of page set layouts - in other words, you start with a viewport with 1, 2, ... N pages. how many do you want to show? What is the visual and logical flow? Latin-req talks about facing page alignment and adjustment, but that is just stitching between things that are together."

Dave: "I grouped that as part of 'Page Display'"
...: "Some of the CSS covers whether the pages are laid out vertically..."

Tzviya: "We can call it REGION not just pages, so it covers things like comic books, and still covers pages."

Dave: "I'm worried things getting confusing if we get too generic. Especially since regions can be non-rectangular and have more complications."

Tzviya: "it could be a separate item. We could talk about the regions being non-rectangular."

Dave: "All sorts of non-rectangular things will be on the pages. Are we worrying about non-rectangular pages? We can make note of that, but I think given right now that our viewports and our pages are rectangular..."

Luc: "I was wondering what we've been doing with typesetting and templates. A title page might be different from following pages. "
... "My 2nd remark is typically digital. We have this case when we can make things appear in popups. It introduces a 3rd dimension. This introduces areas that appear above the page. Is it something we should consider especially if we'd like to manage the space where these popups could appear?"

<dauwhe> popups, overlays, alternate renditions...

Dave: "On the 2nd part, other parts of CSS. Other parts of CSS support layering, stacking, the index. That sort of content would need to be fragmented. As far as describing the types of content/effects with those kinds of books. It feels like something for our 22nd century counterparts to work on."

Tzviya: "Brady/Alan, do you have any comments?"

<astearns> Not really

Dave: "As far as the earlier note on templates. There are some rudimentary ways that you can note in a given way what context the page has been rendered. It is not as complex as some of the proposed ways we want to use it."

Ivan: "My questions are more general: Should I understand it that the first few sections are actually requirements we would like to see as CSS properties - new or existing. Whereas the 2nd part is the runtime DOM access side of it. Or is this whole thing a giant Javascript with all the necessary information."

Dave: "I was broadly thinking of it in the way you mentioned. How that is achieved in the near future is probably through custom CSS properties if houdini brings us fruit."

Ivan: "It may be worth making that clearer on the page itself."

Dave: "I'm also a little wary of putting that sort of info - that it's about implementation rather than..."

<laudrain> +1

Ivan: "We wouldn't like the whole of the pagination issue to be a giant javascript - aside from what CSS does in the traditional way - so it's a general requirement. As much as it is humanly possible..."

Dave: "Given that there needs to be end-user control for some of these issues. Such as reading systems letting users pick font sizes. At least one potential use for that is a "user stylesheet"."

Ivan: "Even our core constituency. Editors and publishers - we do not want to force them to become publishers."

<Ralph> Nick: the Apple watch shows that there is investment in non-rectangular displays

<Ralph> Tzviya: we need to think about whether that is a Digital Publishing concern

<Ralph> Nick: right; we shouldn't reinvent the wheel but let's not assume the world is rectangular

<Ralph> Tzviya: that's why I mention "regions" generally

dave: "Back to the issue of non-rectangular display - for the most part my apple watch IS a rectangle. If we do even 1/2 of what's on this list, it's an accomplishment. I'm worried about putting efforts into things like non-rectangular when there are foundational items."

+1

Brady: "I wanted to bring up Javascript VS CSS. I don't think we know where all of this is going to live. It'd be great for CSS to solve all the problems, but I don't think that's necessarily likely. A lot of pagination will happen in custom javascript. What Houdini is doing is bringing us the ability to do this. Not sure how much is CSS vs JS, but the medium-term will be heavily JS."

Dave: "Hopefully some of that JS can be controlled by custom CSS"

Brady: "As a RS implementer, we sometimes get CSS that is handled by JS. At the end of the day, it's JS, even if the publishers aren't writing it."

Ivan: "This is a set of requirements. it would be important from our point of view - to express that we'd like to be as far away from Javascript as humanly possible. The CSS working group may not completely fulfill it, but there should be a clear case that 'this is the way it is, but we don't like it."

Brady: "If you don't understand that you'll be needing Javascript to do this, then you'll design the spec for CSS, which puts Javascript implementation at a benefit."

Brady: "What i'd like is a tag <book> that lets me miraculously have an ebook reader for my content, but that won't happen, but at least for now, having declarations is a good start. What I really need is enough CSS/JS to allow me to implement this."

<Bill_Kasdorf> a reminder that there are many types of publications that aren't <book>s

Tzviya: "So - this boils down to 'how do we define a package' It would be a very good to put this down in the packaging spec."

Dave: "I have thoughts, so we'll talk."

Tzviya: "Global /book/publiciation/i"
... "Fill free to fill in additional information on this spec, then lets get in touch with Houdini people, and wave our magical wands. Vini Vidi VICI!"

Scholarly publishing

...: "Next item - discussion - Ivan and I discussing whether scholarly publishing is properly addressed in our use-cases. "

<tzviya> http://www.w3.org/dpub/IG/wiki/UseCase_Directory

...: "Many of us are involved in scholarly publishing, which is more about articles and journals (which are collections of articles). There are not focuses on a 'book' or retail channels, most are about publishing through peer-review journals and peer-review. We're not working with Kindles, apples, etc of the world. We're talking about wiley, elsiver, etc."

Bill: "One thing that makes this more urgent is that this content - most scholarly journal content is online."

Tzviya: "A majority of it is online as PDFs. Many have done away with print completely."

Bill: "I think we want schoarly publishing to say: "Lets use dpub instead of PDF.""
...: "The fact that we view scholarly publishing as multi-column isn't really part of what it means to be a journal. I don't see it as being that different from any other pagination issue we have. Our use-cases already addresses multi-column."

Tzviya: "I think we'll find that a lot of our use-cases cover this already."

<HeatherF> Why two columns exist: http://en.wikipedia.org/wiki/Column_%28typography%29

Ivan: "There is a category that is closer to books, but part of scholarly publishing. Those are conference proceedings. Depends on the area. Conferences and workshops, proceedings of some of the big conferences have the same weight as scholarly journals. They are collections of journals, but often published. Not sure I agree with that the issue with columns isn't necessarily about saving paper. For usability, having shorter lines, the document is more readable and better to understand the content. On the side, we have discussions on whether we would make new formats for recommendations for the future. Whether the size of the lines on a W3C recommendation should be a certain size. The 2-column is a disaster if you have to make reviews."

..: "There is one aspect that is very important for Scholarly publishing. More and more scholarly means scientific data. So our package needs to include scientific data - such as fragment identifier, which needs to accomodate the many types of data. There may be a limit where the data is just dumped on the web, and other times where it is part of the publication. When working on astronomy, they have giganitic images that they would like to include. "

<tzviya> we need to build out use cases like http://www.w3.org/dpub/IG/wiki/Publication_with_Date
...: "You get data with different interactive ways of working with the data. There are many online publications that have this feature. Epub is very defensive when it comes to Javascript. I may be imagining that we have to interact with Data in different ways means that the kind of javascripts that we have to allow will be difficult to control."
... "That may have consequences in how we interact with the data."

Bill: "My comment is that the issue of double-columns and "shorter lines" do not need to be related to columns themselves. "

<mgylling> /me it is true…

Tzviya: "If anyone has more information, please add to the use cases"
... "Tables are an issue - it's difficult to present tables on small (or even large) screens. Whether they should be tables,... We can lecture, but some people will do tables anyway."

<tzviya> it would be helpful if someone could address the relationship between DPUB & annotation with respect to scholarly

<tzviya> RDA

<tzviya> NLM

<tzviya> JATS

<tzviya> http://jats.nlm.nih.gov/

Ivan: "I'm just raising one more thing - Not sure if it effects anything up to now - when scholarly articles are published, and are stable, then they are usually - and very often - they are looked up on various search engines & crawlers because they want to have cross references managed. See if the author has references to their publications. Alot of problems that exist in traditional book publishers. My impression is that the community is more active than traditional book publishers. In the case of a scientific publication, the fact that data is made available in search engines seems more often. It's a general point. The other question I have is that the whole workflow for generating the publications, in many respects, is very different. Not sure how it effects the things we've discussed. They seem to have their own XML specification for their items. The items are bound to that type of XML flow."

<HeatherF> and soon-ish, STS, which is a tag suite for standards based on JATS
...: "I don't know whether there are actually anything done to combine that flow with epub3 today - or any kind of epub-like package. Does that community ignore that whole aspect? that is important for us to know."

Tzviya: "Each company does it their own way - as there is a requirement to deliver the papers. It would be good to type up the requirements for things like peer review. One thing we see more in scholarly publishing is the use of semantic web - embedded triple set is much more common in scholarly publishing. Not necessarily something we need to address, but something we have discussed when talking about Metadata.

<HeatherF> Apologies; I need to drop off.

Bill: "An observation - there's a 'BITS' that has - to some extent - removed tables in HTML. It used to accommodate tex, but now it supports MathML. My question is - do we have computable math? Run a calculation on a dataset."

Tzviya: "Computational MathML is what you're referring to I believe."

Markus: "What's interesting for me, How can we get closer to that community - which seems to be very disjoint from traditional book publishing. We have to ensure both these communities are on board. And so I am not absolutely sure how to do it, but it's necessary."

Tzviya: "Who is involved with the scholarly community?"

Tim: "I also wanted to say that - the textbook community is between the scholarly publishing and traditional publishing. "

Nick: "Good for education/outreach"

Tzviya: "If you have use-cases, please add them. What seems to be happening with scholarly publishing is that each organization is doing it their own way."

Markus: "EPUB is not part of the scholarly world today. Is it easy to answer the question Why?"

Tzviya: "I can tell you why epub isn't - epub for scholarly - the PDF is comfortable. At the same time, publishers are still publishing PDFs, they are going ahead with HTML. They have PDF and HTML, so there is a "why do we need epub" as far as a package for things like datasets, where does a large database that is a source-file for content fit?"

Markus: "If scholarly is going to HTML, then there is no fidelity issue anymore, no?"

Tzviya: "That is not as much of an issue for most of the scholarly publishing world."

<TimCole> +1 for Ivan - mathematicians being picky about how their math looks

Ivan: "There is one area where pixel-perfect is important - mathematic. Where authors themselves - mainly mathematicians - are extremely picky on how these things look. That's why, in that part of the world, LaTex rules. The main issue is a perception rather than technical reason. There is a perception that 'epub is for books'. It isn't technical, just a social thing that has to be changed."
... "As nick said, there is an aspect of education that is required. The idea of scholarly results and articles being online - this is something that the scholarly world has proceeded any book going online. In the 90s, what really happened was that the journals were printed, as a collection of articles, but authors did their best to have their papers on the web. The only tool back then was PDF, and that's what stuck. Ever since there was a web, we had PDF for them to publish with. As a first step, journals became online. Subscriptions for jounrals are so high for printed journals, so the online journals because very useful. "

...: "It's only lately that you have primarily HTML journals. Just out of coincidence, there is a journal article that I'm co-author of, and it happens to be in HTML. Then we get into the epub-web, and when I wanted an offline version, they could only give PDF, because they didn't know about epub or ebub-web.

Bye!