Digital Publishing Interest Group Teleconference

25 Jan 2016


See also: IRC log


Markus Gylling, Dave Cramer (dauwe), Charles LaPierre, Ivan Herman, Romain Deltour, Peter, Krautzberger (pkra), Chris Maden (cmaden2), Ben De Meester, Tim Cole, Deborah Kaplan, Nick Barreto, Nick Ruffilo, Bill Kasdorf, Leonard Rosenthol (lrosenth), Vladimir Levantovski (Vlad), Bert Bos, Luc Audrain (laudrain), Daniel Weck, Tzviya Siegman
Phil Archer (phila)
Ayla Stein, Paul Belfanti, Michael Miller,


Markus: "We have a special guest - who is not here yet - so lets begin with two unannounced announcements. The first one is that we've had a change of chairs in the locators task force. Bill Kasdorf has been driving but has had to step off so Ben De Meester will be taking over. We also have a new member of the group, Chris Maden, from Un. of Illinois

Chris: "I'm a w3c old-timer that started with SGML working group back in '96 or so. It's great to be back on some of the communications. I'm here at the library of the university of illinois"

Markus: "Do you have specific interests with our group? Or are you a universal kind of guy?"

Chris: "I'm universal - but I will sign up for a sub-group. I'll do that with the other UIUC members of this group"

Markus: "You work close with Tim?"

Chris: "Yes, I work close with Tim and Ayla Stein"

Approving minutes from last week

<mgylling> http://www.w3.org/2016/01/11-dpub-minutes.html

Markus: "Any objections? Minutes approved"

The POE charter

<mgylling> https://www.w3.org/2016/poe/charter

Phil: "The link is above"

Markus: "Start with an introduction then we'll do questions"

Phil: "I'm Phil Archer and I followed Ivan when he started running Digital Publishing. Looking at the data side of things, I was looking at semantic web and anything that wasn't XML - so that's what I try to co-ordinate. Within that environment, I was thinking about open data and companies exchanging data. We talked also about licensing and rights statements. Mixing multiple datasets. How do you know what rights are linked with what data..."

<phila> ODRL Community group

...: "Many are familiar with ODRL - It seems no one writes it out but it's Open Digital Rights Language - it has been around for a long time and gets used extensively. If you are unfamiliar, there is a link posted in IRC. They are not on version 2.1 of an extensively developed way to express rights - although it is permissions and obligations. Those words get things political though."

<phila> https://www.w3.org/2016/poe/charter

...: "If you're in a publishing environment, there's a combination of different things you can do. There's the permissions and obligations working group. We could have called it ODRL group - and the final outcome will be ODRL, but we wanted to start with an open mind (although not a clean slate). The word licensing and rights both have difficulties around them which is why we chose permissions and obligations. It's about finidng a machine readable way to express permissions and responsibilities. "

...: "Why don't you just use creative commons? - Well - it's good but not good for every situation. There are cases where creative commons is not significantly detailed. Is this DRM - No. Will there be access control - no. There are key areas where we will not tread. Also rules languages. We are going to find a way to express the permissions and obligations. Anything that is built on top of that is out of scope for the group."

...: "Thomson Reuters will be the other chair of the group. The charter is under review - which will come out in 3 weeks, should start next week, possibly march."

Markus: "Questions?"

Bill: "In various sectors of the publishing ecosystem there are groups that are using ODRL as a framework for their own permissions. The IPTC has RightsML based on ODRL - another very interesting things led by Stewart Miles - is open source software based on RightsML based on ODRL. My concern is that this working group doesn't include lots of people who were in the community group. Is it your idea that most of what's in ODRL will still be present at the end of this process? The folks in the magazine industry who were going to do ODRL 2.1 backed off because they were concerned about what might come from this. -- Also - the new terminology is a great choice!"

Phil: "RightsML and that community - a very important one. Awkward fact, we've been trying to get this for years to make ODRL into a full recommendation track... Those interested weren't w3c members. To get around this, I've had meeting with the copyright hub in London and I have been told by them that they plan on joining W3C and participating. Caroline Boyd and others got involved with rightsML and others about joining. We want to work with that group - and we're hoping the copyright hub will be a good advocate and voice. We're trying our best to get those interested people involved as members"

...: "I'd be amazed if the end produced something very far from the existing ODRL."
... "I do need help getting people in the group - and thats part of why I'm here."

Markus: "In the dummies guide to ODRL - what would it say about the relation to RightsML and ODRL?"

Bill: "IPTC.... The Associated Press is using RightsML. ODRL is a way to provide rights, so it's being used as a framework which generates RightsML."

Phil: "So RightsML is an application profile of ODRL. Used heavily by the news industry - although magazines use it as well"

Bill: "The associated press is just ONE member, but the AP produces something like 250,000 assets per day (text and image) every single day. They have a fire-hose of content they need premissions with - so it needs to be machine processable."
...: "ALI - Access and Licensing Indicators - which is a manual lookup process that Scholarly is doing, on much different scale."
... "ALI relies heavily on humans reading the terms."

<laudrain> http://www.linkedcontentcoalition.org/phocadownload/framework/The%20LCC%20Entity%20Model%20v1.0.pdf

Luc: "The RDI project and the copyright hub is involved in this - as I see in the charter, that the goal of the WG is to the semantic model. I was wondering about what was done with the existing LCC model about having a high-level reference model for expressing rights for images, video, text. I'm not sure the difference of what will be covered by POE vs this model"

Phil: "Short answer: I don't know either until we start. My motivation for doing this is broad - census data, financial transaction records. There is not a large scope for variation. I see the LCC as important. How much is any side prepared to change? People are prepared to change if there is a reason to. One possible course of action is to say 'LCC entity model does it all - lets use that' and we'd need to have a good reason to say otherwise if that's the case. But we need to research and listen. We believe there are usecases that would require LCC to be extended. Until we get people around the room and we talk about use cases, we don't exactly know what is missing. We are aware of the importance of the work that has already been done and how it has been adopted."

Ivan: "Several more things. As part of history - we've had relations with IPTC - it was pretty instrumental for the creation of RDFa, for example. That might make it easier to build up an active contact. Phil knows better how it should work out. The fact that we have a co-chair from thomson reuters - that should help. One of the reasons TR joined was for this working group - so that shows how important is. "

Phil: "TR is already using ODRL. I think we're talking about tweaks here"

Ivan: "The current plan is - although Phil is the staff contact - I will probably join that group and lurk. One of the reasons I want to be there is as the link to that group and this group"
...: "One of the pushes that lead to the creation of the working group is this group - the fact that issue came up in the metadata task force is that it was part of our internal discussion. But we now have a committee with members. It also means that - PLEASE - get your AC rep to vote. We have to have a minimum number of votes to have a working group start. If you have an interest or think it is important for the community, please vote."

markus: "Get your AC reps to vote. And then - how can we help once you are up and running? How well covered are you in use-cases, patterns, etc."

<lrosenth> let’s be sure to also address the use cases that straddle Pub+Data - that will be the most interesting to resolve :)

Phil: "The charter went out for formal review 2 weeks ago. We've been talking about it at TPAC... The interest from this community has been strong. I'm pleased that a few have turned that into votes. I would imagine ODRL to be published as a working drafter fairly quickly, but the goal is to have data-centric use-cases. Ivan will be in the group - but we won't let him lurk. Until we start

writing use-cases, I won't know, but we will be knocking on your door saying: 'is your use case covered?'"

Use case and requirements for PWP

Markus: "this is a very fresh start-up to consolidate and cleanup a nice-and-polished notes with global use cases and requirements for Portable Web Publishing. We would follow up this week to see that things are starting to move - and questions from the group?"

<rdeltour> https://w3c.github.io/dpub-pwp-ucr/

<rdeltour> https://github.com/w3c/dpub-pwp-ucr/issues

Romain: "Not much happened, but here is what has. Ivan created a github repo with respec tools so now we have a document to start from. Here's the draft - it's fairly empty. We're using the issue tracker on github to discuss the use-cases, which is pretty empty. What I propose to do is to go over all the use-cases that are currently in the use-case directory on the WIKI and see what is relevant to this document and what should be skipped. Some are already defined elsewhere. Accessibility and personalization are also in another document."

Markus: "Not sure I would say that they are. Deborah?"

Tzviya: "We do have the accessibility taskforce - although we're still trying to figure out what to do as a use-case"

Ivan: "Personalization is important."

Romain: "Please continue and tell us status."
... "Do we try to go over all the use-cases - or do we want to set up a special call?"

Markus: "The wiki - in some areas is really nice, in other areas, not so nice. It would be good to have a few other people to help you understand the perspective on that - so use the list for it, or set up a dedicated call."

<rdeltour> https://github.com/w3c/dpub-pwp-ucr/issues/1

Romain: "I started an issue on the git-hub tracker. And tried to take a first guess on each use-case to see if it makes sense."
...: "I think going by github tracker might be less for the mailing list."

Markus: "You can use the list to let people know."

Leonard: "What does personalization mean in this context? Romain, I'm happy to clean up some of the use cases. Let me know how I can help."

Markus: "Personalization is a long discussion..."

Ivan: "What may help when it comes to the wiki page is that for each item on the issue list, you give a 1-sentence description of the use case. Sometimes the title is not enough. That will help us as the first set of filters.

Markus: "Is there anything else we can help with Romain?"

Romain: "What I can do is try to improve the issue I put in github. I can try to describe the higher-level sections and write the short descriptions - from there we can more easily filter"

CSS use cases

Markus: "No more questions for use-cases group. Moving on to #18 - CSS text - code name for effort to gather screenshots..."

Dave: "I've been spending some time this morning looking into the issue. Rather than figure out what's happening at the end of the process. I started playing around with this issue in InDesign - and what I found doesn't make me happy."
...: "Indesign doesn't try to align - it tries to left-align the 0 point of the glyph."

Markus: " Are the idiosyncracy of a tool what we're looking for?"

Dave: "So far, I don't think I've seen any examples of end products. A) I'm not sure how common a situation this is ... The few books I have with different tables, I couldn't see multiple fonts applied with same alignment context. "

Markus: " You did have Paul and Tzviya looking for examples."

Tzviya: "When I contacted my colleague - he said it sounded familiar. He scanned some 1974 texts with principles behind this. He basically explained that the issue isn't conceptually the alignment, but there was an understanding that went beyond my comprehension. The issue of bolding DOES come up - the issue of different fonts is a non-issue. In hard-typography, you could work out the solution."

Markus: "Are we confident that the scans from 1974 what publishers are doing today?"

Tzviya: "Do the elements range, or align? He was saying that the character doesn't move that much. Certain things should range with the decimal point, and certain things should align to the left. If a row of bolded characters appear - such as in an accounting book - it could show up. But, it is a non-issue from how they see it."

Dave: "A rendering system has to know where to put the character - and this is a question as to where we put the character."

Markus: "We should provide a set of screenshots and use cases and let the CSS working group decide if it's a non-issue or not."

Leonard: "Dave - if you wish to send me the file I can have the engineers take a look and give you a specific explanation as to what is done and why."

Bill: "Since indesign is the gold-standard for typography. If what indesign does is acceptable - then we're done - right?"
... "My suspicion is that the nuances of font-changing ... the industry finds those slight differentiations not important - so it may not be an issue."

Dave: "the engineers care - because the engineers want to know exactly what to do, even if the difference is slight."

<pkra> what's so bad about underspecified specs?

<pkra> *runs away and hides*

Markus: "We'll have examples from Wiley and Pearson - great."

Face to face meeting

Markus: "Tzviya, Ivan and I got together to look at this - but we kind of lost hope for a good time/spot. One somewhat good news in the context is that TPAC is pretty early this year - but that doesn't mean we should be without longer get togethers in the spring
... "What it's looking like right now is that we won't have a face-to-face in the spring - but we can have virtual face-to-face meetings. We can gather from our offices 1 by 1 or in small groups - over a video link and get together for something like 4 hours. It would be a virtual face-to-face devoted to 1 to 2 topics. Instead, we could spent more quality time on one single topic. We should try to plan that for April/May."

Tzviya: "We didn't talk about a specific date - especially when others are traveling very often."

<lrosenth> can we get a set of dates nailed down soon so we can all block off those days from other meetings?

Ivan: "There is IDPF conference which is also the CSS Face to face... That's in May... Then May 16 week is out because the iAnnonate conference and the annotation WG face 2 face. Those two weeks are out... Possibly last 2 weeks of may?"

<lrosenth> 23rd of May is ISO PDF meetings...

Luc: "Have you thought about having face-to-face in France? There is an epub summit in Bordeaux that is rather soon. It is organized by ediaLab. Great opportunity to do something there?"

Ivan: "I have to leave directly from there to another - which would make a F2F hard."
... "What we did discuss, which is something we can see - is that people can sit in their office - or smaller groups can meet up - especially if they are closer located. There might be some in France that could take a call from there."

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.144 (CVS log)
$Date: 2016/01/26 07:31:20 $