Publishing Working Group Telco — Minutes

Date: 2018-06-04

See also the Agenda and the IRC Log

Attendees

Present: Wolfgang Schindler, Garth Conboy, Teenya Franklin, Leonard Rosenthol, George Kerscher, Hadrien Gardeur, Toshiaki Koike, Ric Wright, Vladimir Levantovsky, Nick Ruffilo, Luc Audrain, Derek Jackson, Ben Walters, Jun Gamou, Caitlin Gebhard, Ivan Herman, Benjamin Young, Jeff Buehler, romain, Dave Cramer, Ben Dugas, Joshua Pyle, Marisa DeMeglio, Laurent Le Meur, Tim Cole, Ben Schroeter, Brady Duga, Franco Alvarado, Adam Sisco, Bill Kasdorf

Regrets: Chris Maden, Jasmine Mulliken, Peter Krautzberger, Tzviya Siegman

Guests:

Chair: Garth Conboy

Scribe(s): Nick Ruffilo, Benjamin Young

Content:

Garth Conboy: Lets get started - Tzviya sent out an agenda, but lets approve the minutes from he last meeting first - for the pre F2F meeting

Garth Conboy: https://www.w3.org/publishing/groups/publ-wg/Meetings/Minutes/2018/2018-05-21-pwg.html

Resolution #1: pre-f2f minutes approved

1. F2F review

Garth Conboy: https://www.w3.org/publishing/groups/publ-wg/Meetings/Minutes/2018/2018-05-30-pwg.html

Garth Conboy: https://www.w3.org/publishing/groups/publ-wg/Meetings/Minutes/2018/2018-05-31-pwg.html#option2

Garth Conboy: https://www.w3.org/publishing/groups/publ-wg/Meetings/Minutes/2018/2018-05-31-pwg.html

Garth Conboy: Post the agenda being sent - there were minutes from the 2-days of F2F minutes.
… Ivan do we want to let these sit?

Ivan Herman: We have to give a week for people to read and review and object if they wish.

Garth Conboy: The F2F minutes should be read though - there’s also a nice photo of us all. Kobo hosted and we got an awful lot done. It’s worth reviewing the minutes. Proposed are in blue, others are in red. On the agenda, tzviya had put lots in there. The best thing to do is to read through the minutes and the resolutions. We had lots of discussion about the infosets and boundaries around items. And we have items about defining the bounds of the publications.
… and the bounds for the purposes of off lining, packaging and search, at least the top-level was agreed to. There was another major resolution on the first day which was leaning towards JSON LD serialization of the infoset and using schema.org
… I felt the meetings went well - we spent time on affordances, reading order, there was a resolution the 2nd day that was - that is worth looking it. Resolution #3 on the 2nd day which had to do with reading order VS TOS. Clearly I wrote it, as it’s not something Matt would have written:-) A table of contents should be present, it should be a doc-toc role. It could be elsewhere - but the could be elsewhere might be open to further discussion.
… we spent a lot of time on serialization for reading order and list of resources
… there was a resolution from list of things in a reading order - string or object. Object would refer to mime type.

Garth Conboy: anyone else want to chime in on the meeting high points?

Ivan Herman: one thing that we did also was go through the affordances on the second day
… we actually assigned each affordance to one or more people
… these are to be done as Pull Requests or whatever works
… but these are to be done quickly
… we also looked into schema.org mapping…which we may discuss more today
… another thing, was I think dauwhe suggested we make a list of things to add to schema.org
… so we can get those to them for review
… we also had a long discussion on PWP vs. EPUB4 vs. WP
… where we don’t think there’s any real difference between the required infoset for WP and PWP
… so if there are no real differences, then PWP is just the determination of packaging format to put a WP in
… that being said, a future EPUB4 may use a more defined packaging format
… though we have not yet stated it would be identical to EPUB3’s
… but it will be discussed in the future
… on timing, we do want to have a new document by the end of the month
… so we need those affordances submitted within the next 2 weeks
… Matt is out for at least this week
… and when he returns we’ll have to sort out how to get all these changes into the document

Garth Conboy: Thank you Ivan, the WP, PWP, epub4 was on my list to mention. That did not turn out to be an actual resolution but certainly it seemed to be consensus in the room that we continue to focus on WP, lower the focus on PWP - at least for the moment, figuring that PWP was just a package for WP - so if a packaging format comes up with traction, we can use that - possibly packaged as zip - but the restrictions we apply in epub 3.2 land around mimetype - the various restrictions could be relaxed. It’s unlikely it would be XML but JSON/JSON LD for the infoset…
… The action items are recorded in RED - largely in the minutes, there are things we’re going to expect Matt to turn into spec pros. All of the affordances that Ivan had mentioned had owners. There is a template there for those to be addressed which will be done in the issues or as a pull request in the spec. We spent the last part closing a bunch of issues, which are in the minute and hopefully Matt can turn that into spec-eze…

2. schema.org mapping

Garth Conboy: https://github.com/w3c/wpub/wiki/Descriptive-Infoset-Properties-vs.-Schema.org-table

Garth Conboy: the next (and last) thing on the short agenda #5. How we’re looking to map our infoset items to Schema.org mappings. We have the right-most column reflects the discussions we had at the F2F and there are some of these that are to be discussed from the Schema.org perspective - lots of these we need for our infoset and this table has the initial discussion of the mappings. I believe this was yours initially Ivan - thoughts?

Ivan Herman: https://github.com/w3c/wpub/wiki/Minimal-WPUB

Ivan Herman: https://github.com/w3c/wpub/wiki/Minimal-WPUB-for-a-scholarly-paper-(of-sort)

Ivan Herman: What’s worth mentioning there are two pages that are worth reviewing as well. A wiki page (posted above) which shows a minimal manifest structure that we agreed on for a simple web publication. Something which has only a few chapters (3) and that’s it. The details are not yet absolutely defined. What I did today was to do something a bit more elaborate (also pasted above). I originally wanted to take a scholarly paper but you would have tons of copyright issue, so i took a w3c recommendation that was published a few years ago and thought about how i would turn it into a publication using the schema.org mapping. I think it was Josh who said during the meeting is that what he wanted was to add just a few lines that would turn it into a web publication. And then I made a more elaborate version using schema.org that is more expansive. it’s up to the author to decide what they want. Knowing that schema.org allows even more data to be added because it’s pretty big. They also revealed some problems. Some of us had a discussion about a half an hour ago.
… The reason why it is tricky - and why I wanted it tricky - is that this is a recommendation that uses an external HTML5 document, which is the target for a long-description of an image. There is also a diff file comparing the recommendation to the previous version. It’s referred to in an HREF but it’s not in the reading order.

Garth Conboy: There is no default reading order/spine because everything is in this file?

Ivan Herman: That’s correct.
… There is no title or name in the manifest because - what we said in the document - the title element of the HTML plays that role. And there are CSV and XSL files that are there which are again part of the publication, so if I want to cache the whole thing, I want those to be cached as well - which is why they are there. The big discussion in Toronto is what happens to the CSS and other files that are in the recommendation to render it. They are not listed in the resources in what epub would require and that’s something we have to discuss. That’s still one of the contentions.
… It shows the direction - i believe - where we are going.

Garth Conboy: Indeed…

Tim: Backing up slightly to the wiki page about schema. I was unclear when it came to the emails - are we going to include the bib.schema.org extensions? For cover for example, there is an item for cover art.

Ivan Herman: I believe what we said in Toronto - when we talk about schema.org we talk about the core and all the extensions that are useful to us - including bib. The cover page, nevertheless, we had a long discussion and what we agreed upon - is that the cover page (if there is one) it should be listed as a resource indicating that it is a cover page and not using one of the schema.org values and those are not really what we consider the cover page for a book. It’s not using for example the bib extension you were referring to…

Ivan Herman: but this is clearly something we may want to discuss.

Benjamin Young: I just wanted to thank Ivan for drafting an experiment like that. I’d love to see more. We have a folder int he repo - the wiki is a great start, but lets make sure they get in the repo as well.

Benjamin Young: I also wanted to note that there are a couple of things we’ll need to determine - the editors in schema.org are person typed and not text. We also need to define when and if they become part of the JSON LD. Some things will need to be objects and not strings. But beyond putting that on our roadmap, this is a great start.

Ivan Herman: Just a question - josh, you were happy with the result. Should I put an object with a name?

Joshua Pyle: Were you asking about Ben’s suggestion that authors are always persons? In some cases they are groups or on-behalf-ofs, there is a wide variety. When they are humans we should tag them as humans.

Ivan Herman: But they could be institutions.

Ivan Herman: http://schema.org/Organization

Joshua Pyle: we could have something softer… at the end of the day it would be nice to have some sort of string for a contributor. There will be rare cases where it’s “the working group” of such-and-such. I don’t want people bogged down that we need object definitions for that.

Ivan Herman: You could use valid targets for that.

Tim Cole: The reason why you want - things like authors as objects - the reason you want that is so you can identify a string, but also a URL so you can better identify. Organization is not heavyweight in schema - you don’t have to have a URL but it’s handy to be able to do that. You can provide addresses or additional information that might be useful. I think that as Benjamin noted is that you can’t just put in a string, but you’ll also want to provide a URL.

Joshua Pyle: +1 to Tim

Garth Conboy: For editor it needed to be a person, but for author it’s an Person or Organization.

Benjamin Young: I was going to follow that up with creator, but author is fine. It will just come down to us to iron out what we want and managing expectations. At first glance I don’t see them as out-of-sync.
… We’ll have to maybe have a taskforce…

Garth Conboy: a?

Garth Conboy: Any other discussion on this topic? Or any other discussion on other topic.

Garth Conboy: I think the homework for people is - if you were tagged - to put it together for Matt. A readthrough of the minutes from the F2F, then not have free time at the next meeting.

Garth Conboy: I’m feeling good about progress after the F2F.

Ivan Herman: I will put examples into the Repo.
… The discussion that some of us had this afternoon is that there are a number of minor issues that need to be ironed out. Please look at them.

3. Resolutions