Publishing Working Group Telco — Minutes

Date: 2018-01-08

See also the Agenda and the IRC Log


Present: Peter Krautzberger, Dave Cramer, Tzviya Siegman, Laura Brady, Wolfgang Schindler, Vladimir Levantovsky, Matt Garrish, Ivan Herman, Toshiaki Koike, Deborah Kaplan, Avneesh Singh, Bill Kasdorf, Luc Audrain, Chris Maden, Romain Deltour, George Kerscher, Nick Ruffilo, Benjamin Young, Jun Gamou, Tim Cole, Hadrien Gardeur, Mateus Teixeira, Ben Schroeter, Joshua Pyle, Laurent Le Meur, Brady Duga, Garth Conboy, Ric Wright, Marisa DeMeglio

Regrets: Rachel Comerford, Jasmine Mulliken


Chair: Tzviya Siegman

Scribe(s): Dave Cramer


1. Martin Luther King day

Tzviya Siegman: many offices are closed next Monday
… should we meet?

Deborah Kaplan: -1

Dave Cramer: -1

Vladimir Levantovsky: Tzviya should have her rightfully deserved holiday, -1

Nick Ruffilo: +1 (for MLK being awesome, and me being around)

Peter Krautzberger: +1

Wolfgang Schindler: +1

Luc Audrain: +1

Jeff Buehler: -1

Chris Maden: +1

Ivan Herman: 0

Ivan Herman: +1

Bill Kasdorf: -1

Matt Garrish: 0

Benjamin Young: -1

EvanOwens: -1

Avneesh Singh: -1 not available next week

Tzviya Siegman: 0

Joshua Pyle: -1

Laura Brady: 0

Romain Deltour: 0

Tim Cole: -1

Heather Flanagan: +1

Dave Cramer: I would propose we cancel

George Kerscher: cannot join on 15 -1

Tzviya Siegman: I think I’m with dauwhe

Resolution #1: no meeting next week

2. New member

Tzviya Siegman: we have a new member this week, Laura Brady

Laura Brady: I’m Laura Brady, I’m an ebook developer from the Toronto area

Peter Krautzberger: woohoo! Welcome Laura!

Ivan Herman: welcome Laura
… can you send me the email from the system on the invited expert submission?

3. Minutes of last meeting

Tzviya Siegman:

Tzviya Siegman: any comments?

Resolution #2: minutes approved

4. FPWD publications

Tzviya Siegman:

Tzviya Siegman: congratulations! we have published three FPWDs
… and publishing generates more work
… as you know, these are shell documents :)
… thanks everyone for their hard work

5. JSON encoding for the infoset

Tzviya Siegman: Hadrien was working on some examples

Tzviya Siegman:

Tzviya Siegman:

Hadrien Gardeur: the idea is, now that we have an idea of the infoset for WP
… I’ve been looking in two directions
… one is web app manifest
… and I’ve been looking at what it does, and where we can reuse elements
… and I’ve looked at how we can extend WAM to cover our infoset
… we need to express each thing listed in our infoset
… I think there’s very little in WAM that can be used
… I was mostly inventing new elements using conventions of WAM
… that’s the first experiment
… the second thing is more of a comparison
… in the readium community, we’ve created the web publication manifest for use in readium
… and so I’ve created a doc comparing the w3c infoset with the readium web pub manifest
… and almost everything in the infoset is already in the RWPM
… and the rest can be done with link relations
… in the case of WAM, I also created two examples
… one is a full example
… one is a minimal example

Dave Cramer: note WAM = Web Application Manifest, hard to type a lot

Dave Cramer: RWPM = Readium Web Publication Manifest

Hadrien Gardeur: that’s a quick summary; I’m happy to go into more details

Tzviya Siegman: comments?

Ivan Herman: thanks for doing that
… I’m not surprised that many of the elements are not defined by WAM
… the way I understand WAM, the syntax part is the secondary and easy part
… the core of it is the lifecycle, security, etc, stuff defined around the json
… just the serialization, then I don’t care much about WAM but there are other things
… whether they are useful or not depends on what would be the implementation strategy of WP in a browser
… have you looked into that?

Hadrien Gardeur: part of processing of manifest is tied to syntax, like rules for things that are missing
… this is important for title and language
… for all the other infoset elements, we’d have to define the life cycle ourselves

Hadrien Gardeur: there are at least seven missing elements
… I don’t think we’d get a lot

Hadrien Gardeur: for the second part
… which is more related to UX, that’s not something I addressed
… but there have been discussions started by baldur and benjamin
… I think we have quite different expectations of how things are implemented
… WAM is implemented through a triggered installation
… added to a list of apps
… I think that’s different from what we’re looking for
… I think what we’re looking for is closer to an enhanced reader mode
… so we’re providing a manifest to enhance reader mode
… they don’t have prev/next, offline, etc
… I think we’re more closely related to reader mode than web app installs

Dave Cramer: WAM has much more than just the processing of the members
… including creating browser contexts etc

Benjamin Young: dave stole my thunder :)
… there’s a lot of WAM processing focused on installation on a home screen, and is unrelated to what we want to do
… and the guy who wrote WAM said it wasn’t the same thing at all
… so it’s not the right vehicle for us
… and Hadrien’s experiments confirmed that, so we should mosey on to a different format

Romain Deltour: I have a comment and Q
… comment is the counter-argument to benjamin
… installing a web app on a mobile screen is not far from installing a publication in a library
… launching an app within its own browser chrome is similar to opening a pub in reading mode
… so I think it’s too early to reject WAM
… my Q is for hadrien
… based on my limited knowledge, one of the key features of RWPM, is that it has a model based on collections and links, which is extensible
… if you were to extend WAM with this link and collections model, would then WAM be that different from the readium manifest?

Hadrien Gardeur: in terms of design, the two manifests are very different
… WAM is not designed to truly extensible, or even have a link element
… RWPM is designed to be compatible with EPUB 2 and 3 infosets
… and provide a lossless conversion from OPF
… about extending WAM, you’re right that for some of those things, we could make it more hypermedia friendly
… but I don’t think that’s true for all of them
… I’ve tried to reuse what we’ve learned from readium
… the current principles are not compatible with WAM
… so a partial yes to your last question

Benjamin Young: WAM dictates installation—it’s whole purpose is to install a stripped down browser for a web app
… but there’s no reason to install if you don’t need that
… and if you want reading order, but not at install time, then you shouldn’t use it
… so it’s probably not the right place for that

Tim Cole: this suggests we’re talking about apples and oranges
… the infoset talks about the object that’s being read
… it doesn’t provide help for a generic browser
… for that you need an installable application
… 2 sep. things: what’s the minimum the browser needs to do for us for a reading object
… merging these things is problematic

Ivan Herman: timCole was a bit quicker, as usual :)
… what we need before making a decision:
… we need a general idea (or experimentation) on how this would be implemented in a browser
… the reader mode that Hadrien referred to is a good example
… but we need to have more feeling about what an implementation would be
… and we need to know that before we accept or reject WAM
… there is a different attempt to do manifest in the web payment WG
… where they have to implement in browsers
… and they have a json thing
… in that manifest they have an entry to refer to possible web app manifest
… in case web payments are implemented by a web application
… so the implementation can work either way
… we could invite the staff contact to talk to us
… web app as reference rather than creating an extension

6. Implementations, experimentations

Tzviya Siegman: next agenda item is implementation
… we have google play, kobo, other people from readium
… we would like to see people take the spec as it is now, and work on implementation
… without seeing the implementations, we can’t know if WAM is right for us
… we’ve seen some toy reading systems from bigbluehat and dauwhe
… SUP has an implementation

Hadrien Gardeur: in the case of readium, for RWPM we have many implementations
… epub.js has been using this
… in readium 2 we have something on andriod, ios, windows, linux, macos
… in our case we use that for EPUB, too
… we turn epub into wp locally
… then we display in the navigator

Bill Kasdorf: Do we need one of these meetings to be devoted to an in depth show-and-tell on Readium-2?

Joshua Pyle: I’ve been instructed to make atypon resources available to test this
… our system is based on readium
… I’m volunteering :)

Tzviya Siegman: we need to see non-readium implementations

Hadrien Gardeur: epub.js is not readium
… nypl has built something that wasn’t readium

Ivan Herman: for me the big question is
… do they do offlining?
… in a browser?

Hadrien Gardeur: some are web apps, some are native apps using webviews

Ivan Herman: I mean, its like epub.js, part of my browser but then works offline

Hadrien Gardeur: in the case of NYPL prototype
… it does support offline access, using appcache in ios and SW in android (?)
… offline support is tricky

Tzviya Siegman: we’re fully aware

Ivan Herman: having all this documented would be great

Mateus Teixeira: Norton has an independent platform as well, and the Norton Lab is interested in testing early implementations–we talked a little about this at TPAC. I can’t commit to anything without looking into our plans for the coming months, but I hope to contribute at least on a personal level, even if it needs to be a side project.

Tzviya Siegman: let’s document the implementations

Nick Ruffilo: how in depth does the test need to be

Tzviya Siegman: we’re not testing
… we’re not defining full mechanisms, just showing it could function

Nick Ruffilo: so if i had a manifest and a browser that displayed something?

Ivan Herman: yes, we need framework of what kind of things have to be done for an implementation

Tzviya Siegman: we might do a conglomerate
… Hadrien will contact people who work with readium, Mateus will pull stuff together

Dave Cramer: we’ve been talking about internal formats for reading systems, which might not be the same thing as an authoring format for publishers

Tzviya Siegman: let’s put that on the agenda for next week(ish)

Hadrien Gardeur: producing implementations of offline reading is not testing anything about a format
… these are unrelated to syntax or manifests
… the only thing we test are each browser’s or OS ability to support various web features

Ivan Herman: yes and no, I understand what you say
… my understanding is that current wam is very closely entertwined with service workers
… I don’t know how much of that is a general thing
… and I don’t know if service workers will be the alpha and omega of implementations
… it doesn’t tell us the syntax, but it does tell us whether WAM as a whole is useful

7. Testing & privacy champions

Tzviya Siegman: we still need a testing champion
… we have people who’ve volunteered to do specific tasks
… for example, Avneesh is our a11y champion
… we’ve had lots of new members since we last called for a testing champion
… we’ll buy you a hat!
… we won’t have a spec without testing

Chris Maden: I could do the testing champion thing.

Chris Maden: I volunteer

Ivan Herman: Overall applause for Chris Maden:-)

Tzviya Siegman: we also need work on security and privacy, Bill_Kasdorf may know someone at Atypon

Bill Kasdorf: one obstacle is that he is in Tokyo

Tzviya Siegman: we have an evening call with australia sometime
… anyone want to step up for privacy?
… once we have security we’ll see if privacy can be part of the same champion

8. Packaging mechanism

Tzviya Siegman: there was a hole in the spec
… Ivan or Garth, what should we do at this point? list options? remind people?

Brady Duga: Garth stepped out, so up to Ivan

Ivan Herman: the only thing that we may want to do, we can volunteer garth to find out about JYasskin
… and the google web packaging proposal
… there was talk of meetings with IETF
… we need to know if this is an option at all

Garth Conboy: check in with jeffrey to get status? I can do that.

Jeffrey Yasskin: dauwhe/garth: I have half of the packaging proposal in, and I may be able to convince the IETF HTTP WG to “adopt” it in March. That’s not the part you care most about. The bundling part, which you’ll care about more, is a WIP in

9. General outreach

Tzviya Siegman: we have lots of drafts, we need feedback
… i’ll reach out to TAG
… we want feedback from Readium
… we need to know if this can be implemented in browsers, with existing reading systems, what does amazon think?

Ivan Herman: the 2 most important targets are to talk to the EPUB people at Edge
… and talk to the Chrome team
… I don’t think we can influence apple
… and I don’t know how to reach out to mozilla
… if you have contacts, use them

Hadrien Gardeur: quick feedback
… even for an org like readium
… the offline part is tricky
… it’s not obvious at all that we can have offline for WP on a number of platforms
… I agree we should reach out to browsers, but maybe not just people working on the browsers
… the reader modes might be web apps themselves
… the easy way in could be the reader modes
… chrome is the only one without a reader mode
… safari, opera, edge have one

Tzviya Siegman: if you have a person to contact involved in reading mode, we’d love to know

Hadrien Gardeur: I don’t
… Firefox is interesting in long-form reading

10. Synchronized Media

Tzviya Siegman: marissa had emailed me about the synchronized media group?

Marisa DeMeglio: we have just proposed and created the “synchronized multimedia for publications” community group
… I’ll send a link to our list
… and I’m hoping people will join over the next few weeks and then start work

Tzviya Siegman:

Ivan Herman: marisa, one practical thing that I’ve cleared with the powers that be, I can set up a webex channel for you

Marisa DeMeglio: thanks!

11. Caring of new members

Tzviya Siegman: last item on agenda, before the holidays we asked for feedback from newer members of the group
… sometime in January. I’ll send out an email to the group.
… the first item will be rewriting the guide for newcomers
… as we know it’s hard to dive in
… any other business?

Garth Conboy: are we skipping next Monday?

Tzviya Siegman: yes

Tzviya Siegman: we will meet again on January 22.
… good luck with your implementations

Ivan Herman: email is up and running

Tzviya Siegman: github is there, too :)
… and lots of people here can help with github

12. Resolutions