Publishing Working Group Telco — Minutes

Date: 2018-12-17

See also the Agenda and the IRC Log

Attendees

Present: Benjamin Young, Simon Collinson, Deborah Kaplan, Ivan Herman, Caroline Hayes, Dave Cramer, Joshua Pyle, Wendy Reid, Wolfgang Schindler, Avneesh Singh, Jun Gamou, Franco Alvarado, Marisa DeMeglio, Gregorio Pellegrino, Laurent Le Meur, Romain Deltour, Ric Wright, George Kerscher, Matt Garrish, Charles LaPierre, Ben Schroeter, Lillian Sullam, Garth Conboy, Tim Cole, Brady Duga, Derek Jackson, Mateus Teixeira, Bill Kasdorf

Regrets: Luc Audrain, Yu-Wei Chang (Yanni), Zheng Xu (Jeff), Juan Corona, Vladimir Levantovsky

Guests:

Chair: Wendy Reid, Tzviya Siegman

Scribe(s): George Kerscher, Dave Cramer, Benjamin Young

Content:

Wendy Reid: https://www.w3.org/publishing/groups/publ-wg/Meetings/Minutes/2018/2018-12-10-pwg.html

1. Admin

1.1. Approve last week’s minutes?

Wendy Reid: https://www.w3.org/publishing/groups/publ-wg/Meetings/Minutes/2018/2018-12-10-pwg.html

Resolution #1: last week’s minutes approved

Wendy Reid: Approved last week minutes

1.2. Should the WG publish a draft of WP spec?

Wendy Reid: Two ideas to have a resolution in the meeting, or to send an email out for approval.

Dave Cramer: we should distinguish between editors drafts, which can be done any time.
… However, for a WG draft we should have a resolution.
… In this case it should be a retroactive draft.

Proposed resolution: Approve the publishing of the recent Working Draft of December 2018 (Wendy Reid)

Garth Conboy: +1

Tim Cole: +1

Benjamin Young: +1

Ivan Herman: +1

Deborah Kaplan: +1

Joshua Pyle: +1

Wolfgang Schindler: +1

Jun Gamou: +1

Mateus Teixeira: +1

Evan Owens: +1

Ben Schroeter: +1

Wendy Reid: If anybody objects, let us know. It is approved.

Resolution #2: Approve the publishing of the recent Working Draft of December 2018

Wendy Reid: In the future we will vote in a meeting before, or do an email CFC, for the WG resolution.

2. Entry page and the vanilla browser

Tzviya Siegman: The entry page and what that means in a distinguish browser.
… The entry page needs to open in any user agent.
… More advanced features would need other supported affordances.
… We want to make sure we are all of one mind. We are talking about a user agent with no additions.

Avneesh Singh: +1 entry page should open in browser

Deborah Kaplan: To make it clear, it needs to function in a standard agent that is not WP aware.

Deborah Kaplan: language proposal: ‘Standards-compliant user agent that is not WP-aware (henceforth “minimal browser”)’

Dave Cramer: Are we envisioning additional requirement here?
… We say that the address must resolve to a document that has an embedded document.
… Are we saying the document must be able to access other documents in the publication.

Ivan Herman: No, the only thing we can say is that the author has the possibility to do it.
… However, we cannot say that the other pages may not be meaningful to the browser.
… It is the author’s responsibility to avoid things that will make for a poor reading experience.

Matt Garrish: Ivan covered, I would not turn this into a content specification. If we say that people must do this or that. It is better not to go into the content requirements.

Deborah Kaplan: This not a content spec, but it is important to provide to authors and browsers that there can be best practices.
… We have agreed that in a non-WP-aware browser, you can get to the content. We want to be able to reach the navigation with as few polyfills as is technologically possible.

Laurent Le Meur: a browser that will not be able to process the manifest, then the TOC will be the only way to access the publications.
… If there is no link to the TOC, then there is no way to access the documents in the publications.
… at least something that resembles the reading order must be in the the entry page.
… There must be a way to get to the TOC.

Benjamin Young: +1 to the trend. There is already fallback information.
… If a entry page allows to browse a publication.
… We should take advantage to make sure the publication is available.

Ivan Herman: We must be careful. Deborah said polyfill, which is the magic word.
… What happens when we put in a web publication which has only the declarative.
… What happens when I put in a polyfill which does the min viable behavior.
… We should not confuse these two items.
… If we go down the line of adding content requirements, and shoulds, this is a problem.

Garth Conboy: +1 Ivan

Tzviya Siegman: we’re talking about only the PEP, not the entire publication

Avneesh Singh: you can make the entry page useful for getting into the publication. It’s a fallback mechanism.
… we are not encouraging browsers to render the manifest
… I would be OK with TOC being mandatory

Matt Garrish: we’ve defined the entry page, it’s html with a manifest
… I think we leave it at that
… everything will change over time
… the polyfills were change
… developers will figure out how this works
… people will take this into account.
… we should have best practices and tutorials
… have some fallbacks
… but don’t start mandating content

Ivan Herman: +10000 to Matt

Garth Conboy: + π

Deborah Kaplan: tzviya, could you say the sentence we’re talking about?

Mateus Teixeira: +1 mgarrish & Avneesh

Tzviya Siegman: it is possible to open the entry page in a non-wp-aware browser

Deborah Kaplan: there are later conversations about polyfills
… and we can talk about minimal whatever

Deborah Kaplan: ‘It is possible to open the entry page in any standards-compliant user agent that is not WP-aware (henceforth “minimal browser”)’

Deborah Kaplan: the opening question is that there were questions about what a vanilla browser is
… the simple Q of it is possible to open the entry page in the browser we agree on
… we are hanging all our discussions about polyfills and navigation and mandatory content on this simple question

Deborah Kaplan: ‘It is possible to open the entry page in any standards-compliant user agent that is not WP-aware (henceforth “minimal browser”)’

Ivan Herman: +1 to dkaplan3

Benjamin Young: https://www.w3.org/TR/html/infrastructure.html#conforming-document

Benjamin Young: can we have better terminology than vanilla browser
… we can use something from html conformance classes
… the polyfill discussion is distorting
… you can do anything
… with polyfills
… but one of the conforming classes is something that can’t do scripting
… and we should have user expectations on this document
… and we’re talking about dictating the content
… if this is the publication address, it should send back enough to make sense of it without a polyfill

Mateus Teixeira: +1 bigbluehat

Garth Conboy: plus one to matt ivan dkaplan3
… the content creator decides what happens when that html doc opens
… but we shouldn’t dictate anything

George Kerscher: when I open this page I can bookmark it in my conforming browser, and get back to it
… I don’t have to have any affordances, I can go back to that entry page and continue to explore the publication
… Garth says we’re not going to dictate being able to get to all the publication pages from the entry page

Tim Cole: I agree with dkaplan3’s formulation and bigbluehat’s comment
… I’m a bit concerned by a dead end publication
… we need some way to make clear the content of a publication from the entry page
… even if it’s one by one
… if I’m a user and come to things that come to web publications that are dead ends, that’s bad
… there should be some way in the spec to be able to continue into the publication

Tzviya Siegman: that’s why some of us want an html toc in the entry page

Benjamin Young: https://github.com/w3c/wpub/wiki/Minimal-WPUB#wiki-body

Benjamin Young: I want to share a link from our wiki
… it’s a minimal wpub, if you open in a browser it’s a blank page
… so what tzviya said would solve for that

Dave Cramer: I was concerned about authors who would put WPUBs on the Web knowing that all browsers don’t support them

Garth Conboy: nothing smart will happen with WPUBs without polyfills

Ivan Herman: according to the HTML spec, I can create an html page with a header and an empty body
… why do we want to be more restrictive?
… it’s possible to make something that’s useless. Don’t do that.

Garth Conboy: +1 Ivan

Ivan Herman: mandating content is not how html works

Bill Kasdorf: “render the TOC” vs. “access the TOC”–the latter can be done via the header

Tim Cole: to ivan’s content, if all we’re doing is writing another html spec, he’s right

Deborah Kaplan: +1 timCole

Tim Cole: we’re writing for a new entity that’s distinct from html
… what makes a web publication different from hTML?
… one of the characteristics is that something happens when you go to the entry page

Ivan Herman: that’s a question of where
… if I have a browser with no info about WPUB
… I don’t think that’s a problem
… I can make a document that even in a vanilla browser can do something sensible, but I will need scripting

Tim Cole: I understand your point

Tzviya Siegman: I feel this conversation has opened up a lot of other conversations
… my goal was making the entry page useful to users
… and then make it useful for machines
… and I’m not sure we’ve agreed on this point

Dave Cramer: the original resolution is a tautology–since the entry page is html, it can be opened in a browser

3. WPUB explainer

Tzviya Siegman: https://github.com/w3c/wpub/blob/master/explainers/wpub-explainer.md

Tzviya Siegman: the TAG recommends explainers. Dave wrote the explainer with some help from Ivan and Tzviya

Dave Cramer: I filed an issue which expresses that the most important thing we could do right now
… is get feedback from the TAG
… I started the explainer because that’s a necessary step to get that feedback from the TAG
… we are discussing changes about how the Web works
… and we should get the TAGs feedback on those changes
… to make sure we’re not altering those fundamentals incorrectly
… we’ve been spending a lot of time in our own little publication world
… but since we’re writing for the Web, we really need and could use their help

Deborah Kaplan: when I read that doc, my reaction was “yeah, thank you dauwhe!”
… it does, however, assume the readers have some basic assumed principles
… things from the EPUB community may make people go “huh?”
… one big one is reading order
… we understand that there’s some default reading order
… that doesn’t mean you can’t read the publication other ways, etc.
… but others may not
… so a document that talks this much about reading order, should probably put in a glossary or something explaining those assumptions/understandings

Dave Cramer: thank you so much dkaplan3!

Tzviya Siegman: any other feedback?

Ivan Herman: I have a practical issue
… I’m perfectly agreeing that showing it to the TAG early next year
… there was a TAG review…well, an individual from the TAG reviewing the document
… it was a long time ago and on a very early draft
… so I would think that comments going back to that issue is not necessarily moot, but almost
… because many things are gone…or almost gone
… I’d prefer to streamline a way that it is completely fresh
… and not referring back to the original review

Evan Owens: what is the long term strategic goal of this project?
… are all Web browsers going to read these documents?
… is the entry page just an interim solution?
… what’s the goal?

Tzviya Siegman: if I could jump the queue, I’d say that all browsers are the goal
… but that’s not likely to happen exactly
… chrome or chromium or whoever aren’t likely to implement the whole thing
… they’re more likely to add parts of it
… once we see adoption of a polyfill
… in an ideal world, yes, but in the meantime we’ll need polyfills, etc in order to encourage the browsers to implement

Charles LaPierre: just wanted to point out that when we were doing the a11y task force
… in the digital publishing IG at the time
… we did a gap analysis
… and I’d really like us to see us do that again
… across a Web page, multiple web pages, a publication, etc.
… not just relying on the back button for example
… I’d like to see us spell that out more clearly

Dave Cramer: I’d like to respond to ivan’s thoughts on the TAG feedback
… it seems their original feedback still applies directly
… back then Andrew Betts mentioned the Readium Web Publication Manifest
… and raised concerns about a similar approach
… which we’re still using with our JSON manifest
… we need to explain what are expectations are for the Open Web Platform for implementing these things
… and acknowledging this feedback seems important
… because in some since this isn’t a fresh start

Tzviya Siegman: +1 dauwhe

Ivan Herman: first of all, I accept dauwhe’s comments
… but the reason I was in the queue was going a bit beyond what Tzviya said
… first, agreeing to everything she said
… maybe it’s worth putting into the explainer
… that we understand that browsers will not solve everything under the sun
… that we’re considering a structure where we’re building on top of things built on top of the browsers
… that seems to be the way browser engines want to go
… and that what we’re building can exist on top of those things
… so that we do understand that some of these features may make it into the core
… but that we don’t expect that anytime soon
… and perhaps we can point that out in the explainer

Dave Cramer: that sounds good to me ivan

Avneesh Singh: just one thing
… maybe to revise the document a little…it talks so much about EPUB
… they have a negative reaction to that
… and it looks like we’re building a new version of EPUB
… maybe we should not be saying that, and neutralize it a bit
… we have a cognitive bias in this group
… all the implementers are EPUB RS implementers in this group
… and these folks aren’t, so we should attempt to avoid bias against reading systems

Dave Cramer: I sort of see your point
… but EPUB has had a strong influence on the activity of this group
… because some of these choices were made because these are the kind of things that worked in EPUB
… so trying to explain these things without referring to EPUB would be very challenging
… we’re also chartered to spec EPUB4
… and the TAG has existing perspective on EPUB in general

Garth Conboy: I think “bias” (toward EPUB) sounds pejorative; we need to make sure what we do does provide (eventually) a path forward for EPUB… which is now the dominate format used in eBook land. That’s not a bad thing.

Tzviya Siegman: I agree with dauwhe that acknowledging that bias seems important

Dave Cramer: I can try and do that in a more neutral way

Evan Owens: thank you for the summary

Tzviya Siegman: sorry to cut you off, but we’re out of time
… there are no more 2018 meetings
… next meeting is January 7th
… enjoy your time off!

Benjamin Young: [lots of people]: happy holidays!

4. Resolutions