Publishing BG EPUB-Rec task force telco — Minutes

Date: 2018-11-19

See also the Agenda and the IRC Log

Attendees

Present: Liisa McCloy-Kelley, Luc Audrain, Ivan Herman, Tzviya Siegman, Dave Cramer, Murata Makoto

Regrets:

Guests:

Chair:

Scribe(s): Dave Cramer

Content:

Liisa McCloy-Kelley: we’ve talked about all of this a lot without reaching agreement

Luc Audrain: we could imagine several plans
… and describe them, and these could serve as proposals
… I could imagine 2-3 plans
… one at TPAC: 3.2 as rec in PWG
… and the PWG would be rechartered
… with the new work item. Let’s call it Plan A.
… the second plan would be the least change. EPUB 3.2 stays in CG.
… no rechartering for PWG
… in the middle there may be need to recharter for adding audiopub as module of WP and changing delivery date of EPUB4

Tzviya Siegman: I thought this group was going to try to make sense of the concerns of the larger groups
… maybe we can put to rest some of the concerns, and they might be based on misconceptions
… and we as a group might discuss what we want to do
… we’ve mostly heard that people want to move to REC track

Murata Makoto: Yes, indeed.

Tzviya Siegman: we’re worrying about change

Ivan Herman: beyond what tzviya said
… there are some fundamental questions we have to ask ourselves
… which are necessary for charter, but are important even if we don’t recharter
… what does the publishing community want to achieve?
… what are the needs, and can they happen at w3c?
… we need an explainer for a charter proposal anyway

Luc Audrain: we finished f2f with q from ivan: we’d need a clear answer from PBG before we do anything

Tzviya Siegman: I felt the WG wanted 3.2, but the BG wasn’t informed enough. We need to provide information to them.

Liisa McCloy-Kelley: let’s go to the pros/cons doc and see if we can address concerns

Tzviya Siegman: https://docs.google.com/spreadsheets/d/17CyIqihtjjzT7Abbcq9sSqLNqLGKs2cuXdg4DcFulGY/edit?usp=sharing

Dave Cramer: who is our audience?

Liisa McCloy-Kelley: the audience is 1. the publishing business group 2. all of those parties, perhaps with the exception of the end reader
… we build standards so people can more easily get things out there. it’s governments, education, suppliers
… we know iso is important for some governments and educational institutions
… if we do 3.2 but it’s 3.0.1 that has ISO approval we bifurcate things

Tzviya Siegman: https://github.com/w3ctag/w3ctag.github.io/blob/master/explainers.md

Tzviya Siegman: when we do write this explainer, it will be read by the AC and the TAG when rechartering

Dave Cramer: an early draft: https://github.com/dauwhe/wpub-explainer

Luc Audrain: we’ve brought lots of information to the BG

Liisa McCloy-Kelley: back to the document, do we see any of these things in the pros as inaccurate?

Ivan Herman: there are a lot of comments on the google doc

Luc Audrain: what I hear is that there should be rec track on 3.2 with no change
… so that is not a pro

Ivan Herman: this came out of the need for an editorial rewrite for ISO

Dave Cramer: I think we will make the spec better by doing rec track

Liisa McCloy-Kelley: we want to make sure we get through this without losing features

Murata Makoto: It is very important for JP publishers not to invalidate what they have sold.

Liisa McCloy-Kelley: we understand that

Murata Makoto: Particularly so in Japan, since we have tons of EPUB32
… What kind of adjustments?
… Purely editorial rewrite only?
… Yes, if the normal procedure is used.
… then, will EPUB publications containing such non-normative features non-conforrmant?
… Editorial improvements are fine.

Ivan Herman: we define what we mean by conformance

Murata Makoto: There is a definition!
… I doubt that
… Each spec defines conformance requirements.
… EPUB 3.2 defines conformance requirements. It is already there.

Murata Makoto: https://w3c.github.io/publ-epub-revision/epub32/spec/epub-contentdocs.html#sec-xhtml-conf-content

Luc Audrain: the requirement for two implementations is not conformance?

Ivan Herman: we are looking at edge cases
… the only time this comes up is if there is only a single implementation of a particular feature. That’s the edge case.

Murata Makoto: We cannot remove such a feature if some EPUB publication containing it is already sold.

Tzviya Siegman: the link that makoto posted is to the conformance requirements, and it’s very vague.
… saying it must support “visual rendering of the content documents” doesn’t mean it has to support all of CSS
… we’re worrying about edge cases.

Ivan Herman: the point of the w3c process is to prove that every feature can be implemented

Murata Makoto: I think that it is too late to improse W3C style requirements on EPUB 3

Ivan Herman: based only on what is in the spec.
… if we don’t have that, then we have a problem. If it depends on some secret sauce, we have a problem. I don’t think we’ll face this

Murata Makoto: I think that it is too late to improse W3C style requirements on EPUB 3

Liisa McCloy-Kelley: let’s move on to resolve the point in line 10 about free of charge specs vs ISO

Ivan Herman: does ISO accept a CG report as an equivalent?

Murata Makoto: ISO does not really check technical equivalence.
… If SC34 people argue that it is equivalent, ISO will believe so.
… And make it freely available.

Tzviya Siegman: are we writing this for w3c or ISO? Maybe our goal should be writing a w3c doc

Liisa McCloy-Kelley: my understanding from TPAC was, if we wrote this as w3c REC, then it could be easier to get ISO approval, and we wouldn’t have to make changes for ISO version

Murata Makoto: If W3C has a REC, it can easily be submitted via the PAS procedure.

Ivan Herman: ISO just puts it’s stamp on it, full stop. that’s the PAS procedure.
… it’s an administrative step.

Murata Makoto: Moreover, maintenance procedure is convenient for W3C

Liisa McCloy-Kelley: so we’re not writing an ISO doc

Ivan Herman: we don’t have to

Murata Makoto: For example, ISO/IEC imposes a tough rule about registries.
… If we use the normal procedure, we cannot use Schema.org.
… If we use the fast-track procedure or PAS procedure, we can.

Ivan Herman: there’s no problem with schema via W3C

Murata Makoto: A11Y does.
… Has to be moved to techniques.

Luc Audrain: we are aware of this move to techniques

Luc Audrain: the work in the CG didn’t have real impact on WPUB until now

Ivan Herman: I don’t want to get into personal comments

Tzviya Siegman: row 7: that’s not a concern because we don’t intend to drop features

Murata Makoto: Still concerned.
… I do not understand what you mean by “non-normative”.

Tzviya Siegman: row 8: ocf might be in conflict with packaging
… if we have technical concerns about features it’s better to resolve them within the working group, rather than across groups

Ivan Herman: simpler than coordinating two groups

Tzviya Siegman: I don’t think it’s a concern

Tzviya Siegman: the point about having chairs from Asia is valid but not a con for rec-track EPUB
… if we can people involved in the testing it can force adoption. I dont think it will slow adoption

Murata Makoto: I don’t think that JP publishers care ISO standardization.

Liisa McCloy-Kelley: there’s confusion with people who are still on EPUB 2
… we know it’s fine to go 3, safe to go to 3,0.1 or 3.2

Luc Audrain: the con is slowing adoption of EPUB 3?

Liisa McCloy-Kelley: maybe the con is that it requires a clarification of the timing of all of this

Luc Audrain: my concern is that 3.2 rec track will upset adoption

Tzviya Siegman: we could call what we have now as a beta version of 3.2

Ivan Herman: why is this different than we plan to work on testing in CG?
… if people realize we are testing, it sends the same message
… and most outsiders wont understand the difference between CG and WG

Ivan Herman: do you think we don’t need to test the spec?
… any implementation has effects on the spec

Luc Audrain: one of the first implementation is epubcheck
… so there are some editorial changes needed

Ivan Herman: great!

Dave Cramer: testing will improve bugs and the spec
… it is a part of it to say that the implementation is testable
… testable assertions for reading systems and for publications
… testing the community group will do will focus on the requirements of the reading systems

Murata Makoto: I need more clear definition of what you mean by “making some features non-normative”