Publishing WG Telco, 2017-09-18: SMIL Lite, metadata, fragment locators

See minutes online for a more detailed record of the discussions.


Marisa DeMeglio presented some work that the accessibility task force started on Media Overlay (MO)/Synchronized Multimedia (SMIL) related issues. The task force collected a number of use cases that shows that synchronized multimedia is an important feature for publishing. Use cases include text+sign language, or video+descriptive audio synchronization, for example.

The problem is that while SMIL has been implemented in several EPUB3 readers, its adoption on the Web is very poor. This is a source of problems for Web Publications. The question is how to move forward from there. One avenue is to explore the possibility to define a minimal SMIL (possibly transferring the syntax to JSON) and explore the avenue of creating, eg, polyfills for that level, with the goal of standardizing that level.

Administratively it is not clear whether that should happen within this WG (and whether that is possible, in terms of chartering) or whether a separate CG+WCIG+WG route at W3C is more preferable. To be explored.

Metadata section

A Pull Request has been proposed by Baldur Bjarnason et al on extra metadata issues. This proposal adds some (possibly optional) items to the information set, and also references to external metadata items in their own vocabularies. There has been some discussions on github already, and the resolution is to merge the pull request and separate some of the problems (also discussed on github) into separate github issues.

Fragment Locators

There is a separate discussion around a github issue on what formalism should be used for identifying fragments within resources, as a possibly alternative to EPUB CFI. One approach is to use the Web Annotation Selector and State Model, though this should be explored as for the usability and for its applicability.

The issue includes some questions on details that must be answered. It has been decided that the group would look at the WA document and the issue and a separate Pull Request would be provided.

Publishing WG Telco, 2017-09-01: Web Packaging

See minutes online for a more detailed record of the discussions.

Web Packaging

Apart from some administrative issues, the meeting was around the Web Packaging work, that is currently on the way to define a packaging spec for the Web. This effort replaces the older Web Packaging spec by the TAG. The work has been presented by our guest, Jeffrey Yasskin, who is the main editor of the work.

The work is currently planned to be under the auspices of IETF rather than W3C, although some of the main sections may be, eventually, taken over the W3C. The “Explainer” document on the Web gives an introduction to the technology. The project came out of our emerging markets‚ — a system might have an expensive or limited data plan — so there is peer-to-peer data sharing. Our team wanted to share web pages in the same way. The current sketch is that the whole thing will be a CBOR (binary version of JSON with a few extra features); it has a sequence of features with an index of sections pointing to the offset of the file. The sections are HTTP requests and answers. There will be some mechanisms for sub-packages, as well. The request is where the interesting stuff happens. It has a set of signatures and a certificate on how to trust those signatures. Then there is the manifest – which is the app manifest – and a set of hashes of the sub-resources. There can be a set of hashes for each resource. The thing that is hashed is the concatenation of the request headers, the request, and the body.

The short presentation was followed by some questions and answers, considering issues like relationships to Service Workers (which is very good, ie, a service worker based implementation may hide the details of packaging on the network layer), how to handle this approach with non-browser clients, relationships to (and difficulties with) certificate management, relationship to ZIP, etc.

Publishing WG Telco, 2017-08-28: Issue cleanup, Manifest serialization, WAM, metadata

See minutes online for a more detailed record of the discussions.

Issue cleanup

A number of open issues have accumulated over the weeks. With the latest changes in the Editor’s Draft, and the discussions leading to that point, a number of issues have been closed.

Manifest Serialization

The group discussed whether the serialization of the manifest should follow a JSON syntax (without getting into the details on whether it should be JSON-LD, or any other dialect of JSON). After some discussions the group agreed to go for JSON.

Metadata Proposal

A Task Force has put up a first version of a document on metadata, which had a number of additional discussions off-line. The main question that had to be discussed during the call is to understand what makes a difference between a metadata and a manifest information item; these two coincide a bit in the document.

One approach that was put forward (but not yet decided) is based on what the metadata represents and what we are doing with it:

  • is it for the user agent in order to consume?
  • is it meant for external processors?

However, the borderline are nevertheless fuzzy (e.g., the based accessibility metadata, which is an important group of data).

Discussion will continue with a cleanup of the document.

Relationships to Web App Manifests

The group must answer on what the relationships are between the Web Publication manifest and the Web App Manifest (WAM). No decision can be taken in a short time; instead, the WG sets up a separate Task Force to look into this and come back to the WG. The goal is to be in position to talk to the WAM editors (at the latest at TPAC) but the requires some sort of a consensus in this group.

Publishing WG Telco, 2017-08-21: HTML TOC proposal, milestones

See minutes online for a more detailed record of the discussions.

HTML TOC as manifest proposal

Dave Cramer and Benjamin Young have put forward a proposal, for using HTML as a binding document. It became clear in the discussion that the proposal has several facets that may have to be discussed separately:

  1. Using an HTML file for TOC and its connection to the list of primary/secondary resources
  2. Is HTML a suitable syntax for the Manifest information as a whole

Although no decision has been taken yet on these (and the discussion should continue offline) the more general question on whether it is time to define the serialization of the manifest or whether this should be postponed to cover the abstract manifest first. Some of these issues are also covered in the (currently open) pull request cleaning up the current views on the abstract and concrete manifests.


The chairs have prepared a set of “official” milestones between now and the end of the December when the FPWD is planned. The current issues will be assigned to these milestones in the coming days.

Publishing WG Telco, 2017-08-14: manifest in editor’s draft, usage of URL

See minutes online for a more detailed record of the discussions.

The main underlying goal of the meeting was to consolidate a number of discussion that took place in past few weeks on manifest, so that the Editor’s draft for the Web Publication document could be updated (this will be done under the leadership of the chief editor, Matt Garrish).

Manifest Synthesis

A draft document has been prepared to list the minimum items that should be present in a manifest. That document makes a different between an abstract and a concrete manifest, and lists items into the abstract manifest that reflect current (minimum) consensus. A number of open issues reflect the lack of consensus, for the moment, on whether certain items should be “must” vs. “should”; these issues are explicitly listed in the document.

Some items did come up during the discussion, and the result will be reflected in the consolidated editor’s draft

  1. there should be an item on the necessity (“should”) of including an identifier
  2. it should be “a” reading order, not “default”
  3. there should be a new issue on whether language reference is a “should” or a “must” (and the issue should be referred from the document)

URl-s vs. URI/IRI; identifiers

There was an open issue on whether this (and the other) documents should use the term URL or URI/IRI. Although IRI is more precise and reflects the current RFC-s, the practice of the Web Developers’ community is to stick with the term URL (and possibly make it clear that, syntactically and semantically these include IRI-s). It has been accepted to follow this practice, referring to the HTML5 specification.

A new Pull Request on further terminologies on addresses and identifiers has also been submitted. This will also be incorporated into the editor’s draft.

Publishing WG Telco, 2017-08-07: publ changes, terms, identifiers, minimum manifest

Publishing WG Telco, 2017-08-07: publ changes, terms, identifiers, minimum manifest

See minutes online for a more detailed record of the discussions.

Can a publication change over time?

There was a long email discussion on what the WG should/could do with this problem. Based on an earlier proposal on mail the group agreed that this problem cannot really be controlled by the WP’s, it is a generic Web feature/problem.

Working definitions from Publishing and Linking

The group recognized that a number of definitions are properly defined in a separate Working Group note “Publishing and Linking”. It was decided to reuse these definitions, whenever appropriate, instead of re-defining the same terms.

Identifiers and Locators

A Pull request has been proposed by Tim Cole, Benjamin Young and Bill Kasdorf. The essential point of the proposal is to strictly separate the term “identifier”, with its own definition, and with the requirement that a WP SHOULD (but not MUST) have at least one identifier. It is recognized that, by virtue of being on the Web, a WP does have a locator, but that has a different role. Also, an identifier is defined in such a way that it MUST be possible to convert it into a locator that, through some mechanism, should provide an access to the manifest of a WP.

The WG agreed to merge that pull request, with the understanding that minor editorial changes would still happen.

Minimum Viable Manifest

A proposal was put forward a while ago, refined right before the call. The group has not yet found an agreement; issues discussed were:

  • should it be required to have a “title“
  • should the list of constituent resources include secondary resources as well

The discussion is ongoing on GitHub.

Posted in Activity News, Meeting reports | Comments Off on Publishing WG Telco, 2017-08-07: publ changes, terms, identifiers, minimum manifest

Publishing WG Telco, 2017-07-31: terminology, packaging, locators

See minutes online for a more detailed record of the discussions.


There has been lots of discussions in the past few weeks on various terminology issues. To have a common (albeit not final) baseline, the definition for some terms have been proposed in the current FPWD editors’ draft and discussed on the call. The WG has accepted a baseline definition for

(See the links for the exact values.) These definitions, with some minor wording modifications that were also discussed during the call will be added to the editors’ draft.

Web Packaging Spec

There has been some feedback from the separate work on Web Packaging. The technical details, as well as the division of work between IETF and W3C are still not clear. The plan is to have the editor of the packaging spec join our call in the coming weeks to go for a common understanding. It is expected that this WG will provide use cases for the packaging spec in case there is a clear commonality between the different groups.

Identifiers, locators

There were further discussion on what exactly the role of an identifier for WP-s (as opposed to locators). It has been agreed that we need a clear set of requirements on what WP expects from identifiers; it has been agreed to provide such a set in the days to come.

Publishing WG Telco, 2017-07-24 Code of Conduct, Definition of Web Publication

See minutes online for a more detailed record of the discussions.

Code of Conduct
Garth and Tzviya addressed concerns raised about tone and volume of comments on GitHub. Passion is great, but let’s remember to maintain professionalism. See and for guidance. Please contact chairs or team contact if you have any concerns.

Definition of WP
There has been a lot of discussion in the WG about origin, manifests, and updating and control of the components of a publication. The group decided to revisit the definition of Web Publication and Packaged Web Publication provided by the DPUB IG. See The group is working on refining a definition on the email list so that we can move forward with issues such as addressing, linking, and origin.

2017-07-17 Telco: Draft and TF Update, Issue Triage

See Minutes online for a more detailed record of the discussions.

Review of Authoring Tasks
Tzviya Siegman reviewed the outline of the WP draft and sections assigned to individuals. Editors and task force leads are encouraged to start drafting. Matt Garrish posted a skeleton document, available at and in readable form at

Issue Triage
There has been a lot of activity in our GitHub repo, and we tried to resolve some of the sub-issues that arose. The question raised to the group was what is required in the “manifest” of a Web Publication? Garth Conboy proposed:

  • Identifier of WP (required)
  • Identification as a WP (required)
  • List of publication resources. (required)
  • Reading order(required)
  • Metadata (maybe required)
  • Nav doc (optional)

This led to much debate on the call and additional GitHub issues. Please add your thoughts and feedback at

Posted in Activity News | Comments Off on 2017-07-17 Telco: Draft and TF Update, Issue Triage

2017-07-10 Telco: Testing, Web Packaging, Manifest, A11y TF

See Minutes online for a more detailed record of the discussions.
Shane McCarron of Spec Ops joined us to discuss testing our forthcoming specs. Shane stressed that it’s important to make use of the tools that already exist, such as Web Platform tests. We can also make use of tests that already exist for EPUB, including and the test files for epubcheck. Shane recommends that as we write our specs, we track assertions or testable statements so that we can build a testing strategy.

Web Packaging Update
Garth Conboy and Brady Duga spoke to fellow Googlers about work on Web Packaging. They are happy to work with us, and they plan to split their current work and present formatting and signing to IETF. Several members noted that there are still questions from IETF and Web Apps about the direction of this work.

There has been a lot of discussion on GitHub about manifests. Some threads go into areas that are beyond scope for Manifest. Tim Cole suggested creating separate issues for those items. Dave Cramer will begin drafting some of the manifest spec so that we have something concrete to review.

Accessibility TF
Avneesh Singh sent out a call for participation in the Accessibility TF. Please contact Avneesh if you’d like to participate.