On October 22 and 23 the Publishing WG held 2 days of meetings as part of TPAC in Lyon. This was our second time at the massive W3C gathering (6th if you count our adolescence as an interest group); days were filled with meetings, hallway discussions, and making connections with other parts of the web world. It’s exhilarating and exhausting, but talking to your peers is the best way to make progress, to make what you’re working on better. Special thanks to the organizers of TPAC, who provided us with a room that made the group feel very important.
The working group’s focus for the 2 days was on the following topics:
- Components of Web Publications
- Use Cases for WP
- Interfacing with the Business Group
- Synchronized Media
- EPUB 4
We made progress in a few areas, specifically around better quantifying the relationship between the publication and its manifest. We removed the term “infoset” from the spec in place of reading-order and resources.
Use Cases for WP
A major point of discussion that came up was around use cases and what “WP-aware” means. For context, we wanted to define what a user agent would do with a WP if it didn’t know what it was, and what it would do if it did. We’d been back-and-forth about this for a long time, and we finally settled on the “Minimum Viable Product” (MVP) of a WP and our expectations for a non-WP aware user agent, which then helps us define what a WP-aware one would do. An important part of spec language is defining MUST vs. SHOULD, and we now have a framework for how to do that within our spec. This not only affects how we write the spec, but also how we handle testing when it comes time for that, so it we made a lot of progress there.
We got to a basic definition of the bounds of the publication, which is a point of difference between WP and EPUB 4. It is easy to define the bounds of a packaged document (like EPUB) but harder on the web. We resolved that the essential bounds of a document are the contents of the reading order and resources, anything outside of that is not within the bounds of the publication.
i18n, JSON-LD, and Schema.org
For this session, we were joined by honoured guests Dan Brickley, Richard Ishida, and Addison Phillips. We discussed issues with internationalization in the JSON-LD formatting we use for author, title, and other fields in the manifest. The main issue is around directionality of languages and how to express it, especially when multiple languages with multiple directions are used. We also reviewed some issues we were having with schema.org language, specifically in places where we were unable to find something that suited publishing (`type`) or where a definition didn’t suit the use case we needed it for (`duration`).
Cleaning up Github
The last part of day one was spent discussing and closing github issues that had been sitting in queue. We were able to close older issues that had long been addressed, but we spent some time discussing a longstanding issue on the definition of the HTML TOC format. The proposal was to have two tables of contents within WP (or at least the possibility of two). We are reluctant to have duplication in the spec as this was a big complaint in EPUB 3, but in this case the overwhelming response from publishers was that 2 is the preferred system, a “pretty” human-readable TOC and a machine-readable one. There was a proposal from Juan Corona to create an algorithm that could work with one to create something machine-readable from the human-readable one, which was adopted by the group as an interesting experiment to explore. We look forward to his work.
Updates from the Business Group and Rec Track!
The Publishing Activity has been working in a few different groups since the IDPF came under the W3C. On the second day of TPAC, the Business Group, Community Group, and Working Group had a chance to discuss recent activities together. The CG has been busy working on EPUB 3.2, and whole group discussed how best to advance EPUB. After much discussion, we propose to bring EPUB 3.2 into the WG as a rec track document. This requires BG approval, charter revision, and of course, process-level testing. The group is still discussing exactly how best move forward with Web Publications and EPUB 4.
Marisa DeMeglio and Daniel Weck presented the progress they have made in regards to [synchronized media](https://www.w3.org/community/sync-media-pub/). The focus of their presentation was to present their findings on the research they have done on available technologies already available on the web. Their conclusion was that there was no ready option for synchronized media that meets enough of the use cases required, so they are proposing a variation of existing standards. Draft spec to come. WG TBD.
Garth Conboy led our discussion on EPUB 4, a carry over from the morning. There was a big concern in the group that by introducing EPUB 4 too soon after EPUB 3.2 we will run into adoption issues from the publishing community. Instead of pushing ahead, we proposed putting EPUB 4 in the “freezer” for now, and moving ahead instead with a series of proofs, mini-specs like audiobooks, manga, or sync media. EPUB 4 is still a goal, but not as immediately as thought when we chartered. By moving towards a more modular approach we also solve immediate use cases for the industry while better framing the need for EPUB 4 in future.
Wendy Reid presented an update on the work of the Audiobooks Task Force. The TF has made a lot of progress recently and wanted to bring three issues to the main WG for review. Audiobooks is a real-world use case, and we found out that when people have a clear need in the industry, they spend time on specs. The first issue was resolved by the WG’s discussion with schema.org about the `duration` property. The remaining two issues were bitrate and supplemental content. After much discussion the issues remain open for further debate at another time. The next steps for the audio TF are to reach out to industry experts for more insight on the problems we need to solve, and to address the next big challenge in audiobooks: packaging.
Once the dust has settled and our internal clocks are back in order, we have some work to do! The group is on the verge of a proposed recharter, and the terms of such a major change need to be discussed and shared with the group and the AC. Our work for WP does not stop, as we have use cases to further define and issues yet to be closed. The audiobooks task force has decisions to make and a draft spec to produce. Keep you eyes peeled on future developments from the PWG, there’s much to look forward to.