See minutes online for a more detailed record of the discussions. (The headers below link into the relevant sections of the minutes.)
The CSS WG had a face-to-face meeting last week in Paris; from this group Dave Cramer, Chris Lilley, Bert Bos, and Alan Stearns participated. They gave a report (led by Dave Cramer).
There was a fair amount of discussion on issues of interest for DPUB. The issue of displaying things in a viewport/page came up which was identified as having missing pieces; seems to be missing from the conceptual model of CSS. A new spec will be defined to bridge that gap; that is an important element for pagination in general. The CSS fragmentation spec is also progressing; these two features are essential for a more general pagination solution.
However, on the pagination in general, there is currently a gap among browser vendors and some other participants of the WG. Essentially, browser vendors regard pagination as an application feature (i.e., that should be implemented on top of the browser as a complex Web Application) as opposed to a fundamental feature that should be provided by the platform. Many non-browser participants regard, on the other hand, pagination as a fundamental feature.
For pagination in general (and the needs of the DPUB industry in particular) this means that the core layout part will solve, via the core implementation, about 70% of the needs, and the remaining 30% will have to be solved by external scripts based on the features provided by Houdini. What is in that (rougly) 70% is a very simple pagination which will interact with the viewport spec, but more complex cases like side-by-side, footnoted, sidetones, etc, will not be done.
There are, however, lots of interested people who want to push things forward; maybe subgroups will organize calls and meetings to work on these items and bring them back to the CSS WG.
Another way forward is to see whether there could be some sort of a reference and/or proof-of-concept pagination implementation that could be put out there (maybe even in open source). At the moment, all reading systems implement pagination on their own (and people are usually not proud of the code they produce); but the new possibilities of the CSS core may help in reducing the footprint of such an implementations; a proof-of-concept code would help. The best would be to seek a cooperation with, e.g., the Readium Consortium on this.
Another line of thought is to gather evidence, studies, etc, on the usefulness of pagination for long content, regardless on whether this is used in a browser or a reading system. (An example is a study of the Nielsen Norman Group “Infinite Scrolling Is Not for Every Website”.
There were also some discussion on the feedbacks received on the recent IG draft on CSS priorities at the CSS WG; some reorganization of the document may be necessary. The most important takeaway was that, maybe, a Math may warrant a separate document, due to its complexity and size.
The group also discussed to organize a more structured joint meeting with the CSS WG at the TPAC meeting in October.
Next week’s meeting has been cancelled due to a US holiday.