See minutes online for a more detailed record of the discussions.
The meeting concentrated on the Web Packaging specification draft; Jeffrey Yasskin, the editor of that draft, was a guest at the call.
T¡e spec is now divided into two parts. The part relevant to the publishing work is describing a set of signed HTTP request/response pairs from a given origin. There is a draft of this bundling proposal with a Pull Request ready for issues and comments. At the moment, this document is being produced within the IETF framework, but it may come “back” to W3C. Several issues were discussed on the call.
The spec includes a reference to a Web Application Manifest step, which the “top” of the media bundle. Its “start URL” term is used, for example. (It must be noted that the term “manifest” is a misnomer, because WAM does not list the content of the full bundle collection.) It is not yet clear whether the package should require a Web Application Manifest, or whether a more flexible mechanism would be allowed with other types of manifests. (For example, there were discussions with the Web Performance people on their own “manifests”.) Obviously, this is an issue for publishing, too. (Note that a publication related manifest would include an index of the constituent resources, i.e., the list of the bundles.)
The spec puts a strong emphasis on signing each bundle. However, that may not always work well with a WP/PWP environment where some of the content may come from a file system without having ever been on the Web. One way of handling this could be to rely on either a
file: URL or (preferably…) on a simulated environment using a locally run small server and using
localhost in the URL. In both cases it is not clear how the signature could be implemented without too much hurdle on the publisher. However, it turns out that the Web Packaging spec does not mandate a signature, so there may be other approaches to ensure the authenticity of the content. To be explored.
There were also some discussions about timing of the packaging spec. The publishing WG should produce a CR in about a year; the packaging spec may lead to a final draft around this summer. So that might work out…
This was certainly not the last meeting on the subject!
(And thanks to Jeffrey Yasskin for having joined the call.)