ZedAI Iteration2 Report

From zedwiki

Jump to: navigation, search

Contents

Summary

This is the report from ZedAI WG Iteration 2.

The iteration time span was 19 December 2008 - 17 April 2008. Iteration deliverables were defined in the Iteration 1 report.

The WG decided to extend the time span of this iteration (originally intended to be closed at end february) to satisfactorily complete the iteration deliverables.

Deliverables Report

This iteration includes the following main deliverables:

Profile composition

The profile composition mechanism was refined by introducing the notion of a profile being composed by using a core grammar and zero, one or several optional features. An archetypal feature is MathML support. Note - features are not tied to specific profiles, but intended to be usable by several profiles without modification.

Book Profile

The core grammar and optional features approach means that the previous iterations concept of simple and and advanced book profiles is no longer used. Now there is only one book profile, which instead is varied along the axis of which of the optional features are actually used. As a service to users who are not interested in particular optional features, informative schemas that do not include the optional features are autogenerated.

The second draft of the book profile includes many additions and new features. Please refer to the book profile schema documentation in the distribution zip for details. Notable additions are:

A sample document (nota bene: not exercising all components of the grammar, and still with several suboptimal solutions) is available at the samples directory of the GoogleCode SVN repository.

MODS metadata

RDF Vocabularies

An RDF vocabulary primary targeting print book semantics (referred to for now as the "book vocab") was created. Terminology from Chicago Manual of style was used where applicable. The vocabs serve as the defining instances for values to use with the Role attribute, and when using RDF/a to add inline metadata. The vocabularies are automatically converted to schema enumerations to simplify the manual authoring process.

It is expected that the vocabs will be broken down into several smaller vocabs with a more limited scope, to simplify reuse cross profiles.

Newsfeed profile

A first draft of the newsfeed aggregation profile was created. Schema and documentation is available in the distribution zip. Metadata and content model is still under development.

Interactivity feature research

Research on technologies to use for authoring of interactive documents (educational testing purposes being the main target) was performed. Please see a separate report on this research (and note the example section at the end).

Specification Documents

A first set of skeletons for the specification documents (main spec and a primer) was drafted. These are included in the distribution zip.

Schema Documentation

A system to generate end-user documentation from a combination of the normative schemas and addon prose has been created, and is run automatically when one invokes the build script. The result can be reviewed in the schemadoc subfolder of each profile folder in the distribution zip.

Iteration zip download

An archive with schemas, documentation and specification documents can be downloaded at http://zednext.googlecode.com/files/zednext-iteration-2.zip

Risk analysis

W3C rechartering of XHTML2 and HTML5 WGs
There are discussions ongoing within the W3C on rechartering of the XHTML2 and (X)HTML5 WGs to resolve some percieved overlaps and conflicts between the two groups deliverables. There is a risk that this discussion results in XHTML Modularization 2.0 and XHTML2 no longer being usable as the base framework to use for profile composition, at least not given our 2010 timeframe. Risk level: high. Resolution: ZedAI WG lead to produce and propose a fallback strategy.

Resolved Items from Iteration 1 Issue Log

0-7 Clarification of uses of roles vs. elements
See ZedAI_When_to_use_Elements
0-2 Schema Inline Documentation [Category - Schemas]
See Schema Documentation above
1-1 Content model for frontmatter
A proposed solution for the frontmatter book profile is available in the book profile schema, and can also be reviewed in the schema documentation. The content model is based on using the generic section element, specialized using the role attribute.
1-2 External object inclusion
The WG decided to use the XHTML2 Object element exclusively for external object inclusion.
1-3 Attribution
In the schemas as of this iteration end, the WG settled for the third of three options evaluated for attribution. A typical markup consequence of this approach can be reviewed in the frontmatter section of the available book profile sample.
1-4 Multi-file documents
ZedAI documents may employ XInclude to make use of multiple files during authoring and and interchange stages. Logically however, a ZedAI document is to be considered one infoset.

Iteration 2 Issue Log (to be solved in iteration 3)

Inclusion model for MODS (document-level)
Which is the most appropriate method to include MODS metadata in document instances? The options are discussed in MODS Inclusion Mechanisms
MODS inline using RDF/a
Can atomic MODS properties be used with RDF/a for inline meta annotations?
Autogeneration of XSD
The current autogenerated W3C XML Schemas have some errors (primarily UPA violations). Can this be fixed, or will we eventually have to manually produce XSD schemas (for those profiles where we decide that the provision of XSD is a good thing)?
CURIE prefix binding
Await changes in W3C space to define how CURIE prefix binding (including the default prefix) is to be done.
RDDL or XHTML+RDF/a
A better grammar than RDDL to use for online Resource Directories may be XHTML+RDF/a. This needs to be researched and resolved. We ideally want to use the same langauge as chosen for the DAISY Online Specification (there in the context of service provider discovery).

Deliverables for the next iteration

Iteration 3 time span: 17 April 2009 - 15 May 2009

Main deliverable: a first public working draft of the Z39.86 Authoring and Interchange Framework

Fallback plan
Compose and propose a fallback plan, as per the risk analysis
Testing of book profile, creation of sample documents
The testing subgroup within the ZedAI WG performs testing and further sample document generation.
Third draft of book profile - based on input from testing team
A third draft, mature enough for public review, is produced.
Interactivity feature
A first draft of a feature to support authoring of interactive documents is provided, and incorporated into book profile (see ZedAI_XForms_QTI).
Newsfeeds profile second draft
Includes: addition of source/origin metadata, test, production of samples for all use cases
Periodicals profile, first draft
A first draft of a periodicals profile (superset of newsfeeds profile) is created.
RDF vocab granularity
Define and implement an approach for more granular RDF vocabularies
Further work on profile documentation (RDDL-level and schema-inline)
Documentation of the schema components is taken to a level where it is more or less complete for the profiles which are not in first draft stage (book and newsfeeds profiles).
Usability review of schema documentation
A usability review of the schema documentation format is performed, and fixes/enhancements introduced. Screen reader access optimization is included in this review.
Further work on spec and primer documents
The spec and primer documents are taken to next draft stage. At the end of iteration 3, they will still not be complete.
Establish daisy.org/z3986/2010 web pages, and build upload mechanism
The webpage area (www.daisy.org/z3986/2010) is populated so that the public review can be made through accessing the documents from there.
Review of Generic Document Profile, and decision on whether to include it in the initial profile collection.
The existing draft of a Generic Document Profile is evaluated, and a decision is made on whether it should be included in the shipped profile list.
Tracker issues
Solve as many as possible of the lower level GoogleCode issue tracker issues, with prio level medium and high.

If time allows, work also on these higher level entries:

  • Research on content selection (DISELECT) module
  • Research on speech related (PLS(as link type)/SSML) modules
  • Early drafts of poetry and drama modules
  • Research on computer code markup feature (possibly based on DocBook5)
Retrieved from "http://www.daisy.org/zw/ZedAI_Iteration2_Report"
Personal tools