Instead of adding a whole lot about metadata to the introduction, Liddy suggests making a very small addition to the section on "Programmatically Determined" in the definition of new terms:
The term was created in order to allow the working group to clearly identify those places where information had to be accessible to assistive technologies (and other user agents acting as accessibility aids) without specifying exactly how this needed to be done. This is important because of the continually changing nature of the technologies. It is important neither to declare things as accessible because they might be in the future (when they aren't now) nor to declare things as inaccessible in a permanent way when they may very well become accessible in the future.
The use of the term allows the guidelines to identify what needs to be "programmatically determined" in order to meet the guidelines, and then have separate, updateable documents (the Quick Reference, Understanding, and Technique documents) list those techniques and technologies that meet the requirements over time.
I see no need to add anything to this - we would suggest a technique (see below) .... I believe that the adapted version of 'programmatically determined' covers it -
In ISO and DC work we are finding that we did not take account of different writing systems when we were nomimating langauges - perhaps you should have this accounted for in here somewhere? ie, you apparently need human language, country and writing system for languages such as Japanese .... so maybe that is Level AAA 3.1.7??
Guideline 4.1.2
This seems to cover the situation we would be thinking of with metadata so there should be a technique but no change to this Guideline and I am not sure there is any need for an additional guideline but it is necessary for us to think more about it in general.
1. additions to table of contents
- only necessary if there are additions...
Add definition of "description resource":
Note that the new W3C POWDER protocol states "A group of one or more assertions, known as a description, combined with attribution and a scope of resources that they refer to, together constitute a Content Label, also written as cLabel. This must be able to describe aspects of those resources using terms chosen from different vocabularies. Such vocabularies might include, but are not limited to, those that describe a resource's subject matter, its suitability for children, its conformance with accessibility guidelines and/or Mobile Web Best Practice, its scientific accuracy and the editorial policy applied to its creation" so I propose adding the following definition wherever it needs to go.
A description resource is a group of one or more assertions, often known as metadata, that describe the attributes of a resource, either one that contains the descriptions or another. Metadata descriptions often use terms chosen from defined vocabularies. Metadata, or description resources, are machine-readable in as much as they provide values for common terms which may include such as the resource's subject matter, its date of publication, etc. Well-known specifications for metadata include the Metadata Encoding and Transmission Standard (METS) scheme (http://www.loc.gov/standards/mets/) and the Dublin Core Metadata Initiative (DCMI) terms for cross-disciplinary resources (http://dublincore.org). In general, the aim of metadata is to separate the description of a resource from the resource so the former can be searched without the need for retrieval of the resource, in the same way as catalog records are used in libraries. Like catalog records, metadata does not have to be part of the resource but it can be attached to it by being included, like the label on a tin of food, or stored somewhere else for convenience and searching.Metadata is frequently replicated in a number of places so it needs to be exchanged easily. Thus interoperability is a major characteristic of good metadata and so the commonality of the system for descriptions is critical. In the case of accessibility needs, knowing the exact attributes of the resource of concern to an individual might help them choose what suits them even when it may not suit everyone.
2. Additional techniques for 1.1.1 (Advisory) or 1.2.2??
what we want to do is covered by "Providing text and non-text alternatives for object (future link)" and what we would add is that these may need to be discoverable and in that case metadata may be needed .......
and an example might be
A set of identified alternative components
Access to a "Description Resource" for additional text transcriptions (low-literacy, different languages) is available to replace the video of a lecture.
Note that W3C is using the term 'description resource' in the POWDER specs for a resource that is merely descriptive of another, providing the other's identity etc (what is commonly known as metadata). Such a resource may be in a repository, on the Web, ....
3. Additional techniques for 1.2.5 (Advisory)
A set of identified alternative versions of video of sign language
Access to a "Description Resource" for available of sign language versions to enable choice of sign language.
4. Additional techniques for 3.1.5 (Advisory)
A set of identified texts of different reading levels
Access to a "Description Resource" for available texts to enable choice of reading level.Note that 3.1.5 already has techniques that call for Dublin Core Accessibility metadata but maybe the one above useful as well because it is more general for when other metadata schemes are used?
5. Additional techniques for 4.1.3 (Advisory)
Machine-associable alternative: Provide machine-readable information about alternative or supplementary components so users may receive them according to their individual needs and preferences (e.g. language, content, presentation/display, control, etc.) (Level AAA)
5. Additional techniques for 4.1.5 (Advisory)
A set of identified alternatives, accessible to all or only for some individuals
Access to a "Description Resource" for available alternatives to enable choice among them.