Comments by
Gordon Grace
‘Data provided on the primary PSI site…should be provided in full compliance with WCAG.’
Applying WCAG to PSI data provision may be a misapplication of the guidelines, which AFAIK apply primarily to web content, web pages, user interfaces, and the like.
W3C’s “Four Principles of Accessibility”
Whilst the means by which someone may ‘access’ or discover the metadata about a dataset on the web should be subject to WCAG2 (the shiny front-end of data.gov.au, for example), a raw dataset should be held to alternative standards (e.g. non-propietary, primary, machine-processible, etc.).
Applying WCAG[2] to a CSV file containing thousands upon thousands of records would make full compliance nearly impossible, even though the data may, strictly speaking, still be Perceivable, Operable, Understandable and Robust.
If a dataset is regarded as a web page, web content, or a user interface, all well and good. If not, then this recommendation (along with the requirement to maintain associated accessibility statements on data.gov.au) may need some revision.
Possible typo?
“…Commissioner are set out *in* Clause 9…”
Q11: Are citizen engagement responses appearing on government websites subject to the Archive, Disability Discrimination and FOI acts? If not, should they be?
RDF and RDFa may also be formats worth considering here – particularly when looking to augment existing website properties.
Having responded to one consultation, a user may be more likely to respond to another consultation. A related consultation should be easily visible at the point of completion or commencement of a user’s response.
“Like this consultation? If you’re interested, we’d also like your feedback on consultation X!”
“Including RSS feeds on the consultation site” may require some additional detail. Is it:
> a feed of all consultation responses?
> a feed of upcoming consultations?
> a feed of responses marked as suitable for publication?
> a feed of recently-closed consultations?
Possible improvements:
- Adopt of standardised means for describing (or at least summarising) upcoming, open, and recently closed consultations
- Ensure that consultations seeking to elicit responses from the ‘general public’ are discoverable and/or aggregated via a [government] website targeting the general public (i.e. not necessarily an agency website).
UK Govt. example of technical implementation of consultation metadata:
http://code.google.com/p/argot-hub/wiki/ArgotConsultation
AGIMO’s current Excellence in e-Government awards may consider adding a data re-use / interoperability criteria or category.
http://www.finance.gov.au/publications/excellence-in-e-government-awards-2008-finalist-case-studies/
Additionally, AGIMO could consider expanding the Web Publishing Guide (or a Better Practice Checklist) to include sections devoted to practical [technical], efficient implementations of web-based data accessibility and data distribution case studies and techniques.
When releasing requests for tender, agencies should consider weighting respondees’ proposals based on the flexibility of data re-use (if new data is to form a part of the work, and if existing datasets are used to inform the final deliverables), and expose this criteria as part of the initial tender documentation.
Initial thoughts:
> Tenders (CC-BY)
> Government Job Vacancies (CC-BY)
> Directories (CC-BY)
> Law (CC-BY)
> Service Locations (Government Shopfronts, etc.) (CC-BY)
> Publications / Research Papers (CC-BY)
Other open database-focused licence mechanisms exist in the form of something like the Open Database License (ODbL), which may be appropriate for some purposes.
The licences make a distinction between the rights for database access (sharing, using), and the rights associate with its contents.