I have been aware of Mitchell Whitelaw’s great work on generous interfaces for a while now, so was interested when a colleague recently sent through a link to Mitchell’s 2011 National Digital Forum presentation Generous Interfaces.

The particular example my colleague highlighted from this presentation was Discovering Mildenhall’s Canberra: Photographs from the birth of our capital. And in many ways it is indeed ‘generous’. The different ‘Explore’ options are great, the interface is easy to use, and (with reference to Mitchell’s criteria) the presentation and design allowed me to quickly and easily get a sense of the collection contents.

Even better (and sadly still more rare than it should be) the URLs for each image look concise and unique. Check out this one: http://mildenhall.moadoph.gov.au/photo/4634

But, while the interface is generous, I have a significant issue with the Mildenhall site: stingy metadata.

First, why is it so hard to get from a digitised photograph to the full catalogue record? The National Archives of Australia series and control numbers are provided with the photograph, but they are not a link or a pre-packaged search. Instead, I have to highlight the series number, copy it, then go to the National Archives of Australia site. Once there, the (not very generous) NAA search box on the home screen won’t bring anything up if I paste in the series number. I have to navigate to the RecordSearch – Advanced Search screen and enter it here.

Yet, taking the link above as my example, when I do reach the full catalogue record I see plenty of metadata I can’t get via the Mildenhall site. The picture I was looking at is part of a series of glass plate negatives – useful format information, as the digital image gives me no sense of the physicality of the original. I can also see the dates provided via the generous interface are the ‘Accumulation dates’ (January 1921 to December 1935) rather than the ‘Contents dates’ (January 1913 to December 1935), which could create some doubt in a user’s mind about the accuracy of the date range provided on the Mildenhall site. Perhaps the ‘content’ of the photograph in question falls into the 1913-1921 bracket? A search for the individual item shows the date of the photograph in question is 1929, which is much more useful than a fourteen year range. And the catalogue record goes on.

Yes, I am a dedicated metadata geek. But beyond that, the value of the generous Mildenhall interface would be greatly enhanced if they were equally generous in providing access to all the available metadata, either as a collapsed or tabbed section in the interface itself, or (at the very least) by providing a single active link to the NAA catalogue record.

My second main issue with the Mildenhall metadata is the stingy ‘citation’, which isn’t really a citation at all. There is no title, no series or collection name, no reference to the photographer, creator or collector. What is provided is a reference to an institution and a couple of control codes (which, as noted above, don’t get the user very far without some manual graft).

More to the point, even a full citation of this type would be problematic. The adage ‘cite what you see’ should apply here. Using the same image as an example (http://mildenhall.moadoph.gov.au/photo/4634), what I see is not glass plate negative 5005 in series A3560 of the NAA. I see a titled digital image of the negative, on a site called ‘Discovering Mildenhall’s Canberra’, with a concise, unique URL. Therefore, providing a citation should include the photo name, the name of the site, the URL and the date accessed, as well as the National Archives of Australia series number and control symbol.

You may wonder why I am picking on Discovering Mildenhall’s Canberra. That isn’t my intention. There are many worse sites out there – stingy or confusing interfaces, with missing control codes, poor functionality, or just plain ugly design – which deserve to attract much more criticism. Here the design is great, and Mitchell is right, the interface is generous. But when drawing from well documented archival collections these things should not come at the expense of scholarly and archival qualities such as access to complete catalogue entries and accurate citation information.

I think what bugged me about this particular case was that the creators got so close to creating a fantastic model of generous design, only to be let down by stingy metadata.