GI Usability Conference and the Last Half Meter

Sorry for the absence of posts, I've been too busy over the last week or so but the activities that have kept me away have generated lots of ideas for posts that I'll be getting to shortly.

What is Usability of Geographic Information?: On Tuesday I was at the second GI usability conference (programme and links). The most interesting aspect of the day was comparing my own ideas about what GI usability is against others. To me, its about how data is visualised on screen which is most of what I discuss in this blog. To others it included data structures, metadata, interoperability, topics I think are important but didn't strike me as being 'usablility' before.

The Last Half Meter: Whilst the other definitions of usability I came across are important topics IMHO there is a lack of discussion/knowledge in the area of GI visualisation. To me the situation is nicely summed up by diagrams like this one:

It's a subset of a data flow schema about how a user queries a GI dataset and it's similar to a number of other diagrams I've come across by techies about GISs. in terms of getting the system to work you can clearly see the technical links, what database talks to what etc. but what is missing IMHO is any sort of consideration about how that data is displayed on screen and how the user will interact with it, if you will, the 'last half meter' between the user and the hardware. I want to know more about what will happen in the red box, for example:

• How many screens will they need to navigate to post a query?
• What happens if their query doesn't produce the answer they really wanted, how will they iterate to a better query?
• How will documentation/help/FAQs be offered to them?

All the above are key questions that affect the overall performance of a GIS/user system but they are represented in the figure by a single arrowhead delivering perfect information to the user in a kind of Vulcan mind meld interaction. I think this shows the bias of techies to want to solve the straightforward technical issues like interoperability or speed of display but don't like addressing the messy issue of how the user interacts with the GIS.

In some senses this analysis is a bit unfair, I'm drawing a big conclusion from a single diagram but my argument predicts that you should see GI projects where they author has invested a lot of time in solving technical issues but not tackled some basic rules of cartography/usability. The Hurrican Gustav project that I reviewed is a perfect example and its by no means the only one.