Visual.ly - Drew Skau

In our talk with Drew the one thing that struck me the most was his description of the visualization tool that he was working on. While I expected a more ambiguous answer, he definitively said that it was designed for someone with knowledge about the meaning of the data. While at first this statement seems like a clear focus on one particular user group, I think there a few things going on in this statement. As a tool focused on scientists that have domain knowledge it can either be a flexible presentation tool, or an exploratory visualization tool like Tableau. While the difference between softwares is not a black and white distinction, Drew’s tool doesn’t not seem to be exploratory. Instead of interacting directly with the data, the user applies visual filters to a dataset data. Thus, it seems that Drew’s software is focused on consolidating the workflow between scientist and visualization, where the scientist can make their own visualizations instead of relying on a designer to construct the visualization.

On the other hand Bret Victor’s software seems clearly focused on the designer and not the researcher. While it may seem to occupy the same space as Drew’s software it allows for much more manipulation of the visualizations that is probably outside the scope of knowledge of a researcher. Take his flexible grid example, where the entire grid is based off of the dimensions of a rectangular primitive. Researchers may be able to take advantage of this, but the thing that is more important to them is control of the scale of each side of the graph. If it were truly focused on visualization production from the research perspective the grid would be constructed from two variables that each define the scale of the x and y coordinates.

Both Bret and Drew approach the challenge of constructing visualizations but come as the problem from opposite ends of the problem. Drew focuses on the production of visualizations from the perspective of the researcher and Bret focuses on the ability to create flexible, parametric visualizations without a specific user in mind.