Date: 24 October 2017
Location: Digital Humanities Lab, College of Humanities, University of Exeter
Today I was one of the creative leads at a MozFest Fringe event at the brand spanking new Digital Humanities Lab at the University. Working with students from Exeter Entrepreneurs, we collaborated and played, and hacked our way around code, kit, and play dough.
I took an idea around creating a mobile or VR app to help people experience what it’s like to be colour blind – which I’m not, but have had an interest since Freefall Climate Graffiti – using the jsColorblindSimulator. Early investigations turned up an an existing app, allowing selection from a rainbow spectrum to filter the camera feed. It would be good to be able choose particular types of colour blindness, as eg simulated on Coblis, but we felt that we could probably make more of a difference pursuing other ideas.
Instead I joined Juliane Kahl and the early discussions the team working on wearable tech. They were aiming to create an item that detects and responds to air pollution. So I helped start them on research into sources of air quality observations and forecasts, and accessible air quality monitoring kit. Carbon monoxide was one possibility – it’s a killer and the detectors can be tiny. For example, we found this funky looking shirt that changes pattern with elevated levels of CO. Then I left them in the capable hands of the Python coders to rummage for Raspberry Pis and Arduinos in the boxes of useful looking kit scattered around the lab.
I ended up with JR Carpenter, who brought the idea of hacking some simple Javascript that generates nature poetry. Nick Montfort created Taroko Gorge back in 2009, and it’s been hacked many times since. So a good opportunity to learn a bit of JS, while still being available to answer questions about pollution proxies, and chat about careers and course options.
Taroko Gorge takes simple arrays of subjects, transitive verbs, objects, intransitive verbs, and adjectives, and assembles them randomly. I wanted to see whether I could use the feed of text weather forecasts from the Met Office Datapoint, but rapidly came to the conclusion that parsing the text to create the arrays would be a really difficult problem. Rapid and beautiful failure is one of the lessons of hack days that I was happy to learn! I reverted to entering some words and phrases manually. It also meant I could get a little creative with the text!
Then I thought I’d like to play with the presentation. So I added some code to vary the background colour randomly between greys and blues, and set the text colour to the inverse. That means the text is at times quite obscure and at others stands out clearly. Then I added a transparent cloud as background image, floating slowly in the changing wind.
I also considered using the Shipping Forecast; organiser of the hack-day Michael Saunby reminded me today that the printable low bandwidth version is generative text. But as I set it to Anglican chant a few years back (following in the vocal chords of the Master Singers), I thought I’d go with the regular forecast this time.
I’ve continued to play, and now created “Industrial Revolution” and “Congestion“. Not poems this time. Both variations on a theme, and works in progress. As the atmospheric concentration of greenhouse gases increases on the screen in “Industrial Revolution”, the movement slows; I don’t want this to happen. I adapted the idea and used the slowing as a virtue for “Congestion”, but in this case I want to make the vehicles snake around the screen instead of dispersing. So I need to learn some more Javascript and build some better algorithms.