Created as a collaboration between 9 artists, It’s doing it is an online group exhibition of computer generated images that autonomously updates on a daily basis over the course of 45 days. All of the works in the show are instruction-based artworks expressed through computer programs written by the artists. These programs generate new images once daily that can be viewed on the website.
The project is an attempt to emphasise the expressive, data-driven, and intelligent possibilities of randomness for art creation in the context of the digital. Whereas each each instantiation of the show is similar to the previous one and following ones, it is actually unique.
Every artwork has its own folder which holds a file called Metadata.json. It holds all the metadata about a work, some flags and information about how to run the actual application.
A python script that is scheduled to run once a day (using launchd on OS X) then iterates over all the works and runs the corresponding applications to generate the new images. It then does some resizing (i.e. for retina and none retina images), generates the new website and uploads it to the Webserver via FTP. Afterwards it starts another python script that incrementally posts the current works to Twitter. All of this runs on a Mac Mini. For debugging reasons, a lot of information is logged and emailed so the team can immediately see if something went wrong, here is a screenshot of such a log:
One interesting challenge was that a lot of the programs provided by the artists were different. Neil Mendoza’s work for instance is executed remotely and communicates with the main computer that runs everything through dropbox.
Matthias Dörfelt, one of the artists who initiated the show describes being quite happy with the diversity of works the project ended up with, ranging from physical works such as Neils, to purely generative works like Miguel Nóbregas “Metamorphic Drawings” to iterative works such as David Wicks “Push Pull”, text based works like Allison Parish’s “Auto-Minimalist” to data-driven works like Daniel Schwarz’s “Morning Routine”.
Follow the project on Twitter to get a daily update of all the newly generated images throughout the duration of the show.
Left: Day 1 / Right: Day 2