Matt (Venn)'s project is great - we shared knowledge back in the day since we happened to both be working on the same kind of drawing machine.
I haven't looked closely at the cursive data api, but expect that it could be configured to serve up arbitrary vector graphics instead of data-driven ones.
You could also have a look at the Polarweb project that I did for the Edinburgh Science Festival earlier this year (https://github.com/euphy/polarweb). The interesting bit is that is runs as a web server, drawing things, but when it is finished a particular drawing, it automatically asks for a new image to draw. The actual mechanism of acquiring is not fixed. In the version I have used, I had a webcam looking for faces, and snapped them and traced them on one of the connected machines. You could just as easily have it use something like Beautifulsoup to scrape a website and produce some artwork from that.
It isn't a small task though, unless you've already this kind of thing before. Actually building a list of commands and piping them to a machine is by far the easiest bit, generating the artwork for it is not trivial. I wrote all of the polarweb in Python and Flask. It would be simple enough to have it call out to another http end point to get its next piece of artwork if you did not want to continue coding in python.
The paper indexing - I would love to see that, I have dabbled, but never successfully made anything worth using!
Good luck - and happy to chat more about this next week when I'm back in the workshop.
Sn
|