Sunday, November 19th 2017
Officially in Major Studio I, this week was "User Testing" – but, given the nature of my project (a non-interactive piece that sits somewhere between art and technical research) traditional user testing was not appropriate. Instead I talked to fellow DT students and friends, for a few rounds of feedback "critique-light" that may (I made no promises!) inform the piece.
I distilled the feedback into a few main points, as I don't have time accomodate the outliers:
This was something I ran into several times. I think my prototypes look so underwhelming running on a little laptop screen stuck to a trackpad and keyboard, that everyone's first instinct is to play with it, and it is frustrating to them that they can't.
Solution: I will present this piece via projection in order to reduce the expectation that it is interactive.
What they're describing here is a different project. It might be interesting, but it isn't this one. I set some design values up front – this is a procedural/generative world, not a parametric one.
Solution: No action at this time. I'll just have to defend my decision, right or wrong, in end-of-semester critique.
I agree with the point here. The aesthetics of the piece are tied to my own capabilities as an artist. I found it impossible to pick up 3D modeling in the time I had. Again I'm wrestling with the fact that I truly had zero experience with art and design prior to beginning this program.
One suggestion was to just download more attractive 3D assets from the Unity asset store. My gut reaction is that this feels like cheating.
Solution: If possible, drive out a more attractive 3D prototype using preexisting assets and compare the two.
I've been thinking a lot about this one. The Nature of Code mostly deals with big clusters of nameless, featureless boids, and that's the book that I'm the most familiar with. Conversely, even with zero graphics, Dewdney's Dinner Simulation gives everyone a name, and even a little backstory!
Solution: This is where it makes sense to stay 8-bit for me. It makes it fun, and easy, for me to drive out basic characterization. I'm thinking of naming the agents in the game from a large text of possible names, and either keep them hovering above the figures or at least show them when the camera focuses on each one. I also believe camera framing may allow us to begin to identify with particular agents, especially if each one is unique in both their visual style and AI motivations.
I ported an earlier algorithm (resource gathering) over to Unity. I'd never used it before and I'm out of practice with big, heavy GUI interfaces so I struggled for a bit.
It has some technical and visual hiccups but it successfully marries 8-bit graphics, a moving camera, A* pathfinding, and an earlier algorithm all in one program.
Here's the resource gathering algorithm I used to give my A* pathfinding a series of targets:
I reworked my proposal to better reflect my project's final argument. There is a lot of work I still need to do, because I still need to drive out the details of my project, which will inform the paper.