by Jeff | Sep 27, 2013 | Axon, DataVis, Design/Development, Featured, Games, Our Games, Quench |
We’ve been working on the Quench editor and pipeline for about a month now, and before anything else, I want to know what I can and can’t get away with doing on an Android tablet. Over this past week I’ve performed a semi-scientific study to identify how best to use the A* algorithm in Quench. AI can be incredibly demanding on CPU resources. Having spent time studying robotics, I know the kind of computational power that often goes into academic robot designs with goals no more noble than ensuring even coverage of a surface by a Roomba. It can be surprising how much computation it can take to do something that seems trivial in the human experience. Quench is going to require that groups of animals have herd-level group-think AI and individual-animal-level AI that result in flocking/swarming behaviour to move as a group and also avoid enemies while finding their way through a map of hazards to reach a goal location. Our plan is to implement the group-level AI as an A* algorithm that runs at intervals to identify a clear path to follow. Flocking requires some further study before we can say how exactly that will work. With these goals in mind, I dove into the first assignment for my AI class at Humber to answer some important questions for myself. I wanted to know what factors cause the computation complexity of A* to grow most quickly, so that I could plan to mitigate or sidestep them. And so I wrote myself a simple A* testbed as a C# Console Application that utilized interfaces to specify the necessary...
by Tabby | Jan 4, 2013 | DataVis, Featured |
About a month ago I posted a little survey to a few sites to see if I could gather some data about Flash developers for an upcoming class I’m teaching. I left the survey open to any dev who wanted to take it, but since I was focusing on Flash in particular, the majority of people who responded did actually use Flash in some way. Over the last month I managed to collect 124 responses (way more than I was expecting) so I wanted to share the responses with the community so that we can all benefit from the (admittedly extremely unscientific) research. So, let’s see what we have. I’ve visualized each question and I will point out my observations/caveats as we go. The survey allowed respondents to choose as many categories as they felt appropriate. The survey was aimed to game developers so I mostly wanted to see how many people self-identified with that description. In retrospect, I probably should have added a fourth category for web developers. Regardless, about 3/4 of the respondents identify as game devs. The vast majority of respondents use Flash for work. Obviously since I was targeting Flash devs in particular, this ought to be the case and doesn’t reflect the percentage of developers who use Flash in the whole industry, even specifically games. But it’s reassuring that some 100+ Flash developers exist and are willing to take surveys! This question was meant to determine if the studio/workplace had more than one Flash developer or if the person responding didn’t use Flash but a co-worker did. In retrospect this question was somewhat confusing...
Recent Comments