[theothermkn]

> That story about Carmack applying cutting-edge academic research to video games has always impressed me.... He deserves to be known as the archetypal genius video game programmer for all sorts of reasons, but this episode with the academic papers and the binary space partitioning is the justification I think of first.

I'm not throwing shade on Carmack when I say that the reason the author states seems to me to be evidence, taken alone, of the exact opposite of the case that Carmack is "the archetypal genius video game programmer." Or, if it is, then it impugns the profession of video game programming, because, in most real engineering fields, a literature review is the first step, not the desperate measure taken after "creativity" has been exhausted. It seems to me that a better case can immediately be made that the originators of the technique are the real geniuses, having come up with the idea, and Carmack just adopted it. Geniuses create; the rest of us adopt, right?

Again, don't get me wrong: Carmack could program circles around me. (Or ellipses, or pentagons, or particle clouds, or fractals, or...) But affirming the consequent doesn't make him "a genius." He may or may not be, for other reasons, but 'applying a known technique' just can't be one of them.

[oscargrouch]

You are forgetting a important detail here. If BSP was already succesfully used in games by a pioneer, you would be right. The question is, to recognize an algorithm and apply it, with the proper changes to a novel field, being the first to recognize this, in my experience at least, is not for the average intelligence.

We all have trouble with tagging someone a "genius", because its not clear what this is, or if we are using this too much and in a wrong way (and i think we do)

But i also need to remind you about the early nineties, and how hard was to get into information, papers and research, compared to now. Also the limitation of the computers back then forcing people like Carmack to use very clever algorithms to thrive. We always need to take the context people were in to properly estimate this kind of things.

By the way, he did not just solved that, but a lot of other hard problems with different, successful outcomes.. So its not just because of the BSP he have this level of recognition.

[wiz21c]

I may be wrong, but IIRC, Carmack used a property of the BSP which is that each each leaf of the tree represents a convex volume. A nice property of the BSP makes it easy to locate in which of these convex volume you're inside. A second step is to compute which surfaces are visible from that volume (you compute that off line). So, each convex volume is associated with a list of visible surfaces. So when you're inside a convex volume, you know which part of the world you're in. The fortunate thing is that these lists are rather short, so the whole structure fits in memory.

And that was very clever (dunno if that was invented by Carmack though)

[fxtentacle]

The fact that he DID a literature review puts him above 99% of his peers in my opinion.

Otherwise, there'd be a lot less many hipster databases who now make the exact same mistakes that the standard databases made 10-20 years ago.