Monday, May 12, 2008

Trouble in Polygon Optimization Town

Apparently my Inventions and Great Ideas post stirred some mud. I've gotten a few comments and some e-mails. E-mails from people conveying their experiences and also one from an engineer at Right Hemisphere, with a marketing director's e-mail as a CC no less. Hmm...? Turns out some competitor to Right Hemisphere read into my post a little to deeply and wrote a blog post of their own, claiming that I allege my ideas were "stolen". This isn't true. The point of my post wasn't to scream foul or claim theft, it was that I learned a valuable lesson, a lesson regarding my "ideas". I freely gave my idea, a single idea not ideas as the blog suggests, to an engineer at Right Hemisphere and they listened. Is that so bad? No, in fact its good business. And when they came back, they pitched the idea back at me because of my unique problem and its unique solution. The point I was trying to make was that I hadn't thought out the situation thoroughly and I gave up my idea to someone who could use it to their advantage, potentially with no traceback to me. The idea was anything but stolen.

In fact, I contend that the idea still hasn't been implemented. I know of no tool or plug-in, for Right Hemisphere's Deep Exploration or any other, that does polygon reduction the way I explained it. One person responded suggesting my idea was nothing new, as its just a "Hierarchical Z-Buffer". However, what they failed to realize is I'm working in the modeling side of things, not the pipeline. I "concepted" a way a tool could analyze geometry and throw out useless polygons, similar to a runtime cull, but applied to the polygons in the geometry. My intent is to reduce and remove polygons from the geometry before it enters the pipe.

I feel I should mention some details about the second meeting since the situation was somewhat dynamic and beyond the scope of my original post. Really it is beside the point entirely but I feel that I should elaborate given the amount of attention I've received. I mentioned that the original man I explained my idea to didn't work there anymore (presumably onto greener pastures) and I did so because it is truth, but also because it limited my options and perceptions at the time. Since he left the company, their was no option for me to contact him about the details of the tool/idea implementation, which was partially why I described the situation the way I did. Essentially though, the sales-engineers were listening to me when I explained what I do and how I foresee using their products. One of them was astute enough to recognize my situation and realize that something they had in development (or whatever phrase is appropriate) would apply to me directly in a positive way.


2 comments:

Anonymous said...

Thanks for your clarification on your original post. I find it interesting that the mud has been stirred so deeply, since I merely referenced your post to make a totally different point about Right Hemisphere.

Anonymous said...

By the way, if you combine occlusion culling with selective decimation in 3DVIA Composer, you can get similar results to what you are looking for:

http://www.3dmojo.com/podcasts/3dvia-your-ipod-67-selective-decimation/2008/04/04/