Heh heh Samir, I'm not in San Francisco anymore. I've done the game circuit before and I've got a few arrowheads still in my back from a pack of wild deadlines.
I've got a few dreams of my own I've decided to pursue forever, so I packed up shop, moved it to seattle, and am taking a 'fun' break, to be self indulgent. I'm bulding this model to drum up encouragement for a show I find particularly inspirational, nothing more really...except maybe to show off a little

. I should change my name to "the_big_ego".
For right now in my life, I'm just learning how to write a good script and keep my 3D chops up so I don't get too rusty. I'm collaborating with a few student film makers and donating a little time to other independent projects. I really like telling stories, so I guess I'm driven toward the passive experience instead of an active experience like gaming.
I'll tell you what a bad-a*s dream would be...to work on season 3 and 4 of the big O, and help finish what we all would like to have a little freakin' closure on.
I don't really post very often, but I've spent a lot of the past year reading older threads and going through all of the theories that folks have. It's an interesting community.
Oh, and GOOD EYE R. Daniel! Obviously you do the 3D thing, cuz most folks aren't visualizing the rotational change for an object.You discovered what I found to be a real issue between 2D design and 3D from the original images.
Damn. Busted.
Heh heh, just kidding, I figured out a solution:
The gear itself doesn't rotate, there's a slot down the middle that the cylinders mount to, like an internal gear that rotates on the
inside of the big gear visible.
Because of the compound curvature* that each gear has, the cylinders converge and diverge as you rotate them from vertical to horizontal. They will be slightly angled toward each other when vertical, and angled much more sharply toward each other when rotated horizontally. Based on some guesses as to where the back wall would be, I found that by adding a rotational "gimbal" inset into the end of the cylinder, I could independantly rotate the cylinders, so they're not always parallel with the big gears. This gives the freedom to rotate them along the axis of the gear, and simultaneously animate the individual cylinders staying parallel with the stick track. You can see them in the pics from my previous post with just the back wall, tubes and big gears in it.
It was a wierd thing to deal with, and it took me some time to just decide to take the design liberty to just go with what I wanted.
* The gear has a rotation from global center in Y, and a rotational X axis perpendicular to that Y rotation). Rotating down the gear results in a convergence and a divergence of the cylinders. Also, the dead center for the cylinder will not be in alignment with the gear and still stay in relatively good proportion with the original screen grabs. The gears had to be fattened too disproportionately, and were spread too far apart to keep in line with the originals.
There are a couple of fakes like this in the model, but hey...it works!