Hey that's good to hear. I liked your UQM-HD stuff.Kwayne wrote:That "someone" would be me. I've made most of the UQM-HD crew animations, some critters and captain screens too. I still have those workfiles and I can make new material as well. Check http://kwayne64.deviantart.com/
I can guess what could turn out to be confusing. Bits of code that deal with ``searchfactors". Basically, when looking for links to merge the trees, the code considers small inter-star separations, then works up in scale. This is to avoid searching for links between the most separated stars if possible, because that means calculating lots of inter-star distances. So this is why, in the MST loop, searchfactor is increased until it is sufficiently large to allow the MST. This variable influences the search through ``boxminx" and friends. The utility of ``boxminx" etc is a bit different in the closed loop-making part of the code, but essentially it's still there to hopefully make things run efficiently, by not searching over distances that are unlikely to yield links.Dragon wrote:Nice, I had a brief look at the C code. I'm gonna have to hit it with a compiler/debugger to understand it so I'll play with it this weekend.
Also the comment that goes with the line calculating outerdistance actually belongs with the following line, which is Death 999's linking criterion (now changed). The comment with the line setting closedloopmethod says that my earlier opaque friend method for closed loops is an option, but it isn't (now changed).