I recently came across another case where a long flexible attachment could be usefully defined as ONE segment with the "curve" property, in Poser 5 if the user could define its direction and attitude at various points along its length. Poser 5 could then use use one of the many readily available curve fitting algorithms to fit a route through all those points. The example is the two long tentacles of the Architeuthis dux (giant squid) model, which (according to a description) in the Poser 3/4 version have to be in 17 segments to make them flexible enough. One advantage of having the tentacle as one segment rather than 17 would be that when he wants to change its length (as the squid can easily and quickly do in life, very much so), he only need scaleX (or however it is done) in that one segment instead of having to ferret for 17 segments one by one and scaleX in each. Also, he would be able to move one of his fixed points without having to re-pose everything distal to that point. Possible algorithm: - Hereininder for "R" read whichever of x y z is the part's rotation axis at its proximal joint. - F0 is the part's proximal joint center point. - The user chooses fix points F1 F2 F3 etc. - Each of those points contains x y z (position) and xrot yrot zrot. - L = the part's length along its R-axis, as set by its geometry and its current scaleR setting. Poser 5 could do this:- - At each point, find from xrot yrot zrot: vector P = {Px Py Pz} = direction of the part's R-axis at that point; angle T = how much to twist the part about its R-axis there to get its desired attitude there completely correct. - Leastsquares fit a curved path from F0 through F1 F2 etc, using each point's x y z and P. - Lay the part with its R-axis along that path as far as its current length L will allow. - Twist the object to fit the T values at all the points that it runs through.

It would also be useful in more common objects with several long "tails" of "curve"-d segments, e.g. anything with backpack straps such 3 devices that I put in the Poser Funb Stuff (text search for "appleyard"); and I am planning a scuba set, where the same problem may arise. So far in posing backpack straps I must wrestle with a long serpent of many jointed segments and find that it has many of the properties of a large uncooperative python.

P.S. With two of those devices I had the same bout of python-wrestling trying to pose a hose that runs from the backpack part to the handheld part.

Also, having each strap or hose or whatever as one segment rather than one would make the model's .CR2 and .RSR files much shorter and much lessen the load on Poser of having to keep track of at least eleven settable parameters (some hidden) at each articulation.