Flexible prims and their attached prims

I'm trying to make a very simple little forked tail for myself. I figure it should be straightforward -

- create a long thin cylinder
- create the "fork" from a pyramid (or whatever)
- position the fork, link the items
- add flexibility to the cylinder

But when I add flexibility to the cylinder, it detaches itself from the 'fork', and wobbles around on it's own. Now, a floating fork looks cool, but it's not exactly the effect I wanted. It won't let me add flexibility to the whole item, I have to add it to individual prims.

Suggestions?

Unfortunately, you can't connect anything to the flexible end of a flexi-prim. Linked prim positions and orientations are referenced relative to the root prim, and only change if given a command to do so such as llSetPos. This is also entirely server-side.

Flexibility is a client-side effect, meaning that all the calculations for how the path of a flexi-prim moves are handled by individual users' computers, not the server, and this information is not communicated back to the server, since it would obviously vary to some degree from one client to the next. This is necessary for flex hair/clothing/attachments, since avatar animation is also entirely client-side. The server doesn't know where that loose flexi-prim end is, only where the non-flexi end is and how it's oriented, relative to the root.

This would also require "heirarchal linking", i.e., parenting specific prims to other specific prims in a link set (and in the case of flexis, where along the flex path it's linked). As it is now, there is only a single parent and the rest are it's children.

[Edit to add:]
Although... you can probably fake such an effect by having 2 linked flexi prims of identical length, flex properties and positions. Make a long skinny cylinder for the tail, and a cone with a very wide base for the "fork" at the tip. Apply an alpha texture to the cylinder which makes all but the desired tip length opaque, and the opposite to the cone, making only the tip length opaque. The two will flex almost identically, so the visible tip of the cone will appear to follow the end of the cylinder.

On a related note, Only cylinders and boxes are flexible, is that ever going to change? (like spheres please?)

hi

I used this technique, but they do not move identical. You always notice that there are two different movements. But for some purposes this technique works.

Meni.

Prisms can be flexi too.

But as to spheres, well, that'd be tricky. First off, look at how flexibility is currently implemented: It's a flexible path.

Boxes, cylinders and prisms are linear extrusions of squares, circles (well, technically, icosikaitetragons, at most ) and triangles, respectively. By flexing the path these 2-D shapes are extruded along, the entire object is affected, somewhat like flexing the spine affects the ribcage. When one of these prim types is made flexible, the beginning of the extrusion path (the local -Z end) becomes fixed to the prim's position and orientation, and the other end becomes freely flexible, according to the flexibility settings and the prim's motion.

On the other hand, tubes, torii and rings are extrusions of these same shapes, but along a closed circular path. Spheres are similarly produced, by extruding a half-circle shape around a closed circular path. How would these shapes (the sphere in particular) be made flexible by manipulating that path, without destroying the essence of the shapes they're supposed to be?

Good point, but I wish you could give a half-spherical "cap" to a cylinder