"anchor points" on flexible prims! read!

it would be cool if we would be able to create "anchor points" on a flexible primitive, or maybe "pre-made anchor point" or something, like a grid on a flexible prim with a choice on how much "anchor points" we want on the X-axis and Y-axis, with offsets and more options...

ok let me explain this.

by having "anchor point" on a flexible primitive, we can attack a non-flexible prim to a flexible-primitive.

for example, it's like attaching a ball on the bottom of a piece of fabric.

do you know a character called Mai Shiranui (Fatal Fury series, King of Fighters series)?
http://koti.mbnet.fi/jukka_i/gallery/shiranui_mai___by_funeralwind__and_jukkart.jpg

this is a good picture to show what i mean. the balls would be attached at the bottom part of the flexible primitive, and this ball (non-flexible) will move dependently on how the flexible primitive will move (by moving the avatar or with the wind, etc...), like a real object attached to a long piece of fabric would react!

you want another idea? Ivy's Whip Sword (from Soul Calibur series):
http://www.soularchive.jp/img/wp/ivy1024.jpg

the principle of attaching solid non-flexible prim(s) to a flexible primitive and make those non-flexible prims move along as the flexible primitive moves. we will be able to create this great, cool-looking weapon in SL with that "Anchor-point" idea (yes it would need scripting to make it work right though)

we can't do this properly in SL right now, i've seen scripted furry tails around, but they are made so while the tail-wagging occur, the 2 pieces of the tail would move in sync, so that mean complex scripting, when we could have perfect tails with this "anchor points" idea i had.

anyway, that's my idea, i think it would be great to have this building option in SL.

thanks for reading!

Not to mention being able to make a whole new range of clothes, outdoor signs that hang on flexy chains, making them move in the wind, flexy ropes to which you can actually tie stuff without needing complicated scripting.

This is pretty much a summary of what people have been asking for since before flexi prims even hit beta. The main problem is that flexible effects are only rendered client-side and still appear to be solid prims to the server (with the exception that they contain flexible parameters). This pretty much breaks the idea of flexible prims having any sort of movement bearing on static prims without having to completely change how flexies are handled server/client-wise, and implementing your ideas would no doubt make flexi prims even slower and laggier than they already are. I'd imagine making flexies complex enough to be calculated in real-time on the server itself would require a dramatic change to the physics system. In other words, don't expect to see your ideas being added any time soon.

As a side note, flexible tails don't have complex scripts at all. These are simple "hacks" involving two flexible prims of identical size on top of each other with half-alpha textures to simulate tapering. This takes advantage of the fact that two flexible prims of identical size in the same position will move the same way. The only scripts these might include are scripts that randomly "push" the flexies in a specific direction (which AFAIK only involves repeatedly changing the force_vector flexible parameters) to simulate wagging, but they're definitely not complex or even remotely lag-inducing, or even required. Needless to say, your solution would actually be more complex to SL than the current method.

Edit: I think if LL ever manages to make sculpted prims flexible, we'll be able to do a lot of what you suggested, albeit not quite all. It would still be an impressive functionality change for lots of designers.

I definitely think that being able to use flexiprims to physically move other prims is out of the question, there were things called hinges or joints a while back that do this but they were unstable. Maybe if Havok ever gets upgraded we'll see them again in a more reliable and less-laggy form.

However, being able to attach prims and have them move with a flexiprim client-side would be great. They wouldn't move server-side, just like the flexiprim itself, but to the client you could have a line of beads or some-such by using a long, thin flexi-cube with spheres attached. The visual effects you could achieve would be great, and it would allow us to simulate some flexi-shapes we don't have and can't create using current 'hacks'.

Totally forgot about joints; they actually still exist and function correctly last I checked, it's just impossible to make them now. Can't quite remember why that functionality was removed, but no doubt something was broken about it.

As for additional client-side rendering, I can see that causing performance issues on weaker systems. I can't say for sure how exactly the current system works, but the current rendering method can't be too complex. Best I can figure, the server only needs to transmit the flexible parameters and the client handles the rest; any additional influence on how flexies are rendered shouldn't have any more bearing than standard prim movement and wind calculation. Adding extended flexibility would result many more client-side calculations being performed in order to figure out where attached prims have to move along flexible paths. This on top of the noticeable issues with too many flexies in view could possibly end up slowing down the client, and I can picture prims attached to flexies "drifting" away from their parental path.

(overthinking)

Well at it's simplest you'd attach a prim to one of the flexi-prims 'joints' (where it flexes, you see them more clearly on long flexi-prims with the softness turned down). Since these must be calculated otherwise the prim wouldn't be flexi, thus the prim you've attached is simply moved (client-side) to that location. If the joint is not present (ie the detail level has gone down, thus the only visible joints are joint 0, 2, 4 and 6, and you attached to joint 5) then an average of the two neighbouring joints is taken.

As a client-side effect I don't see it being much strain really, so long as it works with the flexi level of detail algorithm. It would be especially useful for creating shapes like flexi-cylinders that end in a sphere, giving the appearance of a rounded end.


As for joints, while some products did use them to good effect, an improperly used one could cause quite significant strain on the physics engine, and could become rapidly unstable (moving in unexpected ways). I think that without them being performance friendly and reasonably stable that removing the ability to create them was a good move as it was too easy to break or abuse them.