I'm also seeing this error when importing hou into an external python script (Linux Mint).

Just had the same problem. Houdini 16.0.600

The exported alembic camera was connected to a camera which had been linked via vtorigin()/vrotigin(). Also tried with CHOPs. Both crash on import.

A POP id is unique across the whole sim and is assigned by a simple counter of new particles. The ‘gaps’ in your screenshot are caused by particles dying and being removed.

Your fake id is just the raw point number calculated after the sim and is not time dependent. Your scattered points are not related frame-by-frame so the fake id seems meaningless?

Are you perhaps looking to advect points through the pyro sim ( pop advect / vdb advect points ) rather than scattering?

Thanks for the reply Cameron

2) I'll have a look into it if we decide we do want ragdoll recovery for our project. Is there an RFE for this yet?

3) Excellent, thanks!

4) That makes sense - I've turned off limiting speed to gait and this improves the behaviour I was aiming for - A line of agents of varying size following each other along a path at constant speed, without larger agents overtaking.
I suppose if you choose to retime the clip but not fix it to the gait speed you will always get some sliding, especially if your agents vary in size quite a lot?

Thanks

Hi, I have a couple of questions on crowds

1) Is it possible to set the viewport colour of agents without unpacking the sim? At the moment setting Cd will change the colour of the circle beneath it's feet, but this can be difficult to see in large crowds. Unpacking works but this is very slow.

2) Is there a way to implement ‘ragdoll recovery’ as shown here for golaem? http://golaem.com/content/doc/golaem-crowd-documentation/ragdoll-recovering-animation [golaem.com]
Currently setting a crowd trigger & transition from ‘ragdoll’ to an animated state has no effect and the agent stays in the ragdoll state permanently?

3) Is it possible to set a per-agent avoidance forces/weights so that some agents groups avoid other agents aggressively while others do not try to avoid at all?

4) Varying the agent scale (pscale) seems to affect the speed at which agents move? Why does this happen and how can it be prevented?

Thanks

Perfect - Thank you!

I'll keep in mind the power of group conversion

'Unshared Edges' is not present in the Fuse SOP.

I've had a look at Fuse and this is the closest I can get. Increasing the distance threshold further connects corners together.

I have a simple cloth grid shown in A which has been grouped/blasted/poly-extruded. I would like to reproduce the geometry on a smoothed version of the cloth.

Adding a subdivide after the group SOP produces B, and the Polyextrude produces C (Average Positions) or D (No Sharing). It looks like I need to remove the sharing of vertices at the corners of each square, but I'm not sure how to do this. Any suggestions?