Why are DOP networks not a separate network like POP networks?

[magneto]

I am watching Houdini vids and I forgot which one but in one of the vids he mentioned that they are separate but not why, when he used some DOPs. I see that they are inside the SOP network inside a container(?) called AutoDOPNetwork but why are they like this?

I don't know enough to even guess the reasons, but if they had to be part of an existing network, shouldn't that be the POP network?

I am just curious about why it's designed like this, vs DOPs having their own separate network.

Thanks guys.

[ehsan parizi]

you can have multiple dop networks! I think only when you use the shelf tools and it creates the autodopnetwork, everything you add FROM the shelf, will be added to that autodop, but if you make your own dopnet, you can have more than 1 dopnet

[magneto]

you can have multiple dop networks! I think only when you use the shelf tools and it creates the autodopnetwork, everything you add FROM the shelf, will be added to that autodop, but if you make your own dopnet, you can have more than 1 dopnet

Thanks Ehsan, didn't know that was the case. Now it makes sense

[hopbin9]

I don't know enough to even guess the reasons, but if they had to be part of an existing network, shouldn't that be the POP network?

POP networks deal with points only, and designed to work with Point attributes. You can't cook out geometry in a POP network only points.

Where as, DOP networks work with geometry and can access primitive attributes and point attributes. DOP networks can cook out new geometry (like in fracturing dynamic objects or ripping up cloth).

So DOP networks can exist in SOP networks because they work with the same data, but POP networks have their own network.

Like the above comment, AutoDOPNetwork is something managed by the shelf. If you delete some of the merge nodes, then the shelf won't work anymore. So you have to follow a node connection pattern for the shelf to continue working in the AutoDOPNetwork. If it stops adding to the network, just rename it, create a new AutoDOPNetwork and copy over the nodes you need to the one you renamed (make sure your Node paths are adjusted if needed).

The AutoDOPNetwork can be confusing when you first start in Houdini, but after a while it'll start to be clear what's going on.

[magneto]

Thanks hopbin9. Awesome reply. Didn't know these