Hi everybody,

newbie here, learning Houdini “again” after a one week course in… what was it… 1996? It even had a different name at the time… Prisms.

Anyway, I'd like to understand better the modeling workflow in Houdini and how to take advantage of its non-destructive capabilities.

The concept of having previous nodes one can go back to and modify, affecting subsequent nodes, it's clear. And in many cases it certainly works. But as soon as one starts manually adding edgeloops here and there, to get more detail where needed, the IDs of vertices, edges and primitives can change quite drammatically, negatively affecting nodes down the line.

I can see how nodes generating groups can mitigate these issues. But sometimes you are not interested in the named edgeloop group, rather you need one quad between the new edgeloop and a previously existing one.

I was wondering if, well, yes, that's the way it is or if I'm missing something fundamental in the workflow.

Kind regards,

Manu

sops that create geometry should have the option to automatically group what was created, by geometry type, but this hasn't been fully implemented. I think polyextrude is close.

There's a few other issues with sops that can bloat or break a procedural setup, most already reported, just have to wait.
In the meantime its probably cheaper to have a specialist modeller re-build something over and over again, rather than invest in a procedural build.

Ok cpb, thank you for your answer.