tpetrick
The “Geometry” output type stores the geometry directly onto the work item, without writing it to disk. It's kept in memory as part of the work item itself. That geometry can only be used by nodes that support it, which currently includes the Geometry Import and the Invoke node. It's primarily used for in-process geometry processing using the Invoke and compiled blocks. The TOP Geometry in SOPs is also able to pass in-memory geometry data back to SOPs. I'll make you an example today.
There currently isn't a good way to add attributes to the geometry prior to merging it, but I'll log an RFE to add support for that.
Thank you very much for the answer!
I'm doing some experiments at sop level, so I was using the TOP geometry to wedge parameters in a sop tree to create different variations of the same tree without over-complicate things using takes or even to create a full TOP network out of my context. That's why I was needed to have like an identifier to then filter each wedge out of the merged geometry. I make it work but using an invoke over the merge geometry to create point groups or int attributes for the filtering, but it will be nice to propagate task attributes over the geometry.
Thanks,
Alejandro