Attending: Dalai Felinto, Hans Goudey, Jacques Lucke, Pablo Vazquez, Simon Thommes.
Agenda and notes for meeting at 14:00 CEST on blender.chat. Interested developers and artists are welcome to join.
- Confirm priorities for the second half of the current sprint
- Finishing material support.
- Finishing curves for the nest use case.
- Initial attribute list for non-point cloud geometries.
- Technical design for the advanced attribute list.
- Support Fabian on design feedback for the data-set region patch.
Discuss patches / design as necessary
Delete Geometryit is.
- Look into displaying menu entries in node search (e.g. “Mesh Primitives > Circle”).
- Viewer Node
- Why not to use Viewer Node instead of the per-node icon for the spreadsheet?
- Including Ctrl+Select to cycle through the outputs.
- There can be viewer nodes for attributes (or specific attribute drawing settings).
- Attribute Removal and explicit persistent attributes
- Why are the reasons we need an Attribute Removal node for?
- Without a way to explicitly tell local/global attributes apart there is definitively a need for attribute removal.
- It will be easier to reassess this after we have attribute processor – when the need for removal is less common.
- Jacques and Hans think having local/global attributes (work vs stack data) may complicate things.
- Attribute removal is secondary, the main topic is whether it should the separation between local and global attributes should be explicit (Dalai thinks so).
- Ideas on where to make an attribute explicit (global):
- Create Attribute Node ([x] persistent]).
- Node Group Output node (list - or [x] all?])
- Force them to be created outside (similar to how mushroom masks are required for nodetrees - and where node tools could help).
- Selection is fine and Point Separate should follow suit.
- Socket colors
- Pablo will write exact colors in the task (T88176).
Next planning meeting: 2021-05-17 at 14:00 CEST on geometry-nodes on blender.chat.