Page MenuHome

Find a solution to preview of a part of the nodetree
Needs Triage, NormalPublicDESIGN

Event Timeline

Isn't it possible to just use the thing that node wrangler does for shaders nodes, i.e. CTRL + SHIFT + RMB connects the geometry socket of the selected node to the group output?

I second this re: CTRL+SHIFT+LMB. We're all so accustomed to Node Wrangler and the same shortcut creates a viewer in Sverchok to allow you to preview current step.
I think as long as a Geometry Socket is present, it should just auto connect to the Group Output node.

There also should be a text viewer to check numeric output of math nodes as doing multiple complex transforms get's a bit much for mental arithmetic. I think something like the Animation Nodes viewer is neat and practical for this purpose: {F9817727}

Edit: spreadsheet already fulfills the text viewer comment.

Design ideas for how this could look like:



I think none of the existing solutions tick all the boxes:

I think the pop over probes are a really sharp design for within the node tree. My question though would be, how is this displayed in 3D?
For example if you're putting the probe on a Geometry socket, does it change the geometry in the viewport?
In Sverchok the viewer nodes will output the model as an overlay instead of real geometry so there's no risk of you rendering them and / or missing them in the viewport.

Also what would happen when probing a single value like that from a math / vector math node as opposed to an Attribute?
It could be a good opportunity to have some kind of display attached to the probe as things that aren't on an attribute won't appear in the spreadsheet.
The current method of creating a burner attribute on an Attribute Fill node just to debug math nodes feels like a long way around.

Good questions!

Initially, 3d visualization would be just overriding the geometry output with the probed geometry (see #3). As for more complicated overlays, I think that would fall more under T85653.

As for inspecting non-geometry, I'd imagine in the future the spreadsheet editor will be able to display all sorts of data, not just attributes. So when you inspect a wire, you could see it's "contents" in the spreadsheet editor in a similar fashion. One exciting future possibility would be to inspect a node directly, and see it's inputs, implicit conversions, and outputs as columns all in the same spreadsheet.

However, this is probably something for the far future.

Other usability tidbits, instead of the x icon and removing probes, they could also use the pinning method already present in some editors: