Open ImLucasBrown opened 3 years ago
Hi, first of all thank you for all your work on this project. NXT is awesome! I had a question about this issue.
If I wanted to save out individual nodes to a library and then be able to instance them into my current layer what would be the way to do that currently?
Should I just save all nodes to one layer, then reference in that layer and instance the nodes I want from it?
My end goal would be to have a "Node Library" which are nodes that wrap alot of our common operations we do in our pipeline. Also if I got pointed in the right direction, might be interested in contributing this as a feature when I have some time.
I would appreciate any tips/advice thanks!
Hey @jamesbaber1,
Thanks for your kind words, we love to heard from people enjoying NXT! @MichaelAldrich and I talked about this a little and reasoned ourselves into suggesting that layers are the correct way to do this. We have the concept of builtin graphs (layers) which could be used to share layers with a team of people, assuming you forked NXT or otherwise put your layers into that builtins directory.
Another thing that might help you achieve a node library is utilizing the file fallback token which could be used to point to a "builtin" graph that's stored somewhere externally. The NXT_FILE_ROOTS
env var could store the root path to that external repo of graphs.
Ok great to know, Ill just load up a builtin graph with all the nodes we need for our library. For my use case, I might work on a dockable widget that is a QTreeView of the builtin layers and node names that gets filtered down by a search bar, so we can easily find the node we want to drag and drop into a new graph.
If I do, I can share my fork with the new widget here.
Thanks!
Is your feature request related to a problem? Please describe. Node's that can be instanced at all times regardless of the reference layer stack.
Describe the solution you'd like Perhaps availed via the a "tab" menu. Eventually could replace some builtin graphs.