HI!
I tend to work in a way that allows me to prototype in a Test folder outside of the scope of the project. This means I can work non-destructively on assets while I am prototyping. Typically I'll build my shaders like this, in a silo and when they're ready to integrate back into the project I'll copy the data into a newly created shader that can then be wired up.
I should preface by saying that I am working on OSX.
First, when copy\pasting nodes from one shader or function to another I lose all of my custom Property and Variable Names. It's a huge pain to have to go back to the original shader I prototyped and start re-typing in all of the properties and variables. As a result of the variables, typically none of my Variable Get nodes retain their connections, therefore I also have to go back and start determining what was wired where. Some of these shaders are a little complex. Would you guys be able to resolve this type of issue?
Second, if I have nodes selected in the graph and then grab the Graph Menu Bar to move the window to another location on the screen... Whatever nodes are actively selected will shoot off into the ethos of the entire graph in the direction that I pull the window. This can be frustrating from a workflow perspective because I tend to spend a bit too much time trying to get my graphs all neatly arranged and lined up so they're easy to read. Have you encountered this? I can email you a video if you'd like since I can't attach one on here.
Thanks!
Seth