i never looked at PluginBuilder.sln tbh
you can just create the projects directly in vam, like a Custom/Scripts/Case/YourProject1 and then add the references .dlls so that you'll get error checking, autocomplete & references
it will make a YourProject1.cs file, you can delete it and do your own folder structure if you want like
Helpers/
Script/
YourProject.cslist
then after you make a change you just hit reload script in vam and done. i don't think there's any moving necessary.
Okay ... little confused, sorry if my question is stoopid:
You're talking about a "strategy/workflow" at "project level", right? What about the solution that contains the project? Do I need to create a solution ("Solution_VaM_whatever") and add the "VaM-projects a la SPQR" to it? (Not talking about MeshedVR's PluginBuilder.sln)
Thanks a lot for your advice so far, and for the patient explanations! Greatly appreciated - I've "coded" a bit during my time @ University, but that was mostly "texteditor & KDE-console tinkering". With those modern IDE's, I feel like a time-traveller from 1992 ...
If you want to do dll compilation too let me know, there are some extra quirks for that
Fair warning: You may regret having made that offer ...
I think I'm a ways away from the "custom .dll's for VaM" level - but I'm wondering: What are the advantages of compilation into .dll compared to an ordinary "plugin.cs" that's compiled at runtime by VaM?
Better runtime performance?
Asking bcs my "dream-project" would be a kind of spline-library with lots of helper-classes (different spline-definitions, different (global) smoothing options, maybe (global) smoothing with additional boundary constraints, distance of arbitrary point to nearest point on curve, computation of tangent/normal/binormal over whole curve etc). Afaik, the 2017 Unity version doesn't offer much in that regard.
Would smth like that benefit from compilation into a .dll?
And is there a way to tell the system to run code on the GPU rather than the CPU?
(A lot of "spline-related maths" can be formulated as "linear mappings", ie matrix-vector multiplication. That's where GPUs excell over CPUs - also why a lot of physics departments went crazy over CUDA ten years ago)
Would smth like that benefit from compilation into a .dll?
And is there a way to tell the system to run code on the GPU rather than the CPU?
(A lot of "spline-related maths" can be formulated as "linear mappings", ie matrix-vector multiplication. That's where GPUs excell over CPUs - also why a lot of physics departments went crazy over CUDA ten years ago)
Last edited: