Is there currently a way to set a per-linestrip point budget for the Resample POP? Either by a combination of the line POP operations or the GLSL (advanced) POP?
When I have, e.g. 2 linestrips with 10 points each and a small max distance, two points on a linestrip with a large gap can eat up the entire vertex budget, starving the other lines. Is it possible to set a max number of vertices per linestrip rather than a global budget?
I’m also interested in a solution to this. I’ve encountered this issue when exploring Differential Line Growth, and both my example, and other DLG examples (like Mini UV’s) encounter the problem of line strips disappearing from the sim.
I think it would be worth adding features to the Resample POP to manage this better, though some insight into accomplishing this in the Advanced GLSL POP would be helpful in the meantime!
I’ll add a note for the Line Resample POP. Should it switch to a larger spacing so it can still distribute points across the whole line strip, or should it just stop adding points once the minimum‑distance rule can’t be met anymore? It’s definitely something that could eventually be added.
It should be possible to replicate some of the Line Resample POP’s behavior with a GLSL Advanced POP, but it would likely require a fair amount of work. You’d still need to set an absolute max vertex count, but you could assign a budget per line strip.
On a GLSL operator running a thread per input primitive, some line strips metrics could be fed in and the number of new points per input line strip would be outputted based on the minimum distance and the point budget per line strip.
From there, other GLSL operators could fill the vertex and point buffers using those per‑strip counts. You’d also need to generate a line‑strip info buffer, a per‑vertex line‑strip index buffer, and buffers for strip counts and per‑strip vertex counts. A Topology POP can then assemble the final resampled strips. The LineStrips GLSL Advanced POP snippet shows the structure that the Topology POP expects.
Preferably, distributing points along the line even though it can’t adhere to the max distance constraint. Is this how it currently works too?
Yes, this is how it works currently. When the maximum number of vertices is reached while subdividing a line strip, the distance between points starts to exceed the max distance constraint.