You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In version 1.0.0, iOS 15 Safari shows some overlapping pixels where caps meet segments. I'm not sure what's causing this, unless maybe triangle strips require knowledge that it's a triangle strip in order to avoid? I thought as long as the tris met floating-point-correctly that the GPU had some magic way to get this right.
At any rate it's not a blocker and there are bigger self-intersection issues with transparent lines anyway, but it'd be nice to understand.
The text was updated successfully, but these errors were encountered:
Ah, the existence of these on the no-caps test suggests it's a result of degenerate triangles still resulting in a couple rasterized pixels. I think they should be degenerate to within floating point accuracy, so maybe this is just the way it works. https://rreusser.github.io/regl-gpu-lines/docs/tests.html#miter/insert-caps/none
Only seems to happen for endpoint caps. Full opacity or depth testing are not subject to the problem. Backface culling cuts it in half.
rreusser
changed the title
Caps overlaps segments on mobile
Degenerate triangle show up on some devices
Nov 20, 2021
In version 1.0.0, iOS 15 Safari shows some overlapping pixels where caps meet segments. I'm not sure what's causing this, unless maybe triangle strips require knowledge that it's a triangle strip in order to avoid? I thought as long as the tris met floating-point-correctly that the GPU had some magic way to get this right.
At any rate it's not a blocker and there are bigger self-intersection issues with transparent lines anyway, but it'd be nice to understand.
The text was updated successfully, but these errors were encountered: