Volodymyr Agafonkin
Volodymyr Agafonkin
I see the issue now — thanks for the explanation. I don't have immediate suggestions to fix this — this problem might stem from the original tactics from https://www.cosy.sbg.ac.at/~held/projects/triang/triang.html to...
No updates on this yet unfortunately. If anyone has an idea on how to fix the issue without breaking existing test cases, I'll be glad to see a PR!
That's a valid feature request, thanks! My hunch is that `reverse` is not an efficient way to do this — we need to construct triangles in the right order in...
@andreasplesch awesome, thanks for letting me know!
> After getting a triangulation, you can verify its correctness with earcut.deviation: > > ``` > var deviation = earcut.deviation(vertices, holes, dimensions, triangles); > ``` > > Returns the relative...
Oh sorry, I didn't read the PS carefully. I'll think about adding an option to bail out early.
One thought is adding an option to bail out when it reaches the `earcutSplit` phase (which usually means earcut is having trouble and is entering slow desperate mode). Although I'm...
> a PR seems rather excessive Not at all! The biggest advantage is that it will run all the tests against the fix.
This is done for performance reasons. Is this really a problem?
@kylebarron awesome, thanks for working on this! I attempted this in the past but run into the same issue — it demanded a significant increase in complexity and branching not...