Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Tangential circles in a rectangle can break contour filling #303

Closed
natronium opened this issue Nov 18, 2017 · 10 comments
Closed

Tangential circles in a rectangle can break contour filling #303

natronium opened this issue Nov 18, 2017 · 10 comments

Comments

@natronium
Copy link

natronium commented Nov 18, 2017

System information

SolveSpace version: 3.0~13695be0
Operating system: Gentoo Linux

Expected behavior

The contour gets filled in

Actual behavior

This happens:

Additionally, extruding the shape creates naked edges at the points where that same circle touches its neighbours:

All of this is accompanied by the same error that whitequark pointed out in #296:

couldn't find an ear! fail

however, unlike #296, enabling 'force NURB surfaces to triangle mesh' doesn't fix the issue in either case.

Additional information

The file in question: tangent-circles-break-contour-fill.zip

@whitequark whitequark added the bug label Nov 19, 2017
@whitequark
Copy link
Contributor

Yeah, needs a better triangulation algorithm.

@Evil-Spirit
Copy link
Collaborator

Tangential should be handled as error like countour self-intersection. In case of tangential the result mesh will be non-manifold (this is bad for some reason)

@whitequark
Copy link
Contributor

Tangential should be handled as error like countour self-intersection.

I agree. Do you think you can write the fix? I'm not good with numeric code...

@whitequark whitequark added this to the 3.0 milestone Jul 12, 2018
@Evil-Spirit
Copy link
Collaborator

This not look hard to me.

@phkahler
Copy link
Member

Initially could not reproduce using the attached .slvs file of similar construction.
Tangent_Circle_Test.zip

But... It happens if the construction lines attached to the upper right circle-center are H and V constrained. In that configuration 4 Beziers end at a single point where the circles meet. Pinging @ruevs because he's got some interest in NURBS issues and this is a case where something 2D breaks down prior to any surface creation.

@ruevs
Copy link
Member

ruevs commented Jun 25, 2020

Hi @phkahler,

Thanks for pinging me. I know of this one, it is in my list of "interesting things to look at", unfortunately there is always an indeterminately long "someday..." hanging off the end :-)

@phkahler
Copy link
Member

PR #675 only addresses the first part of this, so I'm not going to close this yet. The naked edges remain.

@whitequark
Copy link
Contributor

@phkahler Did you close this by accident?

@phkahler
Copy link
Member

@whitequark Yes by accident, I thought I deleted the reference from the PR before I merged it but this still closed. It's still a good case of naked edge problems though.

@phkahler
Copy link
Member

I've submitted issue 696 which I believe covers the second problem reported here. I'd like to close this, as the title does not match the second issue which has nothing to do with the contour filling and everything to do with coincident edges.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

5 participants