Shatter periodic nurbs curves

Replies to This Discussion

Permalink Reply by Andrew Heumann on April 20, 2012 at 2:19pm

You've reparameterized the curve for the shatter but not for the divide, so the t-parameters you're giving it are outside of its domain.

Permalink Reply by Jon Mirtschin on April 20, 2012 at 2:57pm

You're right.  I changed my test case just before posting.

I hope I haven't slipped in preparing this example that should show the problem.

Attachments:

• shatter periodic curve.gh, 5 KB

Permalink Reply by Andrew Heumann on April 20, 2012 at 9:09pm

Sorry if I'm missing the point, but I'm not sure I understand where the problem is. For a periodic curve, doesn't this behavior make sense? For an open curve I would expect N+1 segments given N division points, and for a periodic one I would expect the values to be equal. When I open your example, the shatter components given 2 parameters return 2 subsegments, and the one given a single parameter returns one curve. What behavior did you expect to see?

Permalink Reply by Jon Mirtschin on April 20, 2012 at 10:24pm

Thanks, your explanation makes perfect sense.  I guess (after a long Friday trouble shooting nurbs at it's mathematical level) I was thinking about the curve seam already  being a split location.  But I agree with your explanation.  Thanks for taking the time.

Permalink Reply by David Rutten on April 21, 2012 at 6:22am

I was thinking about the curve seam already  being a split location

Of course on a mathematical level it is. But that is just a necessity that flows from the Nurbs definition. For end-users, we try very hard to pretend that a periodic curve has no seam.

--

David Rutten

david@mcneel.com

Poprad, Slovakia

Permalink Reply by Jon Mirtschin on April 21, 2012 at 6:29am

Sure, I think you've done the right thing.  It was just I'd earlier run the openNurbs method Split for a periodic oncurve with a single parameter, and that does generate two curves (and not relocate the seam).   Learn something every day.

Cheers,

Jon