4 point surface gives more than 4 edges

hello,
I’m in the middle of an experiment to create a planar quadratic surface in a specific way.
My problem is that although I have the facets created with the 4 point surface component, by coplanar points (checked) they are sometimes divided into 2 faces.
as a result, a mesh that is supposed to have 12 pq faces, ends up with 22 faces
What am I doing wrong?
deterministic2.gh (50.8 KB)

Can you please extract and post the script to check for co-planar? Sometimes tolerance value make a difference.

I’ts not a script just a cluster, if you double click on it you will see.
(the problem is that the points are found ‘mechanically’ through consecutive intersections of planes, lines, etc)

If you can isolate the code for calculating the co-planar points (just use any 4 points as input, then I can help you debug it.

Just do that ! It seems to work

1 Like

Something crazy happened!
I internalized two series of points in the data components you see and once i copy-pasted them on a new gh file to send it to you,it magically worked! I am buffled!
(could it be some kind of bug with the 3 loops of anemone that I have?)

you rock! thank you!!!

NEVER underestimate the power of the White Side (the internalize bit).

BTW: Not an issue in this case … but … If crazy things happen and/or GH/R crashes at opening etc etc, disable solver and remove all things that hold internalized stuff (especially breps).

BTW: You may want the “‘beams’” as well ??? (BUT … are they the right ones?)

LongIsThePathAndHilly_V1.gh (14.2 KB)

Well for the lovers of absolute zero I have good and bad news.

Open and enjoy

LongIsThePathAndHilly_V2.gh (22.1 KB)

Other than the zero thiingy > raise the D > what happens if the offset Mesh is an offset mess?

Moral: life sucks

BTW: Spend 5 minutes more on that one by adding some “visual” aids in order to get the gist of the offset mess (blame Karma and D) issue. So take a (hopefully very long) break from Platonic Love Affairs and address the real-life challenge (for the brave) > see inside def.

LongIsThePathAndHilly_V2A.gh (21.8 KB)

BTW: For offsetting a Mesh for real-life purposes there’s 5 checks that you must perform (DO NOT offset a Mesh “at-once” > the art of pointless).




In plain English (for each candidate Mesh face AND before appending it to the offset Mesh):

  1. Check the DotProduct (in a 1:1 basis) between each donor Mesh face edge VS the candidate Mesh edge.
  2. Check the DotProduct between each donor Mesh face normal VS the candidate Mesh face normal. That said a Mesh face normal IS not the Mesh normal (aggregate of unitized directions from vertex to adjacent Topology vertices).
  3. Check each candidate Mesh face edge Length.
  4. Check CCX events between the candidate Mesh Face VS all the previous created ones (appended into the offset Mesh [on the making]). That said there’s mo Method available for CCX in Meshes with selves (like the one for Curves).
  5. Check all candidate Mesh Face edge pairs angles.

And this … is (supposedly) what you are after:

Mesh-alongCurve_V1.gh (126.5 KB)

Awarded in Code Project (1999) as the most stupid thingy (by far) known to man. Nobody could understand what actually does and why is written anyway (but Greeks know these things because they NEVER listen, he he) .The fact that at the time C# was not yet available to the masses adds to the pleasure.

Note that the offset issue reings supreme meaning that all kings horses and all kings men … blah, blah.

Note: planarity to 0.0001 and planarity to E-666 is the same thingy since nobody can tell the difference in real life.

Moral: life sucks.