All points are on the same plane, no duplicates, so what’s causing it to leave out one point like this:
If I move the “left out point” to the left then it is included…
Test it out: points for meshpatch.3dm (20.6 KB)
And a wish: It would be good if it could force a boundry as illustrated in red (as a setting):
Not everyone is on V6 or V5-SR13 - unable to open file in V5SR12.
You could use Delaunay Mesh with Grasshopper to do this.
Left: result of mesh-patch; right result of Grasshopper Delaunay.
Hi Jorgen - I see that - for now, run testConvexHull on the points, and then MeshPatch with that included in the input.
@tim, it seems to me you once explained me why points can be left out in MeshPatch - if I am not making that up, can you remind me? it probably made some sense at the time but I cannot remember what you told me…
-Pascal
Ooh, is that going to become a real command…? Are you just grabbing the GH algorithm? I don’t see that method yet in RhinoCommon…
–Mitch
Hmmm - not that I know of… it really does not come up very often I guess, Having it in RhinoCommon would be a good thing. All that said, there are comments from the developer, in the code, about how slow this function is, and indeed to bogs down pretty quickly if you give it even a couple of thousand points, so I guess it is really not very ready for prime time.
-Pascal
OK, there already is a convex hull component in GH, and as GH is based in RhinoCommon, I thought… But I guess there are other things also missing in RhinoCommon that are in GH like Delaunay, etc…
BTW feeding 10K points to the GH Convex Hull gets a result instantly. Maybe the developer in question should talk to David… 
–Mitch
Indeed… I think testConvexHull is ancient history and really is a test thing for some other (not top level Rhino) purpose. I’ll dig into this a little more and see what we can come up with.
-Pascal
I had no idea V5sr12 coudln’t open a V5sr13 file, that seems like a bug to me as I didn’t know that sr13 added any new geometry to the file that sr12 couldn’t read.
But thanks for showing the GH component.
It can (well … didn’t try but it would surprise me 
)
The Audit3dmFile command run on this file tell me:
3dm Version: 6
I guess that explains it 
Hi Jorgen - Tim is revising how the outer boundary is computed - it will not need testConvexHull - you can fix it yourself, in a way, by running the dash version of MeshPatch (I know, I know) and setting MainSegmentCount to 1 instead of 10.
thanks,
-Pascal
And that explains the round points to… 
Ah, yeah, I played with the -meshpatch settings but didn’t get a good result, I thought I adjusted MainSegments too, probably didn’t try 1. Thanks for looking into it!
