Hi everyone. i cant offset a surface network in a same directions. some of them go outside and some of them go inside. i want all of them go outside. the base surfaces are hidden. please give me ur advices.
Patch will not take you anywhere, because you cannot join the deformed surfaces back to the main body.
The edges of the surface generated from Patch do not align with the base curves generate them in N-Gon cases like this.
I prepared the definition, hoping somebody can solve this long-term problem.
Bowl-02 Ngon Patch.gh (101.1 KB)
WARNING: SDiff takes two minutes 
Geometry is internalized, no need for Rhino file.
Clipper PolyOffset is used because result is better than Scale.
thank u Quan. it seems u did it well. it takes time to understand what u did . im trying to figure it out. u really changed my vision to the problem. wish u best(s)
thank u Mr.Oster. i wish i can understand what u did. i haven’t the plug-in that contains polyoffset. i have made the empty wall like this
Thanks @Joseph_Oster for sharing your definition! I liked the look of it and tried a quick and dirty (emphasis on the latter) print without supports:
Sorry for going off topic, but is there a way to get rid of the linear seam running up the side of he bowl? It looks like it comes from the original surface made by @Odyssey (OP) – made by surface of revolution?
Yeah, that print is a little “dirty”. I’ve tried something that avoids the seam (it’s still there) by scaling the Voronoi³ cells. Not as clean and consistent as what I got from PolyOffset but I’ll have to ponder this… 
Two minute delay from SDiff takes patience.
It’s not right yet, I think this approach is a dead end. I can see the problems and have ideas (similar to version ‘Aug24b’) but have other things to do so will post what I’ve got and ponder it further…
Got it 
This wasn’t that hard, spent too much time on the dead end path. Simpler, I think? Back to Clipper PolyOffset and slow SDiff (1.9 minutes).
Too bad the bowl surfaces are internalized instead of parameter driven; changing thickness is not an option, though that can easily change. 
P.S. The “trick” here is not splitting the surface with the Voronoi³ cells. The surface seam is still there but it doesn’t affect hole shapes.
