I’m doing an Islamic geometry pattern and want to offset it for laser cutting - right now I’m using geo region and it works sometimes but other times I get this error - any ideas what the problem is? I’ve tried flattening everything, exploding, making sure no duplicate lines, etc.
Hi, its unclear what you want to do. Can you sketch or reconstruct what you want the output to look like in rhino?
You need to separate the boundary curve (four (or 3?) long curves) from the short interior curves.
Why are you asking if you already know?
Because it works on one of my patterns but not the other.. The working one does not need a separate boundary curve, so I can rule that out as the issue.
Done without using the heteroptera component. Rather than using a black box component, its better to understand what geometric regions actually are. They essentially a surface without a hole. So what you can do is -
1- Sort out the boundary curve from the list and make a surface out of it
2- Split this surface with the rest of the curves so you get each individual surface
3 - Connect all these split surfaces to a curve parameter to get your geometric regions
GeoRegion_surface split.gh (79.0 KB)
I would guess that boundary curves would be known but @annie.locke.scherer didn’t provide one. Odd that your GH file opens without error in R7, even when the Area component is missing the ‘C’ (Centroid) output, as could be expected in R8. But R8 works and R7 fails.
Yes good to move away from the black box and back to basics! Good suggestion 
I found an odd workaround - if I fillet the boundary curve a tiny amount then it works just fine - (not sure why) - think its just some odd heteroptera bug, but good to know in case others encounter it…
No, it isn’t (unless you Join Curves first). And you dismissed the idea that it would be useful.
separating the joined boundary curve does not solve the issue. It works fine on one geo but not the other.
Just like the exception that you believed was proof that boundary curves are irrelevant.
Is there a reason you haven’t posted boundary curves? Surely they exist somewhere, unshattered.
Both instances of working and not working include both boundary curves and geo. They are not irrelevant, they are just not the thing causing the issue.
Everything is internalized and screenshotted in the latest file and post, including the boundary curves…
I’m here asking a technical question in good faith. If you’re more interested in sarcasm than actually helping, maybe just don’t reply. The ROFLs and attitude add nothing, unless your goal is to be condescending, in which case, mission accomplished.
My first two replies in this thread were “in good faith”. Both were rejected. After that, what else can I do but
Glad to hear you finally see the wisdom of my advice.
Because your suggestions do not fix the problem… 
Yeah that geometric regions component does not give out any error message other than the generic object instance not set - which is not useful to understand whats happening under the hood. Your workaround probably won’t work for every scenario. If this is a workflow that you find yourself doing consistently then its worth making a cluster for yourself rather than relying on heteroptera (ignoring the fact that if you find yourself in a scenario where you dont want to edit the boundary curve then this is not even a valid workaround)
true true! I’ll likely use a mix of urs and my workarounds - was trying to avoid surface cutting with the gazillions of lines I have since it is so laggy, but overall that is much more robust than my fillet option. Much appreciated
