Offset crv: anomaly?


#1

How come when I choose option bevel, soft or rounded, if I run an offset is generated correctly, however if I make a inset above options are not considered?
Is correct or is it a bug?


(Margaret Becker) #2

Please post an example of what you are seeing. Thank you.


#3

Here’s the test of offset. The red curve (pentagon) is the reference. The offset with soft edges works, but the inset fails to make the edges soft.
Is a bug?
Test Offset incorrect.3dm (41.8 KB)


(Pascal Golay) #4

This is the expected result. What radius would you expect on soft corners to the inside?

-Pascal


#5

Pascal, I do not know if it’s a bug, maybe not. However, if the offset is executed with soft option, because it could not be expected in the inset? The logic is not the same? What will change?


(Wim Dekeyser) #6

I suppose that the radius would have to be the same as the offset distance. The only problem is that this is likely to fail in many cases and what should happen then? Create the biggest possible radius or not have a radius at all?


(Pascal Golay) #7

You could but that is no longer anything like an offset, it seems to me- the arcs added to outside corners are at least offsets (smooth and chamfer are a little fuzzy in that regard, but it still makes sense to fill the gap left by offsetting curve segments on outside corners with something.) There is no such gap to fill on inside corners…

-Pascal


(Wim Dekeyser) #8

Ah, I see. No open rhino in front of me at the moment but then the default outside offset should be corner=none whereby the gap isn’t filled with anything and I don’t think I would ever expect that. … or?


(Pascal Golay) #9

Hi Wim- the default is Corners=Sharp, that is, the offset segments are extended until they intersect- which is perhaps not a true offset in those areas either, I guess, but I don’t know of a better way - it’s what is usually expected.

-Pascal


(Wim Dekeyser) #10

I totally agree with that. By the same token, a ‘true’ offset to the inside would create line segments that are not connected - which also nobody would expect. As trimming is performed in this case, I suppose round should trim with an arc. The question remains what should be done if there is no space for an arc with offset distance radius…


(Pascal Golay) #11

Well, we are picking nits here perhaps, but on an inside corner, every point on the offset, even after trimming, is offset distance away from the closest point on the original. If the corners were rounded, this would not be the case. But also, there is simply no logic that I can see that would lead to a rounded inside corner.

-Pascal


(Wim Dekeyser) #12

Yup, sorry about the picking of nits…
I don’t really see a real-life use for the options anyway (that is, I have never used them as far as I can remember) and we haven’t heard back from the OP…