Strange happenings since installing latest release


#1

I am working V5 with a file that is about a half g in size. Since installing the latest release I have observed in this file that the Boolean operations for combine and split are reversed. If I want to do a Boolean split I have to select the combine function and visa versa. I had opened another simple file and the operations acted normal. Now the color of some of my solid entities has changed to that of the inside surface that I had previously specified. They even changed on locked layers that were hidden.
Do I have a corrupt file and if so how do I salvage it. Or is it something else?


(Willem Derks) #2

Hi Robert,

What you describe: inverse Boolean results and inside surface coloring, is due to the geometry in question not being closed polysurfaces but polysurfaces with an open edge somewhere. The normals are opposite to what you expect and that is why our expected outside is colored as the inside.

To fix this you can:
Make the geometry sold by finding the naked edges and closing them
( command _ShowEdges )

Or simply run the command _Flip on the objects to Flip their normal direction.

HTH
-Willem


#3

Is that with every entity, even simple boxes spheres? Usually when Boolean operations act in the opposite way you expect, at least one of the objects is open and has its normals facing inwards…

–Mitch


#4

Willem, Mitch: This is good info to know, but the OP seems to be saying that what he has is a change in behavior for an existing file between the previous release of Rhino and the current one.

If msbob2010 can be sure that nothing else changed, then it seems to me this is something the McNeel developers would be interested in.


#5

This is what I’m trying to determine. IIRC this is the second report in a couple of weeks with the same “inverted boolean” symptoms, which is why I asked. However neither has gotten back with a definite example where a valid closed solid booleans backwards. This stuff is normally pretty reliable if the objects involved are correct… It would be good to know why the file is also exhibiting the classic symptome of objects being inside-out - i.e. the inside surface (backface) color being on the outside…

–Mitch


(Willem Derks) #6

Hi Al,

Re-reading the topic I see how this is possibly more than the obvious. @msbob2010 can you specify if this happened to unchanged files the worked OK is an earlier version?
Note that the color change is dependant on the displaymode and not bound to the file.
So changing the “inside color” of surfaces will show up in earlier files.

-Willem


#7

The best would be to have at least one object where this is not working to analyze…


(Brian James) #8

Hi Robert,

Booleans shouldn’t have changed to my knowledge as of SR6. If you have a simple case that shows the ‘flipped’ behavior between Union and Difference, please share it with tech@mcneel.com or upload to www.Rhino3D.com/upload. The 500mb file may succeed in uploading but if you can isolate only a couple objects that show the issue and use Export to save those as a separate file that would be best. Thanks for the report, let me know if I can help more please. I agree with Willem here that the geometry used is likely not closed or the normals are inconsistent. It could also be the case that a small fix in another area of Rhino has resulted in this specific Boolean working ‘correctly’ now. I’ll need the objects in question to know for sure.