Grasshopper

algorithmic modeling for Rhino

Hi David, I finally had a chance to report a bunch of bugs I've come across in the last couple of weeks. Some of these seem like deeper RhinoCommon problems, but I'll list them here anyway and they can be moved to the appropriate place. The files are sensitive, but I can email them directly to you if you like.

1/ Contouring (and also Brep/Plane Intersection) generates non-closed curves from a closed brep (the screenshot actually shows a surface instead of a brep, but the same thing happens):

2/ Contour generates non-planar curves (one is also open, see below). This is very disturbing because it cannot be used to create a 'boundary surface'.

3/ Offset doesn't return all results. This seems like more of a rhinocommon problem. It always returns a valid result, but often not the one I want. Better would be to return all results and let me choose what I want.

4/ Fillet issues. See image below, the fillet component works fine up to a certain radius and then the one on the right disappears completely (presumably the radius is too large so it gives up). However, if I use the FilletAtParameter component, the fillet works at each of these points but it won't do all of the fillets at once (regardless of how I arrange the data tree). My work around at this point is to get it to fillet each of the sharp bits separately and then RegionUnion all the curves together, which is incredibly slow.

5/ There is no ExtrudeTapered component, so I wrote a quick VB.Net component to expose this functionality. Firstly: I cannot for the life of me figure out what the "Base Point" input does. This seems to have no impact on the result and the documentation is missing. Secondly: giving it a non-unitized vector does very strange things to the result.

Thank you for your help!

Steven

Views: 466

Replies to This Discussion

#4 is technically not a bug.

The way [Fillet] and we have to be careful here as they are all called [Fillet] so we should specify the major input difference. So (R)Fillet's description is to "Fillet the Sharp Corners of a Curve", The most notable* thing about this phrase is multiple corners, single curve. This means that every corner fillet needs to work within the half length of the two segments that "connect" it. This means that if your Radius exceeds this half length then it fails. 

*by this I don't mean that it stands out but you should take note of the fact.

I guess what would be more useful than either of these would be a rolling ball (or in this case, circle) fillet similar to what we have for the surface fillets. Or else the ability to fillet at multiple parameters simultaneously.

Anyway, the first 3 issues are far more important so it would be good to get them added to the bug list.

Steven

RSS

About

Translate

Search

Videos

  • Add Videos
  • View All

© 2024   Created by Scott Davidson.   Powered by

Badges  |  Report an Issue  |  Terms of Service