Rhino 7 ExtendSrf causes other problems

Hi All

I used the BlendSrf to generate a surface, and I opened the record history.

When I scale the input surface or turn on control points to change the input surface, the output surface is fine.
extendsrf

But when I extend the input surface edges, the output surface results in a problem
extendsrf02

I also tested Rhino 6.35 and had the same problem.


But when I was using an earlier version of Rhino 6, this all worked fine.(Probably in 2018)

↓ This GIF is a screenshot from an earlier video tutorial and you can see that everything is working fine

extendsrf03

So it could be a bug in ExtendSrf?

1 Like

I see the same behavior.
7.19.22137.7001, 2022-05-17

History with Record=Yes Update=Yes BrokenHistoryWarning=Yes
BlendSrf between two surfaces
Modify one the surface by moving a control point, dragging the surface, etc. Do not use ExtendSrf
The blend surface follows the modified input surface
ExtendSrf
The blend surface does not follow the extended input surface
No broken history warning
Modify the extended input surface by moving a control point, dragging the surface, etc.
The blend surface follows the modified surface.

Is this behavior a bug or intentional?

1 Like

Thanks, I’ll check it.

There is an update, but it stays put - that is, the surface stays attached at the original parameter on the edge. The difference is not between V6 and V7, it is in the state of the target edge - trimmed edges behave as in the first case above, unmtrimmed not - it looks like we should be using normalized parameters - then I think both would behave the same way.

https://mcneel.myjetbrains.com/youtrack/issue/RH-53677

-Pascal

1 Like

@pascal If an untrimmed edge used as the target for BlendSrf is extended using ExtendSrf then the blend follows that edge. I understand the rationale that the parameterization of the edge does not change in that situation, but the difference in behavior may appear arbitrary to users who are not familiar with the math behind NURBS.

More inconsistencies:

Loft using the edge of surface as an input curve. ExtendSrf the surface with the input edge. The lofted surface changes as the surface is extended. Why is there an inconsistency between how a blend surface and a loft surface react to changes in the input?

Loft using only curves. Extend one of the input curves. The lofted surface does not change. Why is there an inconsistency between what happens when the edge used as input is extended using ExtendSrf and when an input curve is extended using ExtendCrv…

Yes. That is why I suggested we should use the normalized parameter as suggested by Lowell in the bug report.

-Pascal

Thanks for your help and feedback, cheers!