(Wish) Gumball Location, _GumballRelocate with different Options

[Tom_P]

Above the most simple case that annoys me:
select 3 of 6 CVs of the black Hexagon.
Gumball Position is calculated as center of the red Bounding Box.

my wish

_GumballRelocate
should offer different options to calculate the Gumball Location:

• BoundingBox-Center (current default)
• Average of CVs (best for shown case)
  … other approaches
• best fit Plane
• Align to ControlPolygon
• Normal to Surface
• smart (automatically shows the best of above options)
  …
  it would also be great to have the this exposed to Rhinocommon - if it is not already possible.

I think Gumball Location, Gumball Orientation, non rectangular U-V aligned Gumball for Surfaces, saved Gumball Location, …

this would be a great - and relative straight forward - field of enhancements for V9

Thanks

-tom

[Tom_P]

… well one of this topics, I do not understand the dynamics of this forum.
… maybe friday is a bad day to post stuff.

another example where the alignment is not intuitive …

for an planar edge loop - i expect an “orient to Object” Gumball being sensitive for this planarity.
“best fit Plane” in my list above.

@theoutside any thoughts / comments ?

[theoutside (Kyle Houchens)]

imo this still needs some tuning… the object is a bit of an odd shape so the result is not insane, but it would be nice if it could see the planarity.

[Tom_P]

is there a Gumball-Orientation bug (again) ?

or did we get an update, that made it worth ?
(Mac, Intel, Version 8 (8.12.24254.14002, 2024-09-10))

a simple EdgeLoop on a freely orientated subd-Cylinder with Gumball to Object defaults back to World/C-Plane ??

Am I the only one, unhappy with the gumball orientation ?

[Tom_P]

@mikko
may I ask you to have a look at this - you where working on some gumball alignment / orientation issues

Thanks

[BrianJ (Brian James)]

Hi @Tom_P ,

I’ve added this thread to https://mcneel.myjetbrains.com/youtrack/issue/RH-86215 as more feedback related to alignment wishes. Thanks.

[Tom_P]

another case where the default algorithm somehow gives unexpected result:

Gumball Position for Arcs with small sweep angle

i came across the following - drawing an arc with small sweep angle - the gumball is far of screen [Screenshot 2025-09-29 at 13.30.13] technically correct - the gumball is at the center of the arc / circle [Screenshot 2025-09-29 at 13.30.39] but somehow this does not feel intuitive or user-friendly. I think the only solution / approach would be, to allow the users to choose different algorithms how gumball position is calculated - see link below. Gumball_of_screen_for_arc.3dm (3.1 MB)

[Tom_P]

please - more love for the gumball location:
a planar subd patch with 4x2 faces.
it s planar.
!!!
???

gumball_position.3dm (3.1 MB)

[Lagom]

Uh

[Tom_P]

@Lagom

not sure which emoji captures the emotion mostly…
(and nice overall graphics with your avatar.)

[Lagom]

The Gumball is certainly off kilter. I mean, how hard can it be to orient on a plane?

What I really dislike is that, if you relocate the Gumball to a specific location, next time you’ll want to use it, it is no longer there. I mean, why is that?

[Tom_P]

if ever a developer manages to read this topic:
on top of the actual orientation, there are serveral topics around, that wish to remember a custom gumball orientation.

it would be nice to have a stored gumball location

• entire objects
• … super hard i know: for combined selections / sub-selections
• at least for named selections
• groups
• instances of block s → all instances should have the same new or reset gumball location.

[Joshua_Kennedy]

@Tom_P

entire objects

The gumball does save its location on the object always in object mode. When reselecting after a transformation or gumball relocate the gumball will align itself to the saved orientation.

@Tom_P

… super hard i know: for combined selections / sub-selections

This one is tough for a variety of reasons.

@Tom_P

at least for named selections

We have RH-51735 open about this. It’s difficult for the same reasons as the previous. I added a note to the issue.

@Tom_P

groups

This is implemented in the Rhino WIP.

@Tom_P

instances of block s → all instances should have the same new or reset gumball location.

The default object alignment for block instances is to their instance transform. Is that what you are expecting?

[Tom_P]

@Joshua_Kennedy

@Tom_P

instances of block s → all instances should have the same new or reset gumball location.

The default object alignment for block instances is to their instance transform. Is that what you are expecting?

@Joshua_Kennedy thanks for having a look

I thought of the following sample - but after thinking over it, i believe another feature will be much nicer: “BETTER suggestion - redefine Insert - CPlane” - see at the end:

Gumball to Object, Block

a super simple cylinder as a block - the default Object Gumball Orientation

Relocate Gumball

next Instance

first suggestion

after setting a new Gumball-Location for a single Block Instance - other instances should inherit it. If you reset the Gumball-Location this might also affect / reset all Block instances…

this is just an idea, but i like it, from a user point of view i think it would be great.

BETTER suggestion - redefine Insert Point / Insert-CPlane

ok but the usecase i think of could be solved with a even more powerful feature:
Redefine Block Insert-Cplane / update Insert Point, with the default option of keeping the placed geometry where it is - by updating the instance-Transformations as well.
as far as i remember this feature was already wished somewhere.

kind regards -tom