Rhino 8 Development

Sketchup is also for theatrical and fashion show set designers and retail shop display designers… and there’s something horribly wrong with that.

Those people go to college to learn to draw and for some reason they all end up using the CAD equivalent of Duplo. Somebody make it stop!


I don’t think so. It’s a great and cheap 3D modelling tool for beginners and amateurs. It was my start into 3D back in high school. If I would have had to start with Rhino, I probably would have given up right away! The ease of use of SketchUp is widely unrivalled up to now. Maybe TinkerCAD can match it in terms of accessibility. Rhino is way too complex, clumsy, and bloated, if you ask me.


I have professionals routinely giving me sketchup files that include complex curves rendered in low poly mesh that forces me to redraw a lot of their work in order to CNC cut it.

They take a vector graphic brand asset and import it to sketchup and want me to carve a 3d relief or letters.

If people are going to be designing things they want built, they should use CAD that allows them to model exactly what they want built.

1 Like

Disagree. Its very simple to draw a few lines and extrude them. Learning curve is flat.
Still, there’s a lot of stuff that needs a breeze of 21st century.


As a grad student I had the opportunity to introduce Rhino, Maya, and zBrush to some of the incoming students. I found that the students who had used SketchUp in the past were most frustrated with the new software because the “push-pull” methodology that they relied on didn’t translate to other modeling types.

1 Like

They should be teaching students use the right tool for the job.

Some people are graduating from art, technical, and even engineering schools that don’t understand the difference between a mesh and a CAD model, and when to use one over the other.

Well that’s not really the programs fault, is it. I’d simply charge them extra for their ignorance, after all that’s how the job economy generally works. You know how to do something that somebody else doesn’t and you thus should get paid for whatever service you provide. There are also CNC apps that can generate engraving or cutting toolpath from flat vector files.

Yeah, you say that but I’m not really convinced, being a seasoned CAD and CG all-rounder myself. I’ve helped so many people here over the last few years that I can confidently say that many don’t get that as naturally as pushing and pulling in SketchUp. It’s a stretch to start from a couple of points, to get to a line or curve that then eventually leads to a three-dimensional object.
The learning curve is rather square root-ish, I’d say.

I get that. The push-pull-methodology is really intuitive. Now, with the gumball in Rhino, things are looking up. Extruding and everything related is still a little flimsy, compared to other applications, but I’m optimistic.


That might have been important 5 years ago, but sure is less and less. Also meshes have had a resurrection with the rise of home 3D-printing and CNC-milling.
You probably also mean a mesh versus N.u.r.b.s. geometry, because “CAD” is “Computer Aided Design” not a geometry type.


All this A is great, B is better, person X should not use if Y and Z is nice, but I suggest y’all take that to Meta so in this thread we can go back to moaning exclusively about Rhino and what should be done better now that we devs definitely have been aware of certain issues for decades.


There probably an approved generic term for something that is not a mesh.
There are times where even a Nurbs result is not the best result.
When using some 2D CNC, Nurbs or spline curves should be avoided in favor or arcs

I would like to see improved support for pointclouds from 3D Laser Scanners.
Rhino already does a great job of importing and navigating E.57 pointclouds, but its tool set is a bit limited.
I work as a 3D modeller creating models from pointclouds, mainly for the boat building industry. I have been involved in laser scanning since 2000 and have seen HUGE uptake of the technology recently in architecture, forensics, manufacturing, engineering, surveying and shipbuilding.
Rhino is actually a very good platform for modelling from Pointclouds, but only if the pointclouds are prepared prior to import using something like “Cloudcompare” to apply colour, decimation etc.
It would be great to be able to view pointclouds as colour or intensity at the click of an icon and be able to load a decimated version of the cloud and switch back to a full version at will.

There are Pointcloud Plugins for Rhino, namely Veesus Arena4D for Rhino,(costing more than Rhino itself), but native tools would be great.


In no order…

  • A linear texture mapping like GTKRadiant has had for 20 years, so that material’s need not be remapped after resizing an object. More texture/material just slides out, and stays in scale. You want a longer wall? Just resize it, more wallpaper is added. Rhino has some ability for this on a global scale, but not per object.
  • Mapping tools that allow x,y copies of a material to be placed on an object. Also a fit option that will fit a material to any rectangle surface, such as fitting a wall-hanging to a frame.
  • Natural NURB h/v texture mapping, such as found in GTKRadiant, which allows the simple application of materials on objects such as cable, wires, even while coiled.
  • Some work on the block manager, so that larger projects don’t have to be handed off to McNeal’s competitors.
  • Perhaps some enhancement to the the walk-through, which allows some kind of rudmentary conclusion detection, which would allow a client to walk through a building before it is constructed.

That would be the surface parameter mapping that already exists and that you don’t have to add.

Here a simple piece of wiring, where I added only a material with a simple texture. I only adjusted the repeat a bit, but no fiddling with texture mapping otherwise. If the surfaces of your object are well behaved you get well behaved surface parameter mapping.

I don’t understand what this means. Are you talking about the repeat of a texture?


Hi Nathan,

Yes, that looks promising, but it needs to be a per-object mapping technique instead of a material-property. The trick is: we need to apply the same material to many objects, without becoming lost in material-menu instances nonsense.

If that indeed is the same kind of NURB H/V, technique, the user could make a surface with say 8x8 nodes, and really twist and distort it, and while the mapping would get distorted, it would not fail, like without going off into a situation where we are looking at the thin edge of the mapping, like accidentally rotating the x/y/z in the current mapping techniques.

The user should be able to select a surface, apply a material, select the number of copies on it–all in perhaps 5-10 seconds.

Additional use cases for NURBs H/V-Mapping include: applying a material on a curved sidewalk for landscaping people. The user applies the material and changes the mapping to fit 1 x 8 copies, and they are done. No twittling buttons to align anything. Additional cases include, drapes in buildings, carbon-fiber weave on car parts, applying wood grain to carved handrail. It also is great because Rhino can use arbitrarily-sized materials.

[There was a complimentary capping feature, which generally mapped as untrimmed surface, which is fairly like the Rhino Capped Cylinder Mapping. ]

As for the other mapping features…

Apply Material-Copies or Fit to Rectangular Surface:
Because many things in architecture are rectangle such as doors, windows, decorations such as pictures, it’s handy to be able to select a surface, apply a material to it, and tell it either: how many repeats, or the ability to stretch to fit. This ends up working a little like decals, it’s quicker, but it only works from the edges of the surface. In Rhino, there is nothing other than NURBs, so this is more of the same, but its uses here would be for rectangular objects.

You select a surface, apply out a material, then set the mapping to either Fit, or Fit Number of Copies for H and V. This works good for placing pictures on walls, or signs, labels on boxes for advertising. The mapping feature ends up working like Array/ArrayLinear but with materials instead of geometry. Perhaps there are GUI interpretations, I have not considered.

Linear Resizing Keeping Material Scaling:
As far as the linear thing, it would work with the 1D resize/rescale. For the example, the user set the scale just like they like, but if they want to resize the object, the 1D scale would add/spool-out/issue more material as the object is lengthened, or remove it as it’s shortened, something like the absolute box scaling, but per object.

It would seem that the Scale1D would could be modified to read the original mapping, and recalculate the mapping after a resize, and then reapply it. In most cases it would advantageous if the material mapping was recalculated in a way that is non-integer as far as the material copies. In other words, if you put woodgrain on something, and make it a little larger, you likely wouldn’t want a big texture-sized change, just issue out a little more material, as it does in the world-wide box mapping, currently used, but on a local, per-object scale.

Many things in architecture are made but cutting things to length, such as wood, siding, flooring, ceilings ties, paneling, etc.

Replicate Aligned Mapping
It would also be very helpful to be able to copy a material mapping and paste it to several objects, aligned. In other words: if we are putting some nice Victorian wallpaper in our house, it would be nice if we could match it up easily over the door opening. This

[ For extra credit, a randomize-h, or randomize-v buttons are handy, because architecture or not, if you can see material repeats on stairs it’s awful. It likely would help when applying grain other objects that are replicated as well. Alternatively, it would be astounding to be able to grab the material and drag it accordingly. ]

With Rhino 3D wealth of powerful curve and NURBs tools, the addition of lightning-fast material application features such as these will kick the proverbial door open to not only architecture, but cinema, and game-editing*–which are fairly similar in needs.
[AFAIK, Rhino was used in the new Blade Runner. : ) ]

While these may tools seem like simple in theory, (I know, I know, not in implementation), but to do without them is simply unbearable to do without, once you have gotten used to them–especially if you have used them professionally.* ; )

(Thank you for the reply.)

This is a rather esoteric request but it might be easyish for you to implement.

I have a spherical HDRI file that shows earth from the stratosphere; it actually looks like a view from orbit.

You can see the sun in the image but most of the light comes from the clouds below so my model is dark on the topside and light on the sides and bottom.

So even though the sun is visible, fairly high up, in the image: it isn’t casting enough light for my purposes.

So I tried to put a directional light into the scene and I tried to put it where the sun is.

That turned out to be more difficult than I expected.

I tried:

  • setting camera location to world origin
  • put in shaded mode to find the camera and select it
  • put in rendered mode to find the sun

But where has the light gone?

I’d like a single command to put the camera at the world origin and go into rendered mode then I can rotate the view and click on the center of the sun then when I complete the command a directional light is placed somewhere on a line between sun and the world origin. The light would be pointing at the world origin.

I don’t really care how it’s implemented, just as long as it’s easy to do.

To recap: I’d like some way to place a directional light in a scene that is lit with an image and have the light come from the direction of something that I can click on in the image.

I imagine that my view position would be at the world origin and I click on the screen where I see something in the image and this gives you a direction vector from origin to image that you can reverse to set the direction of any light that has a notion of direction.

I imagine that if I have selected a light b4 I run the command then this just sets the direction vector of the existing light.

Maybe I’d want to be able to create multiple lights in one go. I dunno.

Are you on recent Rhino 7, or still on Rhino 6?

In Rhino 7.10 and later an HDRi that has a very intense spot should give quite a bit of light with the changes I made along the 7 SR road.

Make sure your sun in the HDRi image has values that are very high, say values over 1000.0, (although I suppose anything over 1.0 should work).

You should not need to use a separate light source.

Maybe you can share the image file with me in PM so I could have a look?

I uploaded it using the link that you gave me ages ago. The one for large files.

Is that different from using a PM? If so, how do I do a PM?

Is there a way to edit hdri files? I thought that the sun was just a patch of pixels that is no different than any other patch of pixels in the image.

I didn’t make the HDRI file.

I am using Rhino 7.

Using that link is fine.

Indeed the area giving off light is much larger than the sun. On top you still get hard shadows though.

You could try editing the HDR (I use Blender to do such touch ups) and bump the sun value from its current 175 to say 10000. That should help guide the CDF generation that happens for the environment texture used in lighting.

If you want to help the lighting you can add a directional light. Set the end vector to be at 0,0,0. Then rotate the light so that it goes from the sun in the image to the 0,0,0 coordinate. That gives you lighting control with matching shadows. Attached a quick setup of the directional light matching quite closely the sun position, even though I did it only manually.

There is no need to move the view around like you do, just keep the directional light vector end at 0,0,0 and rotate the light around that to match.

You could add a few points on a surface where there are clear shadow features (sharp corners etc) that help you line up the directional light.

jmcauley_DH_Stratosphere_test.3dm (493.5 KB)

13 posts were split to a new topic: Mac Rhino 8 Crashing

As a Sketchup used since v4, I agree with much of what you say. And I agree the complexity of Rhino (and many other packages) is why SU became so popular in certain fields. Many of us were art students not math/computer science, (very few were both) and just the language of modeling software was foreign. Just the meaning of “curve” for example. That said, I don’t think Rhino is bloated or clumsy. I just offers more features for more creation. When I first got Rhino, I kept jumping back to SU for the speed. I slowly learned, and for my type of work I’m now pretty much up to my SU speed. I miss the functions of SU guidelines, and the ease of applying colors and textures, but now I never go back unless I’m asked to.