WISH : Glossary of test commands

Having discovered by accident the existence of “TestPolycurvePoints”, which in my mind makes CPs behave at last how they always should have, I realize there may be more stuff out there that was just mentionned briefly deep in an obscure thread.

Of course, I don’t expect anything as earth shattering as the gradient hatches which were presented in great fanfare, but who knows ?

And what’s the point in test commands if there’s just a pair of folks who know they exist ? What is the statistical value of their feedback ?

Hi - That’s a frequently asked question, Olivier. Here’s one of the answers that was given in the recent past:

A complete list of all test commands wouldn’t do any good. The majority of them simply wouldn’t help anybody in any situation. Some might help in some situations. Some are frequently used but for some reason left as a test command.

Some of the test commands that some users say they use can be found in this thread:

For the specific case of TestPolycurvePoints, note that its behavior will be the default in the upcoming WIP and that using that test command will then turn off the desired behavior.

Fuzzy way to move forward indeed.
If the commands come and go, you might want to simply update the glossary (bright, huh ?).
I hope you guys at least keep track of this internally…

This is absolutely brilliant! So much easier and more intuitive to edit polycurve.
I agree with your statement on discovering such gems by accident.


_TestGlyphMesher. For whatever reason, for a split second, this command started auto-completing. I just want to know how much space the code for this one is taking up on my C Drive.

Also, y’all created _TestBaseball and have been holding out on us? Well, the cat’s out of the bag now.


Oh please ! I want this one !
Who needs a tropical aquarium now ?
Hey, what about flashing gradient hatches ?

OMG when I run _TestGlyphMesher, the letters are just black. What are you doing differently?

I got the funk in me my friend.

1 Like

Think of these “Test” command like a back stage pass at a concert.
Yes, you can poke around and maybe get some insight into how the magic happens, but don’t get in the way of the show.

If you want access to everything, you need to work here or be a member of the band. There are responsibilities that go along with that.

So please, let it go.
When a test command is a useful way to explore different options for a problem, we may use them to do that exploration. They will not be documented, not announced, and not supported as a released command. They will come and go. They are a tool for us to improve development and get better tools built for you.

You’ll need to trust us to do that.

1 Like

You’re basically telling me I’m not spongeworthy :frowning:

Not at all.
I’m just telling you to please refrain from distracting the band or stage hands too much. We’re trying to get the new show started.
Don’t be offended if they (we) ignore you occasionally.

Naaaah… I’m kind of used to that.
Take my block improvement wishes for example…they’ve been ignored for 15 years !

That’s not true either.
The ideas that seem worthwhile to many users generally, have been captured and saved for future development.
Some (I think) have been implemented.
Some are way too complicated and would constitute a major code overhaul project.
Those get saved until we decide a project of that scale is worth the development effort.


As a bit of fun I had a go at brute force ‘attacking’ the command line in V6 to find more test commands. Currently have found 387, of which 46 also exist as real commands. Definitely not exhaustive as my computer can only test 10k commands / second.
Out of respect to John’s philosphy above, I won’t post the list I’ve found, but here are a few interesting ones:

  • TestBoxOPoints (puts random points inside a box volume)
  • TestAddSmartPoints (i’m keen to know what this was about :smiley: )
  • TestCurveDir (adds an arrowhead to the end of a curve)
  • TestVid (adds labelled dots to mesh vertices)

There’s also a lot that I wouldn’t be game to run, in case it put rhino in a weird state I can’t get it out of. And many that are just random, eg “TestZLFT” or “TestJohnM”. All in all, I’d say the results aren’t as interesting as you’re probably imagining…


Keep in mind that you are taking extreme risk using any “test” command. They are created by developers to test something deep in the core. The developers don’t care if they crash Rhino or trash the model if it helps them work through an idea.


I totally understand that the developers do not want feedback on test commands. With that said, please be mindful of the fact that we all are advanced primates with an innate sense of curiosity. It’s part of our biology. It makes us collectively stronger.

I see threads like this pop up often in one form or another and even asked this myself a few years ago.

Here is my proposition:
Provide a glossary of test commands that are relatively safe to tinker around with and exclude all others from public builds.

1 Like

Business idea:

You launch RhinoCloud Ultimate. A subscription version of Rhino that runs on the cloud, in a cheap and slow computer. And it’s basically just Rhino PLUS an extra toolbar with all the tests commands. So that’s the cutting edge, the newest tech, the more forward thinking 3D for true innovators.

Of course, those test commands they don’t just run. They get sent to another cloud computer, one that is a little faster, nothing fancy, but decent. And that computer sends the results back to your slow and crappy cloud instance. And for this to happen the user needs to buy and use Rhino Credits. So you charge your customer extra money for running the commands in the software you are already charging them to use. Also because you do some switcharoo of computers and you are running test commands that produce results that look like shit, you need to spin it somehow, to make the shit results from those commands be a desirable thing. You say that those turds of bad geometry are very special, they are “AI-generated”

I don’t think anyone has ever those that yet, right?


None are safe because they are always changing.