Uploaded June 2019 | Updated November 2026, 1 hour ago
Lets refactor the destruction system we built in a recent live stream. This lets us destroy "Asteroids" (read: cubes) in Asteroids VR by subdividing them until a minimum size is reached. This initial design relies on a single `MonoBehavior` to perform this and that has a few issues. The most annoying of these is a weird behavior with Unity's prefab system that causes the object to reference itself instead of the prefab we've created. Another thing I'd like to get out of this system that is currently difficult is a way to track and control the creation of new objects more centrally. This will allow us to more easily use an object pool later as well as more easily control how objects are generated in other ways.
To adopt this we'll be refactoring the code that instantiates new "Chunks" into a separate component which we'll discover by using [`GameObject.FindObjectOfType`](https://docs.unity3d.com/ScriptReference/Object.FindObjectsOfType.html). This retrieves a `MonoBehavior` from any `GameObject` in our scene which matches the type we're searching for. This search method is not the final solution we'll end up using later in development, but for testing right now it's more than enough to get us what we want. We can use the reference we retrieve to replace the actual instantiation, effectively moving the parts of the code we want into a separate place.
To make things a little easier and to make sure our new component is more useful we'll still do some of the lifting in the Asteroid's code. This includes calculating the rotation and world position of each new "Chunk" of the Asteroid. We do this so that the controller we built can be used to generate brand new Asteroids and not just create debris from existing ones.
There is still a bit more to do to get a fully functioning destruction system (we don't track health yet for example), but this gets us to a place where object destruction and creation works reliably well.
The base subdivision destruction system was built in a Live Stream which you can find here: youtu.be/POVJ-QIyygE
Join the World of Zero Discord Server: discord.gg/hU5Kq2u
Lets refactor the destruction system we built in a recent live stream. This lets us destroy "Asteroids" (read: cubes) in Asteroids VR by subdividing them until a minimum size is reached. This initial design relies on a single `MonoBehavior` to perform this and that has a few issues. The most annoying of these is a weird behavior with Unity's prefab system that causes the object to reference itself instead of the prefab we've created. Another thing I'd like to get out of this system that is currently difficult is a way to track and control the creation of new objects more centrally. This will allow us to more easily use an object pool later as well as more easily control how objects are generated in other ways.
To adopt this we'll be refactoring the code that instantiates new "Chunks" into a separate component which we'll discover by using [`GameObject.FindObjectOfType`](https://docs.unity3d.com/ScriptReference/Object.FindObjectsOfType.html). This retrieves a `MonoBehavior` from any `GameObject` in our scene which matches the type we're searching for. This search method is not the final solution we'll end up using later in development, but for testing right now it's more than enough to get us what we want. We can use the reference we retrieve to replace the actual instantiation, effectively moving the parts of the code we want into a separate place.
To make things a little easier and to make sure our new component is more useful we'll still do some of the lifting in the Asteroid's code. This includes calculating the rotation and world position of each new "Chunk" of the Asteroid. We do this so that the controller we built can be used to generate brand new Asteroids and not just create debris from existing ones.
There is still a bit more to do to get a fully functioning destruction system (we don't track health yet for example), but this gets us to a place where object destruction and creation works reliably well.
The base subdivision destruction system was built in a Live Stream which you can find here: youtu.be/POVJ-QIyygE
Join the World of Zero Discord Server: discord.gg/hU5Kq2u




![I Learn How To Make Bash Flip Tables When I Break Things
Lets take a look at how to customize a `bash` shell. Well be making our shell print a table flip whenever the last command failed. In order to customize your prompt you can overwrite the `PS1` environment variable.
To see what your current prompt is set to you can run `echo $PS1`
The table flip script looks like this:
```bash
table_flip() {
if [[ $? 0 ]]; then
echo
else
echo (╯°□°)╯︵ ┻━┻
fi
}
```
Here is a tool that can help you customize your prompt more easily: http://ezprompt.net/
The ArchLinux wiki goes into much more detail about prompt customization: https://wiki.archlinux.org/index.php/Bash/Prompt_customization#Bash_escape_sequences
Join the World of Zero Discord Server: https://discord.gg/hU5Kq2u I Learn How To Make Bash Flip Tables When I Break Things](https://i.ytimg.com/vi/XUOti_9WgaM/mqdefault.jpg)

 class that provides some helper functions to make this easier.
The first thing to do is define a set of planes we can use to describe the view of our camera. This is called a frustum and is the white box-like gizmo you see in Unitys scene editor when you select a `Camera`. The frustum defines your cameras view and we can take advantage of this to test if objects are inside or outside of the bounds we calculate.
In order to calculate the frustum planes well want to run `GeometryUtility.CalculateFrustumPlanes(...)` and pass in the camera we want to calculate the Frustum of. In our quick example well use `Camera.main` to grab the main camera in our scene.
Once weve calculated the camera frustum planes we can provide those to the `GeometryUtility.TestPlanesAABB(...)` method to test if the bounding box we provide is inside or outside of the planes we calculated earlier. This method returns true if the bounding box provided overlaps with or is inside of the `Plane[]` boundaries. In this case were testing if a bounding box overlaps or is inside of the cameras viewport.
To get the bounds of an object you can use the `Collider.bounds` field. This calculates a world space axis-aligned bounding box for the collider and returns it to you so you can reference it in your object. The bounds are always a box aligned to the XYZ planes but are supported on all types of `Collider` in Unity.
Once youve passed your bounds into the test function you can use that to perform your behavior for objects in or out of view. This can be used to freeze objects when they come into view (like the Doctor Who Angels) or to move/delete an object when it gets out of sight (like some gibs you only want to be cleaned up when the player looks away). Theres a lot of different uses for functions like this!
The code Im using in this example has an `Update` that looks like this:
```csharp
var bounds = collider.bounds;
cameraFrustum = GeometryUtility.CalculateFrustumPlanes(camera);
if (GeometryUtility.TestPlanesAABB(cameraFrustum, bounds))
{
renderer.sharedMaterial.color = Color.green;
}
else
{
renderer.sharedMaterial.color = Color.red;
}
```
You can learn more about dot product here: https://youtu.be/8cZo-c-f1yc
***
Join the World of Zero Discord Server: https://discord.gg/hU5Kq2u Detect Objects in Your Camera View - Unity Geometry Utilities](https://i.ytimg.com/vi/XrqesjfcitU/mqdefault.jpg)
![Exploring How to Write Custom Pixel Shaders for the Windows Terminal
The Windows Terminal is a new terminal for Windows that allows you to run PowerShell, Command Prompt, the Azure Command Shell or any other terminal environment. One of the cool experimental features it includes is the ability to run a custom DirectX pixel shader when rendering the terminal. This can allow you to customize your prompt and create cool visual effects in your Terminal. Lets explore how you can configure and create your own shader for the Windows Terminal!
To use a shader you will need to update your Windows Terminals `settings.json` file to point to a DirectX pixel shader file. The setting is called experimental.pixelShaderPath and is located in your terminals profiles. In this example well be applying the shader to every profile in our terminal by default by updating our settings file like this:
```diff
profiles:
{
defaults:
{
...
+ experimental.pixelShaderPath: The absolute path to our shader.hlsl,
}
}
```
A basic shader that just rerenders the sampled terminal input might look like this:
```hlsl
Texture2D shaderTexture;
SamplerState samplerState;
float4 main(float4 pos : SV_POSITION, float2 tex : TEXCOORD) : SV_TARGET {
return shaderTexture.Sample(samplerState, tex);
}
```
It includes a `shaderTexture` which is a Texture that contains the rendered content of the Terminal prior to your shader being applied. This means it will include the colored text but none of your effects. A `samplerState` is also provided which provides state information about sampling the texture.
The shader requires a `main` method that returns a four dimensional vector containing the red, green, blue and alpha channels to render. A transparent result will render the background of your terminal profile.
To modify the behavior of the shader, change the results returned from the `main` method to match your desired outcome and update the Windows Terminal. You should be able to see the results of your changes. For example our shader that only renders the red channel of our terminal has a `main` method that looks like this:
```hlsl
float4 main(float4 pos : SV_POSITION, float2 tex : TEXCOORD) : SV_TARGET {
float4 color = shaderTexture.Sample(samplerState, tex);
return float4(color.r, 0, 0, color.a) ;
}
```
Some additional documentation on Windows Terminal pixel shaders is available on the Windows Terminal GitHub page here: https://github.com/microsoft/terminal/blob/main/samples/PixelShaders/README.md
That README includes additional information including how to access the `PixelShaderSettings` which provide information such as a `Time` variable to create animated effects!
***
If youd like to add a shortcut that can toggle shader effects on or off then you can add an action to your Windows Terminals `settings.json` file. In the `actions` block add an entry for `toggleShaderEffects` like this:
```json
actions:
[
...
{
command: toggleShaderEffects,
keys: shift+f10
}
]
```
In my example pressing the Shift and F10 key would toggle shader effects.
***
If you dont have the Windows Terminal you can download it here: https://www.microsoft.com/en-us/p/windows-terminal/9n0dx20hk701
You can find some additional examples of Windows Terminal shaders here: https://github.com/Hammster/windows-terminal-shaders
My pride flag Windows Terminal shader is located here: https://worldofzero.com/posts/pride-windows-terminal/
Join the World of Zero Discord Server: https://discord.gg/hU5Kq2u Exploring How to Write Custom Pixel Shaders for the Windows Terminal](https://i.ytimg.com/vi/YF50axsb-cA/mqdefault.jpg)


).GetAwaiter().GetResult();
```
Find a `Camera` object in the current scene:
```csharp
cameraObject = await CSharpScript.EvaluateAsync[GameObject](GameObject.FindObjectsOfType[Camera]().First().gameObject,
ScriptOptions.Default.WithImports( // Import namespaces to simplify scripts
System,
System.Collections.Generic,
System.Linq,
UnityEngine
).AddReferences( // Include additional assemblies
typeof(System.Linq.Enumerable).Assembly,
typeof(UnityEngine.Transform).Assembly
));
```
The Roslyn Scripting API is documented on GitHub: https://github.com/dotnet/roslyn/wiki/Scripting-API-Samples
You can find the Nuget package here: https://www.nuget.org/packages/Microsoft.CodeAnalysis.CSharp.Scripting/
Join the World of Zero Discord Server: https://discord.gg/hU5Kq2u Runtime C# Scripting - Embedding the .NET Roslyn Compiler in Unity](https://i.ytimg.com/vi/YtzYkoWhrTk/mqdefault.jpg)