archived-code-general

[pure cliff]

doesnt seem to be a big deal atm but I gotta go back and make the two the same type at some point

then wont need to do that

[ashen yoke]

you can use a single buffer

[pure cliff]

?

[ashen yoke]

a list that acts as a buffer

you clear it, use it, repeat

[pure cliff]

oh yeah, that prolly would be much more performant

Ill keep that in mind if this ends up needing the performance boost

[eager ingot]

Hmm...I want to use addforce to move the player, but I ran into a bit of a problem. I want to make it so that there is a maximum amount of force a player can add using WASD input, but if you are airborne, forces like jumping, explosions, etc can make the player move even faster. I want their airborne speed to be essentially limiteless. I tried setting a maximum velocity, but I quickly realized that this would cause the airborne velocity to follow the same rules as the running/strafing velocity. How can I set a max input velocity or acceleration to my rigidbody?

[mental rover]

as an idea, get a normalised vector for the player input and dot it with the current velocity - if the result is greater than the maximum magnitude of velocity you want the player to have direct control of, then discard the input

[eager ingot]

That's not a bad idea actually

I'll give it a shot

[pale bay]

hello, can someone help me understand what im missing in regards to saving:

I have a mobile game that saves a json file to the application.persistentDataPath and this works fine in the editor, however building and running the game onto my android, I find that I cannot find the file

The first image shows the file in my local machine when saving on my local pc

Setting up a log in the saving process, building and running the game on my phone, and viewing said log in LogCat shows the directory of the that file

2023/07/18 20:20:57.977 3978 4040 Info Unity Saving settings to: /storage/emulated/0/Android/data/com.DefaultCompany.marblegame/files/settings.json

I navigate to that file and find nothing (second image)

I know for a fact that it exists because everything I open the game, the save file is loaded and my settings are applied correctly

What might I be missing here? I want to view the file so I can debug an issue

I have windows file explorer set to view hidden files btw

[leaden ice]

How are you sure that the file was actually written?

also that directory doesn't look quite the same

[pale bay]

for example, I have a slider UI element that has its value set when that file is loading as well as saved when the app closes it. it works fine. Ill set more logs up.

from what I understand /storage/emulated/0/ is ignored

I could be wrong about the second part but I do not see the storage directory when navigating my phone in file explorer

[leaden ice]

logging when you write the file would be a good start - as well as logging the file path as that happens

also showing your code for writing the file would be good

[pale bay]

Saving

    public void SaveSettings()
    {
        SettingsSaveData.musicVolume = MusicVolume;
        SettingsSaveData.sfxVolume = SFXVolume;
        SettingsSaveData.vibrationState = VibrationState;

        string jsonData = JsonConvert.SerializeObject(SettingsSaveData);
        File.WriteAllText(_settingsFP, jsonData);

        // This is the log in the log entry that you suggested to implement
        Debug.Log($"Saving settings to: {_settingsFP}");
    }

Loading

    public void LoadSettings()
    {
        if (File.Exists(_settingsFP))
        {
            // If the save file exists then read it
            string jsonData = File.ReadAllText(_settingsFP);
            SettingsSaveData = JsonConvert.DeserializeObject<SettingsSave>(jsonData);

            Debug.Log($"Loading settings from: {_gameStateFP}");

            if (SettingsSaveData == null)
            {
                Debug.Log($"File not found, creating new object");
                SettingsSaveData = new SettingsSave();
            }

            return;
        }

        // Return a new score data object if no save file is found
        SettingsSaveData = new SettingsSave();
    }

I got this log when rebuilidng with the save log
2023/07/18 20:48:53.237 16704 16727 Info Unity Saving settings to: /storage/emulated/0/Android/data/com.DefaultCompany.marblegame/files/settings.json

Save file content:

{"musicVolume":0.0,"sfxVolume":0.389136523,"vibrationState":false}

[leaden ice]

should work

[pale bay]

my thoughts exactly

[mellow copper]

HI, teams, when tried to add an example script, I met an error [No MonoBehaviour scripts in the file], but the file name is already same as class name. What should I do to avoid this?

Here's my code and file name, editor version is 2023.1.3f1c1.
ExampleClass.cs

// Initializes the target variable.
// target is private and thus not editable in the Inspector

// The ExampleClass starts with Awake.  The GameObject class has activeSelf
// set to false.  When activeSelf is set to true the Start() and Update()
// functions will be called causing the ExampleClass to run.
// Note that ExampleClass (Script) in the Inspector is turned off.  It
// needs to be ticked to make script call Start.

using UnityEngine;
using System.Collections;

public class ExampleClass : MonoBehaviour
{
    private float update;

    void Awake()
    {
        Debug.Log("Awake");
        update = 0.0f;
    }

    IEnumerator Start()
    {
        Debug.Log("Start1");
        yield return new WaitForSeconds(2.5f);
        Debug.Log("Start2");
    }

    void Update()
    {
        update += Time.deltaTime;
        if (update > 1.0f)
        {
            update = 0.0f;
            Debug.Log("Update");
        }
    }
}

[pale bay]

ill keep digging, im sure its something small

[quartz folio]

https://unity.huh.how/programming/scripts/loading-issues

[pale bay]

reconnecting my phone and trying again worked

[pulsar isle]

How can I check in a coroutine while something is true keep looping and if it is false then execute the code underneath?

[somber nacelle]

a while loop

that's kind of their whole deal

[pulsar isle]

so something like this?

or do I have to put something inside the while loop?

[somber nacelle]

yeah, but you'll want to yield inside the loop, a yield return null if you just want to wait until the next frame to check the condition again

alternatively, if you don't plan to have any other code in the loop you can yield a WaitUntil

[pulsar isle]

I got it working! But just to understand it better how can I transform that arrow function into a normal function?

[somber nacelle]

WaitUntil takes a parameterless Func that returns bool so you can just use a method that returns a bool. or keep the lamda expression, either one works

[west sparrow]

I didnt know about WaitUntil 🧠 🔫 that's handy

[pulsar isle]

Okay, thanks

I just tested my code and the coroutine works but there is a minor problem

The coroutine runs before the camBrain.isBlending bool is first set to true

So it instantly runs the code after the WaitUntil

[somber nacelle]

when are you starting the coroutine?

[snow geode]

How can I check if a button is being held down and if a button is not being held down anymore through a script where it references the buttons? I can’t just use the onclick() trigger on the button it self because I need this to work with multiple instantiated copies of the player.

[pulsar isle]

On update inside a function when the player presses a key

[somber nacelle]

is this before or after you start whatever transition/blend for the cinemachine brain?

[pulsar isle]

After

I incremented the priority of the camera and then started the coroutine

And I made the code only run once with a bool for it to not repeat endlessly

So right after the player presses the key there is a bool that stops the code from running again so it all only runs once

[somber nacelle]

ah, in that case the actual update of the vcams is probably happening later in the frame. you could try a 1 frame delay at the beginning of the coroutine with yield return null; to see if that gets it working

[pulsar isle]

Oh, so I could use yield return null and then yield return new WaitUntil() in the same coroutine?

[somber nacelle]

yes, you can yield as many times as you'd like in your coroutine

[pulsar isle]

Wow, thanks xd

[somber nacelle]

actually you may want a WaitForEndOfFrame instead of null, but just try it with yield return null first

[pulsar isle]

Then I guess it could also work by using first yield return new waitforseconds(delay), right?

[somber nacelle]

sure, if you want to wait for a number of seconds instead of for waiting until specific parts of the update cycle

[snow geode]

Boxfriend, do you mind trying to help with my issue after you’re done helping geeksand?

[somber nacelle]

if i knew the answer i would have already provided it

[snow geode]

Ok, thanks

[pulsar isle]

Thanks 🙂

[molten sun]

Quick question, how do you change the shadow quality of individual lights at runtime? I don't want to change the shadow quality globally, only for individual lights.

I learned about Light.shadowResolution, which seem to be exactly what I'm looking for, but the documentation doesn't have any examples of how to actually use this.

Does anyone have an idea on how to do this? Thanks (I'm using the built in render pipeline)

[fathom plank]

Thoughts on setting up a health system like this? Basically I was thinking this way avoids the need of getting the component on the target game object directly in a collision event/etc to apply damage, and instead you can just fire an event with the collider.gameObject, or w/e and if it has a health script, great, if not oh well.

i guess a downside is that the FireApplyDamage has to know which game object has the script attached to it to apply properly so it would be awkward to add as a nested child

[cosmic rain]

Imagine you have thousands of objects with health. Each of them would get the method triggered and make a reference comparison.
Not to mention that it's plainly confusing implementation. It implies that Health is some kind of global object or concept.

Getting a damageable component on the go would be way better. Getting components on a gameObject is not that expensive, considering most of them have less than 10 components. If you have the component reference you would be able to do other stuff and make specific checks on it as well. For example, if you only want to apply damage to enemies, you could get the character component and check it's "team" or "faction" before applying damage to it.

[fathom plank]

if a script is disabled, afaik you can still get reference to it, as well as call functions on it, maybe im mistaken? This was also an attempt at basically making the script a no-op in the case its already been disabled

i mean i suppose i could early out, in those cases

[cosmic rain]

I don't see how that argument is relevant to anything I've said.

You can make an enabled check all the same.

[fathom plank]

fair enough

[earnest gazelle]

Please read my question carefully.

A developer presses play button in the editor when the scene is playing. I want to show a dialogue box with yes/no button.If No button is pressed, it should do nothing. isPlaying = true freezes the game.

        EditorApplication.playModeStateChanged += OnplayModeStateChanged;

 private static void OnplayModeStateChanged(PlayModeStateChange state)
    {
 if (_state == PlayModeStateChange.ExitingPlayMode && Application.isPlaying)
        {
            if(EditorUtility.DisplayDialog(DialogTitle, DialogMessage, "Yes", "No")){}
            else{}
        }
    }
}

isPlaying = true is the problem.

[mossy snow]

so you want to abort the mode change? if you're checking for exiting play mode, you're probably too late. Might be smarter to try and target the play button itself

[earnest gazelle]

I did not get

Suppose a level designer in play mode is working on a scene and suddenly press play button before saving it. I want to prevent it

[somber nacelle]

why not just save the scene when the play button is pressed instead of showing a dialog?

[thin aurora]

#archived-networking

[gray mural]

what's the best way to get this difference assuming that white square is gameObject and I have RaycastHit2D info about red object?

note: player should be translated to that red square's position

[verbal cliff]

what do you mean by the difference? the gap between?

[gray mural]

exactly

[lusty gull]

i suggest using distance - collider.bounds

[gray mural]

yes, should it look like this?

(2f - red.size.y - white.size.y) * .5f * dir.y

where 1f is 1 unit in world space

and 1f - red.size is diff between 1 unit and scale (don't mention prefab's initial scale)

[lusty gull]

something like this, it should be like take the AB vector, then you can use this vector to get distance and direction from one another. Then you can take the half collider bounds and multiply for the direction, then subtract from the original vector and get the distance of that one. repeat for as many objects as you have! i don't know if it's clear

[gray mural]

not that clear actually.

[earnest gazelle]

Yes, it is a solution but maybe I do not want to save it always in exiting play mode

[lusty gull]

ok ill make a drawing for you in paint

[gray mural]

actually this is working assuming that object is in center unit. so if it's gap between unit's max y and unit's min y is the same

then is works @lusty gull

if gameObject is in the center of unit

[gray mural]

I see, so it's transform.position.y += A.halfBound + B.halfBounds + x

(I know that transform.position cannot be set like this)

[winged mortar]

[lusty gull]

halfbounds in my calculation is a float. If you need the direction information, you can use:
A + AB.normalized * (AB.magnitude - HalfboundsA - HalfboundsB)

[gray mural]

will this work if gameObject isn't in the center of unit?

if it's position isn't inted (int)?

[lusty gull]

no. if you need dynamics not based on transform, i suggest looking into the Collider API

[gray mural]

I see, thank you, I will try that!

[lusty gull]

the useful methods for you in this instance will be ClosestPoint and ClosestPointOnBounds

[cold egret]

- Given a collision2d, can i get a color of the object i just collided with at the contact points?

[copper rose]

about unity version control i have again same problem, one new project created by going to project settings -> unity project id, but unity version control using another projecct, how can i fix that? because i have changesets from one project pending in another project, a really big problem i made my self

[gray mural]

yes, SpriteRenderer has color field

collision2D.transform.GetComponent<Renderer>.color;

[cold egret]

- It has a texture. It's a tilemap renderer, and tiles do have various colors, they're not solid. So getting the renderer's color might give an approximation, but never the actual color. I mean it's ok if there's no other way around this, but i'd like to get the exact color of the collision spot

[gray mural]

get color of its meterial probably

[rain crane]

Hello is there anyway to make a triggered 2D collider reload on everytime i press fire1 button? i tried enabling and disabling the collider but it doesnt count any collisions with an object with collider

[gray mural]

yeah

retrigger it

[cosmic rain]

It should be possible to get the tile and it's texture and sample it somehow, but it's not simple.

[rain crane]

hmmmm sounds like a good idea gotta try it thanks

[gray mural]

wait, do you mean OnTriggerEnter2D method?

oh, collider

[rain crane]

yeah

[gray mural]

method?

[rain crane]

I use OnTriggerEnter2D method

[gray mural]

yes, you have to retrigger it manually

[rain crane]

is there any way to do that in code? since i cant find any method or either a value to modify

[gray mural]

yes, there is, hold on

[rain crane]

tyty

[gray mural]

I have done it once before

I have to remember it.

[rain crane]

aight

[gray mural]

Physics2D.IgnoreCollision(triggerCollider, otherCollider, true);

Physics.Simulate(Time.fixedDeltaTime);

yield return null;

[rain crane]

ah cool thanks

[gray mural]

IgnoreCollision
Simulate

this should work

we ignore collision for 1 udpate, then simulate physics and retrigger OnTriggerEnter2D this way

[rain crane]

the otherCollider is now the problem since i do not know who can it be.

[gray mural]

you can also use Physics.IgnoreLayerCollision

[rain crane]

tyty

[short badge]

Hey, anyone seeing anything wrong here? My player's not moving at all..
https://pastebin.com/B1Ahddrb

[rain crane]

damn that makes the character to fell off

[gray mural]

yes, you have to work with layers

you have to ignore specific layer e.g. Enemy

not all layers

also you can find collider by tag

or name

there are so many possibilities to do it

[gray mural]

is onMovementInput method even called?

also currentMovement is too small

but it may be enough

[cosmic rain]

Why do you even need to do something like that? It sounds like you're approaching the problem from a wrong angle.

[rain crane]

im trying to create my own melee combat system for the game

but it seems i found an easier solution

[cosmic rain]

That doesn't answer my question.

[short badge]

its called automatically by unity because of the started, canceled and performed.

[rain crane]

But that's basically what you've asked

[cosmic rain]

No, I didn't ask what you're implementing, but why you do it the way you do.

[rain crane]

that was the only solution i found atleast by my own

[cosmic rain]

Maybe explain what exactly you're trying to do? Detect when a sword is overlapping with the enemy or something?

[short badge]

fixed it, needed to add a speed variable

[gray mural]

yes, it was sure too small

[short badge]

well i believe i fixed it* looks fine for now lol

[rain crane]

aight.
I made a box collider 2D and made it with the scale of the swords' range.
i was planning if i could re enable the collider on attack and detect the new collisions.

[gray mural]

also you probably don't wanna enable and disable new Input System like that

private void OnEnable()
{
    playerInput.Movement.Move.started += onMovementInput;
    playerInput.Movement.Move.canceled += onMovementInput;
    playerInput.Movement.Move.performed += onMovementInput;
}

private void OnDisable()
{
    playerInput.Movement.Move.started -= onMovementInput;
    playerInput.Movement.Move.canceled -= onMovementInput;
    playerInput.Movement.Move.performed -= onMovementInput;
}

[short badge]

Thanks!

[gray mural]

don't see any reason to do it in 2 actions

just in start probably

it's just 1 script, so it doesn't make sense

also we write scripts from capital letter

[cosmic rain]

Why would you need to reenable it?

[rain crane]

to detect new collisions ig

[cosmic rain]

Why not keep it enabled all the time?

[rain crane]

it'd be looping on OnCollisionEnter2D or might only fire that event once

[gray mural]

do you know OnCollisionStay2D?

[cosmic rain]

The first is solvable with layers. Only make it collide with the enemy layer.
What's the problem with it firing only once per collision?

[rain crane]

oh yes

[finite panther]

error for me text error fix

me

[gray mural]

do it yourself, as I have mentioned before

[finite panther]

too lazy

and i dont know how to use unity

so fix it

[rain crane]

i want it to fire that event everytime the attack button is being pressed and the collision happens

[cosmic rain]

You could cache the colliding objects and apply your logic to these objects only when you press the button. What's the problem?

[rain crane]

it seems you're right gotta try it thanks

[lusty gull]

guys anybody knows if it's possible to use [BurstCompile] on extention methods?

[late lion]

I believe so.

[gray mural]

How to make gameObject not be affected by Rigidbody2D at all, without making is static or kinematic (in order to trigger OnCollisionEnter2D)?

[winged mortar]

Does the thing you are colliding with have a rigidbody attached?

Iirc you need atlast one of the two to have a rigidbody

If yes, then you can just remove the rigidbody from your gameobject

[gray mural]

no

[winged mortar]

Why do you need the collision?

[gray mural]

to see when my player has collided with another collision and stop the game

[winged mortar]

Then the other collision could have a rigid body right?

[gray mural]

no

it won't

[winged mortar]

[gray mural]

it's moved by tranform.Translate

😭😭 it shouldn't have rigidbody, because it's behaviour will be broken then 😭😭

[winged mortar]

RigidBody.MovePosition

[gray mural]

maybe I should move my player with it?

[winged mortar]

Or

https://docs.unity3d.com/ScriptReference/Rigidbody-position.html

translate is just transform.position + difference iirc

[gray mural]

it's just updated after next physics simulation step

also when my player has rigidbody it's behaviour seems to be quite weird.

it doesn't seem nice

I just cannot move as usually

[winged mortar]

Looks like there's a collider on your playing field

And it's pushing it out of the field as soon as it starts

[gray mural]

but it has different z position

[winged mortar]

Why are doing this weird 2d 3d mix man

[gray mural]

what do you mean?

[winged mortar]

Not an expert on this but if you click 2d as a template when creating your project you shouldn't be getting stuff like this right

[gray mural]

no

[winged mortar]

Also, try using a tilemap!

[gray mural]

it should

[gray mural]

why should I?

[winged mortar]

There's probably someone else on YouTube that can explain that significantly better than I do, but lets just say that it isn't required but it can simplify everything

[gray mural]

I see

I think I should probably jsut use Physics.OverlapBoxAll

[sonic edge]

Hello, I have an issue in my 2D game. I implemented a ledge grab which works with 2 collisions : one to detect the ledge to grab, and another one which verify that there is no "Ledge" above (so the player will grab the ledge at it's top). When I jump it's working pretty well, however if my character is too fast (-20 y velocity is the max), the box starts bugging and will sometimes detect the ledge and sometimes not. Does someone know how to fix this ? https://streamable.com/wfhc7k

[winged mortar]

Although not ideal, you can raycast from your previous position to find these kinds of colliders

Or up the fixedupdate rate

But I am pretty sure this is just a problem with any fast moving object

[sonic edge]

ok I'll try thanks !

[winged mortar]

You can also make your colliders larger

[winged mortar]

Btw double check this somewhere, maybe in #⚛️┃physics

[sonic edge]

ok I asked and I will see if there is a easier way to fix it ...

[gray mural]

hello, any way to implement smth like OnCollisionEnter?

I don't mean functionality, I mean that my method should be "called" like it's

private void Foo(Vector3 smth)
{
    // and here is already all information about `smth`
}

it also doesn't seem to be overriden

ok, it's hard one, don't care about it

I have another question. Why do BoxCollider2Ds not collide?

[winged mortar]

Depends on your setup

[gray mural]

there are 2 gameObjects with BoxCollider2D and they just not collide

their z position is the same

[winged mortar]

Rigidbodys?

[gray mural]

on?

[winged mortar]

That are simulated?

[gray mural]

no

[gray mural]

no rigidbodies

[winged mortar]

One of them needs to be a rigidbody for unity to do anything with them

I think you can also manually call the physics functions from the context of the collider

[gray mural]

I see, it isn't just OnCollisionEnter

so I move my obstacle in Update

private void Update()
{
    transform.Translate(GetDirection(direction) * speed * Time.deltaTime);
}

and I check for collision in LateUpdate in another script

private void LateUpdate()
{
    if (CheckCollision(k_obstacleTag))
        gameManager.EndGame();
}

But late update is a bit too late

it does execute just when colliders are on each other

any suggestions?

[hasty haven]

Wow, looks like I should not be using lists for this

[steady moat]

You can allocate memory in advance by specifying the capacity

[hasty haven]

Yeah, I was able to do that for the vertices but I am now in a puzzle for the triangle data, since there are different submeshes

and I dont want to pre allocate 256 large capacity arrays for worst case

I do have a solution idea though

[steady moat]

Whatever you do, you gonna want to allocate contiguous memory.

List is just a wrapper around an array.

[hasty haven]

My plan is to make one main array with some logic around it, i already know the max case in terms of triangle indices (just integers)

And the type is denoted by a byte

So I think I may count all the types before generating and partition the main array sorta

each partition being the size of the worst case for triangles

worst case meaning all sides of the voxel are rendered

[steady moat]

Are you generating the voxel base on visibility ?

[hasty haven]

Yeah so the original data for voxels is simple byte array,
a 3x3x3 chunk would be 27 bytes.
When I am generating the mesh I check visibility of each voxel that isnt empty

The time is super inflated with the deep compiler open, usually its <10ms for worst case

which is a grid that looks like this

[steady moat]

Did you try without checking visibility ?

[rancid jolt]

I made this code for bullets to move towards the direction of player in 2D environment but they are moving upwards.

[steady moat]

Because I am not sure what is the performance upgrade you gonna gain.

[hasty haven]

Performance upgrade is massive

[steady moat]

You might be doing something wrong then

[hasty haven]

this is the worst case, every face is exposed

[steady moat]

You should not have an issue rending this.

[hasty haven]

I dont, my question is just to see how people would suggest splitting up the submesh types without allocating garbage

[prime sinew]

Maybe because you're rotating them too

[hasty haven]

It works very well but I am just further optimizing, this is 3x3x5 chunks all of the worst case.
Uses <5% cpu at any time and <25% gpu and i have a 1050 ti

moving around and destroying them even

[steady moat]

Then why are you still optimizing ?

[hasty haven]

Its gross seeing 0.6 MB being allocated every time a block is modified

[steady moat]

Also, you could use two step algorithm, one that define the amount of memory needed and the other that do the actual thing.

[hasty haven]

Ah maybe I should do that beforehand and include it with the block data

during generation

Count of each block type

[rancid jolt]

Now no rotation made it simpler now.

[naive narwhal]

Isometric Hexagon Tactics Battlemap: __Grid __vs GameObjects

My goal is to create an isometric hexagon battlemap with units moving across those hexagons on their turns. The hexagons will need to be** re-colored** / highlighted on various circumstances:

• when unit is selected to display where they can move + when I hover mouse on those tiles;
• when skill is selected to display where it can be used + mouse hover;
• etc.

The question is, which should I use to generate those hexagons: GameObjects or Grid?

Grid does sound like the perfect choice, but I struggle to find viable ways to constantly change how tiles are displayed based on different rules (distance from character GameObjects, etc).
Meanwhile, if the hexagons are GameObjects I'll be able to change their Sprites based on game state changes, mouse hover events, etc

[steady moat]

Start by trying out this: https://docs.unity3d.com/Manual/Tilemap-Hexagonal.html

[naive narwhal]

Thanks, on it!
One thing that concerns me with the tilemaps is the constant need to calculate Vector3 mouse position and compare it to the positions of tiles.
I will also need to store an additional array to store info on each tile properties (isPassable, Effects, etc).
Would you say it's not a big deal and it's not worth using GameObjects just for that?

[steady moat]

It is not a big deal.

[leaden ice]

isn't that something you'd only need to do once per frame?

And only write in one place?

[steady moat]

You can add data on tile map tile with https://docs.unity3d.com/Manual/Tilemap-ScriptableTiles-Example.html

[naive narwhal]

With mouse -- probably, yeah, you're right.
With tile properties -- each time something happens, I'd need to apply those changes both on Tilemap, as well as that other array that stores all the properties.
Say, I pressed my mouse button on a specific tile and now it needs to be on fire. I'll need to change both the Tilemap (for visuals), as well as the array (to apply effect properties).
And all of that by comparing world coordinates to tilemap and vice versa.

[leaden ice]

with proper separation of responsibilities that's really not a big deal though

[naive narwhal]

Ohh, so I can actually store all the properties there, right on tiles!
Okay, that probably solves the issue I just described...

[leaden ice]

all that code would not be in one place

[steady moat]

Beware, those are ScriptableTile meaning that they are share by all the tiles. It is not made for Runtime modification.

[naive narwhal]

You are probably right. Maybe I'm just too used to GameObjects that have everything I need in one place.

[naive narwhal]

Ahh, I see. Then a separate entity for properties that is.

[steady moat]

If you want to add data to change at runtime, you gonna want to either:
1- Use GameObject on top of the Grid
2- Use Dictonnary or other data structure to add data
3- Create your own tilebase (And instantiate it): https://docs.unity3d.com/ScriptReference/Tilemaps.TileBase.html

[stuck plaza]

I need a help with my project

Anyone can help me?

[vagrant blade]

You need to ask a proper question, first.

#854851968446365696 if you need to see how to do so

[naive narwhal]

(1.) If I were to Instantiate GameObjects on top of the Grid, I think I could as well give them Sprites and simply use that, couldn't I?

Other than that, perfectly clear!

@leaden ice @steady moat
Thank you both, I have more than enough to research / experiment with for now! 🙂 ❤️

[gray mural]

just understood I could ignore collision between them, then realised collider can just be trigger 🙄

stupid me 3 hours ago

[rigid island]

https://dontasktoask.com

[gray mural]

also you can read this short tip

it explains everything you need in 2 words

[topaz ocean]

Im trying to make a dynamic physic material that belongs to the instance its assigned to. Would I need to make a struct or set of variables to re construct the physic material each time its modified or is there a more straightforward way of doing this?
Ive tried using a custom editor to create and modify a physic material when the instance is validated, but these changes are not ever saved.

[leaden ice]

if you're creating a physic material in memory in the editor you will need to save it as an asset for it to be persisted

but can't you just modify the one you're using as needed?

Why create a new one

[unreal temple]

You can use https://docs.unity3d.com/ScriptReference/ScriptableObject.CreateInstance.html

So probably I'd do something like this:

class PhysicsMaterialController : MonoBehaviour {
  [SerializeField] Rigidbody _rigidbody;
  PhysicsMaterial _material;
  public void SetPhysicsSettings(float bounciness, float whatever, ...) {
    if (_material == null) {
      _material = (PhysicsMaterial) ScriptableObject.CreateInstance(typeof(PhysicsMaterial));
      _rigidbody.physicsMaterial = _material;
    }
    _material.boucninesss = bounciness;
    _material.whatever = whatever;
    // ...
  }

  void OnDestroy() {
    if (_material) Destroy(_material);
  }
}

that kind of thing

[topaz ocean]

Well Im creating a prefab of a collection of objects with joints and colliders to represent a wheel, and I figure it could be useful to have the ability to dynamically adjust the properties at runtime. To do that though I need a material instance for every wheel. I couldnt just modify a material asset because that would modify the base and all of the instances that use it.

[unreal temple]

Yeah, in this case you can do: var newMat = Instantiate(_originalMat);

Just make sure to destroy it when you're done with it

[leaden ice]

ok so - create the instance, assign it to the object, modify it as needed.

[topaz ocean]

How is that mechanically different from constructing one with new()?

[leaden ice]

It registers the SO properly with the C++ side of Unity

[topaz ocean]

I mainly ask in case the way theyre cleaned up is considerably different, because you mention I need to destroy it I would like to know where the heck it actually goes when its created. I thought the garbage collector would handle the one created with new().

[leaden ice]

UnityEngine.Object derrived objects all have a C++ half and a C# half

[leaden ice]

GC only handles the C# half

Destroy is needed for the C++ half

[topaz ocean]

Ohh because c++ doesnt have automatic garbage collection

[leaden ice]

yes, not only that but even if it did - you need to tell it to drop all the references to the object somehow. You can assume the engine tracks these objects somewhere

[topaz ocean]

If I managed to accidently not destroy the instance is that something I have to manually clean up or would that just require an engine relaunch

[leaden ice]

it would be a memory leak

which may or may not be a big deal

[dusk apex]

Pretty sure when the Editor or application terminates, the memory is freed.

[leaden ice]

^ this for sure

but if your game is long running and you do it a lot, it can build up

[ashen yoke]

unity has only MonoBehavior, ScriptableObject and various other injection points, what unity does is it creates a backend Mono object, linked with the c# object, which you extend

[topaz ocean]

Yeah I know this, I just want to make sure there im not accumulating junk while editing.

[ashen yoke]

only way for it to know you want a new native object is through the api it provided

[gray mural]

How do I hide this field in derived class?

[SerializeField] protected Direction direction;

[leaden ice]

Only way would be with a custom inspector

if you want it hidden in a derived class then derived class was probably the wrong design decision

[gray mural]

ok, I see, thank you

I think I just shouldn't care about it

[pure cliff]

private

[gray mural]

?

[pure cliff]

protected means classes that inherit from this can access it
private means they cant

[gray mural]

yeah, I know, why do you say that?

[pure cliff]

or do you mean you dont want it to show up in the inspector?

[gray mural]

I want to show it in inspector for all derived class except one

I just cannot override it

so yeah, I just don't care then

[dusk apex]

You wouldn't be able to inherit the data as private.

[pure cliff]

you want it to be a hard set immutable value for the exception class?

[gray mural]

how long have you been typing it? 🤔

[gray mural]

no, I want it to not be shown in inspector in derived class

[dusk apex]

I was typing something other - was that important?

[pure cliff]

then you probably need to branch your inheritence higher upstream

[trim schooner]

You would have to write custom a inspector for the class you don't want that field showing

[gray mural]

no, just my curiosity, I didn't mean to hurt you

[pure cliff]

no need to overcomplicate it with weird hacky stuff, usually this just means you need to further subdivice your inheritence out to make a branch

[gray mural]

yeah, I know, PraetorBlue has told me that

[dusk apex]

I agree though that you could have the field private and serialized in the child parent and expose a protected property for the inheriting class to have access to the member though.

[gray mural]

yeah, I just need this field in all other derived classes

[pure cliff]

youd have to serialize every single one of the other children, could be unweildy

[gray mural]

what does it mean?

[pure cliff]

No this just sounds like you need to further subdivice your tree

[gray mural]

exactly

[pure cliff]

You need another abstract class

Whats the names of these things so we have common lexicon

[gray mural]

how does it help me?

[gray mural]

names of..?

[pure cliff]

these classes we are talking about

so instead of "the things" I can say what they actually are haha

[dusk apex]

I meant parent @pure cliff

[gray mural]

ObstacleSpike : ObstacleController

ObstacleLog : ObstacleController

EnemyStraight : ObstacleController

ObstacleController : ObjectController

[pure cliff]

kk

and then whats the high level reason you dont want one obstacle to have the field exposed

like what makes it special

[gray mural]

because ObstacleSpike doesn't have any direction

it's just spike

it cannot move like enemy

[pure cliff]

gotcha

[gray mural]

but I have Direction.None anyway

(it's enum)

[pure cliff]

No probs, so what you want is to rename ObstacleController to be like, DirectionalObstacleController or whatever

[gray mural]

what. why.

[pure cliff]

and then you want like

ObjectController -> ObstacleController -> DirectionalObstacleController

[gray mural]

I see, you want me to do one more class

[dusk apex]

#↕️┃editor-extensions

[pure cliff]

And then your spike inherits from just ObstacleController and all the rest inherit from DirectionalObstacleController

[gray mural]

nah, that's so strange to do one more class just to hide one field

[pure cliff]

Now actually, you can pretty much have all your logic in ObstacleController still so

[pure cliff]

99.99% of the time once you start doing this, you start seeing other things that can go between the classes

this was just your first roadbump that made it obvious that you needed 2 layers of fineness for defining your obstacles

[gray mural]

yeah, that really does make sense, but I don't wanna do it just for one field, I just don't gonna use it

[gray mural]

yeah, I gonna do that if I find other things

[pure cliff]

nothing wrong with making a second class that just has a shorthand on it

that then saves you writing the same 1 line of code on a bunch of other classes

[gray mural]

I was also thing about doing

EnemyController : ObstacleController : ObjectContorller

[pure cliff]

yeah EnemyController could be a better name if thats the special logic for "has a direction"

[gray mural]

yeah, but Log is Obstacle and is has a direction

I just haven't invented full logic of my game

that's why it's strange

[pure cliff]

yeah so I think you will have multiple layers of defining your classes, and thats super normal

[gray mural]

yes, probably more of them in the future

[pure cliff]

I also typically subdivide my folders to match these various layers, and I just call it <Category><Type>Base in my folder tree

[dusk apex]

You need to decouple stuff from the base class if they aren't available in every inheriting class

[gray mural]

that's true, I'm still thinking about it

[gray mural]

what's that?

[pure cliff]

For example I very likely will soon have EnemyEntityBase abstract class and UIEntityBase in this folder structure

[gray mural]

I see, I also like ordering everything correctly

[pure cliff]

Though I have largely circumvented that by instead using composition pattern + strategy pattern

so instead of logic being on the entities themselves, all they set is "what is my strategy" and paramets for it, and then I grab the matching strategy class for em, I effectively inject in what their AI is for how they behave

[gray mural]

I see, you have said those classes are abstract?

[pure cliff]

any class I end with Base is an abstract one

[gray mural]

I have never used abstract classes before actually 🤔

[pure cliff]

all an abstract class is, is blocking the ability to new() one up

[gray mural]

yes, that might be useful sometimes

[pure cliff]

Like if you have abstract fruit class, and then apple : fruit, as a normal class it means you are allowed to new apple() but you arent allowed to just new fruit()

[dusk apex]

The band-aid solution would be ```cs
Class A
private something//Field can be serialized and exposed to the inspector but private
protected Something//Property exposed to inheritant

[gray mural]

can property be serialized?

oh, wait

no

we don't need it

you're right

[dusk apex]

Exactly

[pure cliff]

yeah so you wouldnt even have Direction in ObjectController, and only DirectionalObjectController will have it show up

[gray mural]

no, I just need ObjectContorller this way

[pure cliff]

The other option is do what I am doing:

Dont care what "Direction" stuff like a spike is, it can actually be allowed to face left/right/up, but it has zero impact

[gray mural]

yesh, it has Direction.None

[pure cliff]

cause in the end what direction my sprites face is purely handled by the animator

[gray mural]

I'm kinda perfectionist, you see

[pure cliff]

so I would just have the spike have the same sprite for every direction

[gray mural]

yes, it will have

[pure cliff]

yeah then that honestly is fine

[gray mural]

even though my "sprites" are solid colors now

even solid colors as sprites are better than if I draw them myself

[pure cliff]

it mostly depends on

1. how expensive your directional logic is to be running on things that dont care about it
2. How many of them there are

Like its possible breaking out the class to refactor the direction logic apart, may cause a significant or insignificant performance change

[gray mural]

yeah, that's true

making lots of classes and inheriting from them may be quite confusing, but it's much less confusing than if you have 100 half-same classes with repeating behaviour and 5 times more redundant code

[pure cliff]

exactly yup

and like I said, usually once you break it out a bit, you quickly start finding other things that belong in that layer instead of the layer above or below naturally

as you expand the project and start wanting other features you quickly go "Oh wow thank god I broke this out earlier, cuz now I have the perfect spot for <other thing> to go!"

[gray mural]

yeah, half of my code will probably be changed once I add new logic to my game

[gray mural]

that's why first things that I have done were parent classes for obstacles and objects

[pure cliff]

Im currently bumping into this awkward process of dealing with how I want my architecture to work with "bubbling up" events, because I have designed my whole game engine to be a big tree of services that are the "source of truth", and the monobehavior game objects sit at the very bottom of the tree, and their only job is to be controlled by the parent services

[dusk apex]

Should probably focus on writing code once and not having to touch it again unless you're doing so for learning purposes.

[pure cliff]

but... stuff like click events and whatnot occur on those monobehaviors all the way at the bottom, and the stuff that cares about if the event happens are close to the top, so I gotta like, bubble it up

[gray mural]

writing game engine sounds crazy

I think I will be doing it in my currect game

[pure cliff]

eh its just all the core logic for my game

but I like to keep all my code largely aggregated into 1 large stack of services with a single entry point at the very top, and everything cascades down from there

[ashen yoke]

use event bus

[pure cliff]

I can access the event, it just kinda feels messy

I just dislike events but I cant be having my models acting like they are services

[ashen yoke]

you are misunderstanding probably

event bus is not related to c# event

[pure cliff]

I know what an event bus is yes

Pipe<T> would be the modern C# class to use to make a simple one

[ashen yoke]

the separate services/systems stack is common, but what you described seems like overly imperative approach that defeats the purpose of component model

[pure cliff]

Id still have to hand off a reference to the writer to said classes, Id need to figure out how to do that

[ashen yoke]

can you link to Pipe<T>

[pure cliff]

derp not pipe

Channel, my bad

pipes are for interprocess stuff, wrong one haha

[ashen yoke]

channel seems to solve a different set of issues

[pure cliff]

Channel is what I would use for an event bus

[rocky jackal]

I first generate a list of directions the green lines are the directions within the fov and the red ones are outside. The yellow line is the transform.forward and im trying to find the closest unobstructed direction forward in the fov area but for some reason it returns the closest unobstructed world forward transform. can someone help ?

[pure cliff]

async support (but not mandatory), and works fine for a pub/sub model

[ashen yoke]

typically a simple event bus would act as a mediator

alright

[pure cliff]

var eventChannel = Channel.CreateUnbounded<MyEvent>()

and then you have eventChannel.Reader and eventChannel.Writer which you can async or synch read/write off of respectively

and hand off the reader to the thing you push to, and then you can hand the writer off to everyone who has events to push onto the bus

its pretty much just a Queue but it has support for stuff like Streams and Async reading/writing which can be nice

[ashen yoke]

can you show me a simple example of some ui mono raising some event and some system subscribing to that event?

i want to compare apis with mine

[gray mural]

@pure cliff @dusk apex thank you both for help guys btw

[pure cliff]

yeah so I have

public class WaitButtonEntity : EntityBase
{
    ...
    public event EventHandler OnClick;
    public void Clicked()
    {
        if (Pausing)
            return;

        OnClick?.Invoke(this, null);
    }
    ...
}

As an example of the monobehavior (EntityBase is a mono)

It has a button which hooks into the Clicked() method

[ashen yoke]

ok how is OnClick linked to the channel?

where does the initialization happen

[pure cliff]

Im not using a channel atm I just hook into it directly

I was just saying if I wanted to use a bus, Id use Channel<T> but I havent yet

My main issue with event bus is it restricts you to trying to figure out how to make all your events "sameish" so they can use the same bus, and that can cause antipatterns to arise

eventbus mostly makes sense when all your events really are gonna be sameish in some way so thats not a downside

[ashen yoke]

i dont understand

[pure cliff]

That or you have to make a bus for each different type of event which sort of defeats the purpose

[ashen yoke]

struct OnClickedSomeButton : IEvent{}
void Clicked()
{
  EventBus<OnClickedSomeButton>.Raise();
}

this is my current

yes there is no shared channel/pipe

havent required it, yet

[pure cliff]

so right now I just hook into the event like this

public class InputService : IInputService
{
    ...
    public InputService(IEntityService entityService)
    {
        EntityService.WaitBtn().OnClick += OnWaitBtn;
    }

    private void OnWaitBtn(object sender, EventArgs e) { ... }
    ...
}

[magic heron]

Hi guys, i'm trying to make an object stacked when clicked to only take half of it, but when both conditions are met, nothing happens, can someone please help me?

sorry to interrupt btw

[ashen yoke]

public class InputService : IInputService
{
    EventBinding<OnClickedSomeButton> _onClickBind;
    ...
    public InputService(IEntityService entityService)
    {
        _onClickBind = new EventBinding<OnClickedSomeButton>(OnWaitBtn);
    }

    private void OnWaitBtn(OnClickedSomeButton ev) { ... }
    ...
}

here

[pure cliff]

Im not sure what this is

[ashen yoke]

my bus api

showing example

[pure cliff]

ah kk

[brazen narwhal]

Hello, is it possible to set a GameObject to a specific size and not just scale?

[lean sail]

What are you expecting to do? the scale is the size

[brazen narwhal]

Wait, I may be misunderstanding things, Im just new to unity. Does any two objects given the scale of (1,1,1) will be the same size?

[gray thunder]

no

[pulsar dove]

The sizes of the Mesh may differ

[ashen yoke]

you have to understand that "size" is an abstraction

an object "size" can mean different things, for example in most cases you mean visual bounds of the mesh

[gray thunder]

You could have a huge building and a small player

[ashen yoke]

the object can have no mesh, but a collider, then the size will be the bounds of the collider

[gray thunder]

both being the scale 1,1,1

[ashen yoke]

the object can have no bounds from renderers or colliders, so it wont have any size

[brazen narwhal]

Is there a way to make them the same?

[ashen yoke]

what do you want to make same size, 2 meshes?

[brazen narwhal]

I need to make a GameObject change in size based on a given scale

[ashen yoke]

2 meshes?

[brazen narwhal]

Yes? Sorry, I am not used to the terms in unity

[lean sail]

What is the game object? Itll be easier to suggest something if we know more

[brazen narwhal]

The game object is a 2d sprite

[ashen yoke]

so its a sprite

sprites also have bounds and size

what is the problem you are trying to solve?

ie i place one sprite next to another but they are displaying differently

show a screenshot

[brazen narwhal]

I need to change the size of the object so that it will be in matching sizes with a map

[ashen yoke]

so you want to stretch the sprite over the whole map?

as background?

[brazen narwhal]

The size of the object may vary in sizes since it will be uploaded by the user

[ashen yoke]

you have several instruments, first ppu in the sprite import settings

then scale

[brazen narwhal]

Nope, I want it to scale correctly with the map

The user will also give the scale in m(unit)

[ashen yoke]

ok do you have consistent ppu across all sprites?

can the ppu be used as a basis for size?

[brazen narwhal]

I think no. The only thing the user know is the scale in meters. After that, the object should match the size with a map

[ashen yoke]

how does user set the meters?

[brazen narwhal]

The user just knows what the size of the sprite is based on meters. Because it will be made with that on mind

But their sizes will not be uniform

[ashen yoke]

size of the sprite is based on meters

what does it mean

[brazen narwhal]

The user will make a sketch outside the app and will upload it. It will then be overlayed on a real world map

[ashen yoke]

ill rephrase

[brazen narwhal]

Thats why it needs to be correctly scaled to meters

[ashen yoke]

what is the correlation between meters, and the sprite user makes

what exactly should user do to achieve a 2x2 meter sprite

[brazen narwhal]

For example, the user makes an image which should be 2m in length on real world. Of course it will not literally be a 2m image. What I need to do is to change its size to be 2m when overlayed on a real world map on the app. The user knows that it is 2m in size and will be able to input that on the app. But the problem is the images made may not be in uniform sizes. Like one may make a 500x500px for 2m irl size. Another one may make a 700x700px for 2m irl size

[ashen yoke]

so user sets the size in meters

in the app, there is a input field or dropdown

[brazen narwhal]

Yes

An input field

[ashen yoke]

ok, so you have the size in provided by user, you have constructed the sprite at runtime with Sprite.Create()

which accepts ppu value

use some universal ppu across all sprites, for example 100

then simply scale the image by the factor of the meters the user provided using 100 pixels as one meter initially

so a 512*512 sprite will be ~5x5 meters

target is 2x2

the math will be something like

float scaleX = (ppu * targetMeters.x) / sprite.width;

[brazen narwhal]

Thank you so much! I can finally do it

The way you explain things is so easy to understand. Also sorry for not being very clear at asking questions

[static matrix]

How can I set whether or not a particle system loops on runtime?

[leaden ice]

https://docs.unity3d.com/ScriptReference/ParticleSystem.MainModule-loop.html

[simple egret]

It's probably because your buttons are not anchored properly, so their detection area gets moved when the resolution changes.
Ask in #📲┃ui-ux as it's not a code issue.
Post the question as text for folks with limited data plans and slow networks.

[brave harness]

Thank you, but how do amour them properly

[simple egret]

What?

[hasty haven]

Love when one of my methods starts always returning true :(

[simple egret]

Anchor? The #📲┃ui-ux channel has some links to tutorials on how to do it

[brave harness]

K

[soft shard]

Ah, misunderstood what you were trying to do - since this is a Editor script, you want to use EditorApplication.isPlaying instead, with your example:

if (state == PlayModeStateChange.ExitingPlayMode)
        {
            if (EditorUtility.DisplayDialog(DialogTitle, DialogMessage, "Yes", "No")) {  } //do nothing, exit will happen automatically if "yes"
            else { EditorApplication.isPlaying = true; } //prevents exit if "no"
        }

[brisk grove]

Hi

guys i dont know how to use github

can any 1 help me

[leaden ice]

Here ya go https://docs.github.com/en/get-started/quickstart/hello-world

[brisk grove]

1 thig

how to open a unity project from it

...

[dusk apex]

Say what? You'd use the Unity Hub and Editor to open Unity projects.

GitHub is a code hosting platform for version control and collaboration. It lets you and others work together on projects from anywhere.

Relative to working together and from anywhere: you'd have the same versions synced using version control.

[leaden ice]

there's nothing special about a Unity project managed by git and any other unity project. You just open it like normal from the Unity Hub

[modern creek]

Is there a preset or an easier way to set a RectTransforms anchors based on the size of the parent? I'm finding I do the following - and it seems like there should be a better way.

In the video, I show how I get a dialog box to resize properly based on the parent. First I do it without any anchoring, then I do it with the default anchoring, then I show how I get it to work correctly: I have to memorize the size of the parent (in this case, 1920x1080), then set the child to stretch, then manually set the left/right/top/bottom anchors to AnchorMinX = L/W, AnchorMaxX = 1-(R/W), AnchorMinY = B/H AnchorMaxY = T/H

I do this often so that all my elements size properly but it seems like this should be a default anchoring setup somehow? I'm thinking maybe I should/could write an inspector plugin for this? Or am I just doing something incredibly naive/wrong?

[leaden ice]

I don't really know what the math you're doing there is but the resulting values you're getting are just pretty intuitive percentages. E.g. 24% of the screen to 76% of the screen or whatever

could you not just think of it that way?

[soft shard]

There are also additional options if you hold down SHIFT or ALT before clicking, when you open that menu @ 7 seconds of your video, I believe SHIFT lets you do essentially what your doing manually

[modern creek]

Yeah, I'm just saying I have to do the above like... over and over and over when I nest components. If I do a nav bar that's 1920x100, then every button inside it that I want to be scaled properly has to do the above

I feel like there should be a preset? for what I'm doing that I'm not aware of

Or is this anchoring just part of y'alls layout process as well? Ie, figure out some reference resolution and manually lay every item out pixel by pixel..?

[leaden ice]

I defeinitely don't think about pixels

I think in terms of percentages usually

occasionally "pixels" when doing offsets at the top

e.g. like 10 units from the border or whatever

so there's no math - in your case I would have just written .24 and .76 directly

[modern creek]

yeah the only problem with that (using "pixels" instead of percent) is then they break when different scales

ok, i could probably shift my brain into working like that

[thick frost]

I always thought that "code rot" was the tendency of unmaintained code to fall out of date with libraries. But I've seen it used elsewhere in different context. How so you folks use it?

[modern creek]

i could even make my reference resolution something like 10000x8000 (or whatever 16:9 is)

[modern creek]

code that is old ("old" per changing scope/requirements)

[leaden ice]

Something like:

• Comments no longer accurately describing what the code does
• The original nicely designed architecture and abstractions getting polluted over time by small "fixes" and hacks

[modern creek]

Or not even negatively connotated scope changes. IE, you've got a player object with a deepcopy:

public struct Player
{
  public string Name;
  public Player DeepCopy()
  {
    Player p = new()
    {
      Name = this.Name,
    };
    return p;
  }
}

Then your designer says "we need age, too!" so you add public int Age and now your DeepCopy() is "rotten". Simplistic example, perhaps

[brave harness]

Oh i forgot to metion the buttons still show in the build they just don't work

[random oak]

Help my Coroutine is stopping inside an if statement and I have no idea why. It is only in one particular condition in the coroutine, there are no errors and I am not deactivating the script in any way.
Anyone know why this might be and how to check what's wrong?

[modern creek]

paste code

[leaden ice]

you would have to show the code. Can't help based on nothing

[simple egret]

I know, have you fixed the anchoring?
Maybe you have another object that goes on top of the buttons when the size changes, because it's not anchored properly. And that's blocking the mouse clicks from reaching your buttons

[leaden ice]

The only things that will stop a coroutine are:

• Deactivating the GameObject it's running on
• Calling StopCoroutine or StopAllCoroutines
• yield break
• Destroying the GameObject it's running on
• an exception being thrown
  @random oak

[random oak]

https://hatebin.com/qjrkfgdldb

it only stops when reaching line 39

[leaden ice]

Debug.Log($"{player.PlayerName}...Got {card[0]}!");
That's a log statement, it can't stop there

it will only stop on yield statements

unless you get an exception

[random oak]

I mean it prints but nothing else happens after

no errors

[leaden ice]

TurnManager.Instance.NextTurn(); will happen next

add more logs to check

[random oak]

it doesn't reach line 80

[leaden ice]

line 80 is in the else

else won't run

since you're in the if

87 will run after 39

[brave harness]

I watch a quick tutorial changed, but i might have done something wrong. Then it tryed it on a couple resolutions and it all still worked

[random oak]

is it not outside?

sorry line 89

it always reaches line 89 normally but doesn't when inside line 39

[leaden ice]

show your console window when this happens

(the whole window)

[random oak]

ok I will try, it only hits that if statement if the bot has no matching card and has to draw

[leaden ice]

Ok first off you have Collapse turned on in your console

you should turn it off

with Collapse on the logs will not always show in order

second - please show the full console window in these screenshots as it's important to see everything especially the top right and top margin

[random oak]

ahh didn't know sorry

[leaden ice]

See how ENDED ROUTINE has a (2) at the end?

That means that it happened twice

so it is indeed running when you expect

but due to the COllapse setting it is not showing again after the other log, it is merely being added there as a (2) @random oak

when you turn off Collapse it will show as expected

[random oak]

ok let me try again with no collapse, I have to restart editor cause unity flooded me with errors when I did

2022 shenanigans

yeah it's weird now is working after draw??

yeah it keeps working so far, idk so maybe some weird editor bug causing coroutine to act strange?

I'm so confused

[leaden ice]

It was working always

the logs were just obfuscated by the COllapse setting

at least that's how it seems from my outsider perspective

[random oak]

the logs were there somewhat put all over the place to try to find it , but in the game it wasn't switching turn to other player at times

anyway seems to work so far ! so who knows if it was just internal editor bug/crash in unity causing weird behavior .
Never happened on 2021
ty for you help @leaden ice

[brave harness]

so i think i fixed the anhcor but it still dosnt work and the music also isnt there

[daring cove]

Hey,
How to calculate the angle required to turn a vehicle?

Vector3 direction = destination - vehicle.transform.position;

float angle = Vector3.SignedAngle(vehicle.transform.forward, direction, Vector3.up);

int angleNormal = angle > 0 ? -1 : 1;

TurnHull(angleNormal);

This always causes angle to get stuck between 166 and -166 causing angleNormal to be 1 and -1 meaning the vehicle doesn't know where to turn and just turns in place forever.

[static matrix]

c# deciding to use python syntax for switch-case for no reason:

[leaden ice]

isn't this backwards?
int angleNormal = angle > 0 ? -1 : 1;

[lost dove]

Greetings, hoping for someone to have a look at this.
When my player faces direction 0, the determined input axis needed to rotate towards the mouse position is fine. However if I start rotating my player ship, then the axis keeps jumping between 1, 0 and -1.. and I'm not sure how to make it consistently just stay 1, or -1 depending on the easiest path towards target angle.

            var direction = Entity.Transform.Rotation.ToEuler().Y;


            //var direction = Properties[(uint)EntityPropertyTypes.TurretRotation].Int;
            // Get the mouse position in screen coordinates
            Vector3 mousePosition = Input.mousePosition;

            // Calculate the target direction angle based on the mouse position
            float targetDirection = Mathf.Atan2(mousePosition.y - Screen.height * 0.5f, mousePosition.x - Screen.width * 0.5f) * Mathf.Rad2Deg;

            // Adjust the target direction angle to match the desired rotation
            targetDirection -= 90;

            // Normalize the target direction angle to be within the range [0, 360)
            targetDirection = (targetDirection + 360) % 360;

            // Calculate the angle difference between target direction and current direction
            float angleDiff = targetDirection - direction;

            // Normalize the angle difference to be within the range [-180, 180)
            if (angleDiff >= 180)
                angleDiff -= 360;
            else if (angleDiff < -180)
                angleDiff += 360;

            // Determine the direction based on the angle difference
            int rotationDirection = 0;
            if (angleDiff > 5)
                rotationDirection = 1;
            else if (angleDiff < -5)
                rotationDirection = -1;

            // Set the direction value to your desired input or use it as needed
            // For example:
            ksReactor.InputManager.SetAxis(Axes.TURRET_ANGLE, rotationDirection);```

[leaden ice]

rather than reading euler angles use something like this #archived-code-general message

[static matrix]

different question: Can I change the material of a particlesystem on runtime?

[lost dove]

@leaden iceThank you. I cannot use signed angle however, at least not in that manner because I will be dealing with a single angle coming from the server, telling my client that "my turret is rotated to x rotation", and according to that turret rotation, I need to determine whether or not it is facing my mouse, and then determine how to reach it with an axis, so that the player can let the server know where to put it.

I'm using my player ship angle presently as a dummy until it is perfected.

[leaden ice]

your player ship angle is reading an euler angle back

which is the cause of your angst

is that what the server will be doing?

The server should use SignedAngle

[lost dove]

The server is just updating a float that starts from 0.

[leaden ice]

You can also try https://docs.unity3d.com/ScriptReference/Mathf.DeltaAngle.html on it

[leaden ice]

even better

then you won't have euler bouncing nonsense to worry about

[lost dove]

oh wow

delta angle looks promising

[lethal vigil]

How can I make sure my game stays the same on all mobile resolutions. Its works fine most phone devices but if someone was to play it on the ipad, it cuts too much from the left and right side. My game is a endless side scroller. I know there are anchor stuff I can do which works fine for ui but how can I made sure all game objects also fit in all screens

[leaden ice]

The issue is it can't really be the same if the aspect ratio changes

you would have to do letterboxing or something to force an aspect ratio

[lethal vigil]

is that a normal practice?

[leaden ice]

you can also do some math and set the camera's orthographic size according to the aspect ratio to show what you want to show

[leaden ice]

either changing the camera size or letterboxing

one of those

or changing the shape of the world according to aspect ratio

[lethal vigil]

what would u suggest?

[leaden ice]

It's not my game

probably just setting camera size

The size value is explained here which you can use as the basis for any math you want to do with it:
https://docs.unity3d.com/ScriptReference/Camera-orthographicSize.html

[daring cove]

Thanks It was flipped haven't seen that before, its been weeks 🙃

I'm having bit of trouble figuring out pathfinding for vehicles.
Is there a way to make NavMesh.CalculatePath not find the quickest route, but calculate a more central route. I already have baked navmeshes with ideal values but the pathfind algorithm always manages to find a way get off the navmesh by taking the very edge of the navmesh and clipping a fender on the side of house.

[quasi crow]

Hey, just curious - how do you guys structure your game managers? I have seen some enum-based game managers and I also heard about finite state machines, is it a good idea to use one along with a game manager?

[leaden ice]

"Game Manager" is an extremely vague concept.

It's different in every single game I make

some games need a more formal state machine

others don't

[quasi crow]

I see... and considering im making a small endless runner game - do you have some tips for that?

[leaden ice]

no need for a complicated state machine

Probably two scenes

Menu scene and Game scene

[quasi crow]

Yeah, something like that, so I guess the manager should just handle switching the scenes and maybe toggling some pause and game over menu?

[leaden ice]

perhaps

basically oyu're either:

• in the menu scene, menuing
• in the game scene waiting to start the game
• in the game scene playing the game
• in the game scene paused
• in the game scene looking at a gameover UI

so that game scene stuff could be a state machine if you want

but it's also simple enough it doesn't need to be formalized much

[quasi crow]

alright, sounds fair enough, thanks for your help! :D

[random oak]

does unity support nullable types?

public bool? MyNullableBool()
{ return null; }```

[leaden ice]

Unity supports everything in C# except a few noted exceptions https://docs.unity3d.com/Manual/CSharpCompiler.html

[swift falcon]

direction = tempRotation.eulerAngles;
rb.AddForce(direction * 500f * Time.deltaTime * ai.projSpeedBonus);```

[random oak]

thanks!

[swift falcon]

any ideas for why this will only add force upwards?

[leaden ice]

transform.rotation.z is nonsense

rotation is a quaternion

don't read .z from it and expect anything useful

[somber nacelle]

don't multiply your forces by deltaTime, and reduce that amount of force when you remove that multiplication

[leaden ice]

Another thing is that euler angles are NOT a direction vector

[swift falcon]

aight so how would I go about getting the angle from the rotation?

[leaden ice]

what are you trying to do

project the rotation on the x/z plane?

[swift falcon]

essentially just move it forward on it's z axis

it's in 2d so i cant use forward

[leaden ice]

Vector3 forceDir = Vector3.ProjectOnPlane(transform.forward, Vector3.Up);```

oh 2d

then

[swift falcon]

and transform.right gives me issues for some reason

[leaden ice]

what kind of issues

maybe you want transform.up

it dpeends how your sprite is laid out

[swift falcon]

I'll try it out

[leaden ice]

also yes get rid of Time.deltaTime and 500

[swift falcon]

yeah i gotta fix all my rigidbody settings in this project still

this is the first game I made so a lot of the early stuff is messy

I just set the direction to transform.right and it worked this time thanks yall.

[hushed dust]

If you're making a game in which the player gets surrounded by dozens of enemies, and they make footstep sounds when they walk... what's the most common solution to preventing a cacophony of footsteps?

[leaden ice]

Use an audio mixer and make sure you normalize the input

https://docs.unity3d.com/Manual/class-AudioNormalizeEffect.html

Otherwise they'll all add up and break your eardrums

you can of course also do some kind of pooled audiosource thing and limit the number of concurrent active audio sources

[hushed dust]

yeah i do that part already

normalize is a good tip though. do you generally apply normalize to all your mixing groups?

[leaden ice]

I'm going to be real with you, I only ever had one project that got far enough where I needed to care about such things.

[hushed dust]

haha hey thats one more than most 😄

[leaden ice]

For context: in that particular case it was sort of a tetris / bejeweled style game and when the blocks were eliminated they would burst into particle systems (sort of like a firework explosion). Those particles would fly over to the score counter UI and each particle would make a small sound when it finished doing so.

When you scored a big combo all the simultaneous sounds were maxing out the speakers and very unpleasant. Using the Normalize effect made it sound better.

[hushed dust]

@leaden ice oh man... i also did not realize you can add duck volume to a mixer group and add a send above it and send it to itself

[sturdy thorn]

Excuse me, I'm finding an issue when I try to make the enemy wander randomly they get stuck within the obstacles, I wrote the code that sets the destination anywhere else when they touch said obstacles, but they wont avoid them and will just get stuck.
[18:32]
https://pastebin.com/fUkn47gp

[mossy gust]

How could I save a list to a json file?

[leaden ice]

use a json serializer such as Unity's JsonUtility or Newtonsoft JSON

[devout nimbus]

public State.CombatState CheckCombatState() {}```

I assumed you could have methods with return values on interfaces but I get "Not all code paths return value"
The only place the interface is implemented returns a value. Are you just not able to have return values with an interface? Been a while since Ive touched C#

[somber nacelle]

you're not returning anything there

[devout nimbus]

Yes it is the method on the interface

[somber nacelle]

well you've provided a default implementation

which you are not returning in

[devout nimbus]

I have to provide a default return value? Odd

[somber nacelle]

you do if you provide a default implementation for the interface method. if you do not wish to do so remove the curly brackets and replace them with a semi colon

[devout nimbus]

It works. I assumed you wouldnt have to with the default implementation

That works also

Thats stupid imo but it works. Thank you

[somber nacelle]

you either declare the method without a body so it does not have a default implementation and your implementing objects have to provide their own implementation for the method, or you can use the default interface method where you've declared a body for the method so it must be a valid body, including required returns

[devout nimbus]

Makes sense

[dusky cosmos]

Shader error in 'Universal Render Pipeline/Lit': maximum ps_4_0 sampler register index (16) exceeded at INVALID_UTF8_STRING(11) (on d3d11)

I get 5 similar errors while getting build

[rotund moat]

how to lock character in one rotation

[leaden ice]

simple - don't ever change the rotation of the character in your code.

[primal wind]

Also if it's a rigidbody you would need to add a constraint on rotations otherwise the world could interact with it in such a way that it would rotate

[copper blaze]

does anyone know how to autoformat c# code in visual studio code such that the curly braces start on the same line as method declarations? (sorry if i'm asking this in the wrong channel, not sure where else to ask)
example:

private void method1() {
  // i want this
}

not

private void method1()
{
  // not this
}

[somber nacelle]

if only there was some website with some sort of documentation for vs code so that you didn't have to break server rules by answering questions with AI generated responses.
oh wait https://code.visualstudio.com/docs/editor/codebasics#_formatting

[unreal trench]

didnt know, sorry

[fathom plank]

I created this little healthbar rendering manager script, it works well enough, but feels somewhat awkward. Basically the intention is that healthpools are just health amounts current/max, etc for a character. The HealthBarManager class is attached to a canvas, and determines which healthpools should currently be rendered, takes care of making sure the renderers are in place while the healthpool is on screen, recycling them when they are no longer on screen. Anyway, just looking for any feedback on the design of this 🙂

[wintry crescent]

Camera.main is super slow. Don't call it every frame, cache it

[copper blaze]

ok better question what's a good formatter to use to format the code since vsc doesn't have default formatters for c#

[buoyant crane]

https://drlongnecker.com/blog/2018/11/vscode-omnisharp-kr-formatting/

[lean sail]

while it is better to cache it, it is not super slow. It was changed years ago so its not using FindObjectWithTag anymore, it is internally cached.
https://docs.unity3d.com/ScriptReference/Camera-main.html

[wintry crescent]

ah, good to know

I can never remember, help
If I have a class A, and a class B:A
and I have
A b = new B();
and I call b.GetType()
will I get A or B

[buoyant crane]

you will get B

[wintry crescent]

thanks
it'd make no sense if it was A instead, but I keep forgetting anyway

[lean sail]

im not entirely sure but wouldnt it be better to just give everyone a healthbar renderer and simply/enable disable it when they are out of view. Then just keep pooling for assigning/recycling the health bars when they die or are spawned

overall i dont see anything wrong though

[fathom plank]

reason i didn't do it that way was because the healthbar is rendered on the canvas and i didnt want the characters to be children of the canvas, maybe there is a better way of linking 3d objects to canvas object though 🤔

this is the sort of heiarchy ive got currently

[lean sail]

If it introduces complications then I probably wouldnt worry about changing it unless you had a ton of enemies and needed to optimize

You could probably just keep the current setup, and disable the healthbars on the canvas while the enemy is offscreen, while just not updating the healthbars position

When they come back on screen, enable and update position

[fathom plank]

thats currently what happens with the pool manager class. when the healthbar goes offscreen, it gets tossed back in the pool so whenever any other character with health comes on screen, it can be assigned one of the available healthbars, i figure it was a good idea since not all characters will be within the viewport at all times so this keeps the number of objects down a bit

so like if all 4 of those tanks are destroyed, the 4 healthbars will stay attached to the canvas but disabled, when it new tanks are created, as they come on screen, the healthbars will reactivate as needed and get assigned to the tanks as they come on screen

[hoary swift]

anyone knows if there is a method like AssetBundle.LoadAllAssets<T>() for addressables which takes an IResourceLocator as a parameter?

[ivory ginkgo]

Are there some better practices to acquire lists of players/enemies/entities/items that are available in the level other than the basic options like below?

GameObject.FindObjectsOfType(typeof(SomeScript));
GameObject.FindGameObjectsWithTag("SomeTag");

Is there a better way to cache these items in a Singleton List on spawn or something? I'm not sure on the best practice here. Any ideas or references would be great!

[ashen yoke]

which plugin is that?

[quartz folio]

Yes, cache them as you spawn them instead of finding them. Or get your objects to register themselves.

[fathom plank]

? its just the default heiarchy view that comes with unity

[ashen yoke]

its actually pretty cool, and doable

[ashen yoke]

i was joking about the colored lines

[ivory ginkgo]

With a Singleton that handles this? Like a SpawnedManager.cs or something? Or are there better approaches? All of my items already have an OnSpawn Action that gets sent out, so having them register would be easy. I'm just not sure on approach, or best practice of storing those items.

[ashen yoke]

separation of concerns

what keeps track of enemies? what makes sense to keep track of enemies? enemy system? etc

a generic EntityManager is also good, you can use dictionaries by types

[quartz folio]

Sure, whatever you want

The best practices in games are very "do whatever makes sense for you and your setup"

[ivory ginkgo]

that's fair, so the approach really doesn't matter, it just matters that they are cached

[ashen yoke]

depends, you can make a big mess if you dump everything into a single class and use direct references to strongly typed lists

if it can be neatly generalized, abstracted behind a single api its good, healthy

[ivory ginkgo]

I might try an EntityManager with a dictionary or lists of each type and see how that goes. My project doesn't have a lot of different types, it's pretty simple. Appreciate the help. May seem like a simple thing, but I like to use good/best practices where I can, so I was curious.

[native copper]

I need help and I wasn't sure which channel to put this in. I made a lego wall in blender and in unity I wanna make it so when the player shoots a rocket at it the wall blows up into a bunch of lego bricks. I can make it so when the player shoots the wall the wall disappears and a lego block instantiates but I want it to instantiate a bunch of different colored blocks and have them spread out in a bunch of directions

[cosmic rain]

Instantiate a bunch of them the same way you instantiate one. Then make sure they have rb+collider components and apply forces to them.

[lean sail]

you can just have a prefab of lego bricks thats roughly in the same shape of the wall, then replace the disappeared wall with the broken up one. Apply explosive force manually depending on which area of the wall was hit

[native copper]

well how would i get the instantiated ones to have a random pool of colors

[leaden ice]

Use Random.Range to pick a random color from an array or use https://docs.unity3d.com/ScriptReference/Random.ColorHSV.html

[native copper]

thanks

[lean sail]

wouldnt you want it to have the same color as the wall?

or else the user will just see it change colors randomly

[native copper]

yeah i guess your're right

[rancid frost]

does anyone know why my vertical meshes (the blue color shapes) are see through?

[leaden ice]

looks like you only made geometry on one side

triangles in meshes are single-sided

unless you use a mateiral that renders backfaces

[rancid frost]

hmm, shouldnt I be able to not see the blue faces at all then?

When I flip it back, it looks fine

But the order is wrong

top is down and down is up

[ashen yoke]

thats generated?

[leaden ice]

I think the normals in this first image are inverted for some portion of the mesh

[leaden ice]

in this image the normals look correct

[ashen yoke]

no thats not normals

[rancid frost]

yes

[leaden ice]

what does "flip" mean in this context?

[rancid frost]

flipping the x axis

of the meshes

[leaden ice]

that flips it upside down basically

[ashen yoke]

where triangle will be facing is dictated by which order it was constructed

clockwise and counter clockwise

[leaden ice]

so you're seeing the back of many of the faces when it's flipped

which is generally not visible

unless using a special material

[rancid frost]

no, im using the sprite defalt

[leaden ice]

you're using a sprite material for a 3d mesh?

that's probably part of the issue too

[ashen yoke]

are you saying you made a hexagon out of sprites?

[leaden ice]

the normal mateiral would be Lit or Unlit (or Default in BIRP)

[ashen yoke]

the only way it will work is if the material is opaque

or at least cutout

otherwise the alpha sort will fail at specific angles

im ignoring the fact that this is the weirdest way yet ive seen anyone construct a hex

[rancid frost]

the problem is this
The blue verticals are rendered from TOP to bottom,
so a hex that has a higher Y position will add blue verticals to its side

however, the generated mesh, with default orientation it looks as though the blue verticals are generated bottom to top

[rancid frost]

Nah, they are meshes

[ashen yoke]

read on "winding order"

[leaden ice]

ok so you[re not doing procedural mesh generation

you're just using a fuckload of meshrenderers

[rancid frost]

yes

[leaden ice]

btw why are you using Plane

instead of Quad

Quad would be way simpler

[rancid frost]

ok, will use quad

[leaden ice]

anyway it just looks like your code is not positioning/sizing some of these planes properly

and when you flip it upside down (why?) you won;t be able to see the backs of the planes

because you can never see the backs of the default plane mesh

(barring a backface shader)

[rancid frost]

the position are inverted,
for this picture +Y goes downward and -Y goes up, same with the Z axis

If i try and flip the image to properly match the axes, the blue things dont match

[leaden ice]

Is there a question in here somewhere?

[rancid frost]

I dont get why its flipped, my triangles are rendered clockwise

[leaden ice]

it just looks like your code is positioning things incorrectly

it's unclear what data we're starting with to generate this stuff

and what the code is actually generating

and of course, we haven't seen the code at all

[rancid frost]

if (Position.y > hex.Position.y)
                {
                    (int, int) sData = OppositeCorners[i];
                    (int, int) data;

                    data.Item1 = i;
                    data.Item2 = (i + 1) % 6; // so the last one loops back to the first one
                    
                    // the current hex's side vertices
                    Vector3 pos1 = GetSideVertexPosition(data.Item1);
                    Vector3 pos2 = GetSideVertexPosition(data.Item2);
                    
                    // the surrounding hex's side vertices
                    Vector3 sPos1 = hex.GetSideVertexPosition(sData.Item1);
                    Vector3 sPos2 = hex.GetSideVertexPosition(sData.Item2);

                    int sv = SlopeVertices.Count;
                    int v = Vertices.Count;

                    SlopeVertices.Add(pos1);
                    SlopeVertices.Add(pos2);

                    SlopeVertices.Add(sPos1);
                    SlopeVertices.Add(sPos2);
                    

                    //SlopeTriangles.AddRange(new int[6] {
                    //    sv + v + 0, sv + v + 1, sv + v + 2,
                    //    sv + v + 2, sv + v + 1, sv + v + 3
                    //});

                    SlopeTriangles.AddRange(new int[6] {
                        sv + v + 3,  sv + v + 1, sv + v + 2,
                         sv + v + 1,  sv + v + 2, sv + v + 0,
                    });

                    SlopeColors.Add(SlopeColor);
                    SlopeColors.Add(SlopeColor);
                    SlopeColors.Add(SlopeColor);
                    SlopeColors.Add(SlopeColor);
                }

What the code is doing is getting those 4 corners and setting vertices and triangles

it does it for each side of a hex

[leaden ice]

so wait you are doing procedural meshes?

[rancid frost]

So in summary,
for each surrounding hex, we check if the surrounded hex is lower than the current hex, if it is draw the height vertice

[leaden ice]

I'm confused because you showed the Plane mesh before

[rancid frost]

yes,

I want to replicate the tilemap system

but with meshes

[leaden ice]

I mean there's a lot of moving pieces here

[copper blaze]

added the omnisharp.json file to root and pasted in the provided code and still nothing 😦

[leaden ice]

from this image for example it seems possible that GetSideVertexPosition(data.Item1); could be returning the wrong positions

[rancid frost]

let me verify that

what do you at the very least understand my dilema?

[leaden ice]

I mean I understand that your procedural mesh generation is working improperly

I don't really see what you mean as a "dilemma"

I know it took me about three weeks to get a working procedural mesh system for building hesagonal pipes in my factory builder prototype

procedural meshes are hard

I don't really think "bottom to top" or "top to bottom" matters terribly much - either way it's connecting the lower and higher hex

[rancid frost]

everything is as it,
data is the current corner (the top 2 vertex)
sdata is the matching corners (the bottom 2 vertex)

[rancid frost]

Feels good to know its not just me

[leaden ice]

i mean these definitely look like numbers

no way of knowing from this screenshot if the y values are correct

they're certainly internally consistent

[rancid frost]

I see

[leaden ice]

I'd recommend using gizmos/handles to visualize this stuff in the editor

and make sure it matches up with expectations

btw https://docs.unity3d.com/ScriptReference/Handles.Label.html was the magic sauce that let me fix my procedural meshes. Actually rendering the vertex numbers in 3d space is a gamechanger

[leaden ice]

the question is way too vague to be answered honestly

it depends on what these things are, what they do, how they interact with the rest of the game

answers range from basically... strings to enums to ScriptableObjects

it really depends on the needs

inheritance is usually not the best way, I'll say that

could be

that's kinda similar to SOs

[oblique spoke]

You can replace the strict inheritance hierarchy with interfaces. Create item classes based on functionality and add various interfaces based on the features of the item to expose them to other systems in the game.

[oblique spoke]

You could try composing the items with components or by other means, but settings exposed in the inspector can often handle a lot of variants.

[twin yoke]

guys I need help with an error I am getting and I don't why it appears. The issue is that is it is random:ArgumentNullException: Value cannot be null.
Parameter name: _unity_self
UnityEditor.SerializedObject.FindProperty (System.String propertyPath) (at <4911eca47f294e18a7b3306f02701303>:0)
UnityEditor.UIElements.Bindings.SerializedObjectBindingContext.BindPropertyRelative (UnityEngine.UIElements.IBindable field, UnityEditor.SerializedProperty parentProperty) (at <ddc34215cb194bc6b42a1ae3b66800fe>:0)
UnityEditor.UIElements.Bindings.SerializedObjectBindingContext.BindTree (UnityEngine.UIElements.VisualElement element, UnityEditor.SerializedProperty parentProperty) (at <ddc34215cb194bc6b42a1ae3b66800fe>:0)
UnityEditor.UIElements.Bindings.SerializedObjectBindingContext.ContinueBinding (UnityEngine.UIElements.VisualElement element, UnityEditor.SerializedProperty parentProperty) (at <ddc34215cb194bc6b42a1ae3b66800fe>:0)
UnityEditor.UIElements.Bindings.DefaultSerializedObjectBindingImplementation+BindingRequest.Bind (UnityEngine.UIElements.VisualElement element) (at <ddc34215cb194bc6b42a1ae3b66800fe>:0)
UnityEngine.UIElements.VisualTreeBindingsUpdater.Update () (at <73c3b9fa4da644c9a21a8a16d8e2909f>:0)
UnityEngine.UIElements.VisualTreeUpdater.UpdateVisualTreePhase (UnityEngine.UIElements.VisualTreeUpdatePhase phase) (at <73c3b9fa4da644c9a21a8a16d8e2909f>:0)
UnityEngine.UIElements.Panel.UpdateBindings () (at <73c3b9fa4da644c9a21a8a16d8e2909f>:0)
UnityEngine.UIElements.UIElementsUtility.UnityEngine.UIElements.IUIElementsUtility.UpdateSchedulers () (at <73c3b9fa4da644c9a21a8a16d8e2909f>:0)
UnityEngine.UIElements.UIEventRegistration.UpdateSchedulers () (at <73c3b9fa4da644c9a21a8a16d8e2909f>:0)
UnityEditor.RetainedMode.UpdateSchedulers () (at <ddc34215cb194bc6b42a1ae3b66800fe>:0)

[leaden ice]

looks like an internal unity editor bug

[twin yoke]

I remove a variable that goes a little bit like :
[SerializeField] public Transform parent;
and it seems like the issue is gone for now but I don't why

I was dynamically creating sprites and put them as child under a specific parent I wanted

at least for now, it seems like it is fix. Sometime, it work fine for hours then later I got errors like this

[dusk apex]

Parent is a property of Transform that you cannot set to null. If you've got your own variable called parent, make sure you're setting that variable and not Transform's parent property.

[oblique spoke]

This is a common error seen recently and unlikely to really have anything to do with your code

Updating to the latest patch version might address it

[twin yoke]

ah man thx

I will update unity ig

I have worked with c++ before so getting errors gave me a trauma with null errors

[broken fractal]

I am using this method to rotate my camera and the new input system, how can I go about clamping this rotation so it doesn't rotate more than -90f, and 90f. thanks!

(the value from the input event returns from -1 to 1)

[dusk apex]

Accumulate rotation deltas in a variable and subtract the excess amount from your calculation before rotating.

[broken fractal]

hmm thanks

[quartz folio]

If that's mouse input, you shouldn't be scaling it by deltaTime

[spark flower]

how do i get angle diffrence but wit hpositive and negative value?

vector3.angle gives just positive

[lean sail]

https://docs.unity3d.com/ScriptReference/Vector3.SignedAngle.html

[pure cliff]

Is there any easy/handy way I can add a "test" button as a component in the editor or whatnot, that I can click to invoke methods on mono behaviors and etc?

not a button in the scene, but like one I can click in the inspector? Ideally with as minimal effort as possible just so I can test out that when the method is invoked, it does what I need it to do

as an example I wanna test out Kill() methods for dying enemies in my game, so ideally I just click on the enemy in the editor, swap to game view with inspector open, and click a button in the inspector on the component that'll just call the method directly so I can make sure the method does what I want it to do

[somber nacelle]

if you use naughty attributes or odin, both of them have Button attributes to call methods from the component inspector. otherwise you could create your own with some editor scripting

[pure cliff]

do they add any overhead to the actual game I compile or is it purely just editting tools?

[lean sail]

if you put it in a folder named Editor, it wont make it to the build

[somber nacelle]

yeah you keep all your editor scripting in an Editor folder since nothing that uses the editor will actually be compiled in a build

[pure cliff]

gotcha

[quartz folio]

The easy thing to do is just decorate a method with ContextMenu and then you can right-click your component and call it

[abstract nimbus]

https://codeshare.io/mp9RY4 Is this spaghetti code? 🤔

it's for a chess application

[lean sail]

probably, but i doubt anyones really gonna take the time to go through that and try understanding it, this is a pretty bad way of writing it. Its really not readable, not commented, and you definitely dont need to chain so many linqs together. Imagine u are a new developer trying to understand existing code and this code is the first thing you see

[abstract nimbus]

yeah my first version had many nested foreach loops and if statements but I had too many so I tried to use Linq to get less nested stuff but using Linq was also a mess

[quartz folio]

new { variation, move } is this an anonymous type

[abstract nimbus]

a short explanation of what it does: the string moveNotation as the parameter might look like "d6 Nf3 e5" so what my code does is it looks for a variation that has made all those 3 moves in the same order

[lean sail]

you should just stick to the foreach version, i guarantee it'd be more readable

[abstract nimbus]

ye I needed to return variation depending on what move is

[abstract nimbus]

but I don't like to have more than 3 nested brackets like: { { {

[lean sail]

then make a method for some calculation

[abstract nimbus]

yeah I guess

[quartz folio]

anonymous types are truly disgusting, just use a value tuple

[abstract nimbus]

https://codeshare.io/1YmOkn here is my version with foreach loops

oh I did ToList on the split for no reason 😂

[lean sail]

i think you should probably use one of the !code sites listed, it seems like i can edit that link lol

[tawny elk]

Posting code

📃 Large Code Blocks
Large code blocks should be posted as links to services like:
https://gdl.space/, https://paste.ofcode.org/, https://hatebin.com/
https://paste.myst.rs/, https://hastebin.com/

📃 Inline Code
Surround code with three backquotes. Not quotation marks.
To get C# formatting the first line should only contain cs or csharp.
Add a comment with a line number if there is an error message.
```cs
// Your code here
```
Do not share screenshots of code unless requested.

[lean sail]

just incase someone decides to put something bad in there

[abstract nimbus]

https://gdl.space/uxujidogoy.cs

the foreach version is at least more readable than the Linq version 😂

but it has too many nested stuff

uh you can't edit that one right?

[lean sail]

i wouldnt say its so bad that you should refactor purely because of the nested statements here. I also dont know what the code is doing so its hard to say if you even need all this, i understand the goal but most of these values dont really mean much to me.

yea cant edit

[abstract nimbus]

do you know how chess move notations works?

cuz that's part of what my code does

[lean sail]

yes, ive played chess a bit but i dont know what a chapter is

[abstract nimbus]

a chapter contains variations and a variation contains a list of moves

so it's like a book kinda

so a variation is could be e4 e5 Nf3 Nc6

and if moveNotation is e4 e5 the method will return that variation

but it will not return that variation if moveNotation is e5 e4 because that's the wrong order

so currentComparingMoveNotation makes sure I do it in the correct order

[lean sail]

what are you storing this for, predetermined AI moves?

[abstract nimbus]

the moves are from a book that I've purchased so I pulled all those moves in to my app

so this method is like asking "when do I play e4 and d4" for example

and it's a Dictionary so it doesn't get multiple variations with the same position

[winged mortar]

Yes man

Youve definitely misused linq

[abstract nimbus]

https://gdl.space/uxujidogoy.cs have a look at this version which does the same thing

[winged mortar]

For each looks fine to me

[abstract nimbus]

okay

[lean sail]

im still confused what you would actually use that for, sounds like some chess learning app. From the problem it sounds like you do need all the nested statements, but i almost wonder if itd be better storing all the related variations in one collection then searching through that instead of calculating a subset everytime.
Chess related calculations always suffer from performance/memory issues though so you'd probably have issues either way

[abstract nimbus]

the reason I made this method is because while I play chess in the opening stage, I somtimes ask myself "when do I play d6 in this opening?" Or "when do I play both e6 and d6?" and this method gives me all the variations that answers questions like that so it's just a way to find what I'm looking for quicker

[lean sail]

all the sites that do something similar to what youve described just have massive databases, at least from what ive seen

its a memory vs performance tradeoff, where these sites have the money for memory

[abstract nimbus]

there are many apps that don't have all the features that I want so I made my own app with a mix of features that I want lul 😛

it's also for personal use only

[winged mortar]

Honestly you are rpobably better off doing some regex thing

Instead of this

[abstract nimbus]

hmm

that's interesting

[winged mortar]

You have a list of all of the variations

[abstract nimbus]

yeah but the list of moves are in the list of variations tho

[winged mortar]

If you are looking for kd4 rf2

[lean sail]

you would have a list of every variation, which is what i suggested here

[winged mortar]

Just find it

Yeah basically a big database

Which isn't much different from what you have right now

[junior delta]

Hello

[abstract nimbus]

true 🤔

[lean sail]

especially if its just on your pc, memory shouldnt really be an issue

[winged mortar]

Depends on how many pages the book has that he purchased

[abstract nimbus]

this is my "database" 😂

[lean sail]

you can easily hit 1m+ lines on a database, though with chess i imagine that'll be not much

yea thats not much

my first job i opened a "database" of 2 million entries with 50 columns each. This is lightwork

it was in excel too

[winged mortar]

Ye this shouldn't be a problem

[winged mortar]

Hence the "database"

[abstract nimbus]

but I'm using SQLite as well so the first time I run the app I read from the json files and add them to the SQLite database and the next times I don't need to read from the json files

[lean sail]

it did take like 2 minutes to open though so id go get free coffee everytime i had to open it lol

[winged mortar]

And this where your caffeine addiction started...

[winged mortar]

That's cool, I guess

But you can just have it in memory if you have a simple structure

Isn't sql lite in memory as well?

If it isn't you'd still be dealing with your IO

[lean sail]

i would still just add each variation, im mostly sure sqllite will be fine but you should run through it and calculate how many lines you'll be storing to make sure its not like over a few million

ive never used sqllite so i dont really know about it specifically

[winged mortar]

Also if performance is important look into an algo like this one

[gray mural]

Hello, do you guys have any ideas how to make level creator mini-engine in Unity?
Probably like e.g. level creator in Geometry Dash, where players can create their own levels and play them.
Let's not talk about making it multiplayer, how do I implement this so that just I can create levels in my own game?
Everything that is instantiated in the game - stays just in game and doesn't affect editor.
How do I save levels and then make me able to use that levels?

Even e.g. instantiating prefabs on random position when pressing space button. Then when game is finished and prefabs are spawned, this (I think) Scene should be saved somewhere.

Maybe that shouldn't be done in game, just in editor ?

Thanks for reading

[winged mortar]

https://en.m.wikipedia.org/wiki/Knuth–Morris–Pratt_algorithm

You wouldn't create a new scene