Sargas

I used the platform behavior to move the enemy around already. Like I have tried to explain above. It worked to an extent, but it is very buggy.

I had it evaluate if the player is above or below with comparison, and if it was left or right, using comparisons with x and y values.

What you have, I know what direction to take, but it just doesn't seem the right way to go about it. I have looked all over these forums and there is no sort of platformer based pathfinding.

Lordshiva1948

Thank you for your input, but I think you miss interpreted what I tried to say, a platformer style game not top down. Here is an illustration of the type of pathfinding I am trying to achieve

Where the red is the enemy, and green is the player.

Sure, I could hard code the AI for this specific scenario, but it wont be. As the player can make up the platforms obstacles, walls etc. It is dynamic, not a static level.

Something I have been searching the entire internet is how to efficiently do platform path-finding AI.

I have coded some very messy code comparing the x and y positions of the enemy and the player to find the player using those conditions bundled with the platform behavior, but it just seems very buggy and inefficient, is there possibly a solution to this issue?

When just doing horizontal pathfinding, that is fine, but when you start adding the y-axis in the mix as well as obstacles and walls things get very complex quickly.

I am just stumped

Thank you!

QuaziGNRLnose

Thanks for the input, I am going to try out a simple chunk loading system like Minecraft, utilizing arrays, and comparing them to the players position.

Might be a messy way, but could work

QuaziGNRLnose

If the engine only renders what in the frustum, then why does it seem to run so poorly when I create (example 10,000 cube objects, with a diffuse texture in the animation editor), wouldn't it only render whats in view aka frustum?

Occlusion culling shouldn't matter right? due to the 2d platform nature of my project, as all object faces are to be seen in the screen view.

QuaziGNRLnose

For my game, I was thinking I may have to create some sort of "chunk loading" system for a mass amount of objects present in the world.

How does Q3D exactly render? Does it just render what is in the viewport or will I have to make a chunk loading system for many objects spread out through the entire world to make sure performance remains in tact

A better way to explain it possibly, is how can I load many objects without the CPU usage going through the roof?

QuaziGNRLnose

I cant get it to work for some strange reason.

And okay, I just created a box model and set it to invisible for detection, it seems to be working fine

QuaziGNRLnose

Pretty cool demo!

I cant seem to get the bullet behavior working with Q3D sprites. Is it even possible to use the bullet behavior on them?

And also, for me to use the raycast, it seems I have to have a model present on an object in order for that to work. I tried using an object with model set to no, and raycast does not catch it, is this intentional?

QuaziGNRLnose

Ahh so thats where the Q3D sprites come into play for basic effects?

Have you found a solution to the bug regarding the exporting of Q3D projects?

QuaziGNRLnose

Ahhh! There we go, thank you!

It still seems there is a difference between the actual mouse X position relating to the object created however, Maybe I should just guesstimate a constant number to adjust it?

EDIT: Nevermind, I seem to of got it, but one more question, how do I apply the collider when creating a new object? It doesn't seem to have any collision detection on

those blocks

EDIT 2: I figured out how to get colliders, I just used System -> Create object to create a new instance of the block, it gave me the colliders.

Can the Particle system in Construct be used with Q3D?

QuaziGNRLnose

I think I almost have it, I can place blocks but they seem to be placed way too high using the formula you gave me. Maybe I missed something, so here is my Capx

Thank you for the help!

QuaziGNRLnose

I am making a 2.5d game, I would think that the vector math would be minimal since the game is actually functioning in 2d not 3d, I am just using 3d models to give a sense of depth.

Or can the Q3D sprite object give me a sense of depth if I still want to maintain 2d functionality?

Here is the closest example I can give you on what I am trying to achieve.

That is why I was wanting to know how to raycast where my Mouse would be located in 3d space to place blocks. Just the X and Y coordinates would be needed since the z axis is restricted.