Posts by rocketappliances

I tried that before and it seemed to be even slower than just using active objects... Maybe I need to test it out again.

Hi chrilley, you game looks amazing. I was wondering for your layer of tiles I was wondering if you used multiple active objects to make your tiles? If so how do you do multiple layers because I would think that would bring your game to a crawl because you would have to have 4 times the tiles.

Hey King Cool,

I wanted to see if I could implement your push out or event collision methods and I am trying to figure out if I could actually use them. As I agree moving one pixel at a time is not efficient vs/ resolving collisions.

See below my example scenario where the player has tunneled through enemies and walls in a movement.
Please login to see this attachment.

The only solution I can think of is this. Move the object freely applying the forces. Upon detecting a broadphase collision (purple box, from P0 to P1) start a fastloop 1 pixel movement from P0 to P1 until you hit the first collidable.

This example gets even more funky if the enemies are moving too :(.

What are your thoughts on my options on this.

Resolved it by looking for another active that has the same way point movement and then sharing the exact lerp time.

I hear you casleziro. I ran into this problem before. I use my arrays for collision so this might not help me. Once I figure it out I will share my learnings.

Congrats Outcast hope it does well!

This looks like its going to be an amazing game. The artwork style and use of color, effects, is superb. Hats off to what you are doing. You are currently what I want to be. You have a customer in me.

Thanks schrodinger for the example! I am using a similar type of system. Amiman99 I have been using floats this whole time and you are correct if I don't I will get nil instead of a fraction.

Right now I am getting variation of approximately 0.016 to 0.048 between objects which should be syncing. It may be due to the floating point precision not being the same. This may seem like a small amount but when these objects move fast it makes a huge difference.

casleziro - when do you think I should round? I feel that if I round I at all in my calculations that it will just be more off than it already is but I would like hear your thoughts. Currently all my positions are maintained in an array in which I multiply the float by 1000 prior to saving it to the array. I then multiple the array value by 0.001 when bringing it out.

Hi everyone,

I have a game where I dynamically make actives which cycle through way point lerps of various times. These objects are moving platforms. I want to sync all of their movement together. Currently when I make the object I set its lerp percentage to gametime mode timetolerp ÷ timetolerp. This gets me very close to having all objects at the same time but I see slight variation between objects likely due to rounding.

What solutions have you used to address this issue? My next approach might be that I check what percentage an on screen object has and match all new objects to that timing exactly. It would just create another variable. Let me know what you think.

Other notes my game will create and destroy objects based on how far they are away. Hence the need for sync as otherwise platform timings between platforms will be messed up and going in oppositte directions potentially because of the order they were created.

Thank you DaveC! I was trying to use this in the wrong way now I can use it correctly.

Can you post a simple example of how this works. I tried the load array with a file selector and then tried to assign the FileSelReturn$( "File" ) to a string alterable value and it doesn't work.

What are you using these curves for? I am just trying to understand how they are used in games.

King Cool I haven't gotten a chance to deep dive back into this yet. However, during my lunch I read this article that will help both of us and anyone. reading this article. Please login to see this link.

My system does a for each "enemy"
Then does fastloops on each enemy

So the for each scopes down my fastloops.

Ie
Enemy#1
Run AI fastloop
Run move fastloop
Run overlap fastloop
Run resolve overlap fastloop

Enemy#2 do the same

For your engine do you move all the enemies first then do a resolution on overlaps or do you move each and check overlap on each?

Also what do you mean by reseting of inputs? I will checkout the example when I get home tonight King Cool.

Hi Everyone,

I'm working on a platform game which uses custom movement and a custom array level editor. The game stores the tilemap of the level in one array. A list of the objects in the world is in another array.

This system let's me make large world's.

Right now I setup the calculation of the movement by using an ID to point to my object array to calculate how much the character moves.

My Question: how should I structure how I handle the movement, AI, of particles and enemies and obstacles?

Right now I'm thinking doing the following:

0) collect player input
1) move player based on input
2) run AI for each enemy then move each enemy
3) check for overlaps and resolve overlaps
4) move particles and obstacles through a for each loop
5) check and resolve particle overlaps

What do you think of my structure? What is the best practice for how to structure your game.

Thanks King_Cool! I will try to add the rounding function to my event as all my calculations are done in an array.

Update: It worked like a charm!

Also something that I found which could be useful to others: Please login to see this link.

If you are using these global objects does anyone know if it's better to have one massive enemy object with all spirtes of all enemies vs. A bunch of active objects one per type of enemy which would only have the sprites for that particular enemy?

Hi King Cool,

I made a manual AABB check where I see based on coordinates if my player overlaps. They both use floats. Problem is that sometimes it registers a collision when it should because of floating point rounding errors. How does everyone compensate for these rounding errors?

I have successfully implemented my method. Only problem with my line method is that it doesn't work properly if your prior position is inside the obstacle already. So you must ensure that the prior position is fine.

Thanks for the example casleziro! I will have to do my own bounding box checks like you have done to create the zero lag correction like you have. Enjoy your vacation.