oosyrag's Forum Posts

  • It's less about your viewport size and more about how much space you have between obstacles.

  • By default, an undefined cell contains a value of 0. Values outside of the array dimensions also result in 0.

    You can enter your own string or value that you don't plan on using otherwise to define an undefined cell, and compare values to look for your specified value/string.

    For example, in my game I wanted to differentiate between empty or out of bounds in my array. Since I did not plan to use negative values, I filled the array with -1 on start of layout. Then I could compare to see that -1 was empty, 0 was out of bounds, and anything else was a valid data value.

  • So it was just taking them a while to find a path? Consider doing something to ease the burden of the pathfinding algorithm. The A* algorithm should be very fast to the point it is unnoticeable on modern devices.

    Things like increasing your cell sizes or limiting the area and distance of pathfinding can do a lot for performance.

  • Is the delay happening when finding the path before the pathfinding path found triggers?

    You can visually fudge it by spawning them invisible, and enabling visibility on pathfinding path found.

    If they're still getting stuck after pathfinding path found and not moving, you might have some other events interfering with their movement or state.

  • Depending on your game, one of the most data efficient replay systems you could use a deterministic simulation where you only save user inputs to play back. This is similar to lockstep netcode.

    Otherwise, just use local replays. Even big companies generally don't store replays for users.

  • Try editor.construct.net to see if your gamepad works.

    If it does, but not in your project, you probably have an issue in your events.

    If it also doesn't, you might look into troubleshooting your gamepad/os connectivity first.

  • The top face at least can be done with an invisible helper object at the correct z elevation I believe.

    Also not sure if this works, but if you set the 3d object to a negative z elevation equal to it's height the top face should (might?) align with the base plane.

  • Are you sure the trigger is firing?

    To prevent websites tracking you based on the available gamepads, most browsers supporting Gamepad input will report that no controllers are connected until a button is pressed on one of the devices, when On gamepad connected will also run.

    Also note the index is 0 based, so 0 should be the first controller.

  • You might also want to check out the included example.

  • The engine certainly can, but it would be very clunky and inflexible to do it this way.

    Instead, you can simply get the difference of the x position between the ball and the paddle (assuming both are centered) on collision. Then using that difference, apply it to the angle of reflection. This way, you'll effectively have unlimited resolution/precision.

  • Have you tried setting a negative path cost? Although I'm not sure if it works.

  • The line of sight behavior also will allow you to cast rays, and get the angle of reflection from colliding objects. You can then modify this angle with a spin variable for example, or simply distance from the center of the paddle.

    I wouldn't recommend using the built in "bounce of solids" property, as that does not give you much control and treats corners as rounded.

  • Also going to suggest Araknoid has absolutely nothing to do with physics. Use the bullet behavior, set angles of motion based on your own specific set of conditions to achieve the precision you're aiming for.

  • The ideal solution would probably be to get and use unixtime directly on your backend server.

    You may be able to use an api such as helloacm.com/tools/unix-timestamp-converter to do conversions, or build one into your game.

  • Try Construct 3

    Develop games in your browser. Powerful, performant & highly capable.

    Try Now Construct 3 users don't see these ads