R0J0hound's Recent Forum Activity

  • What problems? You forgot to specify what they were. I have some down time to guess though, but I may or may not be around to reply again.

    Last I checked the drawing canvas object doesn’t store depth, it only stores color. So pasting 3d objects results in the 3d faces drawn in an arbitrary order instead of sorted with a z buffer.

    If you mean pasting stuff onto a distorted drawing canvas. I think that never worked, or worked as if the canvas wasn’t distorted. But maybe things have changed with newer updates.

  • You probably could do it with a 3d camera too. That would let you use normal text.

  • The description of what you’re after is still pretty vague. However, you can solve any problem by breaking down what you want to do into smaller doable steps.

    It looks like maybe you want to drag icons from the left onto the top. Maybe you could give the icons the drag and drop behavior. Then when you drop them (the one drop trigger) you’d check if it’s overlapping the top bar, and if it is then do something to position it in a nice way. And it’s not overlapping the top bar on drop just move it back to the left toolbar. You could also just move it back to the left toolbar on drop but create a duplicate on the top bar.

    To position them in a nice way on the top bar you could pick all the icons overlapping the top bar, then “for each” the icons, and set the icons position to something like (topbar.x+icon.width*loopindex, topbar.y). That would put them in a nice evenly spaced horizontal line.

    Just some rough ideas that could be a possible starting point.

  • Set rotateSpeed to lerp(rotateSpeed, 0, 0.05)?

  • Couldn’t you just use the bullet behavior, and rotate behavior but gradually reduce the turning rate to zero?

    Behaviorless is similar and easier for me to write off the top of my head and would be: 

    Every 1 second
— create bullet at sprite.x, sprite.y
— bullet: set angle to random(360)
— bullet: set turningRate to random(30,120)

Every tick
— bullet: move forward 150*dt pixels
— bullet: set turningRate to lerp(self.turningRate, 0, 0.05)
— bullet: rotate self.turningRate*dt degrees

    If you used behaviors then you’d basically reuse the action that sets the turning rate to some extent.

  • Try Construct 3

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

    Try Now Construct 3 users don't see these ads

  • I mean ignoring hierarchy, waits and such to have the fireball shoot one way or another based on if the enemy is mirrored could be simple as: 

    Every 1 seconds
For each enemy
— enemy: spawn fireball at imagepoint “mouth”
— enemy: is mirrored
— — fireball: apply impulse at angle -30
— else
— — fireball: apply impulse at angle 180+30

    Anyways, the relevant part is the “is mirrored” check. And generally if logic works for one enemy but not consistently with multiple then you can put that logic under a “for each enemy” event.

  • Maybe they are trying to make a merge game? And whatever they tried only works for one instance based on the “note”.

    We are missing information.

  • Couldn’t you just do “$”&TotalCash/100

  • Nope. The only vanilla plug-in way to do it is with mesh distort with some math. Beyond that you have to use 3rd party plugins. Reach out to the devs of the plugins on discord. They may very well be able to port it to v2 or the sdk.

  • I can never recall the name of the plugins.

    If you mean the 3d shape object that comes with construct then no, it only lets you rotate around the z axis.

    If you branch out to third party plugins such and “3d object” or third party behaviors such as “3d rotate” then you can rotate around more axis’ fairly easily.

    I haven’t really used them though since I’m more interested in exercising my mind to be more familiar with the math involved.

  • Since I made the example linked in the original post here it is tweaked to allow for tweening:

    dropbox.com/scl/fi/s234ich05y6bkli6uw6ws/card_flip3.c3p

    Hey! In Construct 3, since it's a 2D engine, simulating Z-axis movement usually means faking depth using scale, position, or layering.

    While it's true construct is primarily a 2d engine it indeed has real depth and a depth buffer if you want to utilize it. The original example and the updated example take advantage of it.

  • I mean at a glance it looks to be a long and complex solution. Probably can be simplified a lot. But that would require breaking it down to figure out exactly what it’s doing and why.

    But I guess I’m not a fan of llm ai code. Debugging, testing and cleaning up code is time consuming and I’d rather not do it if I can.