Laserwolve Games's Forum Posts

  • I'm working on an APRG called DaggerQuest. https://github.com/Laserwolve-Games/DaggerQuest

    Pitch: There hasn't been an ARPG that's truly 2D in over 20 years. I'm making one.

    I've made a good deal of progress and would like to onboard other developers to speed things up.

    I'm using Daz Studio to make 2D renders to use as assets. I'm using Selenium Java to automate Construct itself and GitHub for version control/issue tracking/documentation.

    I can dole out specific work if that's what you like, or you can just clone the repo and submit a PR for anything you like. There's a wide range of stuff to be done so any skill level is fine. Knowledge of Git would be good, but you can use a GUI.

    The game is in a tech-demo state at DaggerQuest.com If the player character and enemies don't appear, refresh the page. (This only happens in an HTML5 export... Good example of bugs I find with a project this big)

    Tagged:

  • Issue: github.com/Scirra/Construct-3-bugs/issues/5895

    Each frame in the sprite's animation has a different size, and origin image point location. That's the only thing I can think of that may be causing this issue, although I don't have any theories as in why.

  • Try Construct 3

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

    Try Now Construct 3 users don't see these ads
  • Scroll down on the properties and click "Make 1:1" to correct that.

    Typically you shouldn't get values like that. I've gotten decimals like that when I change sizes of scaled objects or something like that. But I don't understand how that occurs for unscaled objects.

    A R0J0hound reply usually means my problem has been fixed. Unfortunately I just dragged the sprites in, so they were already 1:1. I still smacked that button just to be safe but I still have the same issue.

  • These are called floating point errors. Unfortunate result of how computers do math pretty much.

    I figured as much, it's pretty obvious it's some sort of rounding error. I think it may be a sub-bug in a bigger project-specific bug I'm combating, so I'm trying to find the reason why.

    If the Width value of X frame in Animation Y on Sprite Z is 173 pixels, it should always be 173 pixels... If it was consistently rounded incorrectly, I could work with that (.1 + .2 = 0.30004)

    The longer the animation plays, the more the number changes. The first playthrough actually has correct values. I'm filing a bug report, I was able to re-create in a .c3p with nothing but a single sprite.

  • Hello,

    Does anyone know why sprites that play animations have height and width values different from that frame's actual image size? There's no scaling in effect. Sprites that are just static images display their height and width correctly in the debugger.

  • Update: I've found the issue. If you give a sprite a child during the runtime, and then try to get it's AsJSON value, it throws that error. I can't recreate in stable, only on r287.

  • Hello,

    I'm trying to save a sprite's state as JSON, and get the following error message. I can reproduce the error only on certain sprites in my project. What would cause the properties of a sprite to be undefined, but only when trying to be read by AsJSON? The sprite's properties are visible in the debugger and otherwise functioning normally.

    Tagged:

  • In my game, there are a large set of high-res animations for each player/enemy combination. To provide come context, I expect the total size of the animations for my game (after C3 processes them) to be upwards of 1gb.

    It's hard to say without knowing what kind of game you're making, but you might want to look into a paper doll system. If you've got PNGs for every combination of player and enemy, your memory consumption will increase multiplicatively (or maybe exponentially?) if you ever want to add new enemies.

    The images directory of my project is 1.1gb. Why does the profiler show 3700mb for images? 3700mb image memory is for uncompressed images.

    PNGs are technically compressed; if I remember correctly they take up 4 times as much space in-game. They're also existing in your GPU's dedicated memory. Keep an eye on that under the "Performance" tab of Task Manager in your benchmarks as well.

    Your best bet is to split everything into as few sprites as possible and get really creative with load/unload system actions to keep memory usage low but keeping gameplay smooth.

  • I just realised that both of the features in this thread were added in beta version r277.

    and here I thought that beta version was special for me. haha, thanks everyone!

  • It is indeed a bug, but it was easy to fix so I already fixed it for the next beta. There's no need to file an issue for it.

    Thanks Ashley, you rock!

  • Got it. I usually like to run stuff by the forum first before posting on the github because 9 times out of 10 its user error. I'll do that if no one else has any input.

  • Hello,

    simple .c3p file demonstrating just the problem:

    1drv.ms/u/s!AsdHCyNoGktqnNwMO-bsIEWNYsR8DA

    press S to delete the sprite, then press D to (supposedly) unload its images ~

    The estimated image memory consumption doesn't go down, the below error message is thrown, and the game freezes:

    Using the object picker to unload images works just fine, is anyone else successfully utilizing the "Unload Object Images (by name)" action?

  • Yes, I think this is an excellent candidate for a suggestion that has very strong chances of actually being implemented. Mainly because it is very easy.

    I'll submit that request ASAP.

    Shortcuts for changing between animation would also be quite easy to implement, but I am not sure what should they be? Any suggestions?

    I've noticed that plus (+) and minus (-) are undocumented keyboard shortcuts to cycle through frames of the current animation; maybe repurpose those to toggle through animations? Or perhaps ctrl-(+) and ctrl-(-)? Really anything works, you're the expert.