jayderyu's Forum Posts

  • jayderyu

    I include it. But it will be one of the last things I add. I honestly can't think of many reasons you would want that "billboard". Seems like a niche use case. Why not prerender something onto a spritesheet?

    It might be useful for equipping armor in rpgs. You could create a new sprite-sheet when something is equipped on the 3D model rather than storing hundreds of sprite-sheets for every possible item combination.

    I'm thinking there might be a focus on the wrong part of what I was writing.

    If you render to texture to display an image. You don't need to update the image everytick. You only need to update the image on when there are key moment of change. Then that removes the high cost of rendering multiple objects at once.

  • > Good luck.

    > I had this in mind. I started out with it actually. I will most likely include it too. But the performance hit is quite massive because I have to draw the 3D onto a 2Dcanvas and then create the texture for the "sprite" and apply it. And that 60 times per second. I did a lot of testing and it gives very big performance hits. However I will still include it once I am done

    >

    > Here is an example of how it has to work.

    > http://jsfiddle.net/Schoening/XxzS7/6/

    > You can see how much resources the 2D texture function uses. And this is not even applying it to a Sprite!

    >

    The system is called billboards. Keep in mind you don't need to render every tick. Instead monitor for an update.

    Rotated

    Light change

    camera perspective change

    animated

    unless those key events occur to the 3d object. Then there is no need to render the image again. If you only render on required you will see 3d rendering through the roof. I remember once back in the erly days of JavaMonkeyEngine before billboards, and only render on required. There was a test to render a billboard, but only render to the billboard on any of the above conditions. The render jumped from 100fps to 500fps. It's super cheap and innefeciant to only render to texture when needed.

    I was going to do this stuff myself some day when the need arised for my DOL project(below). But it's months down the road. And if someone does this well ahead of time. All the happier to give credit to the plugin creators

  • GiveYourFontsMono in the forum search here. It's an app that produces SpriteFont format fonts. The default font isn't really meant to be used practically. Unless your doing retro games.

  • Well I'm up for giving this babysitter job and the occasional solution a go around.

  • Try Construct 3

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

    Try Now Construct 3 users don't see these ads

  • I don't know why Ashley doesn't use this method for compiling APK's for C2 IDE. It should be. NodeWebkit uses a similar method to this. So why not. Oh well.

  • Probably because you accessing it as a file in dropbox rather than the web access service.

  • OP, I found your capx to have a lot of small problems. There are a lot of good suggestions in the thread here that you should do.

    C2 doesn't support more than 8/9 collision points. Squares only need 4.

    Family up your events(ie no blue on touch, blue on red.. that's bad, bad bad bad bad. )

    Your Evertick eats too much.

    I've come to testing design for my personal computer. Create a test app. run the test app on your minimal target device. once you get the performance you want by adding features. note down the CPU usage on your computer for the exact same app.

    As an example using my PC and iPodTouch4g.

    My PC will have a CPU usage of 3% to maintain 55+fps. Any more than that will exponentially drop performance.

    So your EveryTick often hits my computer with 5.8% here and there. While often fluctuating between 2.8 to 4%. So when it hits 5.8% it starts to cascade the performance.

  • Good luck.

    But I want to add. I personally would like a double option for 3d. All 3d Plugins so far have been canvas renders based. In which the plugin is built to render worlds. I would really like to see Render To Texture and use the Texture/Image. That way I have more options. So i'm not looking so much as a world renderer, but more of an object renderer.

    Some one recently posted that Nintendo called them. I asked about how long it took. Apx 3 months. Also others thankfully chimed in too. They will get back to you in apx 2 to 6 months.... I'm still waiting

  • This is a demo game for my friends school project. I think it looks good. I suggest playing the game with good speakers and use Chrome or FF.

    https://apps.facebook.com/thebluecode/

  • I kinda agree with the OP, but nothing is going to change unless you can prove to Ashley what your talking about.

    I've voiced that C2 should use an Image library. That way any object can reference any image. Including being able to create dynamic animation sets. Also I voiced that an Image Library can be easily atlased. However the idea was shot down.

    Ashley claims that atlasing does not have worth while improvement gains. Yet atlas is one of Unity pro's big features for the cost. I've seen Unity atlast and the images are rotated every which way to fit best. Also WebGL rendering has no overhead for rendering rotated images. It's just UV coordinates.

    If you want to prove your performance point which I do agree with. I suggest creating your own intensive JS webgl image rendering app( just a test). Run it with a better packed sprite sheet, then run it with a C2 exporter spritesheet. And then display the cpu/gpu/mem usage for both. If you can and I hope you do prove the benefits. Then we will see the change. But this is one of those subjects that Ashley is firm on.

    Also to let you know. Each Sprite Object runs it's own sprite sheet.

  • GameClosure promise was based on better dev and performance for Canvas rendering. however that was all lost when Ludei CJS came out with WebGL. GC also had a few of the same limitations as CJS.

    So

    CocoonJS has, WebGL, C2 Plugin

    GC has, better Canvas renderer and JS engine. and effort to port.

    GC was the top of my list for awhile.

  • Well using the FGL plugin won't work from preview. Because the FGL sdk should be loaded prior to running C2. The plugin doesn't load the sdk as far as I know. My capx force loads the sdk, and my sample holds to a black empty screen until the sdk loaded. I suggest for development using my CAPX force load group, but using FGL plugin. Since there is an official plugin my pure capx implementation should retired. but I'm leaving it up so every can access the Forceload feature.... use it on black screen

    fldr

    Yes. My game passed QA... and using my CAPX only implementation(don't use now). Though it's migrating to use the plugin.

    Is your game landscape? Silk does not lock orientation. During the mobile online testing phase you will be asked what orientation your game plays in. Make sure you choose only landscape. FGL supposedly implemented there own orientation stuff.

  • BluePhaze is that the same spritesheeting in the reference. you can still put the parts of the animation onto an ATLAS. That way the atlas can be used by the webgl renderer for more effeciency. And personally i'm still not convinced that Spriter couldn't use an atlas. An object is not the same as an image. An image can use a frame/UV coordinates of an atlas(that C2 creates) to be more efficient. Where as Spriter treats every single object, and c2 treats all different objects as a different texture for the renderer. Which is unessecarry overhead.

    Personally I think C2 should use more use of atlas-ing. But that's just me.

  • FGLmatt

    Could you please pass on to the programmers.

    Thank you very much for your efforts for a C2 plugin. This plugin will go very well over with the community. Although could please add more "isFeatureXEnabled" also added the expression. That way we as developers could use the plugin with inline coding.m Thanks you guys rock