Vilehead's Forum Posts

  • Hi guys. I'll try to keep this clean and blunt.

    TiledBg object in C2 is very undeveloped. To the point where it misses a lot of its major functionalities.

    Just recently i have switched from CC to C2 so i was incredibly suprised how handicaped the TiledBG is in C2.

    In CC i could Offset the TiledBG image, rotate it, scale it. When in C2 its all gone. I did some research and i know that you had some troubles even coding the whole TiledBG into C2, but the difficulty should not be an excuse when we're dealing with comercial product, such as C2.

    I would like for you guys to finish TiledBg or at least develope it to the point where TiledBG in CC stands. To encourage you i will give you some examples:

    1. Animated BG without the need of extra animation frames

    You just keep offseting/rotating the TiledBG.

    2. Randomisation

    You could take any seamless texture, make it as a TiledBG and set a Random offset for the graphics to make each level diffrent.

    3. Beams, lasers, energy bars.

    In this case you could have a nice looking "animated" laser effect with evergrowing offset on one axis, simulating energy beam.

    4. Size changing, animated elements such as Interface.

    Like lifebar or energy bar that keeps on moving (offseting or rotating)that changed its size but doesnt go outside of set boudaries.

    Those are just a basic examples, that come to my mind in 2 minutes of thinking. I had a long list somewhere but i couldn't find it.

    I request for TiledBG to be developed at least to its CC version funcationality.

    Regardless of the costs. Its vital and crusial and will not only make many thing easier but will also introduce many features impossible without it.

    Especialy as the whole TiledBG idea/concept is propably as old as gaming in general. Its been here from like always and should not be ignored. Especialy when rival softwares has this. So you are like missing a feature that everybody else has.

    Take no offence from my words. I am not trying to insult anybody.

    Just requesting a vital and powerful feature.

    And please don't tell that all those can be emulated in another ways.

    Sure they can be emulated, but its not the case.

    You can use a car to drive 100 miles. And do the same on foot. But if i have a choice - i prefer the car. Thats what progress is.

    Kudos.

  • just swicthed to Unreal 3 engine.

    sh*t never looked so good.

  • Thank you. It actualy is. When testing i offten found myself actualy PLAYING it instead of testing heh.

    Its about Cows in space, eating rotten eggs. You fight various enemies like meatgrinders, horny farmers, flying saucers, butchers or space cowboys or Fat People. You fight all of them by shotting various kinds of milk from your udders and poop from your butt. There are 22 enemy types, 16 cows, 16 perks, 16 power ups and 8 game modes. The most of the stuff has to be unlocked in Milk Bar by spending points you earned from playing.

    The game cannot actualy be "beaten" per se. Its mostly about trying to survive for all long as you can, eating eggs, killing enemies and gaining expirience/Level Ups on the way.

    Its quite simple, kind a old fasion when games doesnt need to have AAA graphics, million dollar budget and 300 people working on it.

    I am trying to go all the way back when games were actualy just... fun to play you know? When it was a source of joy and entertaiment not semi-interactive movie treating players like retards with invisible walls everywhere.

    The name is Rotten Eggs Eating Cows.

    REEC for short.

  • Couldnt ressist. Its hard to dev a game in total secret.

    Without feedback you recive much less motivation.

    I am sharing this here and here only.

    I still need about a week or two to finish the game.

    Then a week+ for establishing a internet store.

    Intro+Menu

  • Have you tried measuring the cpu usage when moving the mouse without previewing? 5-7% is likely just caused by Windows posting notifications of mouse movements.

    this sounds ridiculous but

    my gawd Rojo lol you are right!

    But this is madness O_O

  • a friend of my confirmed that its not only on my end. The mouse movement in general eats about 5-7% CPU.

    This is a serious issue.

  • Try Construct 3

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

    Try Now Construct 3 users don't see these ads

  • okey but what about the CPU increased usage?

  • Okey after some testing, the delayed Pos updating turns out to actualy be a V-sync issue. The pos are actualy updated correctly. Its the visiuals that arent.

    When it switched the V-Sync into unlimited or Fixed (about 100FPS) the updating is perfect. But thethe CPU suffers greatly.

  • Well basicly i almost finished programming the gameplay and now i started working with some UI. So i wanted to replace the original mouse cursor with some nice custom one. So yeah, ive hidden the cursor and made a sprite object and an event

    +Always = Set SpirteX to MouseX and SpriteY to MouseY

    Now heres what happends.

    If the mouse is being moved rapidly and fast the Sprite is falling behind! And not only that! whenever the mouse is being moved in such a rapid manner - the CPU usage grows preety noticeable.

    At first i though its because i have about 1200 events running, so it creates the delay. But no.

    So ive gone and made a new, perfect clean project.

    Added 16^2 sprite

    Same event. Always>set PosX/Y to MouseX/Y

    If the mouse is still, then the CPU wainting on circles around 95% (+/- 0.30%.

    But when ill start moving the mouse like a crazy monkey then blam: down to 88%(!!!!). Thats a Whoping 7% CPU usage just for mouse movement O_O The heck?

    And obviously, even at such empty, newborn project - the sprite falls behind the mouse cursor.

    And its not a Debugger error - ive checked the CPU usage also via windows Task Manager. It shows the same. At worst the usage is 41% without the mouse movement. With the mouse going crazy - up to 51%. And thats with Dual AMD Phenom CPU.

    1. Why position updating leave the sprite behind?

    2. What the heck is wrong with the CPU usage?

    3. How to fix the CUP usage and/or Sprite Pos updating?

  • Thank you Arima.

    You clarified preety much everything i was concerned about.

    I resumed my work :)

  • Well heres the deal.

    I have the Level layout.

    In this layout, which is the game main layout there is the player controled object and the enemies.

    I have about 22 enemy types. All with diffrent graphics.

    Theres virtualy no way that all of them will ever be on one Layout, at once.

    At first i though i can handle this with Unique Sprites plugin and load the needed graphics as they are put in use, to spare the VRAM. But US plugin end up being quite buggy and risky to use. So i replaced it with regular Sprite object.

    Now, as i tried to avoid using Fammilies, i though that i might use one Sprite object which has all the graphics for all the enemies. But as the object is created - all of it frames and graphics are loaded into the VRAM. Regardless are they being used or not.

    This wouldnt be such a big deal if i would have less enemies. But with 22 enemies, where each has graphics wourth of +/- 4-5mb VRAM this ends up with some nasty high VRAM usage. And as i said - theres never a situation where theres more then 4 diffrent enemy types on the level. So the rest would just trash the VRAM with needless graphics.

    So i though - maybe i have to split the enemies into seperate Sprite Objects. Sure, itll be anoying to set all the PVs and whatnot for each and every one of them, but oh well - no one said GameDeving is sweets&suger. I though that i might store them on a diffrent Storage kind Layout and make sprites global.

    This way i can easly spawn the enemies i need, at the given moment, without wasting VRAM.

    BUT.

    First problem is setting up events for all of them. 22 Objects means all the events x 22. So i though that i might have to use Fam's after all. And just control the Enemy Objects with the Fam's based events. But... i am a bit affraid how well itll handle it. Its not just collision detection. Its PV's control mainly. ALOT of PV's control.

    Additionaly i dont really know how it will impact the CPU.

    I mean the graphics are being constantly loaded and unloaded as the enemies are being created and destroyed. The enemies graphics arent that big. At worst, we are talking about 15 frames TOTAL, where biggest are 128^2.

    So the question are:

    1. Will this constant loading/unloading hurt the CPU?

    2. Will it work if ill replace the exact sprites objects with Fam?

    3. Will it work for EACH of the enemy or for all of them?

    4. Do you know any other (better) way of doing this?

    5. Or maybe is there an option to Force/Set CC so he loads only the selected Animations from the Sprite Object? that would be the perfect solution. Letting me avoid Fammilies and having all the graphics nicely organised within one sprite object.

    If anything is unclear - please ask. I still havent drink my morning coffee so i am a bit sleepy and i might not wrote everything enough clear.

  • Unique Sprites used with Bullet beh crashes the app after about 7-9sec of firing. Thats regardless of fire rate or speed.

    UPDATE:

    Actualy this happends without any beh at all too.

    It doesnt happend with rising numbers of the objects. Just with time.

    Funny - as long as they wont MOVE everything is fine O.o

    But make them fly and sh*t hits the fan.

    UPDATE:

    Okey it looks like the problem might be comming from when the Unique Sprite object is being destroyed. If too many of them are being destroyed at once... well. App crashes.

  • That's fine, I never had any problems doing that. Again, family variables should work fine if you don't delete or rename them.

    Ahhhh how nice.

    Thank you Arima :)

  • Hmm. I see.

    I am thinking about using Fam's for colision detections.

    Basicly If Bullet Overlaps Enemy then > something happends

    And as i have many enemies i though to group them with a Fam.

    So instead of the above it would be like:

    If Bullet overlaps Fam >then something

    No actual Fam's related variables. Just using the Fams to wrap many objects into one category.

    What do you think? in terms of safety.

  • hmm then how to get rid of the Vars if ive made an mistake? deleting the whole fammily?

    And what sort of problems you encountered? Can you name them?