Holy crap, you keyframed every single frame. LOL. I can't believe that's what you have to do to fix this. WTF.

Thanks though.

Here's the fla:
http://spamtheweb.com/ul/upload/090709/3 4960_anomaly.fla

Look inside the MC... play the frames. All the lines animate as they should. But when I ctrl+enter to test it out, it cuts the bottom line out for some reason... Is it doing this for anyone else? If so, do you know why? I've never run into this problem before. o.O

Yes, this collision bug is something I've experienced too. I don't know if it has been reported to Adobe yet though. It's been around for a while. (like a year) because I remember it with AS 2.0 even if I'm not mistaken. The fix would be to instead use hitTestPoint instead of hitTestObject(). You would need to hitTestPoint() every corner of the object. The important tweak to make it work though, is to + or - the x or y value of the corner that you are hitTesting. hitTestPoint() will be testing the same points as hitTestObject in your case, but you'll be able to shift the point that is being tested in any direction you need. You'll need to make the point being tested closer to the object to fix the problem.

I don't have time to open Flash right now, but what do the behind, normal, selected things do? Does one of them make it so that you only draw inside of a selection of lines? The problem I have always run into is like, if I've drawn a circle and filled it in, it's really hard to color the shading in around the edges because the paint always goes outside of the edge of the circle. I end up having to draw separate lines and just use the paint bucket filler and then delete the lines just so I can have a nice, clean section.

Can I ask what changes they think are negative ones? I hope AS 3.0 isn't one because I love 3.0.

What, did Macromedia advance Flash more than Adobe has or something? I'm in the dark here. :S

Any code in a swf has to be interpreted and processed by the actual Flash plugin the browser uses. The Flash plugin will not execute foreign code.

Create a variable that will serve as a flag and have the isDown function set it to true while you set it back to false with a keyUp function.

Come on, there has to be something. :\

For a very large game, you want the artwork to take up as little space as possible. There are bitmaps, which take up the most space, vector art which takes up less, and then procedural code through the drawing API to tell Flash what to draw which takes of the least amount of space. I have decided that for the magnitude of my game, bitmaps will take up way too much space and I think procedural code wuold slow the art process down too much as well, even with a map editor. So that leaves me with vector art, which won't take up too much data and is fast to draw. My game will use tiles because the world will be seemless and thus I will only be wanting to load in the section of the world that will appear on the screen, thus the vector art will need to be split into tiles. The only way I can think of to do this is to create a grid using lines that would split everything up and then select everything inside the grid one section at a time, but I have been doing this and it's taking a really long time since the artwork is so detail and there are so many things to select. (and even if I try to select everything and deselect what I don't need)

So is there any addon or plugin or program that will split vector art into tiles from within Flash? I tried Illustrator, but importing back into Flash is a disaster. All the artwork is messed up when I try.

There are also products out there that will obfuscate your code so that it is nearly unreadable and thus no use to anyone, but that's mainly for code protection rather than graphics protection.

Judging by your comments in this topic http://www.newgrounds.com/bbs/topic/1077 398 towards another user and the fact that you're spamming this thread in it trying to find people to answer your question, I don't think anyone is going to help you out any more. You are not a person worth helping.

Buddy, we are trying to help you, but you have to show a little bit of effort on your part or else I bet some people on here are going to think you're just wanting an easy fix or some code you can copy and paste. If you don't understand listeners, learn them. Try to problem-solve this.. your gotoAndPlay statement is executing continuously, so how do you get it to only execute once after you press the a key?

--"The character running animations plays when you press A but he runs on the spot"

--" the character moves but the animation just stays in one place."

ok aren't those two contradicting statements? First you're saying the animation plays yet the character stays in one spot, then a second later you're saying the character moves, but the animation stays still...

By the looks of your code though, I bet you it's that your character is moving on the x axis fine, but the walking animation of him is stuck. This would be because every time your key down code executes, it's telling your movieclip to go to and play at a certain frame. So the animation never actually gets to play because it's going to the first frame of animation each time. (since your key down code is executing like 30 times per second)

It would be like you trying to count from 1 to 10, but 30 times a second I say "start counting" which makes you start counting from 0, so you never even get to 1.

No, that is incorrect, at least considering the fact that you're using push() to add to the array. If you aren't understanding this by now, you need to run through your code line by line increasing i each time to really see what is happening because it doesn't look like you understand right now. I wish I could re-explain it a different way, but dELtaluca pretty much nailed it right on the head.

I'm not sure what exactly you mean when you ask about memorizing code. I mean, I know from memory the syntax of things like hitTestPoint(x,y,boolean); and addEventListener(event,function); if that's what you mean. But when I'm writing code, it's mostly about asking myself "hmm, what do I need to do to make this work? If statements? A for loop? Will I need flag variables?" One thing I am not good at is remembering the properties of certain objects. Things like Text Fields, for instance, will be tricky because even though they have a property like textColor to change the color, the majority of text-altering properties are contained within a style or text format class. In cases like that where I'm not sure what the properties are, I simply look up the class in the help files. But even there, you have to know which class you're looking for sometimes, which requires some memorization. If you were looking for the visible property of an object, you might think to look under the shape, sprite, or movie clip class, but it's actually under the Display Object class. You just have to know some things.

I have done different types of tile engines concerning how graphics are processed, but I have never done one that blits (I think that's the right word) bitmap data onto a bitmap object. I was wondering if anyone knew if this was an efficient method of doing an engine. The idea is that you would have one gigantic bitmap image, or vector image that is .cacheAsBitmap'd, a bitmap object, and you would calculate which pixels to write from the big bitmap to the bitmap object, which would only be the size of your screen of course. I was wondering if this is a good way to do an engine. Furthermore though, would rewriting all the data at once be the best way or would it be better to have a grid of bitmap objects to write to that also scroll so you wouldn't have to rewrite the entire screen region every frame? You would have to write to an entire row or column of tiles and would scroll them all in a direction.

At 6/27/09 07:05 PM, shadowgronds wrote:

At 6/27/09 06:56 PM, Woadraiders wrote: Are you kidding me?

AS3 is the latest scripting language by Adobe, and it is much better than actionscript 1 or 2.

So basicly as3 is action script 3.0?

yes

One thing to note: if you are going to do a rather large area with tiles, displaying them all at once will slow the fps a lot. I learned the hard way that switching the visibility of tiles to false does not solve this issue. You need to make sure you write your engine so that only the tiles that are supposed to be on screen are created and the ones that go off screen are deleted. But again, this is only when you have like 1000 50x50 pixel tiles for a level. If it's a small area, you might not have much of a problem. The smaller the area and the smaller the tiles, the more you will avoid this problem.

Enemy AI should just be based around if statements in my opinion. If the player is in range, get closer to the player and attack, else follow a path back and forth to patrol an area. The way you check to see if a player is in range is by using the distance formula to see how close the player and enemy are to each other. If in range, make the enemy's movieclip increase/decrease x/y coordinates to move closer based on whether the player is to the left or to the right or up or down from the enemy.

That will not work because it would then be impossible to go to a negative value on the z axis, which needs to be possible.

z=100 would be close to the screen
z=0 would be a little far away
z=-100 would be far away

or in x coordinate terms
if z=100 then x=200
if z=0 then x=30
if z=-100 then x=1

although those are only estimated x values.

Does anyone know the relationship between the z coordinate value of a point and the x value of that point that is displayed? I have written code that loads in a 3D mesh and displays the points based only on the x and y coordinates of the points of the cube. The result is a flat looking cube you're looking straight at. In order to work in the z coordinates of the points, I have to know how to recalculate the x and y points based on the z coordinates. Now I understand that the displayed point's x value will be closer to the point of origin the farther away from the screen it is and vice-versa. I understand that the increase or decrease in x value is also exponential. For instance, an point that is -100000 away on the z axis will be at something like .005 on the x axis and if it is -5 away on the z axis, it will have an x value of perhaps 30.

Using Math.tan() seems like the way to go because tangents increase and decrease exponentially based on the angle calculated. I came up with Math.tan((zcoordPoint)*.005)*100; which seems to be close to what it's supposed to be, but not quite. You would of course add 275 to the 100 to offset the point by half the stage width so that the origin appears to be in the middle of the screen. With that calculation, the higher the zcoordPoint, the higher the x value is, exponentially. The problem is that if the z coordinate goes below zero, the x values become negative rather than simply closer to 0. I'm not sure how to handle this calculation though. I thought about using cos() or sin() but I'm not sure how that calculation would be done or if it would even be better than using tan().

The best thing I can come up with is this:

stage.addEventListener(KeyboardEvent.KEY_DOWN, keydown);
stage.addEventListener(Event.ENTER_FRAME,EnterFrameHandler);

var counter = 0;
var counterSave;
var comboque=["","",""];

function keydown(event:KeyboardEvent)
{
	if(event.keyCode==87)//w
	{
		comboUpdate("w");
	}
	if(event.keyCode==65)//a
	{
		comboUpdate("a");
	}
	if(event.keyCode==68)//d
	{
		comboUpdate("d");
	}
}

function comboUpdate(letter)
{
		comboque[2]=comboque[1];
		comboque[1]=comboque[0];
		comboque[0]=letter;

	if(counter-counterSave<25) // 1 second to press each key for a combo
	{
		if(comboque[0]=="w" && comboque[1]=="a" && comboque[2]=="d")
		{
			//doAnAttack
		}
		counterSave=counter;
	}
}

function EnterFrameHandler(e:Event):void
{
	counter++;
}

might not be perfect or what yo'ure looking for, but that's the idea

As far as I can tell, that code looks correct. I will tell you though that no one uses scenes any more. It was a bad idea when they created it. Scenes are no longer used in the newest versions of Flash. Keep everything you create in one scene. There is no reason to use two scenes.

I recall reading an article that stated files coded in 3.0 will run 10 times faster than those coded in 2.0, but that of course is relative to what exactly you're trying to do.

The major differences I noticed when I migrated over to 3.0 were that referencing objects was more difficult, understanding how objects are created was a twist, that the enterFrame shabang was different, and that catching keypresses was also different.

Mouse clicks, key presses, and onEnterFrames are now all handled by what are called event listeners.
1. create an event listener: stage.addEventListener(Event.ENTER_FRAME ,EnterFrameFunction);
2. write the function:
function EnterFrameFunction(e:Event):void
{//docraphere
}
and that's it.

Here's one for a mouse click:
stage.addEventListener(MouseEvent.MOUSE_
UP, mouseFunction);

function mouseFunction(event:MouseEvent)
{//do crap here
}

as far as objects being created, you no longer do createEmptyMovieClip(). You will write code like this:
createBall();
function createBall()
{
var myBall:Sprite = new Sprite();
myBall.name="myBall";
this.addChild(myBall);
}

and as I was saying, referencing objects is more difficult in my opinion. If, in another function, you want to change the x position of myBall, you would go like this:
function changeX()
{
this.getChildByName("myBall").x+=5;
}

That's the jist of AS 3.0

Ok, I'm not fully understanding you I don't think. Are you trying to pass things between classes or pass things between objects that are of a class type? Like, I can create myClass { public var myNumber = new Number();} and then create an object of that class: var myObject:myClass = new myClass(); but are you trying to send a parameter actually INSIDE of the myClass definition so that the myClass constructor is modified based on that parameter? Because that can't be done, at least not to my knowledge. From what I know, class constructors will make properties of whatever object you create as that class the same as what the values are in the constructor and there's no way to change that. If you were wanting to modify properties of an object however, that is manageable through simple functions where parameters are passed. If that is the case, just look up tutorials on functions and passing parameters to them and using "return" to return a value from them.

In order for people to answer your question, you may need to explain in greater detail what exactly is not working about it. Does it throw an error? Does nothing happen? Does some of the code work, but a specific part of it doesn't? I noticed that it's doing something to a movieclip called "legs". Do you have this movie clip or did you just slap this code onto a random movieclip? Your friend would need to send you the appropriate movieclip of the legs in order for your code to work properly.

Hmph. Well, there doesn't seem to be much interest in this. I kind of thought I would have at least one or two people jumping at the chance to be able to work on an rpg with the programming all taken care of, but I guess not. What if I pitched a different idea, some other kind of game that required a large world? We could make a shooting game that kind of acts like an rpg. A shooter-rpg if that's possible. Or just a regular shooting game. I guess I'll try anything because either way, I need artwork from someone to have a game made.

var Zombie:MovieClip = this[enemyName + enemyNumber];

Since you only get that error after removing the object, it probably means you have some code elsewhere that is trying to do something to "back" but can't because at that point "back" doesn't exist any more. It could be that you did not remove the listener for "back" because it can't listen for mouseclicks on an object that doesn't exist any more. Check your code for anything else that is trying to do stuff for "back" though.

About tile engines and arrays though: when you have an array like
var myArray=[
[0,1,4],
[6,4,6],
[9,5,2]
]
the values indicate which tiles are to be used. One way is to have all your tiles inside a MovieClip on frames 0,1,2,3,4,5,etc. So when you are creating the tiles through AS, you do something like

for(y_v=0; y_v<myArray.length; y_v++)
{
for(x_v=0; x_v<myArray.length; x_v++)
{
var blankSprite:tileMC = new tileMC(); //tileMC would be a linkage class from library
blankSprite.name = "tile"+x_v+"_"+y_v;
blankSprite.x = 50*x_v; //50 is tile width/height
blankSprite.y = 50*y_v;
this.addChild(blankSprite);
blankSprite.gotoAndStop(myArray[y_v][x_v ]);
}
}

to lay out all the tiles and have each tile go to the correct frame of the movieclip by looking at what frame number is in each slot of the 2D array.