Game Development

7:01 PM

I'm not sure there is any way to academically bridge the gap between architecture and the programming language.

user4704

Perhaps. I've not given it much thought.

user4704

Thing is, it's not necessary or even always useful to know how the processor works in a lot of detail to be a good programmer.

user4704

Certainly there are concepts, such as cache coherency and whatnot, that you should understand.

user4704

But it's not like you want to be writing specifically to a specific chip these days anyways, in most cases.

melak47

and "a little knowledge is a dangerous thing" probably applies, too. You don't want to be making assumptions with processor archirecture y because you've worked with x :p

John McDonald

7:14 PM

@TreDubZedd yeah, when I went to school, we started with a real language, then took a circuits class, and then an assembly class

Jimmy

@TreDubZedd "enough" is always a dangerous word

TreDubZedd

It's not dangerous, in itself...it's just ambiguous.

And it's the ambiguity that is dangerous :)

Jimmy

LISP is close enough to machine code, for appropriate values of "enough"

melak47

we did functional programming and assembly as the first two languages at uni

Gajet

@JoshPetrie agreed no one can truly get an in depth understanding of how the computer works with learning C++

Jimmy

7:17 PM

well, it's easier to illustrate concepts with C++ than with the other languages you'd teach to first- or 2nd-year CS students (Python or Java)

not so much how the hardware works, but how the software at the lowest level works

melak47

my uni seems to have a different idea of a cs course :)

Jimmy

I never learned assembly but in our systems class we were still expected to be able to disassemble programs and figure out how to exploit buffer overflows

Gajet

@melak47 we started with Java and then went to C++ and then Assembly in uni

of course I prefer python (or C#) as the first language to learn.

never liked Java at all

John McDonald

@Gajet may I as why?

melak47

opal was the first language we learned, a homebrew functional language

Gajet

7:21 PM

@JohnMcDonald I'm not sure why actually,

John McDonald

Good reasoning

Gajet

but I think the biggest problem I have is that operators are not overload-able.

Jimmy

we start with either Java (normal intro) or Scheme (advanced intro, SICP)

melak47

opal and assembly first semester, java and C 2nd, more java and C in the third semester, yet more java and C in the 4th :)

John McDonald

@Gajet You know (one of) the reasons they don't allow that, right? To prevent programmer error

melak47

7:23 PM

also unsigned types. nobody knows what they do anyway :p

tylerrrr07

I went from VB, to C++, to C++, to Java, to C#, and then a little web thrown in here and there (ASP.NET, JSP, HTML, CSS, JS)

ITT focused a lot more on getting you skills to get you jobs versus understanding how it works under the hood (compilers, assembly, etc.)

Jimmy

@melak47 unsigned types are less confusing than signed bytes

tylerrrr07

Though it could be argued knowing that stuff is still quite valuable in a job - my job is high level enough I can live without the info

Jimmy

I don't trust programmers who have no experience in things lower level than Java

melak47

if I go way back, my history looks like this: php, then delphi, prolog, java, assembly, sql, opal, assembly, C, java, assembly, C, and I don't remember when I started using c++

Jimmy

7:28 PM

it's like a driver who doesn't know what Antilock brakes are for. They can drive normally just fine, but you can't rely on them when things go wrong.

TreDubZedd

You can't be a (good) programmer and be scared of low-level languages!

tylerrrr07

I know C++, I just don't know anything about it... if that makes sense

TreDubZedd

Mainly I was just trying to cash-in on the same kind of rhetoric that got one of my other comments starred.

tylerrrr07

I noticed :P

Jimmy

hmmm is that a guaranteed star?

You can't be a (good) programmer and be scared of snakes!

3

testing.

tylerrrr07

7:38 PM

You can't be a (good) programmer and be scared of not getting stars!

TreDubZedd

I've created a monster.

Damir

You can't be a good programmer and a bad programmer

@TreDubZedd I blame you :P

TreDubZedd

@Damir I've had worse.

tylerrrr07

Well jebus. They just added Braid, Super Meat Boy, and Lone Survivor to the humble indie bundle.

$8 average right now

Jimmy

yeah this is the Justice League of Indie Bundles

John McDonald

7:49 PM

wow, that's awesome

tylerrrr07

8 great games, 5 soundtracks, $8 (and that's if you want to go as cheap as possible!)

John McDonald

lol, @notch: $9,999.99

I guess he can afford it

Probably like 1hr's worth of work

Noctrine

Just a shame I have all but one.

tylerrrr07

Which one?

Noctrine

Lone Survivor

John McDonald

7:54 PM

I only had 3, but that's because I didn't really think I'd enjoy playing some of the others

tylerrrr07

I think I'm going to buy it for the music at least

Chorche

8:15 PM

how can i multiply this lwjgl.org/javadoc/org/lwjgl/util/vector/Matrix4f.html with a float?

Jimmy

what do you mean, multiply it by a float?

what kind of matrix is it?

Damir

Two Guys From Andromeda are back together and are making a new game!

kickstarter.com/projects/spaceventure/…

Hopefully there are people here who remember Space Quest :P

Jimmy

they have 300k+? I'm always surprised at the amount of "latent" money floating around in gamerland

tylerrrr07

8:35 PM

Are users really afraid of vertical scrolling on web sites?

John McDonald

yes

Jimmy

I'm more afraid of horizontal scrolling

tylerrrr07

If you don't want vertical scrolling don't design me a form that only fits on an HD monitor

Gajet

9:04 PM

@Jimmy me2

w8 vertical is the same scroll bar right of the windows right?

Noctrine

Titanium: All the fun of CSS, none of the documentation.

tylerrrr07

@Gajet - yes. Like to read old chat messages here you have to (vertically) scroll up

Chorche

Matrix multiplication

In mathematics, matrix multiplication is a binary operation that takes a pair of matrices, and produces another matrix. This term may refer to a number of different ways to multiply matrices, but most commonly refers to the matrix product. This article will use the following notational conventions. Matrices are represented by capital letters in bold, vectors in lowercase bold, and entries of vectors and matrices are italic (since they are scalars). Scalar multiplication The simplest form of multiplication associated with matrices is scalar multiplication. General definition ;Left...

Jimmy

@Chorche I know what matrix multiplication is

I'm asking why you want to scale by a float

Chorche

1

Q: Importing 3D skeletal animations from .smd model

I'm writing an .smd importer, and i'm stuck at the skeletal animation part. The problem is that i dont know exactly how it works. I'm using this this to write the exporter, but it doesn't show how to use the information stored in file. I imagine that all vertexes with the same bone id should be ...

vertexPosAnimated =
boneToWorld[0] * boneAtRest[0] * vertexPosModel * weight[0] +
boneToWorld[1] * boneAtRest[1] * vertexPosModel * weight[1] +
boneToWorld[2] * boneAtRest[2] * vertexPosModel * weight[2];

boneToWorld, boneAtRest and vertexPosModel are matrixes

and weight is a float

Jimmy

9:14 PM

what is vertexPosAnimated

vertexPosModel should be a vector, not a matrix

if it is a matrix, then weight[0] is a vector

TreDubZedd

@Chorche We're back to the same problem you had yesterday (and a few other times, here in chat). You need to provide context and question--not just one or the other.

(or neither)

So...what is it you're trying to do?

Chorche

@TreDubZedd i asked "how can i multiply this lwjgl.org/javadoc/org/lwjgl/… with a float?"

TreDubZedd

Yes...but why?

Jimmy

@Chorche I'm 80% certain that's the wrong question

TreDubZedd

What is it you're trying to do that requires you to multiply (scale) a Java matrix by a float?

Jimmy

9:17 PM

You can multiply a matrix by a float, in the mathematical sense, but it makes no sense for 3D math

Chorche

because i want to set mutiple bone for a vertex

so you want to say that the anwser is wrong?

Jimmy

no, I'm saying you dont' understand that formula you put there

one of those terms has to be a vector

Chorche

no!

TreDubZedd

No, the question is wrong. You should be asking how to scale a matrix by a given weight...

Jimmy

okay

TreDubZedd

9:20 PM

A float, in and of itself, doesn't mean anything. There's no context.

Chorche

boneToWorld[0] * boneAtRest[0] * vertexPosModel - matrixes

weight[0] - float

user4704

That does not make sense.

user4704

That means the result of that entire computation is a matrix.

user4704

Which is wrong.

user4704

Unless I misunderstand what you are trying to do here.

TreDubZedd

9:21 PM

No, @JoshPetrie, That's what he's trying to get.

user4704

vertexPosModel should be a vector.

TreDubZedd

Basically, you're trying to find a weighted sum of matrices, in order to apply that transformation to a vertex.

The variable names, in my opinion, do not correlate well with what's going on.

Jimmy

hmmm, okay I understand

he wants an interpolated transformation

I'm not sure this is how you'd go about it though

if you multiply a transformation matrix by a number, you screw up the W-values

TreDubZedd

Assuming your weights sum to 1, don't the resulting matrix's W-values work?

Jimmy

I'm not sure the relationships work out

since your bottom row values are implicitly multiplied by w-value

specifically the last entry might be an issue

I just did some napkin math, and you might be able to get away with Matrix4.Scale(weight[i]) * boneToWorld[i] * boneAtRest[i] * vertexPosModel

TreDubZedd

9:31 PM

Out of curiosity (since I've not done much 3D), what transformations could cause the 4th row to be anything but 0,0,0,1?

Jimmy

good question. I'm not sure

Chorche

@JoshPetrie is there any difference between translating boneToWorld[0] * boneAtRest[0] by a vector, or translating an identity(vertexPosModel) matrix by a vector and then doing this boneToWorld[0] * boneAtRest[0] * vertexPosModel

user4704

"identity(vertexPosModel)" makes no sense to me.

user4704

You don't "translate" matrices.

Chorche

@JoshPetrie i know

user4704

9:34 PM

Then why did you ask a question that makes no sense?

Jimmy

@TreDubZedd oh wait, Translations aren't affine, so any translation would be X,Y,Z,1 in the 4th row

Chorche

the method is called "translate"

user4704

What method?

Chorche

that's way i say that word all the timw

lwjgl.org/javadoc/org/lwjgl/util/vector/…

user4704

I don't know what that method does. Matrices do not (mathematically) have a "translate" operator so I have no context for what you're trying to do.

Jimmy

9:35 PM

he means interpret vertexPosModel as either a vector or a translation matrix from that vector

user4704

I think you need to brush up on your linear algebra. :|

Chorche

@JoshPetrie i sayd that i know that

the method is called translate

that's why i always say translate

user4704

I recommend this book: amazon.com/The-Geometry-Toolbox-Graphics-Modeling/dp/1568810741

user4704

Well stop saying it, at least in general.

Chorche

ok

user4704

9:36 PM

If the rest of these fellows know what you mean, you should probably continue to talk to them because they understand more about what you've been doing that I do.

Jimmy

I ignored the previous conversations

but as I said above, I'd try Matrix4.Scale(weight[i]) * boneToWorld[i] * boneAtRest[i] * vertexPosModel

Chorche

"identity(vertexPosModel)" by that i wanted to say, that the identity matrix is called vertexPosModel. It's just a varname

Jimmy

vertexPosModel is always an identity matrix?

Chorche

Matrix4f vertexPosModel = new Matrix4f();
vertexPosModel.translate(vertex);

TreDubZedd

That doesn't make sense. The multiplication boneToWorld[0] * boneAtRest[0] being the identity matrix does make sense.

Chorche

9:39 PM

vertex is a Vector3f

TreDubZedd

@Chorche It looks to me like vertextPosModel should be a Vector3f.

Jimmy

yeah that's what I suspected

TreDubZedd

That is, vertextPosModel == vertex

Chorche

Matrix4f.translate(vertex, vertexPosAnimated, vertexPosAnimated);

Jimmy

it smells like wrong, but he's converting all calculations into Matrix mode

including the original position

Chorche

9:42 PM

Matrix4f.mul(bone.boneToWorldTransforms.get(2), bone.atRestTransform, vertexPosAnimated);
Matrix4f.translate(vertex, vertexPosAnimated, vertexPosAnimated);

is that ok?

TreDubZedd

ok, @Chorche, you're just spewing code without context, again.

what is bone.boneToWorldTransforms.get(2)?

Chorche

it is a boneToWorldTransform that is stored in arrayList

TreDubZedd

Ah-HA! I think I see your problem.

Your output vertexPosAnimated needs to be a Vector3f.

Therefore, your input vertexPosModel also needs to be a Vector3f.

So, what you're doing is getting a weighted sum of vertex positions

(or, rather, what the StackOverflow answer is indicating you should do)

Jimmy

well, that's what I said, (and Josh said) earlier

TreDubZedd

Then I missed that, and I apologize.

Jimmy

9:46 PM

but Chorche is obviously trying to make everything a Matrix4 for some reason

TreDubZedd

@Chorche Stop doing that.

Jimmy

26 mins ago, by Josh Petrie

vertexPosModel should be a vector.

33 mins ago, by Jimmy

vertexPosModel should be a vector, not a matrix

TreDubZedd

Look at it this way: You have a transformation associated with each bone, which you are applying to a vertex. Once you have a transformed vertex position associated with each bone, you want to take the (weighted) average of all those positions.

Chorche

so i have to:
Matrix4f.mul(bone.boneToWorldTransforms.get(2), bone.atRestTransform, vertexPosAnimated);
then turn vertexPosAnimated into a Vector3f, and then
Matrix4f.translate(vertex, vertexPosAnimated, vertexPosAnimated);

?

Jimmy

in short, no.

Chorche

9:53 PM

why?

Jimmy

because vertexPosAnimated should not ever be a Matrix

Matrix4f.mul(bone.boneToWorldTransforms.get(2), bone.atRestTransform, vertexPosAnimated);

seems to imply that it is

Chorche

how can i do this Matrix4f.mul(bone.boneToWorldTransforms.get(2), bone.atRestTransform, vertexPosAnimated); if the vertexPosAnimated is a vector3f?

Jimmy

you don't use Matrix4f.mul

or rather, there's so many bad things going on in that code, I'm not even sure where to begin

starting from the fact that you initially put the data for vertex into vertexPosAnimated and you immediately overwrite it with that Matrix4f.mul call

Chorche

@Jimmy i dont

Jimmy

oh right, because of that second line

which does not do what you think it should be doing

Chorche

10:00 PM

i do the Matrix4f.mul and then i do the translation call

Jimmy

yes, the TRANSLATION CALL makes no sense

why would you translate a transformation matrix by the position of the current vertex

Chorche

just explain how should i do the
vertexPosAnimated =
boneToWorld[0] * boneAtRest[0] * vertexPosModel * weight[0] +
boneToWorld[1] * boneAtRest[1] * vertexPosModel * weight[1] +
boneToWorld[2] * boneAtRest[2] * vertexPosModel * weight[2]; part

because now you're just confusing me

Jimmy

okay, lets play detective.

1) boneToWorld[0] is a matrix

2) boneAtRest[0] is a matrix

3) boneToWorld[0] * boneAtRest[0] is also a matrix

Chorche

yes

Jimmy

4) vertexPosModel is a vector. Lets make it a vector4 because that's easy to multiply with a matrix4

5) boneToWorld[0] * boneAtRest[0] * vertexPosModel is another Vector4, since VertexPosModel is a Vector4

6) weight[0] is a float

7) the results of [5], scaled by a float, remains a Vector4

we've just established that vertexPosAnimated is a vector4. if you ignore the W, it is a vector3.

TreDubZedd

10:05 PM

Note that, if W is 1, you can ignore W safely.

Chorche

daah :D

TreDubZedd

(alternatively, converting a Vector3 to a Vector4 is done by making the W term 1)

(x, y, z) == (x, y, z, 1)

Jimmy

I'd guess W should be 1, or at least we should probalby be ignoreing it, since we're ignoring the camera transform matrix so we're not doing any perspective divides.

Chorche

lwjgl.org/javadoc/org/lwjgl/util/vector/Matrix4f.html witch method should i use to multiply a matrix with a vextor4f?

Jimmy

transform

Chorche

10:11 PM

ty

TreDubZedd

Vector4f convertedFromV3 = new Vector4f(v3.getX(), v3.getY(), v3.getZ(), 1);
Vector3f convertedFromV4 = new Vector3f(v4.getX() / v4.getW(), v4.getY() / v4.getW(), v4.getZ() / v4.getW());

Jimmy

this is one of those cases where lack of overloaded operators stinks

Chorche

mul.transform(mul, vecPosMod, vecPosMod);

can i do this?

Gajet

@Jimmy @TreDubZedd seems the pattern is working, You can't be a (good) programmer be scared of *

Chorche

(store the result in the same vector that i was multiplying?)

Jimmy

10:14 PM

don't know, I've never used LWJGL before

user4704

You can't be a (good) programmer and not be me.

user4704

:D (worth a shot right?)

Jimmy

lies! although I'd support a weaker form of that statement and insist that your name begin with "J"

TreDubZedd

Hm...I don't have a 'J' in any of my names or aliases...I guess I'm out of luck.

Jimmy

you must be a terrible programmer....

TreDubZedd

10:15 PM

Depends on who you talk to.

Jimmy

maybe some people misspell your name with a J

TreDubZedd

Foreign Nationals who can't pronounce "Zedd"? Maybe.

I could go by "Jedd" in France, e.g.

Gajet

@JoshPetrie there is only one big problem with that statement

it doesn't have the "scared" part

You can't be a (good) programmer and be scared of me!

that would be a better shot

TreDubZedd

You can't be a (good) programmer and be scared of being me!

user4704

You should be scared of me, I am mean.

user4704

10:18 PM

I make people cry.

TreDubZedd

Only when you tell them they can't play GW2 Beta.

Gajet

@JoshPetrie then I won't send my resume to ArenaNet.

I guess I need some advise here,

user4704

@TreDubZedd All you gotta do is buy the game ;)

Gajet

I think I need to make some changes into my project, since the whole structure now is kinda of a mess.

but the problem is I can't change the classes one by one

everything there depend on each other.

user4704

That is a problem, and you should fix it. This is why encapsulation and isolation are important.

Jimmy

10:22 PM

"everything depending on each other" is usually a sign of bad design

Gajet

the problem is if I want to fix a class, I need to fix 3 or 4 classes at a same time, which will result in changing more than 1000 lines of code

Jimmy

have you run static analysis over your code?

user4704

But I do find it hard to believe that it's actually everything that depends on everything else.

Gajet

let me see

user4704

So you can probably untangle a little bit that you can fix in isolation, and build from there.

Jimmy

10:23 PM

it can give you metrics on classes that depend on too many other classes

Gajet

it's a little bit more complecated...

and the main problem is I don't know where to start with isolation

user4704

Ideally you start when you are initially designing things, but that ship sailed.

user4704

What is an example of something you're trying to change -- a small thing -- and what prevents you from doing so?

Gajet

when I started that thing I didn't think it was going to be big at all.

and little by little I've added new features over each other.

which now resulted into a big pile of mess.

@JoshPetrie I've got two classes representing player.

one is doing everything regarding it's logic, and the other one is an information holder

John McDonald

looks like you need to spend some time refactoring

Gajet

10:27 PM

and there is a third classes which controls those two and also draws players

Jimmy

that's not bad so far

Gajet

it doesn't look bad at all.

but the problem is everything is controlled by hand in there.

there is almost no isolation and I can easily break the whole code with one wrong statement.

user4704

"Information holder?"

Jimmy

it's hard to visualize

user4704

Doesn't sound like you need three things for this.

user4704

10:29 PM

I'm going to a meeting, I'll be back later.

Gajet

I've got two or three duplicates for everything, which I need to keep in sync and so on.

Jimmy

I'm going to a party, will not be back later

Gajet

@JoshPetrie information like position or current hit point and so on. as I you suggest having 3 classes is wrong. But I can't really clear my mind to find out what I want.

and the main problem is I can't really separate access to class members, I think my whole code flow is just a mess :(

Gajet

10:44 PM

@JoshPetrie is it a good practice to separate a class into two classes, for the sole purpose of isolation? In result everyone will get a pointer to parent class (which is holding data) and I put every other member (which should be isolated) into another which inherit that parent?

CaptainRedmuff

11:01 PM

vertex winding ordering is confusing 0.o

calculating normals is also confusing

damn you OpenGL! I give up for the evening.

perhaps sleeping on it will help things sink in ;)

waves

user4704

11:20 PM

@Gajet If you split a class into two classes, but still keep them coupled, you really just created a more complex representation of the same thing. Generally.

user4704

Why not just have a single class of type Player that contains the player's logic data and methods to operate on that logic data?

Gajet

they will not be coupled

one will inherit the other, and have more functionalities.

@JoshPetrie cause they were developed in different times, with different approaches at their time

and that's why it's a little bit hard to handle everything right now.

actually at the time, I was thinking about an AI contest and to prevent cheating I split those classes

user4704

@Gajet That's coupling.

user4704

Why do you need inheritance, why not just modify the original?

user4704

Inheriting is extension and so appending to the base should be fine.

user4704

11:24 PM

I fail to see how this design helps prevent cheating.

Gajet

@JoshPetrie they have the information through some const instances which they can not change

and I just ask their classes to think and report which action should they make

so they can't directly make any change in data section, and I can validate everything before applying .

user4704

const is a language construct, it doesn't prevent cheating

Gajet

I know that,

user4704

cv qualification (and access modifiers) are not security; if you have the code they can be circumvented and they have no impact in the compiled binary.

user4704

So I still don't see how this prevents any kind of cheating.

Gajet

11:31 PM

but still it's a little bit safer, I can create copies and send them

but without separating those classes creating a copy would not be possible.

I don't know what information they need to create a deep copy of an AI class

user4704

It's not any safer, and I don't see how these classes enable creating copies.

user4704

(in a way that can't be done in one class)

Gajet

let me write a psudo code

user4704

I think you should post examples of what you mean, your explanation isn't conveying anything

Gajet

while(true)
{
	World.update();
	for(int i=0;i<n;i++)
	{
		Request r = player[i].think(World);
		World.ApplyRequest(i,r);
	}
}

this is how the function looks like now.

which is not safe as you say.

but changing "Request r = player[i].think(World);" into "Request r = player[i].think(World.copy());"

will result in safe calls.

the AIs can't change anything from actual data since there is no way they can get where data is stored.

and later I can delete that copy part to increase performance

and whole information about each playing in this case is stored in World class.

now if you are still not agreed with me, there should be something wrong with how I think.

user4704

11:41 PM

...

user4704

Okay, so, there is no reason you need two classes just to give one class a "copy" method. That sample does nothing to demonstrate a need for two classes.

user4704

Second, the copy is wasteful, just pass it by const reference.

user4704

Yes, "think" code can const-cast it away and do "bad things"

user4704

But such code is already in your address space and could already do bad things, so you're just wasting time for no reason.

Gajet

you can do that with only one class?

@JoshPetrie just being in address space doesn't guaranty it can do bad things

user4704

11:44 PM

Absolutely it does.

user4704

And yes, you can trivially add a copy method to one class.

user4704

class A { int x; public: A copy () const; };

Gajet

he needs to know where the instances are and I'm not providing it with any pointer.

user4704

Nope.

user4704

The code could maliciously stomp random memory.

user4704

11:46 PM

Run in an infinite loop.

user4704

It could find the original instance by searching memory and modify it.

Gajet

not that it can't

but that's completely out of my audience league.

user4704

You've successfully made everybody's life harder in order to win the sad, lonely little prize of making the life of the one guy who wants to write code in your system that does "bad things" just a teeny tiny bit harder

user4704

It's a terrible plan.

Gajet

(or at least I think so)

user4704

11:48 PM

But regardless that sample does not demonstrate any need for two classes.

Gajet

it's just like installing anti-theft alarms on your car,

user4704

No, it isn't.

Gajet

they can still steal it but it will be a little more tricky, and less people will bother trying to

user4704

int * p = 0; *p = 3;

user4704

DONE.

user4704

11:49 PM

My think() crashed your game. Your protection was trivial.

Gajet

that's not what I'm trying to protect my self

user4704

Anything you can be trying to protect yourself against is less bad than a crash.

user4704

So you're still wasting your time.

Gajet

well, I'm not trying to protect against any kind of crash.

user4704

The copy() function is entirely useless, and there is no reason you need two "player" classes (one a subclass of the other) to do so.

user4704

11:51 PM

So there you go. Refactor that mess.

Gajet

I just want to prevent data manipulation without me being noticed

@JoshPetrie actually I was going to create two world classes

user4704

Whatever, the point still stands.

user4704

You don't need it; it's overly complex.

Gajet

I've already have multiple player classes (for AI, Human, net)

user4704

Which isn't neccessarily good, nor should it be justification for doing it elsewhere where it is known to be unneccessary

Gajet

11:54 PM

the whole point is to later write same code for a generic player and use overriding.

user4704

Uh, what?

user4704

So do that later.

user4704

Don't introduce complexity now that you think you might need later. It just makes things harder.

user4704

Also, I seriously question the need for a distinct local versus network player. AI, even.

user4704

It implies that your player is not just data and methods about a player, but also the things that control the player's actions.

user4704

11:56 PM

Which create multiple responsibilities and thus increase maintainability cost, and increased likelyhood of coupling, which also increases maintainability.

Gajet

I told you players are doing everything about the logic.

for information I've got only one class

and it only contains information.

wait what am I saying, I'm not here to defend my structure.

as I said it's just the fact that I've added features one by one into the code and now I'm stuck with a big mess.

and I'm wondered how can I clean everything up,

Transcript for