Slick Forums

Hi,

I'm looking for some assistance on a problem that I'm trying to solve. I've been looking at the code and the example (starwarrior game) and it appears that the spatials are where you setup the visual component for an entity, however they only contain one graphic or visual representation. How would you implement an animation system based on state? (i.e. Walking, turning, attacking, death, etc...). Should the visual representations and the state be contained within the spatial or components/systems? To make it more confusing, some enemies may have an attacking state and others might not, any ideas on how to implement this within the framework?

thanks,

is this forum still active?

In our case, we have an AnimationComponent with multiple Animations and each Animation is a collection of Sprite. Then, for each specific kind of entity (for example for the super mario character) we have one or more Scripts with the logic of which animation to use, for example, if the character starts to jump then we set the jumping animation, if the character is idle then we set the idle animation, etc.

Hope it could help as an idea, here is an old game I made for a ludum dare following this approach, search in the core module.

_________________

On your blog you post a possible solution for scripting. Wouldn't it be a security hole to pass world and entity into scripts? Might be more secure to create a interface api for adding entities to the world in a way that the host application can control what components an entity has and puts less control in scripts.

I do agree that Artemis is better to use for rendering then logic for more complex games, which you would use scripts for, otherwise you get SYSTEM OVERLOADING [PUN]

When I say System Overloading I mean so many systems doing so many things it creates a debugging nightmare when your sprites are doing funky stuff because they are being affected by one of the many systems you have running. This does not apply to everyone just some of us who prefer to be able to sleep at night.

Security hole ? We are not leaving personal user information available for hackers by using Scripts or anything like that (at least we don't have personal user information yet). On the other side, it is an optional thing, you can use all the systems and components you want to avoid using scripting. I agree, however, that the design and API could be improved.

_________________

Not the type of security hole i was talking about but that would be a very BIG security whole if you did.. I wouldn't play your game

But what I meant was API security. You might not want the end user to be able to Delete(npc) a special NPC therefore creating a chain of butterfly effects to your adventure altering the ending in ways you didn't imagine possible.

Letting the end user get access to the World or Entity instances opens up the entity.Delete() and World.DeleteEntity(entity) methods to the end user.

For now, the end user is not supposed to touch any Script since they are Java classes. If a hacker wants to decompile the code, modify a Script, good for him. If the game doesn't work anymore after that modification, it is not my problem.

I don't really understand the problem here. Maybe you are thinking about leaving the Scripts outside the code, for example, the Lua Scripts of Aquaria. Aquaria is a successful game and, as far as I know, nobody ever said anything about security holes for using Lua for scripting.

End users just play the game but please correct me if I am wrong. Maybe I am just misunderstanding your point.

_________________

I was just trying to help with problems that may arise. Maybe I don't get what you are using Scripts for. I was under the impression they would be scripts loaded from files.

Here is an example of a script:



Scripts are just Java classes, they are used because we can attach them to the Entity's life cycle in an easy way. For now, we are not using external files, nor different languages.

_________________

I'd recommend loading only one instance for each script, compile it, and reuse the instance for all entities sharing the same script.

... post deleted by author

Hi,

I have an issue that I was wondering if someone could help me with that I have no idea on how to solve. Every so often when I run my tests (this is on iOS, I converted Artemis to Objective-C). I have gotten it setup with a some systems in place (currently 5) which have been working good. However during testing iterations of my game (If I kill the game and restart) one of my systems fails to process. I have traced it back to this code here:

- (void) change:(Entity*) entity {
BOOL contains = (systemBit & entity.systemBits) == systemBit;
BOOL interest = (typeFlags & entity.typeBits) == typeFlags;

if (interest && !contains && typeFlags > 0) {
[actives add:entity];
[entity addSystemBit:systemBit];
[self added:entity];
}

what happens is that sometimes through the contains variable evaluates to true and so the following if statement doesn't execute and that system doesn't have any entities to process. However this appears to be occurring sporadically. When the issue occurred the systemBit value as 156848688 and the entitySystemBits (1, 2, 4, 8, 16) respectively. Both systemBit and entity.systemBit are defined as longs. If anybody could provide some guidance that would be great.

thanks,

Well, I don't think there's enough traffic for a dedicated forum. I think 1 thread is sufficient, and Google has already decided this is it
I did put up a forum, you link to, but as I said, it's pretty dead.

Not sure when you started porting, there are some fixes in the svn (http://gamadu.com/svn/artemis), you can check the logs and see if anything is relevant.

It may be a porting issue though, you're probably not resetting the entity.systemBits properly.