Microbe Stage Development

Right, so it's been two weeks since I posted here... again. I'm not going to be able to contribute to this continually, but I'll hopefully get either a day or a long evening in every week or two, and try and get a few significant features done each time.

I had started work on both the input manager and entity system, though Dani has since also done a lot of work implementing the entity system, so I'll need to have a good look at that before I carry on.

I can't say when I'll next get something done, but I'm thinking of stuff to do almost continuously, and I'd really like to get atleast some of it into code soonish...

@scio - don't worry about commenting for now, the work we're doing is setting up the nitty gritty basics, and is only really of interest to coders. I expect this thread to become a lot more interesting to others once we start implementing actual gameplay. We always appreciate the support though.

I've been using the entity framework a bit, adding some things. In particular, colisions. Althrough this is the very first implementation of it, I think it looks fine.

What it does is push objects away from each other if their distance is closer than their combined radius. It doesen't care about how heavy they are, only works for circular shapes, and will punch objects really far if they manage to get really close, but as i said this is a first test. Anyway most of the times when cells colide they will probably engulf one another, instead of bounce away.

You can download it from here

On a side note:
I've been thinking about how to implement species, but i can't come up with a good idea. I would like for it to go inside the engine, and I have some ideas, but nothing that fits our plans completelly.

Daniferrito wrote:

I've been using the entity framework a bit, adding some things. In particular, colisions. Althrough this is the very first implementation of it, I think it looks fine.

What it does is push objects away from each other if their distance is closer than their combined radius. It doesen't care about how heavy they are, only works for circular shapes, and will punch objects really far if they manage to get really close, but as i said this is a first test. Anyway most of the times when cells colide they will probably engulf one another, instead of bounce away.

You can download it from here

On a side note:
I've been thinking about how to implement species, but i can't come up with a good idea. I would like for it to go inside the engine, and I have some ideas, but nothing that fits our plans completelly.

Good Job

So, could the next step be to put some of the protocell models into the game somehow? Maybe we could have an orange centipede (heat eater) and a lily pad (light eater) to brighten up the mood. Has anyone made the little bug (eater)?

MitochondriaBox wrote:

So, could the next step be to put some of the protocell models into the game somehow? Maybe we could have an orange centipede (heat eater) and a lily pad (light eater) to brighten up the mood. Has anyone made the little bug (eater)?

The nest step would be implementing species. Separate models for species come later.

so a few updates on my end:

In windows I am at the point that my error is now an ld error:
ld can't find OGre_Main and OIS
I copied the OPHreMain.dll and OIS.dll into the project's directory
and then liberally pasted it around hopging that ld searches the project's dic for libraries, but to no avail.

In linux I am getting a similar ld error, but it's just a can't find boost_threat_mt_some such version so i just need
to find the boost lib thread in my usr/lib and then create a symbolic between the two, and hopefully that will work.

In other news I am not working 7 days a week right now (looks like I resume work in March and then possibly have April off) so I will be working on these problems
in the coming week.

did not know that target_link_libraries calls on ld, might have windows issue solved.

nope.... but did fix one more error in windows.

Have you tried the instalation guide that i posted here? It's an easy way of setting up the environement on windows and one person alredytold me that it worked for him.

Daniferrito wrote:

Have you tried the instalation guide that i posted here? It's an easy way of setting up the environement on windows and one person alredytold me that it worked for him.

O the installation is fine
The problem it's with compiling the tutorials. I did use your guide and it went fine.

But are you able to compile the code on the repo? The tutorial code is probably outdated.

Daniferrito wrote:

On a side note:
I've been thinking about how to implement species, but i can't come up with a good idea. I would like for it to go inside the engine, and I have some ideas, but nothing that fits our plans completelly.

The way I did the speices for my Ai test system worked fine. I t worked by haveing in a world class an array of actors
Organisms org[10000];
and an array of speices
Species spec[100];
Each actor would be assined a speices with a little id tag and the speices would be a class with all the info like modelshapes speed AI net weights and such.
so
org[0] would have a tag of
org[0]->SpeciesId and then it would look for it's speices in the speices array.

This is they way I would do it.

Ps the dropbox for the sim is here
http://dl.dropbox.com/u/68001178/Thrive.zip

Yes, with that type of structure that would be re route i would take too. However, we are using an entity framework, and so we just have entities. Basically, it's a different problem. Look at the code on the repo (my branch) to see how it works.

I saw your code some time ago, i tried to compile it, but i couldn't. The program looks good. But why the cells go to the top left when left alone?

The reason why is that the search code is of and so when a cell is not tracking something the default place to go is to 0,0 or -1,-1. Also I think when you tried to compile it there may of been errors.

I have looked at your code and the code is very different but i think that it might work but a bit metarmofisized/changed?

Ps: I do like the coding style
On another note couldn't the ogre and the game play layer be separate, so that you can read what is game play oriented and what is ogre based. I think it would read easier ? This would help people who are less experienced at ogre-ing than gameplay could work on that and those who are more proficient at ogreing can work on that.

If you mean the code at the thrive class, that will need to be splitted into other places.

Actually the entity framework is there so all code can be splitted into more manageable parts, called systems. Instead of having a huge updating method, with everything from the game stuffed into it, we have a system for every single separated part. For example, one system to move things around. One system to get the actions from the agents. One system to handle colisions. One system to evolve the species. One system to create the ogre models from the species attributes.

The code will hopefully be somewhat separated, but there needs to be some parts where they meet. At least with the basic things from ogre. Any system that interacts in any way with the world will want to know where stuff is, and that is stored into an ogre node. Speed, direction, distances, all of that will be handled by ogre.

Sorry what I mean is the interaction between the entity and the speices class, the thrive was just a little way to cheack if the AI was working. I was thinking that we could have a similar interaction.
Sorry for the confusion

Oh, yes, cells/creatures have to have either a pointer to the its species or som other way to point to them, like their name. The dilema i was having was where to store the species, either in a vector on its own or as entities, together with all the other entities. I spoke with seregon and he prefers to have them together with the other entities.

Can you guys give us an update on the current status of the code, and what the Task list should look like? Also, we are getting some new programmers, like TropicalMammoth and Tarpy, so try and contact them and explain where we are right now and where we are going.

I just added species to the game. Each specie has a set of attributes, like the organelles it has, the model to display and some other, like mass or acceleration.
From the two cells that there are, each one is from a diferent specie, and so each one has a different model. There are abolutelly no models on the right format (.mesh), so i had to use one from the tutorial resources. If someone posts any cell model, i'll change it.

I've added organelles as well. They are defined from code, not from xml. Maybe someone could do it?

In the end, the only visible change is that one of the cells has a different model.

Download: https://www.dropbox.com/s/qlhcv05kmya679a/dist.rar

I'll add this to the releases thread

Alright, I want to start being helpful with this project, what type of file will models be using in the game, and which program should I use to make/edit them?

I'm fairly sure the models are .mesh files. You can make them using a 3D Modeling program named Blender.

This thread and subforum is for discussing programming. Make sure to ask questions on art on an art thread, or even just the Misc. thread.

Just wanted to add a bit of an update on the current progress since not everyone can be expected to look at (or even find) the github issues page.

Which i will add, is the best way for people to get the absolute latest news of whats currently being developed!


We are currently working hard at creating a decent microbe stage prototype for christmas, and I'll add a few points at what you can expect and can't expect with it:

You can expect:

• A microbe to move around in a randomly generated enviroment - Done
  
• Functionality to move around the enviroment, gather different compounds and process them into other compounds, this includes a system for procucing and basic consumption of ATP (energy) - Done
  
• Limited functionality to custimize your microbe and enviroment through lua scripts - Done
  
• Other microbes moving around in your enviroment with basic AI - Development not started
  
• Very basic UI, showing what compounds your cell contains.
  

What you very likely shouldn't expect:

• Visually appealing or correct graphics. You will see re-used models and a microbes that looks like it belongs in the microbe editor, etc.
  
• Gameplay direction and goals. You won't be able to complete the game or reach the next stage.
  
• Correct representation of microbes. They won't have all the organelles/parts you would expect a microbe to have.
  

What we might be able to squeeze in:

• Advanced microbe interactions, like attacking organelles used to kill other microbes.
  
• Slightly more informative/fancy UI
  
• Menu/startup screen - We should have functionality to make this without much trouble, but low priority
  
• Simple microbe editor - This will be a tough one to make in time, but one i think would be really cool to have for the prototype
  

Note that these lists have been made by me alone and I'm not the sole decision maker, so they are naturally subject to change.

General Development status:

Development is slowly moving towards actual gameplay development as opposed to mostly engine development. But we are still making changes to the engine itself, and there will definitely be more demands on the engine, as new features start popping into the todo-list.

Development is currently being done by Nimbal and myself with some recent help from patowen95/SuperLala. I have let myself be told that Daniferrito might be returning to active development within a week or two.


Feel free to ask questions about the prototype or anything else relevant!

Fantastic! This is bound to get us at least some attention. Once people start playing the game, no matter how simple it is, important that it is actually playable, they will sort of realize that we are capable of making progress and will look at us with less ridicule .

Tarpy wrote:

they will sort of realize that we are capable of making progress and will look at us with less ridicule .

Don't expect too many positive reactions. Today's players are pretty hard to impress, what with the countless indie projects that pop out of nowhere with fancy graphics (which we won't show) and fresh gameplay ideas (which we won't show). The planned milestone is little more than a proof of concept that the engine is more or less capable of keeping graphics, physics, gameplay and (maybe) sound together for a very simple game. The gameplay itself will be something that an experienced game developer would be able to cobble together within a few hours at most.

Don't get me wrong, I don't want to curb your enthusiasm. I just think it's important to manage expectations. I'd rather have people being pleasantly surprised than dissapointed.

I think the best benefits of this will be A), having proof of concept, as Nimbal said, but also B), once we get to a certain point of playability and fun, we will actually have something as payoff for all the work invested into the project. For anyone who begins to feel burnt out working on the project, just being able to play whats been made so far and have fun doing so helps a tremendous amount, plus its just plain fun.

In terms of attracting new game developers, I think it will be gradual, since the early releases still will lack the most important concepts of evolution, editors, stage progression, and dynamic competition. But as we get those implemented, the game will begin to form more of a solid basis.