The nitro-meca project

[Proxx]

Hi,

today I would like to present you the current development progress of my project: a game creator for LBA like games called nitro-meca.

On my webpage http://www.nitro-meca.com I post from time to time the newest implemented features and plans for the future.
If you like, drop by and leave a comment. I would be happy.

To get an idea about the look and feel, here is Twinsen's house created with my game editor:


The overall plan is to create a tool that let player build rooms, houses and islands, plan and document complex stories and model the game flow with a graphical node language. In short, it should be possible to create an own LBA game.
As I am working on this project for quite a long time already and the first pre-alpha release is going to be released in the next weeks, I think it's a good point of time to publish this project now.
Check it out, I'm curious about your feedback.

[Darkflame]

Nice work! Skim-read the blog, looks like interesting stuff.

I'll be interested to see how the node language works.
One of the hard things with making an editor for this sort of game I always thought is flexibility. Making the maps is (sort of) just going to tile and object placement....but making gameplay dev systems that takes thinking about.

Have you considered a object property system (if its not there already), so you can associate values with objects to be changed or queried later? That sort of thing can often help make things flexible for scripting.
I designed a scripting language (no fancy node editor) for a online point and click game engine, and it proved very useful once I added optional property/variable support for any sprite. (hitpoints left, coins stored in chest, switch in position B....all that sort of thing ended up using the same system)
Even ended up defining "flammable" on all wood objects and letting fire spawn on them which was fun :P

[Proxx]

Thanks!
Yes, flexibility is indeed one of the main problems. Another one is finding the balance between hardcore-detail-modeling (whichs gives much freedom in game flow creation) and fast-and-easy-modeling (which is better suited for users that just want the things to run somehow). My system will try to offer both: detailed modeling components will be parameterizable and reusable.

All of the game objects implement some standard intefaces (depending on its type), and using this interfaces in the game flow modeler can than be used to modify them. Example: a wall is (among other things) an IHidable and can thus be hidden at runtime. It could, too, implement IFlamable.
Besides, all objects can be assigned custom properties (in term of a value-key-pair) that can further be used for game flow modeling. So a locker could have the property ("contains red keycard", "true") before Twinsen opens it and then switch this property to "false" after opening.

[Darkflame]

Besides, all objects can be assigned custom properties (in term of a value-key-pair) that can further be used for game flow modeling. So a locker could have the property ("contains red keycard", "true") before Twinsen opens it and then switch this property to "false" after opening.

Ah, perfect.
Thats what I was thinking of.

Look forward to seeing how it comes together.

[Proxx]

I just created a new post in my blog about new set elements and some editor improvements:

http://www.nitro-meca.com

Come along, if you are interested.
Attached, a screen of the new wall types.

[Polaris]

Nice hybrid room !

[Proxx]

Thanks!
I'm trying to keep the Set Editor as flexible as it's possible to allow all modelers to be free in their creativity. On the other hand, no one should be stressed by being forced to model sets on polygon level. There is some kind of balancing needed here.
In the moment each modeled wall can be modified by a set of basic parameters that change the textures and dimensions. This seems to be a quite good trade-off.

[xesf]

Indeed nice work.

Good to see people inspired by this game and make their own tech around it.
I really don't see this as "yet another LBA project", but an inspiration to work hard and learn from this master piece.

Keep up

[marcosmapf]

Your work looks wonderful! As xesf said, its realy nice to see that some people are so inspired to the game that they are creating their own engine based on it! Awesome job, keep up with it!

[DarkOnistar]

Saw this thread recently and I think it's pretty cool; I liked map making for the LBAnet project but the limitations in the editor really hindered progress as I don't have much experience using 3D software such as Blender.

[Proxx]

Thanks, I am happy that you guys like my project.
When I release the editor, I will definetly need some testers to find some bugs and to know where usability can be further improved. Are you then interested in helping me, DarkOnistar? Would be great!

Currently, I am working on windows and doors. I will try to implement some blurry glass for windows. Let's see if this visually fits into the LBA look and feel. Next screenshots are coming soon.

[DarkOnistar]

Sure I don't mind testing. I previously helped out in testing LBAnet and built levels with LBAarchitect; it was fun and easy to build when you have tilesets, 3D modelling and importing them is cool but requires time and skill which I never really got round to sad to say.

On an unrelated note I've been playing around with Unreal Engine 4 that has become free without subscription and I've liked how a system like blueprints simplifies the process for level design.

[Darkflame]

Originally Posted by Proxx

Currently, I am working on windows and doors. I will try to implement some blurry glass for windows.

How is control over shaders done at your end? Are you having to manually write the shader code, or does what your using give some usefull shortcuts?

Been learning glsl recently and while some stuff is simple stuff like bluring/refractions seem quite complex
--
Incidentally, If you are writing your own shader stuff though, and you ever need a good way to have text in 3d this is worth looking at;
http://www.valvesoftware.com/publica...nification.pdf
Very impressive results from tiny bitmap fonts.

[Proxx]

All shaders are written by hand. This way, I can combine several techniques like fbx model bone transformation (learned it in this tutorial by Eric Richards), lighting and normal mapping without having much passes. And I learn a lot.
I use HLSL shader language and currently I am investigating if the Shazzam Shader Editor could be a good tool for me to test my pixel shaders.

I agree that refraction is somewhat complex. Because of this, I will try to just use gaussian blur for glass window effects. It's faster and may be good enough.

Big thanks for the paper link! I will surely make use of it when it comes to text rendering and conversation.

[Proxx]

I just finished my blurry glass pane feauture and added some new elements to my editor in the last days: http://www.nitro-meca.com/.

The blurry glass is implemented with a render-to-texture technique that merges and blurs (using simple box blur) the already rendered part behind the pane and its texture.

I will now work towards my next milestone: improving interaction and game flow modeling. So next time, I will hopefully be able to post some short video about this.

[Darkflame]

Ah, that looks darn good and sounds pretty easy to do compared to some of the stuff I was reading.