PlayNodes ready to kill the bricks

Game Engine Game Engine Support and Discussion
21

Did I read that right, so we would be able to just convert our games to use an all Python system with the HIVE nodes along with our existing scripts?

That would be awesome if it was true, and it may even then be that should any future plans call for the BGE to be replaced by the use of Panda 3D, all that would be needed then is for that engine to read the BGE modules and we’ll be good to go on quickly getting our projects back up to speed while maintaining the advantage of the ‘press P’ system.

Then you would just have OGRE run the graphics and the conversion process for our projects will then become quick and easy. The advantage being that those in the BGE community will pretty much lose nothing should the BGE be retooled as just an interactive environment system :slight_smile:

22

There is also Kupoman’s project for conversion to pure python (the difference is you would need to script the logic bricks from then on)

Whatever happens, Panda integration (or other engines) will always heavily emphasise user-friendliness.

23

The summary of my position is that if things could be done in a way to replace the BGE with another Python-based engine (you mention Panda), that could actually serve to resolve the potential issue of it taking a long time to get projects back up to speed (which includes having OGRE power the GLSL in the viewport when ‘Blender game’ is selected, then I’m all for it.

The HIVE converter idea alone would allow for the removal of the Logic brick system due to the former having much better code design. If nothing else, Panda 3D could at least temporarily have the ability to read the BGE modules so as to serve as a grace period to get our important scripts ported over to their system while keeping our projects fully functional (though I’m not sure if they would have an equivalent to the bullet-specific functions like rayCast()).

So the new way to create games in Blender is through a fully-integrated environment where Blender remains the app. that all the development takes place in, but the development possibilities itself provided by full integration with OGRE, Panda, and HIVE.

One of the key reasons I wanted to see the BGE stay is because of how much time it would take to port my larger projects to another engine, but seeing the possibility of having a FOSS developer environment that actually makes it a rapid process, that would no longer have a case.

Everyone wins in this, the BGE community will still be able to work on their projects and the rest of the community gets their interactive mode.

24

Adding ogre in the viewport, would be fun.

There must be a way to do it like cycles using bpy.types.RenderEngine.
I mean getting the rendered pixels from ogre and pasting them in the view, the same way we integrate ogre in wpf.

viewport render

def view_update(self, context):

25

Indeed, that’s the planned means of integrating Panda3D.

26

Panda3D is so easy to use !

Has the project started, any links ?

27

I support nodes for the BGE in any way, be it Hive, Panda, or Ogre or a new system, or a fusion. Collaboraiton and the like. it’s much easier to work “logic” than write lines! I think it would make it that much more fun. =P Keep at it! I like the sounds of “interactive” central hub called Blender with the outlets to the engines that could be integrated.

+1

28

There was a demo somewhere, but a number of things need to be finalised before proper integration goes smoothly (Python3 official support, etc)

29

You and cubedParadox should be collaborating :smiley:

30

Haha, I just saw “GameNodes” video.

ill post a video too asap !!! The competition is hard !

31

Here a mouse look faster than light.

32

Looks like Unreal Kismet, but better! :smiley:

33

Hey !

I couldnt resist to show you the WASD movement setup.

I really like the clean node graph that we get compared to the logic editor.
Things make so much more sense when you see them connected together without overlaping.

For these moouselook and wasd, i had to create the required nodes when i needed them.

I believe that developping PlayNode this way is funnier.

I mean thinking of a little demo, then designing the nodes that should be used to perform the desired effects, and finally implementing the effects in the runtime.

Do you guys have some suggestion ?

I have a some ideas already, about various little game templates.

34

You try something like delete overlapping instanced objects.Is that too hard to implement?

35

Do you mean deleting objects when they collide ?

36

I have some suggestion for PlayNodes.

  1. Make a button to run PnCompile.py scripts.
    example :

  2. Run the Pnui.py script as addon.
    example :

  3. Replace Exec label with In or Out (Input/Ouput) label.

  4. Replace obj: label with Object: label

37

Yes Its supposed to be an addon with various tools including the node editor, maybe some blend tree/state machine designer.
You dont need to press a button for compilation, this must be automatic when you press p to play for example, it will compile and launch the game.

in/out is a good idea, but in fact there is no need of a label for this socket, in kismet they use a white arrow. Placed Left is in and right is out, intuitive enough, because execution only happens from left to right.

Do you know if its possible to align these in/out sockets in the first line with the layout system ?

38

I suggest not making every single node manually, but instead, creating a node with nothing but a text box where you type a python function in, then based on this function it adds in node inputs as function parameters and outputs as returns. So that way you can even make your own classes that can be used within a node.

39

I don’t know if that’s possible to do like that.

40

I mean delete one of the instanced objects when they collide.