In short, fork the BGE starting with HG1’s branch, cut out the non-relevant portions of Blender, call the new project ‘GameBlender’, set up a Github project site, and port over over any relevant modeling, UVmapping, or texturing code when possible from the main Blender project. This is what you might consider a radical solution, but the BGE (or GB as it will be known by then), could have an even brighter future than it would have remaining stuck to Blender.
Basically the entire idea is that this will allow Ton to do what he wants with the BGE while giving us a vastly enhanced version of the game creation tool that we’ve worked with for years, because the BGE will be an independent project using Github as the code and development repository which would then lead to a new website and community. Above all, it will still have the modeler along with the UV and texturing features, it just won’t be connected to the core Blender project.
In short, we don’t have to give up and all move to Unity or one of those other engines, the BGE is FOSS, we can do something about it, and we can move in a direction that’s under the complete control of a new project leader, the BGE devs., and its users. In final words, it would be better if we got the wheels turning now than on the eve of the current GE disappearing.
Any thoughts on what’s in the linked post and this proposal?
The BGE source code is too integrated with the rest of Blender to just separate it out and connect it back with an addon (or at least it’s not possible to do it without some major rewrites to replace the interfacing that the engine lost).
It’d just be easier to let it stay connected to the Blender code and instead remove the parts that the game developers will not use.
The issue is not this. Firstly, there is no fundamental need for a “major rewrite”. There are places that need work, but they’re well abstracted to avoid causing large issues. Secondly, it’s the concern that Ton’s solution is to remove functionality, without the intention of replacing it.
I dont agree with completly scrapping everything, I think what makes blender so uniqe is that it has blender and bleder the game engine all together in one. I think you would have to have heavy heavy compatibility with the new engine and blender so that you can still export meshes and what not to the new engine.
The new project will still have all of the modeling, texturing, and UVmapping features of Blender trunk, so you’ll still be able to make your assets and your game without the need of exporters. You’ll also still have all of the animation, shading, sound, and Python API features that are currently used in the BGE.
The only reason some level of compatibility would be required is to allow the porting over of various new features, overhauls, and cleanups that concern modeling, UVmapping, texturing, the node editor, the interface, the modifier system, and the material stack (in a way, anything that benefits the BGE developer’s workflow). The overhauls like expected with 2.7x and 2.8x will be a bit trickier, but some of these are things that are currently not used all that much by those using the BGE.
One thing though, the hypothetical GB app. will still be GPL because it was forked from an app. containing GPL code.
Though to ease licensing fears, a dev. could just make the BPPlayer the default gameplayer rather than Blender’s current one (it creates a shell around .blend files so as to make the source almost impossible to access.)
Kupoman has recently said that Ton’s proposal in its current form may not be final which may or may not decide if such a project is even needed, but we will need to see a serious pickup in BGE development inside of Blender trunk if the BGE is to remain an official part.
I think the first step would be to merge all the things that need to be merged? You have the Harmony branch, the Candy branch, HG’s build, all the patches…
The main problem of the bge is the lack of developers.
Splitting the the project into two could even make it worse.
In my opinion the worst thing that could happen is the blender foundation thinking that the bge doesn’t need any more love since the bge guys left for running their own fork.
EDIT:
having read this (http://code.blender.org/index.php/2013/06/blender-roadmap-2-7-2-8-and-beyond/) made me overthink the above.
If the ge in its current form will really be trimmed down to a realtime visualization prototype thingie a fork could be a good idea.
But I would wait until Ton gets a bit more explicit in what he intends to make out of the ge before making any hasty decisions here.
-No more BlenderPlayer.
-Support for multiple external game engines.
-Better Bullet integration into Blender
-“Interactive mode” is something like pressing play in Unity and changing stuff while the game is running.
The BlenderPlayer was geting excluded from Blender since we got the big 2.5 version. It was a huge change/rewrite for Blender, setting the path for future development and BlenderPlayer was just updated to run new blend files (rna/dna).
I don’t think that loosing BlenderPlayer is a bad thing. We already have alternative patched versions that can play blend files, there is Burster for web games, unity has good support for blend files, maybe this will take us even further. And just to mention that even the only BF game project used Crystal Space (engine) integrated in Blender.
That makes sense. It’s just hard for me to believe that Ton was talking about reducing the Blender Game Engine into something basic, just for Blender users to have some virtual reality. I couldn’t imagine that. I only hope functionality will not be reduced. I’m guessing that in the new API old attributes will be rewritten and will have new names maybe. But I just hope none of those will be emitted, except for those that are already labeled deprecated.