Why can’t we have both.Really great non programmer solutions and regular programmers.Just like making your own food as opposed to buying fastfood and tv diners.
Unfortunately, I don’t see it working that way. BGE was never intended to be a full game engine, and has already far surpassed what was originally meant for it.
That would be why I mentioned that Ton Roosendaal’s plans: he wants to make the BGE into a full engine with a full range of use from the Blender platform (Particles, Ray Visibility, etc.) so that it can act as a legitimate engine. But it will never be to the capacity of what we see in engines nowadays.
BGE is kind of a stepping stone for early developers. Everybody here makes something out of it, then hopefully moves on to bigger projects. And BGE will continue to fulfill that purpose, nothing more, nothing less.
So what bottleneck are you addressing?
Object -“wire” a polygon - “wire material”
addobject - ray start point,
scale to ray endpoint = Ray visualization,
particles - add in a “handle” for particles, using single faces, and moving them. (4000 particles at once?)
4000 faces. (done)
it’s just a transform on 4 (or 3) verts that re-iterates.
Render needs love. see ->http://www.clustermonkey.net/Applications/gpu-programming-for-the-rest-of-us.html
Hi there
Every game has something of an artist and programmer.What we can not do just by drag and drop, we have to remedy with some programming. I think all programmers have some artist and vice versa.I really think that the great programmers use their own game engine and BGE is to start with this fantastic gaming world.
Cheers.
Well I can code python, but I have yet to find anything as intuitive as the logic bricks,
any other engine to speak of?
Most engines have a form of visual scripting in the present day.
Engines are nothing like fast food. Unfortunately, and don’t take this personally as it is a general statement, we seem to be making a lot of broad meaningless statements here that don’t actually offer anything to the topic of discussion. Generally speaking, unless you can either offer an insight into how something works, or a criticism, only questions are permissible. Otherwise it’s just a lot of noise.
Sent from my Nexus 5 using Tapatalk
Ok broad general idea, what is the slowest part of the bge?
Honestly that depends upon your game. I suspect that the least performance would be some of the render code. However, the problems faced by developers are closer related to ease of development, rather than performance.
Sent from my Nexus 5 using Tapatalk
It’s the same old discussion. Of course every artist would like to work on the workstation presented in this video, because it looks so neat. And probably Blender could as well have such an interface with nodes, presets, and whatever it would make it more artist-friendly. However the task would be tremendous, and I don’t think anyone or any team of developers would be able to spend years working on it for free. I think it’s a bit disrespectful to the community, and especially to all freelance developers, to complain about the Blender Game Engine being not as ‘user-friendly’. People who do, are forgetting that it’s an open source program. Feel free to join in, become a programmer, and add some code of your own!
Blender game is amazing, I have not even hit a scene bottleneck yet, I am just trying to figure out what would help everyone.
Could we push the render into modern openGL? Or openCL or CUDA?
I don’t know enough about what we have to know where to go, is the render single threaded and on a cpu?
I have honestly no idea where to start, gitHUB?
I personally don’t want all those presets for BGE. The reason I love it so much is because you literally start from nothing, and create something that is truly yours. Another problem would be that as soon as you start to see all those presets, you add more and more weight to the engine. The more weight you add, the slower it gets. If we had all the presets of an engine like CryEngine, (ocean, terrain, terrain paint, particles, etc.) BGE would collapse at the frame. It just doesn’t have the prebuilt strength for realtime work, and I don’t think the argument of making BGE able to compete with is feasible. That doesn’t mean we can’t update it to make it more efficient using some of the advancements in render types we’ve seen over the last few years. 
BluePrintRandom, in 2.6.7 if I’m not mistaken, they changed the render type to something to do with your GPU. All I really know is that it took all the pressure away from the general rasterizer category.
I noticed I have not even been able to make a slowdown yet on my lil Msi afterburner 
The more we can get on the GPU, the more our engine will hold.
imagine processing every object in a scene, at once, then again if there are dependencies, with parrallela
Logic return .001ms
Ok tested the render, no animations or scripts I lost 3 fps at around 98,000 faces…
and for the record those faces are flat… there is just… a lot of them
this is multiexture
…which is exactly what I see? 58 fps with 200,000 faces? (maybe you have a good pc)
PD: I think we deviate a little from the topic.
What are you getting on the actor rig these days carlo?
It looks like we are more then clear, the only issue is armatures, and rasterizer from that,
Can armature skinning be done on the gpu in blender right now?
is it just a large quantity of transformations?
My English is killing me(I have some difficulties to understand at times), You mean the ragdoll rig of the other time? (jacki)
The current Rig, “Jackii’s rig” with the planet etc,
The only thing I see in the Snowdrop engine is , image based HDR lighting , soft shadows , tessellation and instancing extreme. All those things could be added to BGE in one weekend. But , considering how good we are with programming , it will probably take us 300 years to get it right. The node editor can get just as complicated and hard to read as a script , so I’m not sure if it would really help.
Same reason for me that I like the BGE. But from an artists standpoint I understand that an engine preferably would be fit to his needs so he wouldn’t have to be a text-based programmer too.