Blender: A request for opinion

I am writing this thread in regards to a post found in the Game Engine section of this forum.
I for one know, that i rarely move outside of the Game Engine thread, and so i know little of what occurs in threads related to the Blender Internal.
The topic of my post relates to the direction in which the development of Blender is moving. Before i start writing, i dare ask that you bear in mind that this is an opinion, not fact, and also that i acknowledge that a majority of Blender development comes outside of the Blender Foundation, and so moves in a communal direction.

There are arguably two main parts to the Blender family. Firstly, there is the orignal 3D modelling and rendering, whilst the other ‘member’ is the 3D Game Engine.
Over the years, the internal renderer has seen superb advances that keep it very close to that of expensive renders such as 3DS Max. This is because of the work put in by the Development Community. With the new addition of the Cycles render engine (which yes, i know is not complete), one can argue that it’s ‘all guns blazin’) for the Blender Render aspect.

However, now i take a look at the Blender GE. The Game Engine, has not seen major rehauls or development since that of Yo Frankie, and from what i can see, a very small number contribute to the improvement of features (and we will wait to see the fruits of GSOC). In the trunk, we have seen the following noteable improvements.
Soft shadows (although the feature freeze denies this at this moment), Components (not complete and also frozen), and a few other nameable additions.

When you compare the speed, and feature list of Blender’s GE to other noteable Game Engines that are also free, such as Unity Free (who make money by upgrades) and OGRE or UDK (licence, i know), we are not only lacking as you would expect from a contribution based development platform, but we are lagging maybe 3/4 years behind.
Simple features that the Game Engine lack include:

  • SSAO
  • HDR
  • Generic scripts, e.g MouseLook
  • Networking*
  • Armature deformation¬
  • Inferred Lighting
  • Blender Constraints
  • Depth Of Field
  • Reflections
  • In-Game UI

Not only this, but it is SLOW!
Without particles, fluid simulation of any sorts is very ‘hacky’ using displacement maps.
What i wish to ask is, how many of you know of the GE, and where is the development actually coming from?
Many thanks, and i hope you will consider my opinion/questions!
*I have made a networking component, but waiting for feature freeze
¬Does not deform mesh, and therefore physics in Game

I myself am a GE man. I hate rendering, but do it occasionally for clients. I agree, the GE is the ugly stepchild, so to speak, but that’s changing. There are a TON of threads about improving the GE - you are not alone.

I see a lot of requests for really advanced things in the GE. I say before any of that, let’s fix the shadow system. Bloom and SSAO would be nice, but point lamp shadows that don’t cost 25 fps each are, I think, more fundamental.

I was going to add that, but thought it would just increase the length of my post!
Yes, the fact that an GTX 460 OC’d to 900 mhz core clock still can’t handle 3 spot lamps, is just appalling.
But of course, the 3D render community ought to have a say, because - They are going to have contrasting opinion.

There is someone working on the GE for the Google summer of code thing. He/She’s asking for opinions. You’ve probably seen the thread.

I am a render type guy. I will admit though that sometimes I need to make interactive elements for presentations of various models ect. . Right now this is time consuming at best, with no “easy” way to make UIs (I know someone will probably point me to a thread, that gives me a tutorial, that shows a workaround. But honestly why should we need a WORKAROUND?), no ragdoll physics abilities (again, somebody hanging around somewhere is working on that, making a workaround), and no mouse look feature already coded into the Logic bricks.

Put simply,
Although I render mostly, When I do need the game engine it so difficult that I have to give up and get my friends over to code it for me.

Thanks for your veiw Wefyb; it’s balanced and shows some flaws!