YESH!
I always found the current approximate with indirect light looks BAD. Shadows are unrealistic.
The other two I’ve not really head abut but sound great to me!
I only use the BI for my renderings and AO supporting transparency (alpha maps) is something I really would like to have.
I only use the BI for my renderings and AO supporting transparency (alpha maps) is something I really would like to have.
Support AO react to normal (bump), will be nice to.
yeah, and maybe lots of time to render even a simple image*?
Yes it is took long time… but if texture plugins will return, then Photon mapping may be possible as texture… (Environment map texture as example)
(this one)
Use geometric progression in calculation of GI - this is wrong way…
In start of theme i wrote only about returning old functions, but not comparison of renderers… Thanks to trolls like Ace Dragon…
Since i’m afraid i sounded rude, i’ll just answer your initial question:
point 1 and 3:
the whole discussion so far: development in now focused on cycles and those old BI technologies are not worth the time
point 2:
if i was a developer i’d rather spend my time implementing new cycles textures instead of (again) put my hands in involuted code and bring back a platform-specific feature
If only Сycles wil be support in future and Blender Internal will be discontinued, then Cycles must support non realistic non raytrace shaders like Lambert, Phong etc. Cycles is not bad renderer, but not flexible…He too narrowly specialized. It is look like with LUX renderer you try produce fast animation render… Internal must be extended to Cycles and Lux funcionality, but if think try work Cycles and Lux as Internal - is not possible 
@Old_demon
IMO, A new, modern OGL, a GLSL render is more appropriate for non photorealistic renders.
BI is very slow on such tasks.
BI is very slow on such tasks.
Only because he is not develop 
You’re really not getting this. BI cannot just be developed to be made faster. It would be easier to throw it away and start over. This is why Cycles exists. The effort of improving BI is not worth anyone’s time. It is, for all purposes, a dead engine. Attempts to improve BI multiple times have been failed experiments by very capable render developers. No one wants to touch it. No one will touch it. Big, new features will not be coming to it because no one has interest in extending a renderer based on 80s/90s paradigms, unpredictable light transport, poor material definitions, a poor material system in general, and spaghetti code on top of everything else. The rendering world has moved on to bigger and better things. Even Renderman has finally caved and moved into the 21st century full swing. BI is a dead system. It cannot be extended to properly support modern techniques without unrealistic effort from a team of render developers, which Blender doesn’t even have.
Because it is not worth the effort.
From my experience and for the render quality I want, BI is way faster than Cycles, also without any noise.
I don’t do photorealistic images or animations, I do something between cartoon/manga and realism (3DCG manga?), and I can obtain a 1280x720 image rendered (compositing included) in less than 2 minutes with a Core i7 4770 using the BI. I try to do the same in Cycles with 3 or 4 GTX970 and there’s no way I can obtain any clean image in less than 15 minutes (and noise persists on some areas) with the same scene.
I know there are a lot of settings for Cycles in the render panel to reduce samples, or some other in the object panel to disable diffuse/glossy/transmission/… per objects, etc, etc but seriously for my needs, BI and its shaders are way faster to configure. And I don’t believe it’s just a question of being used to this or that render engine…
Because it is not worth the effort.
Well, then another question for Cycles maniacs: Include in Cycles photon mapping function and/or irradiance map function as a primary bounce, it is worth the effort? How long are you willing to wait for the realization?:eyebrowlift:
If you’re sticking with BI because you don’t need photoreal, you’re wasting colossal amounts of time. You could be getting the same quality out of something like the CryEngine movie maker or UE4 (or even something like Panda3D if you have zero budget) hundreds or thousands of times as quickly, with material and lighting systems that are just as simple to use. BI is, in most cases, even LESS capable than modern real time engines in terms of feature set.
Given the availability of render developers to the Blender project, no, it’s not worth the effort. We have one, maybe two truly capable render devs who can work on big features since Brecht left. Photon mapping and cache algorithms have never been intended to be features of Cycles. I don’t doubt that they will come eventually, but now is not the time for them when Cycles in its original configuration still needs work.
Are you telling me to use an external real time engine made for low-poly instead of BI, which allow me to use all Blender features, for my projects which aren’t low-poly and for the one I need sometime soft bodies or clothes and sometimes particles?
If you’re sticking with BI because you don’t need photoreal, you’re wasting colossal amounts of time. You could be getting the same quality out of something like the CryEngine movie maker or UE4 (or even something like Panda3D if you have zero budget) hundreds or thousands of times as quickly, with material and lighting systems that are just as simple to use. BI is, in most cases, even LESS capable than modern real time engines in terms of feature set.
m9105826, he will simply go on 3ds max, Maya or Cinema 4D or another software… but he still use Blender…
As long as needed.
And, reading through this thread, I didn’t notice any maniacs around.
If I’m after a non-photorealistic render, of course, I have to use BI, I never started any war against our old friend.
As m9105826 stated, you’re really not getting why BI further development stopped.
Nobody said that Cycles is the appropriate renderer for every task.
BTW, I’m among the few who found some cycles nodes (like the toon shader) a little pointless.
A new OGL based engine to replace BI, what do you say?
A difficult and challenging task, lot of potential on this development though.
Yes, especially when the modern cinematic tools allow SubD models, and have internal soft body, cloth, and particle tools, the latter being superior to Blender’s in pretty much every way. Again, there is no case where someone is rendering with BI that isn’t wasting valuable artist time in 2015.
The topic is not whether certain features make sense or whether they are worth to be developed for Cycles as far as I know.
Believe it or not, I’m not wasting my time using BI, Blender’s simulations or even its non-destructive modifiers.