This allows group nodes inside other group nodes in cycles and makes the
code more generic for all possible cases, like direct group
input-to-output links and unused group sockets.
Previous code tried to connect external nodes and internal group sockets
by following links until a "real" node input/output. This quickly
becomes complicated in corner cases as described above and can lead to
unexpected behavior when the group socket is of a different type than
the internal/external sockets, but that conversion is skipped.
The new code uses the concept of "proxy nodes" similar to what the new
compositor does. Each group socket is replaced with a proxy node with a
single input and output, to which other nodes in the same tree and
internal nodes can link to. After all groups have been expanded in the
graph, these proxy nodes are removed again, adding converter nodes if
necessary.
Node specially useful for Texture correction.
This is also a nice example of a simple node made from scratch in case someone wants to create their custom nodes.
Review by Brecht.
reviewed by Brecht, with help from Lukas.
Note: dot is reversed compared to Blender.
In Blender Normals point outside, while in Cycles they point inside.
If you use your own custom vector with the Normal Node you will see a difference.
If you feed it with object normals it should work just as good.
lower than 1.3, since we're not officially supporting these. We're already not
providing CUDA binaries for these, so better make it clear when compiling from
source too.
* Added option "WITH_BF_CYCLES_CUDA_THREADED_COMPILE" for the people who have much RAM (8 or more) and can compile several kernels at the same time. If enabled, it uses the general BF_NUMJOBS flag.
* The option is off per default.
Fix#29475: remove node from properties editor crash on windows. This was a bug
in the UI code, which code access removed data.
Fix OpenCL still being used in a case where Experimental was disabled.
Fix msvc debug warning in md5 code.
as with the HSV node the OSL code is relying on the (yet to be implemented) autorename.
Also the svm code could use mix (svm_lerp) instead:
32 . float3 color_inv = make_float3(1.0f, 1.0f, 1.0f) - color;
35 . . stack_store_float3(stack, out_color, svm_lerp(color_inv, color, factor));
I have a feeling that each node 'program' should have the least program as possible. I'll see with Brecht later.
But overall I don't know if that's any fast. And apart from that I think we will need this kind of function to move to a library if multiple functions linked in are not a problem.
----------------------------
reviewed and approved by Brecht
Important note:
the camera Z is reverted compared to Blender render.
Now it goes from zero (camera) to positive (in front of the camera)
.........................
note, the OSL code has a problem.
In the original node the input and output nodes have the same name (Color).
So this will be fixed here once Brecht come up with a nice autorenaming (or we do a doversion patch) for that.
Array indexing doesn't work there.
I'm yet to setup my CUDA computer, in the meantime this proved to work (tested by Daniel Salazar).
If I found other ways of doing it I get back to that.
* Compile all of cycles with -ffast-math again
* Add scons compilation of cuda binaries, tested on mac/linux.
* Add UI option for supported/experimental features, to make it
more clear what is supported, opencl/subdivision is experimental.
* Remove cycles xml exporter, was just for testing.
bit so that it's more symmetric and resulting float ops are compiled the same
way. Otherwise extended float precision being used in one place and not the
other can make comparisons fail.
"The Blender Foundation also sells licenses for use in proprietary software under the Blender Licens"
also remove NaN references from files that have been added since blender went opensource.
by Gaia Clary.
Rationale: the name was confusing and not always used consistently, and this
map itself is not something that can be layered, rather the map can be used
as texture coordinates in some layered setup.
The original intent was to indicate this contained more than just UV's, but
the game engine settings have already been moved out, and apparently users
didn't really get this from the name anyway.
Some drivers don't support passing include paths with spaces in them, nor does
the opencl spec specify anything about how to quote/escape such paths, so for
now we just resolved #includes ourselves. Alternative would have been to use c
preprocessor, but this also resolves all #ifdefs, which we do not want.
* Reduce kernel arguments size, helps compile for apple nvidia.
* Fix use of unitialized variable in displace kernel.
* Use build flags in opencl kernel md5 hash.
* Reorganize code for kernel feature #defines a bit.
