Hi Joseph, Yeah - I just thought it may be nice to have the capacity to provide the argument. That way you're not filling a list of color spaces with variations of the same one. Makes the List a

Nice work. A few comments: OCIOYaml.cpp * There was a missing #include <cstring> * For the Enums, (line 512+) this could could be greatly simplified by relying on the helpers in OpenColorTypes.h. *

I've also cleaned up the enum code. See 1307a8730ffe745daeb53cff4298518ed5ceafc3

Alan, By dynamic I mean c++ or python can create colorspace configurations dynamically. That way if you are using ARRI log C You could dynamically create or alter the colorspaces to handle the

Hi, I think you just fixed this, but yeah there was a bit of code dup. With the current structure it would be possible to have GroupTransform()s containing GroupTransform()s and possibly more

Only addressing the GroupTransform issue... I am in favor of keeping GroupTransform as is. Having a group of transforms (at the API level) be interchangeable with a single transform is a very

Hi, Comments below. +1 this would clean the whole thing up for me. Yeah thats sounds cool, ColorSpace.setTransform(Transform *) clearly makes a GroupTransform just another Transform which I like a

Addressing the YAML topics: * I have pushed your commit, plus my non-contentious updates, to spi/yaml. Once build issues are worked out, this will be moved to spi/master (and delete the yaml

It does compile on osx, so I'll do further linux investigations tomorrow. -- Jeremy

2.8 +1 - OSX 10.6 - CentOS 5.3 - Debian squeeze (testing) do you mean adding this to the yaml-cpp project? set_target_properties(<target_name> PROPERTIES COMPILE_FLAGS -fPIC) No bait intended

Hi, This should fix the link error on linux http://github.com/malcolmhumphreys/OpenColorIO/commit/7b016f274aefd8a4ce2deb7c72639351db8de0b4 I always forget how to create these diffs so I guess other

PyOpenColorIO.so wouldn't compile on linux http://github.com/malcolmhumphreys/OpenColorIO/commit/bb90284fa98161aae4ace74987246ae301ec1822 - Added some const_cast<char*>(..) ie. 'error: invalid

Works great on our linux distro now, thanks! (LGTM) I agree that it may make sense to include boost/shared_ptr.hpp, let me look into this. -- Jeremy <malcolmh...@...> wrote:

LGTM. This was a compilation warning though, right? (It shouldnt have been an error). -- Jeremy <malcolmh...@...> wrote:

Folks, I've deleted the OCIO yaml branch. The main trunk now is up to date with all recent commits. -- Jeremy

* cleaned up malformed profile name (ocs_profile_version vs ocio_profile_version) * added better loading error checks * updated raw profile to new yaml

This makes the naming unified across C++ / python, and also gets rid of "getXML" (which is no longer xml). js/master commits: bd559608bdfbbc930634ef78eebd42cda83647da

LGTM

LGTM, Did you see my comment in one of the threads about also supporting the .__str__() method?

- Added missing NAME arg to add_test() as this was failing to work on linux (worked on OSX, small cmake