I just pushed a change which switches back to setuptools as the distribute fork has long been merged back into setuptools. I have also pushed the versions

No hurry, unless 1.0.9 fixes a major bug... If we're going to bundle something then a (non-beta) version of setuptools would be appropriate but an option to use the system installed version should be

Sorry in the middle of a house move. It was part of the python 3 pull request someone sent. I thought it was a requirement to get it to work with 3. Once I'm back up and running I'm happy to revert it

Not sure why distribute was added. I'm very comfortable switching back if we believe it to be appropriate. -- Jeremy

Jeremy, Do you know why distribute was added? Looking at upstream it appears to be a depreciated fork of Setuptools. https://pypi.python.org/pypi/distribute/0.6.49 Thanks, Richard

Hi! Just tested and added a link to buildlog in #318 issue. Hope it helps, somehow. Cheers. -- Matteo F. Vescovi Debian Maintainer GnuPG KeyID: 0x83B2CF7A

Haven't done any real testing, but the build completes with docs off. Thanks, Richard

https://github.com/imageworks/OpenColorIO/pull/300 Thanks, Richard

Can you clarify what you mean by this? I'm not sure what the "pull 300 patch" refers to. Yah, I'm seeing that too. Looks like the docs build target is hosed for 1.0.9; looking into it... Can you

Ok, a couple of observations when trying to build for Fedora 19 x86_64 1. I was using the pull 300 patch which is still open. It looks like parts of it were committed, but not all of it. I don't think

https://github.com/imageworks/OpenColorIO/tree/v1.0.9 Will send a link from now on! :) (Docs are in the process of rebuilding, the website will be updated momentarily).

Hi again!

Guess I got there too quick, don't see it yet on github... Also, although it's not difficult to find a link would be handy :) Thanks, Richard

I've just tagged the latest OCIO master as 1.0.9. There had been a bunch of bugfixes in there, and it had been hurting our distros to not have a tagged release with the lates. -- Jeremy **Version

We've just put out the 1.0.9 tag. Please see if this fixes things?

Oh - on further inspection I see you're compiling against yaml 0.5.1 This is a known issue, being tracked here: https://github.com/imageworks/OpenColorIO/issues/318 -- Jeremy <mfv.d...@...>

Hi! Good to know ;-) Well... I can test it, obviously ;-P Gonna let you know the results of the build with that parameter set. Thanks for your help. Cheers. -- Matteo F. Vescovi Debian

Awesome, thanks Sean! Does anyone on this list have experience working with the Nuke GPU path? This API is new to us, and we'd love to get some experienced eyes on this code. :) -- Jeremy

Thanks for this report. It looks to be a symbol visibility issue related to our use of yaml-cpp, so this is probably fixable by editing the top-level CMakeLists file's "USE_EXTERNAL_YAML" section. I

Hi, Apologies for bumping an old discussion, but I'm in looking for a similar set of features to the ones Nathan described, and as the discussion seemed to have fizzled out, I wanted to show my