Attendance:

Discussion:

Kimball discusses changes he has made or is contemplating. He’d made a cleanup pass on core, now merged in. But the real thing was a proposal as part of “bringing the goodness of core to the C++ API”. The particular issue was the way that C++ lets you construct headers as metadata repositories outside the context of a file. The group discussed three possible approaches:

1. header and attributes as they are vs
2. header and attributes implementation uses C types vs
3. halfway, with ImfHeader translating old to new [by which I think C types was meant]

Kimball is leaning towards providing some sort of context class to provide state to the stateless C API

Kimball also said he’d entered a PR for the bug that the Gaffer people wanted fixed “last night”
There was something of a discussion of Peter’s page on what linearity means. People were supportive.
As for the upcoming TAC yearly thing, what we’d have to present in an update would be (a) core changes and (b) discussion of new compression types.
Nick started a discussion of OpenEXR as a carrier of fine arts imagery; there was a discussion of what it meant for a digitized artwork to be ‘in scene space’. It turns out Li has considerable involvement in fine arts capture, and he emphasized how much more than just colorimetry (or spectral imagery) was involved, as far as 3D texture went. Nick stated they had taken a photogrammetric approach in the capture; presumably this implied multiple cameras. We talked about the appropriateness of using ICC profiles, but I think the consensus was that at most OpenEXR could be a carrier of a particular rendering for a particular environment with perhaps an ICC profile.