September 19, 2019

Host: Carol Payne

Secretary: Doug Walker

Attendees:

  • Rémi Achard (TSC) - DNEG
  • Mark Boorer (TSC) - Industrial Light & Magic

  • Mei Chu (TSC) - Sony Pictures Imageworks

  • Sean Cooper (TSC ACES TAC Rep) - ARRI

  • Michael Dolan (TSC) - Epic Games

  • Patrick Hodoul (TSC) - Autodesk

  • John Mertic - Academy Software Foundation / Linux Foundation

  • Carol Payne (TSC Chair) - Netflix

  • Mark Titchener (TSC) - Foundry

  • Carl Rand (TSC) - Weta Digital

  • Doug Walker (TSC Chief Architect) - Autodesk

  • Kevin Wheatley (TSC) - Framestore

  • Name - Organization

OCIO TSC Meeting Notes

  • Update on new ACES Configs:
    • Carol: Wanted to update the TSC on the progress of the Configs Working Group.  Plan is to release the configs with the 2.2 library at the end of October.  We're planning one more beta release next week that will incorporate the remaining camera color spaces and fix some other issues.  We've been collecting feedback in issue #67.
    • Mei: I'll take a look and have Chris Davies do so as well.
  • 2.2 Release Update:
    • Carol: Wanted to update the TSC on the progress towards the 2.2 library release at the end of October.
    • Doug: I adjusted the Project board for 2.2 a few weeks ago.  The original board was a place to collect ideas from people about what they wanted to see for the 2.2 release.  But now that we're closing in on the release, I updated it to make it clear which issues had developers who signed up for them.  So I moved a lot of issues to the "No Dev Assigned" column.  The remaining issues are what we hope to achieve for the 2.2 release, assuming things go smoothly.  This list is aligned with what we covered during the ASWF OSD presentation.
    • Michael: Epic has a PR bug fix submitted (#1684), just needs our CLA sorted out, which is in progress.
    • Doug: Where do we stand with Remi's OIIO/OpenEXR PR?  Michael: I reviewed that and it's good to merge.  There are some CI tests failing but it will be better to merge it first before adjusting the CI.  I will take an action item to merge this and look at the tests.
    • Carol: Regarding issue #1686, I will reach out to Sean to submit a PR for the LogC4 Built-In.  Doug: Sounds good.
    • Mei: Are there some good C++ first issues?  Doug/Michael: Issue #1455 (OFX FileTransform) would be a good one.
  • Sourcing FileTransforms from a stream:
    • Doug: As part of the archive feature (#1627), we'll provide a way for a client app to provide LUTs via a stream rather than a file on disk.  We're wondering if anyone is currently using the OCIO API function SetComputeHashFunction to override the way OCIO calculates the fast file hash/fingerprint of files on disk?
    • Michael: This was issue #1150 and implemented in PR #1169.  The idea was to support workflows where the LUTs might come from some external system such as Perforce or a virtual file system, where using the OS file system stat call would be insufficient.  But I'm not aware of anyone actually using it.  Sounds like it would no longer be needed with the new feature being added.
    • Doug: We weren't sure how it would be used in practice, since the rest of the library currently still needs a file on disk for each FileTransform.  But with the archive feature, we will be adding the capability for the FileTransform to be provided by a client application independently from a file on disk.  May wind up deprecating SetComputeHashFunction as a result, looking for feedback on that.
  • No labels

1 Comment

  1. Doug, a data point: as you know we use the stream extensions we added internally to OCIO v1 (and I shared them to your as reference for v2). We do not use SetComputeHashFunction.