TSC Meeting Notes 2020-06-04

Owned by Cary Phillips
Nov 04, 2022
2 min read

Attending:

  • Cary Phillips

  • Christina Tempelaar-Lietz

  • Eskil Steenburg

  • John Mertic

  • Joseph Goldstone

  • Kimball Thurston

  • Larry Gritz

  • Nick Porcino

  • Owen Thompson

  • Peter Hillman

  • Rod Bogart

Discussion:

  • John Mertic revives the discussion of the logo design:

    • Larry referenced https://landscape.aswf.io. Some look great small, some do not.

    • The concept was to demostratine dynamic range in the layers.

    • The original idea was that the logo could appear as an icon on a camera.

    • Eskil: Think of the original photo by Debevec with same image at different stops.

    • Nick: The most important feature is dynamic range; Nick proposes a drawing that includes and image ad different exposure levels: night, day, sun.

    • Could also illustrate "multipart."

    • Rod: Should not look like photoshop layers.

    • Could look like an image coming apart.

    • Could show channels vs. some other multi-dimensional.

    • Needs to be easy to reproduce (t-shirts, etc).

    • Could be an element that’s a response curve, except that exr is linear(?)

    • Logos should be clear: NASA has a continuum of logos.

    • Cmake has really simple logo, has nothing to do with.

    • John will invite the LF graphic designer next time.

    • After the meeting, watch this: https://www.youtube.com/watch?v=pnv5iKB2hl4

  • Imath repo project:

    • Revision history can be done later

    • Need to resolve the handling of exceptions ASAP.

      • Consider the new C++11 exception system categories feature, although that's mroe for OpenEXR itself.

      • Imath deals mostly with math-specific exceptions.

      • Nick: I ported a lot of Occulus’s code but comparisons don’t work across boundaries.

    • IlmThread stays behind in OpenEXR.

    • PyIlmBase goes to a new repo.

    • Autotools? Leave it behind; if someone objects, we can bring it in later if needed.

    • OpenEXR should encorporate Imath as a Git submodule. We’re creating proper “find” cmake files; if it finds it, it’ll use it, if not, it’ll build it.

    • Nick: I tend to write shell scripts to handle all this stuff; I can put together a Gist to illustrate.

    • Or, just build the dependency first?

  • Christina: need Kimball to check her latest PR.