Academy Software Foundation Technical Advisory Council (TAC) Meeting - March 5, 2025

Join the meeting at https://zoom-lfx.platform.linuxfoundation.org/meeting/97880950229?password=81d2940e-c055-43b9-9b5a-6cd7d7090feb

Voting Representative Attendees

Premier Member Representatives

Project Representatives

Carol Payne - OpenCOlorIO / Diversity & Inclusion Working Group Representative
Cary Phillips - OpenEXR Representative
Chris Kulla - Open Shading Language Representative
Diego Tavares Da Silva - OpenCue Representative
Jonathan Stone - MaterialX Representative
Ken Museth - OpenVDB Representative

Industry Representatives

Jean-Francois Panisset - Visual Effects Society

Non-Voting Attendees

Non-Voting Project and Working Group Representatives

LF Staff

David Morin - Academy Software Foundation
Emily Olin - Academy Software Foundation
John Mertic - The Linux Foundation
Michelle Roth - The Linux Foundation
Yarille Ortiz - The Linux Foundation

Other Attendees

Lee Kerley - Apple
JT Nelson - Pasadena Open Source consortium / SoCal Blender group
Olga Avramenko

Antitrust Policy Notice

Linux Foundation meetings involve participation by industry competitors, and it is the intention of the Linux Foundation to conduct all of its activities in accordance with applicable antitrust and competition laws. It is therefore extremely important that attendees adhere to meeting agendas, and be aware of, and not participate in, any activities that are prohibited under applicable US state, federal or foreign antitrust and competition laws.

Examples of types of actions that are prohibited at Linux Foundation meetings and in connection with Linux Foundation activities are described in the Linux Foundation Antitrust Policy available at linuxfoundation.org/antitrust-policy. If you have questions about these matters, please contact your company counsel, or if you are a member of the Linux Foundation, feel free to contact Andrew Updegrove of the firm of Gesmer Updegrove LLP, which provides legal counsel to the Linux Foundation.

Agenda

General Updates
- Dev Days - May 2025 #966
- Security Threat model analysis for ASWF projects #615
Planning Session for TAC #972

Meeting Notes

General Updates
- Dev Days - May 2025 #966
  - Larry: coming up May 15th, every project is expected to participate, but individual projects can decide what they can do. At minimum have a number of issues tagged as “help wanted” or “good first issue”, so people can choose from issues that are appropriately sized / skilled. Try to pay attention to Slack / Mailing Lists for new people who need help. Make sure your build and test documentation is up to date. That’s the sate we would like to think projects are always in. You’ll mostly see good participation and not extra work to do. Encourage everyone to participate!
  - Larry: on the company side, now is the time to recruit juniors in your org who want to try out open source. As leaders in your companies, make sure that anyone who wants to participate has all pre-requisites done such as CLAs.
- Security Threat model analysis for ASWF projects #615
Planning Session for TAC #972
- Google Doc
- Larry: Open brainstorming. We want TAC to be more than just annual reviews and procedural discussions. We want a significant component that is of service to member companies and projects. This is a chance to make suggestions.
- John: also are there initiatives the TAC should take on.
- John: everyone can enter ideas in the document
- JF: collaboration between projects
  - Encourage cross project initiatives
  - Gary: important to OpenFX, how to collaborate with OTIO to represent effects in timelines, OCIO on how to do color
  - John: CI integration, how much should that be leaned into?
  - JF: where’s the best place to share topics potentially relevant to every project
  - Cary: Nanobind vs Pybind11 subject came up, conclusion was “Nanobind is everybody does it or nobody does it”, but kind of died there. That’s the kind of thing that bears more specific investigation.
  - John: common tools and libraries. Cary: things that affect interop between projects. John: less “pick a winner”, makes it easier to consume projects.
  - Larry: OpenVDB went ahead and switched to Nanobind, so we already have different projects supporting both, so switch to Nanobind may be necessary.
  - John: how much of an overlap between toolchain dependencies vs VFX Ref Platform? Larry: there are parallels, but thinking of things that aren’t dictated by VFX Platform. Some projects thinking about CMake minimum version for instance, or the Nanobind issue, which are out of scope for VFX Platform. Cary: VFX Platform isn’t a set of recommendations of what to use, more here are the versions you should use. This forum may be more about recommending. Larry: VFX Platform is a spreadsheet of what the major DCCs are using, every project struggles with supporting this matrix, and how many years back to support. Kimball has given thought on OpenEXR to structure to avoid ABI breakage and reduce the pain of staying on the current version vs supporting old versions. This kind of knowledge sharing could be good in this forum. Lot of us would love to just be told how to handle these issues.
- Talks of general interest to this project (does our project us an interesting technology the rest of us show know about) Hard won best practice knowledge
  - Larry: no one can follow every project, when a project spends a lot of effort figuring something out, could share the knowledge. We could make a talk for Open Source Days at SIGGRAPH, or mention in annual review, but we could have 10 minute talks at a TAC meeting.
  - Larry: if OpenVDB wanted to give a lightning talk about conversion to Nanobind, would avoid every project starting from scratch.
- Gordon: I like the system with upvotes (pollev), we can collect ideas, understand them, find other process to get ideas.
- Gordon: prioritized view of open source gaps. What’s the role we play between “kingmaker” / allocate capacity/resources? Not sure who was able to listen to the steering committee for AI in pipeline. Where do we see open source playing in there, where are the gaps, where are there N projects when there should be 1, what are projects outside the foundation that are in use. A more “directed search”, not necessarily “here’s how you have to do it”, but more building a portfolio, where are the gaps, the opportunities, where do we have too much choice. Maybe a yearly goal to do this?
  - John: I see TACs do similar things, try to figure out what is our stack, what do we need to focus, where are there no good solutions. What are areas of risk, what are areas where there’s too much going on.
- Gordon: Do we need to do anything around Qt? We’re caught between commercial and LGPL world of Qt. As an industry it’s a de-facto project, the LTS changes introduced a few years ago introduce long term supportability challenges. We had some conversations here and in VFX Platform, should we have our own branch and maintain it? Are we at a point where we should think about UI frameworks? Should we look at other technologies, especially with less desktop centric UI scenarios?
  - John: a very topical issue.
  - Larry: to what degree are you referring to Qt use in our projects, or at the member studios? We’re all pretty bound to Qt because the major DCCs are using it. So if you want to write plugins for DCCs, you don’t have a choice. Are you more interesting in opening discussion about what DCCs should be doing?
  - Gordon: anything in the Steering Committee discussions that should be given to TAC?
  - David: collectively the TAC is the core, I’m here to support you, and we’re here to support the engineers who do the open source work. We’ve build a solid foundation in last 7 years to support the projects and infrastructure, we have something good and solid. It’s time to figure out what else we can do apart from supporting the projects, what guides us. Forum was about open source and new technologies, in this case AI. We may have missed opportunities with Virtual Production for instance, some were done by SMPTE RIS work with camera, came to present to us, we asked them to come back with something more structured. They open source OpenTrackIO at SMPTE, it’s good that Open Source is happening there. But if we had had more of a “think tank” segment to think about VP, we could have helped these folks who came to us to grow and have a VP project we could have used to grow in that field. That’s why we put Open Source Forum around new technology and AI, this continues into internal discussion and what we should do about AI. Thank you Gordon for the praise about the polls! The idea of collaboration between projects is important. Another concept could be to develop specific UI elements like a color picker. We should ask ourselves at least once a year, what can we do we’re not already doing. With 15 projects in the foundation, good question to ask ourselves how they can collaborate.
  - Gordon: want to be open at suggestions from Steering Committee.
  - Larry: are you asking about the open source projects, or what the major DCCs do about Qt? Studios have to follow the lead of the DCCs, since Qt is what they are exposing.
  - Gordon: would like to explore a new area for us to explore. We came in with mandate to support Open Source projects, have a thriving ecosystem. Qt is different, not suggesting we start our own UI framework project, but it is open source we all depend on. There is a huge amount of inertia around Qt, both with the DCCs, and I imagine in the studios. Nothing will happen overnight, but we have to start thinking about this sooner rather than later. Maybe ready our own branch so we can fix issues together. Or longer term we want to go over to other framework. Qt has some commercial licensing subtleties, and it’s not super strong on the web / mobile, which are platforms we may need to get to. We talk about this in the VFX Reference Platform, waiting to see if “something bad happens.”
  - John: maybe a parallel conversation to the Linux Workstation conversation from a year ago? It’s a problem in this area how can we help?
  - Gordon: we’ve left a lot of this to the VFX Platform group, but it doesn’t have the resources
  - Nick: I’m gonna have to call out the per-seat licensing of Qt, and the alternative of copy-left make Qt a very tough choice for anyone without Big Corporation budgets. Having worked at Big Corporations where only Blessed Engineers get a license, I think the costs can’t be ignored. inertia = sunk cost fallacy ;) Nick: the costs are also very high, the Qt license is antagonistic to the kind of work we do.
  - Nick: More suitable if you are already in the GPL world (GNOME desktop). Gordon is bringing up lots of points that are important to me. It’s not an issue we should brush off.
  - Larry: VFX Platform is about pulling consensus between vendors and telling people about those versions. It’s not about picking the best technologies.
  - Gordon: it’s all about “versionitis” how do we agree on the versions of a popular set of components, so we can share things. As a studio can hopefully pick a set of versions and products that target a specific year. What are the components that people are using where it’s important to align on the versions.
  - Larry: answering “which version of Qt we should use” rather than “should we look at something else”
  - Gordon: may still be the closest place to have those discussions.
  - JF: there is no longer a place that seems to curate patches to the open source version of Qt
- Gordon: Align on versionitis with USD? (Gordon)
  - Goal of USD shouldn’t be to re-invent the wheel on well established standards, so USD should connect to OCIO / OpenVDB / … USD has an important role to play in connecting these together. On the AOUSD, not sure what we’re doing for versioning. Trying to get MaterialX changes to work in USD, we got trapped in versions. Is this something we would like to have more of an active voice. Encouraging them to follow semantic versioning? Do we want to be more actively engaged?
  - Jonathan: the USD-MaterialX WG is very active on this topic, meeting tomorrow. Gordon: thanks Jonathan to get us out of that pickle! But we need a holistic strategy across projects.
  - Nick: API stability, ABI stability and data compatibility. API and ABI stability are almost contradictory in C++ world due to lack of C++ ABI and the tight coupling with the STL. It’s a very hard problem. API stability is addressed with versioning, but doesn’t address being able to move a plugin between DCCs. So how can the data be interoperable? Working on architectural proposal to provide profiles and capabilities systems, so you can state your compatibility profile. Looking at requirements of other groups like IFC or BIM to get inspiration for best practices, hoping to bring compatibility practices to USD based on experience from other industries.
  - Gordon: hopefully schema helps up wrapping and amplifying value of projects we’ve built here. So do our versioned projects integrate with the USD versioning scheme. Not trying to solve the topic here, but do we think we should have a role? Nick: I would support that. Once we have proposal in a publishable state, would be happy to bring it here.
  - Larry (chat): Nick, maybe in Slack can you follow up with links decoding a couple of those acronyms and how they might guide compatibility issues? Nick: posted to Slack
- Strategies on versionintis more broadly
  - Gordon: move the version coupling we have in our DCCs to a central point (USD) that links nearly everything. It’s going to make it much more important to get right.
- John: any expansion we would want to do?
- John: I should have started a pollev, maybe we can curate this first.
  - Gordon: can be hard to come up with this in 30-40 minutes, give people a chance to add other items
  - JF: share the document link explicitly in TAC channel
- John: homework assignment is to review the document, we will curate it and get it into a more structured state so we can prioritize and execute.
Larry: can you give us a reminder of which project reviews are coming up? Sometimes that slips people’s minds
- OCIO next meeting Mar 19
- Rez Apr 2
- OpenEXR Apr 16
- USD WG Apr 30
- OpenImageIO May 14
- OpenCue May 28
- WG CI May 28
- raw2aces: Larry working to reconnect