Kimball Thurston - Chairperson, Weta Digital Limited
Bill Roberts, Adobe
Gordon Bradley, Autodesk
Roy C Anthony, DNEG
Matthew Low, DreamWorks Animation
Christina Tempelaar-Lietz, Epic Games, Inc.
Brian Cipriano, Google & OpenCue Representative
Sean C McDuffee, Intel Corporation
Larry Gritz, Sony Pictures Imageworks
Jean-Francois Panisset, VES Technology Committee
Cory Omand, The Walt Disney Studios
Daniel Heckenberg - Animal Logic Pty Ltd / Industry Representative
Eric Enderton, NVIDIA & Asset Repo Representative
Sean Looper, Amazon Web Services
Michael Min, Netflix
Michael B. Johnson, Apple
Greg Denton, Microsoft
Sean O’Connell, Advanced Micro Devices
Mark Visser, Unity Technologies
Ken Museth, OpenVDB Representative
Carol Payne, OpenColorIO Representative
Cary Phillips, OpenEXR Representative
Joshua Minor, OpenTimelineIO Representative
Chris Kulla, Open Shading Language Representative
Jonathan Stone, MaterialX Representative
Other Attendees
John Mertic, Linux Foundation
JT Nelson, Pasadena Open Source consortium / SoCal Blender group
Scott Wilson, Rust WG
Allen Rose, Madison Square Garden
Alix Robinson, ASWF marketing
Michelle Martineau, Linux Foundation
Deke Kincaid, Digital Domain
Dennis Adams, Sony Creative Software / OpenFX
Gary Oberbrunner, OFX
Michael R Carroll, Intel, RayTracing Software Technical Enabling
Patrick Hodoul, Autodesk, OCIO
Pierre Jasmin, RE:Vision Effects / OpenFX
John-Paul Smith, Boris FX / OpenFX
Rachel Rose, ILM, D&I WG
Doug Walker, OCIO / Autodesk
Lee Kerley, Sony Imageworks
Morgan Prygrocki, Adobe
Ryan Stelzleni, Netflix
Apologies
Kimball Thurston, TAC Chair
Jonathan Stone, MaterialX
Joshua Minor (OTIO)
Nick Porcino
Agenda
Larry Gritz leading meeting
ASWF Governing Board upcoming meeting
Project / WG leads to provide updates
Google Summer of Code (Carol)
Not happy updates: only received 1 application
Lots of theories about what it happened, not necessarily what we did as AWSF, Google changed a lot to the program last year and this year
Smaller organizations getting a lot less organization from GSoC applicants (like ASWF)
So unless OCIO approves the one student we got, we are done
Larry: we did the same as previously 2 years ago when we got dozens of applicants
On GSoC mailing lists, other projects experienced similar
Not sure we need to do a deep post mortem, seems to be mostly due to changes on the Google side.
Appreciate all the work that got done to get the proposals ready
Larry: what’s the plan B, we still have the room to mentor students, do we have time to do something on our own?
Carol: we would be doing a disservice to try to do something on our own for this summer, also don’t have funding approved. Seeing what’s happened, we’ll keep an eye on it and see if they make substantial corrections, but we should do our own program for next summer and start planning now. The attraction of GSoC is the wider pool of applicants and notoriety of the program. But if we don’t get that anymore, may not be a reason to keep participating.
John: heard similar feedback from other projects that ran into similar situations. LF can help with individual programs, also can run through LFX platform. Can do a spring / summer / fall term mentorship. Sometimes get more candidates during spring / fall even with overlap with school year. Can work help get a proposal to the Governing Board. But yes, probably too late for this summer, but might be possible for Fall, or 2023.
Larry: we now have a longer time to plan for 2023 for a project on our own.
Carol: still have Summer Learning Program, people signed up to be GSoC mentors can consider mentoring for Summer Learning Program if you have time.
Larry: another benefit with going on our own is that we can make the outreach a bit more targeted, and go outside the bounds of who GSoC might have accepted. We can look for more experienced people such as grad students for instance. We have a set of projects that are a bit complicated for beginning students. So maybe this is a shift in our approach.
Project Reviews
There is a mapped our calendar of what’s coming so projects have a heads up, and get feedback on the last project review, OCIO.
JF: OCIO put the bar pretty high
Larry: did that capture the info we needed to hear? Do other projects know what they are in for to prepare their report?
Cary: we need to get our ducks in a row first, experience of a healthy examination of the project, which would be good for everybody.
Carol: was great to present, would like to see a bit more discussion, we were a little rushed, future projects reviews should leave time for questions around the project. Larry: any feedback you were looking for? What can the TAC do for the project?
Carol: brought up a couple of things around CI, how we talk about project health and measure how many developers are involved, should metrics play a larger part in determination of project health. Are other projects aligned on VFX Platform / release timing (need to have 2023 release 6 months early). Doug: more of a 2 way conversation between the projects. Learn from the expertise of the group. How do we maintain the long term health of the project, and every project wrestles with that. Looking for new ideas on how to do that.
John: can you talk about the long term health concerns? Doug: for OCIO, when we look at all the things we’re involved in, not only developing the library but the documentation, educating the community, answering questions, CI… there are lots of tasks, and always struggling to make as much progress as we would like. Also have several WGs going. The more people we can get involved and have them put in serious time, do code reviews… it’s a struggle to find people to do that. As with many ASWF projects, it’s a steep learning curve, a lot of the contributions is from a relatively small number of companies / members. Most of the TSC members are people from large VFX studios and a couple of vendors, would like to get more participation from outside. More students contributing for instance. Student contributions have been difficult to convert them to regular contributors, not just doing a single contribution. Except that other ASWF projects also wrestling with these issues.
John: challenge of Open Source is the “scratch your own itch” model, good in a lot of ways, but converting someone from interest -> contributing -> maintainer is driven by someone’s interest. “We have lots to do and not enough hands to do it”. We should look at what all those things to do are, should we get doc writing help, it is a skill that can be contracted out to a degree. This is a good exercise to do after these annual reviews: identify challenges, getting deeper, can share guidance from other projects (foundations). Sounds like every project may be struggling with these issues, could be a workshop at SIGGRAPH between project leads to come up with some ideas? We are stronger as a group, the more this can be shared, especially in an in person setting… Carol: a separate forum for this discussion would be great. Right now OCIO is doing OK, but looking at the future. This isn’t a warning about the health of the project, but looking longer term, looking at how widely OCIO is being adopted across the industry, it’s growing and want to be good stewards of this project. John: good to think about this now rather than later when it’s a big problem. Sean: when adding Rez to the project, a bit reason to join was to get ASWF support, does the ASWF guarantee support to a project, or is it more a forum of like minded people. John: “support” can be defined in many ways. Around ecosystem building, infrastructure, open source best practices, foundations are most effective in this area, can bring expertise across LF areas, security initiatives… A lot of the model at ASWF / LF is more around “teach someone to fish”, give communities the tools and resources to help be successful. Can be funding for documentation, mentorship, infrastructure, and being available for advice. We aren’t the ones that will “get things done”.
Hopefully people have had a couple of weeks to reflect on what was presented
Pierre, Gary and JP from OpenFX are here
Christina: feel like it’s a good fit, seems to be interest from other projects from OCIO and OTIO to work with OFX, good crossover. Initially wasn’t sure since it was more a specification than an implementation, but we would be able to include specification based projects.
Carol: a reservation, not specific to OpenFX. One thing we brought up is we need to be considerate of our resources, we have a lot of projects in incubation, would be interested in thoughts as to whether it’s good to let other projects coming in at once.
Pierre: the project might offer additional resources, don’t need additional resources from the ASWF.
John: are you thinking more in terms of projects come in, that would distract from people already contributing from other projects who might diminish their contribution to existing projects? Carol: thinking both.
Larry: not sure someone spending time on an existing project would diminish their contribution, more likely to attract new people. But it does add some organization overhead, there will be a time when there are more projects than slots in the year for yearly TAC reviews! But we will have to tackle this at some point.
Cary: I support without reservations, but OpenEXR may not get project review until next year, there’s a lot for this group to consider.
Doug: not specific to OFX, we have a recommendation that members of ASWF are supposed to contribute 1 x FTE to ASWF projects (or open source in general?), that’s another aspect, the more projects we have, the more this effort gets diluted (unless we get more member companies).
Larry: not sure that we’ve ever had a problem with companies wondering where to put in their FTE, they seem to have figured out where to allocate those resources. Not sure we are resource limited in this way. More work to be done to encourage contribution past the FTE commitment.
John: one way to navigate this, some time last year, TAC approved a Sandbox stage, so it can come into incubation stage as a strong project with a roadmap and a community. For projects in incubation stage, first months is getting everything in order, then it’s a quick shift into growth. Sandbox stage can help separate those stages, and ASWF can have to give less support at that stage. If there is a concern from the TAC on the resource side, the sandbox stage might be a good option. The TAC can have a quarterly touch with sandbox stage project.
Larry: sandbox stage seemed to be more for “speculative” projects that are starting out, whereas OpenFX seems like a mature project, been around for a long time, integrated in a variety of products, they already have a foundation. They are way beyond the sandbox stage, as opposed to a project like raw2aces that feels more like a sandbox stage project. OFX seems to be well beyond that. John: yes, I agree, more about reminding the TAC about the sandbox stage, and possibly as a middle ground. But yes, this is a mature spec. Larry: could get confused by abstract notions of what happens when our organization gets too big, we should look at that separately from the OFX proposal. Are we “assets” to each other, is it valuable to other projects.
Eric: are we ready for a vote?
JF: propose a motion as an incubation stage, Sean: second. Ayes, one opposed, looks like a pass. Welcome to OFX, congratulations to project #11!
Larry: Project #11 in our 3rd year is not an alarming growth, should be OK! (compared to CNCF!)
Larry: will need to assemble a TSC, when to hold meetings, meeting list, Slack channel. Pierre: talked to John and David. Need to vote on the OpenFX side as well. John: will follow up with Pierre for next steps for infrastructure. Will make sure it’s on the agenda for the next Governing Board meeting.
Open Discussion for next TAC agendas
Should project reviews be shortened and doubled up to avoid taking up all upcoming meetings with project reviews
Sean: should project reviews be done in a single day, like a BoF at SIGGRAPH?
Doug: projects could record a presentation so the time during the TAC could just be discussion. Eric: need to trust that people do the homework. John: other projects have dedicated time, will schedule over the year when project reviews happen so it’s not too “lumpy” and everyone has an opportunity. Also make sure it’s “timeboxed”. Larry: how does CNCF do it? John: next size up of LF foundations that has 20-30 project reviews, try to focus on high level, key accomplishments, project lifecycle. Larry: is what OCIO did last time overkill? John: for other foundations, their TACs have less of a technical review, more of a project health review. A way for the project to voice concerns and opportunities back to the TAC, ask for support. Carol: OCIO might have approached presentation differently if we had been told 30 minutes instead of a full hour, so we need to make sure that projects know the timeframe. Larry: also first round of reviews for a project that’s been in ASWF since the foundation began may have had more “backlog” to communicate, and a lot has happened in OCIO. So may not be representative. John: very detailed presentation may be more appropriate for Open Source Days. Carol: if we stick to 30 minutes, with 20 minutes for presentation and 10 minutes for discussion, but only have a TAC meeting. John: historically, this group has had the time to do the “around the room”, with 11 projects + 5 WGs, there may not be that time anymore. Structure / format of TAC meetings may need to be adjusted.
Erik: project update may be different if it’s the first time in a year, vs 2 minute update every 2 weeks. John: projects can have just specific announcements. Gordon: Autodesk have joined R&A WG last week, proposal to contribute core of RV to ASWF, will be working with that group to see how it fits with studio contributions, so anyone interested, come join the discussion. John: would be good to rotate R&A sooner rather than later. At some point R&A may need to become a real project. Larry: can be horse trading on the schedule, projects that are scheduled for later who want to come earlier, that can be arranged. Michelle: yes, list as put together arbitrarily, could be more of a “sign up” type list.
Cary: could have a kind of “report card”, not something as formal as CII badge, but a standard set of questions, so that things that are OK (for instance) don’t need to be discussed. Carol: fine line between recording a presentation, vs dedicated time in a meeting.
OpenEXR was first admitted, more recent projects could come later, MaterialX just got their repo moved over. Let’s look at the list and maybe juggle the order.
John: would OpenEXR be up to go next? Cary: wouldn’t mind stalling for time a bit, 2 halves to the project update / review, one is the discussion within the TSC, take a step back from daily churn, then report to the TAC. But would be OK for June 15th. Cary: yes, that’s fine.
Larry: VES Tech committee presenting on May 4th, than May 18th is strategy session with David. So not until June. Can take discussion offline. Christina: need to tell Jonathan Stone that MaterialX is pushing back. John: can communicate with project leads.