Hello Logik Friends!
Autodesk’s Jasmin Frenette (@frenetj) is coming on Logik Live this Sunday. He’s a member of the Dev Team, and I wanted to get some questions for him from the community.
Outside of the micro freezes and M1/Apple Silicon migration, which I’m sure we will cover, what would you like to ask about Flame development? Jasmin works directly with the programmers, so no question is too technical or too deep!
Not sure how far-reaching the convo is supposed to be, but can we talk about many legacy Action inefficiencies?
Totally! Can you be more specific?
I cant make the event but Avid Media Central | Production Management (Interplay) integration is high on my list. Got it for Adobe now so Flame is the last one in our pipeline.
Mike
I’m sure Jasmin will be happy to talk about features and feature requests, but this can also be an excellent opportunity to get some insight as to the underlying technologies and challenges that go into developing flame, too. For example, last year Jasmin explained why the transition to Apple Silicon was so challenging, and the answer was definitely a surprise to me. There’s a lot of work being done under the hood right now, so some insight into the bow and why of that could also be very cool to hear more about.
delete the word “Action”
Action is where the most significant inefficiencies live.
Integrate the ML time warp into the software.
How many people work on Flame full time? How is the Dev team setup?
Do devs at ADSK only work on one software at a time or do they also work on other software packages like Maya at the same time e.t.c?
Is there a bunch of legacy code around that is impossible to work on and to fix/improve because its so old? (paint, text, particles) Like whats the technical issue with those modules.
Do most people in the team decide they want to work on flame or is it more of a “you get assinged to this weird software you have never heard of” situation.
Do you have very very old Discreet era builds /source of flame 1.0 or something around? or How about do you have old legacy systems around that still work ?
Can we get some storytime about that custom ILM Inferno (sabre department) ?
This sounds intriguing.
Training new people on flame I realise just how much stuff is inconsistent. There are still a lot of siloes which made sense in desktop land but not in batch. And certainly when comparing to after effects or nuke they seem prehistoric. Which they are. We could list many things but that seems redundant. I’m sure a result of the siloes there are code inefficiencies all over the place. I’d be very interested to hear what Autodesk propose for the future.
We have to make it easier to pick up. Easier to manage a job. I’m open to new workflows.
I can not love or upvote these enough.
We get that question so often and I still don’t understand why.
Would like to ask about Flame on Windows but I already know the answer sadly 
When I ask another Flame Artist about a project she just completed, I’m fascinated by the scope and scale and almost always ask her how much time something took and how many people touched it. It provides a great amount of context and humanity to a project and that gives me a great deal of insight. I’m fascinated by the culture of developers and daydreaming a little about what your teams look like and how they are shaped and any similar tidbits of humanity really helps me put all what you guys do into perspective.
While I’m sure that there are important financial reasons why there isn’t a windows version of Flame, I think it’d be fascinating to understand the technical challenges to porting Flame over to windows. Perhaps we can ask Jasmin for his thoughts on that.
I think both Linux and macOS are in extension versions of Unix, like the Irix on which early versions of Flame ran. So from the standpoint of implementing an application on similar operating systems would be viable yet it would be a whole different ballgame to get it to run on Windows, or any other non-Unix flavored OS.
totally, its such a mystery and probably thats business relevant so adsk wont tell, but its very interesting if its like 10 or 2 .
cant hurt to ask, dont expect much in terms of answeres 
But the amount is just a small variable that doesn’t mean much. What if it was 2 but they have been working on Flame since the beginning and know the industry, the code base, the user base, etc. instead of 10 that are freshly out of school, have no mentor and must navigate millions of new lines of code. In that situation you may prefer having only 2, but it might also depend on the kind of job required, the talent of each individuals, etc. Even if we would tell you how many people are in the team how does that tell you if it’s small, medium, big, etc. In relation to what?
I was going to post a different question about why I can’t copy bfx directly to my desktop reels anymore, but this strikes me as a slightly more interesting topic.
I take your point that its only one variable in how Flame is developed, but if you’re frequently and repeatedly asked this question then clearly there is something else to it.
I won’t presume to speak for others, but my personal feeling about it is that Flame is a drop in the bucket of ADSK’s business.
My sense of the development historically is that after the anniversary edition there was a significant amount of growing pain which was to be expected. As a userbase I would say that we weathered that change fairly well given the situation. Most of the kinks seemed to get ironed out by about '16/'17 and we had a relatively stable software with some great new features added for the next several years. More recently that hasn’t been the case. We’ve been receiving new features, but the stability of the software has frankly taken a nose-dive the last few years. Micro-freezes, GUI corruption, instability within the motion vector module, this new issue with timeline keyframes in 2023 that @creonovo and @dlevine just pointed out the other day, the resolution and aspect ratio problems that @ALan brought up recently(and that really burned me on a job I was doing last week). I could go on listing things for quite awhile and a bunch of them are showstoppers for many of us.
I think its completely within bounds for us to be inquiring about the level of investment that ADSK is putting into the software given that our livelihoods are completely wrapped up in the program’s success. If Autodesk is slowly letting Flame wither on the vine then what will happen to the userbase going forward. Since we don’t have any direct insight to ADSK’s level of investment, knowing how many people are on the development team is an index of that investment from which we can build a bigger picture about how much long term job security there is in being a Flame artist. Like you said only one variable, but one which when combined with observations about the state of the program more generally is helpful in giving an indication of whether I just need to jump ship and start using Nuke fulltime.
I can understand why you wouldn’t want to divulge that infomation given what I just said, but if you’re confused about why people are asking thats my take on it.
Where that might make sense is to compare against other major apps in the space, such as Nuke, DaVinci Resolve, and some niche players like SGO Mistika.
For me the relevancy is mostly the capacity to have enough resources to do all of:
Those are constant tradeoffs for dev teams. As much as Apple Silicon is technically fascinating and could move some aspects of the industry forward, it has been a huge tech tax on the industry at the cost of other innovation. So if you’re not a Mac fan boy, we paid a definite price for that.
Understanding how much spare capacity there is to absorb some of these hits is material, and a fair comparison with other platforms, where we could make informed decisions on which platforms are a better fit/bet long-term.
I don’t think the total head count matters as much as the sense of how constrained it is, and how balanced across these aspects it is.
(full disclosure, I spent a prior career in tech as a software architect and program manager, so I understand the nuances)
Of course also understand that this is not necessarily information that is often disclosed, and shouldn’t really. Why executive never speak in absolute numbers, but are more comfortable with percentages.