FFmpeg Needs to Change  ## Metadata - Author: [[Theo - t3․gg]] - Full Title: FFmpeg Needs to Change - Category: #articles - Document Note: --- tags: [Open source governance, Developer relations, Video encoding] author: [MISSING] source: YouTube reference: [https://www.youtube.com/watch?v=3VkcAvqV24c] date created: {{date:YYYY-MM-DD}} {{time:HH:mm}} --- # 🏷️ Title: FFmpeg needs to change ## 🧰 Action Inbox - [H1] Publish encoding-constraints primer — context/https://www.youtube.com/watch?v=3VkcAvqV24c&t=0 #articleseed - [H3] Draft comms policy template for FOSS orgs — context/https://www.youtube.com/watch?v=3VkcAvqV24c&t=112 #articleseed #share - [H5] Ship GPU-encoding explainer with NVENC notes — context/https://www.youtube.com/watch?v=3VkcAvqV24c&t=554 #articleseed #xref - [H7] Track VLC 4 status; list alternatives — context/https://www.youtube.com/watch?v=3VkcAvqV24c&t=930 #xref #addendum - [H8] Diagram playback stack: FFmpeg/VLC/MPV/IINA — context/https://www.youtube.com/watch?v=3VkcAvqV24c&t=832 #articleseed - [H6] Create “send patches” moderation macro — context/https://www.youtube.com/watch?v=3VkcAvqV24c&t=680 #share - [H9] Write guide for testing VLC nightlies — context/https://www.youtube.com/watch?v=3VkcAvqV24c&t=987 #articleseed - [H2] Outline donor-relations best practices — context/https://www.youtube.com/watch?v=3VkcAvqV24c&t=20 #articleseed #xref ## 🧭 Table of Contents - [📝 Summary](#-summary) - [📖 Main Content](#-main-content) - [🗒️ Notes](#️-notes) - [📚 References](#-references) - [🗣️ Paraphrase](#️-paraphrase) - [📰 Newsletter](#-newsletter) - [✅ Action Items](#-action-items) - [🧪 Hypotheses](#-hypotheses) - [📊 Methodology](#-methodology) - [❓ Questions](#-questions) - [🔮 Future Research](#-future-research) - [💡 Implications](#-implications) - [🧠 Learning Styles Assessment](#-learning-styles-assessment) - [📊 Accessibility Assessment](#-accessibility-assessment) - [🌍 DEIB Assessment](#-deib-assessment) - [✅ Logical Validity](#-logical-validity) - [🛡️ DARVO Analysis](#-darvo-analysis) - [🎯 Conclusion](#-conclusion) ### 🔗 References - Include links cited in or relevant to the article. ## 📝 Summary - The creator praises FFmpeg’s technical impact and donates funds. He criticizes the project’s social media conduct. See [H1], [H2], [H3]. - He defends Sebastian Alton from perceived harassment by the FFmpeg account. See [H4]. - He explains GPU encoder realities and codec tradeoffs. See [H5]. - He argues VLC’s delays harm modern video workflows and users. See [H7], [H8], [H9]. - Bottom line: great software needs respectful, professional communications and clear roadmaps. ## 📖 Main Content - Encoding remains complex despite mature tooling. Constraints stack across formats and platforms. See [H1]. - FFmpeg is foundational. Many apps embed or wrap it directly. See [H2], [H8]. - Public comms shape contributor health. Aggressive replies deter non-code contributors. See [H3], [H6]. - GPU “hardware encode” uses dedicated blocks like NVENC. Parallelism has limits. See [H5]. - VLC’s release cadence affects codec availability and playback. Developers seek alternatives. See [H7], [H8], [H9]. - Governance and moderation norms matter for OSS sustainability. Respectful tone attracts funding and help. ### 🗒️ Notes #### ⏱️ [00:00](https://www.youtube.com/watch?v=3VkcAvqV24c&t=0) Context and stakes ### [H1] Video encoding remains brutally complex > [!quote] [FFmpeg needs to change](https://www.youtube.com/watch?v=3VkcAvqV24c) by an unnamed author via [[YouTube]]. Published [MISSING]. Accessed {{date:YYYY-MM-DD}}. > “Video encoding is really, really hard.” Complexity persists across formats, devices, and latency budgets. [view highlight](https://www.youtube.com/watch?v=3VkcAvqV24c&t=0) #articleseed ID:rw_[MISSING] ### [H2] FFmpeg underpins modern video infrastructure > [!quote] [FFmpeg needs to change](https://www.youtube.com/watch?v=3VkcAvqV24c) > “I cannot possibly put into words just how essential FFmpeg is…” This frames FFmpeg as critical infrastructure. [view highlight](https://www.youtube.com/watch?v=3VkcAvqV24c&t=88) #xref ID:rw_[MISSING] #### ⏱️ [01:52](https://www.youtube.com/watch?v=3VkcAvqV24c&t=112) Comms and conduct ### [H3] FFmpeg account allegedly harasses community members > [!quote] [FFmpeg needs to change](https://www.youtube.com/watch?v=3VkcAvqV24c) > “I have been harassed by the FFmpeg Twitter account consistently…” Project accounts should avoid personal attacks. [view highlight](https://www.youtube.com/watch?v=3VkcAvqV24c&t=112) #articleseed #share ID:rw_[MISSING] ### [H4] Sebastian Alton receives undue public attacks > [!quote] [FFmpeg needs to change](https://www.youtube.com/watch?v=3VkcAvqV24c) > “why the [ __ ] is FFmpeg beefing with him?” Defense focuses on expertise and contributions. [view highlight](https://www.youtube.com/watch?v=3VkcAvqV24c&t=126) #xref ID:rw_[MISSING] #### ⏱️ [09:14](https://www.youtube.com/watch?v=3VkcAvqV24c&t=554) Hardware realities ### [H5] GPU encoders use dedicated hardware blocks > [!quote] [FFmpeg needs to change](https://www.youtube.com/watch?v=3VkcAvqV24c) > “GPU encoders have a dedicated blackbox chip for video encoding.” Expect NVENC/AMF/QuickSync behavior, not generic GPGPU. [view highlight](https://www.youtube.com/watch?v=3VkcAvqV24c&t=554) #articleseed #xref ID:rw_[MISSING] #### ⏱️ [11:24](https://www.youtube.com/watch?v=3VkcAvqV24c&t=684) Contribution norms ### [H6] “Talk is cheap; send patches” escalates conflicts > [!quote] [FFmpeg needs to change](https://www.youtube.com/watch?v=3VkcAvqV24c) > “Talk is cheap. Send patches.” Policy should value code, docs, funding, and advocacy. [view highlight](https://www.youtube.com/watch?v=3VkcAvqV24c&t=684) #share ID:rw_[MISSING] #### ⏱️ [15:34](https://www.youtube.com/watch?v=3VkcAvqV24c&t=934) Release management ### [H7] VLC 4 delays block modern codec access > [!quote] [FFmpeg needs to change](https://www.youtube.com/watch?v=3VkcAvqV24c) > “VC 4 is coming this year with a new interface…” Delays hinder HEVC/ProRes/HDR workflows for many users. [view highlight](https://www.youtube.com/watch?v=3VkcAvqV24c&t=934) #xref #addendum ID:rw_[MISSING] #### ⏱️ [13:52](https://www.youtube.com/watch?v=3VkcAvqV24c&t=832) Player stacks ### [H8] VLC and IINA both rely on FFmpeg > [!quote] [FFmpeg needs to change](https://www.youtube.com/watch?v=3VkcAvqV24c) > “Both are powered by FFmpeg.” Playback differences stem from wrappers, settings, and versions. [view highlight](https://www.youtube.com/watch?v=3VkcAvqV24c&t=832) #articleseed ID:rw_[MISSING] #### ⏱️ [16:31](https://www.youtube.com/watch?v=3VkcAvqV24c&t=991) Nightlies friction ### [H9] VLC nightlies feel hard to obtain and run > [!quote] [FFmpeg needs to change](https://www.youtube.com/watch?v=3VkcAvqV24c) > “they're impossible to get running.” Friction reduces testing and community feedback. [view highlight](https://www.youtube.com/watch?v=3VkcAvqV24c&t=991) #articleseed ID:rw_[MISSING] ## 📚 References - FFmpeg project. https://ffmpeg.org - VideoLAN VLC. https://www.videolan.org/vlc/ - MPV player. https://mpv.io/ - IINA player (macOS). https://iina.io/ - NVIDIA NVENC overview. https://developer.nvidia.com/nvidia-video-codec-sdk - WorkOS platform. https://workos.com - Primary source video. https://www.youtube.com/watch?v=3VkcAvqV24c ### 🔖 Zettel Candidates - **Video encoding remains brutally complex.** Use to frame latency and bitrate tradeoffs. Source-link: [H1]. - **FFmpeg underpins modern video infrastructure.** Note dependency chains in apps. Source-link: [H2]. - **Comms conduct affects OSS sustainability.** Track norms and contributor churn. Source-link: [H3]. - **GPU encoders use dedicated hardware blocks.** Explain NVENC/QuickSync/AMF limits. Source-link: [H5]. - **“Send patches” culture deters non-code help.** Broaden contribution definitions. Source-link: [H6]. - **VLC delays limit codec availability.** Recommend alternatives and build paths. Source-link: [H7]. - **Wrappers change playback experiences.** Document MPV/IINA/VLC variations. Source-link: [H8]. - **Nightly build friction reduces testing.** Improve docs and signing pipelines. Source-link: [H9]. ## 🗣️ Paraphrase - The speaker admires FFmpeg’s technical excellence but criticizes its social media conduct. H - Summary: FFmpeg is one of the most incredible pieces of software ever built, which is why the way they run their twitter & comms is such a problem... Thank you WorkOS for sponsoring! Check them out at: https://soydev.link/workos Want to sponsor a video? Learn more here: https://soydev.link/sponsor-me Check out my Twitch, Twitter, Discord more at https://t3.gg S/O Ph4se0n3 for the awesome edit 🙏 - URL: https://www.youtube.com/watch?v=3VkcAvqV24c Video encoding is really, really hard. It's a problem that in one hand is solved, but on the other, we're fighting every single day. I know this from a lot of my own history in video. From how I run the channel with custom tooling to things I built like ping to my time at Twitch trying to make video encoding work at the scale of that business. It's crazy. And I don't think we give enough appreciation to the people who put the work in to make video as stable as it is today. is genuinely a miracle that you can watch me live on this video without any crazy issues, even on a not that great internet connection. This is a series of miracles stacked on top of each other by really, really smart people that have resulted in the web working and video working in ways that I never would have imagined when I was a kid. This is best summarized by everyone's favorite XKCD comic, the all modern digital infra with a project some random person in Nebraska has been thanklessly maintaining since 2003 holding up everything. There's no one thing that fits the role of this block better than FFmpeg. FFmpeg is an all-in-one tool set for doing video encoding to and from basically any format you can imagine. And I have spent so much time making FFmpeg do unbelievable things. I genuinely cannot imagine where my career or my channel would be without FFmpeg powering all of it. It's such an incredibly good tool. I cannot say good enough things about it. That's why I personally donated $5,000 to supporting the development of FFmpeg. It's an incredibly important piece of software and I cannot possibly put into words just how essential FFmpeg is to the success of video and arguably the web as a whole. Which is why I am so disappointed by the state of their Twitter and their coms in general. I hate that I have to make this video because I don't want to talk about FFmpeg in a negative light, but I feel like I have to because I'm no longer the only one that is dealing with this problem. I have been harassed by the FFmpeg Twitter account consistently for quite a bit now. And if it was just me, whatever. I'm the JS Soy boy. You can make fun of me all you want, even though I'm contributing and actively supporting the development of your project. When you go after other people, in particular, you go after Sebastian Alton, who is one of the best rendering people in the history of software. There is no person that I know of that has made more impact on how rendering works in game engines in a very long time. At the very least, nobody who shares their understanding publicly as well as he does. I and many others have learned a lot from Sebastian's work on rendering. So, why the [ \_\_ ] is FFmpeg beefing with him? Ah, this is going to be an interesting video. It's not going to be as technical as most of my stuff is, even though, believe me, I want to go in the weeds of FFmpeg. I really do. But the focus today is going to be a lot more so on how communication works at scale and how important it is to pick your fights properly and not just optimize for clout. I am genuinely disappointed in how the FFmpeg account runs its public presence. It feels like they had some good dunks early and now they're just riding the wave. I have feelings. I also have to cover that donation somehow. So, we're going to do a quick break for today's sponsor and then we'll dive in. I have a quick question. Have you ever seen this screen before? Statistically speaking, you probably have. It's the signin for cursor. This is on authenticator.cursor.sh. It's the way you off for cursor. As you probably know, cursor is a very valuable company. So, they probably rolled their O themselves, right? Cuz all big companies roll their own off. Now, let's take a look at the network tab. Huh? That's not their own Oth. That's coming from work OS, huh? Turns out a lot of companies aren't rolling their own O anymore. Everyone from OpenAI to Verscell to Cursor, Plaid, Loom, Socket, FAL, Carta, Source Graph, all of these companies have chosen to not roll their own off. Even though in the era of AI, it's so easy to roll these types of things, right? It is until you have to integrate with another business. Work OS isn't trying to be yet another O library. They're trying to be an O platform. All of the pieces you need to do enterprise O correctly. Whether you're protecting your sign-in page from bots or you're trying to make it easy for a big enterprise to sign a big contract with your company, work OS is everything you need to get it right. It's still kind of crazy to me that OKIT's just out here for free and open source and it's being used directly by our friends over at Cursor for their homepage signin. Like, how cool is that? The same way that billionaires and US devs are using the same MacBook and iPhone, billionaires and us small devs are using the same offplatform and signin. And that's great because it makes your platform compatible with these other giant companies. And in a world where big businesses are adopting tools by two to three people, there's never been a better time to try and set this up right from the start. You might notice the T3 chat signin doesn't look like this. That's cuz we screwed up and rolled our own off. And it has been the bane of my existence since we did it. If I could roll back time and start with work OS, you know I would. and we're actively in the process of moving over now. If you're ready to start selling to enterprise or you're just tired of dealing with Oth, look no further than works and try them out now at soyb.link/workos. So, uh, fun fact, I actually filmed a video like this a while ago as my response to the beef I got into with fmpic a few months ago and I decided to not publish it because it didn't feel necessary for me to go after an open source project like this as the main victim of it. But now that I see them doing this to other people and continuing to cause problems, I feel like it deserves a little more attention on it because I am 100% on Sebastian's side here. It's a little bit absurd he's getting called out the way he is. The discussion started all the way over here with people talking about Google Deep Mind and this new AI game world exploration hallucination thing Google's doing where I can make up a world that you can play. Mike 64 here posted about the challenges here. Just so you guys know, the bottleneck for getting data out of Minecraft is literally ffmpeg. And I can guarantee this model is both slower and worse than actually being in game. The timer speed can be increased. Frame capture can happen in game at scaled rates. And the thing that will make the system come to a crawl is ffmpeg, and there's nothing you can do about it. You can't encode video at 1200 fps with current technology unless you're recording at like 240p or you have a 20 pabyte SSD somewhere and decide to use AVI. Yes, to be very clear, this is not a post saying FFmpeg is bad. This is a post saying that video encoding is so hard that even the industry standard battle tested super performant really good video encoder ffmpeg cannot solve this problem. Turning a ton of hallucinated PGs into a video stream someone can consume on the other side is a complex problem to solve with compute in general. And that is one of those things that is making these like hallucinated video things so hard to make. It's also a big part of why game streaming has struggled to take off because the cost of encoding the video at a high enough quality and low enough latency and also with enough compression to send it over the internet for you to play a game that's running remotely somewhere else is genuinely really really complex. And this became a discussion across other people. Jabrim posted here, I started looking at FFmpig's code the other day and it was atrocious. I find it hilarious. They speak about their code as some of the most optimized on the planet. Then you hear about the low frame rates that they were able to achieve. Of course, these codecs don't help. They frankly weren't designed for real-time performance. Certainly nothing deterministic either in terms of compression guarantees. This first half, eh, I don't like commenting on others code quality. I've learned my lesson there. I'd rather just talk about the capabilities and functionality. If you think FFmpeg could be faster, I would like you to prove it, not just complain about the source code. This part is very good to call out, though. The codecs are not written in a way where they're optimizing for encoding cost. They're optimizing for both the compression amounts and the decoding cost. How can we make the smallest video file without compromising on quality and more importantly decode it cheap enough that most devices can play back the video without too many problems? I wish this is the focus of Jabrim's post. If Jabrim here was the one that was being dunked on by Fmpeg, I'd be a little more sympathetic to them, but it wasn't. They went after Sebastian. Video codecs are such a massive code bloat. Integrating these encoder and decoder libraries in your project is a lot of work and their compile times are massive. Their core compression loops are properly vectorzed, but the rest of the ecosystem is not optimal at all. This is a very correct post. This is not a comment that FFmpeg is bad, that the people building these tools suck or anything along those lines. All this says is that the code bases for these codecs are big enough that integrating them into other things is unpleasant. If you were to embed some of FFmpeg's codecs into your project that isn't a video encoder, you just want to add this feature and you're actually pulling that source code into your codebase and compiling it as part of your builds. Good luck. It's very slow. These code bases are huge and complex and slow to build and they probably have to be. None of this is saying f ofmpeg is bad or sucks. This is commentary on the state of video encoding. It is hard to do these things. And now we see how this was discovered by FFmpeg because Drim here quote tweeted an FFmpeg post talking about how GPU encoding has marketing has all been lies and GPGPU isn't suited for video compression. GPU encoders have a dedicated blackbox chip for video encoding. This is a true statement. My GPU that I'm using right now has an Envank encoder on it. Envank is Nvidia encoder. It's a custom chip that they put on my GPU just for video encoding. And I'm stuck buying the highestend Ned Nvidia GPUs because the cheaper ones only have one invank encoder and I need two, one for my stream and one for my camera. It's annoying, but I accept it because doing parallel video encoding is hard. They're not doing much parallelization, but there's honestly not that much to do. It's pretty hard to parallelize video encoding for most modern formats cuz it's like you have an iframe and then compressed frames from there. Do you have to do a first pass where you find all the scene changes to do I frames for and then in parallel generate the P frames from there? It just doesn't [ \_\_ ] make sense. It's so I do not agree with this Jim guy. I fully agree with Sebastian. These are things that suck to integrate as a developer if you're using the source. And f ofmpeg jumps in quick to complain but not quick to contribute. Get log- author equals alton blank. Not necessary at all. And the worst part is I know where all of this comes from. This post was the first time this format went viral. This one was funny for reasons that are a bit of inside baseball for me. I haven't shared this much publicly. Soundwave was my first employee at Ping. Soundwave was diving deep in both FFmpeg and OBS because we were working on making changes to them for the video products that we were building. Soundwave was no longer employed by me when he made this post. So, this was not an employee of mine getting in a fight with FFmpeg. This is an exemp employee of mine getting in a fight with FFmpeg. Save is a very smart person. I trust his gut if he thinks this codebase is a mess. I also trust that there are good reasons for it to be a mess. Regardless, FFmpeg is the best video encoder suite ever made. I would have advised him to not post this if he was my employee at the time and I had any knowledge that he was planning on posting this. And this is where FFmpeg started the Twitter trend that they cannot stop doing of dunking on a small account by saying, "Talk is cheap. Send patches." You clearly don't know what you're talking about if you've never contributed code to our codebase. Aside from the fact that code contributions are far from the only way to contribute to things, there's also documentation. There's also educational materials. There's also financial. There's also like emotional support, community support, building awareness. There's so many other ways to support a project other than writing code for it. Especially a project that holds their bar so high supposedly for who can and can't contribute and act so elitist about how important it is that everything's in CN assembly and anybody who doesn't know those isn't good enough to contribute. They're holding this high bar for who's allowed to contribute and then using that bar to hit people who can't contribute. It makes no sense and it's really toxic. At this point, it was just a funny thing they did back in February of last year, but it stopped being funny pretty quick, and they can't help themselves, but just get into fights constantly. This donation that I made that I mentioned earlier was part of my argument with FFmpeg. My post was deleted because I was trying at the time to reduce the amount of harassment I and my team were getting by hiding it. It didn't work. My post wasn't about FFmpeg. My post was about VC because VC is an absolute shitow right now if you're using any modern video standards. I made a complaint that VC has outofdate codecs that make it nearly impossible to use for modern video standards because I work with a lot of modern video standards. I generate a terabyte of video per stream. I spend a lot of time inside a video and the amount of problems I have with VC is insane. This is just one random demo that I recorded to demonstrate it. And here you'll see a video playing back in VLC. This is a file straight off of my camera. It's me doing my facial reactions, my like react faces for my thumbnails. And you'll see I'm like skimming through it and nothing's happening. It's just frozen. It's also memory leaking. It's using like 8 gigs of RAM here. It's just entirely frozen. And then I ina, which is a player I highly recommend on Mac. You'll see I can skim, I can fast forward, I can save screenshots, I can do everything I need to with INA. Totally fine. So clearly the problem has to be the encoders, right? Like VC uses FFmpeg, ANA must use something else, right? Nope. Both are powered by FFmpeg. Ina's using an MPV core. MPV is a light wrapper around FF Play, which is part of FFmpeg. VC rolled their own everything. They even have some of their own codecs in VC. The problem with VLC is that you cannot change the codecs in it. Everything is baked into it directly, which means if you want something new, you need to wait for an update. The same way if you want an update for Safari or Apple Music on your iPhone, you have to install an OS update. If you want a new codec or playback method in VC, you need to wait for a new version, which would be totally fine if VC was shipping new versions. The problem is that VC 4 has been indefinitely delayed for a disgusting amount of time. This is an article on The Verge. VC 4 is coming this year with a new interface for the popular video player. This would also include all of the codecs that I need for my job. It would include the ability for me to play an iPhone video without it freezing, memory leaking, and getting all of the color wrong. You know, things you would expect a video player to do. These are all of the reasons why like the entire like video piracy community, the whole like anime dub community, all of these people have moved off of VLC like five plus years ago. Anyone who knows what they're talking about with video [ \_\_ ] shits on VC non-stop. There are literally books written about how to find bugs in software that use VC as their first example because it is in such a rough state. VC deserves the hate. Not cuz the devs aren't working hard, but they're not very transparent. I've tried so hard to get a hold of them just to figure out where to throw money to speed things up because I will pay the VC devs to get their [ \_\_ ] done because this article about VLC 4 coming out later this year is from 2021. It has been almost 5 years and we have not gotten an update about when VLC 4 will actually ship. What the [ \_\_ ] If it was a few months late, fine. If it's a year late, annoying, but whatever. 5 years late. What? And the screenshot that they're showing here is from 2019. What? How? How are we here? Seriously, no new video codec, no improvements to video codecs, no change to video that's happened in the last 7 years has made it into VLC. Every improvement FFmpeg has made has been sitting on the VLC 4 branch that they don't even publish nightly for right now. coming this year. I'm gonna So, somebody was posting about how great VLC is. And I did a quote tweet saying, "Can we please stop recommending software that's this dead? It's obnoxious. It's And yes, there are technically 40 nightlys, but they're impossible to get running. I spent like 2 hours trying to get one to build cuz I don't build them. I had to build it myself to run it on Windows." And it was slightly better. I thought it would be more better because when I first tried it, it did feel significantly better. It fixed this issue where the video like couldn't be sc because it's on an up-to-date codec, so it's a lot better, but it still ended up having a lot of other problems. Like video was flickering. It would move up and down two pixels constantly. It just it seemed a lot better when I first tried it and then when I kept trying to use it, it fell apart fast. So I'm back on Aina on Mac and on Windows I deal with MPV and also use media player classic a decent bit. There's a fork HC that's fine. Aa's really nice on Mac though. Regardless VLCs has so many problems for modern formats, modern like trying attempts to play back with subtitles, all of these things. It's just I don't know how much this will include. I just hope you understand. VC makes my life much harder because I work in video and it does not work with real video. So how did we get here? I complained that the codec in VC was egregiously out ofd and didn't support modern things. I wasn't even thinking of the fact that it was using ffmpeg as some of those codecs internally. It just it was so old and out of date that the ffmpeg was a implementation deal that didn't matter. I did not mention ffmpeg anywhere in any of my posts. That did not stop them from coming at me as a huge fan. genuinely kind of [ \_\_ ] insane of them to take a post that was about VLC being out of date and say it is proof that I don't know what I'm talking about and they're excited for me to go contribute the change. If I submitted a patch to VC, how many years would it take to ship? FFmpeg has run incredibly well. I wish the same dedication was put into VLC. V4 has been needed for 5 years. It's becoming vaporware. I stand behind every [ \_\_ ] word I said here. I am correct. To which they replied, "You would never manage to submit a patch even close to replacing FFmpeg in VLC. I don't want to replace FFmpeg. I just want to bump the [ \_\_ ] version. You clearly have no idea how much my history was in FFmpeg. I built a ton of early Twitch info on top of y'all. I'm not talking about my ass. I'm complaining that a valuable piece of software can't even play a [ \_\_ ] video recorded on an iPhone because V4 soon." Again, I am talking about VC. He is inserting himself making it about FFmpeg when it's not about FFmpeg. Reminder, my original post never mentioned FFmpeg. I did not talk about FFmpeg here other than to say how incredibly well-run it is and how important a project it is. I had no beef with FFmpeg, but FFmpeg chose to have beef with me because to FFmpeg or at least the person who runs their Twitter account who I'm not even confident contributes themselves. This all could just be projection on their part. I kind of represent the JavaScript soy boy influencers that they hate so much. They even subweeted me during all of this. It's important to remember that technology influencers can be bad role models. Many encourage people to be fans of technology instead of actual developers of technology. Put hot takes are easy. Writing good code is not. Influencers talk a big game, but forget we were on mailing list and IRC before they were born. [ \_\_ ] I probably ran one of the IRC channels you were [ \_\_ ] posting in. I This is so stupid. This is This was so dumb of them. I still can't believe that like they came after a lifetime supporter of FFmpeg because it's fun to dunk on the TypeScript influencer. The [ \_\_ ] And again, I said that this is because VLC is out of date. This is the problem. Let us know where your FFmpeg patches are. There's a difference between being a user of FFmpeg and a developer of FFmpeg. There's also a difference between being a dev FFmpeg and being the guy who runs their Twitter like an angry Reddit mod, but that doesn't seem to stop you, [ \_\_ ] So, he couldn't decide which of the dunks he wanted to do. So, he did the first one, then followed up with, "Writing scripts for a hugely unprofitable streaming service is not the flex you think it is." Referring to me working at Twitch. Do you know what else is hugely unprofitable? FFmpeg. I don't think the amount of money the business doesn't make is relevant here. And I am far from a supporter of Twitch as a business model. I think that I'm surprised they're still in business if I'm being real. That doesn't mean they haven't done revolutionary things with on the edge global video encoding and decoding. The things that Twitch did to make live video possible are still talked about in like the deep video circles. You would know this if you went to the events about this. If you went to the most recent video land meetup in SF, you would have heard people who worked at Twitch on these things presenting because they do know it very well. And I don't think you should be talking about profitability when you're an open source project when it has nothing to do with this. It's a funny joke, sure, but this is the problem. He cares way more about being funny on Twitter than representing the project in a good light. And I responded accordingly. Going after a company that's contributed money and patches to y'all instead of addressing the point is disappointing. I expected better from you guys. We would love to see a copy of your maintenance contract and or the full-time members of staff working solely on FFmpeg. That [ \_\_ ] goalpost just keeps moving, man. Non-stop. The total of FFmpeg donations over a year barely account for a fifth of a Silicon Valley engineer. Cool. Whatever. Regardless, I care a lot about FFmpeg. I wanted to show that I care about FFmpeg and that this is all [ \_\_ ] stupid and they're coming at me as a supporter. They just it did not seem to register in this fucker's head that I love FFmpeg and I rely on it heavily. He needs to go back to moderating on Reddit because that's the only place where this mindset makes any sense at all. This is a Reddit mod who got an account with followers that got a power [ \_\_ ] high because of the dunk on my old employee that can't stop himself now. So, I donated five grand because the people who actually own the Twitter account, who actually build FFmpeg are great. I care a lot about them. ffmpeg is so important and I'm happy to put five grand in to help make it better. I rely on this software every day. I'll probably donate more money in the near future. I care a lot. You know what? I'll put my money where my mouth is. If the current admin running the FFmpeg Twitter account quits, I'll donate 20 grand. I'll do it. Mark my words. If the guy who runs his account retires, he gives up. he stops doing this [ \_\_ ] [ \_\_ ] and going after people who do not deserve it for no reason, I will immediately send $20,000 to the FFmpeg organization because I am that confident he is causing problems. And by giving even the five grand, I've contributed more to FFmpeg than that [ \_\_ ] ever has. So yeah, let's see what happens. If the FFmpeg team wants $20,000 more dollars to be donated, they can take care of this problem now. I hate doing incentives like this, but I hate [ \_\_ ] like this guy even more. and I'm so tired of him going after not just me but other people now too. And I jumped in on this one for the same reason. I dropped. I miss when the FFmpeg account dunked on people who deserved it. I've had my disagreements with Seb, but he really knows his [ \_\_ ] and I cannot fathom questioning his expertise. FFmpeg replied to me cuz they can't help themselves. They thought they'd get an easy dunk on me. He went down the same road as you with the whole look how important my business or employer is. I therefore must know what I'm talking about thing. Oh, kind of like the thing you do because you're in a non behind the FFmpeg account. We just assume you know what you're talking about. You're the one doing this, not me. We have historical records, educational materials, shipped products, and so much more. Showing that we understand this [ \_\_ ] He tried quote tweeting me mentioning my history of FFmpeg. Talk is cheap. Show the patches you personally wrote. Would love to see whatever the [ \_\_ ] this guy's actually contributed. We're still waiting for your patches. By the way, this was his second reply here because he deleted the first one. Do you even know C? Yeah. Did you guys know the FFmpeg account deletes tweets? Cuz this is how I learned that. Seriously though, what the [ \_\_ ] Why is my Soy JavaScript ass triple ratioing the FFmpeg account? Literally three in a row. That shouldn't be possible. And I I need them to understand whoever the hell else works on FFmpeg that has any access to his account. You need to calm this [ \_\_ ] down. FFmpeg is too important to have its reputation ruined because this guy goes after the wrong person cuz he doesn't recognize their username. God forbid he accidentally targets somebody who's actually a contributor. Do you know how bad that's going to look even after he deletes it? You need to get rid of this guy. It sucks. The only problem I have with FFmpeg is whoever the [ \_\_ ] running this Twitter account. I miss when they were funny. I miss when they dunked on people who deserve to be dunked on. I missed when this wasn't a problem and I didn't have to worry about people just mentioning codecs or talking about video stuff on Twitter. Like, I'm scared of doing educational stuff about video and how it works now because I'm scared if even I get all the details right, there's going to be some dunk about how influencers should stay out of important code. It's just insane to me. I hope this video does something other than start yet another flame war. I hate that I'm doing this. I did not plan to ever talk about it again. When you're going after good faith, hardworking people doing real work and real education in the space because you want free clout on Twitter, [ \_\_ ] off. FFmpeg is too important to be acting like a 12-year-old on Reddit. You need to get over this [ \_\_ ] guys. I got nothing else. Thank you. Until next time. Peace., FFmpeg Needs to Change, Theo - t3․gg, 01k8jy956yfmcd5wd1t52vz6pk.