Podcasts about HTML5

An open and shut case, the victim raped the man's wife and then the man shot him. But is it legally murder? Nick and Roger take a look at 1959's Anatomy of a Murder. Your browser isn't showing you an HTML5 audio player. Download

Liftoff finally went public this week — at a valuation that tells you exactly what the public market thinks mobile ad networks are worth. That's just one of four stories this week that genuinely matter if you run UA.Matej Lančarič flies solo for the breaking news segment, ranked from biggest to most practical. Liftoff listed on Nasdaq as LFT after a second attempt, raising $437M at a $3.83B valuation — a 25% haircut from the $5B it wanted in January, and below the private valuation General Atlantic paid in 2025. A Niko Partners report buried a number most Western publishers still aren't modeling: minigames are now almost 20% of mobile game spending in China. Akin launched AMF Capital with Makers Fund, opening with a $28M UA financing facility for Birhack. And the throughline of the week — mobile has officially shifted from core-first to event-first, with Monopoly Go's Simpsons crossover, Rovio's own admission, and Supercell's MoCo reboot all pointing the same direction.The bar keeps moving up. The industry is consolidating around scale, capital, and live-ops.━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━⏱️ TIMESTAMPS00:00 Supercell reboots MoCo's live-ops00:30 Liftoff goes public at $3.83B — the IPO breakdown03:00 China minigames are now 20% of mobile spend05:30 AMF Capital launches with a $28M UA financing deal07:00 Mobile shifts from core-first to event-first09:00 What event-first actually means for your UA

We've previously talked about Knives Out and Glass Onion, and now Roger and Nick return for the third Benoit Blanc film, 2025's Wake Up Dead Man. Your browser isn't showing you an HTML5 audio player. Download

The dame is lovely, and very dead. But that's just another day in the big city as Nick and Roger take a look at 1944's Laura. Your browser isn't showing you an HTML5 audio player. Download

A forgotten classic, overrated, or (as usual) something in between? Roger and Nick go from surgery to jet car to concert stage in 1984's The Adventures of Buckaroo Banzai Across the 8th Dimension. Your browser isn't showing you an HTML5 audio player. Download

Nick and Roger take another dip into the barrel of Hammer with Christopher Lee in Wheatley's terribly serious tale of satanism, 1968's The Devil Rides Out. Your browser isn't showing you an HTML5 audio player. Download

Roger and Nick approach a horror classic from different angles as we take a look at 1980's The Shining. Your browser isn't showing you an HTML5 audio player. Download

Nick and Roger take a look at famed cinematic turkey Ishtar (1987), and at another film inspired by the Hope/Crosby Road To series, Spies Like Us (1985). Your browser isn't showing you an HTML5 audio player. Download

Roger and Nick blend cinematic apocalypse with cinematic disaster as we go back to what was at the time the most expensive film ever made, 1995's Waterworld. Your browser isn't showing you an HTML5 audio player. Download

In this episode of the Shift AI Podcast, Scott Roberts, CISO at UiPath, joins host Boaz Ashkenazy for a deep dive into how agentic AI is reshaping enterprise security and automation—both for customers and inside UiPath itself.Scott shares his 25-year security journey spanning Microsoft's early Security Response Center days (including the era that produced Patch Tuesday and the Security Development Lifecycle), product security work across Windows and Xbox, time at AWS, and leadership roles at Google where he helped build the Android Security Assurance and Pixel Security teams and the Android Monthly Security Update process. He also discusses his work in security standards across IPsec, HTML5 encrypted media, GSMA device security, and most recently, contributions to emerging agentic AI security standards.The conversation then explores UiPath's evolution from traditional RPA into a unified platform that combines deterministic automation with agentic workflows. Scott walks through a real-world healthcare billing example where agentic automation increased deduplication accuracy dramatically by handling complex, variable inputs that classic RPA struggled with—while still keeping humans in the loop and feeding outcomes back into the system to improve over time.Boaz and Scott go deep on what's changed for CISOs in the post-LLM world: the need for guardrails, identity and entitlements for AI agents, and the challenge of end users copying sensitive information into consumer AI tools. Scott explains UiPath's approach: enable adoption while using nudges and policy controls to redirect sensitive workflows into enterprise-safe environments rather than relying solely on blocks.The episode closes with an eye-opening look at UiPath's internal “agentic threat analyst” system—an orchestration of 60+ agents that can investigate SIEM alerts end-to-end, generate structured incident writeups, and compress hours of analyst work into roughly a minute and a half. Scott's future-looking takeaway: as AI models evolve beyond “read-only” into potentially “read-write” systems that can update their foundational knowledge, the acceleration could be truly mind-blowing.This episode is essential listening for security leaders, enterprise operators, and automation teams trying to understand how agentic systems change not just productivity, but the entire security operating model.Chapters[00:01] Scott's Security Journey: Microsoft, Google, Coinbase, UiPath[01:33] Security Standards Work: From IPsec to Agentic AI Standards[04:08] What UiPath Does: Process Orchestration, RPA, and Enterprise Automation[06:28] RPA vs Agentic Automation: A Healthcare Billing Deduplication Example[09:17] The Agentic Stack: Canvas, Guardrails, and the AI Trust Layer[10:31] How LLMs Change Security: Data Controls, Access, and Governance[12:14] Internal Adoption at UiPath: AI Tooling by Persona (Legal, Finance, Engineering)[13:13] Code Velocity and Security: Agents Generating Code, Agents Verifying It[15:53] Two AI Security Worlds: Orchestration Platforms vs End-User Chat Interfaces[17:11] Securing End Users: Enterprise LLMs, Nudges, and Browser-Based Controls[19:07] Sovereign AI and Data Boundaries: Keeping Data in the Right Region[21:00] Over-Permissioning Meets Agents: Why AI Makes Old Problems Obvious Fast[22:21] The Next Wave: AI Transforming the Entire SDLC End-to-End[24:53] Security Pitfalls in Agentic SDLC: Misaligned Incentives and Permissions[26:02] UiPath's Agentic Threat Analyst: 60+ Agents, SIEM to Writeup Automation[30:07] What Changes for Humans: Faster “Time to Truth” and Higher-Leverage Work[32:09] Two-Word Future: “Mind Blowing” and Read/Write ModelsConnect with Scott RobertsLinkedIn: https://www.linkedin.com/in/scottroberts6/Connect with Boaz AshkenazyLinkedIn: https://www.linkedin.com/in/boazashkenazy/Email: info@shiftai.fm

Nick and Roger conclude our look into 1950s visions of nuclear apocalypse (for now) with World Without End (1956), Panic in Year Zero! (1962, but it feels very 1950s so we kept it in) and On the Beach (1959). Your browser isn't showing you an HTML5 audio player. Download

In today's podcast episode, we're diving into the world of Poki Games—one of the most popular platforms for playing free online games directly in your browser. If you've ever wanted instant entertainment without downloads, sign-ups, or complicated setups, this episode is for you.Poki has built its reputation on one simple promise: play instantly, anywhere. Using modern HTML5 technology, Pokigame.my lets players jump straight into games with a single click. No apps. No Flash. No waiting. Whether you're on a PC, tablet, Chromebook, or smartphone, the experience stays smooth and accessible.What makes Poki stand out in 2026 is its focus on quality over clutter. Instead of overwhelming users with thousands of low-effort titles, Poki curates a clean library of engaging games—from endless runners like Subway Surfers to action-packed favorites such as Stickman Hook and Venge.io. It's casual gaming done right.In this episode, we also talk about why Poki is especially popular with students and casual players. The platform loads fast, runs well on low-end devices, and works seamlessly even during short breaks. It's entertainment without friction—something many modern gaming apps struggle to deliver.Another key highlight is safety. Poki maintains a family-friendly environment with minimal ads, making it a trusted choice for younger players and parents alike.If you're curious to explore Poki yourself, check out https://pokigame.my to discover browser-based games you can play instantly—no downloads required.Whether you're a longtime fan or completely new to online browser gaming, Poki Games proves that sometimes, the simplest way to play is still the best.

Roger and Nick haul on our lead BVDs and revisit several 1950s visions of nuclear apocalypse and its aftermath. This episode, Five](https://www.imdb.com/title/tt0043539) (1951) and The World, the Flesh and the Devil (1959). Your browser isn't showing you an HTML5 audio player. Download

Nick and Roger aren't going to face 1984's The Ice Pirates sober. Your browser isn't showing you an HTML5 audio player. Download

Is it a bird? Is it a plane? Is it a lovable bumbler? Roger and Nick visit one of the films that turned Disney back towards animation, 1981's Condorman. Your browser isn't showing you an HTML5 audio player. Download

Nick and Roger sew our limbs back on for 1993's The Nightmare Before Christmas. Your browser isn't showing you an HTML5 audio player. Download

CrazyGames is a leading browser-based gaming platform that has quickly become a favorite destination for players who enjoy free online games without downloads or installations. Designed for instant access and smooth performance, CrazyGames offers a modern gaming experience powered by advanced HTML5 and WebGL technology, allowing games to run directly in web browsers on both desktop and mobile devices. This approach makes the platform highly accessible for casual gamers, students, and anyone looking for quick entertainment at any time.One of the biggest strengths of https://crazygame.my/ is its diverse and constantly expanding game library. Players can explore thousands of high-quality titles across popular genres such as action, shooting, racing, puzzle, simulation, strategy, multiplayer, and idle games. From fast-paced shooters like Shell Shockers to relaxing puzzle games and competitive multiplayer experiences, CrazyGame.my caters to a wide range of gaming preferences and age groups.CrazyGames is also known for its clean, user-friendly interface. Games are well-organized by category, popularity, and release date, making it easy for users to discover trending titles or hidden gems. Each game page includes helpful descriptions, controls, and recommendations, enhancing the overall user experience and encouraging longer play sessions.Another key factor behind CrazyGames' popularity is its strong focus on performance and quality. The platform collaborates with independent game developers worldwide, showcasing innovative and creative browser games that rival downloadable titles in both graphics and gameplay depth. Frequent updates ensure fresh content and improved performance across devices.Overall, CrazyGames stands out as a reliable and engaging hub for free online gaming. With instant play, a massive game selection, cross-platform compatibility, and a polished interface, it continues to set the standard for modern browser gaming and remains a top choice for players seeking high-quality entertainment online.

Pour l'épisode #329 je recevais Frédéric Barthelet. On en débrief avec Nicolas.🎙️ Soutenez le podcast If This Then Dev ! 🎙️ Chaque contribution aide à maintenir et améliorer nos épisodes. Cliquez ici pour nous soutenir sur Tipeee 🙏Archives | Site | Boutique | TikTok | Discord | Twitter | LinkedIn | Instagram | Youtube | Twitch | Job Board |Hébergé par Audiomeans. Visitez audiomeans.fr/politique-de-confidentialite pour plus d'informations.

Roger and Nick consider another candidate for "the first Talkie", the first feature-length all-talking film: 1928's Lights of New York. Your browser isn't showing you an HTML5 audio player. Download

We're celebrating the new design of our website. Yeah, up until recently, the general lifespan of websites was between two and three years but now they're saying it's fallen to one year. Now Richard, we've had our website since 2006.我们正在庆祝我们网站的新设计。是的，直到最近，网站的一般寿命是两到三年，但现在人们说已经缩短到一年了。Richard，我们的网站自 2006 年以来就一直存在。Wow! Wow! Yeah, so we must be doing something right to be still going. Yes, there's been a few changes, mainly under the hood as it were. What does that mean? Well, things that the users wouldn't really recognise, things like HTML5 is the new technology that's taken over and also the biggest thing recently is HTTPS.哇！哇！是啊，我们能坚持这么久，说明我们确实做对了一些事情。是的，期间有一些改动，主要是在“引擎盖下面”。那是什么意思？嗯，就是用户不太察觉的东西，比如 HTML5 已经成为新主流技术，而最近最大的更新就是 HTTPS。What do you mean by that? Well, most people will see the little green padlock at the top of the browser bar, which means that the website is safe, basically. OK. Now, for our new design, we've concentrated on a few major areas, haven't we? And I suppose number one has to be the overall look, the overall design.你是什么意思？嗯，大多数人会在浏览器地址栏看到一个小小的绿色锁头，那基本上意味着这个网站是安全的。好的。现在，对于这次新设计，我们专注于几个主要方面，是吧？我想第一点就是整体外观，整体设计。You were mainly responsible for that, Jackie. What were you looking for here? Well, I suppose I wanted a fresher, more modern look and for me the starting point was I wanted an image of England or Britain and we didn't want the typical London look, you know, the double-decker buses and all that kind of stuff. But you wanted something green.Jackie，这部分主要是你负责的。你当时想要呈现什么呢？嗯，我想要更清新、更现代的外观。对我来说，起点是我想呈现英国的形象，但我们不想要典型的伦敦风，比如双层巴士那些东西。而你想要绿色的感觉。Well, I wanted something to represent England's green and pleasant land, so the countryside and I think it's a reflection of our culture and it says something about us. Yes, it's not just the colours though, is it? It's also the navigation and things like that. People are often saying how user-friendly our site is and I really wanted us to stay with that.嗯，我想呈现“英国绿野和美丽乡村”的形象，我觉得那能反映我们的文化，也能代表我们自己。是的，不过不仅仅是颜色，对吧？还有导航之类的设计。人们经常说我们的网站非常易用，我真的希望我们继续保持这一点。And I think we've... I hope we've done that. Yes, it's less cluttered, isn't it? We've just got the one sidebar now rather than two. I think originally, Richard, we were very concerned when people turned on their computers that all the information, the important information, was there to be seen straight away.我觉得我们……我希望我们做到了。是的，现在简洁多了，对吗？我们现在只有一个侧边栏，不再是两个。我记得当初，Richard，我们很担心用户打开电脑时，所有重要的信息能不能第一时间看到。But that's not so necessary nowadays, is it? The long scroll. It used to be important, as you said, to place all your important elements above the fold, as it were, immediately visible, but now people are more willing to scroll down. And I think they're more willing to scroll down because people, when they use mobile phones, they're always scrolling, scrolling, scrolling down, so it's become a familiar thing to do.但如今这已经不再那么必要了，对吧？现在大家习惯长距离滚动页面。以前，正如你说的，把所有重要内容放在“折叠线以上”（即无需滚动即可看到）是很重要的，但现在人们更愿意往下滑。我想这是因为大家用手机时一直在往下滑，所以已经习惯了这种行为。That is a very good point, Jackie, and it actually brings us to point number two, being responsive. What do you mean by that? Well, you talk about people using mobile phones, so the website has got to change size for people using mobile, so mobile phone users can see the website very, very clearly and very easily. So you're saying that the content should be available on every device? Yes, desktops, laptops, tablets and mobile phones.这一点非常好，Jackie，而且它引出了我们的第二点：响应式设计。你指的是什么？嗯，你提到人们使用手机，所以网站必须根据设备自动调整大小，这样手机用户才能清晰、方便地查看内容。也就是说，内容应该在每种设备上都能正常显示？是的，包括台式机、笔电、平板和手机。And also, of course, that helps us because Google favours sites that are mobile-friendly as well, so search engine optimisation is a very important thing. And the third point, Richard, we're thinking about is shareable content, sharing things. So we really want people to be able to share the podcasts that they like very easily on another social media site, for example, Facebook or Twitter.而且，这样做对我们也有好处，因为 Google 更喜欢对手机友好的网站，所以搜索引擎优化非常重要。第三点，Richard，就是可分享的内容。我们希望用户能够非常轻松地将自己喜欢的播客分享到其他社交媒体，比如 Facebook 或 Twitter。So the website, like every other one, is a work in progress, and we're also looking at future improvements as well, such as, as I mentioned, search engine optimisation and also the speed of page downloads as well. So it's been an exciting new project for us. We hope everybody likes it and we look forward to you telling us what you think about it.所以，我们的网站与其他网站一样，是一个持续改进的过程，我们也在考虑未来的优化，例如我刚提到的搜索引擎优化，以及页面下载速度。对我们来说，这是一个令人兴奋的新项目。我们希望大家喜欢，也希望你们告诉我们对它的看法。Yes, especially if you spot any glitches, let us know.是的，特别是如果你发现任何小问题，请告诉我们。

Nick and Roger strangely fail to don blackface as we visit the "first talkie" (I think you'll find it's a bit more complicated than that), 1927's The Jazz Singer. Your browser isn't showing you an HTML5 audio player. Download

Puedes vernos o escucharnos desde ahora en la nueva app Be Native. Descárgala ya desde el App Store: Be Native. El 13 de noviembre de 2025, Apple hizo algo histórico: lanzar el Programa de Mini Apps con comisión del 15% (la mitad de lo habitual) y actualizar las App Review Guidelines de forma radical. En este episodio te explicamos TODO lo que necesitas saber como desarrollador iOS.

Show DescriptionBen's got an updated edition of his book, Responsive Web Design with HTML5 and CSS, to chat about as well as how much AI is being inserted into writing, layers, scope, color contrast vs contrast color, shouldn't AI bots pay for data they slurp, iFrames permissions issues in Chrome, anchor positioning, and where have all the bloggers gone? Listen on WebsiteWatch on YouTubeGuestsBen FrainGuest's Main URL • Guest's SocialI write web development books, make online courses, and publish YouTube videos. Links Ben Frain – author and web developer Responsive Web Design with HTML5 and CSS available at Amazon benfrain (GitHub) Ben Frain - YouTube JavaScript for Everyone - Piccalilli BCD Watch SponsorsAtomic Design Certification CourseMaster tokens AND atomic design to elevate your design systems game. Get access to both courses, which include hours of comprehensive video lessons, sample token architecture for Figma & Code, process diagrams, exercises, and exclusive Slack!

"le modèle nous fait du feedback directement sur la satisfaction de l'utilisateur" Le D.E.V. de la semaine est Frédéric Barthelet, CTO @ Alpic. Frédéric nous fait découvrir le Model Context Protocol (MCP), une innovation technologique qui ouvre les portes de nos applications et plateformes favorites aux agents IA.Si les applications mobiles et les sites web ont permis aux entreprises de proposer leurs produits et services à leurs utilisateurs, et si les API leur ont permis d'être intégrées chez leurs partenaires, ce sont leurs serveurs MCP qui leur permettront demain d'être découvertes et utilisées par des IA.Frédéric nous partage les enjeux de conception d'un serveur MCP pour maximiser son taux de succès, la métrique phare de la qualité d'un serveur. Réduire le nombre d'outils via le polymorphisme, limiter la portée de ses réponses, utiliser les erreurs comme mécanisme de découvrabilité : autant de stratégies pour développer le meilleur serveur possible.Enfin, il évoque l'avenir du protocole et les potentiels mécanismes de découvrabilité (marketplace et régie pub) qui sont mis en place côté client par des géants comme Anthropic, OpenAI ou Mistral pour amener ses clients sur ce support et canal d'acquisition dernière génération.Chapitrages00:00:55 : Model Context Protocol, kezako ?00:03:53 : Tools, resources et prompts d'un serveur MCP00:05:45 : Améliorer le taux de succès de son serveur00:10:08 : Agentic Experience, ou l'art de désigner des parcours pour agents IA00:13:22 : Limitations et Perspectives00:17:30 : MCP, un nouveau canal d'acquisition pour les business00:23:26 : Contrôle et Autonomie des Agents00:25:39 : Ellicitation, ou comment solliciter l'utilisateur quand l'agent ne suffit plus00:29:10 : Passer à la v2 de son serveur00:53:47 : Perspectives Futures du Protocole MCP01:01:18 : Conclusion Liens évoqués pendant l'émission Alpic pour déployer son serveur MCPTalks de MCP Dev Summit sur YoutubeTalk de Laurie Voss sur toutes les alternatives MCP 🎙️ Soutenez le podcast If This Then Dev ! 🎙️ Chaque contribution aide à maintenir et améliorer nos épisodes. Cliquez ici pour nous soutenir sur Tipeee 🙏Archives | Site | Boutique | TikTok | Discord | Twitter | LinkedIn | Instagram | Youtube | Twitch | Job Board |Hébergé par Audiomeans. Visitez audiomeans.fr/politique-de-confidentialite pour plus d'informations.

Francois Daost is a W3C staff member and co-chair of the Web Developer Experience Community Group. We discuss the W3C's role and what it's like to go through the browser standardization process. Related links W3C TC39 Internet Engineering Task Force Web Hypertext Application Technology Working Group (WHATWG) Horizontal Groups Alliance for Open Media What is MPEG-DASH? | HLS vs. DASH Information about W3C and Encrypted Media Extensions (EME) Widevine PlayReady Media Source API Encrypted Media Extensions API requestVideoFrameCallback() Business Benefits of the W3C Patent Policy web.dev Baseline Portable Network Graphics Specification Internet Explorer 6 CSS Vendor Prefix WebRTC Transcript You can help correct transcripts on GitHub. Intro [00:00:00] Jeremy: today I'm talking to Francois Daoust. He's a staff member at the W3C. And we're gonna talk about the W3C and the recommendation process and discuss, Francois's experience with, with how these features end up in our browsers. [00:00:16] Jeremy: So, Francois, welcome [00:00:18] Francois: Thank you Jeremy and uh, many thanks for the invitation. I'm really thrilled to be part of this podcast. What's the W3C? [00:00:26] Jeremy: I think many of our listeners will have heard about the W3C, but they may not actually know what it is. So could you start by explaining what it is? [00:00:37] Francois: Sure. So W3C stands for the Worldwide Web Consortium. It's a standardization organization. I guess that's how people should think about W3C. it was created in 1994. I, by, uh, Tim Berners Lee, who was the inventor of the web. Tim Berners Lee was the, director of W3C for a long, long time. [00:01:00] Francois: He retired not long ago, a few years back. and W3C is, has, uh, a number of, uh. Properties, let's say first the goal is to produce royalty free standards, and that's very important. Uh, we want to make sure that, uh, the standard that get produced can be used and implemented without having to pay, fees to anyone. [00:01:23] Francois: We do web standards. I didn't mention it, but it's from the name. Standards that you find in your web browsers. But not only that, there are a number of other, uh, standards that got developed at W3C including, for example, XML. Data related standards. W3C as an organization is a consortium. [00:01:43] Francois: The, the C stands for consortium. Legally speaking, it's a, it's a 501c3 meaning in, so it's a US based, uh, legal entity not for profit. And the, the little three is important because it means it's public interest. That means we are a consortium, that means we have members, but at the same time, the goal, the mission is to the public. [00:02:05] Francois: So we're not only just, you know, doing what our members want. We are also making sure that what our members want is aligned with what end users in the end, need. and the W3C has a small team. And so I'm part of this, uh, of this team worldwide. Uh, 45 to 55 people, depending on how you count, mostly technical people and some, uh, admin, uh, as well, overseeing the, uh, the work, that we do, uh, at the W3C. Funding through membership fees [00:02:39] Jeremy: So you mentioned there's 45 to 55 people. How is this funded? Is this from governments or commercial companies? [00:02:47] Francois: The main source comes from membership fees. So the W3C has a, so members, uh, roughly 350 members, uh, at the W3C. And, in order to become a member, an organization needs to pay, uh, an annual membership fee. That's pretty common among, uh, standardization, uh, organizations. [00:03:07] Francois: And, we only have, uh, I guess three levels of membership, fees. Uh, well, you may find, uh, additional small levels, but three main ones. the goal is to make sure that, A big player will, not a big player or large company, will not have more rights than, uh, anything, anyone else. So we try to make sure that a member has the, you know, all members have equal, right? [00:03:30] Francois: if it's not perfect, but, uh, uh, that's how things are, are are set. So that's the main source of income for the W3C. And then we try to diversify just a little bit to get, uh, for example, we go to governments. We may go to governments in the u EU. We may, uh, take some, uh, grant for EU research projects that allow us, you know, to, study, explore topics. [00:03:54] Francois: Uh, in the US there, there used to be some, uh, some funding from coming from the government as well. So that, that's, uh, also, uh, a source. But the main one is, uh, membership fees. Relations to TC39, IETF, and WHATWG [00:04:04] Jeremy: And you mentioned that a lot of the W3C'S work is related to web standards. There's other groups like TC 39, which works on the JavaScript spec and the IETF, which I believe worked, with your group on WebRTC, I wonder if you could explain W3C'S connection to other groups like that. [00:04:28] Francois: sure. we try to collaborate with a, a number of, uh, standard other standardization organizations. So in general, everything goes well because you, you have, a clear separation of concerns. So you mentioned TC 39. Indeed. they are the ones who standardize, JavaScript. Proper name of JavaScript is the EcmaScript. [00:04:47] Francois: So that's tc. TC 39 is the technical committee at ecma. and so we have indeed interactions with them because their work directly impact the JavaScript that you're going to find in your, uh, run in your, in your web browser. And we develop a number of JavaScript APIs, uh, actually in W3C. [00:05:05] Francois: So we need to make sure that, the way we develop, uh, you know, these APIs align with the, the language itself. with IETF, the, the, the boundary is, uh, uh, is clear as well. It's a protocol and protocol for our network protocols for our, the IETF and application level. For W3C, that's usually how the distinction is made. [00:05:28] Francois: The boundaries are always a bit fuzzy, but that's how things work. And usually, uh, things work pretty well. Uh, there's also the WHATWG, uh, and the WHATWG is more the, the, the history was more complicated because, uh, t of a fork of the, uh, HTML specification, uh, at the time when it was developed by W3C, a long time ago. [00:05:49] Francois: And there was been some, uh, Well disagreement on the way things should have been done, and the WHATWG took over got created, took, took this the HTML spec and did it a different way. Went in another, another direction, and that other, other direction actually ended up being the direction. [00:06:06] Francois: So, that's a success, uh, from there. And so, W3C no longer works, no longer owns the, uh, HTML spec and the WHATWG has, uh, taken, uh, taken up a number of, uh, of different, core specifications for the web. Uh, doing a lot of work on the, uh, on interopoerability and making sure that, uh, the algorithm specified by the spec, were correct, which, which was something that historically we haven't been very good at at W3C. [00:06:35] Francois: And the way they've been working as a, has a lot of influence on the way we develop now, uh, the APIs, uh, from a W3C perspective. [00:06:44] Jeremy: So, just to make sure I understand correctly, you have TC 39, which is focused on the JavaScript or ECMAScript language itself, and you have APIs that are going to use JavaScript and interact with JavaScript. So you need to coordinate there. The, the have the specification for HTML. then the IATF, they are, I'm not sure if the right term would be, they, they would be one level lower perhaps, than the W3C. [00:07:17] Francois: That's how you, you can formulate it. Yes. The, the one layer, one layer layer in the ISO network in the ISO stack at the network level. How WebRTC spans the IETF and W3C [00:07:30] Jeremy: And so in that case, one place I've heard it mentioned is that webRTC, to, to use it, there is an IETF specification, and then perhaps there's a W3C recommendation and [00:07:43] Francois: Yes. so when we created the webRTC working group, that was in 2011, I think, it was created with a dual head. There was one RTC web, group that got created at IETF and a webRTC group that got created at W3C. And that was done on purpose. Of course, the goal was not to compete on the, on the solution, but actually to, have the two sides of the, uh, solution, be developed in parallel, the API, uh, the application front and the network front. [00:08:15] Francois: And there was a, and there's still a lot of overlap in, uh, participation between both groups, and that's what keep things successful. In the end. It's not, uh, you know, process or organization to organization, uh, relationships, coordination at the organization level. It's really the fact that you have participants that are essentially the same, on both sides of the equation. [00:08:36] Francois: That helps, uh, move things forward. Now, webRTC is, uh, is more complex than just one group at IETF. I mean, web, webRTC is a very complex set of, uh, of technologies, stack of technologies. So when you, when you. Pull a little, uh, protocol from IETFs. Suddenly you have the whole IETF that comes with you with it. [00:08:56] Francois: So you, it's the, you have the feeling that webRTC needs all of the, uh, internet protocols that got, uh, created to work Recommendations [00:09:04] Jeremy: And I think probably a lot of web developers, they may hear words like specification or standard, but I believe the, the official term, at least at the W3C, is this recommendation. And so I wonder if you can explain what that means. [00:09:24] Francois: Well. It means it means standard in the end. and that came from industry. That comes from a time where. As many standardization organizations. W3C was created not to be a standardization organization. It was felt that standard was not the right term because we were not a standardization organization. [00:09:45] Francois: So recommend IETF has the same thing. They call it RFC, request for comment, which, you know, stands for nothing in, and yet it's a standard. So W3C was created with the same kind of, uh thing. We needed some other terminology and we call that recommendation. But in the end, that's standard. It's really, uh, how you should see it. [00:10:08] Francois: And one thing I didn't mention when I, uh, introduced the W3C is there are two types of standards in the end, two main categories. There are, the de jure standards and defacto standards, two families. The de jure standards are the ones that are imposed by some kind of regulation. so it's really usually a standard you see imposed by governments, for example. [00:10:29] Francois: So when you look at your electric plug at home, there's some regulation there that says, this plug needs to have these properties. And that's a standard that gets imposed. It's a de jure standard. and then there are defacto standards which are really, uh, specifications that are out there and people agree to use it to implement it. [00:10:49] Francois: And by virtue of being used and implemented and used by everyone, they become standards. the, W3C really is in the, uh, second part. It's a defacto standard. IETF is the same thing. some of our standards are used in, uh, are referenced in regulations now, but, just a, a minority of them, most of them are defacto standards. [00:11:10] Francois: and that's important because that's in the end, it doesn't matter what the specific specification says, even though it's a bit confusing. What matters is that the, what the specifications says matches what implementations actually implement, and that these implementations are used, and are used interoperably across, you know, across browsers, for example, or across, uh, implementations, across users, across usages. [00:11:36] Francois: So, uh, standardization is a, is a lengthy process. The recommendation is the final stage in that, lengthy process. More and more we don't really reach recommendation anymore. If you look at, uh, at groups, uh, because we have another path, let's say we kind of, uh, we can stop at candidate recommendation, which is in theoretically a step before that. [00:12:02] Francois: But then you, you can stay there and, uh, stay there forever and publish new candidate recommendations. Um, uh, later on. What matters again is that, you know, you get this, virtuous feedback loop, uh, with implementers, and usage. [00:12:18] Jeremy: So if the candidate recommendation ends up being implemented by all the browsers, what's ends up being the distinction between a candidate and one that's a normal recommendation. [00:12:31] Francois: So, today it's mostly a process thing. Some groups actually decide to go to rec Some groups decide to stay at candidate rec and there's no formal difference between the, the two. we've made sure we've adopted, adjusted the process so that the important bits that, applied at the recommendation level now apply at the candidate rec level. Royalty free patent access [00:13:00] Francois: And by important things, I mean the patent commitments typically, uh, the patent policy fully applies at the candidate recommendation level so that you get your, protection, the royalty free patent protection that we, we were aiming at. [00:13:14] Francois: Some people do not care, you know, but most of the world still works with, uh, with patents, uh, for good, uh, or bad reasons. But, uh, uh, that's how things work. So we need to make, we're trying to make sure that we, we secure the right set of, um, of patent commitments from the right set of stakeholders. [00:13:35] Jeremy: Oh, so when someone implements a W3C recommendation or a candidate recommendation, the patent holders related to that recommendation, they basically agree to allow royalty-free use of that patent. [00:13:54] Francois: They do the one that were involved in the working group, of course, I mean, we can't say anything about the companies out there that may have patents and uh, are not part of this standardization process. So there's always, It's a remaining risk. but part of the goal when we create a working group is to make sure that, people understand the scope. [00:14:17] Francois: Lawyers look into it, and the, the legal teams that exist at the all the large companies, basically gave a green light saying, yeah, we, we we're pretty confident that we, we know where the patterns are on this particular, this particular area. And we are fine also, uh, letting go of the, the patterns we own ourselves. Implementations are built in parallel with standardization [00:14:39] Jeremy: And I think you had mentioned. What ends up being the most important is that the browser creators implement these recommendations. So it sounds like maybe the distinction between candidate recommendation and recommendation almost doesn't matter as long as you get the end result you want. [00:15:03] Francois: So, I mean, people will have different opinions, uh, in the, in standardization circles. And I mentioned also W3C is working on other kind of, uh, standards. So, uh, in some other areas, the nuance may be more important when we, but when, when you look at specification, that's target, web browsers. we've switched from a model where, specs were developed first and then implemented to a model where specs and implementing implementations are being, worked in parallel. [00:15:35] Francois: This actually relates to the evolution I was mentioning with the WHATWG taking over the HTML and, uh, focusing on the interoperability issues because the starting point was, yeah, we have an HTML 4.01 spec, uh, but it's not interoperable because it, it's not specified, are number of areas that are gray areas, you can implement them differently. [00:15:59] Francois: And so there are interoperable issues. Back to candidate rec actually, the, the, the, the stage was created, if I remember correctly. uh, if I'm, if I'm not wrong, the stage was created following the, uh, IE problem. In the CSS working group, IE6, uh, shipped with some, version of a CSS that was in the, as specified, you know, the spec was saying, you know, do that for the CSS box model. [00:16:27] Francois: And the IE6 was following that. And then the group decided to change, the box model and suddenly IE6 was no longer compliant. And that created a, a huge mess on the, in the history of, uh, of the web in a way. And so the, we, the, the, the, the candidate recommendation sta uh, stage was introduced following that to try to catch this kind of problems. [00:16:52] Francois: But nowadays, again, we, we switch to another model where it's more live. and so we, you, you'll find a number of specs that are not even at candidate rec level. They are at the, what we call a working draft, and they, they are being implemented, and if all goes well, the standardization process follows the implementation, and then you end up in a situation where you have your candidate rec when the, uh, spec ships. [00:17:18] Francois: a recent example would be a web GPU, for example. It, uh, it has shipped in, uh, in, in Chrome shortly before it transition to a candidate rec. But the, the, the spec was already stable. and now it's shipping uh, in, uh, in different browsers, uh, uh, safari, uh, and uh, and uh, and uh, Firefox. And so that's, uh, and that's a good example of something that follows, uh, things, uh, along pretty well. But then you have other specs such as, uh, in the media space, uh, request video frame back, uh, frame, call back, uh, requestVideoFrameCallback() is a short API that allows you to get, you know, a call back whenever the, the browser renders a video frame, essentially. [00:18:01] Francois: And that spec is implemented across browsers. But from a W3C specific, perspective, it does not even exist. It's not on the standardization track. It's still being incubated in what we call a community group, which is, you know, some something that, uh, usually exists before. we move to the, the standardization process. [00:18:21] Francois: So there, there are examples of things where some things fell through the cracks. All the standardization process, uh, is either too early or too late and things that are in spec are not exactly what what got implemented or implementations are too early in the process. We we're doing a better job, at, Not falling into a trap where someone ships, uh, you know, an implementation and then suddenly everything is frozen. You can no longer, change it because it's too late, it shipped. we've tried, different, path there. Um, mentioned CSS, the, there was this kind of vendor prefixed, uh, properties that used to be, uh, the way, uh, browsers were deploying new features without, you know, taking the final name. [00:19:06] Francois: We are trying also to move away from it because same thing. Then in the end, you end up with, uh, applications that have, uh, to duplicate all the properties, the CSS properties in the style sheets with, uh, the vendor prefixes and nuances in the, in what it does in, in the end. [00:19:23] Jeremy: Yeah, I, I think, is that in CSS where you'll see --mozilla or things like that? Why requestVideoFrameCallback doesn't have a formal specification [00:19:30] Jeremy: The example of the request video frame callback. I, I wonder if you have an opinion or, or, or know why that ended up the way it did, where the browsers all implemented it, even though it was still in the incubation stage. [00:19:49] Francois: On this one, I don't have a particular, uh, insights on whether there was a, you know, a strong reason to implement it,without doing the standardization work. [00:19:58] Francois: I mean, there are, it's not, uh, an IPR (Intellectual Property Rights) issue. It's not, uh, something that, uh, I don't think the, the, the spec triggers, uh, you know, problems that, uh, would be controversial or whatever. [00:20:10] Francois: Uh, so it's just a matter of, uh, there was no one's priority, and in the end, you end up with a, everyone's happy. it's, it has shipped. And so now doing the spec work is a bit,why spend time on something that's already shipped and so on, but the, it may still come back at some point with try to, you know, improve the situation. [00:20:26] Jeremy: Yeah, that's, that's interesting. It's a little counterintuitive because it sounds like you have the, the working group and it, it sounds like perhaps the companies or organizations involved, they maybe agreed on how it should work, and maybe that agreement almost made it so that they felt like they didn't need to move forward with the specification because they came to consensus even before going through that. [00:20:53] Francois: In this particular case, it's probably because it's really, again, it's a small, spec. It's just one function call, you know? I mean, they will definitely want a working group, uh, for larger specifications. by the way, actually now I know re request video frame call back. It's because the, the, the final goal now that it's, uh, shipped, is to merge it into, uh, HTML, uh, the HTML spec. [00:21:17] Francois: So there's a, there's an ongoing issue on the, the WHATWG side to integrate request video frame callback. And it's taking some time but see, it's, it's being, it, it caught up and, uh, someone is doing the, the work to, to do it. I had forgotten about this one. Um, [00:21:33] Jeremy: Tension from specification review (horizontal review) [00:21:33] Francois: so with larger specifications, organizations will want this kind of IPR regime they will want commit commitments from, uh, others, on the scope, on the process, on everything. So they will want, uh, a larger, a, a more formal setting, because that's part of how you ensure that things, uh, will get done properly. [00:21:53] Francois: I didn't mention it, but, uh, something we're really, uh, Pushy on, uh, W3C I mentioned we have principles, we have priorities, and we have, uh, specific several, uh, properties at W3C. And one of them is that we we're very strong on horizontal reviews of our specs. We really want them to be reviewed from an accessibility perspective, from an internationalization perspective, from a privacy and security, uh, perspective, and, and, and a technical architecture perspective as well. [00:22:23] Francois: And that's, these reviews are part of the formal process. So you, all specs need to undergo these reviews. And from time to time, that creates tension. Uh, from time to time. It just works, you know. Goes without problem. a recurring issue is that, privacy and security are hard. I mean, it's not an easy problem, something that can be, uh, solved, uh, easily. [00:22:48] Francois: Uh, so there's a, an ongoing tension and no easy way to resolve it, but there's an ongoing tension between, specifying powerful APIs and preserving privacy without meaning, not exposing too much information to applications in the media space. You can think of the media capabilities, API. So the media space is a complicated space. [00:23:13] Francois: Space because of codecs. codecs are typically not relative free. and so browsers decide which codecs they're going to support, which audio and video codecs they, they're going to support and doing that, that creates additional fragmentation, not in the sense that they're not interoperable, but in the sense that applications need to choose which connect they're going to ship to stream to the end user. [00:23:39] Francois: And, uh, it's all the more complicated that some codecs are going to be hardware supported. So you will have a hardware decoder in your, in your, in your laptop or smartphone. And so that's going to be efficient to decode some, uh, some stream, whereas some code are not, are going to be software, based, supported. [00:23:56] Francois: Uh, and that may consume a lot of CPU and a lot of power and a lot of energy in the end. So you, you want to avoid that if you can, uh, select another thing. Even more complex than, codecs have different profiles, uh, lower end profiles higher end profiles with different capabilities, different features, uh, depending on whether you're going to use this or that color space, for example, this or that resolution, whatever. [00:24:22] Francois: And so you want to surface that to web applications because otherwise, they can't. Select, they can't choose, the right codec and the right, stream that they're going to send to the, uh, client devices. And so they're not going to provide an efficient user experience first, and even a sustainable one in terms of energy because they, they're going to waste energy if they don't send the right stream. [00:24:45] Francois: So you want to surface that to application. That's what the media, media capabilities, APIs, provides. Privacy concerns [00:24:51] Francois: Uh, but at the same time, if you expose that information, you end up with ways to fingerprint the end user's device. And that in turn is often used to track users across, across sites, which is exactly what we don't want to have, uh, for privacy reasons, for obvious privacy reasons. [00:25:09] Francois: So you have to balance that and find ways to, uh, you know, to expose. Capabilities without, without necessarily exposing them too much. Uh, [00:25:21] Jeremy: Can you give an example of how some of those discussions went? Like within the working group? Who are the companies or who are the organizations that are arguing for We shouldn't have this capability because of the privacy concerns, or [00:25:40] Francois: In a way all of the companies, have a vision of, uh, of privacy. I mean, the, you will have a hard time finding, you know, members saying, I don't care about privacy. I just want the feature. Uh, they all have privacy in mind, but they may have a different approach to privacy. [00:25:57] Francois: so if you take, uh, let's say, uh, apple and Google would be the, the, I guess the perfect examples in that, uh, in that space, uh, Google will have a, an approach that is more open-ended thing. The, the user agents has this, uh, should check what the, the, uh, given site is doing. And then if it goes beyond, you know, some kind of threshold, they're going to say, well, okay, well, we'll stop exposing data to that, to that, uh, to that site. [00:26:25] Francois: So that application. So monitor and react in a way. apple has a more, uh, you know, has a stricter view on, uh, on privacy, let's say. And they will say, no, we, the, the, the feature must not exist in the first place. Or, but that's, I mean, I guess, um, it's not always that extreme. And, uh, from time to time it's the opposite. [00:26:45] Francois: You will have, uh, you know, apple arguing in one way, uh, which is more open-ended than the, uh, than, uh, than Google, for example. And they are not the only ones. So in working groups, uh, you will find the, usually the implementers. Uh, so when we talk about APIs that get implemented in browsers, you want the core browsers to be involved. [00:27:04] Francois: Uh, otherwise it's usually not a good sign for, uh, the success of the, uh, of the technology. So in practice, that means Apple, uh, Microsoft, Mozilla which one did I forget? [00:27:15] Jeremy: Google. [00:27:16] Francois: I forgot Google. Of course. Thank you. that's, uh, that the, the core, uh, list of participants you want to have in any, uh, group that develops web standards targeted at web browsers. Who participates in working groups and how much power do they have? [00:27:28] Francois: And then on top of that, you want, organizations and people who are directly going to use it, either because they, well the content providers. So in media, for example, if you look at the media working group, you'll see, uh, so browser vendors, the ones I mentioned, uh, content providers such as the BBC or Netflix. [00:27:46] Francois: Chip set vendors would, uh, would be there as well. Intel, uh, Nvidia again, because you know, there's a hardware decoding in there and encoding. So media is, touches on, on, uh, on hardware, uh, device manufacturer in general. You may, uh, I think, uh, I think Sony is involved in the, in the media working group, for example. [00:28:04] Francois: and these companies are usually less active in the spec development. It depends on the groups, but they're usually less active because the ones developing the specs are usually the browser again, because as I mentioned, we develop the specs in parallel to browsers implementing it. So they have the. [00:28:21] Francois: The feedback on how to formulate the, the algorithms. and so that's this collection of people who are going to discuss first within themselves. W3C pushes for consensual dis decisions. So we hardly take any votes in the working groups, but from time to time, that's not enough. [00:28:41] Francois: And there may be disagreements, but let's say there's agreement in the group, uh, when the spec matches. horizontal review groups will look at the specs. So these are groups I mentioned, accessibility one, uh, privacy, internationalization. And these groups, usually the participants are, it depends. [00:29:00] Francois: It can be anything. It can be, uh, the same companies. It can be, but usually different people from the same companies. But it the, maybe organizations with a that come from very, a very different angle. And that's a good thing because that means the, you know, you enlarge the, the perspectives on your, uh, on the, on the technology. [00:29:19] Francois: and you, that's when you have a discussion between groups, that takes place. And from time to time it goes well from time to time. Again, it can trigger issues that are hard to solve. and the W3C has a, an escalation process in case, uh, you know, in case things degenerate. Uh, starting with, uh, the notion of formal objection. [00:29:42] Jeremy: It makes sense that you would have the, the browser. Vendors and you have all the different companies that would use that browser. All the different horizontal groups like you mentioned, the internationalization, accessibility. I would imagine that you were talking about consensus and there are certain groups or certain companies that maybe have more say or more sway. [00:30:09] Jeremy: For example, if you're a browser, manufacturer, your Google. I'm kind of curious how that works out within the working group. [00:30:15] Francois: Yes, it's, I guess I would be lying if I were saying that, uh, you know, all companies are strictly equal in a, in a, in a group. they are from a process perspective, I mentioned, you know, different membership fees with were design, special specific ethos so that no one could say, I'm, I'm putting in a lot of money, so you, you need to re you need to respect me, uh, and you need to follow what I, what I want to, what I want to do. [00:30:41] Francois: at the same time, if you take a company like, uh, like Google for example, they send, hundreds of engineers to do standardization work. That's absolutely fantastic because that means work progresses and it's, uh, extremely smart people. So that's, uh, that's really a pleasure to work with, uh, with these, uh, people. [00:30:58] Francois: But you need to take a step back and say, well, the problem is. Defacto that gives them more power just by virtue of, uh, injecting more resources into it. So having always someone who can respond to an issue, having always someone, uh, editing a spec defacto that give them more, uh, um, more say on the, on the directions that, get forward. [00:31:22] Francois: And on top of that, of course, they have the, uh, I guess not surprisingly, the, the browser that is, uh, used the most, currently, on the market so there's a little bit of a, the, the, we, we, we, we try very hard to make sure that, uh, things are balanced. it's not a perfect world. [00:31:38] Francois: the the role of the team. I mean, I didn't talk about the role of the team, but part of it is to make sure that. Again, all perspectives are represented and that there's not, such a, such big imbalance that, uh, that something is wrong and that we really need to look into it. so making sure that anyone, if they have something to say, make making sure that they are heard by the rest of the group and not dismissed. [00:32:05] Francois: That usually goes well. There's no problem with that. And again, the escalation process I mentioned here doesn't make any, uh, it doesn't make any difference between, uh, a small player, a large player, a big player, and we have small companies raising formal objections against some of our aspects that happens, uh, all large ones. [00:32:24] Francois: But, uh, that happens too. There's no magical solution, I guess you can tell it by the way. I, uh, I don't know how to formulate the, the process more. It's a human process, and that's very important that it remains a human process as well. [00:32:41] Jeremy: I suppose the role of, of staff and someone in your position, for example, is to try and ensure that these different groups are, are heard and it isn't just one group taking control of it. [00:32:55] Francois: That's part of the role, again, is to make sure that, uh, the, the process is followed. So the, I, I mean, I don't want to give the impression that the process controls everything in the groups. I mean, the, the, the groups are bound by the process, but the process is there to catch problems when they arise. [00:33:14] Francois: most of the time there are no problems. It's just, you know, again, participants talking to each other, talking with the rest of the community. Most of the work happens in public nowadays, in any case. So the groups work in public essentially through asynchronous, uh, discussions on GitHub repositories. [00:33:32] Francois: There are contributions from, you know, non group participants and everything goes well. And so the process doesn't kick in. You just never say, eh, no, you didn't respect the process there. You, you closed the issue. You shouldn't have a, it's pretty rare that you have to do that. Uh, things just proceed naturally because they all, everyone understands where they are, why, what they're doing, and why they're doing it. [00:33:55] Francois: we still have a role, I guess in the, in the sense that from time to time that doesn't work and you have to intervene and you have to make sure that,the, uh, exception is caught and, uh, and processed, uh, in the right way. Discussions are public on github [00:34:10] Jeremy: And you said this process is asynchronous in public, so it sounds like someone, I, I mean, is this in GitHub issues or how, how would somebody go and, and see what the results of [00:34:22] Francois: Yes, there, there are basically a gazillion of, uh, GitHub repositories under the, uh, W3C, uh, organization on GitHub. Most groups are using GitHub. I mean, there's no, it's not mandatory. We don't manage any, uh, any tooling. But the factors that most, we, we've been transitioning to GitHub, uh, for a number of years already. [00:34:45] Francois: Uh, so that's where the work most of the work happens, through issues, through pool requests. Uh, that's where. people can go and raise issues against specifications. Uh, we usually, uh, also some from time to time get feedback from developers and countering, uh, a bug in a particular implementations, which we try to gently redirect to, uh, the actual bug trackers because we're not responsible for the respons implementations of the specs unless the spec is not clear. [00:35:14] Francois: We are responsible for the spec itself, making sure that the spec is clear and that implementers well, understand how they should implement something. Why the W3C doesn't specify a video or audio codec [00:35:25] Jeremy: I can see how people would make that mistake because they, they see it's the feature, but that's not the responsibility of the, the W3C to implement any of the specifications. Something you had mentioned there's the issue of intellectual property rights and how when you have a recommendation, you require the different organizations involved to make their patents available to use freely. [00:35:54] Jeremy: I wonder why there was never any kind of, recommendation for audio or video codecs in browsers since you have certain ones that are considered royalty free. But, I believe that's never been specified. [00:36:11] Francois: At W3C you mean? Yes. we, we've tried, I mean, it's not for lack of trying. Um, uh, we've had a number of discussions with, uh, various stakeholders saying, Hey, we, we really need, an audio or video code for our, for the web. the, uh, png PNG is an example of a, um, an image format which got standardized at W3C and it got standardized at W3C similar reasons. There had to be a royalty free image format for the web, and there was none at the time. of course, nowadays, uh, jpeg, uh, and gif or gif, whatever you call it, are well, you know, no problem with them. But, uh, um, that at the time P PNG was really, uh, meant to address this issue and it worked for PNG for audio and video. [00:37:01] Francois: We haven't managed to secure, commitments by stakeholders. So willingness to do it, so it's not, it's not lack of willingness. We would've loved to, uh, get, uh, a royalty free, uh, audio codec, a royalty free video codec again, audio and video code are extremely complicated because of this. [00:37:20] Francois: not only because of patterns, but also because of the entire business ecosystem that exists around them for good reasons. You, in order for a, a codec to be supported, deployed, effective, it really needs, uh, it needs to mature a lot. It needs to, be, uh, added to at a hardware level, to a number of devices, capturing devices, but also, um, uh, uh, of course players. [00:37:46] Francois: And that takes a hell of a lot of time and that's why you also enter a number of business considerations with business contracts between entities. so I'm personally, on a personal level, I'm, I'm pleased to see, for example, the Alliance for Open Media working on, uh, uh, AV1, uh, which is. At least they, uh, they wanted to be royalty free and they've been adopting actually the W3C patent policy to do this work. [00:38:11] Francois: So, uh, we're pleased to see that, you know, they've been adopting the same process and same thing. AV1 is not yet at the same, support stage, as other, codecs, in the world Yeah, I mean in devices. There's an open question as what, what are we going to do, uh, in the future uh, with that, it's, it's, it's doubtful that, uh, the W3C will be able to work on a, on a royalty free audio, codec or royalty free video codec itself because, uh, probably it's too late now in any case. [00:38:43] Francois: but It's one of these angles in the, in the web platform where we wish we had the, uh, the technology available for, for free. And, uh, it's not exactly, uh, how things work in practice.I mean, the way codecs are developed remains really patent oriented. [00:38:57] Francois: and you will find more codecs being developed. and that's where geopolitics can even enter the, the, uh, the play. Because, uh, if you go to China, you will find new codecs emerging, uh, that get developed within China also, because, the other codecs come mostly from the US so it's a bit of a problem and so on. [00:39:17] Francois: I'm not going to enter details and uh, I would probably say stupid things in any case. Uh, but that, uh, so we continue to see, uh, emerging codecs that are not royalty free, and it's probably going to remain the case for a number of years. unfortunately, unfortunately, from a W3C perspective and my perspective of course. [00:39:38] Jeremy: There's always these new, formats coming out and the, rate at which they get supported in the browser, even on a per browser basis is, is very, there can be a long time between, for example, WebP being released and a browser supporting it. So, seems like maybe we're gonna be in that situation for a while where the codecs will come out and maybe the browsers will support them. Maybe they won't, but the, the timeline is very uncertain. Digital Rights Management (DRM) and Media Source Extensions [00:40:08] Jeremy: Something you had, mentioned, maybe this was in your, email to me earlier, but you had mentioned that some of these specifications, there's, there's business considerations like with, digital rights management and, media source extensions. I wonder if you could talk a little bit about maybe what media source extensions is and encrypted media extensions and, and what the, the considerations or challenges are there. [00:40:33] Francois: I'm going to go very, very quickly over the history of a, video and audio support on the web. Initially it was supported through plugins. you are maybe too young to, remember that. But, uh, we had extensions, added to, uh, a realplayer. [00:40:46] Francois: This kind of things flash as well, uh, supporting, uh, uh, videos, in web pages, but it was not provided by the web browsers themselves. Uh, then HTML5 changed the, the situation. Adding these new tags, audio and video, but that these tags on this, by default, support, uh, you give them a resources, a resource, like an image as it's an audio or a video file. [00:41:10] Francois: They're going to download this, uh, uh, video file or audio file, and they're going to play it. That works well. But as soon as you want to do any kind of real streaming, files are too large and to stream, to, to get, you know, to get just a single fetch on, uh, on them. So you really want to stream them chunk by chunk, and you want to adapt the resolution at which you send the stream based on real time conditions of the user's network. [00:41:37] Francois: If there's plenty of bandwidth you want to send the user, the highest possible resolution. If there's a, some kind of hiccup temporary in the, in the network, you really want to lower the resolution, and that's called adaptive streaming. And to get adaptive streaming on the web, well, there are a number of protocols that exist. [00:41:54] Francois: Same thing. Some many of them are proprietary and actually they remain proprietary, uh, to some extent. and, uh, some of them are over http and they are the ones that are primarily used in, uh, in web contexts. So DASH comes to mind, DASH for Dynamic Adaptive streaming over http. HLS is another one. Uh, initially developed by Apple, I believe, and it's, uh, HTTP live streaming probably. Exactly. And, so there are different protocols that you can, uh, you can use. Uh, so the goal was not to standardize these protocols because again, there were some proprietary aspects to them. And, uh, same thing as with codecs. [00:42:32] Francois: There was no, well, at least people wanted to have the, uh, flexibility to tweak parameters, adaptive streaming parameters the way they wanted for different scenarios. You may want to tweak the parameters differently. So they, they needed to be more flexibility on top of protocols not being truly available for use directly and for implementation directly in browsers. [00:42:53] Francois: It was also about providing applications with, uh, the flexibility they would need to tweak parameters. So media source extensions comes into play for exactly that. Media source extensions is really about you. The application fetches chunks of its audio and video stream the way it wants, and with the parameters it wants, and it adjusts whatever it wants. [00:43:15] Francois: And then it feeds that into the, uh, video or audio tag. and the browser takes care of the rest. So it's really about, doing, you know, the adaptive streaming. let applications do it, and then, uh, let the user agent, uh, the browser takes, take care of the rendering itself. That's media source extensions. [00:43:32] Francois: Initially it was pushed by, uh, Netflix. They were not the only ones of course, but there, there was a, a ma, a major, uh, proponent of this, uh, technical solution, because they wanted, uh, they, uh, they were, expanding all over the world, uh, with, uh, plenty of native, applications on all sorts of, uh, of, uh, devices. [00:43:52] Francois: And they wanted to have a way to stream content on the web as well. both for both, I guess, to expand to, um, a new, um, ecosystem, the web, uh, providing new opportunities, let's say. But at the same time also to have a fallback, in case they, because for native support on different platforms, they sometimes had to enter business agreements with, uh, you know, the hardware manufacturers, the whatever, the, uh, service provider or whatever. [00:44:19] Francois: and so that was a way to have a full back. That kind of work is more open, in case, uh, things take some time and so on. So, and they probably had other reasons. I mean, I'm not, I can't speak on behalf of Netflix, uh, on others, but they were not the only ones of course, uh, supporting this, uh, me, uh, media source extension, uh, uh, specification. [00:44:42] Francois: and that went kind of, well, I think it was creating 2011. I mean, the, the work started in 2011 and the recommendation was published in 2016, which is not too bad from a standardization perspective. It means only five years, you know, it's a very short amount of time. Encrypted Media Extensions [00:44:59] Francois: At the same time, and in parallel and complement to the media source extension specifications, uh, there was work on the encrypted media extensions, and here it was pushed by the same proponent in a way because they wanted to get premium content on the web. [00:45:14] Francois: And by premium content, you think of movies and, uh. These kind of beasts. And the problem with the, I guess the basic issue with, uh, digital asset such as movies, is that they cost hundreds of millions to produce. I mean, some cost less of course. And yet it's super easy to copy them if you have a access to the digital, uh, file. [00:45:35] Francois: You just copy and, uh, and that's it. Piracy uh, is super easy, uh, to achieve. It's illegal of course, but it's super easy to do. And so that's where the different legislations come into play with digital right management. Then the fact is most countries allow system that, can encrypt content and, uh, through what we call DRM systems. [00:45:59] Francois: so content providers, uh, the, the ones that have movies, so the studios here more, more and more, and Netflix is one, uh, one of the studios nowadays. Um, but not only, not only them all major studios will, uh, would, uh, push for, wanted to have something that would allow them to stream encrypted content, encrypted audio and video, uh, mostly video, to, uh, to web applications so that, uh, you. [00:46:25] Francois: Provide the movies, otherwise, they, they are just basically saying, and sorry, but, uh, this premium content will never make it to the web because there's no way we're gonna, uh, send it in clear, to, uh, to the end user. So Encrypting media extensions is, uh, is an API that allows to interface with, uh, what's called the content decryption module, CDM, uh, which itself interacts with, uh, the DR DRM systems that, uh, the browser may, may or may not support. [00:46:52] Francois: And so it provides a way for an application to receive encrypted content, pass it over get the, the, the right keys, the right license keys from a whatever system actually. Pass that logic over to the, and to the user agent, which passes, passes it over to, uh, the CDM system, which is kind of black box in, uh, that does its magic to get the right, uh, decryption key and then the, and to decrypt the content that can be rendered. [00:47:21] Francois: The encrypted media extensions triggered a, a hell of a lot of, uh, controversy. because it's DRM and DRM systems, uh, many people, uh, uh, things should be banned, uh, especially on the web because the, the premise of the web is that the, the user has trusts, a user agent. The, the web browser is called the user agent in all our, all our specifications. [00:47:44] Francois: And that's, uh, that's the trust relationship. And then they interact with a, a content provider. And so whatever they do with the content is their, I guess, actually their problem. And DRM introduces a third party, which is, uh, there's, uh, the, the end user no longer has the control on the content. [00:48:03] Francois: It has to rely on something else that, Restricts what it can achieve with the content. So it's, uh, it's not only a trust relationship with its, uh, user agents, it's also with, uh, with something else, which is the content provider, uh, in the end, the one that has the, uh, the license where provides the license. [00:48:22] Francois: And so that's, that triggers, uh, a hell of a lot of, uh, of discussions in the W3C degenerated, uh, uh, into, uh, formal objections being raised against the specification. and that escalated to, to the, I mean, at all leverage it. It's, it's the, the story in, uh, W3C that, um, really, uh, divided the membership into, opposed camps in a way, if you, that's was not only year, it was not really 50 50 in the sense that not just a huge fights, but the, that's, that triggered a hell of a lot of discussions and a lot of, a lot of, uh, of formal objections at the time. [00:49:00] Francois: Uh, we were still, From a governance perspective, interestingly, um, the W3C used to be a dictatorship. It's not how you should formulate it, of course, and I hope it's not going to be public, this podcast. Uh, but the, uh, it was a benevolent dictatorship. You could see it this way in the sense that, uh, the whole process escalated to one single person was, Tim Burners Lee, who had the final say, on when, when none of the other layers, had managed to catch and to resolve, a conflict. [00:49:32] Francois: Uh, that has hardly ever happened in, uh, the history of the W3C, but that happened to the two for EME, for encrypted media extensions. It had to go to the, uh, director level who, uh, after due consideration, uh, decided to, allow the EME to proceed. and that's why we have a, an EME, uh, uh, standard right now, but still re it remains something on the side. [00:49:56] Francois: EME we're still, uh, it's still in the scope of the media working group, for example. but the scope, if you look at the charter of the working group, we try to scope the, the, the, the, the updates we can make to the specification, uh, to make sure that we don't reopen, reopen, uh, a can of worms, because, well, it's really a, a topic that triggers friction for good and bad reasons again. [00:50:20] Jeremy: And when you talk about the media source extensions, that is the ability to write custom code to stream video in whatever way you want. You mentioned, the MPEG-DASH and http live streaming. So in that case, would that be the developer gets to write that code in JavaScript that's executed by the browser? [00:50:43] Francois: Yep, that's, uh, that would be it. and then typically, I guess the approach nowadays is more and more to develop low level APIs into W3C or web in, in general, I guess. And to let, uh. Libraries emerge that are going to make lives of a, a developer, uh, easier. So for MPEG DASH, we have the DASH.js, which does a fantastic job at, uh, at implementing the complexity of, uh, of adaptive streaming. [00:51:13] Francois: And you just, you just hook it into your, your workflow. And that's, uh, and that's it. Encrypted Media Extensions are closed source [00:51:20] Jeremy: And with the encrypted media extensions I'm trying to picture how those work and how they work differently. [00:51:28] Francois: Well, it's because the, the, the, the key architecture is that the, the stream that you, the stream that you may assemble with a media source extensions, for example. 'cause typically they, they're used in collaboration. When you hook the, hook it into the video tag, you also. Call EME and actually the stream goes to EME. [00:51:49] Francois: And when it goes to EME, actually the user agent hands the encrypted stream. You're still encrypted at this time. Uh, encrypted, uh, stream goes to the CDM content decryption module, and that's a black box well, it has some black, black, uh, black box logic. So it's not, uh, even if you look at the chromium source code, for example, you won't see the implementation of the CDM because it's a, it's a black box, so it's not part of the browser se it's a sand, it's sandboxed, it's execution sandbox. [00:52:17] Francois: That's, uh, the, the EME is kind of unique in, in this way where the, the CDM is not allowed to make network requests, for example, again, for privacy reasons. so anyway, the, the CDM box has the logic to decrypt the content and it hands it over, and then it depends, it depends on the level of protection you. [00:52:37] Francois: You need or that the system supports. It can be against software based protection, in which case actually, a highly motivated, uh, uh, uh, attacker could, uh, actually get access to the decoded stream, or it can be more hardware protected, in which case actually the, it goes to the, uh, to your final screen. [00:52:58] Francois: But it goes, it, it goes through the hardware in a, in a mode that the US supports in a mode that even the user agent doesn't have access to it. So it doesn't, it can't even see the pixels that, uh, gets rendered on the screen. There are, uh, several other, uh, APIs that you could use, for example, to take a screenshot of your, of your application and so on. [00:53:16] Francois: And you cannot apply them to, uh, such content because they're just gonna return a black box. again, because the user agent itself does not see the, uh, the pixels, which is exactly what you want with encrypted content. [00:53:29] Jeremy: And the, the content decryption module, it's, if I understand correctly, it's something that's shipped with the browsers, but you were saying is if you were to look at the public source code of Chromium or of Firefox, you would not see that implementation. Content Decryption Module (Widevine, PlayReady) [00:53:47] Francois: True. I mean, the, the, um, the typical examples are, uh, uh, widevine, so wide Vine. So interestingly, uh, speaking in theory, these, uh, systems could have been provided by anyone in practice. They've been provided by the browser vendors themselves. So Google has Wide Vine. Uh, Microsoft has something called PlayReady. Apple uh, the name, uh, escapes my, uh, sorry. They don't have it on top of my mind. So they, that's basically what they support. So they, they also own that code, but in a way they don't have to. And Firefox actually, uh, they, uh, don't, don't remember which one, they support among these three. but, uh, they, they don't own that code typically. [00:54:29] Francois: They provide a wrapper around, around it. Yeah, that's, that's exactly the, the crux of the, uh, issue that, people have with, uh, with DRMs, right? It's, uh, the fact that, uh, suddenly you have a bit of code running there that is, uh, that, okay, you can send box, but, uh, you cannot inspect and you don't have, uh, access to its, uh, source code. [00:54:52] Jeremy: That's interesting. So the, almost the entire browser is open source, but if you wanna watch a Netflix movie for example, then you, you need to, run this, this CDM, in addition to just the browser code. I, I think, you know, we've kind of covered a lot. Documenting what's available in browsers for developers [00:55:13] Jeremy: I wonder if there's any other examples or anything else you thought would be important to mention in, in the context of the W3C. [00:55:23] Francois: There, there's one thing which, uh, relates to, uh, activities I'm doing also at W3C. Um. Here, we've been talking a lot about, uh, standards and, implementations in browsers, but there's also, uh, adoption of these browser, of these technology standards by developers in general and making sure that developers are aware of what exists, making sure that they understand what exists and one of the, key pain points that people, uh. [00:55:54] Francois: Uh, keep raising on, uh, the web platform is first. Well, the, the, the web platform is unique in the sense that there are different implementations. I mean, if you, [00:56:03] Francois: Uh, anyway, there are different, uh, context, different run times where there, there's just one provided by the company that owns the, uh, the, the, the system. The web platform is implemented by different, uh, organizations. and so you end up the system where no one, there's what's in the specs is not necessarily supported. [00:56:22] Francois: And of course, MDN tries, uh, to document what's what's supported, uh, thoroughly. But for MDN to work, there's a hell of a lot of needs for data that, tracks browser support. And this, uh, this data is typically in a project called the Browser Compat Data, BCD owned by, uh, MDN as well. But, the Open Web Docs collective is a, uh, is, uh, the one, maintaining that, uh, that data under the hoods. [00:56:50] Francois: anyway, all of that to say that, uh, to make sure that, we track things beyond work on technical specifications, because if you look at it from W3C perspective, life ends when the spec reaches standards, uh, you know, candidate rec or rec, you could just say, oh, done with my work. but that's not how things work. [00:57:10] Francois: There's always, you need the feedback loop and, in order to make sure that developers get the information and can provide the, the feedback that standardization can benefit from and browser vendors can benefit from. We've been working on a project called web Features with browser vendors mainly, and, uh, a few of the folks and MDN and can I use and different, uh, different people, to catalog, the web in terms of features that speak to developers and from that catalog. [00:57:40] Francois: So it's a set of, uh, it's a set of, uh, feature IDs with a feature name and feature description that say, you know, this is how developers would, uh, understand, uh, instead of going too fine grained in terms of, uh, there's this one function call that does this because that's where you, the, the kind of support data you may get from browser data and MDN initially, and having some kind of a coarser grained, uh, structure that says these are the, features that make sense. [00:58:09] Francois: They talk to developers. That's what developers talk about, and that's the info. So the, we need to have data on these particular features because that's how developers are going approach the specs. Uh. and from that we've derived the notion of baseline badges that you have, uh, are now, uh, shown on MDN on can I use and integrated in, uh, IDE tool, IDE Tools such as visual, visual studio, and, uh, uh, libraries, uh, linked, some linters have started to, um, to integrate that data. [00:58:41] Francois: Uh, so, the way it works is, uh, we've been mapping these coarser grained features to BCDs finer grained support data, and from there we've been deriving a kind of a, a batch that says, yeah, this, this feature is implemented well, has limited availability because it's only implemented in one or two browsers, for example. [00:59:07] Francois: It's, newly available because. It was implemented. It's been, it's implemented across the main browser vendor, um, across the main browsers that people use. But it's recent, and widely available, which we try to, uh, well, there's been lots of discussion in the, in the group to, uh, come up with a definition which essentially ends up being 30 months after, a feature become, became newly available. [00:59:34] Francois: And that's when, that's the time it takes for the, for the versions of the, the different versions of the browser to propagate. Uh, because you, it's not because there's a new version of a, of a browser that, uh, people just, Ima immediately, uh, get it. So it takes a while, to propagate, uh, across the, uh, the, the user, uh, user base. [00:59:56] Francois: And so the, the goal is to have a, a, a signal that. Developers can rely on saying, okay, well it's widely available so I can really use that feature. And of course, if that doesn't work, then we need to know about it. And so we are also working with, uh, people doing so developer surveys such as state of, uh, CSS, state of HTML, state of JavaScript. [01:00:15] Francois: That's I guess, the main ones. But also we are also running, uh, MDN short surveys with the MDN people to gather feedback on. On the, on these same features, and to feed the loop and to, uh, to complete the loop. and these data is also used by, internally, by browser vendors to inform, prioritization process, their prioritization process, and typically as part of the interop project that they're also running, uh, on the site [01:00:43] Francois: So a, a number of different, I've mentioned, uh, I guess a number of different projects, uh, coming along together. But that's the goal is to create links, across all of these, um, uh, ongoing projects with a view to integrating developers, more, and gathering feedback as early as possible and inform decision. [01:01:04] Francois: We take at the standardization level that can affect the, the lives of the developers and making sure that it's, uh, it affects them in a, in a positive way. [01:01:14] Jeremy: just trying to understand, 'cause you had mentioned that there's the web features and the baseline, and I was, I was trying to picture where developers would actually, um, see these things. And it sounds like from what you're saying is W3C comes up with what stage some of these features are at, and then developers would end up seeing it on MDN or, or some other site. [01:01:37] Francois: So, uh, I'm working on it, but that doesn't mean it's a W3C thing. It's a, it's a, again, it's a, we have different types of group. It's a community group, so it's the Web DX Community group at W3C, which means it's a community owned thing. so that's why I'm mentioning a working with a representative from, and people from MDN people, from open Web docs. [01:02:05] Francois: so that's the first point. The second point is, so it's, indeed this data is now being integrated. If you, and you look, uh, you'll, you'll see it in on top of the MDN pages on most of them. If you look at, uh, any kind of feature, you'll see a, a few logos, uh, a baseline banner. and then can I use, it's the same thing. [01:02:24] Francois: You're going to get a baseline, banner. It's more on, can I use, and it's meant to capture the fact that the feature is widely available or if you may need to pay attention to it. Of course, it's a simplification, and the goal is not to the way it's, the way the messaging is done to developers is meant to capture the fact that, they may want to look, uh, into more than just this, baseline status, because. [01:02:54] Francois: If you take a look at web platform tests, for example, and if you were to base your assessment of whether a feature is supported based on test results, you'll end up saying the web platform has no supported technology because there are absolutely no API that, uh, where browsers pass 100% of the, of the, of the test suite. [01:03:18] Francois: There may be a few of them, I don't know. But, there's a simplification in the, in the process when a feature is, uh, set to be baseline, there may be more things to look at nevertheless, but it's meant to provide a signal that, uh, still developers can rely on their day-to-day, uh, lives. [01:03:36] Francois: if they use the, the feature, let's say, as a reasonably intended and not, uh, using to advance the logic. [01:03:48] Jeremy: I see. Yeah. I'm looking at one of the pages on MDN right now, and I can see at the top there's the, the baseline and it, it mentions that this feature works across many browsers and devices, and then they say how long it's been available. And so that's a way that people at a glance can, can tell, which APIs they can use. [01:04:08] Francois: it also started, uh, out of a desire to summarize this, uh, browser compatibility table that you see at the end of the page of the, the bottom of the page in on MDN. but there are where developers were saying, well, it's, it's fine, but it's, it goes too much into detail. So we don't know in the end, can we, can we use that feature or can we, can we not use that feature? [01:04:28] Francois: So it's meant as a informed summary of, uh, of, of that it relies on the same data again. and more importantly, we're beyond MDN, we're working with tools providers to integrate that as well. So I mentioned the, uh, visual Studio is one of them. So recently they shipped a new version where when you use a feature, you can, you can have some contextual, uh. [01:04:53] Francois: A menu that tells you, yeah, uh, that's fine. You, this CSS property, you can, you can use it, it's widely available or be aware this one is limited Availability only, availability only available in Firefox or, or Chrome or Safari work kit, whatever. [01:05:08] Jeremy: I think that's a good place to wrap it up, if people want to learn more about the work you're doing or learn more about sort of this whole recommendations process, where, where should they head? [01:05:23] Francois: Generally speaking, we're extremely open to, uh, people contributing to the W3C. and where should they go if they, it depends on what they want. So I guess the, the in usually where, how things start for someone getting involved in the W3C is that they have some

Is HTML5 gaming finally ready for prime time? After multiple failed attempts over the past decade, something fundamental has changed.In this episode, we sit down with Dmitry Kachmar, CEO of Playgama, who recently raised $3M betting that HTML5 gaming is about to explode. Thanks to GPU advances from the AI boom, 5G networks, and platforms desperate to break free from app store monopolies, browser-based games can now rival native mobile experiences.We dive deep into why Discord, Telegram, and every platform with a web browser are racing to add games, how developers can leverage this distribution revolution, and what it means for the future of mobile gaming. Dmitry shares candid insights on platform-specific design strategies, the challenges of retention without app store lock-in, and why he believes over 50% of new games will be HTML5-based within five years.Whether you're a game developer, studio executive, or industry observer, this conversation reveals why the next chapter of gaming might not be in app stores—but everywhere else.Key topics covered:The technology convergence enabling AAA-quality browser gamesHow to monetize without the 30% app store taxPlatform-specific strategies for Discord vs Telegram vs web portalsWhy viral mechanics are no longer optionalThe distribution-as-a-service model reshaping game publishingAbout the guest: Dmitry Kachmar is CEO of Playgama, former Yandex executive with multiple exits, Harvard grad, and adtech veteran now focused on democratizing game distribution.

In this repeat episode, Chris Coyier, co-founder of CodePen, talks about the evolving landscape of HTML heading into 2025. He delves into topics like the slow evolution of HTML compared to CSS and JavaScript, the importance of backwards compatibility, new HTML elements and pseudo-elements, and the potential of declarative shadow DOM for server-side rendering in web components. Links Website: https://chriscoyier.net Codepen: https://codepen.io/chriscoyier Frontend Social: https://front-end.social/@chriscoyier Github: https://github.com/chriscoyier Threads: https://www.threads.net/@chriscoyier Bluesky: https://bsky.app/profile/chriscoyier.net We want to hear from you! How did you find us? Did you see us on Twitter? In a newsletter? Or maybe we were recommended by a friend? Let us know by sending an email to our producer, Em, at emily.kochanek@logrocket.com (mailto:emily.kochanek@logrocket.com), or tweet at us at PodRocketPod (https://twitter.com/PodRocketpod). Follow us. Get free stickers. Follow us on Apple Podcasts, fill out this form (https://podrocket.logrocket.com/get-podrocket-stickers), and we'll send you free PodRocket stickers! What does LogRocket do? LogRocket provides AI-first session replay and analytics that surfaces the UX and technical issues impacting user experiences. Start understanding where your users are struggling by trying it for free at LogRocket.com. Try LogRocket for free today. (https://logrocket.com/signup/?pdr) Special Guest: Chris Coyier.

The 16:9 PODCAST IS SPONSORED BY SCREENFEED – DIGITAL SIGNAGE CONTENT Digichief has been helping digital signage and DOOH network operators feed the so-called content beast for a bunch of years. While the Kentucky-based company started up in 2007, its roots go back another decade to a tech start-up that did similar graphics-driven content work for broadcast TV. I've known co-founder Gene Hamm forever, but this podcast was the first time we had a detailed chat about what Digichief does and offers. We get into a bunch of things, including what's widely used and what seems like perfect contextual content, but hasn't caught on. We talk in detail, as well, about more customized content, and about a new service called Mercury that Digichief spent more than a year developing and recently rolled out. If you hear thumping sounds in the background on my end, that's the roofers. It wasn't until the morning we recorded this that I remembered about the racket they'd be making. Big job. Big bill. Subscribe from wherever you pick up new podcasts. TRANSCRIPT Gene Hamm, thank you for joining me. For those people who don't know much about Digichief, could you gimme the elevator pitch on what you guys do?  Gene Hamm: Absolutely. Thanks Dave. Long-time listener, first-time caller. Am I the first one to say that?  Probably not, among the first.  Gene Hamm: My kids always say I've got a lot of dad jokes, so I oh, no, I won't bore with that. But thanks for having me today. I'm Gene Hamm, one of the founders of Digichief.  In a nutshell we're a content solutions provider. Basically, a one-source solution for all things content. We work in a number of capacities. We have a white labeled solution for data feeds for those clients who want to control the designs themselves. Or we can provide an integrated solution with HTML5, our widgets for clients that don't want to do the heavy lifting on the design. We already have it baked into our APIs, and so we've built up a library of content over the years. All the staples, weather news, sports info, that sort of thing. We also have some short-form, video series, and some other products that we work as distribution partners, with digital art, things like that. But in a nutshell, we aggregate, we curate, and we create content for you, and we provide it in a consistent manner. We take care of the licensing, and we keep up with the inevitable changes in the source, data feeds, and put it out in a highly scalable, cloud infrastructure. So I would say in the early days or earlier days of digital signage, a lot of companies, I shouldn't say a lot because there weren't many, and there still aren't that many, but the companies that were doing the sort of work that you do, I would describe as aggregators that they were collecting and harmonizing data feeds from news gathering organizations, government organizations like National Weather Service and so on, and getting in a format that's structured, reliable and all those sorts of things so that CMS companies or end users could tap into your feeds and have something that's reliable, organized, and curated to some degree. Is that a fair way of describing things?  Gene Hamm: That is a fair assessment, and I think it's evolved over time. I think early on, it was basically, just kind of an aggregation model. We actually started the company, it's an offshoot of another company we'd started back in the 90s where we worked in the broadcast television space, where we were doing lower third tickers, turnkey systems.  So kinda like Chiron?  Gene Hamm: Yeah, we were third-party developers for Chiron. So we worked a lot with Chiron early on, but a lot of the stuff you saw on the lower thirds and newscasts around the country was our stuff.  The dreaded tickers.  Gene Hamm: The dreaded tickers that kind of blew up in the 90s, yeah. We did news headlines, we were doing integrations with AP Weather. We actually ended up doing elections, school closings, and internet chat. We were all over the board on that.  So that's how we got our feet wet on integrating and aggregating content. In the mid 2000s, we saw the digital signage kind of take off, and we said, look, we've already got these connections with these sources, so why don't we just license these and license this vertical? So that's kind of how it started, but it's evolved over time. We certainly still do that and provide those in a consistent format, but then it's also moved into kind of bespoke projects where people will say, we've got this data, we've got, we want this, maybe we have to go out and do research on specific topics for “Cold weather starting tips for Automotive Dealerships”, things like that. So there's really a research arm to it that we can go out and create stuff for custom projects.  So if you had to give a percentage of from a third party versus what you guys are developing internally, what roughly would that be?  Gene Hamm: I would say about 60 to 70% of it is aggregating. All the staples, traffic, transit, flight data, news headlines, sports scores, the stuff that people want to display most often. So yeah, I would say roughly 60 to 70% of it, and then the other stuff is, a lot of stuff on the infotainment route is data-based that we've created over time and this could be for like “This day in history” trivia, fun facts, jokes, clean jokes of the day, holidays, whimsical, eye-catching things to get eyeballs up on the screen.  The challenge I've always seen with using third-party sources for things like tickers and full-screen presentations, whether it's from the AP, Canadian Press, or Reuters, is that they typically don't write headlines for digital signage or digital at home or anything else, and they don't even really do it in a lot of cases online. So what you end up with are headlines that don't really say anything. It'll say, “This week's top news is this…” and that'll show up on screens. I see it on broadcast still, and I'm going, why are you even using this? Why don't you curate stuff that you know has fully formed thoughts and says in a headline what you need to know versus kind of a teaser?  Have you guys struggled with that, or has it gotten better?  Gene Hamm: We've absolutely run into that. You're speaking to the choir here. We've knocked our head against the wall so many times, and I just think that for these news organizations, digital signage is an afterthought. Believe me, over the last 20 years, we've seen so many stories come out that we just scratch our heads, and I've had conversations with the editors to try to plead my case, and it just goes on deaf ears.  So basically what we have to do with our news, we have two formats. We have one that's filtered, and we've got lookups and intelligence written in where if something comes out misformed or certain key phrases, we just kick them out. And then we have basically a curated version where we actually go in and manually approve and post. We look at the image, we look at the images is another problem with it, but we look at the story, and we say, this doesn't make sense, or maybe we change a few words around to make it flow better and fit into a kind of concise title and description. So yeah, it's been a big problem and honestly it hasn't gotten any better in my viewpoint. Does AI present an opportunity to clean things up? Because I will take the odd story that I write and dump it into Claude and just say, “Give me 10 suggested headlines” and it'll knock out ten headline headlines in 15 seconds, and I'll look at it and go, oh, that one's pretty good and I'll take that one and maybe massage it a little bit. But it does a pretty good job with that sort of thing.  Gene Hamm: It absolutely will be a tool that we can utilize, and we're certainly looking into it right now to try to inject on our backend tools that you can request a specific, character-limited title that makes sense. One of the nuances to AI, which I know you're aware of, is that it's all in the phrasing of how you ask the question for the format that you wanted back in.  Prompt engineering.  Gene Hamm: Yeah. It's an art in itself, and what we see is that we think that AI can help this curation service to look at the headlines that we're getting and spit them out in more of a usable, readable, concise form.  But it's not gonna be autonomous anytime soon.  Gene Hamm: We'll see.  Yeah, not reliably autonomous, it's still gonna give you some weird headlines and all that, but then again, you could hire somebody and they'll give you weird headlines.  Gene Hamm: That's true. That's absolutely true. We try to say that our Soft News, which is our curated version, and we try to bill it as G-rated content that's not going to tick somebody off, but that's next to impossible these days because whatever you think is G-rated and is not going to satisfy everyone. We try to stay away from the political end of it, but there's always gonna be somebody that's offended. Yeah. I've talked to a few people who just said, you know what, we don't even do politics on our feeds anymore, or what we show on our screens, because somebody's gonna be irritated, somebody's gonna complain, and it's just not worth it.  Gene Hamm: Oh, the stories I can tell. It's funny. We have a custom bad word filter for stuff that we don't want to come across in the AP and so we've built that over time, and I could never let that see the light of day that the things that we've seen come across the wire that we now omit. Even the images as well. There are a lot of times we'll get images that don't really explain the story, it doesn't make sense, maybe they aren't centered on the right focal point of the image, and we think maybe AI could definitely benefit, maybe being able to zone in on what the main cue is of the image that we get with the AP stories or any of the news images.  Have the demands and the uses, usage trends evolved through the years, like when I got into digital, more than 25 years ago now, there weren't really even smartphones, and the internet was still fairly new-ish, and you could have public screens in elevators or walkways or shopping malls or whatever that were running news and weather on there, and those would be a primary source for that information, you fast forward to now, and you can't get away from news, you can't get away from weather data, that sort of thing.  I've always wondered, do those things need to be on screens anymore?  Gene Hamm: That's definitely a good debatable topic. There are so many of these black screens in our hands that fight for attention. We work in the automotive space in dealer showrooms and you walk into the showroom there and people are in the waiting area, and they've got screens up with content on it, news headlines, weather, things like that, and everybody is looking at their phone. So you're always thinking how do we compete with getting eyeballs up on the screen to get the messaging and whatnot for the client, as opposed to the ubiquitous news headlines and things like that. So yeah, it's something that our clients definitely have to deal with. Is that something you coach to, to tell both your resellers and your end users, that it's important to really think through what you're using in terms of content feeds or your content mix so that it's hyper relevant and contextual to where you are versus just “We need stuff to run on this lower third” or “We need stuff to run in between our dealer promotional messages” or whatever it may be, whatever the venue is. Gene Hamm: Absolutely. As you said, it's all in the content mix. If you're trying to get eyeballs up there on the screen, you gotta have relevant hyper-local content, whether that be local traffic maps or local sports scores or things like that for the market.  But yeah, the dwell time and how long the content is on the screen, you want to get the eyeballs up there and then move on to what your marketing message is. So it's definitely a delicate balance between, you can't just inundate someone with all the news, all weather. You definitely have to make it in short, concise forms because people's attention spans go elsewhere. They go back to their phone or something else.  A few months ago, you announced a partnership with a company called Stream, and I've done a podcast with those folks and laid out what they do and all that.  How do you work with them, and could you kinda run down what they do and how that's resonating with your user base? Gene Hamm: Yeah, so we met Anthony Nerantzis at one of the trade shows, and he came by and explained his interest. He's kind of a broadcaster, newsroom journalist. So basically, what it is they do is a presenter-led, concise, short-form video of bespoke custom news, right? And it can be catered to the industry.  So if it's medical, financial, or automotive, or what have you. They can go back, write the scripts, and of course, Anthony can describe this company better than I can, so hopefully he's not gonna be mad at me for giving this kind of dissertation. But yeah, I just thought it brought to the table something that we could really customize for our clients, and it's very professional, the workflow is great, you can provide some of the background, what you know the company's looking to do, what type of information they're trying to get across, their team can go back and write a script that's engaging and they can automate the product to put it out on whatever the interval you need, whether it be weekly or monthly.  Originally, when they came out, it was a closed caption type thing with lower third supers on the bottom of the screen and I had mentioned to them, “Hey, there are too many graphics on the screen. Maybe, you might wanna streamline that a little bit.” They did that because they're very good about taking feedback, and now they've moved in. It was more of a no-volume type environment product, and now they've, they're able to do audio voiceover as well from the on-air talent actually speaking and you can actually hear it.  Now they're getting into kind of the marketing communication end of it where, let's say it's a pharmaceutical company or something that wants to talk about things that like the president or the CEO wants to talk about certain things to their employees that they have going on, his team's able to go out and produce that and deliver that information and they can get eyeballs up on the screen, educate and inform the client. It's been very well received and we're also looking to work with them on some of our feeds, whether it's health-related type content, maybe we can work in some of the real, day-to-day, hyper-local information on the tail end of the video segment. Say if it's a medical facility and they're talking about medical health tips, things like that, maybe it comes in and we can integrate with one of our APIs and follow the levels of the flu levels there are for the specific area, so we can really hyper-localize it.  So in a lot of respects, it's a variation on the sort of work that you've been doing, particularly on the custom side of it. But instead of it just being text and visuals, they can do a full video with on-air talent and they do that by green screening, on-air hosts, and then mashing that up with AI so that it's a human talking to you and doing a custom presentation as opposed to an anime avatar look that I think looks ghastly in most cases? Gene Hamm: Absolutely. I think going to the presenter-led approach is advantageous and some of the early ones, like you said, that we've seen are just creepy. But I think what they're doing with their technology is amazing. I think it looks spot on.  Yeah, I've looked at it a couple of times for extended periods, just paying attention to see if it's glitchy at all, and it's very smooth, and if you didn't know, you'd be hard pressed to know, this is AI-generated, but it's absolutely human. But the movements and lips and all that stuff are being massaged through AI.  Gene Hamm: Yeah, and the neat thing about it, too, is just it's so scalable and they can automate it, and they can really like its bespoke content, so they can create the script, have it produce it in very short order. So more recently, you've announced something else called Mercury. Can you walk through what that is?  Gene Hamm: Mercury was created basically to give our users a more robust way to onboard our HTML content. We were getting requests for more of a web portal that gives more granular design choices such as colors, backgrounds, logos, the transitions. They can go in and micromanage the news they wanna see, or the sports they want to see, the duration that it's on the screen, and then, they can compile that into a playlist and then output it to a URL and that URL can be scheduled. It's quite a long time coming. We certainly had HTML55 widgets before, but this just gives people a little bit more granular decisions and a web portal, and then we also thought it was a good way to showcase our widget library. We built up these designs over time. Many of the products that we have, there's multiple designs, and so for, we think it might be a growth area for new prospects, that it lowers the barrier of entry to go out and actually, sign up for a free trial, take a look at, it's an all you can eat type model where we've got all the staples, the news, the weather, the sports, the stocks, the infotainment and we're adding new designs and widgets all the time. I think it's intuitive where we spent well over a year designing the system, and I think it really gives people a way to sample our products and see how it works with their systems.  Could you give an example of how a typical client would use it and what they do? Gene Hamm: Yeah, so they sign up for the product. It's a subscription service, with volume discounts that they can go in, and we've got a kind of smorgasbord of content, a widget library and it's all categorized by, like I said, news, weather, things like that, and they can pick and choose what content they wanna build into a playlist? Now that could be just a single piece of content, whether, say, weather, and they've got a bunch of different designs, whether they wanna do a 5K five-day forecast, if they wanna do a full-screen weather map, they can choose their locations, and then they can output it as a URL that URL can be a plugged into a playlist and that pluglist can have their content or they can massage their own local content, through their own platform, so it just gives them the ability to do this kind of infotainment type stuff in between their other messaging.  But yeah, they can build a playlist with a single asset, or they can build a playlist with 30 and build a longer duration, say, a 20-minute loop if they want. So yeah, that's the typical workflow.  So more normally or in the past, if I were a corporate entity and I had a corporate campus in three cities in South Carolina. If I were buying that from a typical subscription content service or weather provider, it's going to have a certain look and color schemes, everything else, and you can't really deviate from that, versus with Mercury, you can choose your fonts, choose your background, colors, everything else, and tweak it so it fits the way you want, maybe has the company's corporate colors and or just fits in with the overall look of the network.  Is that a clear way of saying this? Gene Hamm: Yeah. To make it very granular, the layout of, let's say, a five-day forecast, the data itself is set on the screen, but all the other elements around it like if they wanted to upload their own. company logo, if they wanna match their corporate colors, they can choose certain fonts that may match what you know they're using. So yeah, they can make different transitions to it, so they can really make granular choices with it to fall in line with what they're looking for, but be on the same thing across the same board. We have stocks, if they wanna put their own company stock up there, they can do that. If they wanna do infotainment like trivia or whatnot, we have a number of different trivia categories that they can choose. So yeah, they can really hyper-localize.  Do you put guardrails in terms of design choices that can be made? Like thinking particularly of font choices and Lord knows we've all seen online, particularly, and less so on digital signage, here somebody decides I'm going to use this font, and it's just the wrong choice.  Gene Hamm: We have chosen a list of fonts that we have in a dropdown box that they can choose from. As you can imagine, this was our initial decision when we debuted this release system a few months ago, and our thought is that we wanna give them these options to an extent, right? So we have several fonts that we think we deem look good, and we certainly can add additional fonts as we go. But yes, I agree there's some god awful fonts up there that we don't think would at the end of the day look great on particular design.  Is this the way to deal with the demand that can scale up so that if you were just doing this through managed services, where you would have companies come to you and say, “Hey, we would like a live custom feed that presents ou  weather and other information in these fonts, this background and everything else.” That's hard to do and hard to charge because if it's a one-off, you gotta charge a lot more for it, versus a service where you log in and you do it yourself, by and large, that makes it possible to do more.  Gene Hamm: Yeah, I think so. I think with the pricing model, how we have it, they can use everything. It's all you can eat, in terms of all these different designs and content categories that they can go in and it's not gonna cost them anymore if they put the news or the weather up there. I think the value proposition to Mercury is that we're doing the heavy lifting on the backend, and that these local networks don't have to go out and find different sources, and like you mentioned, the National Weather Service.  Early on, we were integrating with the National Weather Service and that got to be just an overwhelming task because of stages and formats, and changes in the designs and things like that. It just made more sense for us to go out and get an aggregated list. Actually, we have a couple of different aggregated services. So, like a lot of our staples, we have a primary source and a backup source. So if one goes inevitably, these sources have issues, and if one goes down. It really streamlines the whole process.  Has the whole business of getting data from different sources improved? Have they started to, or maybe not started, but long since understood that you can't keep changing the structure. You've gotta stick to something.  Gene Hamm: Yes and no. With sports specifically, they're good about giving us a heads up when things are gonna change. In the olden days, we would find out about it after it happened. So I think a lot of the source APIs that we have do a good job of giving us kind of a change. But there's repercussions. If they do a full change of their structure, we have to integrate that, and if it has any changes to how we do content, we have to let our clients know, and we have to make sure the widgets are changed. We have to make sure they know that the structure's changed.  During the pandemic, we really moved our cloud infrastructure from one cloud service to another. We added a lot of data points to our structure, and so that was really an uphill battle in terms of having to communicate to our current client base that had already done the design work and had already integrated with our APIs to let them know that's coming. So we don't take these things lightly and we've communicated to our sources over time about the repercussions to this. You can't just pull the trigger and give us a two-week notice.  What about social media? If I go back 10-15 years, there were a lot of subscription content providers and CMS companies developing widgets so that you could display Twitter (now X) or Facebook post or whatever maybe on screens and I think over time people realize, oh boy, that's a dangerous thing to do unless you've got somebody sitting right on top of it all the time.  Gene Hamm: It's absolutely the case. In fact, we were one of the ones early on that were doing native integrations with the APIs from Twitter and Facebook and whatnot, and it got to be a full-time job for our developers, changing not only the licensing, but the structure, and we finally threw in the towel on it and outsourced it to a company where that's all they do, and so we work with this particular company, and they take care of it. They've got a team of developers that don't do anything else, and they keep up on all the backend changes, the licensing, and so we're able to not only provide Facebook, Instagram, Twitter or X, LinkedIn, all this as a concise data feed with different data points and assets, and then we also have an HTML version that integrates with it.  So yeah, we've definitely gone the route of outsourcing that to someone who could keep up with it. Is there a most popular resource and one that you thought would have traction and that just never worked out, and you've since dropped or rarely see sold? Gene Hamm: About a year ago, we started with a health API, so seasonal and patient level data, and by seasonal, we mean pollen which is a big one and we have multiple sources for that. But, RSV levels, COVID-19 numbers, cold and cough, and flu. And then we can even get granular with patients. We can go and say a zip code in the United States, and say, what are the ten highest levels of obesity? And they can customize a message or an ad campaign towards that. Those particular zip codes we thought would take off at least the patient-level stuff and it was just really slow out of the gate. We've had a lot of interest and we've made a lot of presentations, but I think there are a lot of these companies that are still trying to figure out how they might use it. Flight data is one that we work with, and we have some clients using it.  There are certain sources that are very expensive to keep up with. That's something that we thought would be selling more than it does. A lot of times, the people that you know that put the flight data up are probably going directly to the source as opposed to going through somebody like us.  Is there one that everybody uses, or almost everybody?  Gene Hamm: Everybody uses weather, of course, that's the big one. Everybody uses sports scores, and everybody uses news. That's news, weather, sports are the big dogs.  Just a couple of final questions. Where are you guys based, and how big is your company? I'm thinking you don't have that big of a headcount because you don't need to, because you're using external resources. Gene Hamm: Yeah, so we're based in Lexington, Kentucky. We also have partners spread across the world. But I got a partner in California. There are a few of us here, and then we've got a couple in Ukraine. So we've been working with a couple of developers who are now employees in Ukraine, well before the war. So it's been interesting seeing that side of it from an employee.  It gives you a perspective on a drone flying over, and bombings and things like that. So there are five of us. We run a small operation, but like you said, we don't really need an extensive team. We certainly have worked with or contracted out some design work in terms of the graphical design. We've worked with the same designers for well over a decade. All right, so thank you. If people wanna find out more, it's just Digichief.com, right?  Gene Hamm: Yeah, Digichief.com, and then if someone wants to sample Mercury for a free trial, there's a Mercury link on there that they can go and sign up for, and give it a whirl.  Gene, thank you. Gene Hamm: Thank you, Dave. I appreciate your time.

Kacper "prefetcher" Staroń created the PinkSea oekaki BBS on top of the AT Protocol. He also made the online multiplayer game MicroWorks with Noam "noam 2000" Rubin. He's currently studying Computer Science at the Lublin University of Technology. We discuss the appeal of oekaki BBSs, why and how PinkSea was created, web design of the early 2000s, flash animations, and building an application on top of the AT Protocol. Prefetcher Bluesky Github Personal site Microworks (Free multiplayer game) PinkSea and Harbor PinkSea PinkSea Bluesky Account PinkSea repository Harbor image proxy repository Harbor post from bnewbold.net imgproxy (Image proxy used by Bluesky) Early web design Web Design Museum Pixel Art in Web Design Kaliber10000 Eboy Assembler 2advanced epuls.pl (Polish social networking site) Wipeout 3 aesthetic Restorativland (Geocities archive) Flash sites and animations My Flash Archive (Run by prefetcher) dagobah Z0r Juicy Panic - Otarie IOSYS - Marisa Stole the Precious Thing Geocities style web hosts Neocities Nekoweb AT Protocol / Bluesky PDS Relay AppViews PLC directory Decentralized Identifier lexicon Jetstream XRPC ATProto scraping (List of custom PDS and did:web) Tools to view PDS data PDSls atp.tools ATProto browser Posters mentioned vertigris (Artist that promoted PinkSea) Mary (AT Protocol enthusiast) Brian Newbold (Bluesky employee) Oekaki drawing applets Tegaki chickenpaint Group drawing canvas Drawpile Aggie Other links Bringing Geocities back with Kyle Drake (Interview with creator of Neocities) firesky.tv (View all bluesky posts) ATFile (Use PDS as a file store) PinkSky (Instagram clone) front page (Hacker news clone) Smoke Signal (Meetup clone) -- Transcript You can help correct transcripts on GitHub. Intro [00:00:00] Jeremy: Today I am talking to Kacper Staroń.  He created an oekaki BBS called PinkSea built on top of the AT protocol, and he's currently studying computer science at the Lublin University of Technology. We are gonna discuss the appeal of oekaki BBS, the web design of the early 2000s, flash animations, and building an application on top of the AT protocol. Kacper, thanks for talking with me today. [00:00:16] Prefetcher: Hello. Thank you for having me on. I'm Kacper Staroń also probably you know me as Prefetcher online. And as Jeremy's mentioned, PinkSea is an oekaki drawing bulletin board. You log in with your Bluesky account and you can draw and post images. It's styled like a mid to late 2000s website to keep it in the spirit. What's an oekaki BBS? [00:00:43] Jeremy: For someone who isn't familiar with oekaki BBSs what is different about them as opposed to say, a photo sharing website? [00:00:53] Prefetcher: The difference is that a photo sharing website you have the image already premade be it a photo or a drawing made in a separate application. And you basically log in and you upload that image. For example on Instagram or pixiv for artists even Flickr. But in the case of an oekaki BBS the thing that sets it apart is that oekaki BBSes already have the drawing tools built in. You cannot upload an already pre-made image with there being some caveats. Some different oekaki boards allow you to upload your already pre-made work. But Pinksea restricts you to a tool called Tegaki. Tegaki being a drawing applet that was built for one of the other BBSes and all of the drawing tools are inside of it. So you draw from within PinkSea and you upload it to the atmosphere. Every image that's on PinkSea is basically drawn right on it by the artists. No one can technically upload any images from elsewhere. How PinkSea got started and grew [00:01:56] Jeremy: You released this to the world. How did people find it and how many people are using it? [00:02:02] Prefetcher: I'll actually begin with how I've made it 'cause it kind of ties into how PinkSea got semi-popular. One day I was just browsing through Bluesky somewhere in the late 2024s. I was really interested in the AT Protocol and while browsing, one of the artists that I follow vertigris posted a post basically saying they'd really want to see something a drawing canvas like Drawpile or Aggie on AT Protocol or something like an oekaki board. And considering that I was really looking forward to make something on the AT Protocol. I'm like, that sounds fun. I used to be a member of some oekaki boards. I don't draw well but it's an activity that I was thinking this sounds like a fun thing to do. I'm absolutely down for it. From like, the initial idea to what I'd say was the first time I was proud to let someone else use it. I think it was like two weeks. I was posting progress on Bluesky and people seemed eager to use it. That kept me motivated. And yeah. Right as I approached the finish I posted about it as a response to vertigris' posts and people seemed to like it. I sent the early version to a bunch of artists. I basically just made a post calling for them. Got really positive feedback, things to fix, and I released it. And thanks to vertigris the post went semi-viral. The launch I got a lot of people which I would also tie to the fact that it was right after one of the user waves that came to Bluesky from other platforms. The website also seemed really popular in Japan. I remember going to sleep, waking up the next day, and I saw like a Japanese post about PinkSea and it had 2000 reposts and 3000 likes and I was just unable to believe it. Within I think the first week we got like 1000 posts overall which to me is just insane. For a week straight I just kept looking at my phone and clicking, refresh, refresh, refresh, just seeing the new posts flow in. There was a bunch of like really insane talented artists just posting their works. And I just could not believe it. PinkSea got I'd say fairly popular as an alternative AppView. People seem to really want oekaki boards back and I saw people going, oh look, it's like one of those 2000s oekaki boards! Oh, that's so cool! I haven't seen them in forever! The art stands out because it's human made [00:04:58] Prefetcher: And it made me so happy every single time seeing it. It's been since November, like four months, give or take. And today alone we got five posts. That doesn't sound seem like a lot but given that every single post is hand drawn it's still insane. People go on there and spend their time to produce their own original artworks. [00:05:26] Jeremy: This is especially relevant now when you have so much image generation stuff and they're making images that look polished but you're kind of like well... did you draw it? [00:05:39] Prefetcher: Yeah. [00:05:40] Jeremy: And when you see people draw with these oekaki boards using the tools that are there I think there's something very human and very nostalgic about oh... This came from you. [00:05:53] Prefetcher: Honestly, yeah. To me seeing even beginner artists 'cause PinkSea has a lot of really, really talented and popular people (and) also beginner artists that do it as a hobby. Ones that haven't been drawing for a long time. And no matter what you look at you just get like that homely feeling that, oh, that's someone that just spent time. That's someone that just wanted to draw for fun. And at least to me, with generative AI like images it really lacks that human aspect to it. You generate an image, you go, oh, that's cool. And it just fades away. But in this case you see people that spent their time drawing it spent their own personal time. And no matter if it's a masterpiece or not it's still incredibly nice to see people just do it for fun. [00:06:54] Jeremy: Yeah. I think whether it's drawing or writing or anything now more than ever people wanna see something that you made yourself right? They wanna know that a human did this. [00:07:09] Prefetcher: Yeah. absolutely. [00:07:11] Jeremy: So it sounds like, in terms of getting the initial users and the ones that are there now, it really all came out of a single Bluesky posts that an existing artist (vertigris) noticed and boosted. And like you said, you were lucky enough to go viral and that carried you all the way to now and then it just keeps going from there, [00:07:36] Prefetcher: Basically if not for vertigris PinkSea (would) just not exist because I honestly did not think about it. My initial idea on making something on ATProto and maybe in the future I'll do something like that would be a platform like StumbleUpon -- Something that would just allow you to go on a website, press a button, and it gets uploaded to your repo and your friends would be able to see oh -- you visited that website and there would be an AppView that would just recommend you sites based on those categories. I really liked that idea and I was dead set on making it but then like I noticed that post (from vertigris) and I'm like, no, that's better. I really wanna make that. And yeah. So right here I want to give a massive shout out to vertigris 'cause they've been incredibly nice to me. They've even contributed the German translation of PinkSea which was just insane to me. And yeah, massive shout out to every single other artist that, Reposted it, liked it, used it because, it's all just snowballed from there and even recently I've had another wave of new users from the PinkSea account. So there are periods where it goes up and it like goes chill -- and then popular again. Old internet and flash [00:08:59] Jeremy: Yeah. And so something that you mentioned is that some people who came across it they mentioned how it was nostalgic or it looked like the old oekaki BBSs from the early internet. And I noticed that that was something that you posted on your own website that you have an interest in that specifically. I wonder what about that part of the internet interests you? [00:09:26] Prefetcher: That is a really good question. Like, to me, even before PinkSea my interests lie in the early internet. I run on Twitter and also on Bluesky now an account called My Flash Archive, which was an archive of very random, like flash animations. And I still do that just not as much anymore 'cause I have a lot of other things to do. I used to on Google just type in Flash and look through the oldest archived random folders just having flash videos. And I would just go over them save all of that or go on like the dagobah or Z0r or swfchan. 'cause the early internet to me, it was really like more explorative. 'cause like now you have, people just concentrated in those big platforms like Twitter, Instagram, Facebook, whatever. And back then at least to me you had more websites that you would just go on, you would find cool stuff. And the designs were like sometimes very minimal, aesthetically pleasing. I'd named here one of my favorite sites, Kaliber10000 which had just fantastic web design. Like, I, I also spend a lot of time on like the web design museum just like looking at old web design and just in awe. My flash archive on Twitter at least got very popular. I kind of abandoned that account, but I think it was sitting at 12,000 followers if not more? And showed that people also yearn for that early internet vibe. And to me it feels really warm. Really different from the internet nowadays. Even with the death of flash you don't really have interactive experiences like it anymore. 'cause flash was supposed to be replaced by HTML5 and JavaScript and whatever but you don't really make interactive experiences that just come packaged in a single file like flash. You need a website and everything. In flash, it just had a single file. It could be shared on multiple sites and just experienced. That kind of propelled my interest. Plus I, I dunno, I just really like the old internet design aesthetics it really warms me (and really close..?) Flash loops [00:12:01] Jeremy: The flash one specifically. Were they animations or games or was there a specific type of a flash project that spoke to you? [00:12:15] Prefetcher: Something we call loops. Basically, it's sometimes animations. 'cause, surprisingly while I like flash games they weren't my main collection. What spoke to me more were loops. Basically someone would take a song, find a gif they liked, and they would just pair it together. Something like YTMND did. At least from loops I found some of my favorite musical artists, some of my favorite songs, a lot of interesting series, be it anime or TV or whatever. And you basically saw people make stuff about their favorite series and they would just share it online. I would go over those. For example, a good website as an example is z0r.de, which is surprisingly still active and updated to this day. And you would see people making loops about members of that community or whatever they like. And you would for example see like 10 posts about the same thing. So you would know someone decided to make 10 loops and just upload them at once. And yeah, to me, loops basically were like, I mean, they weren't always the highest quality or the most unique thing, but you would see someone liked something enough that they decided to make something about it. And I always found that really cool. I would late at night just browse for loops and I'm like, oh, oh, this series, I remember it. I liked it (laughs)! But of course flash games as well. I mean, I used to play a lot when I was younger, but specifically loops, even animations and especially like when someone took like their time to animate something like really in depth. My favorite example is, the music video to a song by the band Juicy Panic called Otari. Someone liked that song enough that they made an entire flash animated music video, which was basically vectorized art of various series like Azumanga Daioh or Neon Genesis Evangelion as well, and other things. And it was so cool, at least to me, like a lot of these loops just basically have an intense, like immense feeling on me (laughs). I just really liked collecting them. [00:14:38] Jeremy: And in that last example, it sounded more like it was a complete music video, not just a brief loop? [00:14:45] Prefetcher: No, it was like a five minute long music video that someone else made. [00:14:48] Jeremy: Five. Oh my gosh. [00:14:49] Prefetcher: Yeah. You would really see people's creativity shine through on just making those weird things that not a lot of people have seen, but you look at it and it's like, wow. It's different than YouTube (Sharable single file, vectorized) [00:15:01] Jeremy: It's interesting because you can technically do and see a lot of these things on, say, YouTube today, but I think it does feel a little different for some reason. [00:15:16] Prefetcher: It really is. Of course I'm not denying on YouTube you see a lot of creative things and whatever. But first and foremost, the fact that Flash is scalable. You don't lose the quality. So be able to open, I don't know, any of the IOSYS flash music videos for like their Touhou songs and the thing would just scale and you would see like in 4K and it's like, wow. And yeah, the fact that on YouTube you have like a central place where you just like put something and it just stays there. Of course not counting reuploads, but with Flash you just had like this one animation file that you would just be able to share everywhere and I don't know, like the aspect of sharing, just like having those massive collections, you would see this flash right here on this website and on that website and also on this website. And also seeing people's personal collections of flash videos and jrandomly online and you would also see this file and this file that you haven't seen it -- it really gives it, it's like explorative to me and that's what I like. You put in the effort to like go over all those websites and you just like find new and new cool stuff. [00:16:32] Jeremy: Yeah, that's a good point too that I hadn't thought about. You can open these files and you have basically the primitives of how it was made and since, like you said, it's vector based, there's no, oh, can you please upload it in 1080 p or 4K? You can make it as big as you want. [00:16:53] Prefetcher: Yeah. Web design differences, pixel art, non-responsive [00:16:55] Jeremy: I think web design as well it was very distinct. Maybe because the tools just weren't there, so a lot of people were building things more from scratch rather than pulling a template or using a framework. A lot of people were just making the design theirs I think rather than putting words on a page and filling into some template. [00:17:21] Prefetcher: Honestly, you raise a good point here that I did not think much about. 'cause like nowadays we have all of this tooling to make web design easier and you have design languages and whatnot. And you see people make really, in my opinion, still pretty websites, very usable websites on top of that. But all of them have like the same vibes to them. All of them have like a unified design language and all of them look very similar. And you kind of lose that creativity that some people had. Of course, you still find pretty websites that were made from scratch. But you don't really get the same vibes that you did get like back then. Like my favorite, for example, trend that used to be back on like the old internet is pixel art in web design. For example, Kaliber10000, or going off the top of my head, you had the Eboy or all the sites and then Poland, for example, ... (polish website) those websites use minimal graphics, like pixel graphics and everything to build really interesting looking websites. They had their own very massive charm to them that, I don't know, I don't see a lot in more modern internet. And it's also because back then you were limited by screen size, so you didn't have to worry about someone being on a Mac with high DPI or on a 32x9 monitor like I am right now. And just having to scale it up. So you would see people go more for images, like UI elements, images instead of just like building everything from scratch and CSS and whatnot. So, yeah, internet design had to accommodate the change. So we couldn't stay how it was forever 'cause technology changed. Design language has changed, but to me it's really lost its charm. Every single website was different, specific, the web design had like this weird form, at least on websites where it was like. I like to call it futuristic minimalism. They looked very modern and also very minimal and sort of dated. And I dunno, I just really like it. I absolutely recommend checking, on the web design museum fantastic website. I love them and the pixel art in web design sub page. Like those websites to me they just look fantastic. [00:19:52] Jeremy: Yeah, and that's a good point you brought up about the screen sizes where now you have to make sure your website looks good on a phone, on a tablet, on any number of monitor sizes. Back then in the late 90s, early 2000s, I think most people were looking at these websites on their 4x3 small CRT monitors. [00:20:20] Prefetcher: My favorite this website is best viewed with an 800 by 600 monitor. It's like ... what? [00:20:28] Jeremy: Exactly. Even if you open your personal site now the design is very reminiscent of those times and it looks really cool but at the same time on a lot of monitors it's a small box in the middle of the monitor, so it's like -- [00:20:49] Prefetcher: I saw that issue, 'cause I was making it on a 1080p monitor and now I have a 32x9 monitor and it does not scale. I've been working on reworking that website, but, also on the topic of my website, I, I wanna shout out a website from the 2000s that still exists today. 'cause, my website was really inspired by a website called Assembler. And Assembler, from what I could gather, was like a net art or like internet design collective. And the website still works to this day. You still had like, all of their projects, including the website that my website was based off of. [00:21:28] Jeremy: Yeah, I mean there, there definitely was an aesthetic to that time. And it's probably, like you said, it's probably people seeing someone else's site in this case, what, what did you call it? Assem? Assembler? [00:21:42] Prefetcher: Assembler. [00:21:42] Jeremy: Yeah. You see someone else's website and then maybe you try to copy some of the design language or you look at the HTML and the CSS and I mean, really at the time, these websites weren't being made with a ton of JavaScript. There weren't the minifiers, so you really could view source and just pull whatever you wanted from there. [00:22:06] Prefetcher: We also had those design studios, design agencies, notably 2advanced which check in now, their website still works, and their website is still in the same aesthetic as it was those 20 so years ago just dictating this futuristic design style that people really like. 'cause a lot of people nowadays also really like this old futurism minimalism for example a lot of people still love the Wipeout 3 aesthetic that was designed by one of my favorite studios overall the designers republic. And yeah, it's just hard for me to explain, but it feels so soulful in a way. [00:22:53] Jeremy: I think there are some trade offs. There's what we were talking about earlier with the flexibility of screen size. But there used to be with a lot of websites that used Flash, there used to be these very elaborate intros where the site is loading and there's these really neat animations. But at the same time, it's sort of like, well, to actually get to the content, it's a bit much, but, everything is a trade off. [00:23:25] Prefetcher: People had flash at their disposal and they just wanted to make, I have the tooling, I'm going to use all of the tooling and all of it. [00:23:33] Jeremy: Yeah. Yeah. but yeah, I definitely get what you're saying where when I went to make my own website I made it very utilitarian and in some ways boring, right? I think we do kind of miss some of what we used to have. [00:23:54] Prefetcher: I mean, in my opinion, utilitarian websites are just as fine. Like in some cases you don't really need a lot of flashy things and a lot of very modern very CPU intensive or whatever animations. Sometimes it is better to go on a website and just like, see, oh, there's the play button and that's it. [00:24:17] Jeremy: Yeah. Well definitely the animations and the intro and all that stuff. I guess more in terms of the aesthetics or the designs. It's tricky because there's definitely people making very cool things now things that weren't even possible back then. But it does feel like maybe the default is I'll pick this existing style sheet or this existing framework and just go with that. [00:24:47] Prefetcher: A lot of modern websites just go for similar aesthetics, similar designs, which they aren't bad, but they are also very just bland. They, they are futuristic, they are very well designed. But when you see the same website. The same -- five websites have the same feel. And this is especially, at least in my opinion, visible with websites built on top of NextJS or other frameworks. And it just feels corporate kind of dead. Like someone just makes a website that they want to sell something to you and not for fun. [00:25:26] Jeremy: With landing pages especially it's like, wow, this looks the same as every other site, but I guess it must work. [00:25:38] Prefetcher: It works. And it really cuts down on development time. You don't need to think much about it. You just already have a lot of well-established design rules that you just follow and you get a cohesive and responsive design system. Designing the PinkSea look and feel [00:25:56] Jeremy: Let's talk about that in connection with PinkSea. What was your thinking when you designed how PinkSea would look and feel? [00:26:06] Prefetcher: Honestly, at first I have to admit I looked at other websites. I looked at Bluesky first and foremost. I looked at, front page. I looked at Smoke Signal, and I thought that I might also build something that's modern and sleek and I sketched it out in an application and I showed it to some friends. One of them suggested I go for more like a 2000 aesthetic. I'm like, yeah, okay. I like that. As the website was built, I just saw more and more of how much I feel this could sit with others. Especially with the fact that it's an oekaki page an oekaki BBS and as you scroll through oekaki has a very distinct style to it. And as you scroll and you see all of those, pixel shaded, all those dithered images, non anti-aliased pens and whatnot. It feels really really cohesive somehow with the design aesthetic. But of course, PinkSea in itself is a modern website. Like if you were to go to my PinkSea repository. It's a modern website built up on top of Vue3, which talks via like XRPC API calls in real time and it's a single page app and whatever. That's kind of the thing I merged the modern way of making sites with a very oldish design language. And I feel, in my opinion, it somehow just really works. And especially it sets PinkSea apart from the other websites. It gives it that really weird aesthetic. You would go on it and you would not be like, oh, this is a modern site that connects with a modern protocol on top of a big decentralized network. This is just someone's weird BBS stuck in the 2000s that they forgot to shut down. (laughs) [00:28:00] Jeremy: Yeah. And I think that's a good reminder too, that when people are intentional about design, the tools we have now are so much better than what we used to have. There's nothing stopping us from making websites that when people go to them they really feel like something's different. I know I did not just land on Instagram. [00:28:27] Prefetcher: Yeah. And making PinkSea taught me that it's really easy to fall into that full string of thought that every site has to look modern. Because I was like, oh yeah, this is a modern protocol, a modern everything, and it has to look the part. It has to look interesting to people and everything. And after talking with a bunch of friends and other people and just going, huh, that's maybe like the 2000s isn't as bad as I thought. And yeah, the website especially it's design people seem to just really like it. Me too. I, I just absolutely love how PinkSea turned out it is really a reminder that you don't need modernness in web design always. And people really appreciate quirky looking pages, so to say, quirky like interesting. [00:29:23] Jeremy: I interviewed the, the creator of Neocities which is like kind of a modern version of GeoCities and yeah, that's really what one of the aspects that I think makes things so interesting to people from that era is, is that it really felt like you're creating your own thing, and not just everything looks the same. The term I think he used is homesteading. You're taking care of your place and it can match your sensibilities, your style, your likes, rather than having to, like you said, try to force everything to be this, this sort of base modern, look. The old spirit of the internet is coming back [00:30:08] Prefetcher: I mean Neocities and by extension also Nekoweb are websites that I often when I don't have much to do -- I like just going through them because you see a bunch of people just make their own places. And you see that even in 2025 when we have those big social media sites. You have platforms where you can get a ton of followers. You can get a ton of attention and everything. People to some extent still want that aspect of self-expression. They want to be able to make something that's uniquely theirs and you see people just make just really amazing websites build insane things on those old Geocities-like platforms using nothing but a code editor. You see them basically just wanting thing to express, oh, that's mine and no one else has it. So to say that's why. Yeah. I feel like to some extent the old school train of thought when it comes to the internet is slowly coming back. Especially with the advent of protocols like ATProto. And you'll experience more websites that just allow people to make their own homes on the internet. Cause in my opinion, one of the biggest problems is that people do not really want to register on a lot of platforms. 'cause you already have this place where you get all of your followers, you have all of your connections, and then you want to move and then you'll lose all of your connections and everything. But with something like ATProto, you can use the social graph of, for example, Bluesky. I want to add followers on PinkSea. So for example, you have an artist that has like 30,000 followers for example, I can just click import my following from Bluesky. And just like that they would already get all of the artists that they follow on Bluesky already added as followers on PinkSea. And for example, someone else joins and they followed that big artist and they instantly followed them on PinkSea as well. I think that we are slowly coming back to the advent of people owning their place online. PinkSea and ATProto (PDS) [00:32:24] Jeremy: Yeah. So let's talk a little bit more about how PinkSea fits into ATProto. For people who aren't super familiar with ATProto, maybe you could talk about how it's split up. You've got the PDS, the relays, the AppView. What are those and how do those fit into what PinkSea is? [00:32:48] Prefetcher: My favorite analogy, ATProto is a massive network, and at least me, when I saw the initial graph I was just very confused. I absolutely did not know what I'm looking at. But let's start with the base building block, something that ATProto wouldn't exist with. And it's the PDS. Think of the PDS as like a filing cabinet. You have a bunch of folders in which you have files, so to say. So you have a filing cabinet with your ID, this is the DID part that sometimes shows up and scares people. It's what we call a decentralized identifier. Basically that identifier is not really tied to the PDS, it just exists somewhere. And the end goal is that every user controls their DID. So for example, if your PDS shuts down, you can always move to somewhere else. Still keep like, for example, that you are prefetcher.miku.place. But in that filing cabinet the PDS going back to it you have your own little zone, your own cabinets, and that has your identifier, it's uniquely yours. Every single application on the AT protocol creates data. They create data and they store the data in a structured format called a record. A record is basically just a bunch of data that explains what that thing is, be it a like, a post on Bluesky an oekaki on PinkSea and an upvote on front page, or even a pixel on place.blue. And all of those records are organized into folders in your cabinet. And that folder is named with something we call a collection id. So for example, a like is, if I remember correctly, it's app.bsky.feed.like, so you see that it belongs to Bluesky. The app.bsky part. it's a feed thing, and the same way, PinkSea, for example, the oekaki and PinkSea uses com.shinolabs.pinksea.oekaki with com.shinolabs being the the collective that I use as a, pen name, so to say. PinkSea being, well, PinkSea and oekaki just being the name. It's an oekaki. If you want to see that there are a lot of tools, for example, PDSls or atp.tools or ATProto browser, if you had to go into one of those and you would type in for example, prefetcher.miku.place, you would see all of your records, the things that, you've created on the AT protocol network. Relay [00:35:19] Prefetcher: So you have a PDS, you have your data, but for example, imagine you have a PDS that you made yourself, you hosted yourself. How will, for example, Bluesky know that you exist? 'cause it won't, it's just a server in the middle of nowhere. That's where we have a relay. A relay is an application that listens to every single server. So every time you create something or you delete something, or for example, you edit a post, you delete an oekaki. You create a new, like -- Your PDS, your filing cabinet generates a record of that. It generates an event, something we call a commit. So, anytime you do something, your PDS goes, Hey, I did that thing. And relays function as big servers that a PDS can connect to. And it's a massive shout box. The PDS goes, Hey, I made this. Then the relay aggregates all of those PDSs into one and creates a massive stream of every single event that's going on the network at once. That's also where the name firehose comes from. 'cause the, the end result, the stream is like a firehose. It just shoots a lot of data directly at anyone who can connect to it. And the thing that makes AT Protocol open and able to be built on is that anyone can just go, I want to connect to jetstream1.west.bluesky.network. They just make a connection to it and boom they just get everything that's happening. You can, for example, see that via firesky.tv. If you go to it, you would open it in your browser. Every single Bluesky post being made in real time right directly in your computer. So you have the PDSs that store data, you have the relay that aggregates every, like, builds a stream of every single event on the network. AppViews [00:37:26] Prefetcher: You just get records. You can't interact with it. You can see that someone made a new record with that name, but to a human, you won't really understand what a cid is or what property something else is. That's why you have what we call AppViews. An AppView, or in full an application view is an application that runs on the AT protocol network. It connects to the relay and it transforms the network into a state that it can be used by people. That's why it's called an application view. 'cause it's a, a specialized view into the whole network. So, for example, PinkSea connects, and then it goes, hey, I want to listen on every single thing that's happening to com.shinolabs.pinksea.oekaki, and it sees all of those, new records coming in and PinkSea understands, oh, I can turn it into this, and then I can take this thing, store it in the database, and then someone can connect with a PinkSea front end. And then it can like, transform those things, those records into something that the front end understands. And then the front end can just display, for example, the timeline, the same way Bluesky, for example -- Bluesky gets every single event, every single new file, new record coming in from the network. And it goes. Okay, so this will translate into one more like on this post. And this post is a reply to that post. So I should chain it together. Oh. And this is a new feed, so I should probably display it to the user if they ask for feeds. And it basically just gets a lot of those disjoint records and it makes sense of them all. The end user has a different API to the Bluesky AppView. And then they can get a more specialized view into Bluesky. PinkSea does not store the original images, the PDS does [00:39:26] Jeremy: And so in that example, the PDSs, they can be hosted by Bluesky the company, or they could be hosted by any person. And so PinkSea itself, when somebody posts a new oekaki, a new image, they're actually telling PinkSea to go create the image in the user's PDS, right? PinkSea is itself not the the source of truth I guess you could say. [00:40:00] Prefetcher: PinkSea in itself. I don't remember which Bluesky team member said it, but I like the analogy that AppViews are like Google. So in Google, when you search something, Google doesn't have those websites. Google just knows that this thing is on that website. In the same vein, PinkSea, when you create a new oekaki, you tell PinkSea, Hey, go to my PDS and create that record for me. And then the person owns the PDS. So for example, let's say that in a year, of course I won't do it, but hypothetically here, I just go rogue and I shut down PinkSea, I delete the database. You still own the things. So for example, if someone else would clone the PinkSea repository and go here, there's PinkSea 2. They can still use all of those images that were already on the network. So, AppViews in a way basically just work as a search engine for the network. PinkSea doesn't store anything. PinkSea just indexes that a user made a thing on that server. And here I can show you how to get to it somehow. Those images aren't stored by PinkSea, but instead, I know that the image itself is stored, for example, on pds.example.com, and of course to reduce the load, we have a proxy. PinkSea asks the proxy to go to pds.example.com and fetch the image, and then it just returns it to the user. [00:41:37] Jeremy: And so what it sounds like then is if someone were to create oekaki on their own PDS completely independently of Pink Sea the fact that they had created that image would be sent to one of the relays, and then PinkSea would receive an event that says oh, this person created a new image then at that point your index could see, oh, somebody created a new image and they didn't even have to go through the PinkSea website or call the PinkSea APIs. Is that right? Sharing PDS records with other applications [00:42:14] Prefetcher: Yep. That is exactly right. For example, someone could now go, Hey, I'm making my own PinkSea-like application. And then they would go, I want to be compatible with PinkSea. So I'm using the same record. Or what we call a lexicon, basically describe how records look like. I forgot to mention that, but every single record has an attached lexicon. And lexicons serve as a blueprint. So a lexicon specifies, oh, this has an image, this has a for example, the tags attached to it, a description of the image. Validate that the record is correct, that you don't get someone just making up random stuff. But yeah, someone could just go, Hey, I'm making another website. Let's call it GreenForest for example. And GreenForest is also an oekaki website, but it uses, for example, chickenpaint instead of tegaki but I want to be able to interoperate with PinkSea. so I'm also gonna use com.shinolabs.pinksea.oekaki the collection, the same record, the same lexicon. And for example, they have their own servers and the servers just create regular oekaki records. So for example, GreenForest gets a new user, they log in, create, draw their beautiful image, and then they click upload it. So GreenForest goes to that person's PDS and tells the PDS, Hey, I want to make a new. com.shinolabs.pinksea.oekaki record. The PDS goes okay, I've done it for you. Let me just inform the relay that I did so, relay gets the notification that someone made that new PinkSea oekaki record. And so the main PinkSea instance, pinksea.art, which is listening in on the relay, gets a notification from the relay going, Hey, there is this new oekaki record. And PinkSea goes, sure, I'll index it. And so PinkSea just gets that GreenForest image directly in itself. And in the same vein, someone at PinkSea could draw something in tegaki -- their own beautiful character. And the same thing would happen with GreenForest. GreenForest would get that PinkSea image, that PinkSea record, and index it locally. So the two platforms, despite being completely different, doing completely different things, they would still be able to share images with each other. Bluesky PDS stores other AppView's data but they could stop at anytime [00:44:38] Jeremy: And these images, since they're stored in the PDS, what that would mean is that anybody building an application on ATProto, they can basically use Bluesky's PDS or the user's PDS as their storage. They could put any number of images in there and they could get into gigabytes of images. And that's the responsibility of the PDS and not yourself to keep track of. [00:45:12] Prefetcher: Yes, that can be the case. Of course, there is a hard limit on how big a single upload can be, which is, if I remember correctly, I don't wanna lie, I think it's 50 megabytes, I don't recall there being a hard cap on how big a single repository can be. I know of some people whose repositories are in the single gigabyte digits but this kind of is a thing scares app developers. 'cause you never know when Bluesky the company -- 'cause most people registering, are registering on Bluesky. We don't really know whether Bluesky, the company will want to keep it for free. Forever allow us to do something like that. You already have projects like, for example, ATFile, which just allow you to upload any arbitrary data just to store it, on their servers and they are paying for you. So we'll never know whether Bluesky will decide, okay, our services are only for Bluesky if you want to use PinkSea you have to deal with it. Or whether they go, okay, if you want to use alternative AppViews you have to pay us in order to host them. So, that also leads me to the fact that decentralization is an important part of AT protocol as Bluesky themselves say that they are a potential adversary. You cannot trust them in the long term. Right now they are benign right now, they're very nice, but, we never know how Bluesky will end up in a year or two. So if you want to be in the full control of your data, you need to sadly host it by yourself. And it's honestly really easy in order to do so. There is a ton of really useful online content blogs and whatever. I think I've set up my PDS in 10 minutes on a break between classes and university. But to a person that's non-technical that doesn't know much I'd say around an hour to two hours The liability and potential abuse from running a PDS [00:47:14] Jeremy: Yeah, I think the scary thing for a lot of people is technical or not, is even if it's easy to set up, you gotta make sure it keeps running. You gotta have backups. And so it could be a lot. [00:47:30] Prefetcher: Yeah. This is to be expected by the fact that you're in control of your data. Keeping it secure the same way, for your personal photos or your documents, for example, your master's diploma or whatever. And it's on you to keep your Bluesky interaction secure. On one hand, it's easier to get someone to do it, and I expect in the future we'll get people that are hosting public PDSes I sometimes thought of doing that for PinkSea, just like allowing people to register by PinkSea. But, doing so as a person, you also have to be constantly on call for abuse. So if someone decides to register via PinkSea and do some illicit activities, you are solely responsible for it. PDS and AppView moderation liability [00:48:17] Jeremy: So if they were to upload content that's illegal, for example, it's hosted on your servers so then it's your problem. [00:48:27] Prefetcher: Yeah, it is my problem. [00:48:29] Jeremy: At least the way that it works now, the majority of the people, their PDS is gonna be hosted by Bluesky. So if they upload content that's breaks the law, then that's the Bluesky company's problem at least currently. [00:48:44] Prefetcher: Yeah. That is something that Bluesky has to deal with. But I do believe that in the future we are going to have, more like independent entities just building infrastructure for ATProto, not even the relay it's just like PDSs for people to be able to join the atmosphere, but not directly via Bluesky. [00:49:06] Jeremy: I'm kind of curious also with the current PDSs, if it's hosted by Bluesky, are they, are they moderating what people upload to their PDSs? [00:49:16] Prefetcher: Good question. Honestly, I don't think they're moderating everything 'cause, it's infeasible for them to, for example, other than moderate Bluesky to also moderate PinkSea and moderate front page and whatnot. So it's the obvious responsibility to moderate itself and to report abuse. I'd say that if someone started uploading illicit material, I do not think, and this is not legal advice, I do not think that they would catch on until some point let's say. [00:49:52] Jeremy: I mean, from what you were describing too, it seems like the AppViews would also, have issues with this because if, let's say someone created a PinkSea record in their PDS directly and the image they put in was not an oekaki image, it's instead something pretty illegal in the country that your AppView is hosted then, Wouldn't that go straight to the PinkSea users viewing the website? [00:50:20] Prefetcher: Yes, sadly, this is something that you have to sign up as you're making an AppView and especially one with images. Sooner or later you are going to get material that you have to moderate and it's entirely on you. That's why, you have to think of moderation while you're working on an AppView. Bluesky has an insanely complicated, at least in my opinion, moderation system, which is composable and everything, which I like. But for smaller AppViews, I think it's too much to build the same level of tooling. So you have to rely more on manual work. Thankfully so far the user base on PinkSea has been nothing but stellar. I didn't have to deal with any law breaking stuff, but I am absolutely ready for one day where I'll have to sadly make some drastic moderation issues. [00:51:18] Jeremy: Yeah. I think to me that's the most terrifying thing about making any application that's open to user content. [00:51:29] Prefetcher: I get it, sadly. I'm no stranger to having issues with people, abusing my websites. Because since 2016, my, first major project was a text board based off of, a text board in a video game called DANGER/U/. It was semi-popular, during the biggest spike in activity in like 2017 and 2016, it had in the tens of thousands of monthly visitors. And sadly, yeah, even though it was only text, I've had to deal with a lot of annoying issues. So to say the worst I think was I remember waking up and people are telling me that DANGER/U/ is down. So I log in the activity logs and someone hit me with two terabytes of traffic in a day. There was a really dedicated person that just hated my website and just either spam me with posts or just with traffic. So, yeah, sadly I have experience with that. I know what to expect that's something that you sadly have to sign up for making a website that allows user content. Pinksea is a single server [00:52:42] Jeremy: To my understanding so far, PinkSea is just a single server. Is that right? [00:52:47] Prefetcher: It is a single server. Yeah. [00:52:48] Jeremy: That's kind of interesting in that, I think a lot of people when they make a project, they worry about scaling and things like that. But, was it a case where you just had a existing VPS and you're like, well hopefully this is, this is good enough? [00:53:03] Prefetcher: I actually ordered a new one even though it's not really powerful, but my train of thought was that I didn't expect it to blow up. I didn't expect it to require more than a single VPS with 8 gigabytes of RAM and whatnot. And so far it's handling it pretty well. I do not expect ever to reach the amounts of traffic that Bluesky does, so I do not really have to worry about insane scalability and whatever. But yeah. I thought of it always as a toy project until the day I released it and realized that it's a bit more than a toy project at this point. To this day, I just kind of think that that website even if it were popular, I would never expect it to have -- And in the best, most amazing case scenario, like a hundred posts a day. I do not have to deal with the amount of traffic that Bluesky does. So one VPS it is. [00:53:59] Jeremy: Yeah, that makes a lot of sense. I mean the application is also mostly reads, right? Most people are coming to see the posts and like you said, you get a few submissions a day, but all the read stuff can probably be cached. Harbor image proxy [00:54:15] Prefetcher: Yeah. The heaviest, thing that PinkSea requires is the image proxy harbor, and that's something that right now only runs on that server. It's in Luxembourg. I think that's where my coprovider hosts it but yeah, that gets the most reads. 'cause in most cases, PinkSea, all it does, all you get is reads from a database, which is just, it's a solved problem. It's really lightweight. But with something like image proxying, you have this whole new problem. 'cause it's a lot of data, and you somehow have to send it -- it's enough for me to just host it locally on that PinkSea server and just direct people to it. But sooner or later, I can always just put it behind something like Bunny CDN or whatnot to have it be worldwide. [00:55:09] Jeremy: So Harbor is something I think you added recently. How did the images work before and what is Harbor doing in its place? [00:55:18] Prefetcher: Before I did what a lot of us currently do and I just freeload atop of Bluesky CDN 'cause Bluesky CDN is just open so far. But it's something that personally irked me. 'cause, I want PinkSea to be completely independent of Bluesky Corporation. I, I wanted to persevere even if Bluesky just decides to randomly, for example, close, the CDN to others or the relay to others or the PLC directory in the worst case scenario. So I wanted to make my own CDN more like proxy. You can't really call it a CDN because it's not worldwide. It's just a single server but let's just say image proxy. So Harbor whenever a person goes to PinkSea, they start loading in all of the images and every single image instead of going to, for example, the PDS or to cdn.bluesky.app. They go to harbor.pinksea.art, you get attached the identifier of the user and what we call a content identifier. Every single, thing uploaded to a PDS has an attached content identifier, which identifies it in a secure way so to say. So Harbor does in reality a really simple set of things. First and foremost, if the user has not seen it, like, not loaded it before first Harbor asks the local cache, do I have this file? If they do, if Harbor does, it just sends the file and it tells the browser, Hey, by the way, please don't ask me about this file for the next day. And in most cases, after one refresh, the user, all of the images load instantly because the web browser just goes, of those files were already sent. And Harbor asked me not to like, ask it more about the same file. So in the case of the image isn't in harbor's local cache, Harbor, first does a lot of those steps to resolve, the users identifier through their PDS, basically resolving that identifier, the DID to a DID document, which is a document basically explaining how that user, what is their, alias, what is their handle and where can we find them, which PDS. So we find the PDS and we then ask the PDS, Hey, send us this file for this user. The PDS sends it or doesn't, in which case we just throw an error and, Harbor just saves it locally and it sends it to the client. It basically just that. But to my knowledge, it's the first non Bluesky image proxy that's deployed for any AppView. Which also caught the attention of Brian Newbold one of the Bluesky employees and made me really happy. DID PLC Lookup [00:58:14] Jeremy: The lookup when you have the user's, DID and you wanna find out where their PDS is that's talking to something called, I think it's the PLC directory? [00:58:25] Prefetcher: Actually there are two different ways. First is PLC directory, PLC originally standed for a placeholder, and then Bluesky realized that it's not a placeholder anymore, and they stealthily changed it to public ledger of credentials. So we have PLC and we have web, the most common version is PLC. The document, the DID document is stored on Bluesky controlled servers under the moniker of PLC directory. They expose a web API that basically just allows you to say, Hey, give me the document for did:plc, whatever. And, the directory goes, have it. And this is the less decentralized version. You can host your own PLC directory and you can basically ask (their) PLC directory to just send you every single document and just you can have your local copy, which some people already do, you kind of sacrifice the fact that you are not in control of the document. It's still on a centralized server, even if you control the keys. 'cause every single DID document also has a key. And that key is used to sign changes to the document. So technically, if you define your own set of keys, you can prevent anyone else from modifying your document, even Bluesky. 'cause every single document is verifiable back and forth. You can see the previous document and its key is used to sign the next document and the chain of trust is visible and no one can just make random changes to your identity, but yeah, it's still on Bluesky to control service and it's a point of contention. Bluesky eventually wants to move it to a nonprofit standards organization, but we have yet to see anything come out of it, sadly. DID WEB lookup The next method is web. And web instead of -- 'cause in did:plc, you have did:plc, and a random string of characters. [01:00:30] Prefetcher: Web relies on domains. So for example, the domain would already like be the sole authority of where the file is. So for example, if I had did:web:example.com, I would parse the DID and I would see it's hosted at example.com. So I go to example.com, I go to /.wellknown/did.json which is the well-known location for the file. And I would have the same DID document as I would have if I used, for example, a PLC DID resolved via the PLC directory. the web method, you are in control of the document entirely. It's on your server under your domain. While it's the more decentralized version, it's just kind of hard for non-technical people to make them. 'cause it relies on a bunch of things. And also the problem is that if you lose your domain, you also lose your identity. [01:01:23] Jeremy: Yeah. So unlike the PLC where it's not really tied to a specific domain, you can change domains. With the web way, you have to always keep the same domain 'cause it's a part of the DID and yeah, like you said, you can't let your renewal lapse or your credit card not work. 'cause then you just lose everything. [01:01:49] Prefetcher: Yeah. You would still be able to change handles, but you would be tied for that domain to forever send your DID otherwise you would just lose it forever. [01:01:57] Jeremy: Yeah, I had mostly only seen the PLC and I wasn't too familiar with the web, form of identification, but yeah that makes sense. [01:02:06] Prefetcher: I think the web if I remember correctly, there is slightly over 300 accounts total on the entire network that use it. Mary who is a person on Bluesky that does a lot of like, ATProto related things, has a GitHub repository that basically gives insight into the network. And on her GitHub repository, you can find the list of every single custom PDS and also how many DID webs there are in existence. And I think it was slightly over 300. [01:02:38] Jeremy: So are you on that list? [01:02:40] Prefetcher: My PDS Yeah. If you were to scroll down. I don't use a web DID 'cause I registered my account before when I was brand new to ATProto, so I didn't know anything. But if you had to scroll down, you would see pds.ata.moe, which is my custom PDS just running. [01:02:55] Jeremy: Cool. [01:02:57] Prefetcher: Yeah. Harbor image proxy can cache any image blob [01:02:58] Jeremy: So something I noticed about harbor, you take the, I believe you take the DID and then you take the CID, the content identifier. I noticed if you take any of those pairs from the ATProto network, like I go find a image somebody posted on Bluesky, I pass that post DID and CID for the image into harbor. Harbor downloads it and caches it. So it's like, does that mean anybody could technically use you as a ATProto CDN? [01:03:38] Prefetcher: Yes, the same way anyone could use like the Bluesky CDN to for example, run PinkSea like I did. cause I do not know if there is a good way to check if a CID of an image or a blob basically. 'cause files on ATProto are called blobs. I do not think there is a nice way to check if that blob is directly tied to a specific record. But that also allows you to make cool, interesting things. Crossposting to Bluesky talks directly to the PDS [01:04:06] Prefetcher: 'cause for example, PinkSea has that, cross post to Bluesky thing. So when you create an image, You already have an option to cross post it to Bluesky, which a lot of people liked. And it was a suggestion from one of the early users of PinkSea. And the way it works is that when we create a PinkSea record, we upload that image, right? And then PinkSea goes, okay, I'm gonna use that same image, the same content identifier, and just create a Bluesky post. So Bluesky and PinkSea all share the same image. I don't upload it twice, I just upload it once. use it in PinkSea and I also use it in Bluesky. And the same way Bluesky its CDN, can just fetch the image. I can also fetch the image from mine, 'cause blobs aren't tied to specific records. They just exist outside of that realm. And you could just query anything. Not even images. You could probably query a video or even a text file. [01:05:04] Jeremy: So when you cross post to Bluesky, you're creating a record directly in the person's PDS, not going through bluesky's API. [01:05:14] Prefetcher: No, I sidestep Bluesky's API completely. And, I basically directly talk to the PDS at all times. I just tell them, Hey, please, for me, create a app.bsky.feed.post record. And you have the image, the text, which also required me to manually parse text into rich text. 'cause like, Bluesky doesn't automatically detect for example, links or tags And you basically get -- like PinkSea creates a record directly with the link to the image. And all of those tags, like the PinkSea tag and whatever, And I completely sidestep. Bluesky's API. If Bluesky, the AppView would cease to exist, PinkSea would still happily create Bluesky crossposts for you. Other applications put metadata into Bluesky posts so they can treat them differently [01:06:02] Jeremy: And since you're creating the records yourself, then you can include additional metadata or fields where you know that this was a PinkSea post, or originally came from PinkSea. [01:06:13] Prefetcher: I could do that. I don't really do that right now 'cause I don't really have much of a reason other than adding a PinkSea hashtag to every single oekaki. But I, noticed, for example, I think it was PinkSky, interesting name, PinkSky, which is like (a) Bluesky Instagram client. Any single time you make a post via PinkSky it uses the Bluesky APIs. It's Bluesky, but it attaches a hidden hashtag like PinkSky underscore some random letters. In its feed building algorithm, it basically detects posts with that hashtag, that specific hashtag, and it builds a PinkSky only timeline. 'cause it's still a Bluesky post, but it has hidden additional metadata that identifies, Hey, it came from PinkSky. [01:07:02] Jeremy: It's pretty interesting how much control you have over what to put in the PDS. So, I'm sure there's a lot of interesting use cases that people are gonna come up with. [01:07:14] Prefetcher: Yeah, of course. You still lose some of the data when you go through the Bluesky API. 'cause of course it stores the record and it's all in formats and whatnot. But you can attach a lot of metadata that can identify posts and build micro networks within Bluesky itself. I see it like that. Bluesky CDN compression [01:07:37] Jeremy: And I think, this might have been a post from you. I think I saw somebody saying that when you view an image from the CDN that the Bluesky CDN specifically, there's some kind of compression going on that that messes with certain types of art. [01:07:55] Prefetcher: It's especially noticeable artists are complaining about it all the time, left and right. Bluesky is very happy with jpeg compression, by default, their CDN, -- like to every single image it applies a really not good amount of jpeg compression which is especially not small. If you compare an image that's uploaded via PinkSea, view an image on PinkSea, and view the same image, which is, it's the same content id. It's the same blob. And you view it on Bluesky, it loses so much fidelity, it loses so much of that aliasing on the pen. You just see everything become really blurry. And on top of that, when you upload an image via Bluesky itself, if I remember correctly, I don't wanna lie here, but they also downscale the image to 1024 pixels by default. So every single image, not only big ones, and artists usually work with really big canvases, they get, downscaled and also additionally they get jpegified. So for example, PinkSea directly uploads PNG files to the PDS. And for example, Harbor gives back the original file. It does no transformations on it, but Bluesky transforms all of them into JPEG compressed images and for photos, it's fine sometimes. 'cause I've also seen people just compare directly, downloaded images of the PDS versus images viewed on Bluesky. But for art it's especially noticible. And people really (do) not like that. [01:09:31] Jeremy: Yeah, that's kind of odd. 'cause if, if I understand correctly, then if you post directly to your PDS and Bluesky pulls it in you'll avoid that, that 1024 resizing. So your images will be higher quality? [01:09:47] Prefetcher: I actually do not know. That's an interesting question. Cause I know that the maybe their CDN also does that 'cause that's what I've heard from others, that on upload the image gets processed and squashed down. So I don't know if doing it via an alternative AppView would change it or would Bluesky just directly reject this post? Because for example, PinkSea, I have built-in which I think I might change in the future -- PinkSea will reject your post if it's bigger than 800x800. 'cause then it'll notice that something is off. This could not have been made with PinkSea. [01:10:26] Jeremy: Yeah, that's a good point I suppose we know at the very least, they have some third party and internal moderation tools that they feed the images through to, so they, they can do some automatic content tagging. But yeah, I, I don't know, like you said, whether, the resizing and all that stuff is at the CDN level [01:10:50] Prefetcher: The jpegification is definitely at the CDN level. 'cause, Bluesky is actually running an open source image proxy. It's called imgproxy. Brian Newbold talked about it a bit on that harbor post. And, yeah, so a lot of the compression, the end user things are done via image proxy, but that, downscaling, I don't know, you'd have to ask someone who's a bit more intimate with Bluesky's internals. [01:11:19] Jeremy: Cool. yeah, I think we've, we've covered a lot. Is there, is there anything else, you wanted to mention or thought we should have talked about? [01:11:26] Prefetcher: Regarding PinkSea I think I've mentioned a ton both the behind the scenes things and, the user things, the design principles. What I'd want to absolutely say, and it will sound cheesy, and, is that I'm eternally grateful to anyone who's actually visited PinkSea. It's definitely grown outta all of my like dreams for the platform, to the point where I'm sitting here just talking about it. I definitely hope that the future will bring us more applications (in) ATProto. I definitely have ideas on how to expand PinkSea, a lot of ideas, a lot of things I want to do, and I'm also a very busy person, so I never get around them. But yeah, think that's it, at least regarding PinkSea. [01:12:15] Jeremy: Cool. Well, if people want to check out PinkSea or see what you're up to, where can they find you? [01:12:22] Prefetcher: So PinkSea is at pinksea.art. That's the website and Bluesky Handle is at pinksea.art and me, well, search prefetcher on Bluesky, you'll probably find me. My tag is at prefetcher.miku.place. all of my socials are probably there. I'm Prefetcher pretty much every single platform except for the platforms that already had someone called Prefetcher. GitHub, github.com/purifetchi because Prefetcher was taken. And, yeah, hit me up. I'm always eager to talk. I don't bite. [01:13:00] Jeremy: Very cool. Well, Kacper thanks. Thanks for taking the time. This was fun. [01:13:04] Prefetcher: Thank you so much, Jeremy, for having me over. It was a pleasure.

What is happening with Musk and DOGE and government, who does the USA owe money to, o3-mini vs Google Flash vs avante thoughts, what does Dax watch instead of sports, Adam makes a case for why sports matter, and should Adam trust Marc Andreessen or Ben Horowitz?Links:OpenAI o3-mini | OpenAIClaudeGemini Flash - Google DeepMindGitHub - yetone/avante.nvim: Use your Neovim like using Cursor AI IDE!Can I use... Support tables for HTML5, CSS3, etcWes Bos on X: "this website is nuts! I downloaded the entire codebase and was surprised to find out it's built with web components, Vanilla.js, Shopify and Tailwind! Full vid diving into the stack up nowThe Last DanceBelow Deck - WikipediaMarc Andreessen - WikipediaBen Horowitz - WikipediaAndreessen Horowitz | Software Is Eating the WorldThe Hard Thing About Hard Things | Andreessen HorowitzSponsor: Terminal now offers a monthly box called Cron.Want to carry on the conversation? Join us in Discord. Or send us an email at sliceoffalittlepieceofbacon@tomorrow.fm.Topics:(00:00) - We're so prepared (00:35) - Will the real time zone please stand up? (02:09) - DOGE, debt, and democracy (16:11) - o3-mini, Google Flash 2, avante thoughts (27:38) - Manic and bipolar (29:58) - Superbowl, NBA, and watching sports (37:55) - Reality tv and real life (45:52) - Sponsor: Terminal.shop (46:26) - Should Adam trust Marc Andreessen or Ben Horowtiz? ★ Support this podcast ★

Chris Coyier, co-founder of CodePen, talks about the evolving landscape of HTML heading into 2025. He delves into topics like the slow evolution of HTML compared to CSS and JavaScript, the importance of backwards compatibility, new HTML elements and pseudo-elements, and the potential of declarative shadow DOM for server-side rendering in web components. Links https://chriscoyier.net https://codepen.io/chriscoyier https://front-end.social/@chriscoyier https://github.com/chriscoyier https://www.threads.net/@chriscoyier https://bsky.app/profile/chriscoyier.net We want to hear from you! How did you find us? Did you see us on Twitter? In a newsletter? Or maybe we were recommended by a friend? Let us know by sending an email to our producer, Emily, at emily.kochanekketner@logrocket.com (mailto:emily.kochanekketner@logrocket.com), or tweet at us at PodRocketPod (https://twitter.com/PodRocketpod). Follow us. Get free stickers. Follow us on Apple Podcasts, fill out this form (https://podrocket.logrocket.com/get-podrocket-stickers), and we'll send you free PodRocket stickers! What does LogRocket do? LogRocket provides AI-first session replay and analytics that surfaces the UX and technical issues impacting user experiences. Start understand where your users are struggling by trying it for free at [LogRocket.com]. Try LogRocket for free today.(https://logrocket.com/signup/?pdr) Special Guest: Chris Coyier.

In the final episode of 2024, Bill and Anna take the helm for a hilarious and heartfelt year-end review. From their favorite marketing trends and pop culture moments to bold predictions for 2025, this episode is packed with insights, laughs, and a touch of nostalgia.

Introduction: In this episode, Greg Posner welcomes Jennifer Javornik, the Chief Partnerships Officer at Filament Games, to discuss the impact of educational gaming and the thriving Midwest game development community. Jennifer shares insights into how Filament creates games that go beyond entertainment by transforming players and shaping the future of education through immersive, meaningful experiences. This conversation also highlights the Midwest as an emerging hub for game development, and the key role the Wisconsin Games Alliance plays in that growth.Key Takeaways: The Mission of Filament GamesJennifer emphasizes that Filament Games is a mission-driven, for-profit studio that focuses on creating playful experiences that improve people's lives. Their goal is to make learning fun and engaging, with players emerging from their games transformed in some way—whether it's learning new concepts, developing confidence, or making behavior changes. Transformative Power of GamesResearch backs the effectiveness of well-designed games in education. Games provide an interactive problem space where players can experiment, explore, and gain deeper, longer-lasting understanding compared to traditional learning methods. Jennifer explains how Filament Games integrates this approach into their educational products to encourage players to engage with and retain the material. Impact of COVID on Game DesignThe pandemic led to a shift in educational needs, as many students struggled with virtual learning. Jennifer shares how Filament Games adapted by exploring how their games could support students and teachers during this time. The focus has increasingly been on building multiplayer or collaborative experiences to enhance the learning environment. Innovative Platforms for Learning GamesFilament Games develops across a variety of platforms, from standard tools like Unity and Unreal to HTML5 and mobile devices. Jennifer highlights their work on more unconventional platforms, such as Roblox, and even details their creation of a unique game for pediatric MRI machines to help children remain calm during scans. Building the Midwest Game Development CommunityJennifer discusses her work with the Wisconsin Games Alliance and how the Midwest is becoming a hub for game development. She notes the growing number of indie and AAA studios in the region and the importance of initiatives like MDev to unite the game development community and showcase the talent and innovation coming out of the Midwest.Timestamps: [01:10] – Introduction to Filament Games and its mission-driven approach to game development [03:33] – How games can provide better learning experiences through interactive problem-solving [05:23] – The impact of COVID on education and game design at Filament Games [07:32] – Filament's work on multiple platforms, including Unity, Unreal, Roblox, and even MRI machines [13:18] – Jennifer discusses the growing game development scene in the Midwest and the role of the Wisconsin Games Alliance

The Broken Plane Podcast Rookie Review Saints @ Cowboys NFL Offensive Leaders Week 3 Preview HTML5 Audio Player

The Broken Plane Podcast Rookie Review Chiefs Ravens Seahawks Broncos Other Notes HTML5 Audio Player

The Broken Plane Podcast AFC South Houston Texans Indianapolis Colts Jacksonville Jaguars Tennessee Titans HTML5 Audio Player

The Broken Plane Podcast AFC East New England Patriots Buffalo Bills New York Jets Miami Dolphins HTML5 Audio Player

The Broken Plane Podcast AFC North Baltimore Ravens Cincinnati Bengals Cleveland Browns Pittsburgh Steelers HTML5 Audio Player

The Broken Plane Podcast AFC West Los Diegos San Angeles Chargers Denver Broncos Kansas City Chiefs Las Vegas Raiders HTML5 Audio Player

The Broken Plane Podcast NFC North Chicago Bears Minnesota Vikings Detroit Lions Green Bay Packers HTML5 Audio Player

In this episode of Syntax, Wes and Scott talk about HTML fundamentals — from basic structure and semantics to practical tips for better accessibility and SEO. They also discuss the difference between block and inline elements, form functionalities, HTML5 elements like dialog and canvas, and more. Show Notes 00:00 Welcome to Syntax! 02:33 Brought to you by Sentry.io 03:25 Why HTML is important 06:52 Semantic vs non-semantic 07:58 The basic structure of an HTML page HTML elements reference The Main element 08:45 Doctype 15:24 Nav 18:47 Section 20:41 Aside 22:09 Article 22:54 Span 27:18 Why use a span when you have a div and a paragraph tag? 29:29 Figure and Caption 31:16 Fieldset 31:53 UL vs OL 32:44 DFN The Definition element 34:16 Form 36:56 Button vs Anchor 38:22 Headings 674 - A11y Treats - Heading Design 40:21 Output The Output element 41:46 Dialog 42:04 Tables 44:03 Media media-chrome 45:06 Canvas https://githubuniverse.com/ https://maximeheckel.com/ 46:07 On graphics programming 47:38 Search 354 - The Surprisingly Exciting World of HTML Elements 48:27 Sick Picks + Shameless Plugs Sick Picks Scott: 2Pack Traditional Natural Bamboo Wok Brushes Wes: Logitech MX Master 3S Shameless Plugs Syntax YouTube Channel The Easiest Way to Infinite Scroll with React | Full Example Hit us up on Socials! Syntax: X Instagram Tiktok LinkedIn Threads Wes: X Instagram Tiktok LinkedIn Threads Scott: X Instagram Tiktok LinkedIn Threads Randy: X Instagram YouTube Threads

The Broken Plane Podcast NFC West Seattle Seahawks San Francisco 49ers Los Angeles Rams Arizona Cardinals HTML5 Audio Player

The Broken Plane Podcast NFC South Atlanta Falcons Carolina Panthers Tampa Bay Buccaneers New Orleans Saints HTML5 Audio Player

The Broken Plane Podcast NFC East Washington Commanders - DCFC Dallas Cowboys Philadelphia Eagles New York Giants HTML5 Audio Player

The Broken Plane Podcast Minshew Mania! Season 7 Coming Soon HTML5 Audio Player

The Broken Plane Podcast Run This Back Season 7 Teaser HTML5 Audio Player

Guest: Marissa Mayer, CEO and Founder of Sunshine and former CEO of YahooWhen Marissa Mayer was first hired as the CEO of Yahoo, the company had lost nearly a quarter of its workforce in the preceding six months. Early on, she was chatting with employees in the cafeteria and one of them got her attention by smacking her tray. “Is it go time?” he asked. He was asking if the board and C-suite were ready to lead the company forward, but Marissa thought he had one foot out the door. “I had just come out of this meeting where they were like, ‘Everyone's leaving!'” she recalled. “And I was like, ‘Oh no, please don't go, I've only been here for four days!'”In this episode, Marissa and Joubin discuss the number 12, contacts and photo sharing, fear of AI, soccer moms, maternity as a “disability,” mothers' rooms, Jim Citrin, Project Cardinal, HTML5 vs. native apps, Ross Levinsohn, Lori Puccinelli Stern, Joe Montana, David Karp, Mark Zuckerberg, Taylor Swift, hiring at Google, Amit Patel, Hamilton, John Doerr, and the Google APM program.Chapters:(00:52) - Reading your own press (04:55) - Marissa's lucky number (07:19) - Her latest startup, Sunshine (15:03) - Burnout, resentment, and rhythm (21:46) - The opportunity to become CEO of Yahoo (27:00) - Inverting maternity leave (31:14) - The big interview (36:44) - An epic dinner party (42:51) - The voicemail (47:18) - Farzad “Zod” Nazem and David Philo (50:25) - Last day at Google (53:52) - “Is it go time?” (59:03) - Buying Tumblr (01:04:46) - Alibaba and Verizon (01:06:24) - Larry and Sergey bucks (01:11:05) - Eric Schmidt's advice (01:12:59) - In the room at Google (01:18:36) - Teaching and identifying talent (01:24:32) - Who Sunshine is hiring Links:Connect with MarissaTwitterLinkedInConnect with JoubinTwitterLinkedInEmail: grit@kleinerperkins.com Learn more about Kleiner PerkinsThis episode was edited by Eric Johnson from LightningPod.fm

Show DescriptionJen Simmons, Apple Evangelist on the Web Developer Experience team for Safari & WebKit, stops by to talk about what Interop is, and a look ahead at new CSS features in Webkit and Safari such as JPEG XL, masks, a round function, JavaScript improvements, styling form controls, content unblocks, masonry, and more! Listen on Website →GuestsJen SimmonsGuest's Main URL • Guest's Twitter Apple Evangelist on the Web Developer Experience team for Safari & @webkit. Member of CSS Working Group. Links Can I use... Support tables for HTML5, CSS3, etc Apple Developer JPEG - JPEG XL WebKit Bugzilla Main Page SponsorsJam.devYou've probably heard of Jam.dev, it's used by more than 60,000 people. It's a free tool that saves developers a ton of frustration. It forces your teammates to make the perfect bug report. They can't do it wrong because it automatically includes a video of the bug, console logs, network requests, everything you need to debug. It automatically lists out the steps to reproduce. It's so easy to get your teammates to use. It's just a Chrome extension. When they see a bug, they click a button and right away it creates a ticket. So it saves time for them.

In this episode of Syntax, Wes and Scott talk about the process they are using for extending the Syntax website with themes, including a dark mode and secret custom themes. They also talk about issues encountered with theming and CSS, including flashes of un-themed content. Show Notes 00:00 Syntax + Sentry Announcement! 01:35 Welcome 02:23 Syntax Brought to you by Sentry 03:16 How to implement a theme 10:27 Writing the CSS 11:38 Glasses wearers protip 13:02 Class on the body and server side rendering issue 14:02 CSS Variables 15:05 Color variables in CSS 21:07 Working in half pixel sizes 22:40 Variable usage 25:23 Naming variables after what they style 30:42 Component level variables 33:27 Using zones 38:41 Themes should be defined as light or dark 39:20 Issues: Moving from light to dark 42:29 Issues: Drop shadow in dark mode 44:00 Issues: Flash of unthemed content 44:40 Issues: Opacity values 49:45 Issues: SVG need change color 55:56 Help is on the way! 59:57 SIIIIICK ××× PIIIICKS ××× website/src/styles/themes/level-up.css at v2 The 2023 State of CSS Survey Part 2 × CSS Frameworks × Tooling × Browser Usage “color-contrast” | Can I use… Support tables for HTML5, CSS3, etc The Future of CSS: Easy Light-Dark Mode Color Switching with light-dark() – Bram.us ××× SIIIIICK ××× PIIIICKS ××× Scott: Kala (@engineer.everything) Shameless Plugs Scott: Sentry Wes: Wes Bos Tutorials Hit us up on Socials! Syntax: X Instagram Tiktok LinkedIn Threads Wes: X Instagram Tiktok LinkedIn Threads Scott: X Instagram Tiktok LinkedIn Threads

In this episode of Syntax, Wes and Scott go over some of the results the State of CSS survey for 2023, including drop-shadow, subgrid, logical properties, media range, and more! Show Notes 00:25:21 Welcome 01:31:22 Soccer update 03:52:05 Syntax Brought to you by Sentry 04:03:18 StateOfCSS.com State of CSS 2023 04:21:11 Top 5 Most Used Features 09:13:09 Drop-shadow 13:00:11 Subgrid 18:31:21 Logical properties 20:11:20 Media range 20:52:11 Object view box 23:10:14 Anchor positioning 28:21:09 Intrinsic sizing keywords 29:32:00 View Transitions API Can I use… Support tables for HTML5, CSS3, etc Astro Supper Club × Astro 2.0 with Fred Schott — Syntax Podcast 580 31:43:18 Updates to colors in CSS 35:07:05 Accent color, current color 36:36:11 Scroll snap and overscroll behaviour 39:04:22 Touch action 40:16:01 Line clamp or variable font Variable Fonts 45:28:11 Accessibility features 48:33:12 :Has has jumped 49:32:11 @Property 53:52:13 SIIIIICK ××× PIIIICKS ××× ××× SIIIIICK ××× PIIIICKS ××× Scott: Kodiak Cakes Strawberries & Cream Oatmeal Packets (Pack of 6) Wes: Food Storage Containers, Baby Food Storage Containers with Lid, Baby Food and Toddler Snack Containers Shameless Plugs Scott: Syntax Podcast (@syntaxfm) | TikTok Wes: Web Unleashed 2023 - FITC Tweet us your tasty treats Scott's Instagram LevelUpTutorials Instagram Wes' Instagram Wes' Twitter Wes' Facebook Scott's Twitter Make sure to include @SyntaxFM in your tweets Wes Bos on Bluesky Scott on Bluesky Syntax on Bluesky