Why Did Anthropic Buy Bun - CourtneyR.dev's 2nd Brain

Why Did Anthropic Buy Bun? 🔗 **Source:** [Why Did Anthropic Buy Bun?](https://www.youtube.com/watch?v=zvSJsHmOCXc) > [!note]+ Document Notes > tags: [javascript, ai, bun, anthropic, devtools] > author: Syntax > source: YouTube > reference: https://www.youtube.com/watch?v=zvSJsHmOCXc > date created: 2025-01-27 11:00 > > # 🏷️ Title: Why Did Anthropic Buy Bun? - Syntax Potluck >  ## 🧰 Action Inbox - [H1] Evaluate Bun for development projects — faster startup than Node, SQLite built-in 00:45 #javascript - [H2] Set up proper Git GUI workflow — VS Code staging better than CLI commands 05:23 #devtools - [H3] Implement ASN blocking for web security — GitHub bad-asn-list for malicious hosting providers 42:15 #security ## 🧭 Table of Contents - [📝 Summary](#-summary) - [📖 Main Content](#-main-content) - [🗒️ Notes](#️-notes) - [📚 References](#-references) - [🗣️ Paraphrase](#️-paraphrase) - [📰 Newsletter](#-newsletter) - [✅ Action Items](#-action-items) - [🧪 Hypotheses](#-hypotheses) - [📊 Methodology](#-methodology) - [❓ Questions](#-questions) - [🔮 Future Research](#-future-research) - [💡 Implications](#-implications) - [🧠 Learning Styles Assessment](#-learning-styles-assessment) - [📊 Accessibility Assessment](#-accessibility-assessment) - [🌍 DEIB Assessment](#-deib-assessment) - [✅ Logical Validity](#-logical-validity) - [🛡️ DARVO Analysis](#-darvo-analysis) - [🎯 Conclusion](#-conclusion) ### 🔗 References - [Syntax FM](https://syntax.fm) - [Anthropic](https://anthropic.com) ## 📝 Summary - Anthropic acquired Bun JavaScript runtime to optimize Claude Code SDK performance - Discussion covers developer tools including Git workflows, video recording setups, and accessibility practices - Key stakeholders: JavaScript developers, AI companies, content creators - Tradeoffs: Performance gains vs. ecosystem compatibility, GUI convenience vs. terminal control - Bottom line: Strategic acquisition demonstrates AI companies investing in developer tooling infrastructure ## 📖 Main Content Anthropic's acquisition of Bun represents a strategic move to optimize their Claude Code SDK, which runs on the Bun runtime. By bringing Bun in-house, Anthropic can ship smaller, faster applications and add features they specifically need. This reflects a broader trend of AI companies investing in developer infrastructure. The hosts also address practical developer concerns around Git workflows, noting the advantages of GUI tools like VS Code's built-in Git features over command-line interfaces for common operations. They emphasize that visual staging, diffing, and commit management often provide better user experience than terminal commands. For security-conscious developers, they discuss ASN (Autonomous System Number) blocking as a strategy to prevent malicious traffic from known bad hosting providers, using Cloudflare rules and community-maintained blocklists. ## 🗒️ Notes #### ⏱️ [00:45](https://youtu.be/zvSJsHmOCXc?t=45) — [H1] Anthropic's Bun Acquisition Strategy - Key idea: Anthropic bought Bun to optimize Claude Code SDK performance and shipping - Evidence: "The cloud code SDK runs on bun and they want to be able to ship the cloud code...and make that thing as small and as fast as possible" #### ⏱️ [03:20](https://youtu.be/zvSJsHmOCXc?t=200) — [H2] Bun's Unique Approach to JavaScript Runtime - Key idea: Bun differentiates by adding non-standard APIs that should exist - Evidence: "Everything about Bun feels just a different approach...they're unapologetically adding features that JavaScript, Node.js non-standard APIs" #### ⏱️ [05:23](https://youtu.be/zvSJsHmOCXc?t=323) — [H2] Git GUI vs Command Line Workflow - Key idea: Visual Git tools provide better user experience for common operations - Evidence: "I like clicking the button that says track all files and I like maybe like clicking and seeing the diff on the files" #### ⏱️ [08:45](https://youtu.be/zvSJsHmOCXc?t=525) — [H3] VS Code Terminal Performance Issues - Key idea: VS Code had race condition causing paste delays on Windows - Evidence: "Every 50 characters they're delaying it what five milliseconds...if you were to paste a lot of text into the terminal, it would just take forever" #### ⏱️ [12:10](https://youtu.be/zvSJsHmOCXc?t=730) — [H2] Video Recording Setup Fundamentals - Key idea: Audio quality and code readability matter more than expensive equipment - Evidence: "The fundamentals here are truly making sure the code is big enough to read...making sure that your audio sounds good" #### ⏱️ [15:30](https://youtu.be/zvSJsHmOCXc?t=930) — [H3] Recording App Sources vs Screen Recording - Key idea: Recording individual application sources enables better editing flexibility - Evidence: "Recording sources and not screens...you can record the app, if you tab away or you bring over something over top of it, you don't have to show that" #### ⏱️ [20:15](https://youtu.be/zvSJsHmOCXc?t=1215) — [H2] Handling Difficult Coworkers - Key idea: Document issues and establish formal processes rather than personal conflicts - Evidence: "Document them very clearly...be honest. Don't pull any punches...These are the things that have happened and it's not acceptable to me" #### ⏱️ [25:40](https://youtu.be/zvSJsHmOCXc?t=1540) — [H1] Accessibility vs Creative Coding Balance - Key idea: Creative experiments don't need to be 100% accessible if properly labeled - Evidence: "Not every experience can be the exact same for all abled or disabled people...it's all about how you package those explorations and works" #### ⏱️ [28:20](https://youtu.be/zvSJsHmOCXc?t=1700) — [H3] 2026 WebGL and 3D Web Prediction - Key idea: Visual experiences will differentiate websites as current designs become homogeneous - Evidence: "2026 is the year of WebGL. It's the year of 3D. Everybody is going to care a lot more about the design and the like wow experience" #### ⏱️ [32:05](https://youtu.be/zvSJsHmOCXc?t=1925) — [H3] APCA Color Contrast Algorithm - Key idea: New contrast algorithm provides more accurate accessibility measurements - Evidence: "There's like this new color contrast algorithm that's being worked on that is way more...better" #### ⏱️ [35:10](https://youtu.be/zvSJsHmOCXc?t=2110) — [H2] Home Assistant Smart Home Setup - Key idea: Home Assistant on Raspberry Pi provides powerful automation without heavy hardware - Evidence: "I use Home Assistant running it on my Synology...it's just handling MQTT streams and setting variables" #### ⏱️ [42:15](https://youtu.be/zvSJsHmOCXc?t=2535) — [H2] ASN Blocking for Web Security - Key idea: Block malicious traffic by Autonomous System Number rather than individual IPs - Evidence: "There's a GitHub repo called bad ASN list...you can simply just say requests from these ASNs are going to be either totally blocked" #### ⏱️ [46:30](https://youtu.be/zvSJsHmOCXc?t=2790) — [H3] SAP UI5 Enterprise JavaScript Framework - Key idea: Large enterprises build custom frameworks for specific business needs - Evidence: "SAP UI5...they moved everything over to their own custom MVC JavaScript framework...it's not built on react or angular or svelte" ## 📚 References - [Bad ASN List GitHub](https://github.com/brianhama/bad-asn-list) - [Color Contrast App](https://colorcontrast.app) - [Sentry Error Tracking](https://sentry.io/syntax) - [Inside Archaeology YouTube](https://youtube.com/@insidearchaeology) ## 🗣️ Paraphrase - Anthropic acquired Bun to create faster, smaller applications for their Claude Code SDK while gaining control over development priorities - Developer tooling benefits from visual interfaces for Git operations, though terminal usage remains important for specific tasks - Security-focused developers can use ASN blocking to prevent malicious traffic from known bad hosting providers rather than playing IP address whack-a-mole ## 📰 Newsletter Anthropic's acquisition of Bun highlights how AI companies are investing in developer infrastructure. This strategic move optimizes their Claude Code SDK while demonstrating Bun's unique value proposition with built-in APIs and performance focus. For developers, the discussion reinforces practical workflow improvements: using Git GUIs for common operations, recording individual app sources for better video content, and implementing ASN-based security blocking. These tools reduce friction while maintaining professional standards. The accessibility conversation provides important nuance - creative coding projects can coexist with accessibility expertise when properly labeled and contextualized. Resources: [Bun Runtime](https://bun.sh), [Anthropic](https://anthropic.com), [VS Code Git Features](https://code.visualstudio.com/docs/editor/versioncontrol) ## ✅ Action Items #### DevRel - Developer advocacy: Showcase Bun adoption patterns in AI workflows, create runtime comparison demos - Education: Workshop series on Git GUI workflows, accessibility-first creative coding - Developer experience: ASN blocking implementation guides, video recording best practices - Developer marketing: Case studies on JavaScript runtime choices in production - Community management: Facilitate discussions around accessibility vs. creative expression #### Community - Maintainers: Document runtime migration strategies, establish accessibility labeling standards - Contributors: Create ASN blocklist maintenance workflows, improve Git tooling documentation - Ecosystem: Cross-runtime compatibility testing, security practice sharing #### JavaScript - Runtime evaluation: Bun vs Node performance testing, migration path documentation - Tooling integration: Git GUI workflow optimization, recording setup automation - Security implementation: ASN blocking rule creation, malicious traffic monitoring ## 🧪 Hypotheses - AI companies will increasingly acquire JavaScript runtime and tooling companies to optimize their SDKs and development workflows - Visual Git interfaces will become standard for most developers while CLI remains preferred for complex operations and automation - ASN-based blocking provides more effective security than IP-based approaches for preventing malicious automated traffic - Creative coding portfolios can enhance accessibility engineer careers when properly contextualized and labeled ### 📊 Methodology - Runtime performance comparison: Benchmark Bun vs Node.js startup time, memory usage, and built-in API performance - Git workflow analysis: Time-study visual vs CLI operations for common tasks like staging, diffing, and committing - Security effectiveness measurement: Compare blocked request volumes and false positive rates between ASN and IP blocking - Career impact assessment: Survey accessibility professionals on portfolio diversification strategies ### ❓ Questions - How will Anthropic's ownership change Bun's development priorities and community governance? - What specific features will Anthropic add to Bun that wouldn't have been prioritized in open-source development? - How can the accessibility community better support creative coding while maintaining standards? - What other JavaScript infrastructure will AI companies acquire to optimize their development tools? ### 🔮 Future Research - Impact analysis of AI company acquisitions on open-source JavaScript ecosystem sustainability - Longitudinal study of developer workflow efficiency across different Git interface approaches - Security effectiveness comparison of various automated threat blocking strategies - Career pathway analysis for developers transitioning between specialized and generalist roles ### 💡 Implications - Practice: Developers should evaluate Bun for performance-critical applications while maintaining Node.js compatibility - Policy: Organizations need guidelines for balancing accessibility requirements with creative experimentation - Security: ASN blocking should become standard practice for high-traffic applications facing automated threats - Education: Developer training should include both GUI and CLI Git workflows based on use case appropriateness - Funding: AI companies' infrastructure investments may reduce open-source project sustainability ### 🧠 Learning Styles Assessment Reading level estimate: Flesch-Kincaid Grade Level 10 #### 👀 Visual learners - Present: Git GUI screenshots, ASN blocking configuration examples - Improve: Runtime performance charts, security architecture diagrams #### 🔊 Aural learners - Present: Full podcast episode with clear audio quality - Improve: Timestamped audio summaries for each major topic #### 🗣️ Verbal learners - Present: Clear technical definitions, conversational explanations - Improve: Glossary of terms like ASN, runtime, GUI #### 🧩 Kinesthetic learners - Present: Step-by-step Git workflow examples, code snippets - Improve: Interactive Bun vs Node comparisons, hands-on security setup guides #### 🧠 Logical learners - Present: Acquisition rationale, performance comparisons, decision trees - Improve: Cost-benefit analysis charts, workflow efficiency metrics #### 🧑‍🤝‍🧑 Social learners - Present: Community discussion format, multiple perspectives - Improve: Developer community polls, collaborative best practices #### 🧘‍♀️ Solitary learners - Present: Individual tool evaluations, personal workflow optimization - Improve: Self-assessment checklists, private experimentation guides #### 🌿 Naturalistic learners - Present: Ecosystem thinking, sustainable development practices - Improve: Open-source sustainability connections, community health metrics ### 📊 Accessibility Assessment #### Screen reader semantics - Present: Clear heading hierarchy, timestamped navigation - Improve: Alt text for any visual examples, ARIA landmarks #### Keyboard navigation - Present: Sequential tabbing through content sections - Improve: Skip links to major sections, keyboard shortcuts for timestamps #### Contrast and readability - Present: High contrast text, readable font sizes - Improve: Follow APCA guidelines mentioned in content #### Text resizing and reflow - Present: Responsive design at standard zoom levels - Improve: Test at 200% zoom, ensure no content clipping #### Media and captions - Present: Full transcript provided - Improve: Timestamped captions, downloadable transcript #### Forms and interactions - Present: Standard link interactions - Improve: Focus indicators for interactive elements #### Standards reference - Meets WCAG 2.2 AA for text alternatives and keyboard accessibility ### 🌍 DEIB Assessment #### Cultural breadth - Present: Global perspective on developer tools, international hosting providers - Improve: Multi-region security examples, diverse developer workflows #### Inclusive language - Present: Gender-neutral terms, clear technical explanations - Improve: Avoid assumptions about hearing/vision abilities in audio discussions #### Neurodiversity support - Present: Clear section breaks, consistent structure - Improve: Content warnings for rapid topic changes, processing time considerations #### Representation - Present: Multiple developer perspectives and workflows - Improve: Diverse voices in community examples, varied use cases #### Psychological safety - Present: Respectful disagreement on accessibility approaches - Improve: Acknowledge different accessibility needs without judgment ### ✅ Logical Validity #### Strong reasoning - Deductive strength: Strong - Anthropic's business needs logically lead to Bun acquisition - Inductive strength: Moderate - Git GUI benefits generalized from personal experience - Causal reasoning: Strong - Clear cause-effect relationships in security blocking - Analogical strength: Moderate - Comparisons between different developer tools #### Fallacies - Appeal to popularity: "Everybody watches with captions now" - overgeneralization - False dichotomy: Accessibility vs. creativity presented as either/or rather than balance #### Tone of voice - Overall tone: Conversational and informative with technical expertise and practical focus ### 🛡️ DARVO Analysis - Deny: No patterns of denial observed - Attack: No attacks on individuals or groups - Reverse victim/offender: No victim/offender reversals present - Emotional cues and rhetoric: Balanced discussion with acknowledgment of different perspectives ## 🎯 Conclusion Anthropic's Bun acquisition signals a strategic shift toward AI companies controlling their development infrastructure. For developers, this highlights the importance of evaluating runtime choices based on specific use cases while maintaining ecosystem compatibility. The practical discussions around Git workflows, video creation, and security practices provide immediately actionable insights --- ## Full Transcript [00:00] Welcome to syntax. We got a potluck [00:02] episode for you today where it's where [00:03] you bring the questions, we bring the [00:05] answers. Today we got really good [00:06] questions on why did anthropic by bun? [00:09] What UI are we using for git? You know, [00:11] are using in the terminal? Are you using [00:13] lazy git? Are you using the VS Code [00:15] guey? Are gueies better than the [00:17] terminal? We're going to talk about [00:18] that. Do all web experiences need to be [00:20] accessible? Somebody was saying like, "I [00:23] love accessibility, but I I want to I'm [00:25] building these like crazy experiments [00:27] and and not everything is 100% [00:29] accessible. Should I be putting this [00:30] stuff out there?" Another question on [00:33] how do I block malicious requests? Um, [00:35] so like bad ASN's, what is an ASN? How's [00:39] that different from IP addresses? Really [00:41] good questions. Keep your questions [00:43] coming, by the way. Go to syntax.fm. [00:45] There's a button that says potluck. Ask [00:46] us your question and we'll answer on a [00:48] future episode. But let's get on into [00:50] it. First question from Brett Perryi. [00:54] Why did Anthropic buy Bun? So if you [00:56] didn't hear the news a couple weeks ago, [00:58] Anthropic bought Bun Bun the JavaScript [01:02] runtime. Why would they do that when [01:04] they can just ask Claude to make a run [01:06] JavaScript runtime? But this was this is [01:08] a pretty interesting move. And the [01:11] reason that they did this was because [01:14] the cloud code SDK runs on bun and they [01:19] want to be able to ship the cloud code. [01:23] I'm not going to say binary, but like [01:25] the app itself and they want to ship it [01:28] ideally with the JavaScript runtime. I'm [01:31] pretty sure that's what they do right [01:32] now as well. And in order to make that [01:34] thing as small and as fast as possible [01:38] and I'm assuming get the features in [01:40] that they they necessarily need, it's [01:43] great to to bring Bun under their wing [01:45] and and have them continue developing [01:47] that. Yeah, it's it was it was [01:49] interesting to me because it's like Bun [01:51] is such a unique case where everything [01:54] about Bun feels just a different [01:57] approach than than like whether that is [01:59] the fact that it's written with Zigg or [02:01] the fact that um they're [02:02] unapologetically [02:03] adding features that JavaScript, [02:06] Node.js, [02:07] >> non-standard APIs. [02:08] >> Yeah, non-standard APIs just because [02:11] they should be there. I personally I do [02:13] like using bun. There are enough [02:15] mysterious times when I'm using bun [02:18] where something isn't working quite [02:19] right and I have to go back. But if I'm [02:22] throwing up a script here and there, I'm [02:24] using bun every time. And I I do feel [02:28] like uh that plays well into their [02:31] entire beyond the fact that they they [02:34] use it. I think it plays well into them [02:37] wanting to be the authority on vibe [02:41] coding there, right? building scripts, [02:43] building kinds of stuff. It bun is so [02:45] much less effort to get set up with than [02:48] these other ones. [02:49] >> There's sort of like another aspect to [02:51] this that is probably going to happen is [02:54] that when you ask chat GPT or or cloud [02:59] what is it just called cloud? They're [03:00] chat apps. When you ask them to do a lot [03:02] of stuff, they're literally just running [03:04] Python scripts uh behind the scenes. You [03:07] know, the AI is just running code and [03:09] and spitting out the result for you. So [03:12] if that can be as fast as possible or if [03:16] they can make this amazing little [03:18] nucleus that you can just throw code [03:21] into and and have it spit out exactly [03:24] what you need. That's probably another [03:26] use case. Like there's the obviously the [03:27] cloud code use case where it runs on [03:30] your machine, but some sort of [03:34] compute that's running somewhere [03:36] is is probably the other use case that [03:39] it will do calculations. It will do [03:41] actually running code for the LLM when [03:44] it's needed. [03:45] >> Yeah, it's it's interesting nonetheless. [03:48] You know, I'm not surprised that someone [03:49] bought bun, honestly. Uh yeah, I I don't [03:52] know if I would have predicted [03:53] Anthropic, but I I I do think it's [03:56] fascinating. [03:57] >> We had predicted that Bun would put its [04:00] own hosting competitor together and like [04:03] that [04:04] >> like the other player in this space is [04:06] Dino, right? And and both of these first [04:08] came at like at Node being like we're [04:11] going to be a better node. [04:13] >> And the uptake has been good, but like [04:15] not amazing. I think most people are [04:17] just fine with node. So there's this [04:21] other use case is like oh well maybe we [04:23] can have like hosting or maybe we can [04:25] have like a like an app that we run and [04:28] and those things haven't haven't really [04:30] picked up for Dino either. So they're [04:33] probably looking at being like we [04:34] probably are not going to like crush [04:36] Versel or AWS or anything like this. So [04:41] our other option to like obviously Bun [04:44] is has been VC funded and they need to [04:46] get their money back at some point. The [04:48] other option is to like sell to a larger [04:51] company who has the cash to help them [04:53] out. [04:54] >> Yeah. You know what I wonder about Dino [04:57] and all of this? Like this is seems like [04:59] crazy to say given how much of a change [05:02] Dino was when it was first dropped, but [05:04] do you think Dino [05:07] is too much in the mushy middle between [05:10] bun or [05:12] nojs to really stand out? Right. cuz [05:14] Dino [05:16] >> now feels like bun is kind of the the [05:20] hyper performant. It is the yolo just [05:23] add SQLite to the thing you know kind of [05:26] thing where where Dino feels so much [05:28] more measured and format so much more [05:30] like NodeJS 2.0 that maybe it's not [05:35] poised to break out as much. I think the [05:37] the reason why Bun got so popular was [05:39] because they did non-standard stuff. And [05:42] if you're going to use something that's [05:44] not Node, I think a lot of people sort [05:45] of looked at that and was like, well, [05:47] I'm going to use the thing that has all [05:48] these like weird APIs built in. So like [05:52] if if you're not sure what we're talking [05:54] about is like bun shipped like um SQL uh [05:58] database adapter built in, right? Now [06:00] node node has that as well. Um, they [06:02] also are like competing with Vit in [06:04] being like a bundler, but then they also [06:06] have like a router and they can also [06:07] just import HTML and like return it and [06:10] it will will serve it up and like that's [06:13] that's kind of cool. Those are kind of [06:15] nice APIs for especially for things that [06:17] are need to get up and running very [06:19] quickly which often is these tiny little [06:22] vibecoded apps. So I don't know I don't [06:25] know what the process is. I was hoping [06:27] that Dino would go into the space of [06:30] like web GPU because like so much of [06:34] this whole data science, ML, AI world is [06:38] all written in Python scripts and and a [06:41] lot of that is because [06:42] >> they had some of the really early [06:44] libraries like PyTorch. [06:46] >> Um, but also a lot of it is because like [06:48] this stuff comes out of academia and and [06:51] academics love them some Python. So, I [06:54] was kind of hoping that like Dino would [06:56] go [06:58] like maybe in the model training world [07:00] with all this GPU stuff. They just [07:02] released support for for web GPU in [07:05] Dino. [07:06] >> So, I don't know what the what the move [07:08] is there. The the whole like better [07:10] standards thing for Dino and [07:12] unfortunately is not working out um [07:14] because they they released JSR which is [07:16] the npm competitor. That didn't really [07:18] work all that well. you know, they had [07:20] several other sort of foray into this [07:23] this world, but it's just so hard to [07:26] beat the monolith that is node. Yeah, [07:29] it's actually it's it is it's we're in [07:31] an interesting situation. All right, [07:33] next question here from get commit [07:35] struggler. What do you use to commit [07:37] your changes on your editor? I have a [07:39] weird struggle where I am using an [07:41] external terminal to run my commands, [07:43] but I always use the internal editor to [07:46] run my git commands. get commit, get [07:48] push, etc. I am looking for a better way [07:51] that feels natural. Hey, I do this as [07:55] well. I run all of my terminal commands [07:57] on Ghosty on my terminal, external, like [08:00] no matter what it is, I I'm running all [08:02] my processes starting up there. I code [08:04] in my editor, but then I do use my [08:08] editor and usually just straight up the [08:12] editor's UI for all Git operations. why [08:16] I like clicking the button that says [08:18] track all files and I like maybe like [08:20] clicking and seeing the diff on the [08:22] files and kind of browsing it that way. [08:25] >> And that was one of the reasons why I [08:26] didn't move to the zed editor for a long [08:29] time because their git tools weren't [08:30] there. But zed actually has [08:32] >> really good uh git diff tools and and [08:35] you you can do everything that I was [08:37] doing in VS Code before. So me, I'm [08:39] still using that and and VS Code even [08:41] has like using AI to generate commit [08:44] titles and stuff like that directly in [08:46] the editor. There there's like a lot of [08:47] nice little features that you can have [08:49] via the UI that I just prefer to do it [08:52] that way. Now there is other means to do [08:56] this. Um there was what is that? Git [08:59] like different UIs here. You look it up. [09:00] I'll tell you where I'm at right now [09:02] because like for ever I've been like a [09:04] get in the CLI guy and everybody for [09:08] years has been like you got to use a [09:10] guey. You got to use a guey and I've [09:12] been I switched over mostly to using the [09:15] one that's built into VS Code cursor [09:17] whatever [09:18] >> and it's it honestly it's better like [09:20] having a guey where you can literally [09:22] click on and and like stage the files. [09:24] You can you can click on them and see [09:26] the diffs. It's honestly better for for [09:29] that whether you like the AI commit [09:31] messages or or not. [09:33] >> Yeah, [09:33] >> I find that to be really really nice. [09:35] I'll still every now and then use I I [09:38] have so hardcoded in the terminal. If I [09:40] need to do anything like change an [09:42] origin or list out all of those things, [09:44] I'll just write back to it. But the [09:46] simple adding, committing, pushing, [09:48] those big three are are happening mostly [09:52] in my editor these days. And I also I [09:54] want to move a lot of my terminal usage [09:57] into the editor as well. I just can't [10:00] >> can't do it. I just it feels I don't [10:02] know what it is, but it's just doesn't [10:04] feel good. I like the separate terminal. [10:06] >> Well, uh I mean that does it might not [10:09] feel good because did you see that whole [10:11] drama about the VS Code terminal over [10:14] the past? [10:14] >> Well, yeah. Every 50 characters they're [10:16] delaying it what five milliseconds. [10:18] >> Yeah. Yeah. So, which is funny. So, by [10:22] the time you're hearing this, this is [10:23] old news, right? This happened in the [10:26] last year, uh, because we're recording [10:28] this ahead of time, but there was like a [10:29] race condition that when you pasted text [10:33] into the term, there would be a sleep [10:35] because of wind something on Windows [10:38] apparently. And because of that, if you [10:40] were to paste a lot of text into the [10:41] terminal, it would just take forever. [10:44] Now, what's funny, people are really [10:46] ragging on this, and I understand why, [10:49] but this is a problem that only rears [10:51] its head in a vibecoded world where [10:54] we're pasting a ton of stuff into the [10:56] term, right? Like in the past, how how [10:58] often were you ever pasting in 20, you [11:02] know, 30 line, 40 lines and of code into [11:05] your terminal? You were never doing that [11:06] ever. [11:07] >> No. Yeah. very rarely ever doing those [11:10] things because like at most a curl [11:13] command that's several hundred has [11:16] characters. [11:18] >> Yeah. Right. Yeah. So very funny. [11:20] [laughter] The the thing that I was [11:21] thinking of was lazy git. Lazy git which [11:24] comes on Omari as well as it's like a [11:26] it's a TUI a terminal UI for git. And I [11:31] I I'm going to be honest. I think lazy [11:33] git people really seem to love, but to [11:36] me this is definitely like a weaker [11:38] experience than what I get out of VS [11:42] Code or whatever. And I think it's nice [11:45] if you're the person who likes a [11:46] terminal UI, but ultimately the learning [11:49] curve was pretty high on this overall, [11:51] but it is a terminalbased UI if you want [11:54] kind of to ride that line. Man, I love [11:56] the terminal, but I I do not support [11:59] trying to everybody trying to make a UI [12:02] for absolutely everything in the [12:04] terminal. Man, you know what is good [12:06] >> at making UIs? HTML and CSS or whatever [12:10] native UI you know if you want to put a [12:13] button somewhere you want to click it if [12:15] you want to be able to select text [12:18] >> by word by word that's that's [12:21] >> I want to be able to resize the the [12:23] window without everything kind of [12:24] getting all [laughter] angry at me all [12:26] the time [12:28] >> yeah yeah some of these these terminal [12:29] UIs are great and and I love them for [12:31] little utility tools but some stuff is [12:34] better as a guey [12:36] >> even even uh open code folks just [12:40] released a a guey for that because they [12:42] realize yeah the terminal is the best [12:44] but if you want like normies and regular [12:46] people to use your software you got to [12:48] give them a button. [12:49] >> Yes. Yeah. Just just a button. Just one [12:52] button. [laughter] Yeah. [12:54] >> Frenchie says, "I would like to make [12:57] YouTube videos for code tutorials. Is [12:59] wondering if you could give us your [13:00] setup tools and software to make good [13:02] videos with code at an affordable price [13:04] if possible. Maybe it's more Ry's [13:06] expertise right now, but when you guys [13:08] were already doing a good job, any tips [13:10] would be appreciated. Yes. What are your [13:12] tips, Scott? [13:14] Oh, yeah. My tips. So, I I think a lot [13:18] of it will come down to beyond just the [13:21] software that you're using. The [13:23] fundamentals here are truly making sure [13:25] the code is big enough to read. Uh, [13:28] making sure that your windows are are [13:30] organized, that like it is pleasant to [13:32] watch, that you're not making people [13:34] struggle for it. Uh, making sure that [13:37] your audio sounds good. You can get a [13:39] decent sure microphone that's USB [13:42] powered for 150 bucks. That's going to [13:45] sound endlessly better than what so many [13:47] people try to use. you got that you have [13:50] your code nice and clear and then you [13:53] cut out the dead space with editing and [13:55] stuff like that will get you way further [13:58] than really like futing around the [14:00] edges. So, as far as like recording [14:04] goes, if you want the easiest possible [14:07] setup, I I still think screen something [14:09] like ScreenFlow is the easiest possible [14:11] setup because it includes you can record [14:14] everything. You know, you can record any [14:16] screens or or video that you want audio [14:18] and then edit it right in there. I [14:21] personally use Da Vinci Resolve. It is [14:24] free. It is a professional linear [14:27] editing software, but it is very [14:30] powerful and it is very free. Uh, so if [14:34] you don't want to spend any money and [14:36] you don't mind a bit of a learning curve [14:38] and you you can gain a skill, um, Da [14:41] Vinci Resolve is fantastic. Now, if you [14:43] want my pro setup, the pro setup is to [14:48] record all of your sources some way. [14:50] People use OBS. It's a pain. I made my [14:52] own app, Vramer, that I'll I'll post [14:55] here. It's not available, but it will be [14:56] by the time you're hearing this, that [14:58] records all individual sources. My pro [15:00] setup is to record all your sources. You [15:03] drop them into an app called Recut. [15:06] ReCut then finds all of the blank space [15:09] where you're not talking and it makes a [15:11] cut. You can then go right from ReCut to [15:14] Da Vinci Resolve. So I can do a first [15:18] pass edit with recut to say hey remove [15:20] all of the blank space maybe the ums the [15:22] o's whatever makes it really nice and [15:24] easy for many layers at once and then [15:27] you say from recut send to Da Vinci that [15:30] sends it to Da Vinci and then I just [15:31] finish it off there where I do all my [15:34] fit and finish and polish there. So that [15:36] is my my current flow is Vramer to recut [15:40] to Da Vinci. Yeah, recording sources and [15:43] not screens is something that has been [15:48] doable in the last I don't know couple [15:50] years and it's it's a game changer [15:52] because what I have done previously is [15:54] you just try to have all of your windows [15:56] open and then what I'll duplicate the [15:59] tracks and then I'll just cut each one. [16:01] have a one track terminal, one track for [16:03] code, one track for the browser, and [16:05] that's great because [16:06] >> yes, I [16:07] >> if you can record the app, if you tab [16:10] away or you bring over something over [16:12] top of it, you don't have to show that. [16:14] Um, and and that's that's sometimes very [16:16] frustrating when people tab away and you [16:18] have to try scrub the video back [16:20] >> to where they were or if something is [16:22] over top of of them or or the really [16:24] annoying one is like when people's [16:26] >> face is over top of the part that you [16:29] want. M. Mhm. [16:30] >> Also think about captions. That's [16:32] something as well as I try not to put [16:34] important code in the bottom like maybe [16:37] 20% of it [16:39] >> because people who have captions on lots [16:42] of people everybody watches with [16:44] captions now is just covers over top of [16:47] the the important bits. I'm actually [16:49] thinking about my course platform is [16:51] just moving the captions [16:53] >> off the video. You know, just put them [16:55] just put them in a div underneath the [16:57] the video player. Wouldn't that that [16:59] seems like a better idea? I'm probably [17:01] going to do that. [17:02] >> Yeah, I I I will say what's I got to get [17:05] UV framer because you can just grab as [17:08] many sources as you want. Whatever the [17:10] sources are, you just spit those out as [17:12] MKV. There's no processing time. And I I [17:16] did my last one of my last videos with [17:18] it. And it was like you said, you get to [17:21] then focus on the uh when you're trying [17:24] to record your screen, you're you're [17:27] doing your your performance, you're [17:29] speaking, you're talking, you're coding [17:30] while trying to manage and move around [17:32] the windows. That a way that makes sense [17:34] instead of like focusing on what you're [17:37] doing and then assembling it later. So [17:40] recording the sources I think is like [17:42] you mentioned a huge boon. Is boon a [17:44] good thing or a bad thing? [17:45] >> Uh I think it's a good thing. Yeah. It's [17:47] really good. And like audio wise as [17:49] well, [17:50] >> Scott mentioned this as well, get a [17:52] decent mic. Figure out how to how to [17:56] like make it work well. Put a freaking [17:58] t-shirt under your keyboard or whatever. [18:01] The most annoying thing is when you [18:03] somebody has vibrations in it, like [18:06] understand how sound works, but if [18:07] you're banging away on your keyboard, [18:10] not like like mechanical keyboard [18:11] sounds, a lot of people like that, but [18:13] they're just like [18:16] >> Yeah. you know, like just coming into [18:17] the microphone. That is so obnoxious to [18:19] hear. And as well as like [18:21] >> if if you hear people's like mouth [18:23] sounds or whatever. [18:24] >> Oh, mouth sounds [18:25] >> or [gasps] [18:27] >> like when they're you're breathing in [18:28] every single time. Those are explosives [18:31] and s's. All of that stuff can be taken [18:33] out with software or hardware. But don't [18:36] don't go down the rabbit hole too much [18:38] of just hardware software learning video [18:41] editing as well. focus on like making a [18:45] good video first. Like the content is [18:48] really what matters at the end of the [18:49] day. Some of my stu like worst quality [18:51] videos ever have been been like our [18:54] highest views and then some of the stuff [18:56] we spend forever on get no views. So [18:59] it's it's all about the content, right? [19:01] >> Yeah. Content and then the 8020 stuff [19:03] like audio that's not going to make your [19:06] viewers turn away. [19:08] >> Yeah. Exactly. get that stuff just like [19:09] dialed in and then and also make it easy [19:12] as well so that you could just hit [19:14] record and go. You know, it's not like, [19:16] oh, I'm going to spend some time like I [19:17] know people that have separate rooms for [19:19] recording or they have to book a [19:21] conference room. You're never going to [19:22] make content if it's that if there's [19:24] that many hurdles. [19:25] >> I'm going to throw out a mic [19:26] recommendation here. We just got this [19:28] mic for Courtney and it's a Sure MV7 [19:31] Plus. It has USBC into the microphone, [19:35] but it also has a dynam like it has a [19:38] compressor in the mic. It has a pop [19:40] filter. It has like digital D onboard [19:43] DSP. It has all kinds of stuff. So that [19:46] way it like has like gating and and and [19:48] those types of things. So you'll sound [19:50] good. Just plug it in USB. Yeah, this [19:52] microphone is effortly easy. So my wife [19:55] is sitting outside my door right now and [19:57] listening to me. She says, she texted [19:59] me, "Not you complaining about breathing [20:01] sounds. Tell them about how you had to [20:03] get a tent when you went camping." Yeah, [20:04] I went camping. My poor sister has [20:06] asthma. So, she would sit in front of [20:08] the campfire and have like asthma [20:09] attacks. [20:10] >> And then I would get so mad at her for [20:12] breathing cuz I cannot stand the sound [20:16] of someone breathing makes me want to go [20:20] baddy. [20:21] >> That's me for eating in and the mouth [20:24] sounds when people are eating. Oh, [20:25] >> it just Oh, it like gives me the [20:28] >> sounds of life. [laughter] [20:30] >> Yeah. [20:31] >> Sir Jason Yaml 3 says, "How do I go [20:34] about a disagreement with another [20:37] developer who's been at the company [20:38] longer than me? My coworker crashed [20:41] three of our apps because they did a big [20:44] database production without telling me [20:46] and without testing it. They deleted my [20:48] feature branches, changed deploy [20:50] settings, [laughter] [20:53] which caused more prod downtime. They [20:55] can they cannot take feedback. They seem [20:57] to just do whatever because they've been [21:00] there so long and we can't come to an [21:01] agreement on change of jurisdiction or [21:03] communication. They get angry and snippy [21:06] whenever I bring up major issues like [21:07] this and it turns into a back and forth [21:10] matter of the cause. O, that sounds [21:12] awful. First of all, you probably [21:14] shouldn't be in a situation where [21:18] somebody else can just nuke your entire [21:20] thing. So I guess ideally you'd have [21:22] have like a local implementation of [21:25] something like that. What do you think [21:26] Scott? Like what somebody is this awful? [21:29] I don't know. I've I've never worked [21:30] with awful people like this. [21:32] >> Yeah, that it is tough. I've never [21:34] worked with somebody this awful. Uh but [21:38] there has to be some kind of so [21:40] managerally there has to be some kind of [21:43] process in place where this isn't just a [21:46] like fighting back and forth kind of [21:48] thing like oh don't do that do this do [21:50] that like you there needs to be some [21:53] process in place where you work whether [21:55] that's from top down or whether that's [21:57] from something you're getting agreement [21:58] upon from your boss. So that way when [22:00] these things happen, you have a sign to [22:03] point at that says you can't do that. [22:06] You're not allowed to do that. And then [22:08] if they're breaching that, then it [22:10] becomes an on them type of thing. But if [22:14] this is just kind of ethereal and like, [22:17] oh, we're just kind of just throwing [22:20] code at the wall here and there and [22:21] oops, they deleted my thing. then yes it [22:24] sucks but your amount of actual control [22:29] over the situation becomes way more [22:32] immature and way more like less [22:36] processbased. So I would push for some [22:41] kind of actual structure here whether [22:44] that is from above. I would talk to [22:47] their manager and let them know this has [22:49] been happening and document these [22:52] things. [22:53] >> Yeah. [22:54] >> The one thing that I think always comes [22:56] in handy in these situations is when [22:59] they happen [23:01] document them very clearly. And when you [23:06] speak to your manager, when you speak to [23:09] somebody, the best advice I got was [23:12] like, be honest. Don't pull any punches. [23:16] Be honest. This is how it is. This is [23:18] how it's been going. These are the [23:20] things that have happened and it's not [23:23] acceptable to me. Like what what's the [23:25] path forward here? But if you're just [23:27] like, "Ah, it's been bad. This person's [23:29] been a jerk to me." Then the resolution [23:31] is just not going to happen. There's a [23:33] pretty clear way forward here. The fact [23:37] that they crashed the app three times [23:39] and caused [23:40] >> almost like what is like it's hurting [23:43] the business, right? I think you just go [23:45] forward be like look we need to put [23:46] processes in place so that this does not [23:49] happen again and that we don't lose we [23:51] don't have an even more expensive loss [23:54] of data or something like this. It's [23:56] it's pretty you go whoever is this [23:59] manager be like look I'm trying to [24:00] figure to make it so that we don't have [24:03] this anymore. We don't crash our apps [24:06] and have prod going downtime and and me [24:08] wasting a week of my work. This is what [24:10] I would like to implement and I can't [24:12] get this guy to come on board. So [24:14] [music] how do we fix this? Yeah. Right. [24:16] And if you want to see all of the errors [24:18] in your application, you'll want to [24:20] check out Sentry at centry.io/sax. [24:25] You don't want a production application [24:26] out there that well you have no [24:28] visibility into in case something is [24:30] blowing up and you might not even know [24:32] it. So head on over to centry.io/sax. [24:35] Again, we've been using this tool for a [24:37] long time and it totally rules. All [24:39] right. [music] [24:41] All right. Next question here from [24:43] Sarah. I'm an experienced web dev [24:44] specializing in design systems and [24:46] accessibility for the last 6ish years. I [24:50] feel like I've been typ cast as an [24:52] accessibility engineer and I'm [24:54] struggling with an identity crisis on [24:55] how to portray myself on my personal [24:58] website. Lately, I've developed a strong [25:00] passion for more experimental and [25:02] creative coding projects. I'd love to [25:03] showcase this side of my work, WebGL, [25:06] gamedev, interactive art, generative [25:08] projects on my personal site and attract [25:10] new opportunities to break out of my [25:12] role that I find myself stuck in. The [25:15] problem is creative coding often [25:16] includes visual experimental projects [25:18] that are not and sometimes cannot be [25:20] easily 100% accessible. My professional [25:22] work, however, is deeply rooted in [25:24] promoting and implementing accessible [25:26] web experiences. My question is, am I [25:28] fundamentally jeopardizing my [25:30] professional reputation and design [25:32] systems by pursuing and sharing [25:34] inaccessible creative projects? How can [25:35] I showcase my wider skill set and [25:37] passion for experimental coding without [25:39] undermining my credibility I've built as [25:41] an accessibility expert? You know, this [25:43] one uh is so fascinating to me. I think [25:46] it's all in how you package it. I think [25:48] if you're labeling things like this is a [25:50] creative exploration, this is a creative [25:52] project. I don't think people are [25:54] expecting a WebGL visualizer to be [25:57] accessible to blind folks, you know, [26:00] vision impaired folks, right? That's [26:02] actually something that bothers me with [26:04] uh some and not all all accessibility [26:06] folks, but some accessibility folks will [26:08] look at any experience and be like, "Oh, [26:11] you made a you made a music player. [26:14] Well, what about hearing impaired folks? [26:16] They can't use [26:16] >> Yeah. You you hate blind people. [26:19] >> Yeah. Like so like [laughter] [26:21] that drives me nuts. Not every [26:24] experience can be the exact same for all [26:28] abled or disabled people. Just straight [26:30] up. it can't be. So, I think it's all [26:33] about how you package those explorations [26:36] and works. And by all means, I think [26:38] this is a cool avenue for you and I [26:40] think you should uh lean into it. But [26:42] just label your stuff. Say this is [26:44] accessibility work, this is creative [26:46] coding, this is that, this is this and [26:48] and just make that very clear. I don't [26:50] think if you put a creative coding [26:51] project and and label it as like this is [26:53] a creative exploration or whatever on [26:55] your site, I don't think anybody's going [26:57] to care about that. I think that is a a [26:59] kickass way to get a new job because [27:02] I'll tell you I'll tell you what as [27:04] someone who's hired several people. You [27:06] know what the Slack room looks like when [27:08] people are going through resumes? Check [27:09] out this sick website. That's all it is [27:12] is if there somebody's portfolio has a [27:15] list of really cool things or check out [27:18] this tweet of somebody that built this [27:20] wild thing. And honestly, I think the [27:23] accessibility stuff can be like a bit of [27:25] a Trojan horse into the company because [27:28] you can like win them over with all the [27:30] interactive, all the the artistic, all [27:33] of the next level JavaScript stuff [27:34] you've been building and then bam, guess [27:37] what? Also care about accessibility once [27:39] you you get into the the company. So, I [27:42] don't think I think I think absolutely [27:44] go for it. This is a a great way to [27:46] expand your career, especially right [27:48] now. We said it on our predictions. 2026 [27:51] is the year of WebGL. It's the year of [27:53] 3D. Everybody is going to care a lot [27:56] more about the design and the like wow [27:59] experience of a website cuz every single [28:01] website currently looks exactly the same [28:03] and we're all sick of it. [28:05] >> Yeah. Right. Just think about like that [28:08] Lando Nora site. People saying that's [28:10] not accessible. That's like an art [28:11] piece. That uh Shopify's thing, that's [28:14] an art piece. And like those types of [28:16] experiences are so stinking cool. And if [28:19] you can get in on some of that territory [28:21] right now, I think it's a great time to [28:23] do so. [28:24] >> By the way, Scott, I can hear you [28:25] typing. Your your is going into your [28:28] microphone. You need a shock mount. [28:30] >> I'm trying to get this the It's actually [28:32] my microphone's way over here on a shock [28:34] mount. My computer's right here. I think [28:35] it's just the clickies are too close to [28:37] where the microphone is. [28:38] >> Oh, I see. I [28:39] >> I need to move it over here a little bit [28:40] more. [laughter and gasps] I was talking [28:41] to Killian who works on Poly Pane. [28:45] That's the browser for web development. [28:48] That's amazing. If you want to check out [28:50] building responsive and accessible and [28:52] everything, but um I tweeted something [28:54] out about like, you know, like if you [28:56] look at the Chrome dev tools, if you put [28:59] like white text on a red background, [29:02] it'll be like that's not accessible. And [29:04] you're just like looking at it being [29:05] like that's extremely easy to read. Why [29:08] are these accessibility contrasts [29:11] algorithms [29:12] so crappy, you know, and it turns out, [29:14] we talked to Jen Simmons from Apple as [29:16] well. We talked about the color contrast [29:18] function, why that's not going on. He [29:20] was telling me this whole world of [29:23] there's like this new color contrast [29:27] algorithm that's being worked on [29:29] >> that is way more [29:32] I don't want to say permissible, but [29:34] it's like it's better, you know? It's [29:35] like you look at something where like I [29:37] can read that. I think most people could [29:39] read that. It's a lot better. But he was [29:41] telling me about all the like drama [29:44] going on in the accessibility world as [29:46] to if this is good or not. And it's it's [29:49] cool. So I all I'll tell you right now [29:51] is like the the rules are still like the [29:53] color contrast stuff is still still [29:55] sucks. And a lot of people still have to [29:57] follow those cuz they their company will [29:59] get sued. But hopefully soon we will [30:02] have better algorithms for calculating [30:04] color contrast. There's a really good [30:06] blog post from Leia Veru on this talking [30:09] all about uh some of these different [30:11] strategies, but also outlining uh how to [30:14] be using modern CSS. And it's funny, I [30:16] actually took some of her work and uh I [30:20] I'll share my code as well. I took some [30:22] of her work and like reapplied that with [30:25] some clamps and some light dark [30:26] functions and have what I think is a [30:29] pretty dang good like 90% of the time [30:33] auto color contrast CSS where you can [30:36] just do like auto contrast on this and [30:39] it works with transparent backgrounds. [30:42] It works with any colors and uh I I'll [30:44] share the the code for this thing, but [30:46] like I think it we're getting closer to [30:48] being able to like I I think that's like [30:50] one of the big things especially with [30:52] with automatic color contrasting in the [30:54] browser is that like [30:57] >> you're not going to be able to usually [30:59] apply a blanket function right now that [31:02] just takes care of all situations. But [31:04] if you can if you can fix 99% of [31:07] situations and then like do the rest by [31:09] hand, that gets you a long a decent way. [31:11] Yeah. Killian actually has a website [31:13] colorcontrast.app [31:15] and there's like three levels of color [31:17] contrast, right? There's WC A2 [31:21] >> and you can you either have to be AA or [31:23] AAA compliant. Um, and those are the [31:25] ones that are like, oh, you can't put [31:27] white on red button. That's that's [31:29] annoying. And the new one, which is is [31:31] not a standard yet, but it's what [31:32] they're working on is called the APCA. [31:35] Um, that one seems to be a lot more [31:37] permissible. So, check out [31:38] colorcontrast.app from Killian. Yes, the [31:41] APCA. [31:43] Yes. And the APCA is also um talked [31:46] about quite a bit in the slaver blog [31:48] post, so you can uh pick some of that [31:50] up, too. John Ray says, "What are you [31:52] using for your smart home?" I have an [31:54] old open HAB, open HAB 2.4 from 2018, [31:58] but I really need to migrate to [32:00] something newer. I'm on Home Assistant [32:02] running it on my Synology. I've talked [32:04] about that in the past, but you don't [32:06] need a lot of heavy hardware to run this [32:09] stuff is [32:11] >> it's just handling MQTT streams and and [32:14] setting variables across the the world. [32:17] So, [32:18] >> that is a very good one. I also run [32:21] Google Home as well, just cuz I love [32:23] having the screens all over the house [32:25] and being able to just talk to it. But [32:26] for most of my like scheduling [32:28] automation, I use Home Assistant. Yeah, [32:31] I use Home Assistant as well. I have it [32:34] running on a Home Assistant OS that just [32:36] runs on a Raspberry Pi that I just [32:38] always have on over there plugged into [32:40] that. I have like a Ziggby adapter into [32:42] that directly and that's what I run. I [32:44] like it a lot. I actually just recently [32:46] just wrote a little automation the other [32:48] day that automatically turns on my [32:49] bathroom exhaust fan when the humidity [32:51] gets too high in the primary bathroom. [32:54] It was nice and easy. Took no time at [32:56] all in Home Assistant. And that's just [32:58] the kind of stuff you want with a smart [33:00] home, right? It's like, [33:02] >> yeah, [33:03] >> now I don't got to turn on the exhaust. [33:05] [laughter] I can just [33:07] >> Home Assistant is not for the the faint [33:09] of heart. I think they need to figure [33:10] out some better. It's gotten a little [33:12] bit better [33:12] >> UI and some better like what happens [33:15] with Home Assistant is you add like like [33:17] a light bulb [33:18] >> and then it'll be like here's 600 [33:20] sensors that this light bulb can do, you [33:22] know, and it's just you get to a point [33:23] where you add like I have I have [33:25] probably 40 50 smart things in my house [33:29] >> and then each of them have seven or [33:30] eight different things. You have to [33:32] spend some time really culling it down [33:35] and getting it to as simple as you want. [33:37] So hopefully I don't know. I don't know [33:39] what it that's the thing with any of [33:41] these Swiss Army tools. You let people [33:43] do literally anything [33:44] >> and then it gets to be a little bit [33:47] overwhelming. [33:48] >> That's your home assistant dashboard. [33:50] >> Yeah, it's pretty nicely tuned. And one [33:52] thing that I like about mine is on this [33:54] homepage, like the homepage, I have like [33:56] a bunch of just like presets where it's [33:59] like wa uh watching a movie, eating [34:02] dinner, whatever. And a lot of those are [34:04] on timers, but at any given point, I can [34:06] enable one. And then on any given room, [34:08] I just have all the individual lights [34:11] and whatever set up or like my office is [34:14] a little bit different cuz it's my [34:15] office where I can have like master [34:17] toggles with everything in there and all [34:19] of my studio lights at one. [34:21] >> I have one for my office as well. I have [34:22] it hooked up to my stream deck as well. [34:24] I can turn the lights on and off. [34:26] >> Yeah, love that. [34:27] >> I really want to build some sort of like [34:29] family hub. Um, and on Marketplace the [34:32] other day, somebody had one of those [34:34] Samsung fridges with the TV in it. [34:36] >> Yeah. [34:36] >> And it was had like a blown compressor, [34:38] and I was like, "Oh, I could take the [34:40] screen out of the door." But he wanted [34:43] like 500 bucks for it. So, [34:45] >> that's an absurd. [34:46] >> You could just buy a little tablet for [34:48] like 70 bucks. [34:50] >> Yeah. Yeah. I'm waiting for him to like [34:52] that thing's trash. Nobody wants your [34:54] broken fridge, you know? I'm just [34:56] waiting for him to come down and realize [34:58] that. and I'll give him a hundred bucks [34:59] for the the door or something like that [35:03] >> instead of buying Instead of buying a [35:05] $50 tablet that works just [35:08] >> No way. What $50 tablet? I want a I [35:10] don't want a tiny tablet. I want a huge [35:12] ass screen that has all of our [35:15] appointments, all of our buttons, [35:17] everything. You know, [35:18] >> the problem for me is I have nowhere to [35:19] power that. Like, so what? Like you're [35:21] running power through the wall and then [35:24] >> Oh, yeah. [35:25] >> Easy. And the stuff runs on on low [35:28] voltage, so you can you could probably [35:30] even do it over PoE. [35:33] >> Probably. I don't know. [35:34] >> H maybe not. The screen those screens [35:36] are pretty hungry. We'll see. Solazoo [35:38] says, "I'm listening to potluck 946 and [35:41] Wes mentions he blocks a large number of [35:44] ISPs for bad hosts from all of his [35:46] sites. Where does such a list come from [35:49] that he cured himself?" So there's a [35:52] GitHub repo called bad ASN list. So, an [35:55] ASN is, you can go deep into what an ASN [35:59] is. It stands for autonomous system, but [36:01] typically a hosting provider or an ISP, [36:05] which which a hosting provider is, [36:08] they'll all have their a unique number [36:10] that's attached to it. You know, Comcast [36:12] is 7922. Digital Ocean is 14061. [36:17] And [36:19] depending on the type of site that [36:21] you're operating from, it makes sense to [36:23] simply just block all people that are [36:26] coming from or all requests, not even [36:29] people. So like if you're getting [36:30] requests to your website from like a [36:33] shady ISP in Russia that are like this [36:37] is not people's homes, you know, this is [36:39] not like a cell phone network. This is [36:41] like somebody is making requests from a [36:43] data center somewhere. And if that's why [36:46] is that the case? either somebody's on a [36:47] VPN trying to hide something or or [36:50] somebody's running some sort of bot farm [36:52] and they're they're sending uh requests [36:54] your way. So instead of blocking [36:56] individual IPs or or or blocks of IPs, [37:00] you can simply just say requests from [37:02] these ASNs are going to be either [37:05] totally blocked or you could throw a [37:07] capture at them. Um, I think that's what [37:09] I do right now because it's like there [37:11] is a case where somebody would be using [37:12] a VPN um that's running on one of these, [37:15] but in my case, I've never heard from a [37:18] single customer saying, "Hey, your [37:21] website's blocked." That's never been [37:23] the case. It's it's it's only blocked [37:26] shady people and shady bots. And the [37:28] same thing with like the way that you [37:30] implement it is you go into your [37:31] Cloudflare and you can create a rule and [37:33] it's just this massive rule where he [37:34] says if the ASN is XYZ [snorts] [37:38] you can block it. Um but you can also [37:40] block like countries as well which you [37:43] probably shouldn't block countries but [37:45] you can block everybody who is from the [37:48] XX country or the I think the tour [37:51] country. And anybody who's who's [37:53] visiting your website via tour is [37:56] probably doing shady stuff as well, [37:58] trying to hide their tails. So you can [38:00] you can hide those. All right, next [38:02] question from Sadique. Hey guys, big [38:05] fan. I really enjoy the podcast. In my [38:07] daily job, I work with modern web [38:09] technologies, but over the last year, [38:11] due to a new direction in our company, [38:13] I've had to start learning SAP UI5. It's [38:17] basically a web world inside SAP. Have [38:21] you ever come across this technology? [38:23] I'm still surprised how big the [38:25] ecosystem of languages and frameworks is [38:27] inside of larger enterprises and how [38:30] little it's talked about. It feels like [38:32] a hidden world with a lot of potential. [38:34] Wes, I'm going to say I have zero clue. [38:38] I have no clue. Nothing. [38:40] >> Yeah. I thought this was an interesting [38:42] question because when I was in school, a [38:45] big thing that we had to learn was [38:46] called ERP. Um, and like these large [38:49] companies who do accounting and [38:52] inventory, sourcing products, like I [38:55] remember the one example we had to do [38:56] for SAP was somebody was making granola [39:00] and you had 18 vendors from around the [39:04] world to get all of the pieces to make [39:06] your granola. you know, you had to get [39:08] almonds from Mexico and you had to get, [39:11] I don't know, like chocolate chips from [39:13] Canada and like they all had different [39:15] prices and they all have different lead [39:16] times and in order to make a simple [39:19] granola, it was a very complex thing. [39:21] And then you had um accounts payable, [39:24] ledgers, all of that stuff sort of built [39:26] into it. And these huge companies [39:29] >> are built on SAP. SAP is a massive, [39:32] massive company. And I remember hating [39:34] my life doing this because it was a Java [39:37] app where it was just looked awful and [39:39] you never knew where to click and and [39:42] nothing ever worked and it was errors [39:43] out the wazoo. Um and you had to do it [39:46] on a desktop. And I was thinking about [39:48] that. I was like, I wonder where SAP's [39:49] at, you know. So this guy says that SAP [39:52] and I just looked it up. There's this [39:54] thing called SAP UI5. It looks like they [39:57] moved everything over to their own [40:01] custom MVC JavaScript framework. And [40:04] that's cool because like obviously this [40:06] stuff needs to be able to run [40:07] everywhere. This is why you should bet [40:08] on the web. This like massive company [40:11] which was built on Java is now [40:14] >> the entire UI is all done in JavaScript. [40:16] I went to the docs and I thought it was [40:17] hilarious because they have a a chart of [40:20] like how things work together and that [40:23] chart is done in a PNG that's probably [40:26] exported from [laughter] [40:28] Photoshop or something. So, I just [40:30] thought it was so funny like these big [40:32] companies are just like they can't even [40:34] make labels and boxes with text. They [40:36] have to export it. [40:38] >> Yeah. Right. Yeah. Why why does this [40:40] have to be an image? [40:42] >> What in the world? But it's it kind of [40:46] they built their own like it's not built [40:48] on react or angular or spelt or [40:50] anything. They ma made their own MVC [40:51] framework. Probably makes sense for how [40:54] big of a piece of software this is. But [40:56] that is certainly an area where you can [40:59] I remember in school everyone's like you [41:00] can make a bank if you know SAP similar [41:03] to like SharePoint as well you know [41:06] where it's just like I'm specializing in [41:08] this odd thing and I know I guess React [41:11] is the same way. I know all the ins and [41:13] outs of this thing and I'm gonna make [41:16] bank charging these massive companies [41:18] who just want to make granola on how to [41:20] implement it. [41:22] >> Yeah. Crazy. Yeah. This is a whole world [41:25] where I am just completely out to lunch [41:27] on. Cool. Well, now is the part of the [41:29] show where we get into sick pics and [41:31] shameless plugs. Sick pics for those of [41:33] you who don't know are the things that [41:34] we just like and enjoy in this world. [41:36] Stuff that we've been feeling. And you [41:38] know what? Before we do this, we don't [41:40] ask people to like and subscribe. Like [41:42] and subscribe to Syntax, y'all. If [41:43] you're listening to this, you enjoyed [41:45] this show, like and subscribe. Hit that [41:47] button, you know, do all that stuff for [41:49] us. That'd be great. We We love a little [41:51] subscriptions here and there. I am going [41:53] to sick pick a YouTube channel. And this [41:56] is a YouTube channel that I just found [41:58] out about last night. You know, there's [41:59] always like when you're like discover a [42:02] YouTube channel and you're like, "Wow, [42:03] this is really high quality information. [42:06] I wonder how long they've been doing [42:07] this." And then you go and you see it's [42:09] like, "Oh, they have a video that's 12 [42:10] years old. They've been doing it for a [42:12] long time and this channel has 45,000 [42:16] subscribers and they've been it it was [42:18] really really good. This is called [42:20] Inside Archaeology and it was just [42:24] really nice archaeology information. [42:26] Sometimes it's like new archaeological [42:28] discoveries in October 2025, fascinating [42:32] uh burials, you know, the hunt for Queen [42:34] uh Cleopatra's lost tomb. And like the [42:37] hardest thing with this kind of content [42:39] on YouTube is that so many of the [42:41] accounts that blow up is like this like [42:44] hyperbolic [42:46] like was the pyramids built on giant [42:49] tubes underground like [42:51] >> like just idiotic Joe Rogan brain kind [42:53] of stuff instead of like actual [42:55] information. So uh inside archaeology [42:58] really great channel and uh it just [43:01] nononsense real stuff. So, uh, check it [43:04] out if you are like that. If you like [43:06] history or archaeology or any of that [43:07] kind of stuff, this is my ideal type of [43:09] YouTube, honestly. [43:10] >> I'm going to sick pick a YouTube channel [43:13] as well. And this is a guy called [43:15] Professor Boots. And this guy makes 3D [43:19] printed like RC cars. And it's just the [43:23] most beautiful thing to watch because he [43:27] goes through the like the the 3D [43:30] modeling of these RC cars. like he makes [43:33] like a a skid steer and a bulldozer and [43:35] a log truck and a what else? Basically [43:39] any any vehicle that is in like heavy [43:41] equipment. He's building it. He's [43:43] modeling it, 3D printing it. Then he's [43:46] he's making the electronics to make it [43:48] work with like an ESP32 and then he [43:51] writes the software to control it. And [43:53] then he puts them all into these very [43:55] like digestible, [43:58] fun to watch YouTube videos. And then he [44:01] also has like a like a workshop. You can [44:03] subscribe for 15 bucks a month and you [44:05] can get access to all the guides on how [44:08] to build them. And I was like, "Oh, over [44:10] Christmas I am going to be building I'm [44:12] going to be building a bulldozer or [44:13] something fun like that cuz I have [44:15] experience with all of these things. You [44:17] know, I can 3D design, I can code, I can [44:19] do hardware, and I would love to make [44:21] like a custom syntax bulldozer or [44:24] something like that." [44:25] >> Wow. What a a 5-year-old sentence. a [44:28] ghost in tax or a bulldozer. [laughter] [44:32] >> And the other thing is that like I watch [44:33] it with my kids and they're just like [44:36] >> we got to do that, you know, like they [44:39] actually work, you know? They can you [44:40] can pick things up and push it around. [44:42] >> It looks pretty dope. [44:43] >> Yeah, [44:45] on YouTube. [44:46] >> All right, I'll try this out. Sick. Heck [44:48] yeah. Cool. Shameless plugs. Check us [44:50] out on YouTube. Syntax YouTube. Check us [44:53] out. We're doing all kinds of stuff on [44:54] there. We're doing crazy stuff. So, uh, [44:56] as always, we'll catch you in the next [44:58] one. Peace. [44:59] >> Peace.