PHP Alive And Kicking Episode 17 - TJ Miller

PHP Alive And Kicking: Episode 17 - TJ Miller 🔗 **Source:** [PHP Alive And Kicking: Episode 17 - TJ Miller](https://www.youtube.com/watch?v=nzeTSCQLd3M) > [!note]+ Document Notes > tags: [php, laravel, prism, ai-integration] > author: PHP Architect > source: YouTube > reference: [https://www.youtube.com/watch?v=nzeTSCQLd3M] > date created: 2024-12-03 14:30 > > # 🏷️ Title: PHP Alive And Kicking: Episode 17 - TJ Miller >  ## 🧰 Action Inbox - [H1] Consider PHP Foundation AI integration opportunities — Prism success shows demand #ai #community - [H2] Research local model deployment with Ollama — reduce token costs for testing #development #cost-optimization - [H3] Pitch Laravel-focused AI talk for conferences — bridge PHP/JS communities #speaking #outreach ## 🧭 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 - [Prism - AI SDK for PHP](https://github.com/prism-ai) - [Geocodio](https://geocodio.com) - [PHP Tech Conference](https://phptech.dev) ## 📝 Summary - TJ Miller discusses his journey from HTML/CSS to PHP, creation of Prism AI SDK, and current role at Geocodio - Prism emerged from need for Laravel-native AI tooling, inspired by Vercel's AI SDK - PHP community faces image problem affecting junior developer recruitment - Strong emphasis on community building and conference participation as key to PHP's growth - Technical discussion covers Neo Vim configuration, local model testing, and multi-provider AI integration ## 📖 Main Content TJ Miller, creator of the Prism AI SDK and developer at Geocodio, shares his evolution from basic web development to building sophisticated AI integrations in PHP. Starting with simple HTML templating needs in high school, Miller discovered PHP's power for dynamic content generation and database connectivity. His journey led through WordPress development, a brief consideration of Ruby on Rails, and ultimately deep commitment to Laravel. The turning point came when Taylor Otwell mentioned wondering what a Laravel version of Vercel's AI SDK would look like, sparking the creation of Prism. The conversation addresses PHP's ongoing image problem, particularly in recruiting junior developers who gravitate toward JavaScript and Python bootcamps. Miller advocates for positive messaging about PHP's capabilities rather than defensive responses to "is PHP dead" narratives. Technical insights include strategies for managing AI token costs through local model testing with Ollama, the importance of API-first design in developer tools, and the challenge of supporting multiple AI providers with evolving APIs. ## 🗒️ Notes #### ⏱️ [00:00](https://www.youtube.com/watch?v=nzeTSCQLd3M?t=0) — [H1] Introduction and Setup - Key idea: Live podcast format with Discord integration for audience questions - Evidence: "Please head over into our discord. You can ask questions of our guests" #### ⏱️ [02:30](https://www.youtube.com/watch?v=nzeTSCQLd3M?t=150) — [H2] TJ's PHP Origin Story - Key idea: Started with basic templating needs in high school web development - Evidence: "I wanted a way to templatize some of the work instead of having to copy and paste the sidebar, the header" #### ⏱️ [05:15](https://www.youtube.com/watch?v=nzeTSCQLd3M?t=315) — [H2] The Laravel Discovery - Key idea: Laravel transformed Miller's perception of what PHP could be - Evidence: "I was blown away. I'm like you could write PHP this way. That changed everything for me" #### ⏱️ [08:45](https://www.youtube.com/watch?v=nzeTSCQLd3M?t=525) — [H2] Vim/Neovim Mastery - Key idea: 20-year commitment to Vim has created deep muscle memory dependencies - Evidence: "There are key bindings that have been in there for 20 years that are so ingrained in muscle memory" #### ⏱️ [12:30](https://www.youtube.com/watch?v=nzeTSCQLd3M?t=750) — [H1] Prism Genesis Story - Key idea: Prism emerged from Taylor Otwell wondering about a Laravel version of Vercel's AI SDK - Evidence: "He said in passing, 'I wonder what a Laravel version of that would look like' and the API just kind of hit me" #### ⏱️ [15:20](https://www.youtube.com/watch?v=nzeTSCQLd3M?t=920) — [H2] API-First Development Philosophy - Key idea: Developer experience drives design decisions before implementation - Evidence: "It was very much API first and then backfilling the functionality to meet that API" #### ⏱️ [18:10](https://www.youtube.com/watch?v=nzeTSCQLd3M?t=1090) — [H2] Geocodio Employment Reality - Key idea: Prism is passion project, not full-time income source - Evidence: "I chew through a ton of tokens every month running tests and experiments" #### ⏱️ [21:45](https://www.youtube.com/watch?v=nzeTSCQLd3M?t=1305) — [H2] Token Cost Management Strategy - Key idea: Local models with Ollama reduce development costs significantly - Evidence: "The OpenAI GPT-O OSS model is very impressive. So for building out integrations you can get pretty decent mileage" #### ⏱️ [25:30](https://www.youtube.com/watch?v=nzeTSCQLd3M?t=1530) — [H1] PHP's Image Problem Analysis - Key idea: PHP suffers from perception issues affecting junior developer recruitment - Evidence: "PHP suffers from an image problem. That's where JS is snagging everyone at the top of the funnel" #### ⏱️ [29:15](https://www.youtube.com/watch?v=nzeTSCQLd3M?t=1755) — [H2] Negative Messaging Impact - Key idea: "Is PHP dead" posts harm perception even when defending PHP - Evidence: "People are just seeing the 'is PHP dead' and assuming PHP is dead. They're not going to open it up and read" #### ⏱️ [33:20](https://www.youtube.com/watch?v=nzeTSCQLd3M?t=2000) — [H2] Demographic Challenge - Key idea: PHP conferences show concerning age demographics - Evidence: "They asked anyone over 35 to put their hands up and pretty much the whole room put their hands up" #### ⏱️ [37:45](https://www.youtube.com/watch?v=nzeTSCQLd3M?t=2265) — [H2] Technology Accessibility Revolution - Key idea: Modern PHP frameworks make cutting-edge technologies accessible - Evidence: "With Prism, now you're not bound to Python. You want to work with AI, not a problem working in PHP" #### ⏱️ [41:30](https://www.youtube.com/watch?v=nzeTSCQLd3M?t=2490) — [H1] Community as Core Value - Key idea: PHP community relationships drive long-term engagement and innovation - Evidence: "The thing that gets me most excited about PHP is the community. Every time I go to conferences I come back so inspired" #### ⏱️ [44:10](https://www.youtube.com/watch?v=nzeTSCQLd3M?t=2650) — [H2] Conference Accessibility Culture - Key idea: PHP conferences maintain inclusive, approachable environments - Evidence: "There's no barrier between speakers and general public. Everyone mingles and there's no clicks" ## 📚 References - [PHP Architect Magazine](https://phparchitect.com) - [Prism AI SDK](https://github.com/prism-ai) - [PHP Tech Conference](https://phptech.dev) - [Geocodio API](https://geocodio.com) - [Laravel Framework](https://laravel.com) - [Ollama Local Models](https://ollama.ai) ## 🗣️ Paraphrase - TJ Miller's path to PHP began with simple templating needs in high school, evolving through WordPress development until Laravel revealed PHP's true potential for elegant code architecture - The Prism AI SDK originated from Taylor Otwell's casual suggestion about creating a Laravel equivalent to Vercel's AI SDK, leading to API-first development focused on developer experience - PHP faces significant challenges in attracting junior developers due to perception issues and negative messaging, requiring positive community initiatives and better marketing of modern capabilities ## 📰 Newsletter **TJ Miller Shares Prism AI Journey and PHP Community Insights** This week's episode features TJ Miller, creator of the popular Prism AI SDK for Laravel, discussing his evolution from basic web templating to building sophisticated AI integrations in PHP. Miller's story demonstrates how modern PHP frameworks enable developers to work with cutting-edge technologies without leaving their preferred ecosystem. Key takeaways for WordPress developers include strategies for managing API costs through local model testing and the importance of community engagement for professional growth. Miller's emphasis on API-first design offers valuable lessons for plugin and theme development. The discussion also addresses PHP's demographic challenges and the need for positive messaging about the language's modern capabilities, particularly relevant for agencies positioning themselves in the current market. Resources: [Learn WordPress](https://learn.wordpress.org/), [Developer Docs](https://developer.wordpress.org/), [WP-CLI](https://wp-cli.org/) ## ✅ Action Items #### DevRel - Developer advocacy: Promote positive PHP messaging in content and presentations - Education: Create workshops bridging traditional PHP and modern AI integration - Developer experience: Document local model testing workflows for cost-effective development - Developer marketing: Position PHP as modern choice for AI/ML applications - Community management: Foster inclusive conference environments that welcome newcomers #### OSPO - Policy & Governance: Evaluate AI SDK licensing compatibility with enterprise requirements - Software Inventory: Track AI integration dependencies in SBOM processes - Developer Support: Provide guidance on managing API costs and local model deployment - Community Engagement: Support PHP Foundation AI initiatives and related projects #### Community - Contributors: Support Prism development through testing and documentation - Sponsors: Consider sponsoring key infrastructure projects like Prism - Ecosystem: Coordinate AI integration standards across PHP frameworks #### WordPress - Core AI: Evaluate Prism integration possibilities for WordPress AI features - Community: Address age demographic concerns through targeted junior developer outreach - Documentation: Create guides for modern PHP development practices - Marketing: Develop positive messaging strategies highlighting WordPress PHP advantages ## 🧪 Hypotheses - Local model testing with Ollama can reduce AI development costs by 70-90% during prototyping phases - API-first design methodology produces more intuitive developer experiences than implementation-first approaches - Positive PHP messaging increases junior developer interest more effectively than defensive responses to criticism - Conference participation significantly increases long-term community engagement and project contributions ### 📊 Methodology - Compare token costs between cloud and local model development workflows across multiple projects - A/B test different messaging strategies in PHP recruitment materials and measure application rates - Survey conference attendees on community engagement levels before and after event participation - Analyze developer experience metrics for API-first vs implementation-first package designs ### ❓ Questions - How can PHP better compete with JavaScript/Python bootcamps for junior developer mindshare? - What specific features would make Prism attractive for WordPress core AI integration? - How might demographic challenges affect PHP's long-term sustainability and innovation capacity? ### 🔮 Future Research - Quantitative analysis of PHP community age demographics across different regions and specializations - Comparative study of AI SDK adoption patterns across different programming language ecosystems - Investigation of local model performance characteristics for different development use cases ### 💡 Implications - Practice: Local model testing becomes viable cost-reduction strategy for AI integration development - Policy: PHP Foundation should prioritize initiatives addressing junior developer recruitment challenges - Security: Multi-provider AI integration requires robust API key management and rotation practices - Education: Developer tooling education should emphasize cost optimization and efficiency techniques - Funding: Community sponsorship models need refinement to support critical infrastructure projects sustainably ### 🧠 Learning Styles Assessment Reading level estimate: 12-14 (technical discussion with specialized terminology) #### 👀 Visual learners - Present: Conference footage, code examples in discussion - Improve: API architecture diagrams, cost comparison charts, community growth visualizations #### 🔊 Aural learners - Present: Full podcast audio with multiple speakers - Improve: 5-minute audio summary highlighting key technical concepts #### 🗣️ Verbal learners - Present: Extensive analogies, clear technical term definitions - Improve: Glossary of AI/ML terms, framework comparison narratives #### 🧩 Kinesthetic learners - Present: Step-by-step development stories, hands-on tool discussions - Improve: Interactive Prism tutorials, local model setup checklists #### 🧠 Logical learners - Present: Clear cause-effect relationships in technology adoption - Improve: Decision trees for AI provider selection, cost-benefit analysis frameworks #### 🧑‍🤝‍🧑 Social learners - Present: Strong emphasis on community interaction and conference networking - Improve: Structured mentorship program guidance, community onboarding checklists #### 🧘‍♀️ Solitary learners - Present: Individual developer journey narratives, self-directed learning paths - Improve: Self-assessment tools for AI integration readiness, reflection prompts #### 🌿 Naturalistic learners - Present: Ecosystem analogies for PHP community growth - Improve: Sustainability considerations for long-term community health ### 📊 Accessibility Assessment #### Screen reader semantics - Present: Clear heading structure, descriptive link text in show notes - Improve: Detailed alt text for any referenced visual materials, ARIA landmarks #### Keyboard navigation - Present: Standard podcast controls accessible via keyboard - Improve: Transcript navigation shortcuts, chapter jumping functionality #### Contrast and readability - Present: Standard text meets basic contrast requirements - Improve: High contrast theme option, adjustable text sizing #### Text resizing and reflow - Present: Responsive layout supports zoom up to 200% - Improve: Optimize for 400% zoom, ensure no content clipping #### Media and captions - Present: Full transcript available - Improve: Timestamped captions, downloadable transcript formats #### Forms and interactions - Present: Simple interaction model with clear feedback - Improve: Error handling for media playback issues #### Standards reference - Align with WCAG 2.2 AA standards for media accessibility ### 🌍 DEIB Assessment #### Cultural breadth - Present: Discussion of global PHP community, conference diversity - Improve: Include international perspective on PHP adoption patterns #### Inclusive language - Present: Gender-neutral terms, welcoming community messaging - Improve: Define technical jargon consistently, avoid cultural idioms #### Neurodiversity support - Present: Well-structured conversation with clear topic transitions - Improve: Provide content warnings for rapid topic changes, offer transcript navigation #### Representation - Present: Discussion of community accessibility and welcoming culture - Improve: Highlight diverse voices in PHP community, inclusive imagery #### Psychological safety - Present: Respectful technical disagreement, supportive community framing - Improve: Explicit community guidelines reference, safe reporting mechanisms ### ✅ Logical Validity #### Strong reasoning - Deductive strength: Clear logical progression from community demographics to sustainability concerns - Inductive strength: Well-supported generalizations based on conference experience and industry observation - Causal reasoning: Solid connection between messaging strategies and perception outcomes - Analogical strength: Effective comparisons between PHP and other language ecosystems #### Fallacies - Survivorship bias: Focus on successful community members may not represent broader developer experience - Hasty generalization: Conference demographics may not represent entire PHP --- ## Full Transcript [00:00] Heat. [00:11] Heat. [00:31] Heat. Heat. [00:59] Heat. Hey, Heat. [01:07] Hey. [01:34] Heat. [01:50] Heat. [02:16] Heat. Heat. [02:21] Heat up [02:43] here. [02:55] Do you lie awake at night? Are you [02:58] afraid to turn out the lights? Is there [03:01] something you're scared of? [03:05] What's there out in the dark? Make sure [03:09] to keep all your windows locked. Do you [03:12] do you surrender? [03:15] Try to try to stay awake. These walls [03:18] you can't escape. [03:21] Here's one for the lonely hearts. For [03:24] the ones that fear the dark. You better [03:26] run out. Run out. Out of here. Bet you [03:29] can't look the other way. Here's one for [03:32] the broken road. That's the only one you [03:36] know. You better run out. Run out. Truth [03:38] or dare. Get me out. Take me anywhere [03:41] but here. [03:46] anywhere but here. [03:52] Wanting something good to come. Waiting [03:56] for things to run away from. Are you Are [03:59] you just hiding? [04:02] Yeah. Walking into the fire, talking [04:06] like you never die. Like you're like [04:09] you're immortal. [04:13] Try and try to stay awake. These walls [04:16] you can't escape. [04:18] Here's one for the lonely hearts. For [04:21] the ones that fear the dark. You better [04:23] run out. Run out of here. Bet you can't [04:27] look the other way. Here's one for the [04:30] broken road. That's the only one you [04:33] know. You better run out. Truth or dare. [04:36] Get me out. Take me anywhere but here. [05:00] Do you lie awake at night? Are you [05:04] afraid to turn out the lights? Is there [05:06] something you're scared of? [05:13] Make sure to keep all your windows. [05:41] Hello and welcome to PHP Alive and [05:43] Kicking brought to you by PHP Architect [05:46] and our partners over at PHP Score. More [05:49] about them later. This is a podcast that [05:51] explores the developments in PHP and [05:54] what it's like to earn a living in PHP [05:56] today. I'm one of your hosts, Mike [05:58] Paige. I'm a developer at PHP Architect [06:02] and my co-host is a longtime contributor [06:04] to our magazine and now team member [06:06] Chris Miller. Hi Chris. [06:08] >> Hey Mike. How you doing? [06:09] >> I'm good, thank you. I've been trying [06:11] you've been chat putting messages in the [06:14] private chat on the left hand side but [06:15] all of a sudden I can't open it. So I [06:17] don't know what you're saying to me [06:18] anymore. [06:19] >> I haven't said anything for [06:22] >> There's messages popping up but I can't [06:23] open it for some reason. [06:25] >> I'm ling I guess I guess. [06:27] >> Yeah. Good, good old good old uh reream. [06:30] Um anyway, today is live everyone. So if [06:33] you are listening right now, um please [06:36] head over into our discord. Uh you can [06:38] ask questions of our guests or us if you [06:41] want to know anything about us. Uh but [06:43] do uh do head over there. You don't have [06:45] to stay there just for the show. You can [06:47] we're in there all week long. There's [06:49] lots of great people you can interact [06:51] with and have conversations with. Uh [06:54] please do like and subscribe to our [06:55] YouTube channel. [06:58] And um that [07:00] >> that is not the YouTube channel. [07:01] >> That's not the YouTube channel. [07:02] >> That's the YouTube channel. [07:04] >> That's the YouTube channel. There we go. [07:06] Um I did mention in the intro there that [07:09] we do have a magazine and we do have a [07:13] discount code for our listeners. A live [07:15] three will give you the first three [07:17] months of a yearly dis [07:21] digital that's the one digital [07:23] subscription. [07:24] Uh, so yeah, do head over there, check [07:26] out the articles in the the magazine. [07:28] Um, always great content each month. I [07:32] cannot talk. I think I should just stop. [07:35] >> No. Okay, I'll carry on. We're not going [07:37] to we're not going to in keep swapping [07:39] over like Eric and John tried on on [07:41] Thursday. No, [07:42] >> no, no, [07:42] >> no, you're leaving to me. All right, [07:43] then. Okay. As well as uh being a a [07:47] development studio and having the [07:48] magazine, we also run annual conference [07:52] PHP Tech uh tickets are now on sale for [07:57] next year. So it'd be great to see as [07:59] many people there as possible. But not [08:03] only that, alongside PHP tech, we are [08:05] also going to be running JStech. The [08:08] same time, same location, [08:11] but purely about uh about JavaScript. [08:14] And I believe the call for papers is [08:17] still open. [08:19] >> It is. [08:19] >> Yes, it is. It's in Is it the first week [08:22] of December it closes? [08:23] >> I have no idea because the date's moved [08:25] twice. [08:26] >> Well, at the moment it's open. So, get [08:28] those talks in. Uh if you've got some [08:31] ideas about uh uh Java uh JavaScript [08:34] ccentric talks, um we also have a swag [08:38] store where you can get some lovely [08:40] t-shirts, hats, caps, all sorts of [08:42] things. Um I think Eric keeps adding new [08:45] new things every now and then. Every [08:47] time he comes up with something that he [08:48] wants, he'll add it to the store. So [08:51] >> um badge him. If there's something not [08:53] in the store, badge him. He might [08:54] actually stick it in there. [08:55] >> Okay, that's all our housekeeping, uh [08:58] Chris. So, [08:59] >> so without further ado comment, [09:04] >> uh, not quite yet. Not yet. A little bit [09:08] early. I know I'm in the UK and it's in [09:10] the evening, but I'm not quite yet. Not [09:11] yet. No, it's it's it's actually the [09:14] opposite. I'm just tired. [09:16] >> So, we're definitely not going to try [09:19] and convert you to become a PHP [09:21] developer at JStech right next to PHP [09:23] Tech. [09:24] >> Oh, we're not. I thought that was our [09:25] whole idea. That was the plan. [09:26] >> Shush. We're not letting that bit out in [09:28] public. [09:29] >> I just I I just thought we were going to [09:30] have to hire security to have PHP [09:32] developers and J and JS developers in [09:36] the same vicinity at the same time is a [09:38] bit scary. [09:39] >> I mean to be honest, PHP will just win. [09:42] So it's fine. And and JStech CFP closes [09:45] on December the 1st. [09:47] >> Oh, thank you. Thank you very much. Um, [09:51] so we we better stop nering because we [09:54] got someone in the green room that's [09:56] sitting there waiting. [09:58] >> Um, they were even dialed in to reream [10:00] before either of us. [10:02] >> They were sitting here waiting when I [10:03] came and I was in there early early. So [10:06] uh Chris over to you to uh introduce our [10:09] guest. So, we have a fellow PHP [10:13] developer, creator of Prism, works at [10:15] Geocodio, [10:17] does streaming, does basically [10:20] everything. We have TJ Miller. Hey, how [10:23] you doing? [10:24] >> Hey, how's it going? [10:25] >> Hi, TD. [10:28] It's been uh it's been one of these this [10:30] is one we've been trying to get in the [10:31] diary for a while. [10:32] >> Yeah. [10:33] >> Yeah. [10:33] >> And then it see once we finally got in [10:35] the diary, it seemed to take ages to [10:37] actually get around to today. But we're [10:40] here and we have [10:42] >> we're here. We're doing it. [10:43] >> Yeah. And of course the important things [10:47] are caffeine. [10:48] >> Yes. Absolutely. [10:51] >> So TJ, you've been watching this for a [10:54] while. You're going to know what the [10:55] first question is. How did you get into [10:56] BHP? [10:58] >> Oh man, this is such a super boring [11:01] story. So um this is many years ago. I [11:05] think like 2004ish, [11:09] maybe a little sooner. Uh I was like I [11:13] was building websites with just like [11:15] HTML and CSS for like lots of local [11:17] businesses. I was in high school and I [11:21] wanted I wanted a way to like templatize [11:24] some of the work instead of having to [11:26] like you know copy and paste over all [11:28] the sidebar, the header and everything [11:30] over to these like across all the HTML [11:32] templates and everything. Um, I wanted [11:34] something like dynamic and that's when I [11:37] like discovered PHP and was able to [11:40] start templatizing some of these [11:42] websites that I was building so that [11:43] like the header and the sidebar could [11:45] stay static but we could swap out the [11:48] contents on the inside of the page. Uh, [11:50] and like that was so cool to me. And [11:53] then like started getting into like [11:55] database stuff a little bit. I'm like, [11:57] oh my gosh, you can like connect [11:58] databases to this. Well, that's a whole [12:00] another world to like learn on top of [12:02] it. Uh, so that's kind of like my entry [12:05] point into PHP was just like needing [12:08] some dynamicism to like that's not even [12:10] a word, but uh I needed just to kind of [12:13] like be a little bit more efficient [12:14] about what I was doing. Um, and [12:18] needed at some point like access to like [12:20] data layers. How am I going to do that? [12:22] Well, we can do that through PHP. So [12:24] dove further into that but all just got [12:26] started building like websites for local [12:28] businesses. [12:30] >> Yeah. So that would have been at that [12:32] point you would have been using uh PHP [12:34] includes. [12:36] >> Mhm. [12:36] >> Oh yeah. [12:38] >> I remember building websites like that. [12:41] My entire website was just one big [12:43] include file. [12:44] >> It was [12:46] >> I mean to be fair that is what most of [12:48] my code looks like now. [12:53] So, interestingly, you seem to have [12:56] journeyied on just a little bit from [12:58] there, seeing as you're now writing one [13:00] of the greatest AI integrations in [13:03] Laravel. [13:04] >> Oh, thank you. [13:06] Yeah. Yeah. We've come a long way from [13:08] there for sure. Like, you know, I I [13:12] early career like I got started in [13:14] WordPress doing like lots of just [13:16] pumping out like marketing websites. [13:18] Here's PSDs, turn them into websites. [13:21] um did that for quite a while and that's [13:23] where I dove like much much deeper into [13:26] PHP and uh I was actually at one point [13:31] um working on a uh the agency I was [13:35] working for uh had a a cake PHP product [13:38] that they they were working on and I [13:41] hated it. Uh just like the way it was [13:43] written was like kind of crazy. I I [13:46] don't know. I just like I hated it. Uh, [13:47] so I was about ready to jump ship for [13:50] Ruby on Rails, but I decided to go look [13:52] at like the other PHP frameworks out [13:54] there. So I went and looked at uh like [13:56] Falcon looked really interesting. I was [13:58] thinking about going down that route and [14:00] then I discovered Laravel and I was [14:03] blown away. I'm like you could write PHP [14:05] this way. Like that changed everything [14:08] for me. And from that point on, I just [14:10] made a commitment to be inside of PHP [14:13] and work with work with it like [14:16] explicitly ever since. [14:19] >> Yeah. Cake PHP is an interesting one, [14:22] isn't it? It's you either love it or [14:25] hate it and nowhere in between. [14:29] >> Yeah, I haven't looked at it since. Um, [14:31] but it was it the I I think the one [14:34] thing that drove me crazy about it, uh, [14:36] and this was all like pre-omposer and [14:38] everything too. Um, but what drove me [14:42] crazy about it is like the the developer [14:44] that worked on it at the time kind of [14:46] made it his life's mission to make the [14:48] the templating engine was like an [14:50] object-oriented templating engine that [14:51] you could opt into and he tried to to [14:54] make every template like a single echo [14:56] call. So it was these deeply nested [14:58] chained like this div this like the span [15:04] like just nested all the way down into a [15:07] single echo statement for like building [15:09] out an entire templated page. And so [15:11] that just drove me insane. [15:14] >> Yeah, that is not how you use [15:17] templating. Yeah, it was it was fun to [15:20] to play some code golf with it, but it [15:22] was very frustrating when it was like, [15:24] yeah, you got to add a whole new like [15:26] div in here to do some styling. Oh no, [15:29] like where the hell is this closing [15:30] bracket going to go? 25 files is further [15:34] on, obviously. [15:34] >> Yeah, basically. [15:37] >> Yes, I have worked on sites like that. [15:41] They're they're interesting beasts to [15:42] work on when you when you're looking at [15:44] code and you're going, "Right, I know [15:47] that the other half of this is [15:48] somewhere, but it isn't in this file or [15:51] that one or that one." And by the time [15:53] you've got halfway down the list, it's [15:55] like, "Where did I even start?" [15:59] >> Yep. Yep. I don't know what I would do [16:01] without my uh uh my find files in Neoam. [16:04] Like, I just I don't I don't look [16:06] through directories ever anymore. It's [16:08] just Yeah. Start finding me a file. [16:10] fuzzy finding it or I like grip through [16:11] it and find it. Uh but yeah, back back [16:15] then I didn't have any of those tools. [16:17] So it was definitely like clicking [16:18] through and deep deep searching nested [16:21] files [16:23] >> and of course not knowing how to search [16:24] inside the file. You having to read the [16:26] whole thing? [16:27] >> Yep. Oh yeah. [16:29] >> So So when did you learn how to quit [16:31] Vim? [16:32] Uh very early on I uh my early even [16:37] before I really started building [16:38] websites uh my grandmother had given me [16:41] an old laptop and I just started like [16:44] playing with DOSs like yeah I'm like let [16:46] me let me get into the command line and [16:48] felt like such a cool little hacker that [16:50] I could like list directories and stuff [16:52] and uh kind of broke into like nano and [16:55] vim from there like as soon as I got [16:57] into Linux and you got got stuck once [17:00] figured out how to get out of it and uh [17:02] but was hooked. I I've been uh a Vim and [17:05] Neo Vim user for [17:09] 15 20 years, maybe 20 years now. So big [17:14] fan. I got hooked. [17:16] >> I think the thing I like about it is the [17:18] fact that it is so customizable. [17:21] >> If if you want to change something about [17:23] VI, go ahead. You can just change it. [17:24] It's fine. [17:27] Yeah, it's funny to go through like my [17:29] my Vim config has evolved a lot over the [17:31] years, but uh there are there are key [17:34] bindings that have been in there for [17:35] yeah 20 years that are just so ingrained [17:38] in muscle memory that it makes it hard [17:40] for me to really use any other editor at [17:42] this point. Yeah, I I occasionally I log [17:46] into PHP Storm and it's like right [17:50] >> how do I do anything and and I keep [17:52] doing like control S to save the file [17:54] and oh yeah this is PHP Storm I don't [17:57] need to do that and [17:59] >> yeah [18:01] oh yeah yeah like losing my leader key I [18:04] get so lost in like any other ID at this [18:06] point like I I've tried I've tried a [18:08] whole bunch recently with all the like [18:10] the wave of AI editors and everything. [18:13] You know, I wanted to try them out, give [18:15] them a shot, but I'm so unproductive in [18:17] them. [18:18] >> Uh, thank God for cloud code and I can [18:21] stick in my terminals. [18:24] >> Well, I um I I use PHP Storm, but for a [18:28] while I did use uh Neovim and um then I [18:35] can't remember why I wanted to try PHP [18:37] Storm. I think it was because a lot of I [18:39] I was very new to Neoim. I'd followed [18:42] Jess Archer's um tutorial on Larcast on [18:46] how to set up near Vim for PHP [18:49] developers. I knew there was just so [18:51] much more that could be done, but I [18:54] didn't have the time to do the research [18:56] and figure out how to get it working in [18:58] near them. So, I just went [19:00] >> go to PHP Storm because there's a lot of [19:01] features built into PHP Storm. But one [19:04] of the first things I did was um change [19:07] it over to use um Vim bindings. So I've [19:11] I've brought a lot of my bindings from [19:12] Neo Vim into PHP Storm and that's always [19:16] still evolving and and getting bigger [19:18] and bigger and bigger as I'm discovering [19:20] more and more things. [19:21] >> Yeah, [19:22] >> there's a lot more I can be doing. I [19:24] know that. Um, but yeah, I like I [19:28] unfortunately it means that my Neovim [19:30] setup has been neglected for so long and [19:32] I went to go and try and do some stuff [19:33] in my dot files and my Neim just doesn't [19:36] work. [19:38] >> Cannot get it working anymore. I was [19:39] like, uh, okay. [19:42] So, um, my Oh, um, just just a quick [19:45] one. I think that Eric's got the uh call [19:49] for papers for JStech wrong because [19:52] John's just jumped in. John John has [19:55] said the JST closes on the 7th. Eric has [19:58] said the first. So what we'll do is [20:01] we'll say the fourth and then everybody [20:02] will get them in early. [20:05] >> Yeah, we could say well. [20:08] >> Yes. Right. [20:09] >> The fourth, but but if Yeah, [20:13] >> John is being very clear there. It [20:15] closes on the 7th of December. [20:17] >> Right. Seventh. [20:18] >> It is the seventh. [20:19] >> Okay. [20:20] >> Unless we move it again. back to back to [20:23] Neim. So yeah, unfortunately I went into [20:25] Nevin the other week to try and uh do [20:28] some dot file changes and just try [20:30] change a few text files which I don't [20:32] actually do very often outside of PHP at [20:34] the re recently. Um and even things like [20:38] markdown I do I've been doing in PHP [20:41] store but yeah um it's all broken so I'm [20:44] going to have to call on someone to help [20:45] me fix it. I did try to I did try to do [20:47] it in uh get it fixed in Claude, I [20:50] think, but it just got a bit messy. It [20:53] didn't quite work. [20:55] >> Claude has no idea how to do Neo Vim [20:57] configs. I've noticed this. [21:00] >> I'm I'm pretty sure none of the edit [21:02] none of the staff there use Neoim. [21:05] >> Nothing about it. [21:06] >> Yeah. And all configed in Lua and yeah, [21:10] it's it did uh I had it update. had [21:13] write a couple plugins for me recently [21:15] to uh to offload a few like things I [21:18] wanted to try out and it it did a pretty [21:20] okay job, but yeah, it's it's not super [21:24] well educated on configuring Neoim with [21:26] with Lua for sure. [21:31] >> So, mommy and daddy are fighting. [21:33] >> Yes. [21:35] John says he moved the conference [21:37] because I'm speaking at a diff moved the [21:39] call for papers because I'm speaking at [21:40] a different conference which is a JS [21:43] conference [21:44] and um [21:47] >> we could just do that. Yeah. And [21:49] >> we'll do that. So we we can follow [21:50] follow the follow the argument. [21:54] >> So the rest of this is just going to be [21:55] watching John and Eric decide who moved [21:57] it. Um [22:00] so MVPOP says that Claude Code is [22:02] actually good at Neoim configs. Right. [22:04] What what's your secret? Because [22:06] whenever I ask it to configure anything, [22:07] it just gets it wrong. [22:09] >> And I'm glad you say that, Chris, [22:10] because I started to think it was me. [22:18] >> Oh, yeah. Yeah. I've had I've had to [22:19] build um two I was experimenting with [22:23] some like Prism and AI stuff and trying [22:26] to integrate it into like some Neoim [22:29] things and I was having a really hard [22:31] time with it. So, I had Claude like [22:33] knock it out. It's nothing interesting. [22:36] I've like abandoned the project at this [22:37] point, but uh it did a pretty good job [22:40] being able to like put most of the [22:42] things together and you can always point [22:44] it over to the docs now and you know [22:46] have it go crawl the docs and it's got [22:47] it's got web fetch so it can go go crawl [22:50] web pages. [22:52] >> So, so was this the um ability to use [22:55] Prism in Neoim that you've abandoned? [22:59] Uh it was a way to [23:02] it was a way that I could like get a [23:04] popover window with prism output uh [23:07] using like a PHP script invoked by [23:10] Neoim. [23:11] >> Nice. [23:12] >> And like passing in like like buffer [23:14] context and all sorts of other [23:16] information. So it was it was a I just [23:18] wanted to see if I could do it and I [23:21] could do it and then I've never used it [23:23] since but it was just kind of a itch to [23:27] scratch. [23:28] Well, I I really want to see what that [23:31] code looks like now. [23:33] >> Yeah, [23:36] I I'm mad like that. I like seeing [23:37] things that are just weird experiments [23:39] that usually end up being something [23:41] cool. You you look at something like um [23:45] not claude code and cursor that started [23:48] as somebody's idea of what happens if we [23:50] put AI in an IDA. [23:52] >> Yeah, sure. M [23:55] >> and if you could put AI inside a good [23:57] IDE like Neoim, [23:59] you you end up with quite a powerful [24:01] little tool. [24:03] >> Yeah, it's uh there's much better like [24:06] AI integrations into Neoim than than [24:09] that. It was uh at the time I was like I [24:12] I have these system prompts that I [24:15] really like to use these like personas [24:17] and uh [24:20] I I leverage those a lot through through [24:22] Prism. I've got like a little open web [24:24] UI app that I have like running [24:26] persistently on my computer and then [24:28] that uses a Laravel like that connects [24:31] to like a Laravel API uh through Prism [24:34] and so I can like chat these custom [24:36] personas without having to you know do [24:38] weird things about trying to like reset [24:40] system prompts or or other stuff. I can [24:42] just kind of like interact with them [24:44] just like I'd interact with like pulling [24:45] up like claw desktop or something. So, [24:48] I uh I like to like pipe a lot of stuff [24:51] like through Prism because I can just, [24:53] you know, make make them do whatever I'd [24:55] like super easily. [24:58] >> I I threw that up because Eric said to [25:02] one of the people watching on YouTube, [25:03] "Come and join us in Discord." And [25:05] they're now in Discord, which is cool. [25:08] >> Yeah. to it. Um, on that note, I think [25:11] it's time to um um hang on, find the [25:15] right video. Uh, go to our uh our [25:18] partners for today. [25:23] Thank you to our partners at [25:24] phpscore.com. [25:26] Every app builds up technical debt over [25:28] time. It's the price we pay for shipping [25:30] new features and moving fast. When we [25:33] build up too much of it, though, it can [25:34] start to impact how we work. Team [25:37] velocity suffers. bugs become more [25:39] frequent and take longer to fix and [25:41] everyone starts to get a little [25:42] frustrated. The key to managing debt is [25:45] to measure it. A credit score can help [25:48] you understand how well you're managing [25:49] your financial debt. And now there's a [25:52] credit score for your technical debt. Go [25:54] to phpsscore.com to get a free technical [25:57] debt score and monitoring for all of [25:59] your PHP applications today. [26:04] >> Thank you. PHP school. PHP score. [26:09] >> I'll do I'll do the echo today. [26:13] I was reading some of the chat. I was [26:14] catching up with chat, but you know what [26:16] it's what it's like for me. I'm always [26:18] three or four messages behind. [26:21] >> So am I. And they're usually my [26:22] messages. [26:26] >> Yes. Well, so um I think it's a good [26:29] transition to start talking about this [26:31] package that you've been [26:33] >> working on recently that's got everyone [26:37] excited. [26:39] >> How did that come about? Where did you [26:41] where did the idea come from to begin [26:43] with? [26:43] >> So it started uh when I did the a little [26:47] while before the Laravel worldwide [26:49] meetup I did. Uh I forgot what year that [26:52] was. Um I think it was a couple years [26:53] ago at this point and like sort of [26:58] getting ready for that talk but um kind [27:01] of my journey getting into AI led me to [27:05] a point where um this was like very [27:08] early in AI. This was uh when I got [27:10] started with it. It was uh maybe like 6 [27:12] months before chat GBT came out [27:15] and I just kind of got bit by the AI bug [27:20] and I could I saw like a ton of [27:22] potential with it. Um I've got like [27:24] pretty severe ADHD and I saw so much [27:26] potential for me to be able to like [27:28] build things to help augment my reality [27:31] and like things to be able to like help [27:33] me through like day-to-day work life, [27:36] whatever. um through through AI and like [27:40] the potential I could see it having as [27:42] it got more powerful and uh I didn't [27:45] know Python at the time. All the A IML [27:47] stuff at the that point was like pretty [27:51] solidly in Python. There was some [27:52] TypeScript stuff starting to come out. [27:54] Um Langchain was like getting getting [27:58] popular. There was a a TypeScript [28:00] version of that that came out. Um, but I [28:04] missed spending all of my time in [28:06] Laraveo. Like I've spent so much time [28:09] like so much of my development career [28:11] has been spent with it. I love the [28:13] framework. I live and bleed it. Um, [28:16] but I I wanted to be back in Laravel and [28:20] like working on these experiments and [28:21] continuing to learn more about AI and [28:23] like building like these tool sets. Uh, [28:26] I just really I I wanted to be back [28:29] inside of like PHP and Laravel and like [28:31] there wasn't really any tools at the [28:33] time. Um, so I decided to like start [28:37] working on a little framework. At the [28:39] time I called it Sparkle and [28:42] it was kind like this like similar to [28:45] Prism but definitely different. It was [28:46] like an agent focused library. um it [28:50] didn't it had nowhere near as like well [28:54] ststructured uh like multi-provider [28:56] support but it did support multiple [28:58] providers um so it was kind of this like [29:00] early iteration of Prism but uh I had [29:04] been showing progress on it to Taylor [29:07] and Taylor had just like we were talking [29:09] about the AI SDK from Verscell and he [29:12] said uh he just like in passing was like [29:14] yeah I wonder what a Laravel version of [29:16] that would look like and I sat there for [29:19] a second and the API just kind of like [29:22] hit me and I sat down and like started [29:24] sketching out the API and like how you [29:26] would interact with it and how like [29:28] multi-provider support would kind of [29:30] work and uh then just started like [29:33] fulfilling that contract. So it was like [29:35] very very much API first and then [29:37] backfilling the functionality to meet [29:39] that API. [29:42] >> Nice. [29:43] >> So the right way around you thought [29:45] about the the developer experience [29:47] first. Oh, for sure. And then just it's [29:49] always it's always kind of started [29:51] there, you know, and as I'm facing like [29:53] developing a few new features for it [29:55] now, it's it's right back to that API [29:58] development. It's like how do I want to [29:59] make these interactions work? How is it [30:01] going to feel? Um, and I think that's [30:03] something that I've just gleaned from [30:05] working with Laravel for so long, like [30:07] that that really clean, uh, fluent, easy [30:11] to work with, like where you start [30:13] assuming you work with it long enough, [30:14] you start like assuming methods are [30:16] there and then magically they're just [30:18] there for you. There's like it's a very [30:20] intuitive thing to work with. And so I [30:22] took a lot of that inspiration and [30:23] applied that to the API of Prism. [30:28] >> And that that is one of the things that [30:30] Laravel always does. [30:34] The the developer experience is the [30:36] first thing of Laravel. So much so that [30:39] we actually for a while had people who [30:42] would call themselves Laravel developers [30:44] and not PHP developers. [30:46] >> Oh, sure. [30:47] >> I mean, you see that in like the [30:48] JavaScript community, too. [30:49] >> Oh, yeah. Yeah. [30:50] >> Yeah. like you get you get really like [30:52] locked in like framework lock in um you [30:55] know where that's you you learned you [30:58] learned PHP by working with Laravel and [31:00] like I know a ton of people that that [31:02] that's how that's how they learned PHP [31:03] was by starting like and picking up [31:05] working with Laravel and then they [31:07] learned PHP from there. [31:09] >> Yeah. And and Laravel makes it easy to [31:12] make that bridge because helps you with [31:15] stuff until you need to do more than the [31:19] framework. [31:21] And that's when you start diving into [31:24] into the more complex stuff of okay now [31:28] we're into business logic and how we all [31:30] that then we're into how do we do [31:33] multi-threaded in PHP and [31:36] >> the answer is I don't think you can yet [31:42] >> we've got Nico on Twitter saying thanks [31:44] for prison. [31:45] >> Oh very welcome. It's it's such a [31:47] passion project for me. Like I I love [31:50] working on it. It's it's something I [31:52] like I think about and and am constantly [31:54] tinkering with. So um very welcome. Do [31:59] >> do you know the point I realized that [32:01] Prism wasn't your full-time job? [32:04] >> Which was that? [32:06] >> When I saw an advert for a job at [32:08] Geocodio that you were recruiting for. [32:14] >> Yeah. [32:18] So, yes, that that's when I found out [32:19] that Prism wasn't actually making you [32:22] any money. [32:23] >> No, I have uh I do have like a handful [32:26] of of sponsors that I I deeply [32:29] appreciate and I'm I'm very thankful [32:30] for. So, uh but I I do I chew through a [32:34] ton of tokens every month like running [32:36] tests and experiments and, you know, [32:38] trying to like push Prism further and [32:40] like, you know, even just driving out a [32:43] new provider. um or like testing like [32:47] with Gemini 3.0 Pro that came out, there [32:50] were a bunch of changes in the API that [32:52] kind of needed to be made and we're [32:54] still working through those with other [32:55] contributors because that's it's just [32:57] not a model and provider that I I use [32:59] very often. Um so I don't have a lot of [33:02] reasons to be testing it. So, uh, [33:04] there's been a lot of really solid [33:06] contributors coming back in, uh, and and [33:09] bringing more support for Gemini, but, [33:11] uh, I have to like manually test all of [33:14] that stuff, too. So, it's I I've chew [33:17] through a ton of tokens with Google over [33:19] the last like month or so, you know, [33:21] driving out functionality, then they [33:23] drop a new model, and then I got to go [33:25] test all of that stuff. So, the [33:27] sponsorships go a long way, but it's far [33:29] from being a full-time gig. [33:31] >> Yeah. And any sponsorship is going to [33:34] help. And [33:35] >> oh, it makes such a huge difference for [33:37] sure. Like if I could get to a point [33:39] like I think [33:41] >> for me the dream would be to be able to [33:45] make enough in sponsorships that I could [33:47] like maybe drop a day a week at [33:49] Geocodio. Like be able to just like just [33:52] dedicate Fridays to Prism all day every [33:56] week, you know? Like that would be [33:58] awesome if I could get to that point. Um [34:00] but we'll see. or or talk Taylor into [34:03] taking it on as first party. Oh, [34:04] >> sure. [34:08] >> Yeah. I mean, I you say about you you [34:10] churn through lots of credits. I I I can [34:14] imagine because I I with um chat GPT [34:18] I've only got the free account and I'm [34:20] always running out. I'm always asking [34:22] random questions and it's just things I [34:24] come up with. I just like ask it a [34:26] question then leave it running for a [34:28] little while and I'll go and do [34:28] something else. come back to it churned [34:30] out and I'm always running out of credit [34:32] so I can imagine you going through quite [34:34] a lot. [34:36] >> Yeah. Yeah. I definitely chew through [34:38] them and you know it's it's [34:41] like spinning up accounts at all the [34:43] different providers and like you know [34:46] having to juggle all the API keys for [34:47] everyone and [34:49] >> you know it's remembering which one [34:51] supports what thing and it's it's a it's [34:55] a wild time out there. I mean, look at [34:56] the last week alone, we've had, you [34:59] know, three new premier models drop. [35:02] >> And with that, new new features and [35:05] functionality, too. So, it's it's been [35:08] uh a busy busy couple weeks with with [35:11] Prism and and we're like I said, we're [35:13] still churning through some of the API [35:15] changes for the new stuff. [35:19] >> Yeah, I can I can I can imagine. I [35:21] hadn't thought I hadn't appreciated or [35:23] thought really about the the the cost of [35:26] all the testing you must go through. And [35:28] I'm now thinking about the the little [35:30] idea that I we were talking about just [35:31] before the show that I'm hoping at some [35:33] point uh probably in the new year to get [35:35] time to play around with. I'm thinking, [35:37] yeah, I'm going to go through quite a [35:38] lot of credits just testing that to get [35:40] make sure it works. So, yeah. Yeah, [35:43] that's that's that's very interesting. [35:45] >> Yeah. One of the cool things I found out [35:47] recently, I started playing well, I [35:49] started playing a little bit more with [35:50] um Olama and local models for test [35:54] experiments [35:55] >> because uh a lot of times what I'll end [35:59] up doing is like I'll I'll build a [36:01] streaming chat app just to like get the [36:04] feel for how Prism behaves in there. Uh [36:06] like one of the things I'm working on [36:08] right now is uh like message [36:10] persistence. like how can we make [36:12] persisting these conversations through [36:14] building chat apps like a really clean [36:17] experience. [36:18] >> So like I I would be like chewing [36:23] through tokens building that thing out [36:25] and like testing different [36:26] configurations and different things, but [36:29] um you know I I made the investment in a [36:31] pretty like maxed out MacBook uh [36:35] somewhat recently. So, I have the [36:36] affordance to be able to run some of the [36:38] bigger the the bigger local models and [36:41] I've been pretty blown away like the the [36:44] OpenAI's GPTO OSS model is very [36:48] impressive. So for building out [36:51] integrations and like experimenting with [36:53] stuff, I think you can get pretty decent [36:55] mileage and like a pretty good idea of [36:57] how a large language model is going to [36:59] handle the task at hand through some [37:03] local models and then you can kind of [37:05] upgrade to like test on some of the more [37:07] powerful stuff or if you're running into [37:09] it just like not being effective, you [37:11] know, scale up a little bit, you know, [37:13] try try the next level up model. Um [37:16] because I think from the efficiency side [37:18] of things, you really want to nail on [37:21] using the smallest model possible to [37:23] complete the task. So you can kind of [37:25] scale up that way, kind of starting with [37:27] local and then kind of upgrading through [37:29] the smaller models. And the smaller the [37:31] model, the less expensive they are. So [37:33] you can kind of get away with, you know, [37:35] sort of testing a little bit less [37:37] expensive. And then you can also kind [37:39] through that process of like leveling up [37:42] a little bit at a time as you're trying [37:44] to figure out what it's capable of, [37:46] you'll also probably land on the [37:47] smallest capable model for the job, too. [37:51] >> Yeah. [37:51] >> Yeah. That's really interesting, really [37:54] helpful advice. [37:55] >> And a lot of people don't realize you [37:57] can even do that in some of the IDEs [37:59] like claw code. [38:00] >> You can actually tell it which to use. [38:05] the [38:06] it was either yesterday uh tight article [38:11] on their blog about how to use like O [38:13] Lama and self-hosted. So that's that's [38:16] like a great resource for kind of like [38:18] getting into that as well. [38:21] >> Oh, definitely. So we've had um a lot of [38:24] conversation in a very short space of [38:25] time, which is good. Um first up is [38:28] Jimmy um who has a question. I'm a PHP [38:32] developer for 15 years. Sure, I know [38:35] other languages. I began with assembly [38:36] and C. Now I'm tired with the bullies of [38:39] PHP. I see JS like React and other stuff [38:42] very complicated. I've got the [38:44] impression some juniors don't see PHP [38:47] because PHP is old and not the actual [38:49] hype. [38:51] Oh man, there is [38:54] this is so loaded. Um, yeah. [38:58] >> Yeah. This is this is a gold mine of of [39:00] of [39:01] stuff for like hot takes around PHP. [39:04] Like I I definitely think PHP suffers [39:06] from an image problem. Uh I've thought [39:09] that for a long time. Like that's one of [39:11] the things that I appreciate about [39:14] Laravel and the Laravel community is [39:16] like I think everyone really pushes hard [39:18] to like elevate the language alongside [39:20] like the the frameworks and like you [39:22] don't you don't get to be using PHP [39:25] without you know using a framework at [39:26] this point for a lot of things. So um [39:32] I think a lot of PHP suffers from an [39:34] image problem and I think we suffer [39:37] because of that we end up sort of [39:39] suffering at the top of the funnel right [39:41] when we get new junior devs and people [39:43] entering the industry like we suffer at [39:46] that top of the funnel and that's where [39:48] like JS is kind of snagging everyone at [39:50] this point um is like that's where all [39:52] the boot camps are like using like [39:54] Python or JavaScript like that I think [39:57] that's like 90% of the boot camps that I [40:01] see. Uh I see I mean I'm in based in [40:03] Metro Detroit with a lot of like auto [40:05] industry. So we see a lot of like Java [40:06] boot camps around here too. [40:08] >> Um but the vast majority of it is like [40:11] JavaScript and Python. [40:13] >> So you know we we as a [40:17] brand PHP right we need to figure out [40:19] how we can like capture at that top of [40:21] the funnel. And I think part of that is [40:23] like sort of with like the image issue. [40:25] Um, so I was like really happy to see [40:28] with uh the 8.5 release like they had [40:32] like a really nice branded web page for [40:34] that. I think that was like super clean. [40:36] I love how they approached that. Uh I [40:39] think more more of things like that. Um [40:43] are going to be helpful. So [40:46] >> and I think some of the image we don't [40:49] help ourselves with this and I keep [40:52] seeing I've seen another one the other [40:54] day. um where someone's posting [40:59] maybe in LinkedIn for example this one [41:00] was in LinkedIn and the the like the [41:03] headline says something about PHP [41:06] was actually the question is PHP dead [41:09] and there's a massive sort of the post [41:11] then go if you extend the post there's a [41:12] big explanation why it's not dead but [41:15] the problem is no people are just seeing [41:17] the is PHP dead and assuming PHP is dead [41:21] that's it they're not going to open it [41:23] up and read the rest of it. So those [41:25] things we we've got to try and turn [41:27] those around. Um and that's kind of [41:30] where this this show was born from was a [41:33] a me I was at PHP UK earlier this year [41:37] and there was there was a conversation [41:38] about that about using more positive [41:42] languages language to talk about PHP and [41:45] not keep saying no PHP isn't dead. We we [41:47] should just forget about that and not [41:49] say that anymore. certainly don't title [41:51] websites or posts with those words. Even [41:54] if it's PHP isn't dead, people just read [41:57] PHP dead. They don't get the rest of the [42:00] the rest of the the sentence. So, I [42:02] think as a a community, we we need to [42:05] change the way that we're trying to [42:06] present. [42:09] >> These posts aren't trying to damage PHP [42:11] and I understand that they are trying to [42:13] encourage people, but it's not working [42:15] because people are seeing the negative [42:17] before they see the positive. I think [42:19] it's a little bit of like chicken and [42:21] egg though too. Um I think those types [42:24] of posts exist because there is this [42:26] like connotation that it's dead [42:29] >> and so I I think it it you have to be [42:31] careful that it doesn't get too [42:32] cyclical. But [42:34] >> um [42:36] >> yeah, I don't know. It's it's been [42:38] something that has been on my mind for [42:40] for years as like someone who's like [42:42] passionate about education and like [42:44] sharing this stuff. Um it's like how [42:47] yeah how can we get get more people in [42:50] the door because [42:54] there [42:55] it's like just so advantageous like I I [42:59] look at like setting up Ruby on like [43:02] even just from like a CIS admin level [43:03] like how much it takes to set up Ruby on [43:06] a server, right? Verse how much it sets [43:08] like takes to set up like PHP on a [43:10] server. Maybe I've just been doing it [43:11] for so long, but like it's [43:16] it's like an easy process to walk [43:19] through. It's like it's well documented. [43:21] It's it's because PHP has been around [43:24] for so long. [43:25] >> Um [43:27] granted with that too, there's a lot of [43:29] bad PHP around as well. [43:31] >> Um but I don't know. I don't know. It's [43:35] been something on my mind for a while [43:36] and I love seeing that there is such an [43:39] initiative now kind of across the board, [43:42] you know, internally in PHP, you know, [43:45] podcasts like this, you know, the [43:48] initiative for the 8.5 website, like I [43:50] think I think there's a really healthy [43:53] conversation going on about it now. And [43:56] I think one of the things that's that's [43:58] helping [43:59] >> and you'll know the conversation we had [44:01] in Chicago, [44:02] >> but one of the things that I think is [44:04] helping is actually the marketing [44:06] machine that goes around a lot of the [44:08] frameworks. So you've got Laravel [44:10] drawing people in because they're good [44:13] at marketing for themselves, which in [44:15] turn promotes PHP [44:18] and then you got native PHP who just [44:20] goes, "By the way, we can do mobile apps [44:22] now." [44:22] >> Yeah. And it's like, okay, cool. And [44:26] >> yeah, and and with Prism, like I I don't [44:28] want to like toot my own horn too much, [44:30] but with like with Prism, like now [44:33] you're not bound to like look at my use [44:36] case. Now you're not bound to Python, [44:38] >> you know? You want to work with AI, [44:41] >> not a problem working in PHP. And I know [44:43] there's like a fantastic symphony [44:45] package for working with AI as well. [44:47] Mhm. [44:47] >> So it's it's like more accessible than [44:49] ever inside the language to be able to [44:51] like be be accessing like new hot [44:53] technologies, you know, [44:55] >> and you've got the the official M MCP [44:58] SDK. [44:59] >> Yeah. Yeah. The investment from from the [45:02] actual PHP Foundation to, you know, be [45:06] involved inside of of of AI. I think [45:09] that's [45:10] >> so fantastic. And then and then you get [45:13] some weirdo who's started a series of [45:15] articles in a magazine that's published [45:17] every month by PHP architect on the [45:20] basics of machine learning in PHP. So [45:22] teaching linear regression and all sorts [45:24] in PHP. I have no idea who's writing [45:26] that series. Me. [45:32] >> Yeah. [45:33] And these sort of things are all [45:36] possible in PHP. And it's almost like we [45:40] don't shout about it enough. We don't it [45:44] at the lowest level we don't say [45:45] actually all of the cool things you can [45:47] do in JavaScript. Yeah, we can do that. [45:51] >> Yeah. [45:52] >> And if there's not like direct solutions [45:54] with it, there's a lot of affordances [45:56] inside of frameworks to be able to do [45:57] those things. [45:59] So, I mean, you look at like Livewire, [46:02] like being able to being able to do that [46:05] inside of Laravel gets you so close to [46:07] the JavaScript experience for a lot of [46:09] things. [46:10] >> Um, [46:12] you know, I [46:14] >> think it's a more exciting time than [46:16] ever to be getting into PHP with like [46:17] all all the things that are are at our [46:20] fingertips now. [46:24] >> Yeah. [46:24] >> Yeah. Yeah, I mean um just to just to [46:27] point on what Nico just said about the [46:29] PHP is dead title being a clickbait and [46:32] when you then read the article it says [46:35] it isn't. I I get that the the [46:39] concern that I think a lot of people [46:41] have that I talk to is that people read [46:43] that and they don't click on it. They [46:45] don't read the article to find out. [46:47] >> You're just you're doom scrolling, [46:48] right? You scroll past the title, you [46:50] read the title, but you're not reading [46:51] the article. And so you see the post on [46:53] Reddit. You scroll past the title and [46:55] it's like, "Yeah, is PHP dead?" You're [46:57] like, "Well, I don't know. Is it?" But [46:59] you don't, you're not clicking through [47:00] to read the article because we don't [47:02] have the attention spans anymore. But [47:03] you're scrolling past it. You're seeing [47:05] those titles like, "Is it dead? Is it [47:07] dead? Is it dead?" Well, now you're [47:08] starting to question it. So, I see where [47:10] you're coming from, too. [47:11] >> Yeah. Yeah. Um, there was another [47:14] question, but I now can't find it. [47:18] >> There was It was um [47:22] Yes, that's one [47:26] of [47:27] >> is joining us because he's finished [47:28] dinner and he's done the dishes and he [47:31] wonders if some kind of PR um which Pton [47:35] the Python light LLM model details would [47:37] be of value [47:39] maybe. Uh I'd be open to like checking [47:41] out the PR like we don't have uh [47:44] explicit light LLM support inside of [47:47] Prism. um I think if you're going [47:50] through it as like an open AI [47:52] compatibility layer um [47:55] you might have access to I I don't know [47:57] check it out. I open the PR like I'll [47:59] I'll be happy to take a look at it or [48:02] shoot me a DM on any platform. You go to [48:05] my website. I've got a links page. You [48:07] can find me everywhere. Shoot me a DM [48:09] and we can talk a little bit more about [48:10] it. Um but I don't I don't see why I'd [48:14] be opposed to it. [48:15] >> Yeah. And uh he went on to say that he [48:18] uses light LLM a lot for prototyping and [48:20] it includes all the stats and the [48:22] costings. I think that's more what he's [48:24] getting out of the fact that you could [48:25] pull out the stats and costings. [48:27] >> Yeah. Yeah. You can um it we we do [48:31] expose like all the like tokens and [48:34] stuff that you're using through your [48:36] requests with Prism, but you're you're [48:38] it's still on you to be like totaling [48:40] and exposing them. uh light LLM [48:42] operating as a proxy to providers gives [48:45] it that opportunity to like kind of [48:48] build those stats up for you and store [48:50] them. Um it's I I've not used light LLM [48:54] in a long time, but um it looks pretty [48:57] nice. It's come a long way. [49:01] >> Nico wants to know if you've ever [49:02] accidentally committed an API key. [49:04] >> Yep, 100%. [49:08] >> I think that's a good everyone's [49:10] response. Yeah, [49:11] >> it's 100%. [49:14] >> I I definitely haven't. [49:17] I definitely haven't. And I'm now going [49:19] to check every reaper. [49:24] >> And uh Marian. [49:26] >> Yeah. What is it? What is it? The the [49:28] BFG. Uh is that is that what you use to [49:31] to scrub repos of things? [49:33] >> I'm pretty sure that's the CLI. Yeah, [49:35] I'm I'm I'm familiar. Definitely [49:38] committed keys. So Marian says, "Let [49:41] them think PHP is dead." Um I I get your [49:45] point, but actually I'd rather they [49:48] didn't think PHP was dead and came and [49:50] programmed with us. [49:53] >> Yeah. No, I mean uh the one of the scary [49:56] things I I experienced was that being at [49:58] the PHP UK conference when they were [50:01] doing a um [50:04] a panel and that was one of the [50:06] questions was talking about bringing in [50:08] junior devs and before they answered the [50:10] question um they did ask anyone over the [50:13] age of I think it was 35 if I remember [50:16] rightly to put their hands up that are [50:18] full-time dead and pretty much the whole [50:20] room put their hands up. [50:22] That's a scary position to be in when [50:24] you've not got anyone in their 20ies and [50:27] coming into into the language. What's [50:30] gonna happen in 30 years time [50:34] >> when all those 35 year olds are now 65 [50:37] and trying to retire and no one else is [50:39] coming back in? So I would say we we [50:42] don't want to let people think PHP is [50:44] dead. [50:46] We want to bring in bring bring in the [50:48] new blood. We want to bring in new [50:49] people and that's the way how how [50:52] society and and communities thrive is [50:54] bringing in new blood. [50:56] >> Yeah. [50:57] >> No. [50:58] >> One of two things will happen if we [51:00] don't bring in 20 year olds. One, the [51:03] language will eventually go away. Two, [51:07] those of us that are in our 60s are [51:09] going to be paid a hell of a lot more [51:11] money. [51:14] >> There is that. There is that. Okay. [51:16] PHP's dead. Yeah, [51:22] >> I know. This is exactly why I've been [51:24] all in. I've been in the Strat for [51:26] years, right? I'm just waiting for it to [51:27] die so I can capitalize on all the [51:29] legacy code. [51:30] >> That's exactly how Cobalt did it. [51:35] >> I I have seen Cobalt jobs now advertised [51:38] at £150,000, [51:39] about $200,000 [51:42] for editing cobalt. [51:44] >> Oh, sure. All right, I'm going to start [51:45] googling some tutorials. [51:47] >> Yeah, [51:49] >> I I will give you a clue. Anything [51:51] that's on YouTube about Coal is going to [51:53] be at least 15 years old. [51:57] >> Yes, that's fine. [51:59] >> Yeah, I haven't changed since like 1985. [52:02] So, [52:03] >> right. [52:07] >> Nico makes a fair point. Cobalt guys [52:09] make bank with banks. Yeah, [52:10] >> it's true. [52:11] >> Yeah. [52:12] >> Yeah. I mean, [52:13] >> the the real money's in Forran. That's [52:15] that's where you make the real money. [52:17] >> Jeez. [52:19] >> I I I was once responsible for writing [52:22] Fortran for a very short space of time, [52:27] mainly because it was bloody awful. [52:33] It It's It's like Python in that the [52:35] intend indentation of your code matters. [52:38] >> Oh, yeah. It's It's weird because it's [52:41] like got to be in the right columns and [52:43] it's [52:44] >> Oh, sure. [52:45] >> No, I My brain doesn't work that way. [52:49] Don't Don't give me columns. [52:51] >> Yeah. This is This is also why I don't [52:52] like YAML, you know. [52:54] >> Yeah. [52:57] >> I I do love YAML though. I mean, I I [53:00] hate working with it, but I love it. [53:04] >> I I don't know who wants to tell Marian. [53:07] >> He can't do maths. [53:10] He says he's from 89, so he's 26 years [53:13] old, in which case I'm only 30. [53:19] Yes. [53:21] I wish [53:24] I wish maths work that way. [53:27] >> I mean, we we could just make up maths. [53:31] >> I think I was doing that earlier. [53:37] >> Oh, was it yesterday? I can't remember. [53:39] I'm I'm I'm lost. The last last couple [53:41] of days have merged a bit. Um [53:46] >> maybe I am drunk. Maybe that's what's [53:49] going on. I sniffed the whiskey when I [53:53] went downstairs. That was enough. [53:54] >> There you go. Gotcha. [53:56] >> That was it. It's got me. Yeah. [53:59] >> So, [54:00] >> so I guess the next important question [54:02] then TJ is, are you coming to tech this [54:05] year? [54:05] >> Oh, yeah. I'll be there. [54:07] >> Good. Cool. [54:08] >> Yeah. Yeah. I missed uh I unfortunately [54:11] uh I missed the CFP so I don't I don't [54:15] think I'll be talking this year. Uh but [54:17] I will be there for sure. Have another [54:20] CFP open. [54:21] >> Yeah. What you should do is do a talk on [54:24] PHP for JavaScript developers and put it [54:26] into JStech. [54:28] >> I thought about it. I'm not going to [54:30] lie. [54:32] Not going to lie. Um, I I mean I almost [54:35] thought about doing a like a Versel AI [54:38] SDK talk as well cuz it is such a [54:41] similar realm that like I should be able [54:43] to like pick up and run with it. But uh [54:47] >> I don't know. I'm my my JavaScript is so [54:49] rusty at this point. I think I'd be [54:50] doing myself a disservice to like try to [54:53] do a JavaScript talk and re like relearn [54:56] the language. [54:58] If only you had some sort of system that [55:00] let you integrate with multiple AIs that [55:02] would help you. [55:03] >> I know, right? [55:06] >> That'd be great. [55:08] >> Just Yeah, just get AI to write it for [55:11] you. No one will know. Honest. [55:14] >> Yeah. No, it should be gone. [55:15] >> You're absolutely right. [55:17] >> Yeah. [55:18] I just got to memorize it, right? That's [55:20] good to go. [55:21] >> No, no, no, no. Then you [55:23] >> What you do then is you go to 11 Labs. [55:25] You teach it your voice and [55:29] Let me just get on stage and hit play [55:31] like DJ style, right? [55:32] >> And then just go and sit in the front [55:34] row and then ask yourself some [55:35] questions. [55:39] >> Oh, I should have pitched that one. [55:41] >> Oh, that'd be great. [55:43] >> I may do it. [55:44] >> I may pitch the AI talk. [55:47] >> Yeah, [55:49] >> it's totally doable. [55:52] I think the funny the funniest [55:54] conference talk I ever gave was one on [55:56] how to do conference talks. I said [55:59] precisely nothing in the talk. It was [56:02] just on like raising your voice and [56:04] lowering your voice and all I was doing [56:07] was demonstrating it with and now I'm [56:08] going to raise my voice and then I'm [56:10] going to lower it so that people think [56:11] I'm talking about something important [56:13] and [56:16] >> nice. I love that. [56:20] So, um, we're we're getting up to to our [56:24] time now, unfortunately. It's been a [56:26] been a a laugh and, uh, but we do have [56:29] one more question for you. [56:31] >> Yeah, that's um, what I mean, you've [56:35] probably touched on quite a bit of it [56:36] already, but what excites you about PHP [56:39] now today or what's coming up? [56:43] >> Oh, man. Uh, there's a whole bunch of [56:47] interesting stuff, right? Like I want to [56:49] I want to play with some pipe operators [56:51] a little bit and like brand new [56:53] >> feature out. I want to mess with those [56:54] some more. Um that's that's an exciting [56:57] exciting feature ad for me. [57:00] >> Uh I don't know. [57:02] I thought about this question a little [57:04] bit um early on like while we were [57:07] chatting before the show and honestly [57:10] like the thing that gets me the most [57:12] excited about PHP and like it's the same [57:14] thing that's made me the most excited [57:16] about PHP [57:18] >> for years is just the community. It's [57:21] it's the people that are involved in it. [57:23] It's the, [57:26] you know, the the things that we're [57:28] doing. Like every time I go to like PHP [57:31] Tech or Laracon or, you know, something, [57:36] I come back so inspired to like do more [57:40] and better things like especially now [57:42] that I'm like working with Prism, you [57:44] know, I came back from like all these [57:46] conferences so inspired to like make it [57:48] better and like contribute more. Um, I [57:53] don't know. It's like the [57:55] it's always been the community for me. [57:57] Like that's that's something that bit me [57:59] early on in my like serious PHP career [58:02] when I started, you know, got out of the [58:03] marketing stuff and got into like [58:05] building real world apps. Um, I went to [58:08] Laracon. Let me check the badge. [58:12] 2016 was my first Laracon. So that was [58:14] like my first conference I had ever been [58:16] to. Um, anything. And I walked away from [58:20] that conference just so committed to [58:23] PHP, to Laravel, to the community. Um, I [58:27] was just blown away by that and and I've [58:30] been to a bunch of conferences since [58:32] then. Uh, spoke at a handful and I just [58:36] always come back to for me the most like [58:38] one of the most exciting things is the [58:39] community. So, like if you're watching [58:42] this, like if you can get if you haven't [58:44] been to a conference, like [58:46] >> try and get to one. Like PHP Tech is [58:49] like one of my favorites because uh it's [58:54] it's a wellorganized conference. It's I [58:57] love the venue's like I think great for [58:59] a conference. I've always enjoyed [59:01] speaking at it, but it's such a great [59:04] blend of like, you know, you got [59:07] Laracon, which is like very Laravel [59:09] heavy. I've been to a bunch of Laracons. [59:11] I love it. Um, but this like PHP tech, I [59:14] think, does such a good job of like [59:15] giving you the exposure to like PHP as a [59:18] broader language and like the broader [59:20] PHP community. It's [59:23] >> such a great entry point into that. So, [59:25] like I think that's a great conference [59:26] to go to. Laracon, highly recommend. [59:29] love it. They're all over. The Laravel [59:31] live conferences are fantastic, too. Um, [59:34] so if you can like get out there and get [59:35] involved in the community, like highly [59:38] recommend it. It's I think one of like [59:40] the very special pieces about the [59:42] language. [59:44] >> Yeah, you I love the fact that you said [59:46] the community. I'm afraid you have to [59:48] forgive me. My dog is now barking at [59:49] some idiot setting off fireworks. [59:52] So, if you can hear hear my dogs [59:54] barking, I do apologize, but I wanted to [59:57] weigh in here and not just keep muted [59:59] because um I've said this on the show [60:01] before on other episodes, but the PHP [60:06] tech community um was so welcoming to me [60:10] when I went to it first time last year. [60:13] Um, [60:15] I was obviously I was fly flown over as [60:17] part of the team. Um, and I was working [60:19] on it, but it was everyone was just just [60:23] sit there chatting away with you and you [60:25] would you would be sitting talking to [60:27] someone then you realize it's it's some [60:30] someone from internals or or you're [60:34] talking to Derek or or you're talking to [60:37] Ben Ramsey and you go, "This been Ben [60:40] Ramsey I'm talking to." And and one of [60:42] the big things I think from that from [60:44] especially from tech um is that there's [60:49] no [60:51] there's no clicks is what we would say [60:54] there there's no sort of barrier between [60:55] the speakers and the the general public [60:58] that going to watch the speakers mingle [61:00] um all and there's so many people there. [61:02] It's it's it is such a great experience. [61:04] Um, I would [61:06] >> highly recommend [61:09] that anyone can go. Even if you have to [61:11] get your your company to sponsor you to [61:13] come over, come over and if you don't [61:16] know anyone, just talk to one of us. [61:18] >> You've seen our faces enough. You know [61:20] who we are. We're we're all going to be [61:22] there working and TJ is going to [61:24] obviously be there. Just come and say [61:25] hi. [61:26] >> Say hi to Eric or John or any of any of [61:28] the team and we'll we'll introduce you [61:30] to people. [61:31] >> Um, and I I understand what it's like. I [61:34] I have um social anxiety, so I I know [61:37] what it's like trying to be in a in a in [61:38] a building full of people that I don't [61:40] know. Um and I had never felt at ease so [61:44] quickly when I turned up to tech last [61:47] year when I came to the first one. Um [61:50] and I think it's I think it helps that [61:51] we we do over a spread of a few over [61:54] three days as well. So you start to see [61:56] the same faces the next day and the next [61:58] day and start talking to people. uh not [62:00] knocking any of these on one day [62:01] conferences because I've been to quite a [62:03] few and you do pick up a lot of stuff [62:05] you don't get that social time to [62:08] actually um meet people properly um I [62:12] feel um although I the the conference [62:16] Chris was talking at was it last week or [62:18] the week before um and I went along I [62:20] went along [62:20] >> last game [62:21] >> yeah I can't remember when it was the [62:23] days are emerging now um we I I've I [62:27] obviously sp met a of Chris's con sort [62:31] of people he knows because I didn't know [62:33] anyone there um and he introduced and um [62:35] luckily got to chatting to a few people [62:37] that was it was nice but I found when [62:40] I've gone to conferences on my own when [62:43] there's only it's only one day you don't [62:45] always get that that connection with [62:47] people um but again I've learned a lot [62:50] from those conferences so I'm not [62:52] knocking them it's they're brilliant to [62:53] go to um I wouldn't say don't go to them [62:56] um you might be more easy to talk to [62:59] than I am [63:01] and meet people. But uh yeah, no tech [63:04] tech over the three days with the [63:05] socializing in the evening. [63:07] >> Um it I've not experienced anything like [63:11] it since um all the same. So I I [63:14] absolutely absolutely love it and really [63:16] really recommend it. Really recommend it [63:19] to anyone. Um it's a long flight from [63:21] where we are, but if you can get over [63:23] there, get over there. Submit to submit [63:26] to JST because we'll fly you over if you [63:28] get selected. [63:30] >> Not only fly you over, we'll put you up [63:32] in the hotel. [63:33] >> Oh, yeah. Well, no. Yeah, we'll put you [63:34] in the hotel as well. We won't make you [63:35] sleep on the streets. You can you can [63:37] have a hotel room as well. [63:38] >> You can if you want to, but [63:40] >> Well, yeah, it's it's optional. [63:43] >> But yeah, no, definitely definitely get [63:46] in there. You got a few more weeks [63:47] before they that call for papers close. [63:50] And if you've missed it this year and [63:52] you don't get this the um the ability to [63:55] come over, uh there is virtual tickets. [63:58] Um so you can get a get a virtual ticket [64:00] to to watch the talks, but try try for [64:03] I'm sure I'm sure 27's going to happen [64:06] as well. [64:08] It's a it's a coming up to the stage [64:09] where Eric and John were never doing it [64:11] again, but they will still do it again, [64:12] I'm sure. [64:14] >> But yeah, so try again next time and and [64:17] come along. It is it is a great [64:19] experience. Um, [64:20] >> I mean, as stupid as it sounds, going [64:23] from not knowing a single person in that [64:25] room other than to have met them on [64:27] Discord to sitting around a table eating [64:30] sushi with TJ and Daniel Cobborn and [64:33] it's like, what am I even doing here in [64:35] the middle of Chicago eating sushi with [64:38] people that wrote packages I use? Oh [64:40] man, I'll tell you what, like my like [64:43] handful of best friends that I've the [64:46] longest these these are friends that [64:48] I've had for the longest amount of time [64:50] in my life. Um people that I talk to [64:53] like all the time. Uh I met them in 2016 [64:56] at a Laracon going out to like we went [64:59] out to like dinner together. Like [65:00] someone had just like posted on Twitter [65:02] that like, "Yeah, hey, we're at Laracon. [65:04] We're going out to dinner." And I think [65:06] I met Eric at the same dinner. like it [65:08] was just this like wild experience where [65:11] uh you know just I went along for the [65:13] ride on a whim. I didn't know anybody at [65:15] the conference beforehand and then [65:16] walked away with like four or five like [65:18] my my best friends. [65:20] >> Um so like it's it's [65:23] >> the magic of of of being at some of [65:25] these conferences. So, and it's for me [65:28] it's been ubiquitous across uh Laracon, [65:33] Laravel Live Denmark, uh PHP Tech, like [65:37] this like the PHP community. It's so [65:40] welcoming across the board. Um it's not [65:43] it's not just PHP tech, but like just [65:46] this community in general. Uh you can [65:48] like I landed in uh Larvo Live Denmark. [65:51] I didn't know anybody. first time [65:53] overseas. Uh, and I left with like a [65:56] whole bunch of new friends. Um, it's [65:58] just like the scene like the PHP scene [66:01] is just welcoming. It's friendly. It's [66:03] inviting. Um, [66:05] >> you know, I can't tell you like how many [66:06] conferences we go up to and there's like [66:08] gaps and circles of people talking. Like [66:10] plenty of room for you to just like walk [66:11] in, you know, just listen in, drop in. [66:14] Um, and yeah, just, you know, you see [66:16] the people out there that are are doing [66:18] live streams that are like, you know, [66:20] talking. And if you find them, just go [66:22] like introduce yourself, say hi. Like I [66:24] think one of the the most rewarding [66:26] experiences for me was of of Laracon [66:28] this year was I was walking between the [66:31] venue and the expo center and someone [66:33] ran up to me and they were like, "Hey, [66:34] let me show you what I built with [66:35] Prism." It blew my mind. Like just like [66:39] somebody is using this thing that I [66:41] built. Um but I had never never met them [66:44] before. They just ran up to me and was [66:45] like, "Hey, you got a second? Let me [66:46] show you what I built with Prism." And [66:48] like that was one of the more meaningful [66:50] experiences that I walked away from the [66:52] conference with. So yeah, just walk up, [66:54] introduce yourself, and say hi. [66:56] >> Yeah. [66:57] >> And I I have yet to find a person in PHP [67:00] who doesn't say hi back. [67:02] >> Yep. [67:03] >> For sure. [67:05] >> And that is one of the things that makes [67:07] this such a good language. The the [67:10] community is the center of it. like you [67:12] said it, you can't not have the [67:16] community in PHP [67:18] and not see success. [67:21] >> I mean, I can put a question on Twitter [67:23] and it's highly likely that the person [67:26] that wrote the package I'm asking about [67:27] replies to me [67:29] >> and that that's just weird. [67:33] >> But also, I think being involved in the [67:35] community, going to conferences and [67:36] things like that, how many people have [67:38] got jobs through that those networking [67:40] as well? [67:42] Oh, my entire career was based on my My [67:46] entire career has been based on the [67:47] decision to be involved in the Laravel [67:48] community. [67:51] >> You're going to have to give us some [67:53] context to this one, TJ. [67:59] >> We know I I I think I might know some of [68:01] it. [68:03] >> I definitely didn't crash at his place [68:05] over the summer for a long weekend and [68:08] it was not creepy. Mostly. [68:12] >> Was this the weekend where you didn't [68:13] appear on the podcast, too? [68:15] >> Yes, that was definitely the weekend I [68:17] didn't appear on the podcast. [68:20] >> Yeah, he's just CGIed you into the back [68:23] of the video. [68:26] >> Yes. [68:28] Okay. Um, so, um, we do need to start [68:32] wrapping things up. Um, I think we could [68:35] sit here for hours longer just chatting [68:37] away and like a lot of most of our [68:40] guests we or all of our guests we could [68:42] carry on talking for ages. So, we will [68:44] definitely have you on the short list [68:46] for next year to come on the show. We'll [68:49] catch up with you, see what Prism's up [68:51] to, what new stuff is coming. [68:53] >> Yeah, if you want if if you want, we're [68:55] not going to kidnap you and force you to [68:56] be on the show. [68:58] >> You're not. I clearly am. [69:02] No, I'd love to come back anytime. [69:03] >> You know who's good cop bad cop now. [69:05] >> Yeah, there we go. [69:07] >> Um, so yeah, I mean we we'll we'll start [69:10] wrapping thing things up, but uh Tina, [69:12] thank you very much for coming on. It's [69:15] been one of those episodes that we've [69:16] we've been trying to get in. It's not [69:18] been very easy to find when we you're [69:20] available to jump on and I love the fact [69:22] we've done live uh with you and we've [69:24] been able to get some questions in. Um, [69:27] and definitely we'll get you in again [69:29] next year and we'll we'll chat about all [69:31] sorts. Um, [69:32] >> perfect. [69:34] >> So, I need to bring my script up because [69:35] I will forget something otherwise. [69:38] So, um, yeah. So, thank you very much. [69:40] Uh, thank you very much everyone for [69:41] watching or listening or however you are [69:44] consuming this uh this show. Um, one [69:47] thing we are going to and have I got it [69:49] ready? Have I got it ready? It's there. [69:50] It's there. [69:52] It is there. But I've got to remember [69:55] how to share my screen. [70:02] There it is. Ah, there we go. [70:07] Speakers are starting to be announced [70:09] for tech next year. Um, I probably [70:13] should scroll down the page a bit. I [70:14] can't see what I'm doing when I'm in [70:16] this screen, but yeah. So, we have a a [70:20] really good selection of to uh speakers [70:23] already already selected. [70:27] So, if you haven't got your ticket yet, [70:30] get your tickets [70:31] >> and come and join us. [70:33] >> There's some there's some blooming good [70:34] speakers in amongst who've already been [70:36] selected and there's still more to go. [70:39] >> Yep. Yep. So, keep an eye on the on the [70:42] the website. Um, have you got the link [70:44] there to show Chris? Ready? [70:47] >> You mean that one? [70:48] >> This is This is This is going so smooth, [70:51] so professional. [70:52] >> We We're just learning from John and [70:54] Eric. That's all. [70:55] >> We We've got the best teachers in the [70:57] world. Uh, phpech.io. [71:00] Head over there. You can keep an eye on [71:01] what uh who's uh who's going to be [71:03] appearing. Um, also don't forget to join [71:06] us in Discord. Uh, people are in there [71:09] all the time. Uh, including TJ will be [71:12] in there. If you if you tag him, I'm [71:14] sure he'll answer. [71:15] >> Yeah. [71:15] >> Um, [71:15] >> I I summoned him the other day. [71:17] >> Yeah. If you summon me, I'll show up. [71:19] >> Yeah. Do you have to say your name [71:21] twice? Three times. It's three. You have [71:23] to say your name three times before you [71:24] show up. [71:24] >> It's five times into a mirror. It's like [71:27] handyman. [71:28] >> Right. It's that one. Right. There's so [71:30] many different rules. I can't remember [71:32] who goes by which rules. [71:34] >> It's bloody Mary if you want it three [71:36] times into a mirror. [71:37] >> I'm I'm more preferential to to [71:39] Beetlejuice. So, three will work. [71:41] >> Yeah, [71:43] >> that's what I was going off of. There [71:44] you go. [71:45] >> So, discord.phparch.com. [71:47] Come and join us for the chat. And uh [71:49] we're loads of great people in there all [71:51] week long chatting along. Um so, do join [71:54] us. Uh if you're on our YouTube channel, [71:56] please uh like and subscribe. If you're [71:58] not on the YouTube channel, go over [72:00] there, like and subscribe, and you can [72:02] see all the other episodes. And uh the [72:05] other podcasts that we we do as part of [72:07] the same channel. We have the main show [72:08] on a Thursday evening with Eric and [72:11] John. Got us, which I don't know what [72:13] our schedule is anymore because it's [72:14] like [72:16] >> one out Thursday. I don't know who it is [72:18] yet. [72:19] >> Yeah. Yeah. We'll we'll be taking place. [72:21] We'll have something going out on [72:22] Thursday. Taking place as the main one. [72:24] Uh because our fellow friends over the [72:27] pond will be celebrating its [72:29] Thanksgiving. Have I got that one right? [72:31] >> Yep. [72:32] >> Thanksgiving over there. [72:33] >> It is. [72:34] >> Yeah. So, we'll be we'll be uh putting [72:36] something up on Thursday. Are we doing [72:38] something Friday as well then? As well [72:41] as Thursday. We're just doing Thursday, [72:43] not Friday. We're just putting [72:44] >> Not Friday. We We'll bring it forward. [72:45] >> We're just doing So, we're just doing [72:47] one more this week. Originally, we were [72:49] going to do one every other week and now [72:52] we're doing two a week. [72:54] >> Yeah. This is episode 17 [72:58] and we've been going for three weeks. [73:02] Um so, okay. So that's our YouTube [73:04] channel. Uh don't forget magazine [73:06] discount for our lovely listeners. Alive [73:08] 3 will give you the first three months [73:10] of a annual digital subscription. [73:14] We keep talking about it. I'm going to [73:15] say it again. Tech PHP Tech 26. Do come [73:20] and join us over in Chicago. It is a [73:22] it's a lovely lovely week. Um great [73:25] bunch of people as we've we've tal [73:27] talked about already. So I won't keep [73:29] pushing but don't forget JStech. Uh, [73:32] call for papers is still open till the [73:34] 7th of December. TJ, 7th of December. [73:37] You can learn JavaScript by then, can't [73:39] you? [73:39] >> I think I can probably pull that off. [73:42] >> There you go. There's There's your talk. [73:43] There's your talk. How to learn [73:45] JavaScript in two weeks. [73:47] >> Claude, teach me JavaScript. [73:50] >> Exactly. [73:50] >> No, don't do that. [73:55] >> Yeah, that'll be interesting. Um, and [73:56] then we've got our swag sort. We have [74:00] >> head over there and you can see t-shirts [74:03] and uh some hats and caps and things at [74:07] store at p uh sorry store.phparch.com. [74:11] Head over there and grab some swag for [74:13] it. And uh that is everything. We have [74:17] ticked all the boxes. [74:19] >> Thank you very much again TJ for joining [74:22] us today. Uh thank you everyone. Thank [74:24] you everyone for watching live and we  ## 📺 Video ![](https://www.youtube.com/watch?v=nzeTSCQLd3M) --- # 📄 Full Transcript *Auto-generated from YouTube captions · timestamps roughly every 30 seconds · [watch on YouTube](https://www.youtube.com/watch?v=nzeTSCQLd3M)* **[00:00]** Heat. Heat. **[00:59]** Heat. Hey, Heat. Hey. **[01:34]** Heat. Heat. **[02:21]** Heat up here. **[02:55]** Do you lie awake [music] at night? Are you afraid to turn out the lights? Is there something you're scared of? What's there [music] out in [singing] the dark? Make sure to keep all your windows locked. Do you do you surrender? Try to try to stay awake. These walls you [music] can't escape. Here's one for the lonely hearts. For the ones that fear the dark. You better **[03:26]** run out. [music] Run out. Out of here. Bet you can't look the other way. Here's one for the broken road. That's the only one you know. You better run out. [music] Run out. Truth or dare. Get me out. Take me anywhere but here. anywhere but here. Wanting something good [music] to come. Waiting for things to run away from. Are **[03:58]** you Are you just hiding? [music] Yeah. Walking into the fire, talking [music] like you never die. Like you're like you're immortal. [music] Try and try to stay awake. These walls you can't escape. Here's one [music] for the lonely hearts. For the ones that fear the dark. You better run out. Run out of here. Bet you [music] can't look the other way. **[04:29]** Here's one for the broken road. That's the only [music] one you know. You better run out. Truth or dare. Get me out. Take me anywhere but here. **[05:00]** Do you lie awake at night? Are you afraid to turn out the lights? Is there something you're scared of? Make sure to keep all your windows. **[05:30]** [music] Hello and welcome to PHP Alive and Kicking brought to you by PHP Architect and our partners over at PHP Score. More about them later. This is a podcast that explores the developments in PHP and what it's like to earn a living in PHP today. I'm one of your hosts, Mike Paige. I'm a developer at PHP Architect **[06:02]** and my co-host is a longtime contributor to our magazine and now team member Chris Miller. Hi Chris. >> Hey Mike. How you doing? >> I'm good, thank you. I've been trying you've been chat putting messages in the private chat on the left hand side but all of a sudden I can't open it. So I don't know what you're saying to me anymore. >> I haven't said anything for [laughter] >> There's messages popping up but I can't open it for some reason. >> I'm ling I guess I guess. >> Yeah. Good, good old good old uh reream. Um anyway, today is live everyone. So if **[06:33]** you are listening right now, um please head over into our discord. Uh you can ask questions of our guests or us if you want to know anything about us. Uh but do uh do head over there. You don't have to stay there just for the show. You can we're in there all week long. There's lots of great people you can interact with and have conversations with. Uh please do like and subscribe to our YouTube channel. And um that >> that is not the YouTube channel. >> That's not the YouTube channel. >> That's the YouTube channel. There we go. **[07:06]** Um I did mention in the intro there that we do have a magazine and we do have a discount code for our listeners. A live three will give you the first three months of a yearly dis digital that's the one digital subscription. Uh, so yeah, do head over there, check out the articles in the the magazine. Um, always great content each month. I cannot talk. I think I should just stop. >> No. Okay, I'll carry on. We're not going **[07:37]** to we're not going to in keep swapping over like Eric and John tried on on Thursday. No, >> no, no, >> no, you're leaving to me. All right, then. Okay. As well as uh being a a development studio and having the magazine, we also run annual conference PHP Tech uh tickets are now on sale for next year. So it'd be great to see as many people there as possible. But not only that, alongside PHP tech, we are also going to be running JStech. The **[08:08]** same time, same location, but purely about uh about JavaScript. And I believe the call for papers is still open. >> It is. >> Yes, it is. It's in Is it the first week of December it closes? >> I have no idea because the date's moved twice. [laughter] >> Well, at the moment it's open. So, get those talks in. Uh if you've got some ideas about uh uh Java uh JavaScript ccentric talks, um we also have a swag store where you can get some lovely **[08:40]** t-shirts, hats, caps, all sorts of things. Um I think Eric keeps adding new new things every now and then. Every time he comes up with something that he wants, he'll add it to the store. So >> um badge him. If there's something not in the store, badge him. He might actually stick it in there. >> Okay, that's all our housekeeping, uh Chris. So, >> so without further ado comment, >> uh, not quite yet. Not yet. A little bit early. I know I'm in the UK and it's in **[09:10]** the evening, but I'm not quite yet. Not yet. No, it's it's it's actually the opposite. I'm just tired. [laughter] >> So, we're definitely not going to try and convert you to become a PHP developer at JStech right next to PHP Tech. >> Oh, we're not. I thought that was our whole idea. That was the plan. >> Shush. We're not letting that bit out in public. >> I just I I just thought we were going to have to hire security to have PHP developers and J and JS developers in the same vicinity at the same time is a bit scary. >> I mean to be honest, PHP will just win. **[09:42]** So it's fine. And and JStech CFP closes on December the 1st. >> Oh, thank you. Thank you very much. Um, so we we better stop nering because we got someone in the green room that's sitting there waiting. >> Um, they were even dialed in to reream before either of us. >> They were sitting here waiting when I came and I was in there early early. So uh Chris over to you to uh introduce our guest. So, we have a fellow PHP **[10:13]** developer, creator of Prism, works at Geocodio, does streaming, does basically everything. We have TJ Miller. Hey, how you doing? >> Hey, how's it going? >> Hi, TD. It's been uh it's been one of these this is one we've been trying to get in the diary for a while. >> Yeah. >> And then it see once we finally got in the diary, it seemed to take ages to actually get around to today. But we're here and we have >> we're here. We're doing it. **[10:43]** >> Yeah. And of course the important things are caffeine. >> Yes. Absolutely. >> So TJ, you've been watching this for a while. You're going to know what the first question is. How did you get into BHP? >> Oh man, this is such a super boring story. So um this is many years ago. I think like 2004ish, maybe a little sooner. Uh I was like I **[11:13]** was building websites with just like HTML and CSS for like lots of local businesses. I was in high school and I wanted [clears throat and cough] I wanted a way to like templatize some of the work instead of having to like you know copy and paste over all the sidebar, the header and everything over to these like across all the HTML templates and everything. Um, I wanted something like dynamic and that's when I like discovered PHP and was able to start templatizing some of these websites that I was building so that like the header and the sidebar could **[11:45]** stay static but we could swap out the contents on the inside of the page. Uh, and like that was so cool to me. And then like started getting into like database stuff a little bit. I'm like, oh my gosh, you can like connect databases to this. Well, that's a whole another world to like learn on top of it. Uh, so that's kind of like my entry point into PHP was just like needing some dynamicism to like that's not even a word, but uh I needed just to kind of like be a little bit more efficient about what I was doing. Um, and **[12:18]** needed at some point like access to like data layers. How am I going to do that? Well, we can do that through PHP. So dove further into that but all just got started building like websites for local businesses. >> Yeah. So that would have been at that point you would have been using uh PHP includes. >> Mhm. >> Oh yeah. >> I remember building websites like that. My entire website was just one big include file. >> It was [laughter] >> I mean to be fair that is what most of my code looks like now. [laughter] **[12:53]** So, interestingly, you seem to have journeyied on just a little bit from there, seeing as you're now writing one of the greatest AI integrations in Laravel. >> Oh, thank you. [laughter] Yeah. Yeah. We've come a long way from there for sure. Like, you know, I I early career like I got started in WordPress doing like lots of just pumping out like marketing websites. Here's PSDs, turn them into websites. um did that for quite a while and that's **[13:23]** where I dove like much much deeper into PHP and uh I was actually at one point um working on a uh the agency I was working for uh had a a cake PHP product that they they were working on and I hated it. Uh just like the way it was written was like kind of crazy. I I don't know. I just like I hated it. Uh, so I was about ready to jump ship for Ruby on Rails, but I decided to go look at like the other PHP frameworks out **[13:54]** there. So I went and looked at uh like Falcon looked really interesting. I was thinking about going down that route and then I discovered Laravel and I was blown away. I'm like you could write PHP this way. Like that changed everything for me. And from that point on, I just made a commitment to be inside of PHP and work with work with it like explicitly ever since. >> Yeah. Cake PHP is an interesting one, isn't it? It's you either love it or **[14:25]** hate it and nowhere in between. >> Yeah, I haven't looked at it since. Um, but it was it the I I think the one thing that drove me crazy about it, uh, and this was all like pre-omposer and everything too. Um, but what drove me crazy about it is like the the developer that worked on it at the time kind of made it his life's mission to make the the templating engine was like an object-oriented templating engine that you could opt into and he tried to to make every template like a single echo **[14:56]** call. So it was these deeply nested chained like this div this like the span like just nested all the way down into a single echo statement for like building out an entire templated page. And so that just drove me insane. >> Yeah, that is not how you use templating. Yeah, it was it was fun to to play some code golf with it, but it was very frustrating when it was like, yeah, you got to add a whole new like div in here to do some styling. Oh no, **[15:29]** like [laughter] where the hell is this closing bracket going to go? 25 files is further on, obviously. >> Yeah, basically. >> Yes, I have worked on sites like that. [clears throat] They're they're interesting beasts to work on when you when you're looking at code and you're going, "Right, I know that the other half of this is somewhere, but it isn't in this file or that one or that one." And by the time you've got halfway down the list, it's like, "Where did I even start?" [gasps] >> Yep. Yep. I don't know what I would do **[16:01]** without my uh uh my find files in Neoam. Like, I just I don't I don't look through directories ever anymore. It's just Yeah. Start finding me a file. fuzzy finding it or I like grip through it and find it. Uh but yeah, back back then I didn't have any of those tools. So it was definitely like clicking through and deep deep searching nested files >> and of course not knowing how to search inside the file. You having to read the whole thing? >> Yep. Oh yeah. >> So So when did you learn how to quit Vim? **[16:32]** Uh very early on I uh my early even before I really started building websites uh my grandmother had given me an old laptop and I just started like playing with DOSs like yeah I'm like let me let me get into the command line and felt like such a cool little hacker that I could like list directories and stuff and uh kind of broke into like nano and vim from there like as soon as I got into Linux and you got got stuck once figured out how to get out of it and uh but was hooked. I I've been uh a Vim and **[17:05]** Neo Vim user for 15 20 years, maybe 20 years now. [sighs and gasps] So big fan. I got hooked. >> I think the thing I like about it is the fact that it is so customizable. >> If if you want to change something about VI, go ahead. You can just change it. It's fine. [laughter] Yeah, it's funny to go through like my my Vim config has evolved a lot over the years, but uh there are there are key bindings that have been in there for yeah 20 years that are just so ingrained **[17:38]** in muscle memory that it makes it hard for me to really use any other editor at this point. Yeah, I I occasionally I log into PHP Storm and it's like right >> how do I do anything and and I keep doing like control S to save the file and oh yeah this is PHP Storm I don't need to do that and [laughter] >> yeah oh yeah yeah like losing my leader key I get so lost in like any other ID at this point like I I've tried I've tried a **[18:08]** whole bunch recently with all the like the wave of AI editors and everything. You know, I wanted to try them out, give them a shot, but I'm so unproductive in them. >> Uh, thank God for cloud code and I can stick in my terminals. [laughter] >> Well, I um I I use PHP Storm, but for a while I did use uh Neovim and um then I can't remember why I wanted to try PHP Storm. I think it was because a lot of I **[18:39]** I was very new to Neoim. I'd followed Jess Archer's um tutorial on Larcast on how to set up near Vim for PHP developers. I knew there was just so much more that could be done, but I didn't have the time to do the research and figure out how to get it working in near them. So, I just went >> go to PHP Storm because there's a lot of features built into PHP Storm. But one of the first things I did was um change it over to use um Vim bindings. So I've **[19:11]** I've brought a lot of my bindings from Neo Vim into PHP Storm and that's always still evolving and and getting bigger and bigger and bigger as I'm discovering more and more things. >> Yeah, >> there's a lot more I can be doing. I know that. Um, but yeah, I like I unfortunately it means that my Neovim setup has been neglected for so long and I went to go and try and do some stuff in my dot files and my Neim just doesn't work. >> Cannot get it working anymore. I was like, uh, okay. **[19:42]** So, um, my Oh, um, just just a quick one. I think that Eric's got the uh call for papers for JStech wrong because John's just jumped in. John John has said the JST closes on the 7th. Eric has said the first. So what we'll do is we'll say the fourth and then everybody will get them in early. >> Yeah, we could say well. >> Yes. Right. >> The fourth, but but if Yeah, **[20:13]** >> John is being very clear there. It closes on the 7th of December. >> Right. Seventh. >> It is the seventh. >> Okay. >> Unless we move it again. [laughter] back to back to Neim. So yeah, unfortunately I went into Nevin the other week to try and uh do some dot file changes and just try change a few text files which I don't actually do very often outside of PHP at the re recently. Um and even things like markdown I do I've been doing in PHP store but yeah um it's all broken so I'm **[20:44]** going to have to call on someone to help me fix it. I did try to I did try to do it in uh get it fixed in Claude, I think, but it just got a bit messy. It didn't quite work. [laughter] >> Claude has no idea how to do Neo Vim configs. I've noticed this. [laughter] >> I'm I'm pretty sure none of the edit none of the staff there use Neoim. >> Nothing about it. >> Yeah. And all configed in Lua and yeah, it's it did uh I had it update. had write a couple plugins for me recently **[21:15]** to uh to offload a few like things I wanted to try out and it it did a pretty okay job, but yeah, it's it's not super well educated on configuring Neoim with with Lua for sure. >> So, mommy and daddy are fighting. >> Yes. [laughter] John says he moved the conference because I'm speaking at a diff moved the call for papers because I'm speaking at a different conference which is a JS conference and um **[21:47]** >> we could just do that. Yeah. And >> we'll do that. So we we can follow follow the follow the argument. [laughter] >> So the rest of this is just going to be watching John and Eric decide who moved it. Um [laughter] so MVPOP says that Claude Code is actually good at Neoim configs. Right. What what's your secret? Because whenever I ask it to configure anything, it just gets it wrong. >> And I'm glad you say that, Chris, because I started to think it was me. [laughter] **[22:18]** >> Oh, yeah. Yeah. I've had I've had to build um two I was experimenting with some like Prism and AI stuff and trying to integrate it into like some Neoim things and I was having a really hard time with it. So, I had Claude like knock it out. It's nothing interesting. I've like abandoned the project at this point, but uh it did a pretty good job being able to like put most of the things together and you can always point it over to the docs now and you know have it go crawl the docs and it's got it's got web fetch so it can go go crawl **[22:50]** web pages. >> So, so was this the um ability to use Prism in Neoim that you've abandoned? Uh it was a way to it was a way that I could like get a popover window with prism output uh using like a PHP script invoked by Neoim. >> Nice. >> And like passing in like like buffer context and all sorts of other information. So it was it was a I just wanted to see if I could do it [laughter] **[23:20]** and I could do it and then I've never used it since but it was just kind of a itch to scratch. Well, I I really want to see what that code looks like now. [laughter] >> Yeah, I I'm mad like that. I like seeing things that are just weird experiments that usually end up being something cool. You you look at something like um not claude code and cursor that started as somebody's idea of what happens if we put AI in an IDA. **[23:52]** >> Yeah, sure. M >> and if you could put AI inside a good IDE like Neoim, you you end up with quite a powerful little tool. >> Yeah, it's uh there's much better like AI integrations into Neoim than than that. It was uh at the time I was like I I have these system prompts that I really like to use these like personas and uh I I leverage those a lot through through Prism. I've got like a little open web **[24:24]** UI app that I have like running persistently on my computer and then that uses a Laravel like that connects to like a Laravel API uh through Prism and so I can like chat these custom personas without having to you know do weird things about trying to like reset system prompts or or other stuff. I can just kind of like interact with them just like I'd interact with like pulling up like claw desktop or something. So, I uh I like to like pipe a lot of stuff like through Prism because I can just, you know, make make them do whatever I'd **[24:55]** like super easily. >> I I threw that up because Eric said to one of the people watching on YouTube, "Come and join us in Discord." And they're now in Discord, which is cool. >> Yeah. to it. Um, on that note, I think it's time to um um hang on, find the right video. Uh, go to our uh our partners for today. Thank you to our partners at phpscore.com. **[25:26]** Every app builds up technical debt over time. It's the price we pay for shipping new [music] features and moving fast. When we build up too much of it, though, it can start to impact how we work. [music] Team velocity suffers. bugs become more frequent and take longer to fix and everyone starts to get a little frustrated. The key to managing debt is to measure it. A credit score can help you understand how well you're managing your financial debt. And now there's [music] a credit score for your technical debt. Go to phpsscore.com to get a free technical debt score and **[25:58]** monitoring for all of your PHP applications today. >> Thank you. PHP school. PHP score. >> I'll do I'll do the echo today. [laughter] I was reading some of the chat. I was catching up with chat, but you know what it's what it's like for me. I'm always three or four messages behind. >> So am I. And they're usually my messages. [laughter] >> Yes. Well, so um I think it's a good **[26:29]** transition to start talking about this package that you've been >> working on recently that's got everyone excited. >> How did that come about? Where did you where did the idea come from to begin with? >> So it started uh when I did the a little while before the Laravel worldwide meetup I did. Uh I forgot what year that was. Um I think it was a couple years ago at this point and like sort of getting ready for that talk but um kind **[27:01]** of my journey getting into AI led me to a point where um this was like very early in AI. This was uh when I got started with it. It was uh maybe like 6 months before chat GBT came out and I just kind of got bit by the AI bug and I could I saw like a ton of potential with it. Um I've got like pretty severe ADHD and I saw so much potential for me to be able to like build things to help augment my reality and like things to be able to like help **[27:33]** me through like day-to-day work life, whatever. um through through AI and like the potential I could see it having as it got more powerful and uh I didn't know Python at the time. All the A IML stuff at the that point was like pretty solidly in Python. There was some TypeScript stuff starting to come out. Um Langchain was like getting getting popular. There was a a TypeScript version of that that came out. Um, but I **[28:04]** missed spending all of my time in Laraveo. Like I've spent so much time like so much of my development career has been spent with it. I love the framework. I live and bleed it. Um, but I I wanted to be back in Laravel and like working on these experiments and continuing to learn more about AI and like building like these tool sets. Uh, I just really I I wanted to be back inside of like PHP and Laravel and like there wasn't really any tools at the time. Um, so I decided to like start **[28:37]** working on a little framework. At the time I called it Sparkle and it was kind like this like similar to Prism but definitely different. It was like an agent focused library. um it didn't it had nowhere near as like well ststructured uh like multi-provider support but it did support multiple providers um so it was kind of this like early iteration of Prism but uh I had been showing progress on it to Taylor and Taylor had just like we were talking **[29:09]** about the AI SDK from Verscell and he said uh he just like in passing was like yeah I wonder what a Laravel version of that would look like and I sat there for a second and the API just kind of like hit me and I sat down and like started sketching out the API and like how you would interact with it and how like multi-provider support would kind of work and uh then just started like fulfilling that contract. So it was like very very much API first and then backfilling the functionality to meet **[29:39]** that API. >> Nice. >> So the right way around you thought about the the developer experience first. Oh, for sure. And then just it's always it's always kind of started there, you know, and as I'm facing like developing a few new features for it now, it's it's right back to that API development. It's like how do I want to make these interactions work? How is it going to feel? Um, and I think that's something that I've just gleaned from working with Laravel for so long, like that that really clean, uh, fluent, easy **[30:11]** to work with, like where you start assuming you work with it long enough, you start like assuming methods are there and then magically they're just there for you. There's like it's a very intuitive thing to work with. And so I took a lot of that inspiration and applied that to the API of Prism. >> And that that is one of the things that Laravel always does. The the developer experience is the first thing of Laravel. So much so that we actually for a while had people who **[30:42]** would call themselves Laravel developers and not PHP developers. >> Oh, sure. >> I mean, you see that in like the JavaScript community, too. >> Oh, yeah. Yeah. >> Yeah. like you get you get really like locked in like framework lock in um you know where that's you you learned you learned PHP by working with Laravel and like I know a ton of people that that that's how that's how they learned PHP was by starting like and picking up working with Laravel and then they learned PHP from there. >> Yeah. And and Laravel makes it easy to **[31:12]** make that bridge because helps you with stuff until you need to do more than the framework. And that's when you start diving into into the more complex stuff of okay now we're into business logic and how we all that then we're into how do we do multi-threaded in PHP and >> the answer is I don't think you can yet [laughter] [snorts] **[31:42]** >> we've got Nico on Twitter saying thanks for prison. >> Oh very welcome. It's it's such a passion project for me. Like I I love working on it. It's it's something I like I think about and and am constantly tinkering with. So um very welcome. Do >> do you know the point I realized that Prism wasn't your full-time job? >> Which was that? [laughter] >> When I saw an advert for a job at Geocodio that you were recruiting for. [laughter] **[32:14]** >> Yeah. So, yes, that that's when I found out that Prism wasn't actually making you any money. >> No, I have uh I do have like a handful of of sponsors that I I deeply appreciate and I'm I'm very thankful for. So, uh but I I do I chew through a ton of tokens every month like running tests and experiments and, you know, trying to like push Prism further and like, you know, even just driving out a new provider. um or like testing like **[32:47]** with Gemini 3.0 Pro that came out, there were a bunch of changes in the API that kind of needed to be made and we're still working through those with other contributors because that's it's just not a model and provider that I I use very often. Um so I don't have a lot of reasons to be testing it. So, uh, there's been a lot of really solid contributors coming back in, uh, and and bringing more support for Gemini, but, uh, I have to like manually test all of that stuff, too. So, it's I I've chew through a ton of tokens with Google over **[33:19]** the last like month or so, you know, driving out functionality, then they drop a new model, and then I got to go test all of that stuff. So, the sponsorships go a long way, but it's far from being a full-time gig. >> Yeah. And any sponsorship is going to help. And >> oh, it makes such a huge difference for sure. Like if I could get to a point like I think >> for me the dream would be to be able to make enough in sponsorships that I could like maybe drop a day a week at **[33:49]** Geocodio. Like be able to just like just dedicate Fridays to Prism all day every week, you know? Like that would be awesome if I could get to that point. Um but we'll see. or or talk Taylor into taking it on as first party. Oh, >> sure. [laughter] >> Yeah. I mean, I you say about you you churn through lots of credits. I I I can imagine because I I with um chat GPT I've only got the free account and I'm **[34:20]** always running out. I'm always asking random questions and it's just things I come up with. I just like ask it a question then leave it running for a little while and I'll go and do something else. come back to it churned out and I'm always running out of credit so I can imagine you going through quite a lot. >> Yeah. Yeah. I definitely chew through them and you know it's it's like spinning up accounts at all the different providers and like you know having to juggle all the API keys for everyone and >> you know it's remembering which one **[34:51]** supports what thing and it's it's a it's a wild time out there. I mean, look at the last week alone, we've had, you know, three new premier models drop. >> And with that, new new features and functionality, too. So, it's it's been uh a busy busy couple weeks with with Prism and and we're like I said, we're still churning through some of the API changes for the new stuff. >> Yeah, I can I can I can imagine. I hadn't thought I hadn't appreciated or **[35:23]** thought really about the the the cost of all the testing you must go through. And I'm now thinking about the the little idea that I we were talking about just before the show that I'm hoping at some point uh probably in the new year to get time to play around with. I'm thinking, yeah, I'm going to go through quite a lot of credits just testing that to get make sure it works. [clears throat] So, yeah. Yeah, that's that's that's very interesting. >> Yeah. One of the cool things I found out recently, I started playing well, I started playing a little bit more with um Olama and local models for test **[35:54]** experiments >> because uh a lot of times what I'll end up doing is like I'll I'll build a streaming chat app just to like get the feel for how Prism behaves in there. Uh like one of the things I'm working on right now is uh like message persistence. like how can we make persisting these conversations through building chat apps like a really clean experience. >> So like I I would be like chewing through tokens building that thing out **[36:25]** and like testing different configurations and different things, but um you know I I made the investment in a pretty like maxed out MacBook uh somewhat recently. So, I have the affordance to be able to run some of the bigger the the bigger local models and I've been pretty blown away like the the OpenAI's GPTO OSS model is very impressive. So for building out integrations and like experimenting with stuff, I think you can get pretty decent **[36:55]** mileage and like a pretty good idea of how a large language model is going to handle the task at hand through some local models and then you can kind of upgrade to like test on some of the more powerful stuff or if you're running into it just like not being effective, you know, scale up a little bit, you know, try try the next level up model. Um because I think from the efficiency side of things, you really want to nail on using the smallest model possible to complete the task. So you can kind of **[37:25]** scale up that way, kind of starting with local and then kind of upgrading through the smaller models. And the smaller the model, the less expensive they are. So you can kind of get away with, you know, sort of testing a little bit less expensive. And then you can also kind through that process of like leveling up a little bit at a time as you're trying to figure out what it's capable of, you'll also probably land on the smallest capable model for the job, too. >> Yeah. That's really interesting, really helpful advice. **[37:55]** >> And a lot of people don't realize you can even do that in some of the IDEs like claw code. >> You can actually tell it which to use. the it was either yesterday uh tight article on their blog about how to use like O Lama and self-hosted. So that's that's like a great resource for kind of like getting into that as well. >> Oh, definitely. So we've had um a lot of conversation in a very short space of time, which is good. Um first up is **[38:28]** Jimmy um who has a question. I'm a PHP developer for 15 years. Sure, I know other languages. I began with assembly and C. Now I'm tired with the bullies of PHP. I see JS like React and other stuff very complicated. [gasps] I've got the impression some juniors don't see PHP because PHP is old and not the actual hype. Oh man, there is this is so loaded. [laughter] Um, yeah. >> Yeah. This [snorts] is this is a gold **[38:59]** mine of of of stuff for like hot takes around PHP. Like I I definitely think PHP suffers from an image problem. Uh I've thought that for a long time. Like that's one of the things that I appreciate about Laravel and the Laravel community is like I think everyone really pushes hard to like elevate the language alongside like the the frameworks and like you don't you don't get to be using PHP without you know using a framework at this point for a lot of things. So um **[39:32]** I think a lot of PHP suffers from an image problem and I think we suffer because of that we end up sort of suffering at the top of the funnel right when we get new junior devs and people entering the industry like we suffer at that top of the funnel and that's where like JS is kind of snagging everyone at this point um is like that's where all the boot camps are like using like Python or JavaScript like that I think that's like 90% of the boot camps that I see. Uh I see I mean I'm in based in **[40:03]** Metro Detroit with a lot of like auto industry. So we see a lot of like Java boot camps around here too. >> Um but the vast majority of it is like JavaScript and Python. >> So you know we we as a brand PHP right we need to figure out how we can like capture at that top of the funnel. And I think part of that is like sort of with like the image issue. Um, so I was like really happy to see with uh the 8.5 release like they had like a really nice branded web page for **[40:34]** that. I think that was like super clean. I love how they approached that. Uh I think more more of things like that. Um are going to be helpful. So >> and I think some of the image we don't help ourselves with this and I keep seeing I've seen another one the other day. um where someone's posting maybe in LinkedIn for example this one was in LinkedIn and the the like the headline says something about PHP **[41:06]** was actually the question is PHP dead and there's a massive sort of the post then go if you extend the post there's a big explanation why it's not dead but the problem is no people are just seeing the is PHP dead and assuming PHP is dead that's it they're not going to open it up and read the rest of it. So those things we we've got to try and turn those around. Um and that's kind of where this this show was born from was a a me I was at PHP UK earlier this year **[41:37]** and there was there was a conversation about that about using more positive languages language to talk about PHP and not keep saying no PHP isn't dead. We we should just forget about that and not say that anymore. certainly don't title websites or posts with those words. Even if it's PHP isn't dead, people just read PHP dead. They don't get the rest of the the rest of the the sentence. So, I think as a a community, we we need to change the way that we're trying to present. **[42:09]** >> These posts aren't trying to damage PHP and I understand that they are trying to encourage people, but it's not working because [clears throat] people are seeing the negative before they see the positive. I think it's a little bit of like chicken and egg though too. Um I think those types of posts exist because there is this like connotation that it's dead >> and so I I think it it you have to be careful that it doesn't get too cyclical. But >> um >> yeah, I don't know. It's it's been something that has been on my mind for **[42:40]** for years as like someone who's like passionate about education and like sharing this stuff. Um it's like how yeah how can we get get more people in the door because there it's like just so advantageous like I I look at like setting up Ruby on like even just from like a CIS admin level like how much it takes to set up Ruby on a server, right? Verse how much it sets like takes to set up like PHP on a server. Maybe I've just been doing it **[43:11]** for so long, but like it's it's like an easy process to walk through. It's like it's well documented. It's it's because PHP has been around for so long. >> Um granted with that too, there's a lot of bad PHP around as well. >> Um but I don't know. I don't know. It's been something on my mind for a while and I love seeing that there is such an initiative now kind of across the board, **[43:42]** you know, internally in PHP, you know, podcasts like this, you know, the initiative for the 8.5 website, like I think I think there's a really healthy conversation going on about it now. And I think one of the things that's that's helping >> and you'll know the conversation we had in Chicago, >> but one of the things that I think is helping is actually the marketing machine that goes around a lot of the frameworks. So you've got Laravel drawing people in because they're good **[44:13]** at marketing for themselves, which in turn promotes PHP and then you got native PHP who just goes, "By the way, we can do mobile apps now." >> Yeah. And it's like, okay, cool. And >> yeah, and and with Prism, like I I don't want to like toot my own horn too much, but with like with Prism, like now you're not bound to like look at my use case. Now you're not bound to Python, >> you know? You want to work with AI, >> not a problem working in PHP. And I know **[44:43]** there's like a fantastic symphony package for working with AI as well. Mhm. >> So it's it's like more accessible than ever inside the language to be able to like be be accessing like new hot technologies, you know, >> and you've got the the official M MCP SDK. >> Yeah. Yeah. The investment from from the actual PHP Foundation to, you know, be involved inside of of of AI. I think that's >> so fantastic. And then and then you get some weirdo who's started a series of **[45:15]** articles in a magazine that's published every month by PHP architect on the basics of machine learning in PHP. So teaching linear regression and all sorts in PHP. I have no idea who's writing that series. Me. [laughter] >> Yeah. And these sort of things are all possible in PHP. And it's almost like we don't shout about it enough. We don't it at the lowest level we don't say **[45:45]** actually all of the cool things you can do in JavaScript. Yeah, we can do that. >> Yeah. >> And if there's not like direct solutions with it, there's a lot of affordances inside of frameworks to be able to do those things. So, I mean, you look at like Livewire, like being able to being able to do that inside of Laravel gets you so close to the JavaScript experience for a lot of things. >> Um, you know, I >> think it's a more exciting time than **[46:16]** ever to be getting into PHP with like all all the things that are are at our fingertips now. >> Yeah. Yeah, I mean um just to just to point on what Nico just said about the PHP is dead title being a clickbait and when you then read the article it says it isn't. I I get that the the concern that I think a lot of people have that I talk to is that people read that and they don't click on it. They don't read the article to find out. **[46:47]** >> You're just you're doom scrolling, right? You scroll past the title, you read the title, but you're not reading the article. And so you see the post on Reddit. You scroll past the title and it's like, "Yeah, is PHP dead?" You're like, "Well, I don't know. Is it?" But you don't, you're not clicking through to read the article because we don't have the attention spans anymore. But you're scrolling past it. You're seeing those titles like, "Is it dead? Is it dead? Is it dead?" Well, now you're starting to question it. So, I see where you're coming from, too. >> Yeah. Yeah. Um, there was another question, but I now can't find it. **[47:18]** >> There was It was um Yes, that's one [laughter] of >> is joining us because he's finished dinner and he's done the dishes and he wonders if some kind of PR um which Pton the Python light LLM model details would be of value maybe. Uh I'd be open to like checking out the PR like we don't have uh explicit light LLM support inside of Prism. um I think if you're going **[47:50]** through it as like an open AI compatibility layer um you might have access to I I don't know check it out. I open the PR like I'll I'll be happy to take a look at it or shoot me a DM on any platform. You go to my website. I've got a links page. You can find me everywhere. Shoot me a DM and we can talk a little bit more about it. Um but I don't I don't see why I'd be opposed to it. >> Yeah. And uh he went on to say that he uses light LLM a lot for prototyping and **[48:20]** it includes all the stats and the costings. I think that's more what he's getting out of the fact that you could pull out the stats and costings. >> Yeah. Yeah. You can um it we we do expose like all the like tokens and stuff that you're using through your requests with Prism, but you're you're it's still on you to be like totaling and exposing them. uh light LLM operating as a proxy to providers gives it that opportunity to like kind of build those stats up for you and store them. Um it's I I've not used light LLM **[48:54]** in a long time, but um it looks pretty nice. It's come a long way. >> Nico wants to know if you've ever accidentally committed an API key. >> Yep, 100%. [laughter] >> I think that's a good everyone's response. Yeah, [laughter] >> it's 100%. >> I I definitely haven't. I definitely haven't. And I'm now going to check every reaper. [laughter] **[49:24]** >> And uh Marian. >> Yeah. What is it? What is it? The the BFG. Uh is that is that what you use to to scrub repos of things? >> I'm pretty sure that's the CLI. Yeah, I'm I'm I'm familiar. [laughter] Definitely committed keys. So Marian says, "Let them think PHP is dead." Um I I get your point, but actually I'd rather they didn't think PHP was dead and came and programmed with us. >> Yeah. No, I mean uh the one of the scary **[49:56]** things I I experienced was that being at the PHP UK conference when they were doing a um a panel and that was one of the questions was talking about bringing in junior devs and before they answered the question um they did ask anyone over the age of I think it was 35 if I remember rightly to put their hands up that are full-time dead and pretty much the whole room put their hands up. That's a scary position to be in when you've not got anyone in their 20ies and **[50:27]** coming into into the language. What's gonna happen in 30 years time >> when all those 35 year olds are now 65 and trying to retire and no one else is coming back in? So I would say we we don't want to let people think PHP is dead. [laughter] We want to bring in bring bring in the new blood. We want to bring in new people and that's the way how how society and and communities thrive is bringing in new blood. >> Yeah. >> No. **[50:58]** >> One of two things will happen if we don't bring in 20 year olds. One, the language will eventually go away. Two, those of us that are in our 60s are going to be paid a hell of a lot more money. [laughter] >> There is that. There is that. Okay. PHP's dead. [laughter] Yeah, >> I know. This is exactly why I've been all in. I've been in the Strat for years, right? I'm just waiting for it to die so I can capitalize on all the **[51:29]** legacy code. >> That's exactly how Cobalt did it. [laughter] >> I I have seen Cobalt jobs now advertised at £150,000, about $200,000 for editing cobalt. >> Oh, sure. All right, I'm going to start googling some tutorials. >> Yeah, [laughter] >> I I will give you a clue. Anything that's on YouTube about Coal is going to be at least 15 years old. >> Yes, that's fine. >> Yeah, I haven't changed since like 1985. **[52:02]** So, [laughter] >> right. >> Nico makes a fair point. Cobalt guys make bank with banks. Yeah, >> it's true. >> Yeah. I mean, >> the the real money's in Forran. That's that's where you make the real money. >> Jeez. >> I I I was once responsible for writing Fortran for a very short space of time, mainly because it was bloody awful. [laughter] **[52:33]** It It's It's like Python in that the intend indentation of your code matters. >> Oh, yeah. It's It's weird because it's like got to be in the right columns and it's >> Oh, sure. >> No, I My brain doesn't work that way. Don't Don't give me columns. [laughter] >> Yeah. This is This is also why I don't like YAML, you know. >> Yeah. [laughter] >> I I do love YAML though. I mean, I I hate working with it, but I love it. **[53:04]** >> I I don't know who wants to tell Marian. >> He can't do maths. >> [laughter] >> He says he's from 89, so he's 26 years old, in which case I'm only 30. Yes. I wish [laughter] I wish maths work that way. >> I mean, we we could just make up maths. >> I think I was doing that earlier. [laughter] **[53:37]** >> Oh, was it yesterday? I can't remember. I'm I'm I'm lost. The last last couple of days have merged a bit. Um [laughter] >> maybe I am drunk. Maybe that's what's going on. I sniffed the whiskey when I went downstairs. That was enough. >> There you go. Gotcha. >> That was it. It's got me. Yeah. >> So, >> so I guess the next important question then TJ is, are you coming to tech this year? >> Oh, yeah. I'll be there. >> Good. Cool. **[54:08]** >> Yeah. Yeah. I missed uh I unfortunately uh I missed the CFP so I don't I don't think I'll be talking this year. Uh [snorts] but I will be there for sure. Have another CFP open. >> Yeah. What you should do is do a talk on PHP for JavaScript developers and put it into JStech. >> I thought about it. I'm not [laughter] going to lie. Not going to lie. Um, I I mean I almost thought about doing a like a Versel AI SDK talk as well cuz it is such a **[54:41]** similar realm that like I should be able to like pick up and run with it. But uh >> I don't know. I'm my my JavaScript is so rusty at this point. I think I'd be doing myself a disservice to like try to do a JavaScript talk and re like relearn the language. [laughter] If only you had some sort of system that let you integrate with multiple AIs that would help you. >> I know, right? [laughter] >> That'd be great. [clears throat] >> Just Yeah, just get AI to write it for you. No one will know. Honest. **[55:14]** >> Yeah. No, it should be gone. >> You're absolutely right. >> Yeah. [laughter] I just got to memorize it, right? That's good to go. >> No, no, no, no. Then you >> What you do then is you go to 11 Labs. You teach it your voice and >> [laughter] >> Let me just get on stage and hit play like DJ style, right? >> And then just go and sit in the front row and then ask yourself some questions. [laughter] >> Oh, I should have pitched that one. [laughter] >> Oh, that'd be great. >> I may do it. **[55:44]** >> I may pitch the AI talk. >> Yeah, >> it's totally doable. [laughter] I think the funny the funniest conference talk I ever gave was one on how to do conference talks. I said precisely nothing in the talk. It was just on like raising your voice and lowering your voice and all I was doing was demonstrating it with and now I'm going to raise my voice and then I'm going to lower it so that people think I'm talking about something important and [laughter] **[56:16]** >> nice. I love that. So, um, we're we're getting up to to our time now, unfortunately. It's been a been a a laugh and, uh, but we do have one more question for you. >> Yeah, that's um, what I mean, you've probably touched on quite a bit of it already, but what excites you about PHP now today or what's coming up? >> Oh, man. [clears throat] Uh, there's a whole bunch of interesting **[56:47]** stuff, right? Like I want to I want to play with some pipe operators a little bit and like brand new >> feature out. I want to mess with those some more. Um that's that's an exciting exciting feature ad for me. >> Uh I don't know. I thought about this question a little bit um early on like while we were chatting before the show and honestly like the thing that gets me the most excited about PHP and like it's the same thing that's made me the most excited about PHP **[57:18]** >> for years is just the community. It's it's the people that are involved in it. It's the, you know, the the things that we're doing. Like every time I go to like PHP Tech or Laracon or, you know, something, I come back so inspired to like do more and better things like especially now that I'm like working with Prism, you know, I came back from like all these conferences so inspired to like make it **[57:48]** better and like contribute more. Um, I don't know. It's like the it's always been the community for me. Like that's that's something that bit me early on in my like serious PHP career when I started, you know, got out of the marketing stuff and got into like building real world apps. Um, I went to Laracon. Let me check the badge. 2016 was my first Laracon. So that was like my first conference I had ever been to. Um, anything. And I walked away from **[58:20]** that conference just so committed to PHP, to Laravel, to the community. Um, I was just blown away by that and and I've been to a bunch of conferences since then. Uh, spoke at a handful and I just always come back to for me the most like one of the most exciting things is the community. So, like if you're watching this, like if you can get if you haven't been to a conference, like >> try and get to one. Like PHP Tech is like one of my favorites because uh it's **[58:54]** it's a wellorganized conference. It's I love the venue's like I think great for a conference. I've always enjoyed speaking at it, but it's such a great blend of like, you know, you got Laracon, which is like very Laravel heavy. I've been to a bunch of Laracons. I love it. Um, but this like PHP tech, I think, does such a good job of like giving you the exposure to like PHP as a broader language and like the broader PHP community. It's >> such a great entry point into that. So, **[59:25]** like I think that's a great conference to go to. Laracon, highly recommend. love it. They're all over. The Laravel live conferences are fantastic, too. Um, so if you can like get out there and get involved in the community, like highly recommend it. It's I think one of like the very special pieces about the language. >> Yeah, you I love the fact that you said the community. I'm afraid you have to forgive me. My dog is now barking at some idiot setting off fireworks. So, if you can hear hear my dogs barking, I do apologize, but I wanted to **[59:57]** weigh in here and not just keep muted because um I've said this on the show before on other episodes, but the PHP tech community um was so welcoming to me when I went to it first time last year. Um, I was obviously I was fly flown over as part of the team. Um, and I was working on it, but it was everyone was just just sit there chatting away with you and you would you would be sitting talking to **[60:27]** someone then you realize it's it's some someone from internals or or you're talking to Derek or or you're talking to Ben Ramsey and you go, "This been Ben Ramsey I'm talking to." And and one of the big things I think from that from especially from tech um is that there's no [clears throat] there's no clicks is what we would say there there's no sort of barrier between the speakers and the the general public that going to watch **[60:59]** the speakers mingle um all and there's so many people there. It's it's it is such a great experience. Um, I would >> highly recommend that anyone can go. Even if you have to get your your company to sponsor you to come over, come over and if you don't know anyone, just talk to one of us. >> You've seen our faces enough. You know who we are. We're we're all going to be there working and TJ is going to obviously be there. Just come and say hi. >> Say hi to Eric or John or any of any of the team and we'll we'll introduce you **[61:30]** to people. >> Um, and I I understand what it's like. I I have um social anxiety, so I I know what it's like trying to be in a in a in a building full of people that I don't know. Um and I had never felt at ease so quickly when I turned up to tech last year when I came to the first one. Um and I think it's I think it helps that we we do over a spread of a few over three days as well. So you start to see the same faces the next day and the next day and start talking to people. uh not **[62:00]** knocking any of these on one day conferences because I've been to quite a few and you do pick up a lot of stuff you don't get that social time to actually um meet people properly um I feel um although I the the conference Chris was talking at was it last week or the week before um and I went along I went along >> last game >> yeah I can't remember when it was the days are emerging now um we I I've I obviously sp met a of Chris's con sort **[62:31]** of people he knows because I didn't know anyone there um and he introduced and um luckily got to chatting to a few people that was it was nice but I found when I've gone to conferences on my own when there's only it's only one day you don't always get that that connection with people um but again I've learned a lot from those conferences so I'm not knocking them it's they're brilliant to go to um I wouldn't say don't go to them um you might be more easy to talk to than I am **[63:01]** and meet people. But uh yeah, no tech tech over the three days with the socializing in the evening. >> Um [clears throat] it I've not experienced anything like it since um all the same. So I I absolutely absolutely love it and really really recommend it. Really recommend it to anyone. Um it's a long flight from where we are, but if you can get over there, get over there. Submit to submit to JST because we'll fly you over if you get selected. >> Not only fly you over, we'll put you up **[63:32]** in the hotel. >> Oh, yeah. Well, no. Yeah, we'll put you in the hotel as well. We won't make you sleep on the streets. You can you can have a hotel room as well. >> You can if you want to, but >> Well, yeah, it's it's optional. [laughter] >> But yeah, no, definitely definitely get in there. You got a few more weeks before they that call for papers close. And if you've missed it this year and you don't get this the um the ability to come over, uh there is virtual tickets. Um so you can get a get a virtual ticket to to watch the talks, but try try for **[64:03]** I'm sure I'm sure 27's going to happen as well. It's a it's a coming up to the stage where Eric and John were never doing it again, but they will still do it again, I'm sure. [laughter] >> But yeah, so try again next time and and come along. It is it is a great experience. Um, >> I mean, as stupid as it sounds, going from not knowing a single person in that room other than to have met them on Discord to sitting around a table eating sushi with TJ and Daniel Cobborn and it's like, what am I even doing here in **[64:35]** the middle of Chicago eating sushi with people that wrote packages I use? [laughter] Oh man, I'll tell you what, like my like handful of best friends that I've the longest these these are friends that I've had for the longest amount of time in my life. Um people that I talk to like all the time. Uh I met them in 2016 at a Laracon going out to like we went out to like dinner together. Like someone had just like posted on Twitter that like, "Yeah, hey, we're at Laracon. We're going out to dinner." And I think I met Eric at the **[65:07]** same dinner. like it was just this like wild experience where uh you know just I went along for the ride on a whim. I didn't know anybody at the conference beforehand and then walked away with like four or five like my my best friends. >> Um so like it's it's >> the magic of of of being at some of these conferences. So, and it's for me it's been ubiquitous across uh Laracon, Laravel Live Denmark, uh PHP Tech, like **[65:37]** this like the PHP community. It's so welcoming across the board. Um it's not it's not just PHP tech, but like just this community in general. Uh you can like I landed in uh Larvo Live Denmark. I didn't know anybody. first time overseas. Uh, and I left with like a whole bunch of new friends. Um, it's just like the scene like the PHP scene is just welcoming. It's friendly. It's inviting. Um, >> you know, I can't tell you like how many conferences we go up to and there's like **[66:08]** gaps and circles of people talking. Like plenty of room for you to just like walk in, you know, just listen in, drop in. Um, and yeah, just, you know, you see the people out there that are are doing live streams that are like, you know, talking. And if you find them, just go like introduce yourself, say hi. Like I think one of the the most rewarding experiences for me was of of Laracon this year was I was walking between the venue and the expo center and someone ran up to me and they were like, "Hey, let me show you what I built with Prism." It blew my mind. Like just like **[66:39]** somebody is using this thing that I built. Um but I had never never met them before. They just ran up to me and was like, "Hey, you got a second? Let me show you what I built with Prism." And like that was one of the more meaningful experiences that I walked away from the conference with. So yeah, just walk up, introduce yourself, and say hi. >> Yeah. >> And I I have yet to find a person in PHP who doesn't say hi back. >> Yep. >> For sure. >> And that is one of the things that makes this such a good language. The the **[67:10]** community is the center of it. like you said it, you can't not have the community in PHP and not see success. >> I mean, I can put a question on Twitter and it's highly likely that the person that wrote the package I'm asking about replies to me >> and that that's just weird. [laughter] >> But also, I think being involved in the community, going to conferences and things like that, how many people have got jobs through that those networking **[67:40]** as well? Oh, my entire career was based on my My entire career has been based on the decision to be involved in the Laravel community. [clears throat and laughter] >> You're going to have to give us some context to this one, TJ. >> We know I I I think I might know some of it. [laughter] [snorts] >> I definitely didn't crash at his place over the summer for a long weekend and it was not creepy. Mostly. [laughter] **[68:12]** >> Was this the weekend where you didn't appear on the podcast, too? >> Yes, that was definitely the weekend [laughter] I didn't appear on the podcast. >> Yeah, he's just CGIed you into the back [laughter] of the video. >> Yes. Okay. Um, so, um, we do need to start wrapping things up. Um, I think we could sit here for hours longer just chatting away and like a lot of most of our guests we or all of our guests we could **[68:42]** carry on talking for ages. So, we will definitely have you on the short list for next year to come on the show. We'll catch up with you, see what Prism's up to, what new stuff is coming. >> Yeah, if you want if if you want, we're not going to kidnap you and force you to be on the show. >> You're not. I clearly am. [laughter] No, I'd love to come back anytime. >> You know who's good cop bad cop now. [laughter] >> Yeah, there we go. >> Um, so yeah, I mean we we'll we'll start wrapping thing things up, but uh Tina, thank you very much for coming on. It's **[69:15]** been one of those episodes that we've we've been trying to get in. It's not been very easy to find when we you're available to jump on and I love the fact we've done live uh with you and we've been able to get some questions in. Um, and definitely we'll get you in again next year and we'll we'll chat about all sorts. Um, >> perfect. >> So, I need to bring my script up because I will forget something otherwise. So, um, yeah. So, thank you very much. Uh, thank you very much everyone for watching or listening or however you are consuming this uh this show. Um, one **[69:47]** thing we are going to and have I got it ready? Have I got it ready? It's there. It's there. It is there. But I've got to remember how to share my screen. There it is. Ah, there we go. Speakers are starting to be announced for tech next year. Um, I probably should scroll down the page a bit. I can't see what I'm doing when I'm in this screen, but yeah. So, we have a a **[70:20]** really good selection of to uh speakers already already selected. So, if you haven't got your ticket yet, get your tickets >> and come and join us. >> There's some there's some blooming good speakers in amongst who've already been selected and there's still more to go. >> Yep. Yep. So, keep an eye on the on the the website. Um, have you got the link there to show Chris? Ready? >> You mean that one? >> This is This is This is going so smooth, **[70:51]** so professional. >> We We're just learning from John and Eric. That's all. [laughter] >> We We've got the best teachers in the world. Uh, phpech.io. Head over there. You can keep an eye on what uh who's uh who's going to be appearing. Um, also don't forget to join us in Discord. Uh, people are in there all the time. Uh, including TJ will be in there. If you if you tag him, I'm sure he'll answer. >> Yeah. >> Um, >> I I summoned him the other day. >> Yeah. If you summon me, I'll show up. [laughter] >> Yeah. Do you have to say your name **[71:21]** twice? Three times. It's three. You have to say your name three times before you show up. >> It's five times into a mirror. It's like handyman. >> Right. It's that one. [laughter] Right. There's so many different rules. I can't remember who goes by which rules. >> It's bloody Mary if you want it three times into a mirror. >> I'm I'm more preferential to to Beetlejuice. So, three will work. >> Yeah, [laughter] >> that's what I was going off of. There you go. >> So, discord.phparch.com. Come and join us for the chat. And uh we're loads of great people in there all **[71:51]** week long chatting along. Um so, do join us. Uh if you're on our YouTube channel, please uh like and subscribe. If you're not on the YouTube channel, go over there, like and subscribe, and you can see all the other episodes. And uh the other podcasts that we we do as part of the same channel. We have the main show on a Thursday evening with Eric and John. Got us, which I don't know what our schedule is anymore because it's like >> one out Thursday. I don't know who it is [laughter] yet. >> Yeah. Yeah. We'll we'll be taking place. We'll have something going out on **[72:22]** Thursday. Taking place as the main one. Uh because our fellow friends over the pond will be celebrating its Thanksgiving. Have I got that one right? >> Yep. >> Thanksgiving over there. >> It is. >> Yeah. So, we'll be we'll be uh putting something up on Thursday. Are we doing something Friday as well then? As well as Thursday. We're just doing Thursday, not Friday. We're just putting >> Not Friday. We We'll bring it forward. >> We're just doing So, we're just doing one more this week. Originally, we were going to do one every other week and now we're doing two a week. **[72:54]** >> Yeah. This is episode 17 [laughter] and we've been going for three weeks. Um so, okay. So that's our YouTube channel. Uh don't forget magazine discount for our lovely listeners. Alive 3 will give you the first three months of a annual digital subscription. We keep talking about it. I'm going to say it again. Tech PHP Tech 26. Do come and join us over in Chicago. It is a it's a lovely lovely week. Um great **[73:25]** bunch of people as we've we've tal talked about already. So I won't keep pushing but don't forget JStech. Uh, call for papers is still open till the 7th of December. TJ, 7th of December. You can learn JavaScript by then, can't you? >> I think I can probably [laughter] pull that off. >> There you go. There's There's your talk. There's your talk. How to learn JavaScript in two weeks. >> Claude, teach me JavaScript. >> Exactly. >> No, don't do that. [laughter] >> Yeah, that'll be interesting. Um, and **[73:56]** then we've got our swag sort. We have >> head over there and you can see t-shirts and uh some hats and caps and things at store at p uh sorry store.phparch.com. Head over there and grab some swag for it. And uh that is everything. We have ticked all the boxes. >> Thank you very much again TJ for joining us today. Uh thank you everyone. Thank you everyone for watching live and we