|
Hi there!
Welcome to the 200th freek.dev newsletter! 🥳
Last week, on X, I announced the next Spatie product I'm working on: There There. It's a helpdesk where AI surfaces context and drafts replies, so your team responds faster and every user gets a thoughtful answer.
"There There" is an expression used to calm, comfort, and soothe someone who is feeling bad. It's also the title of a beautiful Radiohead song. After seeing the band perform in Berlin at the end of last year, I had the thought of using that song title as the product name. We also considered the name "Hello, There", but I felt that "There There" has a bit more personality to it.
We're already using There There internally for a couple of weeks, polishing it a lot, so even early access users will have a great experience. You can apply for early access now.
Creating and publishing a big project like There There always feels like putting a bit of myself, my personality out there. I'm not afraid to say that it's a bit frightening and exciting at the same time. I'm sure those of you who publish stuff publicly can relate.
That being said, here are a couple of links I hope you'll enjoy as much as I did.
Prove that you're human
In this post, Justin Jackson reviews the There There homepage. Key takeaway: to hint to your readers that you're not putting out AI slop, add your name / story to your marketing.
⭐ Instant view switches with Inertia v3 prefetching
There There uses Inertia v3's features to provide a very fast UI.
⭐ How we use Inertia v3 optimistic updates in There There
Another cool Inertia v3 feature being used at our upcoming AI powered helpdesk.
Oh Dear monitors your entire website, so you don't have to (sponsored post)
My service can monitor uptime, SSL certificates, broken links, scheduled tasks, DNS, and much more! You get instant notifications when something breaks, a developer-friendly API, and a public status page you can set up in under a minute.
Building Multi-Agent Workflows with the Laravel AI SDK
The Laravel blog walks through how to implement the five multi-agent patterns from Anthropic's "Building Effective Agents" research using the Laravel AI SDK. Prompt chaining, parallelization, routing, orchestrator-workers, and evaluator-optimizer loops, all built with just the agent() helper.
Designing with Claude Code
Steve Schoger shows how he uses Claude Code to design and build UIs, turning natural language prompts into polished interfaces.
Under the Hood: How Blaze Speeds Up Blade Templates
A deep dive into how Blaze works internally. Matt Stauffer builds two toy versions from scratch to show how Blaze shifts Blade component rendering from runtime to compile time.
Two Soups, Two Cookies
A nice analogy about software craft. Same ingredients, same features, but the invisible process behind the decisions is what separates "this works" from "this feels right."
Quantization from the ground up
A thorough explainer on how quantization makes LLMs 4x smaller and 2x faster while losing only 5-10% accuracy. Covers floating point precision, compression techniques, and how to measure quality loss, with interactive examples throughout.
Overcoming AI anxiety
Ryan Chandler shares his honest journey from unease to acceptance with AI coding tools. A thoughtful reflection on how your value as a software engineer is not in writing every line, but knowing which lines should exist at all.
My Take on Vibe Coding VS Agentic Engineering
My thoughts on why Agentic Engineering is a better path than Vibe Coding, and the workflow I use to turn AI agents into a structured engineering process.
SlideWire: Build Presentations with Livewire and Blade
SlideWire is a Laravel package for building browser-based presentation decks using Livewire components and Blade templates. It comes with built-in navigation, transitions, syntax highlighting, and Mermaid diagrams.
Context Anchoring
Martin Fowler explores why AI coding sessions degrade over time and how externalizing decisions into structured documents keeps context reliable across sessions.
Agent responsibly
Vercel shares their internal framework for shipping agent-generated code safely. The core argument: green CI is no longer proof of safety, because agents produce code that looks flawless while remaining blind to production realities. The post outlines how to build systems where agents can act with high autonomy because deployment is safe by default.
Community links
In this section you'll find links submitted by others. Let me know if you wrote or stumbled across a blog post, tutorial or video that might be interesting to appear in this section.
Using the Screen Capture API to share screen regions on video calls (submitted by Paul Conroy)
Under the Hood: How Blaze Speeds Up Blade Templates (submitted by Kayla Helmick)
Old posts
Here are a couple of links from a while ago!
Implementing Multitenancy in Laravel
Stripe Is My DNS Provider Now: When Good APIs Meet Bad Ideas
A Sneaky Phish Just Grabbed my Mailchimp Mailing List
Goodbye reCAPTCHA, hello Turnstile
How to Write Blog Posts that Developers Read
A cookieless, cache-friendly image proxy in Laravel (inspired by Cloudflare)
Using PHP as a (Terrible) Video Player
Enjoy this newsletter? Help it grow!
If you found this edition useful, here are two quick ways to support the newsletter:
👉 Write a short recommendation. Approved recommendations appear on the newsletter page for others to see.
👉 Forward this email to a colleague or friend who might be interested.
Alternatively, you could consider picking up one of the paid products my team and I have worked on:
If you have any questions, remarks or thoughts about this newsletter, simply hit reply!
Thank you so much for reading!
Freek
P.S. Know someone who'd enjoy this newsletter? Send them here to subscribe.
You are receiving this newsletter because you subscribed at freek.dev
Unsubscribe from this newsletter
This mail was sent using Mailcoach
|