In partnership with
Your creative brief is due Friday. Viktor wrote it Tuesday.
Tell him the campaign. Viktor pulls last quarter's performance from Meta and TikTok, scrapes competitor ads, drafts the brief, posts it for review. You edit, he ships the creative requests to your designer. Inside Slack.
What's up, everyone? This week Vercel bought your favorite auth library and Figma bought a vibe-coding team. shadcn changed its default, and OpenAI is back in the open with a new voice. Let's dive in.
You currently have {{rp_num_referrals}} referrals.
Vercel bought Better Auth. Vercel acquired Better Auth, the open source TypeScript auth library with 4.7 million weekly downloads and more than 850 contributors. Founder Bereket Engida and his team are joining, too. The library keeps its open contribution model, which is the reassurance you want when a large platform company buys a tool you depend on. The team has been building Agent Auth so each agent carries its own scoped, revocable permissions instead of acting under your full account, and that work now moves into Vercel Connect. Acquisitions of open source usually make me nervous, but Vercel has normally done a good job keeping things open.
shadcn made Base UI the default. New projects created with shadcn/ui now pick Base UI over Radix. Base UI comes from the same people who built Radix, and projects made through shadcn/create were already choosing it two to one. Radix stays fully supported, every new component still ships for both, and you can switch back with a single -b radix flag. You do not need to migrate an existing app, but if you want to, shadcn released a skill that moves one component at a time while your project stays production-ready.
Figma bought a vibe-coding team. Figma acquired the team behind Bud (formerly Orchids), a Y Combinator startup that built a platform for making apps from a prompt and later an agent product that browses the web and writes code. If you use Bud or Orchids, both shut down on July 18. You need to migrate your projects before then. Figma has spent the past year pushing past static design into building, with Figma Make and integrations with Claude Code and Codex, and this brings coding and prototyping closer to the canvas.
Want to learn more about Payload and how to apply it to your project? Check out my course, Payload Essentials. The course walks you step-by-step through a full Payload and Next.js project so you can go from zero lines of code to a production website ready to deploy.
I believe high-quality education should be as accessible as possible to everyone, no matter where in the world they live. So, purchasing power parity is built in to the price. If you don't see a banner with a discount and believe you should, please let me know!
OpenAI is back in the open, with a new voice. Two weeks after limiting GPT-5.6 to vetted partners at the government's request, OpenAI released the models publicly. Sol, Terra, and Luna are open to everyone, with Sol positioned as its strongest model yet for coding, biology, and cybersecurity. Alongside that, OpenAI launched GPT-Live, a voice model built to listen and speak at the same time so a conversation feels less like waiting your turn. When a question needs search or deeper reasoning, it hands the work to a frontier model in the background and keeps talking while that runs. Two versions are rolling out in ChatGPT now, with API access coming.
Railway put outbound networking in the CLI. The Railway team added a railway outbound-network command for static outbound IPs and outbound IPv6, both previously dashboard-only toggles. Static outbound IPs are important when something outside Railway only accepts traffic from specific IP addresses, like a payment provider or a partner's firewall, and now an engineer or an agent can read and flip that state from the terminal. Railway also moved its plugins for Claude Code, Codex, Cursor, and Grok onto a hosted MCP server, so setup is one click instead of a local config.
Dokploy can finally build in parallel. For two years the self-hosted Vercel alternative ran every build through one global queue. Version 0.29.9 gives each server its own queue with up to two concurrent builds, and that concurrency is a free, open source feature rather than an enterprise upsell. The releases since have added SCIM user provisioning for enterprise teams and a long list of security and permission fixes. If you self-host your deployments, this cuts the time you spend waiting behind your own queue.
RECENT VIDEO
The pressure to keep up with every new AI model, tool, and framework is manufactured, and here's why I finally stopped trying.
What did I miss? There's so much happening in modern web dev that I know I missed something. Please share your thoughts in the comments or reply to this email. I want to address your suggestions and may include them in future newsletters.
Thanks for reading. See you next time.