Essentially rewrote my developer blog completely today, it's meant to be very simplistic so it wasn't awfully much. Now it's powered by Cloudflare Pages functions and D1 (I understand the risks with using D1 in production and I don't mind the risks, nonetheless my database is read-only with only 200kb data (: This is just a one day project, I am using Workers, D1, and R2 for my main hobby project, might post something about it some other day, it deserves its own post. Anyways! https://nora-soderlund.se/ Still need to update some of the old article styling and previews. 😅

Hey folks, My buddy and I just released Garph 0.5. Garph enables you to build end-to-end type-safe GraphQL APIs using TypeScript. You can deploy it with workers. Would be amazing to get your feedback on it and your support, if you like it ⭐ https://github.com/stepci/garph

Hi everyone! I've been working on a project called Mashin, a cloud-agnostic infrastructure as code tool built with Rust, TypeScript, and leveraging Deno's V8 runtime. I've recently written a blog post on how we've leveraged Cloudflare Workers, server-side rendering, and a fantastic router called

Hono

for the Mashin Registry, which plays a crucial role in the project. Hono, built by Yusuke Wada, is incredibly fast and versatile, making it an ideal choice for our project. It's been instrumental in achieving the performance we needed. 👏 The project is still in the pre-alpha stage, and as part of our commitment to building in public, we're documenting the journey and sharing our progress. Here's the blog post: https://blog.mashin.land/building-in-public-the-mashin-registry-leveraging-cloudflare-workers-and-server-side-rendering-2bad5927e113 Looking forward to hearing your thoughts! 🙏

Howdy ! I am building with my colleagues a tool to build static websites (and soon more advanced portals) based on a short questionaire (no it is not ChatGPT based, we have our components library :P). We are using Next.js and plan to integrate among others with Cloudflare Pages and Workers. For now, I invite you into grabbing an ebook about building websites. Fill in a few closed questions and leave an email address. We will be sending ebooks on 29th May. https://generativly.com/survey

inspired on whatthecommit, but for excuses. cloudflare pages with funcitons for generating random excuses for work https://desculpe.me (translation in english: forgive me) https://github.com/desculpe-me/desculpe-me-app

instead of sorry-me, i think "forgive me" would work better

thanks xD

If anyone feels the urge, here is a really basic and very early release of a wordpress plugin that uses the Access JWT to verify the user as they login to avoid re-authenticating. The site needs to be behind Access and you need a Application set up to protect /wp-admin/* Only auths existing users. https://github.com/aarhus/cloudflare-access-login

Do you have the source code of the website?

There are some really cool things possible with the new TCP sockets (https://blog.cloudflare.com/workers-tcp-socket-api-connect-databases/). One of them, is connecting to a Minecraft server ! This demo pings a Minecraft server from Pages to get the MOTD, favicon, server version, and the player count. It can also be used as an easy (and cheap!) API to get this info. It looks very ugly, but it works great! This should support most server versions, and even colours the MOTD for some servers (like Hypixel). Demo: https://mcping.pages.dev/ Git repo: https://github.com/zegevlier/mcping

Built a plugin to completely manage simple webhooks from ChatGPT, using Workers for Platforms, KV, and Wrangler 3 https://twitter.com/adam_janis/status/1659001755726741504

would love to hear what folks have already built on top of the tools that were announced / released this week 👀

Another Hacker News web app https://hacker-news.penportal.net/ HN thread: https://news.ycombinator.com/item?id=36010498 Looking for feedbacks

We have created a small project for our search purposes (small - medium datasets, to replace algolia, which is quite expensive when your users search a lot) https://github.com/TM9657/serverless-cloudflare-search Provides: - an indexing endpoint - serverless search with R2, Cache API, - Queues and Worker - low latency for small - medium datasets [we have tested a movie dataset, 20k movies] (initial 100 - 500ms, cached response 20-40 ms) - multiple parallel indices - arbitrary data (you defined the searchable objects fields and the ID name of your object) Maybe you find this useful for your own purposes (remember to leave a star if you do so :)) There are probably ways to speed this up, feel free to create a pull request, we have found this architecture to be the cheapest for our use-case at the moment.

Another project we built yesterday: https://github.com/TM9657/cloudflare-images-r2-cache Images is a really great product. But the delivery fee is a bit too high, especially for smaller frequently accessed images. We offer a R2 cache for your images. This does not always make sense (large images that are rarely accessed). In these cases you pay extra for the R2 and worker resources. But for smaller, often accessed images, like profile pictures, you save some delivery fees. The project is very easy to set up (CLI) and you can set a maximum image size for the cache. By default this is 400 KB. The more often the images are retrieved on average, the higher you can set this value. Images above this limit are stored as 1 byte in R2 and for these cases the worker forwards to an images domain (custom based on the CDN domain). Let us know what you think and if our assumptions regarding the pricing are correct for your use-cases! **Features**: - R2 Cached Image Requests - Delete endpoint (flushing variants for ID) - Awesome Repo Icon Cheers, Felix

https://npmdiff.dev/ - Powered entirely by Cloudflare Pages and Astro Hybrid Rendering. It lets you quickly and easily diff any 2 versions of an npm package, in its entirety, or file by file. What folks publish on GitHub isn't what's necessarily published to npm. Malicious packages and account compromises happen all the time. And changelogs are also just inconsistent. It's nice to quickly and easily see exactly what changed between versions. https://twitter.com/CherryJimbo/status/1660344066729467906 (CSS isn't the best still, please excuse 😅)

I made a DO that is kept alive for 60 seconds and is reusable with render api: https://github.com/advany/cloudflare-render-api-template/blob/main/src/worker.js Thanks to @celso @Unsmart | Tech debt

nice!

hi folks. I wrote an article about how to use Cloudflare Zero Trust: Access using Service Tokens. https://medium.com/@gpedro/tl-dr-cloudflare-zero-trust-service-token-aa8c83c9ac46 (PT/BR) 🇧🇷

I have created a lightweight zod-like validator of less than 1kb. It doesn't seem to be directly related to cloudflare workers, but using remix+zod with cloudflare workers, it was almost 300kb, which was a big problem under the 1MB free size plan. https://github.com/mizchi/lizod/ https://dev.to/mizchi/lizod-spiritual-successor-of-zod-less-than-1kb-4i67

thanks! Will be sharing my insights in the #1110268932788588624 channel

(the phone recording is not using the Browser Rendering API, though it is using Cloudflare Workers and Images 😉 (edit: we'll ignore that it crashed at the end there)) This is a tiny part of my hobby project but here's my use of the new Browser Rendering API for uses of my avatar service outside of the mobile app environment, the only use right now is automated tests (for mocked user avatars) but I still had this on my list to do, and was for the moment serving rendered avatars through my Node.js server, via my avatar service. Using Durable Objects to keep the Browser instance alive for at least 1 minute, allowing for 2 instances in 2 minutes if the first instance expires before that. D1 database for avatar sprite layers and color information (what avatar layers has what default color/category type, etc). No R2 bucket use yet, still considering the pros/cons about using it for the avatar service sprites. Images for storing both rendered avatars and avatar sprites. https://cdn.discordapp.com/attachments/783765338692386886/1110632756846801046/render2.mp4

I have built https://kwirk.io, a text-editor with inline AI support. Supporting all operating systems and is available as a Web App. The initial goal was to build a lightweight, privacy focused, offline first web editor and experiment with LLM. Inline AI: You select a section of your text and use AI functionality on that isolated part, which will be changed in-place. E.g Translation, Summarizing, Translation, etc. You can import and export docx, or any text format. The website is already the editor, you can use it without logging in. Future Plans I plan to integrate support for kwirk and kwirk-files for all of my other projects. Lessons learned - marketing is hard, even tho release was december last year and I was offering one of the first editors to integrate AI (before notion, etc) - I need to standardize as much of our software as possible (currently working on a global Account and Payment System across all my products, as well as some global AI endpoints) - Getting good feedback is difficult - Having a good Payment provider handling all sales tax is worth a lot, especially as a solo company - Falling in love with your products is not always a good thing Technical Details Desktop Apps are built with tauri and nextjs frontend. Mobile Apps are slightly customized PWA wrapper. Web App is also a PWA. The AI backend is mostly ChatGPT, but I also have an equally good drop in, on premise solution. Payment via Stripe and Paypal. Database: Prisma + CockroachDB. Sentry for logging errors. Infrastructed was hosted on AWS only in the beginning, but I moved some functions to workers as well as the whole DNS. Goal in the future is to move as much as possible over to cf.

I've just opened the waitlist for my upcoming ☁️ cloud service, devyl. The current landing page is hosted on Cloudflare Pages and uses Workers, KV, and Turnstile for the sign up form. The API is built on top of cheetah () and there's a plan for a generous free tier on launch! If you're interested in a serverless API for file scanning, mail checking, deploying & more you might want to **join the waitlist**: https://devyl.net

Workers for Platforms on autopilot (ChatGPT plugin) 🚀 Need a quick webhook to do something simple? Let AI write and deploy one for you! ✨ https://twitter.com/adam_janis/status/1662465298165907456

https://gptstat.us/ made a super simple dashboard for monitoring openai's api performance using cloudflare workers that ping each endpoint every 5m and save it to vercel postgres to be used by the nextjs site

I love this! Where are you pulling the product data from?

Wow, that's really amazing! Is the source code available?

This is really cool! I like it. One suggestion: Display the average rating of each season of the show, perhaps all the way to the right of the table

Is this open source?