🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Branching data sounds cool but creates more problems than it solves. For common use cases (eg. testing schema migrations, validating new code versions) it's overkill and it results in significant privacy/compliance risk. Schema branching with controlled test data gives you

We let down our customers at Cloudflare today. Our Workers KV service failed and the downstream products that rely on that service had outages of their own. We will publish a full postmortem soon. I know that these kinds of incidents have real and serious impact for teams

Don’t constrain the metaverse

Zawinski's Law for databases: "Every database attempts to expand until it can store all types of data. Those databases which cannot so expand are replaced by ones which can."

cloudflare is really doing to AI what they did to internet: taking something complex and making it just work everywhere and for everyone

I'm biased, but if you're not building on Cloudflare Developers you're doing yourself and your team a disservice

This is such a brilliant blog, I think I've read it 4 times now. The system behind DO's with SQLite is just brilliant. blog.cloudflare.com/sqlite-in-dura…

in our weekly @cloudflare Workers team demo session where Matt is showing off how he builds (complex!) Grafana dashboards on top of our internal analytics/metrics schemas using Claude Code. saving literal hours of writing queries panel by panel.

Cloudflare Workers are incredibly powerful, easy to operate, and cheap. But not everything can be written in JS or compiled to WASM. Until now that meant running some applications in another cloud and bridging to CF, but no more!

Huge launch! Excited to get your feedback.

run arbitrary AI-generated code on demand on Cloudflare Code sandboxes this is just the beginning

Another perfect example of Cloudflare reducing a whole AWS service to `fetch`. Load balancing requests across containers with a simple rand() and fetch() 🤯

It's honestly impressive how well positioned Cloudflare is for the agentic future. This stuff just makes so much sense for where we're going.

Surprised at the +1s here. Auth is hard, bugs can be extinction events, enterprise customers demand roadmaps that eat engineering bandwidth at small startups. There's clearly a company size where it makes sense to build in-house, but outsource until then.

An open Internet depends on a model that incentivizes great content to continue to be published. Proud to partner with some of the best publishers and AI services to launch a new model built that rewards high-quality material in a new AI era.

I may use vlookup more than any other human on the planet 👴

Use marimo notebooks to rapidly explore Cloudflare data using Python and SQL. We've partnered with Cloudflare Developers to integrate with Cloudflare authentication and to run marimo on Cloudflare Containers. Get started with examples for R2, workers AI, D1, and more.

R2 is becoming the obvious place to store your data lake, with low per-byte costs, no egress fees, and a managed Iceberg catalog. Amazing to see companies like marimo realizing that and building first-class support for R2!

Starlink on long flights is an absolute game changer.

Directionally correct, but overly simplistic. Think about latency as experienced by your customers, not just internal app latency. E.g. if customers are far away and data is cacheable, you may want app servers near customers, not data. Ideal infra gives maximum flexibility.