Reading up on cloud-computing
100 deep · digging since nov 19, 25
- How we rebuilt Cloudflare Workers’ module registry for Node.js compatibility
Cloudflare rebuilt the Workers module registry to use URL-based resolution, lazy compilation, shared caches, and Node.js-compatible module handling.
- Cloud in a Bottle: making self-hosting accessible to everyone
Cloud in a Bottle simplifies self‑hosting by packaging cloud services into a portable, easy‑to‑deploy solution for everyone and reduces the technical barrier to running personal infrastructure.
- What comes after git // ERSC
ERSC contends Git’s 2005 design limits modern scaling and proposes a pluggable storage engine that speaks Git protocol while enabling Jujutsu-based migration.
- 洪 民憙 (Hong Minhee): Email, decoupled from where and how: A cross-runtime, cross-provider email library for JavaScript and TypeScript
Upyo provides a runtime‑agnostic email library for JavaScript/TypeScript that uses standard web APIs to send mail via any provider through a unified Transport interface.
- My YC app: Dropbox - Throw away your USB drive
Dropbox announced upcoming AI-powered search, drafting, and summarization tools integrated with ChatGPT and Claude to improve file organization and sharing.
- Understanding the Recent DDoS Attack Against Read the Docs - Read the Docs
Read the Docs details how a June 2026 DDoS attack peaked at 5.5 M requests/min, evaded caching, and was mitigated via edge caching, targeted rate limits, and Terraform‑managed WAF rules.
- Cloud in a Bottle: making self-hosting accessible to everyone
Cloud in a Bottle provides an open-source personal cloud with containerized apps, unified authentication, and polished UX to make self-hosting as easy as using a smartphone.
- The Grammar of Data: From Definition to Execution – Xorq
The article shows how xorq’s grammar of data lets you define a data pipeline once and execute it on DuckDB, DataFusion, or Snowflake without changes.
- Have the frontier labs mixed up AI safety and security? - Martin Alderson
Frontier labs confuse AI safety with security, treating safeguards as good enough most of the time, which caused sandbox escapes and shows security must be deterministic.
- Signals and Threads
Andrew Hunter contrasts hyperscale performance tuning, where tiny changes yield big savings, with latency‑focused trading systems where profiling must capture rare, critical moments.
- Ask HN: Why were OpenAI, Claude, and Grok simultaneously down?
An Ask HN post questions the cause of a concurrent outage affecting OpenAI's services, Anthropic's Claude model, and xAI's Grok model.
- Getting Started with Resend
The guide walks developers through installing the Resend SDK, setting up API keys, and sending their first email with code examples.
- Launching Route 53 Files
Route 53 Files exposes AWS Route 53 hosted zones as an NFS file system, allowing standard Unix tools to edit DNS records with near real‑time synchronization.
- tailcat
Tailscale lets you securely connect any internet resource using WireGuard‑based, zero‑trust, end‑to‑end encrypted tunnels that are fully configurable across any infrastructure.
- Run cloud agents on machines you manage
Cursor now lets teams run its AI coding agents on self‑hosted machine pools they manage, keeping agent logic in the cloud while execution occurs on their own infrastructure.
- DSHR's Blog: Small Is Beautiful
The article argues that AI firms must generate $2 trillion by 2030 to justify their massive capex, but rising costs, cheaper Chinese open‑weight models, and efficient local inference threaten that goal.
- Compute that takes any shape - Vercel
Vercel's Fluid compute architecture dynamically assembles tailored machines for builds, sandboxes, and functions, enabling instant provisioning and efficient resource use.
- SQLite in Production with Built-In Litestream Replication – OpenRun
OpenRun integrates Litestream to provide automatic SQLite database backup and restore on Docker, Podman, and Kubernetes without requiring app changes.
- AWS Acquires DuckLabs
Amazon Web Services announced it has acquired DuckLabs, a startup that builds developer‑focused tooling for data workflows, to expand its cloud‑native analytics offerings.
- Saving 100 terabytes of memory by optimizing 1.1.1.1's DNS cache
Cloudflare cut over 100 terabytes of RAM used by its 1.1.1.1 DNS cache by redesigning cache layout and eviction policies without hurting query latency.
- PicoMQ
PicoMQ provides durable, real‑time data streams over HTTP by leveraging S3‑compatible object storage, allowing unlimited, independently addressable streams that scale from idle to high throughput.
- How we saved 100 terabytes of memory by optimizing 1.1.1.1’s DNS cache
Cloudflare reduced its Big Pineapple DNS cache entry size by 56% through five Rust optimizations, freeing ~100 TB of memory while boosting insert throughput 43% and cutting lookup latency 19%.
- Firebase for agents – OpenComputer
OpenComputer lets developers deploy TypeScript agents on isolated Linux VMs with pay‑as‑you‑go pricing, handling model keys securely and offering checkpointable sandboxes.
- Who Wins As Intelligence Commodifies? - by Rohit Krishnan
AI model commodification is eroding labs' moats, pushing them to pursue utility clouds, frontier‑oracle R&D, or conglomerate diversification to sustain profits.
- URL Source: https://saasylaunch.com/
SaasyLaunch offers a done‑for‑you service that manages product launch distribution for SaaS firms, including press release writing, outreach, and channel placement to boost visibility and user acquisition.
- The Cloudflare Blog – Brought to you by EmDash
Cloudflare moved its blog to EmDash on Cloudflare Workers, added caching, redesigned the UI, and rolled out traffic gradually to boost performance and enable agent tools.
- Skydive
Skydive provides a cloud‑based platform for creating AI agents that users can interact with via Slack, email, or iMessage, and which run on their own compute to execute tasks.
- Nvidia dramatically reduces amount of OpenAI infra financing it may guarantee
Nvidia has cut back the amount of OpenAI infrastructure financing it is willing to guarantee, signaling a pullback in its support for the AI startup's compute needs.
- PostgreSQL for Everything
The article argues that PostgreSQL can serve as a versatile backend for diverse applications, from caching to analytics, reducing the need for multiple specialized databases.
- A County Got Rich From Data Centers. Some Question ‘At What Cost?’
Loudoun County, Virginia, earns substantial tax revenue from over 250 data centers, but officials and residents warn that reliance on this income may be risky.
- Git at any scale
Cursor argues Git’s packfile‑based design hinders scalable hosting and presents Continuity, a S3‑backed write‑ahead log system that enables linear, consistent replication.
- The 6-Stage AI Infrastructure Journey: Navigating the Three FinOps & Hardware Crises with ACE Gateway — ACE Blog
The piece outlines a six‑stage AI infrastructure journey, maps three predictable cost/latency crises, and demonstrates how ACE Gateway’s control‑plane features cut spend and prevent hardware failures at each stage.
- Extensible Software in the age of LLMs
Extensible web software can combine a stable core, LLM-driven extensions, and capability-based sandboxes to let users safely create and share personalized features.
- $1 million hacker challenge for Vercel Sandbox - Vercel
Vercel launches a two‑week public HackerOne bounty offering up to $1 million for researchers who can escape its Sandbox microVM isolation.
- Git at any scale
Cursor explains why Git's packfile design hinders scalable hosting, reviews past solutions like Spokes, and introduces Continuity, a S3‑based write‑ahead log system that provides linearizable, elastically scalable Git storage.
- Cursor launches Origin code hosting platform as GitHub outage exposes opening in AI coding race
Cursor launched Origin, an AI‑native code‑hosting platform that mirrors GitHub, aiming to capture developer attention as GitHub’s reliability falters amid rising AI‑agent usage.
- Nvidia’s Risky Business – Stratechery by Ben Thompson
Nvidia is partnering with major financial firms to create $500 billion in financing platforms for AI infrastructure, transforming compute into an investable asset class and expanding systemic risk in the AI buildout.
- Posts tagged "Agents Week" — Cloudflare Blog
Cloudflare's Agents Week highlights tools like AI Search, Kitesurf browser, and MCP v2, advancing agent-first infrastructure with continuous trust evaluation and open platforms for AI-driven work and data exploration.
- Observability for every Supabase project with Grafana Cloud
Supabase introduces a one-click Grafana Cloud integration offering pre-built observability dashboards, alerts, and metrics across all plans, including free, using the Supabase Metrics API.
- Introducing Kitesurf: The agent-first browser that runs in V8 isolates on Cloudflare Workers
Cloudflare introduces Kitesurf, an agent-first browser built on Workers that prioritizes CPU and memory efficiency over human-centric features, designed to lower costs for AI agents while maintaining compatibility with automation tools via CDP.
- Your agent needs a computer, not a container — introducing @cloudflare/computer
Cloudflare launches @cloudflare/computer, an agent runtime that isolates code in lightweight workers and on-demand containers to give each agent its own scalable compute environment.
- Qwen
Qwen releases Qwen 3.8-Max, a 2.4‑trillion‑parameter model with open weights coming next week, demonstrating autonomous coding, research improvement, contest wins, and broad work automation.
- Cloudflare Workers and Hyperdrive with SvelteKit – Master.dev Blog
Integrating Cloudflare Workers and Hyperdrive with SvelteKit requires adjusting build scripts and using server hooks to provide per-request DB connections.
- Running npm in a Separate Container when using Claude
The author demonstrates using a separate temporary Docker container with shared mounts to run npm install and generate package-lock.json without polluting the host or Claude container.
- China’s free Kimi K3 AI model shakes up global tech market - Rest of World
China's free Kimi K3 AI model lets governments run top-tier AI locally, avoiding costly US cloud rentals and enabling sovereign data control.
- South Korea’s Memory Chip Giant Defies A.I. Market Jitters
South Korea's memory chip giant SK Hynix announced stratospheric quarterly sales and profits, but the results failed to ease investor anxiety over AI‑driven market volatility.
- Alphabet Quadruples Profit to $112 Billion, Fueled by A.I. Investments
Alphabet reported a profit of $112 billion, four times higher, driven by AI spending and strong cloud growth across its businesses.
- Requests for Startups
Y Combinator’s Fall 2026 Request for Startups outlines areas they want founders to tackle, ranging from AI tutors and defense tech to offshore compute and trust infrastructure.
- Durable Objects are Made for Agents
The author claims Cloudflare Durable Objects provide a cheap, single‑threaded, SQLite‑backed primitive that simplifies building AI agents, while noting concurrency and language‑support challenges.
- Inside Roblox’s Bet on World Models - ByteByteGo Newsletter
Roblox is pairing its game engine with a video world model (Super Upsampler) to deliver photorealistic, scalable multiplayer games while solving latency, consistency, and creator‑control challenges.
- Openship - Open Source, Self-Hostable Deployment Platform
Openship is an open-source, self-hostable deployment platform that automates builds, SSL, rollbacks, and multi‑service provisioning, supporting any language and offering cloud, self‑hosted, or hybrid options.
- Postgres for time-series workloads at any scale.
Tiger Cloud provides a fully managed PostgreSQL platform that independently scales compute and storage for time-series workloads, ensuring high availability, compliance, and deep observability.
- The Cloudflare Blog
Cloudflare’s July‑June 2026 blog posts detail advances in cache optimization, post‑quantum crypto, consensus research, AI tooling, monetization, and security initiatives.
- Introducing EmDash — the spiritual successor to WordPress that solves plugin security
Cloudflare launches EmDash, a serverless TypeScript CMS built on Astro that sandboxed plugins via Workers isolates to solve WordPress security issues while offering MIT licensing and x402 payments.
- Show HN: PostgreSQL performance and cost across 23 EC2 instance types
PostgreSQL performance and cost were measured across 23 EC2 instance types, revealing clear trade‑offs between query throughput, latency, and hourly pricing.
- Cloudflare gives OpenAI network signals covering 20% of the web
Cloudflare and OpenAI launched a research pilot that feeds Cloudflare's real‑time network freshness and traffic data into OpenAI's search system to speed up and improve AI web indexing.
- ZeroFS — A log-structured filesystem for S3
ZeroFS provides a log‑structured filesystem over S3‑compatible storage, exposing POSIX via NFS/9P and block devices via NBD with built‑in encryption and caching.
- Upyo | Cross-runtime email library
Upyo is a small, cross‑runtime email library for JavaScript/TypeScript that lets you write a message once and send it via SMTP or providers on Node.js, Deno, Bun, and edge functions.
- Cloudflare Workers and Hyperdrive with TanStack Start – Master.dev Blog
Using Cloudflare Workers with Hyperdrive and TanStack Start requires per-request database objects to avoid runtime errors and connection limits.
- Cloudflare Drop
Cloudflare Drop lets anyone drag a folder or ZIP to instantly preview a static site for an hour and then claim it to make the deployment permanent.
- Workflows | Render
Render's Workflows lets developers run reliable application logic and parallel tasks without handling queues, worker pools, or custom retry logic.
- Introducing Meerkat: an experiment in global consensus
Cloudflare Research introduces Meerkat, a global consensus service using the QuePaxa algorithm to provide strong consistency and high availability for control‑plane data across its 330+ data centers.
Takes
Interesting, I killed my OpenClaw VPS servers months ago But today I got a billing alert on its separate Claude account (claudeforopenclaw@), it didn't have auto reload on but I guess at some point it got hacked and the Claude key inside OpenClaw got exposed and then they waited for months before using it on Fable 5.1 I checked all my terminals and VPS sessions and none of them use API so it's interesting Anyway I nuked the Claude account, good I kept it on its own separate one
@levelsio
The free tiers I'm actually using right now: • Clerk — 50k users • Supabase — 50k MAU • PostHog — 1M events • Cloudflare Workers — 100k req/day • Resend — 3k emails (but really 100/day) • Sentry — 5k errors • RevenueCat — free until $2.5k MRR What are yours?
@pfernan95dev
It's time to go bare metal. On Hetzner, similar budget: ~$80/mo VPS → 8 shared vCPU, 16 GB shared RAM, 320 GB SSD ~$68/mo Dedicated → 6c/12t, 64 GB RAM, 1 TB NVMe Yes, I have to set everything up by myself, but that's only once. I'd rather have 4x the RAM for less money.
@DeepakNesss
Introducing Townies – a relaxing place where up to 50 players can keep a quaint, yet blossoming, town thriving. Built on Workers & Durable Objects. Watch others in the town mow grass, deliver papers, clean up trash, picnic at the park, or buy a house. https://townies.town
@BraydenWilmoth
It's killing me that @Cloudflare has invented a new set of rules called "Precursor Rules" that are separate from every other bot protection rule. To exempt an API, two rules now need to be added under both "Security rules" and "Precursor rules" Please fix @eastdakota 🥺
@iangcarroll
You can now go viral on 𝕏 without the stress of runaway CDN bills. Crucially, without sacrificing our S-tier network performance. AI is making devs and startups take off crazy fast. Flat Rate CDN brings you both scale and peace of mind.
@rauchg
53.5 million tokens per second are moving through OpenRouter’s top 20 apps right now. I scraped their public usage pages to show which models are being used https://bentossell.com/realtime-tokens
@bentossell
Wow this took off… you all should feel the same way about your database…. https://convex.ai
@waynesutton
so true, tons of innovation in building agentic workflows etc with so much willingness to pay for simple and effective APIs for services in the cloud that are still stuck behind SaaS UIs and business models (HT @zachmccullough )
@scottbelsky
I made https://infiniteslop.ai yesterday completely on my phone with @TermiusHQ on my @Hetzner_Online VPS with @claudeai Code in the sauna while in a group chat of my friends with @marcantoinefon in it
@levelsio
Git at Scale (by cursor) has been one of the most interesting blog posts i've read in a while. It came right when I was frustrated with Shopify's internal git system. As an exercise, I've implemented it over the weekend as open source. It's a single rust binary that you can point at any S3 type object store. It uses WAL and CAS primitives and requires no other data store. It also implements bundle-uri so large git repos (like our mono) are very fast to download as a chain of static bundles. Also comes with basic familiar UX.
@tobi
Very excited to announce that general sign ups are now open at code[dot]storage We built this to be the fastest, most robust code storage platform available, and they only one made specifically for you and your agent
@pierrecomputer
My SEO setup so far: • sitemap.xml submitted and cleaned up • robots.txt configured • Google Search Console + Bing Webmaster connected • Meta titles and descriptions • Canonical issues fixed • Duplicate/redirect issues cleaned up • www → non-www 301 redirect fixed • Cloudflare DNS / SSL issue resolved • Internal links cleaned up • Structured data / JSON-LD added • Page speed + Core Web Vitals optimized • llms.txt + openapi.json added • Pillar pages + content clusters • Crawl depth under 4 + thin content check • Entity optimization • Quality backlinks What else am I missing?
@TeeDevh
When someday people ask how Cloudflare got into tax prep…
@eastdakota
🤯 Holy shit… I can’t believe I’m giving this away for free. A month ago, 175K people watched me open-source Larasend Instead of slowing down, I added a full team inbox, @Cloudflare sending + receiving, and rebuilt the product around it. The biggest update yet 👇
@vijaytupakula
We bought 19 PB of flash storage for our AWS S3 exit in 2025 for about $1.5m. The list price for this setup today would be $19 MILLION 🤑. Don't know if I've ever timed a trade this well in my life. https://world.hey.com/dhh/it-s-five-grand-a-day-to-miss-our-s3-exit-b8293563
@dhh
properly installed the @cloudflare MCP yesterday and it's already paying off. was setting up Sign in With Apple for https://granite.co and there was an SPF issue and claude just fixed it immediately for me. 🧙♂️
@Shpigford
Agent Plugins are the future of Agent Skills
@GoogleCloudTech
we moved a lot of our data pipeline to cloudflare and i made a quick video showing it off yes this video is NOT ABOUT AI incredible
@thdxr
cli was a year ago. apps maybe 6 months. now it’s services, web, cloud sessions.
@steipete
Another cloud provider down. ❌ Tailscale ✅ Headscale Now my company and products run on my self-hosted, private VPN.
@SimonHoiberg
On Bending Spoons: I still think about when they bought Evernote in 2023, it had user data stored on 750 manually provisioned (!!) VMs on GCP, sharded, with ongoing performance + reliability issues, running a Java 11 monolith (!!!) Then they went ahead and fixed it in ~6 months
@GergelyOrosz
I just reserved my Cloudflare Wallet tag: http://eastdakota.cloudflare.pay. Reserve yours now at http://cloudflare.pay
@eastdakota
Read the full blog post here https://earendil.com/posts/pi-autoresearch-and-databricks/
@pidotdev
I just reserved my Cloudflare Wallet tag: http://nickgray.cloudflare.pay Reserve yours now at http://cloudflare.pay
@nickgraynews
After developing the app locally (two prompts), deploying it took one prompt: > Push this to Vercel. Use Stripe Projects to create the account. Store the state somewhere safe. Claude chose to use Upstash as its datastore, which seems to be working well.
@patrickc
Introducing Conductor Cloud! Out with worktrees, in with multiplayer cloud workspaces. Bring your subscriptions, invite your teammates, start your agents, and shut your laptop. Conduct from iPhone and API, too. http://conductor.build
@charlieholtz
using @inngest is like a little gift to myself
@clairevo
current monthly expenses for http://knockoff.co (most of which *scale up linearly* as more people use it)
@Shpigford
MCP 2026-07-28 is live and it's the largest update to the protocol since launch. MCP is now stateless, making it easier to deploy and scale remote servers. https://claude.com/blog/bringing-mcp-2026-07-28-to-claude
@ClaudeDevs
box is the cheapest full VM sandbox for agents each box gets 4vCPU, 8gb ram, >50gb storage, free disk snapshots, desktop, ssh, https hosting, devtools, docker run >1000 concurrent boxes, self-serve at only $0.00001/s/box imagine all the agents you could run! try box
@AniC_dev
for those of you that have very serious applications deployed on cloudflare, do you really manage it all with a wrangler config?
@thdxr
If you wanna do it yourself, this is how: Buy (~15 min) 1. Cheap VPS from Hetzner or DigitalOcean (~€5-10/mo), Ubuntu 24.04, tick automatic backups at checkout 2. Add your domain to Cloudflare (free plan), switch nameservers at your registrar 3. Install Termius (SSH app) + Tailscale (private network) on laptop and phone, free tiers Lock it down (~20 min) 4. Generate an SSH key in Termius, add it to the VPS at creation. Keys only, never passwords 5. SSH in once via public IP, run updates, install Tailscale on the server, log it in 6. Disable Tailscale key expiry for the server (admin console, one click) 7. Verify you can SSH via the server's Tailscale 100.x address BEFORE the next step 8. Provider firewall: delete all inbound rules, allow only port 443 from Cloudflare's published IP ranges. No public SSH at all. You enter through the tunnel 9. Test from outside: public IP times out on everything, Tailscale IP connects. Server is now invisible 10. One SSH key per device. Phone gets its own key added to authorized_keys Install the brain (~5 min) 11. apt install tmux then install Claude Code (official native installer, one curl command) 12. Run Claude Code inside a tmux session so it survives disconnects and keeps working while your laptop is closed Hand over everything else 13. Write ONE long handover prompt telling Claude Code: the server facts, the security model (so it doesn't "fix" it), folder conventions (/srv/http/domain per project, one tmux session each), your preferences, and standing rules (confirm before destructive actions, new services bind to localhost/Tailscale only) 14. Make it write all of that into CLAUDE.md first, so every future session already knows everything 15. Backups before features: nightly job pushing your data to GitHub, tested, before a single page exists 16. Then let it install the web server (Caddy + Cloudflare DNS plugin plays nicest with the locked firewall), set up SQLite, deploy the first page From then on you never administer the server again. You open Termius from anywhere, on any device, and just say what you want. BOSH
@robj3d3
I've built and sold 2 saas for $8m. now I run a saas portfolio doing $1m mrr churn has cost us more growth than anything else so I made my team a list of 12 reasons users churn. sharing them here too: 1. you sold to bad-fit customers. that churn was booked the day they signed. write down the 3 traits your best accounts share and disqualify leads who miss them, even when you need the cash 2. onboarding never hits the aha moment fast enough. find the one action that predicts retention and rebuild onboarding so every new user does it in session one 3. failed payments cause 10-20% of churn and nobody's watching. add dunning emails, card-expiry reminders, and smart retries. it's the easiest churn you'll ever win back 4. silent churn. they went quiet months before cancelling. build a health score from logins and core-feature usage and get pinged the moment it drops 5. usage isn't value, results are. define the outcome each customer came for and measure whether they hit it, not whether they logged in 6. "budget" is never the real reason. it's the polite version of "I stopped seeing value." when someone leaves on price, dig one layer down and close the value gap 7. you watch net retention and ignore gross. expansion hides a leaking base for a year. if the retention curve never flattens, fix the product before you spend another dollar on ads 8. no expansion revenue, so normal churn drags you backward. add one upsell or usage-based lever that lets good accounts grow, not just renew 9. the product stopped evolving. their needs moved, yours didn't. talk to churned power users and ship against what they outgrew 10. cancelling is all or nothing, so they pick nothing. offer a pause, a downgrade, or a heavy discount, plus a prompt asking why. save who you can, learn from who you can't. read every cancellation reason weekly and then act on it 11. no win-back motion. nearly half of returning customers come back within 30 days. fire a sequence the week after they leave and lead with something they never saw, a new feature or workflow 12. shallow usage means zero switching cost. drive adoption of a 2nd and 3rd feature until you're wired into their weekly workflow. depth is the moat hope it helps 🙏
@tibo_maker
All you really need today to build a business is 100% free open source software that charges you $0/mo + a VPS server + an API to do some required AI stuff + some R2/S3 file hosting They don't want me to tweet that because it destroys their businesses but I can't lie to you!
@levelsio
The guide to software factories
@zachlloydtweets
Today we're publicly launching Cloud Run sandboxes. Here, I start, execute, and stop 1,000 sandboxes in 5s with an average of 500ms latency:
@steren
🤯 Wow I can't believe I'm open sourcing the email platform we built internally. Self-hosted, runs on your own AWS SES. You pay @awscloud $0.10 per 1,000 emails instead of a SaaS markup, and your email data never leaves your infra. MIT licensed. Here's what it does 🧵
@vijaytupakula