1512 posts tagged with "PaaS"
Platform-as-a-Service tooling, economics, and self-hosted alternatives
Your Coding Agent Runs npm install Unattended. Refuse Is the Open-Source Gate That Says No.
Refuse is an open-source, self-hostable gate that blocks known-vulnerable package installs across 18 package managers before anything hits disk — including installs your coding agent runs. Here is how it works, where it belongs in a PaaS build pipeline, and what it cannot catch.
Render Cut Median Builds From 38s to 21s: What Matching It on a Hetzner Builder You Own Actually Costs
Render's August 2026 changelog cut median builds from 38 to 21 seconds with faster disks. Priced three ways against a $10/month Hetzner builder — Render, Railway, and Depot-metered rates — with breakeven volumes, cache strategy, and the burst caveat.
Render's Agentic Playbook, Read as a Platform Spec: What the Platform Ships vs What Stays Your Code
Render's July 2026 agentic-applications playbook draws an honest line between what a deploy platform ships — queues, retries, per-run history — and what stays in your code: idempotency, compensation, and join policy. This post reads that line as a spec for any Render-compatible API.
Render Shipped Official Python and TypeScript SDKs: SDK Parity Is Now the Render-Compatibility Bar
Render's September 10 release of official Python and TypeScript SDKs moves the Render-compatibility bar from matching REST endpoints to matching the full client surface: typed methods, per-item cursor pagination, and machine-readable error codes.
The Self-Hosted Governance Gap: Why Your Compliance Team Says No to Coolify, Dokku, CapRover, and Dokploy
Coolify, Dokku, CapRover, and Dokploy win engineering evaluations and lose security reviews: no enterprise SSO, no audit trail. A capability-by-capability look at what compliance-bound teams need and what a self-hosted PaaS must build to win them.
From One Box to a Fleet: A Second-Machine Migration Playbook for Coolify, Dokploy, and CapRover Teams Moving to Cluster API
Outgrown your single-box PaaS? A phase-by-phase playbook for moving live workloads from Coolify, Dokploy, or CapRover onto a Cluster-API-managed Hetzner fleet — with a stay-or-go checklist, data-migration options, and a weighted DNS cutover that keeps the old box as rollback.
Spotify Portal for Backstage Is GA: The Build-vs-Buy Math for Your Golden Path Just Changed
Spotify Portal for Backstage is GA: onboarding wizards, bundled premium plugins, and an AI assistant in one box. What the DIY cost numbers say, who already switched, and a build-buy-skip verdict for self-hosted platform teams.
The Month-13 Bill: What $500 in Fly.io and Render Startup Credits Really Buys
Fly.io's $500 credit buys 6-12 months and Render's buys a meaningful Pro window — then the full meter starts. A line-by-line pricing of the post-credit bill against flat-rate dedicated hardware, and how to spend credits with an exit planned.
Green Status Page, Broken Deploy: What Two July 2026 Vercel Incidents Teach About Scoped Failures in Multi-Tenant PaaS
Vercel's July 2026 status history holds two scoped failures — four hours of Hobby-only invocation errors and a six-minute telemetry gap disclosed the next day — with two more recurrences that quarter. What they share structurally, and four incident-communication guarantees a multi-tenant deploy platform should build instead.