<?xml version="1.0" encoding="UTF-8"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Mergify Blog</title><description>Stories, insights, and opinions from the team behind Mergify</description><link>https://mergify.com/</link><atom:link href="https://mergify.com/blog/rss.xml" rel="self" type="application/rss+xml"/><item><title>pytest-xdist makes the suite faster and the flakes weirder</title><link>https://mergify.com/blog/pytest-xdist-ordering-races/</link><guid isPermaLink="false">mergify-blog:pytest-xdist-ordering-races</guid><description>Why a test that always passes alone fails on `pytest -n auto`, the fixture-scope rule that prevents most worker races, and the worker_id pattern for genuinely shared resources.</description><pubDate>Sat, 09 May 2026 00:00:00 GMT</pubDate><category>testing</category><author>noreply@mergify.com (Rémy Duthu)</author></item><item><title>RSpec randomized order is the messenger, not the bug</title><link>https://mergify.com/blog/rspec-order-dependent-specs-random-seed/</link><guid isPermaLink="false">mergify-blog:rspec-order-dependent-specs-random-seed</guid><description>Why a green spec on `--order defined` fails on `--order random` with a different seed every CI run, how to bisect to the dependent pair, and the cleanup hooks that fix the underlying coupling.</description><pubDate>Thu, 07 May 2026 00:00:00 GMT</pubDate><category>testing</category><author>noreply@mergify.com (Rémy Duthu)</author></item><item><title>I&apos;m the new hire on a codebase I didn&apos;t write</title><link>https://mergify.com/blog/new-hire-on-a-codebase-i-didnt-write/</link><guid isPermaLink="false">mergify-blog:new-hire-on-a-codebase-i-didnt-write</guid><description>A new engineer at Mergify built a daily quiz, grounded in real test assertions, to learn the codebase he ships into. He scored three out of ten on it, and it&apos;s the cheapest learning loop he has found for the era when AI types most of the code.</description><pubDate>Wed, 06 May 2026 00:00:00 GMT</pubDate><category>ai</category><author>noreply@mergify.com (Thomas Berdy)</author></item><item><title>Playwright auto-wait is great, until your component re-renders mid-action</title><link>https://mergify.com/blog/playwright-auto-wait-element-rerenders/</link><guid isPermaLink="false">mergify-blog:playwright-auto-wait-element-rerenders</guid><description>Why Playwright&apos;s actionability check can fire on a stale element, the React state-update pattern that triggers it, and the locator strategy that survives the re-render.</description><pubDate>Tue, 05 May 2026 00:00:00 GMT</pubDate><category>testing</category><author>noreply@mergify.com (Rémy Duthu)</author></item><item><title>Vitest&apos;s `threads` pool is fast. It is also why your suite leaks state.</title><link>https://mergify.com/blog/vitest-thread-pool-state-leakage/</link><guid isPermaLink="false">mergify-blog:vitest-thread-pool-state-leakage</guid><description>Why module singletons survive across tests inside a Vitest worker thread, what the failure looks like, and the two-line config change that gives you per-file isolation.</description><pubDate>Sun, 03 May 2026 00:00:00 GMT</pubDate><category>testing</category><author>noreply@mergify.com (Rémy Duthu)</author></item><item><title>pytest fixture teardown races: when yield-style cleanup eats itself</title><link>https://mergify.com/blog/pytest-fixture-teardown-races/</link><guid isPermaLink="false">mergify-blog:pytest-fixture-teardown-races</guid><description>Why a session-scoped fixture&apos;s teardown can pull state out from under tests still using it, and the per-fixture lifecycle discipline that fixes the leak for good.</description><pubDate>Fri, 01 May 2026 00:00:00 GMT</pubDate><category>testing</category><author>noreply@mergify.com (Rémy Duthu)</author></item><item><title>What GitHub Webhook Latency Actually Looks Like</title><link>https://mergify.com/blog/what-github-webhook-latency-actually-looks-like/</link><guid isPermaLink="false">mergify-blog:what-github-webhook-latency-actually-looks-like</guid><description>We instrumented GitHub webhook delivery latency for ourselves. The p95 stays under 60 seconds in steady state but climbs close to 40 minutes during check-run incidents.</description><pubDate>Wed, 29 Apr 2026 00:00:00 GMT</pubDate><category>deep-dives</category><author>noreply@mergify.com (Mehdi Abaakouk)</author></item><item><title>How to Cut Your GitHub Actions CI Bill (Without Compromising Tests)</title><link>https://mergify.com/blog/cut-your-github-actions-ci-bill/</link><guid isPermaLink="false">mergify-blog:cut-your-github-actions-ci-bill</guid><description>Five concrete levers that move the GitHub Actions bill: runner sizing, caching, test selection, batching, and two-step CI. The math, the tradeoffs, and how to pick the two that actually fit your codebase.</description><pubDate>Sun, 26 Apr 2026 00:00:00 GMT</pubDate><category>merge-queue-ci</category><author>noreply@mergify.com (Julien Danjou)</author></item><item><title>How to Enable GitHub Merge Queue with GitHub Actions: The 5-Minute Setup</title><link>https://mergify.com/blog/enable-github-merge-queue-actions-setup/</link><guid isPermaLink="false">mergify-blog:enable-github-merge-queue-actions-setup</guid><description>Enabling GitHub&apos;s merge queue is a five-line YAML change plus one repository setting. Here&apos;s the working config, the four things that will break the first week, and how to know when you&apos;ve outgrown it.</description><pubDate>Sun, 26 Apr 2026 00:00:00 GMT</pubDate><category>merge-queue-ci</category><author>noreply@mergify.com (Julien Danjou)</author></item><item><title>GitHub Auto-Merge: When the Native Button Is Enough, and When You Outgrow It</title><link>https://mergify.com/blog/github-auto-merge-when-native-is-enough/</link><guid isPermaLink="false">mergify-blog:github-auto-merge-when-native-is-enough</guid><description>GitHub has had auto-merge since 2021. The native button handles the simple case well. Here&apos;s an honest breakdown of what it does, where it stops being enough, and what to reach for when it does.</description><pubDate>Sun, 26 Apr 2026 00:00:00 GMT</pubDate><category>merge-queue-ci</category><author>noreply@mergify.com (Julien Danjou)</author></item></channel></rss>