<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Testing on RadarModel.com</title>
    <link>https://radarmodel.com/tags/testing/</link>
    <description>Recent content in Testing on RadarModel.com</description>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Sat, 17 Oct 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://radarmodel.com/tags/testing/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>How RadarModel Schema Was Tested Against Real Databases</title>
      <link>https://radarmodel.com/how-radarmodel-schema-was-tested-against-real-databases/</link>
      <pubDate>Sat, 17 Oct 2026 00:00:00 +0000</pubDate>
      <guid>https://radarmodel.com/how-radarmodel-schema-was-tested-against-real-databases/</guid>
      <description>&lt;p&gt;A drift detector earns trust slowly and loses it in one night. Report a dropped table that&amp;rsquo;s still there and the alert channel gets muted. So RadarModel Schema&amp;rsquo;s tests spend as much effort on what it must not report as on what it must.&lt;/p&gt;&#xA;&lt;p&gt;None of them use a fake database. SQLite runs on Linux, macOS and Windows. Postgres 12, 16 and 17, MySQL 8.0 and 8.4, and MariaDB 10.6 and 11.4 run in containers on every push. Each test creates its own database on the server, does real DDL to it, and drops it afterwards.&lt;/p&gt;</description>
    </item>
    <item>
      <title>How RadarModel AI Was Tested Against Broken Provider Answers</title>
      <link>https://radarmodel.com/how-radarmodel-ai-was-tested-against-broken-provider-answers/</link>
      <pubDate>Sat, 17 Oct 2026 00:00:00 +0000</pubDate>
      <guid>https://radarmodel.com/how-radarmodel-ai-was-tested-against-broken-provider-answers/</guid>
      <description>&lt;p&gt;A tool that watches for removed models has one way to be useless: telling you a model is gone when it isn&amp;rsquo;t. Do it twice and the alerts get muted, and then the real removal goes by unseen. So most of RadarModel AI&amp;rsquo;s tests aren&amp;rsquo;t about finding changes. They&amp;rsquo;re about not inventing them.&lt;/p&gt;&#xA;&lt;h2 id=&#34;the-fake-provider&#34;&gt;The fake provider&lt;/h2&gt;&#xA;&lt;p&gt;The tests run against a fake API that answers like all four providers at once: OpenAI&amp;rsquo;s list, Anthropic&amp;rsquo;s with its cursor paging, Google&amp;rsquo;s with page tokens, OpenRouter&amp;rsquo;s with prices and limits. Each test takes a baseline, then breaks the next answer from one provider in one specific way.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
