<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Schema on API Coding</title>
    <link>https://apicoding.com/tags/schema/</link>
    <description>Recent content in Schema on API Coding</description>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Mon, 05 Oct 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://apicoding.com/tags/schema/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Infer Your API&#39;s Real Contract From Traffic, Then Diff It Against the Docs</title>
      <link>https://apicoding.com/infer-your-apis-real-contract-from-traffic-then-diff-it-against-the-docs/</link>
      <pubDate>Mon, 05 Oct 2026 00:00:00 +0000</pubDate>
      <guid>https://apicoding.com/infer-your-apis-real-contract-from-traffic-then-diff-it-against-the-docs/</guid>
      <description>&lt;p&gt;Your OpenAPI file marks &lt;code&gt;shipping_address&lt;/code&gt; as required on &lt;code&gt;GET /orders/{id}&lt;/code&gt;. After Tuesday&amp;rsquo;s deploy, about one response in 300 leaves it out, because a new code path for orders created by the import job skips the field. The docs still say required. The tests still pass, since nobody wrote one for import-job orders. A client with a strict deserializer starts throwing on 0.3% of order pages, and the first report you get is &amp;ldquo;sometimes the order screen is blank&amp;rdquo;.&lt;/p&gt;</description>
    </item>
    <item>
      <title>Audit an API Field&#39;s Completeness Before You Build a Feature on It</title>
      <link>https://apicoding.com/audit-an-api-fields-completeness-before-you-build-a-feature-on-it/</link>
      <pubDate>Mon, 21 Sep 2026 00:00:00 +0000</pubDate>
      <guid>https://apicoding.com/audit-an-api-fields-completeness-before-you-build-a-feature-on-it/</guid>
      <description>&lt;p&gt;An API&amp;rsquo;s documentation lists the fields a record can have. It rarely tells you how many records actually have them. Those are different numbers, and on open data APIs the gap is often enormous.&lt;/p&gt;&#xA;&lt;p&gt;The failure looks like this. You read the schema, spot a field that would make a great feature, build the feature, and then discover the field is populated on four percent of the data. By then the feature is written and the pitch has been made.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
