<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Reese Cooper</title>
    <description>The latest articles on DEV Community by Reese Cooper (@reese_hacks).</description>
    <link>https://dev.to/reese_hacks</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4214128%2F58a96329-d5fc-4d1c-86ef-4439fc0c6b33.jpg</url>
      <title>DEV Community: Reese Cooper</title>
      <link>https://dev.to/reese_hacks</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/reese_hacks"/>
    <language>en</language>
    <item>
      <title>One small win from caching responses</title>
      <dc:creator>Reese Cooper</dc:creator>
      <pubDate>Thu, 29 Oct 2026 02:23:22 +0000</pubDate>
      <link>https://dev.to/reese_hacks/one-small-win-from-caching-responses-22jj</link>
      <guid>https://dev.to/reese_hacks/one-small-win-from-caching-responses-22jj</guid>
      <description>&lt;p&gt;Added a five-second cache to one expensive endpoint; median latency dropped by more than half. Not a deep insight: it was slow, we cached it, it got faster.&lt;/p&gt;

&lt;p&gt;The interesting part was the failure mode. Stale data looked fine in tests and confused everyone in production once an hour. We now return the cache age in a header, which costs nothing and saved a lot of confused debugging.&lt;/p&gt;

</description>
      <category>backend</category>
      <category>performance</category>
      <category>caching</category>
    </item>
  </channel>
</rss>
