<?xml version="1.0" encoding="utf-8"?>
<feed xmlns="http://www.w3.org/2005/Atom">
  <title>Food for Bot</title>
  <link href="https://ender672.github.io/"/>
  <link href="https://ender672.github.io/feed.xml" rel="self"/>
  <updated>2026-05-26T00:00:00Z</updated>
  <id>https://ender672.github.io/</id>
  <author><name>Tim Elliott</name></author>
  <entry>
    <title>Tracing libjxl decoding patterns, with JPEG XL as the trace data container</title>
    <link href="https://ender672.github.io/posts/2026-05-26-tracing-libjxl-decoding-patterns/"/>
    <updated>2026-05-26T00:00:00Z</updated>
    <id>https://ender672.github.io/posts/2026-05-26-tracing-libjxl-decoding-patterns/</id>
    <summary>Visualizing the spatial order in which libjxl decodes strips across 1–16 threads, with the trace data itself stored in a JPEG XL container.</summary>
  </entry>
  <entry>
    <title>Packing Revisions Small and Reading them Fast</title>
    <link href="https://ender672.github.io/posts/2026-03-26-revision-storage-benchmarks/"/>
    <updated>2026-03-20T00:00:00Z</updated>
    <id>https://ender672.github.io/posts/2026-03-26-revision-storage-benchmarks/</id>
    <summary>A benchmark of revision storage strategies, delta encoders, and compressors.</summary>
  </entry>
  <entry>
    <title>An Ode to Merge Join</title>
    <link href="https://ender672.github.io/posts/2026-02-10-merge-join/"/>
    <updated>2026-02-10T00:00:00Z</updated>
    <id>https://ender672.github.io/posts/2026-02-10-merge-join/</id>
    <summary>How the merge join algorithm syncs millions of rows in constant memory using two sorted inputs and 25 lines of Python.</summary>
  </entry>
</feed>
