<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Inspect on BitSpeed.org</title>
    <link>https://bitspeed.org/tags/inspect/</link>
    <description>Recent content in Inspect on BitSpeed.org</description>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Tue, 20 Oct 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://bitspeed.org/tags/inspect/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>BitSpeed Inspect Explains HTTP Transfers and Compares Saved Runs</title>
      <link>https://bitspeed.org/bitspeed-inspect-explains-http-transfers-and-compares-saved-runs/</link>
      <pubDate>Tue, 20 Oct 2026 00:00:00 +0000</pubDate>
      <guid>https://bitspeed.org/bitspeed-inspect-explains-http-transfers-and-compares-saved-runs/</guid>
      <description>&lt;p&gt;A slow request gives you a duration. The next question is where that duration came from, and whether the same interval changed after you adjusted something. BitSpeed Inspect records the client events behind one HTTP transfer and turns them into an explanation you can revisit.&lt;/p&gt;&#xA;&lt;p&gt;Inspect 0.1.1 is available as an independent Go command. It saves JSON, explains one recording and compares two sets of samples. Transfer is a separate tool under development, so there is no need to wait for a resumable-transfer engine to use Inspect.&lt;/p&gt;</description>
    </item>
    <item>
      <title>How BitSpeed Inspect Was Tested</title>
      <link>https://bitspeed.org/how-bitspeed-inspect-was-tested/</link>
      <pubDate>Tue, 20 Oct 2026 00:00:00 +0000</pubDate>
      <guid>https://bitspeed.org/how-bitspeed-inspect-was-tested/</guid>
      <description>&lt;p&gt;Inspect&amp;rsquo;s tests challenge its explanations as well as its ability to complete a request. Controlled local endpoints delay headers, stream bodies, redirect, return empty responses, disconnect early and wait for cancellation. A report that explains only a successful request would miss a large part of the job.&lt;/p&gt;&#xA;&lt;p&gt;The test suite checks timing arithmetic, missing events, partial observations, source evidence links and privacy boundaries. Comparison tests cover changed payloads and encodings, mixed failures, zero baselines, insufficient coverage and overlapping ranges. The conditions and outcomes are kept in &lt;a href=&#34;https://github.com/softdevstudio/bitspeed/tree/main/test/results&#34;&gt;recorded results&lt;/a&gt;, rather than being turned into a generic speed claim.&lt;/p&gt;</description>
    </item>
    <item>
      <title>HTTP Transfer Timing Needs More Than a Stopwatch</title>
      <link>https://bitspeed.org/http-transfer-timing-needs-more-than-a-stopwatch/</link>
      <pubDate>Tue, 20 Oct 2026 00:00:00 +0000</pubDate>
      <guid>https://bitspeed.org/http-transfer-timing-needs-more-than-a-stopwatch/</guid>
      <description>&lt;p&gt;Two HTTP requests can take the same total time for very different reasons. One may spend most of its observed interval establishing a connection. Another may receive its first response byte quickly and then spend longer reading the body. A stopwatch hides that distinction.&lt;/p&gt;&#xA;&lt;p&gt;Inspect records trace events and keeps the interval boundaries explicit. Some events can be missing or ambiguous. Concurrent connection attempts also make a neat additive waterfall less honest than it looks. The report leaves unavailable values unavailable instead of filling them with zeroes.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
