<?xml version="1.0" encoding="UTF-8"?><rss version="2.0" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>programming &amp;mdash; Erik I</title>
    <link>https://erik.itland.no/tag:programming</link>
    <description>My public writing. You can reach me at @eitland@mstdn.io </description>
    <pubDate>Tue, 21 Apr 2026 22:35:55 +0000</pubDate>
    <item>
      <title>On code quality, survivorship bias and programmer hubris</title>
      <link>https://erik.itland.no/on-code-quality-survivorship-bias-and-programmer-hubris?pk_campaign=rss-feed</link>
      <description>&lt;![CDATA[... and yes, I am a programmer &#xA;&#xA;I think there are a number of reasons why we think the code we are maintaining is often so crazy, and I’m trying to summarize it for a blog post, but for now:&#xA;&#xA;survivorship bias: a number of the teams that kept arguing about the perfect way to solve the problem never shipped, meaning no one needs to maintain their almost perfect code. The code that survived is the code that was shipped in time -for some definition of &#34;in time&#34;.&#xA;&#xA;the code that was actually flawless of course doesn’t have to be maintained.&#xA;&#xA;I think we programmers are way to full of ourselves when it comes to judging other peoples work. We can nitpick about code lines or code style in code that has run more or less flawlessly in distributed systems for years or even decades, instead of seeing the beauty of someone being able to achieve that with 80ies technology.&#xA;&#xA;Filed under #programming and #perfectionism]]&gt;</description>
      <content:encoded><![CDATA[<h2 id="and-yes-i-am-a-programmer" id="and-yes-i-am-a-programmer">... and yes, I am a programmer</h2>

<p>I think there are a number of reasons why we think the code we are maintaining is often so crazy, and I’m trying to summarize it for a blog post, but for now:</p>
<ul><li><p>survivorship bias: a number of the teams that kept arguing about the perfect way to solve the problem never shipped, meaning no one needs to maintain their almost perfect code. The code that survived is the code that was shipped in time -for some definition of “in time”.</p></li>

<li><p>the code that was actually flawless of course doesn’t have to be maintained.</p></li>

<li><p>I think we programmers are way to full of ourselves when it comes to judging other peoples work. We can nitpick about code lines or code style in code that has run more or less flawlessly in distributed systems for years or even decades, instead of seeing the beauty of someone being able to achieve that with 80ies technology.</p></li></ul>

<p>Filed under <a href="https://erik.itland.no/tag:programming" class="hashtag"><span>#</span><span class="p-category">programming</span></a> and <a href="https://erik.itland.no/tag:perfectionism" class="hashtag"><span>#</span><span class="p-category">perfectionism</span></a></p>
]]></content:encoded>
      <guid>https://erik.itland.no/on-code-quality-survivorship-bias-and-programmer-hubris</guid>
      <pubDate>Fri, 19 Jul 2019 17:49:07 +0000</pubDate>
    </item>
  </channel>
</rss>