<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Refactoring on MacWorks</title><link>https://macworks.dev/tags/refactoring/</link><description>Recent content in Refactoring on MacWorks</description><generator>Hugo</generator><language>en</language><atom:link href="https://macworks.dev/tags/refactoring/index.xml" rel="self" type="application/rss+xml"/><item><title>Engineer Reads</title><link>https://macworks.dev/docs/today/engineer-blogs-2026-07-30/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://macworks.dev/docs/today/engineer-blogs-2026-07-30/</guid><description>&lt;h1 id="engineering-reads--2026-07-30"&gt;Engineering Reads — 2026-07-30&lt;a class="anchor" href="#engineering-reads--2026-07-30"&gt;#&lt;/a&gt;&lt;/h1&gt;
&lt;h2 id="the-big-idea"&gt;The Big Idea&lt;a class="anchor" href="#the-big-idea"&gt;#&lt;/a&gt;&lt;/h2&gt;
&lt;p&gt;The economic incentives of clean code are shifting from long-term maintainability to immediate operational cost reduction, as decomposing large functions can quantifiably reduce LLM token consumption.&lt;/p&gt;
&lt;h2 id="deep-reads"&gt;Deep Reads&lt;a class="anchor" href="#deep-reads"&gt;#&lt;/a&gt;&lt;/h2&gt;
&lt;p&gt;&lt;strong&gt;&lt;a href="https://martinfowler.com/articles/exploring-gen-ai/refactoring-economic-benefit.html"&gt;The Economic Benefit of Refactoring&lt;/a&gt;&lt;/strong&gt; · Giles Edwards-Alexander
Historically, the ROI of refactoring has been notoriously difficult to measure, often relegated to subjective debates about maintainability and developer velocity. Giles Edwards-Alexander introduces a pragmatic shift by experimenting with code decomposition to explicitly reduce generative AI token costs. The core mechanism hinges on the idea that feeding massive, monolithic functions into LLMs incurs a heavy financial penalty. By decomposing these large functions, developers might directly observe a measurable economic benefit by supplying smaller, focused contexts to the models. While this approach risks over-optimizing code structure purely for AI context windows rather than human readability, it offers a fascinating new metric for code quality. Engineers integrating LLMs into their workflows should read this to understand how architectural decisions directly impact AI operational budgets.&lt;/p&gt;</description></item></channel></rss>