<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Chromium on Thiago Avelino</title><link>https://avelino.run/tags/chromium/</link><description>Recent content in Chromium on Thiago Avelino</description><generator>Hugo</generator><language>en-us</language><copyright>© Avelino</copyright><lastBuildDate>Mon, 17 Aug 2026 09:51:27 -0300</lastBuildDate><atom:link href="https://avelino.run/tags/chromium/index.xml" rel="self" type="application/rss+xml"/><item><title>An engine is not a benchmark</title><link>https://avelino.run/an-engine-is-not-a-benchmark/</link><pubDate>Mon, 17 Aug 2026 00:00:00 +0000</pubDate><guid>https://avelino.run/an-engine-is-not-a-benchmark/</guid><description>&lt;p&gt;&lt;a href="https://avelino.run/taste-is-not-a-method/"&gt;The last post&lt;/a&gt; was only the decisions where something got measured and the measurement changed the drawing. This one is about the decision underneath all of them, the one taken before any of that existed, where nothing was measured because nothing could be.&lt;/p&gt;
&lt;p&gt;zer0 renders with &lt;a href="https://github.com/WebKit/WebKit"&gt;WebKit&lt;/a&gt;. There is no &lt;a href="https://www.chromium.org/Home/"&gt;Chromium&lt;/a&gt; path, not in parallel and not as a fallback.&lt;/p&gt;
&lt;p&gt;I cannot show you a benchmark where WebKit beats Blink. I never ran one, and if I had run one it would not have changed the answer, because rendering speed is not what this choice is about. An engine is not a component you evaluate on a chart and swap when a better one appears. It is the set of decisions you agree to inherit for as long as the project exists, and the only useful question is whose decisions those are and what they cost you when you disagree.&lt;/p&gt;</description></item></channel></rss>