<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0"><channel><description>Backend engineer — Python, FastAPI, PostgreSQL.&#xA;CS at BITS Pilani, working toward AI systems.&#xA;I post what I&#39;m building, what broke, and what I learned fixing it.&#xA;&#xA;subhambhattacharya.com</description><link>https://bsky.app/profile/subham-hq.bsky.social</link><title>@subham-hq.bsky.social - Subham</title><item><link>https://bsky.app/profile/subham-hq.bsky.social/post/3mslmic3mu22g</link><description>An error status is not an error.&#xA;&#xA;Building a link checker, the instinct is raise_for_status(). But a 404 is the thing I&#39;m looking for.&#xA;Only transport failures raise — DNS, refused, timeout. Those mean the request never happened. A 404 means it happened perfectly and the answer is bad news.&#xA;&#xA;#python</description><pubDate>08 Aug 2026 17:40 +0000</pubDate><guid isPermaLink="false">at://did:plc:utglnjohluybfqkryehbgp2d/app.bsky.feed.post/3mslmic3mu22g</guid></item><item><link>https://bsky.app/profile/subham-hq.bsky.social/post/3mslmct4kwk2g</link><description>The most expensive bugs don&#39;t raise.&#xA;&#xA;urlsplit(&#34;example.com&#34;).hostname returns None. So my crawler&#39;s same-host check rejected every link, finished cleanly, exit 0, and reported on zero pages.&#xA;Nothing to catch. Nothing in the logs. Just a confident answer that was wrong.&#xA;&#xA;#python #backend #asyncio</description><pubDate>08 Aug 2026 17:37 +0000</pubDate><guid isPermaLink="false">at://did:plc:utglnjohluybfqkryehbgp2d/app.bsky.feed.post/3mslmct4kwk2g</guid></item></channel></rss>