<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Hacker News: fr3on</title><link>https://news.ycombinator.com/user?id=fr3on</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 23 Apr 2026 18:23:56 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=fr3on" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by fr3on in "Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving"]]></title><description><![CDATA[
<p>The irony of this announcement is in the name: Max-Preview is proprietary, cloud-only. The Qwen models that actually matter — the ones running on real hardware people own — are the open weights series. I run the 32B and 72B variants locally on dual A4000s. The gap between those and the hosted Max is real, but it's shrinking with every release. The interesting question isn't how Max compares to Opus. It's how long until the open-weight tier makes the cloud tier irrelevant for most workloads.</p>
]]></description><pubDate>Mon, 20 Apr 2026 20:00:38 +0000</pubDate><link>https://news.ycombinator.com/item?id=47839756</link><dc:creator>fr3on</dc:creator><comments>https://news.ycombinator.com/item?id=47839756</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47839756</guid></item><item><title><![CDATA[New comment by fr3on in "GitHub's Fake Star Economy"]]></title><description><![CDATA[
<p>Stars measure attention. Packagist downloads measure automation. Neither measures trust. The only signal that's hard to fake is: does something real depend on this?</p>
]]></description><pubDate>Mon, 20 Apr 2026 19:59:07 +0000</pubDate><link>https://news.ycombinator.com/item?id=47839734</link><dc:creator>fr3on</dc:creator><comments>https://news.ycombinator.com/item?id=47839734</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47839734</guid></item></channel></rss>