<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: rockinghigh</title><link>https://news.ycombinator.com/user?id=rockinghigh</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 23 Apr 2026 10:52:41 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=rockinghigh" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by rockinghigh in "Kimi K2.6: Advancing open-source coding"]]></title><description><![CDATA[
<p>The MoE experts are quantized to int4, all other weights like the shared expert weights are excluded from quantization and use bf16.</p>
]]></description><pubDate>Mon, 20 Apr 2026 19:23:57 +0000</pubDate><link>https://news.ycombinator.com/item?id=47839279</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=47839279</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47839279</guid></item><item><title><![CDATA[New comment by rockinghigh in "Kimi K2.6: Advancing open-source coding"]]></title><description><![CDATA[
<p>They could release data to back up that claim.</p>
]]></description><pubDate>Mon, 20 Apr 2026 19:17:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=47839204</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=47839204</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47839204</guid></item><item><title><![CDATA[New comment by rockinghigh in "Kimi K2.6: Advancing open-source coding"]]></title><description><![CDATA[
<p>Are there any protections from industrial espionage when using Anthropic, Cursor, Gemini, or OpenAI?</p>
]]></description><pubDate>Mon, 20 Apr 2026 19:16:04 +0000</pubDate><link>https://news.ycombinator.com/item?id=47839180</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=47839180</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47839180</guid></item><item><title><![CDATA[New comment by rockinghigh in "Oracle slashes 30k jobs"]]></title><description><![CDATA[
<p>Their revenue was $57.4 billion last year. Just in Q4; cloud revenue $6.7 billion, cloud infrastructure $3.0 billion, cloud application $3.7 billion, Fusion Cloud ERP $1.0 billion, NetSuite cloud ERP $1.0 billion.</p>
]]></description><pubDate>Tue, 31 Mar 2026 16:06:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=47589465</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=47589465</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47589465</guid></item><item><title><![CDATA[New comment by rockinghigh in "Leanstral: Open-source agent for trustworthy coding and formal proof engineering"]]></title><description><![CDATA[
<p>It's the number of attempts at answering the question.</p>
]]></description><pubDate>Mon, 16 Mar 2026 22:18:54 +0000</pubDate><link>https://news.ycombinator.com/item?id=47405760</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=47405760</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47405760</guid></item><item><title><![CDATA[New comment by rockinghigh in "Yann LeCun raises $1B to build AI that understands the physical world"]]></title><description><![CDATA[
<p>He founded the team that worked on fasttext, llama and other similarly impactful projects.</p>
]]></description><pubDate>Wed, 11 Mar 2026 00:24:13 +0000</pubDate><link>https://news.ycombinator.com/item?id=47330466</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=47330466</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47330466</guid></item><item><title><![CDATA[New comment by rockinghigh in "Yann LeCun raises $1B to build AI that understands the physical world"]]></title><description><![CDATA[
<p>He founded FAIR and the team in Paris that ultimately worked on the early Llama versions.</p>
]]></description><pubDate>Wed, 11 Mar 2026 00:22:41 +0000</pubDate><link>https://news.ycombinator.com/item?id=47330457</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=47330457</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47330457</guid></item><item><title><![CDATA[New comment by rockinghigh in "Tony Hoare has died"]]></title><description><![CDATA[
<p>It can also be used to simplify existing code bases.</p>
]]></description><pubDate>Tue, 10 Mar 2026 21:10:52 +0000</pubDate><link>https://news.ycombinator.com/item?id=47328826</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=47328826</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47328826</guid></item><item><title><![CDATA[New comment by rockinghigh in "Claude Sonnet 4.6"]]></title><description><![CDATA[
<p>It's a lot simpler. These models are not optimized for ambiguous riddles.</p>
]]></description><pubDate>Wed, 18 Feb 2026 06:32:47 +0000</pubDate><link>https://news.ycombinator.com/item?id=47057907</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=47057907</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47057907</guid></item><item><title><![CDATA[New comment by rockinghigh in "GLM-5: Targeting complex systems engineering and long-horizon agentic tasks"]]></title><description><![CDATA[
<p>How is this riddle relevant to a coding model?</p>
]]></description><pubDate>Wed, 11 Feb 2026 18:14:45 +0000</pubDate><link>https://news.ycombinator.com/item?id=46978611</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=46978611</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46978611</guid></item><item><title><![CDATA[New comment by rockinghigh in "GLM-5: Targeting complex systems engineering and long-horizon agentic tasks"]]></title><description><![CDATA[
<p>They do. Pretty much all agentic models call linting, compiling and testing tools as part of their flow.</p>
]]></description><pubDate>Wed, 11 Feb 2026 18:12:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=46978579</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=46978579</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46978579</guid></item><item><title><![CDATA[New comment by rockinghigh in "AI makes the easy part easier and the hard part harder"]]></title><description><![CDATA[
<p>It's called problem decomposition and agentic coding systems do some of this by themselves now: generate a plan, break the tasks into subgoals, implement first subgoal, test if it works, continue.</p>
]]></description><pubDate>Mon, 09 Feb 2026 00:55:51 +0000</pubDate><link>https://news.ycombinator.com/item?id=46940303</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=46940303</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46940303</guid></item><item><title><![CDATA[New comment by rockinghigh in "Experts Have World Models. LLMs Have Word Models"]]></title><description><![CDATA[
<p>A language model in computer science is a model that predicts the probability of a sentence or a word given a sentence. This definition predates LLMs.</p>
]]></description><pubDate>Sun, 08 Feb 2026 21:12:14 +0000</pubDate><link>https://news.ycombinator.com/item?id=46938562</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=46938562</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46938562</guid></item><item><title><![CDATA[New comment by rockinghigh in "Making geo joins faster with H3 indexes"]]></title><description><![CDATA[
<p>How do you join two datasets using r-trees? In a business setting, having a static and constant projection is critical. As long as you agree on zoom level, joining two datasets with S2 and H3 is really easy.</p>
]]></description><pubDate>Sat, 07 Feb 2026 20:43:16 +0000</pubDate><link>https://news.ycombinator.com/item?id=46927764</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=46927764</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46927764</guid></item><item><title><![CDATA[New comment by rockinghigh in "Making geo joins faster with H3 indexes"]]></title><description><![CDATA[
<p>I wouldn't say R-trees solve the problem better. Joining multiple spatial dataset indexed with r-trees is more complex as the nodes are dynamic and data dependent. Neighborhood search is also more complicated because parent nodes overlap.</p>
]]></description><pubDate>Sat, 07 Feb 2026 20:40:05 +0000</pubDate><link>https://news.ycombinator.com/item?id=46927717</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=46927717</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46927717</guid></item><item><title><![CDATA[New comment by rockinghigh in "Making geo joins faster with H3 indexes"]]></title><description><![CDATA[
<p>That's not true when tiling the Earth though. You need 12 pentagons to close the shape on every zoom level, you can't tile the Earth with just hexagons. That's also why footballs stitch together pentagons and hexagons.</p>
]]></description><pubDate>Sat, 07 Feb 2026 20:34:31 +0000</pubDate><link>https://news.ycombinator.com/item?id=46927665</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=46927665</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46927665</guid></item><item><title><![CDATA[New comment by rockinghigh in "France's homegrown open source online office suite"]]></title><description><![CDATA[
<p>Instagram uses it as their main backend. They have hundreds of million of daily users. Some of the critical backend services are in C++.</p>
]]></description><pubDate>Sat, 07 Feb 2026 15:40:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=46924699</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=46924699</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46924699</guid></item><item><title><![CDATA[New comment by rockinghigh in "Lessons learned shipping 500 units of my first hardware product"]]></title><description><![CDATA[
<p>Tariffs were a huge point of debate in his first administration. The government had to pay $30 billion to farmers to offset the impact of tariffs.<p>> China implemented retaliatory tariffs equivalent to the $34 billion tariff imposed on it by the U.S. In July 2018, the Trump administration announced it would use a Great Depression-era program, the Commodity Credit Corporation (CCC), to pay farmers up to $12 billion, increasing the transfers to farmers to $ 28 billion in May 2019. The USDA estimated that aid payments constituted more than one-third of total farm income in 2019 and 2020.</p>
]]></description><pubDate>Wed, 04 Feb 2026 02:37:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=46880733</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=46880733</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46880733</guid></item><item><title><![CDATA[New comment by rockinghigh in "Cloudflare CEO on the Italy fines"]]></title><description><![CDATA[
<p>That's definitely happening. The US does this through massive government spending on American solutions. The EU is only starting to go that route as well.</p>
]]></description><pubDate>Fri, 09 Jan 2026 18:54:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=46557571</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=46557571</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46557571</guid></item><item><title><![CDATA[New comment by rockinghigh in "iPhone Typos? It's Not Just You – The iOS Keyboard Is Broken [video]"]]></title><description><![CDATA[
<p>Completely agreed. Apple seriously regressed the multi-lingual experience. They probably have a model per language. If you have to mix languages in a sentence, well, good luck!</p>
]]></description><pubDate>Fri, 12 Dec 2025 01:23:57 +0000</pubDate><link>https://news.ycombinator.com/item?id=46239789</link><dc:creator>rockinghigh</dc:creator><comments>https://news.ycombinator.com/item?id=46239789</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46239789</guid></item></channel></rss>