<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: Zerot</title><link>https://news.ycombinator.com/user?id=Zerot</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Fri, 28 Aug 2026 14:05:58 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=Zerot" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by Zerot in "Writing a good Claude.md"]]></title><description><![CDATA[
<p>I would recommend using it, yeah. You have limited context and it will be compacted/summarized occasionally. The compaction/summary will lose some information and it is easy for it to forget certain instructions you gave it. Afaik claude.md will be loaded into the context on every compaction which allows you to use it for instructions that should always be included in the context.</p>
]]></description><pubDate>Mon, 01 Dec 2025 02:09:21 +0000</pubDate><link>https://news.ycombinator.com/item?id=46102689</link><dc:creator>Zerot</dc:creator><comments>https://news.ycombinator.com/item?id=46102689</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46102689</guid></item><item><title><![CDATA[New comment by Zerot in "Writing a good Claude.md"]]></title><description><![CDATA[
<p>> Why is that good advice? If that thing is eventually supposed to do the most tricky coding tasks, and already a year ago could have won a medal at the informatics olympics, then why wouldn't it eventually be able to tell if I'm using 2 or 4 spaces and format my code accordingly? Either it's going to change the world, then this is a trivial task, or it's all vaporware, then what are we even discussing..<p>This is the exact reason for the advice: The LLM already is able to follow coding conventions by just looking at the surrounding code which was already included in the context. So by adding your coding conventions to the claude.md, you are just using more context for no gain.<p>And another reason to not use an agent for linting/formatting(i.e. prompting to "format this code for me") is that dedicated linters/formatters are faster and only take maybe a single cent of electricity to run whereas using an LLM to do that job will cost multiple dollars if not more.</p>
]]></description><pubDate>Mon, 01 Dec 2025 01:59:31 +0000</pubDate><link>https://news.ycombinator.com/item?id=46102617</link><dc:creator>Zerot</dc:creator><comments>https://news.ycombinator.com/item?id=46102617</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46102617</guid></item><item><title><![CDATA[New comment by Zerot in "Internet's biggest annoyance: Cookie laws should target browsers, not websites"]]></title><description><![CDATA[
<p>Sorry, but no. Those functionalities fall under "functional cookies" and as such do not require consent. Also, there is no tracking needed for the dark mode at all. And "logging in" does not mean "tracking"</p>
]]></description><pubDate>Thu, 23 Oct 2025 04:52:16 +0000</pubDate><link>https://news.ycombinator.com/item?id=45678262</link><dc:creator>Zerot</dc:creator><comments>https://news.ycombinator.com/item?id=45678262</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45678262</guid></item><item><title><![CDATA[New comment by Zerot in "Android's sideloading limits are its most anti-consumer move"]]></title><description><![CDATA[
<p>No. You can still install apks through ADB, which is how you would do it during development. But you won't be able to distribute it without signing it through google.</p>
]]></description><pubDate>Tue, 14 Oct 2025 03:13:10 +0000</pubDate><link>https://news.ycombinator.com/item?id=45575872</link><dc:creator>Zerot</dc:creator><comments>https://news.ycombinator.com/item?id=45575872</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45575872</guid></item><item><title><![CDATA[New comment by Zerot in "Want to piss off your IT department? Are the links not malicious looking enough?"]]></title><description><![CDATA[
<p>Seems that the url validation is broken. It says that `<a href="http://test.example" rel="nofollow">http://test.example</a>` is not a valid url</p>
]]></description><pubDate>Fri, 19 Sep 2025 00:41:38 +0000</pubDate><link>https://news.ycombinator.com/item?id=45296709</link><dc:creator>Zerot</dc:creator><comments>https://news.ycombinator.com/item?id=45296709</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45296709</guid></item><item><title><![CDATA[New comment by Zerot in "Some thoughts on LLMs and software development"]]></title><description><![CDATA[
<p>I think in part the issue is that the LLM does not have enough context. The difference between a bug in the test or a bug in the implementation is purely based on the requirements which are often not in the source code and stored somewhere else(ticket system, documentation platform).<p>Without providing the actual feature requirements to the LLM(or the developer) it is impossible to determine which is wrong.<p>Which is why I think it is also sort of stupid by having the LLM generate tests by just giving it access to the implementation. That is at best testing the implementation as it is, but tests should be based on the requirements.</p>
]]></description><pubDate>Fri, 29 Aug 2025 04:40:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=45060274</link><dc:creator>Zerot</dc:creator><comments>https://news.ycombinator.com/item?id=45060274</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45060274</guid></item><item><title><![CDATA[New comment by Zerot in "AccountingBench: Evaluating LLMs on real long-horizon business tasks"]]></title><description><![CDATA[
<p>None. That is the point. 9.9 can be both bigger or smaller than 9.11 depending on context</p>
]]></description><pubDate>Thu, 24 Jul 2025 05:12:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=44667166</link><dc:creator>Zerot</dc:creator><comments>https://news.ycombinator.com/item?id=44667166</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44667166</guid></item><item><title><![CDATA[New comment by Zerot in "AccountingBench: Evaluating LLMs on real long-horizon business tasks"]]></title><description><![CDATA[
<p>Which is the bigger version number? Version 9.9 or version 9.11?
Which is the bigger dollar amount? $9.9 or $9.11?<p>Periods are not always used for the decimal separator but also as a separator for multiple sets of semi-independent numbers.</p>
]]></description><pubDate>Mon, 21 Jul 2025 23:48:07 +0000</pubDate><link>https://news.ycombinator.com/item?id=44641695</link><dc:creator>Zerot</dc:creator><comments>https://news.ycombinator.com/item?id=44641695</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44641695</guid></item><item><title><![CDATA[New comment by Zerot in "Why is F# code robust and reliable?"]]></title><description><![CDATA[
<p>Closures are possible, yeah. But F# also has partial application(and currying). So you don't need to use a closure to do this.</p>
]]></description><pubDate>Tue, 24 Sep 2024 07:03:57 +0000</pubDate><link>https://news.ycombinator.com/item?id=41633834</link><dc:creator>Zerot</dc:creator><comments>https://news.ycombinator.com/item?id=41633834</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41633834</guid></item><item><title><![CDATA[New comment by Zerot in "We fine-tuned Llama 405B on AMD GPUs"]]></title><description><![CDATA[
<p>They aren't. This company is using obsidian publish to publish documents.</p>
]]></description><pubDate>Tue, 24 Sep 2024 06:13:22 +0000</pubDate><link>https://news.ycombinator.com/item?id=41633555</link><dc:creator>Zerot</dc:creator><comments>https://news.ycombinator.com/item?id=41633555</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41633555</guid></item></channel></rss>