<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: cobolexpert</title><link>https://news.ycombinator.com/user?id=cobolexpert</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 23 Apr 2026 04:32:39 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=cobolexpert" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by cobolexpert in "Show HN: A context-aware permission guard for Claude Code"]]></title><description><![CDATA[
<p>Makes sense! Thanks for sharing.</p>
]]></description><pubDate>Thu, 12 Mar 2026 13:29:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=47350264</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=47350264</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47350264</guid></item><item><title><![CDATA[New comment by cobolexpert in "Show HN: A context-aware permission guard for Claude Code"]]></title><description><![CDATA[
<p>How does the classifier work? I see some JSON files with commands in them.</p>
]]></description><pubDate>Thu, 12 Mar 2026 07:39:27 +0000</pubDate><link>https://news.ycombinator.com/item?id=47347632</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=47347632</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47347632</guid></item><item><title><![CDATA[New comment by cobolexpert in "Will vibe coding end like the maker movement?"]]></title><description><![CDATA[
<p>What do you mean I'm not helping?</p>
]]></description><pubDate>Thu, 26 Feb 2026 22:57:36 +0000</pubDate><link>https://news.ycombinator.com/item?id=47173311</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=47173311</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47173311</guid></item><item><title><![CDATA[New comment by cobolexpert in "Claude Sonnet 4.6"]]></title><description><![CDATA[
<p>I like this tree visualization! The background with little squares is making the text difficult to read, though.</p>
]]></description><pubDate>Wed, 18 Feb 2026 00:35:44 +0000</pubDate><link>https://news.ycombinator.com/item?id=47055505</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=47055505</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47055505</guid></item><item><title><![CDATA[New comment by cobolexpert in "Is Show HN dead? No, but it's drowning"]]></title><description><![CDATA[
<p>Had a funny experience with this some weeks ago. I started developing a small side project and after a week I wondered if this existed already. To my surprise, someone had already built something relatively similar _with the exact same name_ (though I had chosen mine as a placeholder, still funny though) only 2 weeks before, and posted it in Show HN.<p>I took a look at the project and it was a 100k+ LoC vibe-coded repository. The project itself looked good, but it seemed quite excessive in terms of what it was solving. It made me think, I wonder if this exists because it is explicitly needed, or simply because it is so easy for it to exist?</p>
]]></description><pubDate>Tue, 17 Feb 2026 11:17:45 +0000</pubDate><link>https://news.ycombinator.com/item?id=47046237</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=47046237</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47046237</guid></item><item><title><![CDATA[New comment by cobolexpert in "I want to wash my car. The car wash is 50 meters away. Should I walk or drive?"]]></title><description><![CDATA[
<p>Changed 50 meters to 43 meters with Opus 4.6:<p>“Walk. 43 meters is basically crossing a parking lot.                ”</p>
]]></description><pubDate>Mon, 16 Feb 2026 09:13:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=47032735</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=47032735</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47032735</guid></item><item><title><![CDATA[New comment by cobolexpert in "65 Lines of Markdown, a Claude Code Sensation"]]></title><description><![CDATA[
<p>Same here, I find most of these skills/prompts a bit redundant. Some people argue that in including these in the conversation, one is doing latent space management of sorts and bringing the model closer to where one wants it to be.<p>I wonder what will happen with new LLMs that contain all of these in their training data.</p>
]]></description><pubDate>Thu, 12 Feb 2026 09:48:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=46986786</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=46986786</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46986786</guid></item><item><title><![CDATA[New comment by cobolexpert in "Ex-GitHub CEO launches a new developer platform for AI agents"]]></title><description><![CDATA[
<p>Yep exactly, a Perl script</p>
]]></description><pubDate>Tue, 10 Feb 2026 23:10:44 +0000</pubDate><link>https://news.ycombinator.com/item?id=46968350</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=46968350</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46968350</guid></item><item><title><![CDATA[New comment by cobolexpert in "Bazzite Post-Mortem"]]></title><description><![CDATA[
<p>As it slowly starts working on another platforms, it can fill in Bazzite's role (a bit ironic I guess, given Bazzite is inspired on SteamOS)</p>
]]></description><pubDate>Tue, 10 Feb 2026 21:42:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=46967327</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=46967327</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46967327</guid></item><item><title><![CDATA[New comment by cobolexpert in "Bazzite Post-Mortem"]]></title><description><![CDATA[
<p>Maybe SteamOS will help with this!</p>
]]></description><pubDate>Tue, 10 Feb 2026 17:03:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=46962975</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=46962975</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46962975</guid></item><item><title><![CDATA[New comment by cobolexpert in "Bazzite Post-Mortem"]]></title><description><![CDATA[
<p>Relevant comment from some months ago [1].<p>Personally I find that sticking to distros backed by companies or very large communities is just easier in the long term (Debian/Ubuntu/Fedora/Arch).<p>[1] <a href="https://news.ycombinator.com/item?id=46092225">https://news.ycombinator.com/item?id=46092225</a></p>
]]></description><pubDate>Tue, 10 Feb 2026 16:36:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=46962404</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=46962404</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46962404</guid></item><item><title><![CDATA[New comment by cobolexpert in "Attention at Constant Cost per Token via Symmetry-Aware Taylor Approximation"]]></title><description><![CDATA[
<p>Dumb question: is the quadratic time complexity for training, inference, or both?</p>
]]></description><pubDate>Wed, 04 Feb 2026 16:05:46 +0000</pubDate><link>https://news.ycombinator.com/item?id=46887522</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=46887522</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46887522</guid></item><item><title><![CDATA[New comment by cobolexpert in "The Codex App"]]></title><description><![CDATA[
<p>In the end this and all other 89372304 AI projects are just OpenAPI/Anthropic API wrappers, but at least one has 1st party support which maybe gives it a slight advantage?</p>
]]></description><pubDate>Mon, 02 Feb 2026 18:40:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=46859555</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=46859555</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46859555</guid></item><item><title><![CDATA[New comment by cobolexpert in "Winapp, the Windows App Development CLI"]]></title><description><![CDATA[
<p>Copilot wanting to name everything Copilot is a funny thought, for some reason.</p>
]]></description><pubDate>Thu, 29 Jan 2026 12:27:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=46809252</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=46809252</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46809252</guid></item><item><title><![CDATA[New comment by cobolexpert in "ChatGPT Containers can now run bash, pip/npm install packages and download files"]]></title><description><![CDATA[
<p>Indeed, but nowadays it’s common to add the annotations to claw back a bit of more powerful code linting.</p>
]]></description><pubDate>Tue, 27 Jan 2026 00:41:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=46773914</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=46773914</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46773914</guid></item><item><title><![CDATA[New comment by cobolexpert in "ChatGPT Containers can now run bash, pip/npm install packages and download files"]]></title><description><![CDATA[
<p>I was also thinking this some days ago. The scaffolding that static languages provide is a good fit for LLMs in general.<p>Interestingly, since we are talking about Go specifically, I never found that I was spending too much typing... types. Obviously more than with a Python script, but never at a level where I would consider it a problem. And now with newer Python projects using type annotations, the difference got smaller.</p>
]]></description><pubDate>Mon, 26 Jan 2026 21:36:11 +0000</pubDate><link>https://news.ycombinator.com/item?id=46771902</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=46771902</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46771902</guid></item><item><title><![CDATA[New comment by cobolexpert in "The browser is the sandbox"]]></title><description><![CDATA[
<p>In which way does native UI have the upper hand, do you think? To me it seems like a lot of users are largely indifferent to this aspect (e.g. so many applications nowadays being Electron/browser based). If browsers keep gaining capabilities then it seems like this gap will get even smaller.</p>
]]></description><pubDate>Mon, 26 Jan 2026 11:08:19 +0000</pubDate><link>https://news.ycombinator.com/item?id=46764230</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=46764230</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46764230</guid></item><item><title><![CDATA[New comment by cobolexpert in "Why Does Destroying Resources via TF Suck?"]]></title><description><![CDATA[
<p>At $WORK we have a Git repo set up by the devops team, where we can manage our junk by creating Terraform resources in our main AWS account.<p>The state however is always stored in a _separate AWS account_ that only the devops team can manage. I find this to be a reasonable way of working with TF. I agree that it is confusing though, because one is using $PROVIDER to both create things and manage those things at the same time, but conceptually from TF’s perspective they are very different things.</p>
]]></description><pubDate>Sat, 24 Jan 2026 20:42:00 +0000</pubDate><link>https://news.ycombinator.com/item?id=46747449</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=46747449</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46747449</guid></item><item><title><![CDATA[New comment by cobolexpert in "The Agentic AI Handbook: Production-Ready Patterns"]]></title><description><![CDATA[
<p>Maybe my initial message was overly harsh, I mostly agree with your points here. I think maybe the point of disagreement is exactly _how much_ extra prompt is necessary to approach 100% of the job, but this is quite hard to measure (obviously). Your point about latent space management is a good mental model to have IMO.</p>
]]></description><pubDate>Wed, 21 Jan 2026 14:40:16 +0000</pubDate><link>https://news.ycombinator.com/item?id=46706335</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=46706335</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46706335</guid></item><item><title><![CDATA[New comment by cobolexpert in "The Agentic AI Handbook: Production-Ready Patterns"]]></title><description><![CDATA[
<p>Ah do you mean sub-agents? I do understand that if I summon a sub-agent and give it e.g. code reviewing instructions, it will not fill up the context of the main conversation. But my point is that giving the sub-agent the instruction "review this code as if you were a staff engineer" (literally those words) should cover most use cases (but I can't prove this, unfortunately).</p>
]]></description><pubDate>Wed, 21 Jan 2026 10:35:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=46703667</link><dc:creator>cobolexpert</dc:creator><comments>https://news.ycombinator.com/item?id=46703667</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46703667</guid></item></channel></rss>