<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: _andrei_</title><link>https://news.ycombinator.com/user?id=_andrei_</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Wed, 15 Apr 2026 02:18:21 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=_andrei_" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by _andrei_ in "Show HN: Ismcpdead.com – Live dashboard tracking MCP adoption and sentiment"]]></title><description><![CDATA[
<p>What's the purpose of this? There is no replacement for MCP.
We need a protocol for calling tools that works with structured outputs, this is what we have.</p>
]]></description><pubDate>Fri, 03 Apr 2026 21:44:57 +0000</pubDate><link>https://news.ycombinator.com/item?id=47632732</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=47632732</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47632732</guid></item><item><title><![CDATA[New comment by _andrei_ in "Why are executives enamored with AI, but ICs aren't?"]]></title><description><![CDATA[
<p>you're a slop maker</p>
]]></description><pubDate>Sat, 28 Mar 2026 22:41:34 +0000</pubDate><link>https://news.ycombinator.com/item?id=47558724</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=47558724</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47558724</guid></item><item><title><![CDATA[New comment by _andrei_ in "If DSPy is so great, why isn't anyone using it?"]]></title><description><![CDATA[
<p>Almost all the points are not about what DSPy is mainly supposed to offer.
What's supposedly great at is automatic optimization, for everything else... who the hell puts Python in production just to make some API calls?
There are "frameworks" available in all the better languages, but the constructs behind are not that complicated. And why does DSPy even try to compete with LangChain/Graph/crap?</p>
]]></description><pubDate>Mon, 23 Mar 2026 16:48:41 +0000</pubDate><link>https://news.ycombinator.com/item?id=47491972</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=47491972</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47491972</guid></item><item><title><![CDATA[New comment by _andrei_ in "How to code Claude Code in 200 lines of code"]]></title><description><![CDATA[
<p>yes, it's an agent</p>
]]></description><pubDate>Thu, 08 Jan 2026 20:48:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=46546234</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=46546234</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46546234</guid></item><item><title><![CDATA[New comment by _andrei_ in "Anthropic acquires Bun"]]></title><description><![CDATA[
<p>All vendors will have to implement test time code execution, solution exploration, etc. as it's a low hanging fruit with huge gains, so I see it as a great hire.
Love Bun, happy for you guys!</p>
]]></description><pubDate>Wed, 03 Dec 2025 00:34:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=46128851</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=46128851</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46128851</guid></item><item><title><![CDATA[New comment by _andrei_ in "Show HN: A live release of available domain names our generator found"]]></title><description><![CDATA[
<p>Hi Rob, thanks for your thoughts!
We're not planning on expanding to watching expiring domains yet, we're focused more on finding domains that were never registered.
But what you're interested in can be done by watching all the registered domains that you're interested in and either catching them when they expire (ex. <a href="https://www.expireddomains.net" rel="nofollow">https://www.expireddomains.net</a>) or by registering a backorder (higher success rate but you have to pay, ex. <a href="https://www.gname.com/backorder" rel="nofollow">https://www.gname.com/backorder</a>)</p>
]]></description><pubDate>Fri, 07 Nov 2025 14:52:50 +0000</pubDate><link>https://news.ycombinator.com/item?id=45847003</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=45847003</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45847003</guid></item><item><title><![CDATA[New comment by _andrei_ in "Claude Is Down"]]></title><description><![CDATA[
<p>what are we gonna dooo?</p>
]]></description><pubDate>Fri, 07 Nov 2025 14:34:39 +0000</pubDate><link>https://news.ycombinator.com/item?id=45846813</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=45846813</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45846813</guid></item><item><title><![CDATA[Show HN: A live release of available domain names our generator found]]></title><description><![CDATA[
<p>Hi HN!<p>We've been having fun building a cool domain name generator and we'd like you to try it out.<p>We're also releasing a bunch of available domain names ideas that our product generated in what we call "freebie sessions".<p>The link above points to our first freebie session, where every few seconds a batch of available domain names will be released.<p>What makes SneakyDomains different?<p>- Shows only available domains (non-premium)
- Real-time availability checking, no cached/stale data
- Credit-based pricing (no subscriptions, no expiry)<p>We're giving out 500 credits (1 credit / generated domain) for the first 1000 users.
Register an account and use this coupon code to redeem your 500 credits: HN500<p>Looking forward to your feedback!</p>
<hr>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=45846598">https://news.ycombinator.com/item?id=45846598</a></p>
<p>Points: 5</p>
<p># Comments: 3</p>
]]></description><pubDate>Fri, 07 Nov 2025 14:14:10 +0000</pubDate><link>https://sneakydomains.com/freebies/qltkukgfy8ykton</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=45846598</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45846598</guid></item><item><title><![CDATA[New comment by _andrei_ in "Tell HN: Azure outage"]]></title><description><![CDATA[
<p><a href="https://www.reddit.com/r/cscareerquestions/comments/1ojbebq/just_pushed_my_first_pr_for_my_new_job_at_azure/" rel="nofollow">https://www.reddit.com/r/cscareerquestions/comments/1ojbebq/...</a></p>
]]></description><pubDate>Wed, 29 Oct 2025 19:11:15 +0000</pubDate><link>https://news.ycombinator.com/item?id=45751589</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=45751589</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45751589</guid></item><item><title><![CDATA[New comment by _andrei_ in "Sampling and structured outputs in LLMs"]]></title><description><![CDATA[
<p>pydantic is a _validation_ library, it does not do any kind of constraints by itself</p>
]]></description><pubDate>Tue, 23 Sep 2025 17:57:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=45350572</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=45350572</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45350572</guid></item><item><title><![CDATA[New comment by _andrei_ in "Type checking is a symptom, not a solution"]]></title><description><![CDATA[
<p>> we need to recognize that the felt necessity of sophisticated type systems is a signal that we’re using the wrong architectural foundations<p>I'd be forever ashamed if I wrote that.</p>
]]></description><pubDate>Fri, 05 Sep 2025 19:08:39 +0000</pubDate><link>https://news.ycombinator.com/item?id=45142395</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=45142395</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45142395</guid></item><item><title><![CDATA[New comment by _andrei_ in "Claude Code Checkpoints"]]></title><description><![CDATA[
<p>this is coming to Claude Code</p>
]]></description><pubDate>Thu, 28 Aug 2025 13:32:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=45051947</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=45051947</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45051947</guid></item><item><title><![CDATA[New comment by _andrei_ in "How I use Claude Code to implement new features in an existing complex codebase"]]></title><description><![CDATA[
<p>> So my argument is that removing the context that is entirely irrelevant for the agent improves performance dramatically.<p>100% agree on building the optimal context, just have not seen parallel agents do better at sequential tasks. the documentation agent may have better initial context about how to write documentation, but it doesn't have the context of the changes, apart from what is passed to it / can explore. if we don't spawn a new session, and instead throw it a /document command - that would still get all the guidelines and rules for writing documentation, and it should have the same weight since it's at the bottom of the context.<p>for me the highest 'model jump level' performance booster is externalizing context and controlling the process - having claude initialize a plan file with a pre-defined template that makes it use that as its to do list and documentation place, and getting it to use that as its primary working area</p>
]]></description><pubDate>Sun, 03 Aug 2025 09:14:05 +0000</pubDate><link>https://news.ycombinator.com/item?id=44775254</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=44775254</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44775254</guid></item><item><title><![CDATA[New comment by _andrei_ in "How I use Claude Code to implement new features in an existing complex codebase"]]></title><description><![CDATA[
<p>> If you don't discuss hooks and subagents, i'm not sure what you're doing right now.<p>Agents have their own context and can be useful for tasks that can be parallelized, which is a minority of tasks. How are they critical to better performance for you?</p>
]]></description><pubDate>Sun, 03 Aug 2025 06:51:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=44774627</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=44774627</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44774627</guid></item><item><title><![CDATA[New comment by _andrei_ in "Deep Agents"]]></title><description><![CDATA[
<p>ah, deep agents = agents with planning + agents as tools => so regular agents.<p>i hate how LangChain has always tried to make things that are simple seem very complicated, and all the unnecessary new terminology and concepts they've pushed,  but whatever sells LangSmith.</p>
]]></description><pubDate>Fri, 01 Aug 2025 21:56:16 +0000</pubDate><link>https://news.ycombinator.com/item?id=44762899</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=44762899</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44762899</guid></item><item><title><![CDATA[New comment by _andrei_ in "Chatterbox TTS"]]></title><description><![CDATA[
<p>very cherry picked</p>
]]></description><pubDate>Thu, 12 Jun 2025 12:52:41 +0000</pubDate><link>https://news.ycombinator.com/item?id=44257201</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=44257201</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44257201</guid></item><item><title><![CDATA[New comment by _andrei_ in "MCP – Flash in the Pan or Future Standard?"]]></title><description><![CDATA[
<p>Funny of this to come out from LangChain.</p>
]]></description><pubDate>Sat, 22 Mar 2025 20:40:30 +0000</pubDate><link>https://news.ycombinator.com/item?id=43448445</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=43448445</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43448445</guid></item><item><title><![CDATA[New comment by _andrei_ in "Show HN: "Git who" – A new CLI tool for industrial-scale Git blaming"]]></title><description><![CDATA[
<p>go install github.com/sinclairtarget/git-who@latest</p>
]]></description><pubDate>Wed, 19 Mar 2025 18:42:00 +0000</pubDate><link>https://news.ycombinator.com/item?id=43415791</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=43415791</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43415791</guid></item><item><title><![CDATA[New comment by _andrei_ in "Character Prefix Conditioning"]]></title><description><![CDATA[
<p>Do they have the solution (I'd assume Supermaven which they recently acquired does) and are they just giving out an interesting challenge to readers?
Or... what is this?</p>
]]></description><pubDate>Sun, 12 Jan 2025 22:33:55 +0000</pubDate><link>https://news.ycombinator.com/item?id=42677655</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=42677655</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42677655</guid></item><item><title><![CDATA[New comment by _andrei_ in "Show HN: Made a small JavaScript benchmarking app – BenchJS"]]></title><description><![CDATA[
<p>Clinic.js was promising but it's not maintained anymore, would be great if we had a modern alternative: <a href="https://clinicjs.org/" rel="nofollow">https://clinicjs.org/</a></p>
]]></description><pubDate>Fri, 03 Jan 2025 15:23:43 +0000</pubDate><link>https://news.ycombinator.com/item?id=42586398</link><dc:creator>_andrei_</dc:creator><comments>https://news.ycombinator.com/item?id=42586398</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42586398</guid></item></channel></rss>