<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: Tsarp</title><link>https://news.ycombinator.com/user?id=Tsarp</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 29 Oct 2026 21:08:59 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=Tsarp" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by Tsarp in "Rewriting Prime Agent in Rust"]]></title><description><![CDATA[
<p>I know people tend to hate on rust rewrites. But having something that compiles to a single binary that you can just copy over and get started has its advantages especially when working with sandboxes etc.</p>
]]></description><pubDate>Sat, 10 Oct 2026 02:09:25 +0000</pubDate><link>https://news.ycombinator.com/item?id=50028827</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=50028827</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50028827</guid></item><item><title><![CDATA[New comment by Tsarp in "RIP, vector database"]]></title><description><![CDATA[
<p>I've really liked lancedb for similar use cases. Not just that it is OSS. But Lance treats ANN as a secondary index similar to what turbopuffer v3 does. Rows sit in fragments, and the vector index never moves them.</p>
]]></description><pubDate>Thu, 01 Oct 2026 17:41:56 +0000</pubDate><link>https://news.ycombinator.com/item?id=49924724</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=49924724</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49924724</guid></item><item><title><![CDATA[New comment by Tsarp in "Opus 5.5 is good at explainer videos"]]></title><description><![CDATA[
<p>Lot of these demos tend to squeeze in as many animations and transitions. Its usually hard to take away anything from the video at the end.<p>I wish more people optimized for learning than "how fancy can i make it look"</p>
]]></description><pubDate>Fri, 25 Sep 2026 01:57:10 +0000</pubDate><link>https://news.ycombinator.com/item?id=49839356</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=49839356</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49839356</guid></item><item><title><![CDATA[New comment by Tsarp in "Grok 4.7"]]></title><description><![CDATA[
<p>Waiting on simonw "Generate an SVG of a pelican riding a bicycle " benchmark to judge this model</p>
]]></description><pubDate>Mon, 21 Sep 2026 16:50:14 +0000</pubDate><link>https://news.ycombinator.com/item?id=49789798</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=49789798</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49789798</guid></item><item><title><![CDATA[New comment by Tsarp in "Show HN: Cactus Needle 3: 8-29MB automation models can match DeepSeek V4 Flash"]]></title><description><![CDATA[
<p>Apart from fictional use cases, what is the real use case here? The pricing on some open models are absurdly low for generic tasks. For the privacy conscious it makes sense to run something like a 8-27B on local network and get the work done.<p>Are there perhaps some industrial or agri use cases?</p>
]]></description><pubDate>Fri, 18 Sep 2026 16:57:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=49757091</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=49757091</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49757091</guid></item><item><title><![CDATA[New comment by Tsarp in "When LLM judges agree, should we believe them?"]]></title><description><![CDATA[
<p>Kinda weird to generalize "LLM". Every lab, every model is different. Has its own biases, reward functions etc.</p>
]]></description><pubDate>Mon, 14 Sep 2026 17:11:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=49700405</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=49700405</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49700405</guid></item><item><title><![CDATA[New comment by Tsarp in "Principles for Fast Tokio Applications"]]></title><description><![CDATA[
<p>One great use of agentic coding is being able to add and very granular tracing instrumentation to help with these sort of optimizations.</p>
]]></description><pubDate>Mon, 14 Sep 2026 16:04:00 +0000</pubDate><link>https://news.ycombinator.com/item?id=49699187</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=49699187</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49699187</guid></item><item><title><![CDATA[New comment by Tsarp in "How to write an effective software design document"]]></title><description><![CDATA[
<p>Its just far easier to ask an agent to do the first past ( with references to actual code ) and then start to dig in.<p>A good implementation of something like this is <a href="https://deepwiki.com/" rel="nofollow">https://deepwiki.com/</a> for OSS projects.</p>
]]></description><pubDate>Mon, 14 Sep 2026 15:53:30 +0000</pubDate><link>https://news.ycombinator.com/item?id=49699024</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=49699024</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49699024</guid></item><item><title><![CDATA[New comment by Tsarp in "How to write an effective software design document"]]></title><description><![CDATA[
<p>Lot of this is outdated. Including stuff that made a lot of sense in the past like diataxis and the google guidance.<p>I've started to build docs structured as a skill. Since every model/harness is trained to handle skills well today. The entire docs of a project or modeled as a skill.<p>I also have additional frontmatter on the md files. Specifically 2 keys -> when should you read, when you should not read this doc. This with a simple cli to help parse this has made the docs experience much quicker and nicer.<p>You can also generate svgs, mermaids etc on the fly now when you are really stuck on something.</p>
]]></description><pubDate>Mon, 14 Sep 2026 15:33:30 +0000</pubDate><link>https://news.ycombinator.com/item?id=49698707</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=49698707</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49698707</guid></item><item><title><![CDATA[New comment by Tsarp in "Claude is no longer available for minors"]]></title><description><![CDATA[
<p>Ah yes, the usual drop down where we need to confirm we are above 18.</p>
]]></description><pubDate>Fri, 11 Sep 2026 11:54:27 +0000</pubDate><link>https://news.ycombinator.com/item?id=49656899</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=49656899</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49656899</guid></item><item><title><![CDATA[New comment by Tsarp in "The Death of Shame Is Tearing Us Apart"]]></title><description><![CDATA[
<p>Might have to do with a lot of thought leaders and influencers pushing this.<p>In my own life/career I have seen that over a long enough horizon that lower shame seems to generally work out for people on the long run if you can back it up with some amount of confidence and competence.</p>
]]></description><pubDate>Fri, 11 Sep 2026 11:53:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=49656879</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=49656879</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49656879</guid></item><item><title><![CDATA[New comment by Tsarp in "YuE2 · Frontier Music with Symbolic Planning"]]></title><description><![CDATA[
<p>I wish folks spend more time on agentic music genertion instead of a single black box prompt to music.<p>Similar to coding, where you are have a bunch of tools that can do the drudge work for you (the mundane things in mixing mastering for example), but still retain full creative freedom and expression.<p>I am not articulating this well, but basically transitioning from IDE->Coding harnesses what is the DAW-> ??</p>
]]></description><pubDate>Fri, 11 Sep 2026 04:03:14 +0000</pubDate><link>https://news.ycombinator.com/item?id=49653403</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=49653403</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49653403</guid></item><item><title><![CDATA[New comment by Tsarp in "Cognition launches new SWE-2 model, Rivaling Fable 5.1 and GPT-Astra"]]></title><description><![CDATA[
<p>With the Devin subscription even at the 20$ plan, they offered unlimited SWE 1.7 usage. Wondering if they do the same for SWE 2.</p>
]]></description><pubDate>Thu, 10 Sep 2026 16:32:46 +0000</pubDate><link>https://news.ycombinator.com/item?id=49646484</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=49646484</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49646484</guid></item><item><title><![CDATA[New comment by Tsarp in "Cognition launches new SWE-2 model, Rivaling Fable 5.1 and GPT-Astra"]]></title><description><![CDATA[
<p>"SWE-2 is post-trained from Kimi K3"</p>
]]></description><pubDate>Thu, 10 Sep 2026 16:11:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=49646156</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=49646156</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49646156</guid></item><item><title><![CDATA[New comment by Tsarp in "PostgreSQL for Everything"]]></title><description><![CDATA[
<p>sqlite for everything<p>NVMe drives + Litestream + object storage(S3/R2..). sqlite simplifies things for the entire long tail of apps/services that aren't the Ubers and AirBNBs of the world.</p>
]]></description><pubDate>Wed, 19 Aug 2026 14:01:48 +0000</pubDate><link>https://news.ycombinator.com/item?id=49361755</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=49361755</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49361755</guid></item><item><title><![CDATA[New comment by Tsarp in "Open Source, Free Tier Capable Whispr Using Cloudflare AI"]]></title><description><![CDATA[
<p>whisper.cpp</p>
]]></description><pubDate>Thu, 23 Jul 2026 03:09:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=49016401</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=49016401</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49016401</guid></item><item><title><![CDATA[New comment by Tsarp in "Open Source, Free Tier Capable Whispr Using Cloudflare AI"]]></title><description><![CDATA[
<p><a href="https://carelesswhisper.app" rel="nofollow">https://carelesswhisper.app</a></p>
]]></description><pubDate>Thu, 16 Jul 2026 06:03:32 +0000</pubDate><link>https://news.ycombinator.com/item?id=48930861</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=48930861</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48930861</guid></item><item><title><![CDATA[New comment by Tsarp in "Apple's new SpeechAnalyzer API, benchmarked against Whisper and its predecessor"]]></title><description><![CDATA[
<p>Any chance you can benchmark against whisper large and large v3 turbo? These run comfortably on older Macbooks and are still far more accurate in real life dictation compared to even the parakeet models( despite ASR leaderboards) with an RTF < 1.</p>
]]></description><pubDate>Mon, 13 Jul 2026 16:55:48 +0000</pubDate><link>https://news.ycombinator.com/item?id=48895468</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=48895468</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48895468</guid></item><item><title><![CDATA[New comment by Tsarp in "Files over tools: how we built our agent with a virtual filesystem and bash"]]></title><description><![CDATA[
<p>What harness are you guys running?</p>
]]></description><pubDate>Thu, 09 Jul 2026 14:46:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=48846782</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=48846782</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48846782</guid></item><item><title><![CDATA[New comment by Tsarp in "Ask HN: Best Podcasts of 2026 [So Far]"]]></title><description><![CDATA[
<p>+1 Latent Space (and AI Engineering ). I hope it stays like it is. Minimal fluff, great guest list and very relevant.</p>
]]></description><pubDate>Wed, 08 Jul 2026 04:44:16 +0000</pubDate><link>https://news.ycombinator.com/item?id=48827629</link><dc:creator>Tsarp</dc:creator><comments>https://news.ycombinator.com/item?id=48827629</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48827629</guid></item></channel></rss>