<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: Zambyte</title><link>https://news.ycombinator.com/user?id=Zambyte</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Tue, 30 Jun 2026 22:28:42 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=Zambyte" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by Zambyte in "Nano Banana 2 Lite"]]></title><description><![CDATA[
<p>I read it as "Google is indifferent to me" rather than "Google specifically dislikes me", as a point of clarification that they didn't get access by being personally selected.</p>
]]></description><pubDate>Tue, 30 Jun 2026 18:55:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=48737568</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48737568</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48737568</guid></item><item><title><![CDATA[New comment by Zambyte in "Anthropic has embedded hidden spyware-like code in Claude Code"]]></title><description><![CDATA[
<p>Copying data without consent is only cool when Americans do it</p>
]]></description><pubDate>Tue, 30 Jun 2026 18:42:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=48737362</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48737362</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48737362</guid></item><item><title><![CDATA[New comment by Zambyte in "Zig – SPIR-V Backend Progress"]]></title><description><![CDATA[
<p>> To an untrained eye, it seems like go, but with manual allocation.<p>I've written a decent amount of Zig, Go, and C, and I often describe Zig as in between Go and C. So your untrained eye seems pretty accurate. Maybe it's just not the tool you need, and that's fine! I could get excited about a new tractor that is well built and leans heavily into right to repair activism, but I live in a city and wouldn't use it.</p>
]]></description><pubDate>Tue, 30 Jun 2026 11:49:41 +0000</pubDate><link>https://news.ycombinator.com/item?id=48731335</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48731335</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48731335</guid></item><item><title><![CDATA[New comment by Zambyte in "Zig – SPIR-V Backend Progress"]]></title><description><![CDATA[
<p><a href="https://ziglang.org/devlog/2025/#2025-06-08" rel="nofollow">https://ziglang.org/devlog/2025/#2025-06-08</a></p>
]]></description><pubDate>Tue, 30 Jun 2026 11:45:10 +0000</pubDate><link>https://news.ycombinator.com/item?id=48731288</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48731288</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48731288</guid></item><item><title><![CDATA[New comment by Zambyte in "Qwen 3.6 27B is the sweet spot for local development"]]></title><description><![CDATA[
<p>Can you link the model? I also have a 24gb card (7900 XTX). I've been having success with the dense 27b model, but I'd like to see if the 35b iq4 is any better.</p>
]]></description><pubDate>Mon, 29 Jun 2026 19:47:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=48724195</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48724195</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48724195</guid></item><item><title><![CDATA[New comment by Zambyte in "Qwen 3.6 27B is the sweet spot for local development"]]></title><description><![CDATA[
<p>I have been using pi (and previously the codex cli) with Qwen 3.6 27b with 100k context for my development at work, and I have been very blown away by how well it works. It's not perfect, but it's enough to accelerate my normal development flow. I mostly use it for writing Go and C#.</p>
]]></description><pubDate>Mon, 29 Jun 2026 19:22:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=48723871</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48723871</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48723871</guid></item><item><title><![CDATA[New comment by Zambyte in "The KIDS Act would require age checks to get online"]]></title><description><![CDATA[
<p>Parents would have a better chance at fighting back. Both with the time gained to communicate their disapproval, and a better relationship with their kids to be more likely to disapprove of it.</p>
]]></description><pubDate>Sun, 28 Jun 2026 14:22:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=48707540</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48707540</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48707540</guid></item><item><title><![CDATA[New comment by Zambyte in "Google limits Meta's use of its Gemini AI models"]]></title><description><![CDATA[
<p>Facebook does seem to be falling behind. Does anyone here use Llama over more recent options for any technical reasons?</p>
]]></description><pubDate>Sun, 28 Jun 2026 14:10:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=48707415</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48707415</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48707415</guid></item><item><title><![CDATA[New comment by Zambyte in "The KIDS Act would require age checks to get online"]]></title><description><![CDATA[
<p>Or we could have UBI and parents could parent their kids without corporations babysitting for them.</p>
]]></description><pubDate>Sun, 28 Jun 2026 14:05:32 +0000</pubDate><link>https://news.ycombinator.com/item?id=48707356</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48707356</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48707356</guid></item><item><title><![CDATA[New comment by Zambyte in "Evan's Jujutsu Tutorial"]]></title><description><![CDATA[
<p>Thanks for this link. I've been using jj for 2 years now. Whenever people ask me why, it's hard to explain. I'll have to keep this in my pocket for the future.</p>
]]></description><pubDate>Sat, 27 Jun 2026 13:07:47 +0000</pubDate><link>https://news.ycombinator.com/item?id=48697928</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48697928</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48697928</guid></item><item><title><![CDATA[New comment by Zambyte in "Previewing GPT‑5.6 Sol: a next-generation model"]]></title><description><![CDATA[
<p>Other people's computers famously can't be taken away.</p>
]]></description><pubDate>Fri, 26 Jun 2026 22:26:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=48692795</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48692795</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48692795</guid></item><item><title><![CDATA[New comment by Zambyte in "Previewing GPT‑5.6 Sol: a next-generation model"]]></title><description><![CDATA[
<p>Yes. The difference is obviously that full, fat Linux runs on a superset of anything a layperson would call a computer, and can be built from source on roughly the same set of hardware. Running the full, fat Deepseek (as in the 1.6T model, unquantized) is too big to run on anything a layperson would call a computer, and being able to actually build it is even harder.</p>
]]></description><pubDate>Fri, 26 Jun 2026 19:50:15 +0000</pubDate><link>https://news.ycombinator.com/item?id=48691164</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48691164</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48691164</guid></item><item><title><![CDATA[Ask HN: What are your favorite CLIs to use as LLM tools?]]></title><description><![CDATA[
<p>I've recently been getting into building out skills for agentic coding. The two main ones I have built out are using the jira CLI[0] and the gitlab CLI[1]. Both work well for what I want to use them for (interacting with issues and with MRs).<p>What have been the CLIs that you have found to be the most useful to use with agents? Have you found any that an agent has been able to compose in interesting ways?<p>[0] https://github.com/ankitpokhrel/jira-cli<p>[1] https://gitlab.com/gitlab-org/cli</p>
<hr>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=48672379">https://news.ycombinator.com/item?id=48672379</a></p>
<p>Points: 3</p>
<p># Comments: 0</p>
]]></description><pubDate>Thu, 25 Jun 2026 12:29:41 +0000</pubDate><link>https://news.ycombinator.com/item?id=48672379</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48672379</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48672379</guid></item><item><title><![CDATA[New comment by Zambyte in "Ask HN: Why don't LLM harnesses enable/expose custom middleware hooks?"]]></title><description><![CDATA[
<p><a href="https://pi.dev" rel="nofollow">https://pi.dev</a> specifically prides itself on having many extensions hook points. Is that one you've looked at? If so, it might be something they'd be willing to add a hook for if you open an issue.</p>
]]></description><pubDate>Thu, 25 Jun 2026 12:05:49 +0000</pubDate><link>https://news.ycombinator.com/item?id=48672179</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48672179</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48672179</guid></item><item><title><![CDATA[New comment by Zambyte in "Show HN: Nub – A Bun-like all-in-one toolkit for Node.js"]]></title><description><![CDATA[
<p>They'll get bought out by OpenAI and convert the project to Zig</p>
]]></description><pubDate>Wed, 24 Jun 2026 15:59:22 +0000</pubDate><link>https://news.ycombinator.com/item?id=48661861</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48661861</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48661861</guid></item><item><title><![CDATA[New comment by Zambyte in "Polymarket has flooded social media with deceptive videos by paid creators"]]></title><description><![CDATA[
<p>This is like a food commercial where someone shows an extreme reaction to food that isn't the product they're selling, or acting like they're eating something that isn't food at all. There are laws regarding food advertising that require it to present the real food (though the presentation of it is usually way better).</p>
]]></description><pubDate>Tue, 23 Jun 2026 02:20:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=48639381</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48639381</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48639381</guid></item><item><title><![CDATA[New comment by Zambyte in "Excessive nil pointer checks in Go"]]></title><description><![CDATA[
<p>Zig 0.16.0 was the big, recent async update. Go channels are handled as std.Io.Queue, the select keyword is handled with std.Io.Select, and goroutines are handled with std.Io.async and std.Io.concurrent. The difference between async and concurrent calls are essentially: async calls <i>may</i> be called concurrently, or they may not (the execution of the async function is not dependent on the caller continuing to execute - like reading a set of files), and concurrent <i>must</i> execute concurrently (the execution of the concurrent function depends on the caller continuing to execute - like a client+server interaction, and will crash if the environment cannot spawn concurrent processes).<p>The fully userspace implementation is a bit more syntactically clunky than the concurrency primitives in Go, but it is very similar semantically.</p>
]]></description><pubDate>Mon, 22 Jun 2026 21:57:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=48636824</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48636824</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48636824</guid></item><item><title><![CDATA[New comment by Zambyte in "Canyon HUD helmet for road riding"]]></title><description><![CDATA[
<p>Well, a wall will. It might even dissuade them from trying.</p>
]]></description><pubDate>Mon, 22 Jun 2026 21:41:00 +0000</pubDate><link>https://news.ycombinator.com/item?id=48636598</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48636598</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48636598</guid></item><item><title><![CDATA[New comment by Zambyte in "Good results fine tuning a local LLM like Qwen 3:0.6B to categorize questions"]]></title><description><![CDATA[
<p>While a fun poc, surely it would be better to just use the API (see the footer)? Or just `curl | x2j | jq` and map the HTML directly to JSON?</p>
]]></description><pubDate>Mon, 22 Jun 2026 16:46:28 +0000</pubDate><link>https://news.ycombinator.com/item?id=48632622</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48632622</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48632622</guid></item><item><title><![CDATA[New comment by Zambyte in "Excessive nil pointer checks in Go"]]></title><description><![CDATA[
<p>I've written a couple of fairly large projects in Zig, and never really felt like I needed anything more than what comptime provides. What kind of pattern do you think Zig would benefit from supporting?</p>
]]></description><pubDate>Mon, 22 Jun 2026 09:04:34 +0000</pubDate><link>https://news.ycombinator.com/item?id=48627666</link><dc:creator>Zambyte</dc:creator><comments>https://news.ycombinator.com/item?id=48627666</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48627666</guid></item></channel></rss>