<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: bellowsgulch</title><link>https://news.ycombinator.com/user?id=bellowsgulch</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 29 Oct 2026 18:22:24 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=bellowsgulch" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by bellowsgulch in "Everyone has ADHD because we made it up"]]></title><description><![CDATA[
<p>After watching for 20 or more years an explosion of the diagnosis, I’m not convinced it’s real because we can’t measure it in any <i>meaningful</i> way.<p>> Everyone gets distracted, forgets things, procrastinates, and loses their keys; what matters clinically is the severity, persistence, combination, and impairment associated with those experiences. Everyone pisses, too. But if you’re doing it 45 times a day, I’d probably go see a doctor.<p>OR. Here’s a controversial idea. We have a cultural problem in that if you have an issue it MUST be health related instead of you not having matured enough yet or built the experiences required to handle a situation.<p>Everyone has a disorder? Or do people just need to learn skills that genuinely take time to develop as a part of a life that requires a lot from you?</p>
]]></description><pubDate>Mon, 28 Sep 2026 12:28:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=49876936</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49876936</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49876936</guid></item><item><title><![CDATA[New comment by bellowsgulch in "MiMo v2.6"]]></title><description><![CDATA[
<p>Yes, an expensive American LLM has zero capabilities as far as I’m concerned because I’m never going to pay for it.</p>
]]></description><pubDate>Mon, 21 Sep 2026 22:07:38 +0000</pubDate><link>https://news.ycombinator.com/item?id=49794101</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49794101</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49794101</guid></item><item><title><![CDATA[New comment by bellowsgulch in "MiMo v2.6"]]></title><description><![CDATA[
<p>That’s not the argument.</p>
]]></description><pubDate>Mon, 21 Sep 2026 22:05:24 +0000</pubDate><link>https://news.ycombinator.com/item?id=49794075</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49794075</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49794075</guid></item><item><title><![CDATA[New comment by bellowsgulch in "OpenArm: An open-source 7DOF humanoid arm"]]></title><description><![CDATA[
<p>Because we're already designing machines with reinforcement learning. It would be two steps forward and one step back to design robots who can walk but can't fold laundry. Why would anyone want a third machine next to their washer and dryer if you're already going to have a household robot?<p>God, not to mention how home are designed around the two-machine space.</p>
]]></description><pubDate>Wed, 16 Sep 2026 17:49:04 +0000</pubDate><link>https://news.ycombinator.com/item?id=49730511</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49730511</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49730511</guid></item><item><title><![CDATA[New comment by bellowsgulch in "Show HN: Swift-Qwen3.8-27B, -58.3% thinking, x1.95 speed, accuracy of xhigh"]]></title><description><![CDATA[
<p>I so wanted to be able to use Qwen3.8-27B locally on my fully loaded M1 Max 64GB when spider-mario recommended I use MTPLX[1], but I still found the general MLX MTP acceleration to be so poor, I'd rather just use cheap tokens from OpenCode Zen and Go.<p>It's just too slow of a model. I know, I know there's new hardware, but my business paid like 4-5 grand for this MBP at the time, and I just don't feel the need to pay 7-8 grand to step up to current hardware when token spend is what it is.<p>[1]: <a href="https://news.ycombinator.com/item?id=49611128#49612229">https://news.ycombinator.com/item?id=49611128#49612229</a></p>
]]></description><pubDate>Wed, 16 Sep 2026 17:17:16 +0000</pubDate><link>https://news.ycombinator.com/item?id=49730090</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49730090</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49730090</guid></item><item><title><![CDATA[New comment by bellowsgulch in "Linux from Scratch"]]></title><description><![CDATA[
<p>LFS is really, really bad. Also, reading the Linux kernel documentation to understand how to do it right is infuriating because it too is really, really bad.</p>
]]></description><pubDate>Tue, 15 Sep 2026 06:46:32 +0000</pubDate><link>https://news.ycombinator.com/item?id=49708658</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49708658</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49708658</guid></item><item><title><![CDATA[New comment by bellowsgulch in "Nike exits the S&P 100 after 18 years and a $200B market-cap wipeout"]]></title><description><![CDATA[
<p>Jesus this, and the shit I do want, they price at like $300. Fuck! I guess I'll just never give them a dollar.</p>
]]></description><pubDate>Mon, 14 Sep 2026 23:31:05 +0000</pubDate><link>https://news.ycombinator.com/item?id=49705680</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49705680</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49705680</guid></item><item><title><![CDATA[New comment by bellowsgulch in "So you want to use OpenRouter?"]]></title><description><![CDATA[
<p>Absolutely no where is that shown to customers, and OpenRouter is going to be in a world of shit when people learn this by HN post and the business asks themselves why its reputation keeps taking hits.<p>What a fucking ridiculous thing for me to learn by social news.</p>
]]></description><pubDate>Sat, 12 Sep 2026 17:11:00 +0000</pubDate><link>https://news.ycombinator.com/item?id=49674561</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49674561</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49674561</guid></item><item><title><![CDATA[New comment by bellowsgulch in "Ask HN: What default model do you use and why?"]]></title><description><![CDATA[
<p>opencode zen</p>
]]></description><pubDate>Sat, 12 Sep 2026 16:18:00 +0000</pubDate><link>https://news.ycombinator.com/item?id=49673955</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49673955</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49673955</guid></item><item><title><![CDATA[New comment by bellowsgulch in "Ask HN: What default model do you use and why?"]]></title><description><![CDATA[
<p>mimo-v2.5-free, mimo-v2.5, deepseek-flash in that order, honestly don’t even bother using qwen3.6-35b-a3b now unless i need uncensored tasks finished, mostly reverse engineering<p>most engineering tasks don’t require frontier llms<p>when they get stuck, then i consider moving up to more capable models<p>purchasing a claude plan seems widely unnecessary to me<p>the tasks they do better than the average engineer cut both ways: unless you have an existing portfolio of well written and designed work done pre-llms, it looks like you’re producing slop that pretends to be well designed<p>poor typography choices despite using the mode,<p>poor layout choices despite using popular CSS frameworks<p>etc<p>bad engineers will always be bad engineers<p>tools don’t make up for it<p>edit: a follow up to this— everyone is using eyebrows in their layouts and have no fucking clue why it was done to begin with<p>everyone has a status pill floating above their front page hero display text and its not fucking status related<p>so gross</p>
]]></description><pubDate>Sat, 12 Sep 2026 15:19:28 +0000</pubDate><link>https://news.ycombinator.com/item?id=49673218</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49673218</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49673218</guid></item><item><title><![CDATA[New comment by bellowsgulch in "DeepSeek v4.1 Flash"]]></title><description><![CDATA[
<p>Yeah, unfortunately. But you get an additional $5.</p>
]]></description><pubDate>Thu, 10 Sep 2026 19:16:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=49648918</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49648918</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49648918</guid></item><item><title><![CDATA[New comment by bellowsgulch in "DeepSeek v4.1 Flash"]]></title><description><![CDATA[
<p>OpenCode Go referral code, if you want to try it. <a href="https://opencode.ai/go?ref=QDJQMTGP5Q">https://opencode.ai/go?ref=QDJQMTGP5Q</a></p>
]]></description><pubDate>Thu, 10 Sep 2026 16:06:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=49646075</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49646075</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49646075</guid></item><item><title><![CDATA[New comment by bellowsgulch in "Automattic's board forces CEO Matt Mullenweg into leave of absence"]]></title><description><![CDATA[
<p>Good, get him the fuck away from Wordpress. Though frankly, a better solution might be for businesses to ask themselves whether or not they need Wordpress anymore.<p>In the mean time, if you're unable to derisk Mullenweg out of your business stack, this is a good first step.</p>
]]></description><pubDate>Wed, 09 Sep 2026 21:50:42 +0000</pubDate><link>https://news.ycombinator.com/item?id=49634921</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49634921</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49634921</guid></item><item><title><![CDATA[New comment by bellowsgulch in "iPhone Duo"]]></title><description><![CDATA[
<p>I want it, but for $2,000, I want to be able to do so much more.</p>
]]></description><pubDate>Wed, 09 Sep 2026 18:20:13 +0000</pubDate><link>https://news.ycombinator.com/item?id=49631069</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49631069</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49631069</guid></item><item><title><![CDATA[New comment by bellowsgulch in "Navier-Stokes – Tristan Buckmaster [pdf]"]]></title><description><![CDATA[
<p>Tell us his name.</p>
]]></description><pubDate>Tue, 08 Sep 2026 22:44:08 +0000</pubDate><link>https://news.ycombinator.com/item?id=49618183</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49618183</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49618183</guid></item><item><title><![CDATA[New comment by bellowsgulch in "C*: Unifying Programming and Verification in C (2025)"]]></title><description><![CDATA[
<p>I suspect that there is a body of C programmers out here that use C on a regular basis that aren’t posting who know that using C safely is 99% just using Valgrind and some specific GCC or llvm flags and you’re done.<p>They don’t reply to these threads to share that knowledge because these threads devolve into cesspools about how you should just use Rust despite the language not meeting specific requirements authors have that you have to do additional work and maintenance for in “safe” languages that you get for free in C because you don’t have to emulate it.</p>
]]></description><pubDate>Tue, 08 Sep 2026 17:14:43 +0000</pubDate><link>https://news.ycombinator.com/item?id=49613283</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49613283</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49613283</guid></item><item><title><![CDATA[New comment by bellowsgulch in "Benchmarking Qwen3.8 27B quantizations: 4-bit holds up, 1-bit collapses"]]></title><description><![CDATA[
<p>Thanks, man! I’ll go use that now that I know. llama-server the last time I used it for inference with this model wasn’t able to produce work fast enough to reach those numbers.</p>
]]></description><pubDate>Tue, 08 Sep 2026 17:10:27 +0000</pubDate><link>https://news.ycombinator.com/item?id=49613222</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49613222</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49613222</guid></item><item><title><![CDATA[New comment by bellowsgulch in "Ask HN: Are others seeing Google's reCAPTCHA rejecting Firefox users?"]]></title><description><![CDATA[
<p>There is no raised cost anymore. Every professional doing this work has perfect alignment with regular consumer heuristics.<p>OS, browser, fingerprinting, networked bytes, residential address spaces.<p>All of it is done.</p>
]]></description><pubDate>Tue, 08 Sep 2026 17:08:02 +0000</pubDate><link>https://news.ycombinator.com/item?id=49613187</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49613187</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49613187</guid></item><item><title><![CDATA[New comment by bellowsgulch in "Benchmarking Qwen3.8 27B quantizations: 4-bit holds up, 1-bit collapses"]]></title><description><![CDATA[
<p>Qwen3.8 27B seems like it was clearly supposed to be a high-end consumer open-weights model, but the t/s is so low for me on my old M1 Max 64GB that I hope others are getting use out of it.<p>Unfortunately, the calculus has changed and it seems cheaper to me to just use MiMo V2.5 for pennies or DeepSeek V4 Flash instead of using Qwen anymore unless I need a local model specifically for doing reverse engineering work that gets otherwise rejected.</p>
]]></description><pubDate>Tue, 08 Sep 2026 15:47:26 +0000</pubDate><link>https://news.ycombinator.com/item?id=49612053</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49612053</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49612053</guid></item><item><title><![CDATA[New comment by bellowsgulch in "GPT-6 Astra on OpenRouter"]]></title><description><![CDATA[
<p>Thanks, might move our corporate accounts over to proxy billing now.</p>
]]></description><pubDate>Sat, 05 Sep 2026 17:31:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=49578720</link><dc:creator>bellowsgulch</dc:creator><comments>https://news.ycombinator.com/item?id=49578720</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49578720</guid></item></channel></rss>