<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: xKingfisher</title><link>https://news.ycombinator.com/user?id=xKingfisher</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Tue, 19 May 2026 00:57:46 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=xKingfisher" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by xKingfisher in "Garry Tan, the CEO of YC, accused me of unethical reporting"]]></title><description><![CDATA[
<p>How does it seem that Radley is talking past Gary?<p>All discussion of the 'Misrepresentations' article is responsive to Gary's mention of it in the original article. And at no point does Radley appear to endorse its contents.</p>
]]></description><pubDate>Mon, 18 May 2026 19:35:10 +0000</pubDate><link>https://news.ycombinator.com/item?id=48184433</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=48184433</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48184433</guid></item><item><title><![CDATA[Why would ChatGPT "confess" to a crime it didn't commit?]]></title><description><![CDATA[
<p>Article URL: <a href="https://radleybalko.substack.com/p/why-would-chatgpt-confess-to-a-crime">https://radleybalko.substack.com/p/why-would-chatgpt-confess-to-a-crime</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=47905438">https://news.ycombinator.com/item?id=47905438</a></p>
<p>Points: 2</p>
<p># Comments: 0</p>
]]></description><pubDate>Sat, 25 Apr 2026 22:58:57 +0000</pubDate><link>https://radleybalko.substack.com/p/why-would-chatgpt-confess-to-a-crime</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=47905438</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47905438</guid></item><item><title><![CDATA[New comment by xKingfisher in "Tinkering is a way to acquire good taste"]]></title><description><![CDATA[
<p>> Coding for others is not art.<p>It absolutely is, and I think it's what separates good from bad and junior from senior devs.<p>Most devs can produce an artifact that more or less works. But one that has an internal consistency others can understand and extend, one which accurately captures the problem as it exists and ways it will likely change, is much more of an art form.<p>A big part of that is knowing which situations are worth making a stand. Every you write code or leave feedback, your doing it for your team current and future.</p>
]]></description><pubDate>Wed, 29 Oct 2025 03:46:56 +0000</pubDate><link>https://news.ycombinator.com/item?id=45742391</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=45742391</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45742391</guid></item><item><title><![CDATA[New comment by xKingfisher in "Tinkering is a way to acquire good taste"]]></title><description><![CDATA[
<p>I think a distinction can be made between bad taste and different tastes.<p>One of the greatest developers I've worked with, who I learned a lot from and respect immensely, has extremely different tastes in software from me. To the point where I wouldn't say I think he has good taste.<p>But, his work still has a distinct style and intention. I can tell anytime I come across libraries he had a hand in. I understand what the code is doing and why is is correct, even when I disagree with it.<p>And I think that is what is important. When working with more junior people, I'll ask them why they did things a certain way and will generally me be with a "well, idk" of some variant of path dependence.<p>I think developing that intentionality as a developer is important. Which does come with some amount of aesthetic, and I think taste is a defensible metaphor.</p>
]]></description><pubDate>Wed, 29 Oct 2025 03:37:49 +0000</pubDate><link>https://news.ycombinator.com/item?id=45742341</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=45742341</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45742341</guid></item><item><title><![CDATA[New comment by xKingfisher in "Kilauea volcano errupts, lava more than 1k feet high [video]"]]></title><description><![CDATA[
<p>It's a "rain shadow"[0]<p>The predominant wind is from the east, and the air cools aid forms rainclouds as it tries to rise over the mountains in the center of the island. Then warms again as it descends down the eastern slopes.<p>So the eastern (Hilo) side is pretty lush jungle, and the west(Kona) is desert. 
With snowy mountains in between.<p>[0]<a href="https://www.climate.gov/news-features/featured-images/rain-shadows-summits-hawaii" rel="nofollow">https://www.climate.gov/news-features/featured-images/rain-s...</a></p>
]]></description><pubDate>Sun, 22 Jun 2025 20:54:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=44350199</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=44350199</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44350199</guid></item><item><title><![CDATA[New comment by xKingfisher in "Waymo rides cost more than Uber or Lyft and people are paying anyway"]]></title><description><![CDATA[
<p>I've had the opposite experience recently.<p>Going from downtown Los Angeles to Santa Monica is $30s for a Waymo and runs up to $50s for Uber/Lyft (sometimes). Otherwise, they tend to be within a few dollars.<p>I figured it was a combination of Google subsidizing rides and a lack of a "traffic tax".<p>They're a significantly better experience for 45+ minute rides.</p>
]]></description><pubDate>Sun, 15 Jun 2025 00:16:57 +0000</pubDate><link>https://news.ycombinator.com/item?id=44279710</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=44279710</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44279710</guid></item><item><title><![CDATA[New comment by xKingfisher in "We Can't Have Serious Discussions About Section 230 If People Misrepresent It"]]></title><description><![CDATA[
<p>We have common carrier laws to cover cases like Comcast. Which Section 230 explicitly says social media platforms are not.<p>And the problem with Comcast is not necessarily that they will make speech decisions we don't like (an inherent part of free speech). The problem is that they're a monopoly and people don't have other options. So we should fix that problem instead of further limiting 1A rights.</p>
]]></description><pubDate>Sat, 02 Mar 2024 21:44:32 +0000</pubDate><link>https://news.ycombinator.com/item?id=39576037</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=39576037</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=39576037</guid></item><item><title><![CDATA[New comment by xKingfisher in "America lost the chestnut, its "perfect tree""]]></title><description><![CDATA[
<p>The issues with the Darling 58 project had some good discussion here two weeks ago[0].<p>Tangentially: A distillery in Los Angeles, The Obscure, is making a rye whisky aged in American Chestnut barrels, with some of the proceeds going to the American Chestnut Foundation. It's quite good.<p>[0]<a href="https://news.ycombinator.com/item?id=38573765">https://news.ycombinator.com/item?id=38573765</a></p>
]]></description><pubDate>Sun, 24 Dec 2023 17:08:49 +0000</pubDate><link>https://news.ycombinator.com/item?id=38754836</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=38754836</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=38754836</guid></item><item><title><![CDATA[New comment by xKingfisher in "A study of Google's code review tooling (Critique)"]]></title><description><![CDATA[
<p>I try to avoid nits totally unrelated to the changes at hand, since on a subconscious level they may discourage people from even wanting to touch older/less loved files at all.<p>The critical exception being avoiding issues due to path dependence.<p>E.g while a change is "correct" is doing X poorly because of surrounding issue Y. So we should fix Y now instead of building atop it.</p>
]]></description><pubDate>Mon, 04 Dec 2023 22:23:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=38524025</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=38524025</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=38524025</guid></item><item><title><![CDATA[New comment by xKingfisher in "A Proposal Fix for C/C++ Relaxed Atomics in Practice"]]></title><description><![CDATA[
<p>I think it's just a very confusing, hard to use correctly behavior. And according to the article, very bug prone in implementations.<p>In the example given, the result is that both writes happened before both reads, which directly contradicts the source. There's a valid explanation for why it happens, but it's still paradoxical.<p>I remember at work a C++ standards committee member was giving examples of atomic and how to use them safely with different memory models, when someone pointed out his toy example for relaxed order was wrong. It took 5 people debating for a week to figure out what a safe/correct behavior would be. For a 10 line sample class.<p>As times your second question, the article recommends Hans Boehm's proposal to add a no-op conditional branch after the read. I guess it forces the load to be resolved and enforces a sequenced-before behavior on the individual threads. So at least one read must resolve before one write in the example.</p>
]]></description><pubDate>Sun, 05 Nov 2023 16:00:46 +0000</pubDate><link>https://news.ycombinator.com/item?id=38152265</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=38152265</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=38152265</guid></item><item><title><![CDATA[New comment by xKingfisher in "Ken Thompson on the bug that exposed his compiler trojan (2021)"]]></title><description><![CDATA[
<p>There is more than just hope to rely on.<p>Diverse Double-Compiling[0] can provably detect this class of attack.<p>[0]<a href="https://dwheeler.com/trusting-trust/" rel="nofollow noreferrer">https://dwheeler.com/trusting-trust/</a></p>
]]></description><pubDate>Tue, 20 Jun 2023 00:11:43 +0000</pubDate><link>https://news.ycombinator.com/item?id=36398083</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=36398083</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=36398083</guid></item><item><title><![CDATA[New comment by xKingfisher in "Sacrifice the first 13 years of your life to Google for 2M"]]></title><description><![CDATA[
<p>But you can do almost all that while working at Google. And imo, the stability makes it even easier to do so.<p>It's been my experience that Google is a super chill place to work, especially when you take advantage of the fact that you're just one cog among many.<p>You get 5 weeks pto plus 4 weeks work from anywhere. This year alone I'm going to the Arctic, Antarctica, and the Carribean. 
Including a 40 day stretch without internet. I couldn't imagine doing the latter in a startup.<p>And all the while getting paid and having the comfort to make plans knowing that I'll (probably) continue to do so in the future.</p>
]]></description><pubDate>Fri, 16 Jun 2023 23:34:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=36365435</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=36365435</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=36365435</guid></item><item><title><![CDATA[New comment by xKingfisher in "Do not taunt happy fun branch predictor"]]></title><description><![CDATA[
<p>It's mentioned in passing at the end of the "the trouble with interpreter loops" section.<p>A traditional switch/goto loop can thrash the branch predictor. Separating into different tail calling functions gives you more slots and allows the branch predictor to learn relationships between ops.<p>Not to discount the many other benefits of tail calls.<p>*Edit: I misspoke slightly, computed gotos can also split the patch jump, but less reliably[0].<p>[0]<a href="https://gcc.gnu.org/pipermail/gcc/2021-April/235891.html" rel="nofollow">https://gcc.gnu.org/pipermail/gcc/2021-April/235891.html</a></p>
]]></description><pubDate>Thu, 26 Jan 2023 03:48:44 +0000</pubDate><link>https://news.ycombinator.com/item?id=34527464</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=34527464</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=34527464</guid></item><item><title><![CDATA[New comment by xKingfisher in "Do not taunt happy fun branch predictor"]]></title><description><![CDATA[
<p>An interesting application of this is massaging the branch predictor using tail calls to speed up a parser/interpreter:<p><a href="https://blog.reverberate.org/2021/04/21/musttail-efficient-interpreters.html" rel="nofollow">https://blog.reverberate.org/2021/04/21/musttail-efficient-i...</a></p>
]]></description><pubDate>Thu, 26 Jan 2023 01:19:03 +0000</pubDate><link>https://news.ycombinator.com/item?id=34526583</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=34526583</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=34526583</guid></item><item><title><![CDATA[New comment by xKingfisher in "Elizabeth Holmes is sentenced to more than 11 years for fraud"]]></title><description><![CDATA[
<p>Federal prison doesn't have parole, it is a fixed 'good conduct' credit of up to 54 days per year. So she's serving a little more than 9 years minimum.</p>
]]></description><pubDate>Sat, 19 Nov 2022 01:20:41 +0000</pubDate><link>https://news.ycombinator.com/item?id=33665603</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=33665603</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=33665603</guid></item><item><title><![CDATA[New comment by xKingfisher in "What can mRNA treat next?"]]></title><description><![CDATA[
<p>They still have a number of issues.<p>* They're expensive (~4-8 dollars per follicle) with a single transplant being 1000-5000 follicles. Severely bald men may need multiple<p>* There is a finite number of follicles that can be transplanted.<p>* Permanent scarring in the donor zone.<p>The ultimate treatment would be either hair cloning (still a ways out, very expensive) or a way to reactivate the dormant follicles.  Presumably if you could solve the latter you would become very rich very quickly.</p>
]]></description><pubDate>Sun, 02 May 2021 22:45:49 +0000</pubDate><link>https://news.ycombinator.com/item?id=27019733</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=27019733</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=27019733</guid></item><item><title><![CDATA[New comment by xKingfisher in "Don’t blindly prefer emplace_back to push_back"]]></title><description><![CDATA[
<p>It's definitely a lot to process.<p>I found these two articles really helpful when trying to understand moves:<p><a href="https://abseil.io/tips/55" rel="nofollow">https://abseil.io/tips/55</a>
<a href="https://abseil.io/tips/77" rel="nofollow">https://abseil.io/tips/77</a><p>Though I'd also say don't worry about it too much, especially at first. If you're copying a lot of temporary objects moving can get you some performance wins, but that's something profiling should be telling you m</p>
]]></description><pubDate>Thu, 04 Mar 2021 12:23:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=26341456</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=26341456</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=26341456</guid></item><item><title><![CDATA[New comment by xKingfisher in "Don’t blindly prefer emplace_back to push_back"]]></title><description><![CDATA[
<p>The article did touch on a possible reason for the misunderstanding.<p>If you already know from previous versions that push_back makes a copy and that C++11 added move semantics and emplace_back, it's not a huge leap to connect the two. Especially if you don't notice push_back's rvalue overload (or understand rvalues).<p>And if you're new to C++ and are having all of this thrown at you at once it'd definitely be easy to get some crossed wires.</p>
]]></description><pubDate>Thu, 04 Mar 2021 10:29:28 +0000</pubDate><link>https://news.ycombinator.com/item?id=26340785</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=26340785</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=26340785</guid></item><item><title><![CDATA[New comment by xKingfisher in "Don’t blindly prefer emplace_back to push_back"]]></title><description><![CDATA[
<p>And explicit constructors won't save you from this one.</p>
]]></description><pubDate>Thu, 04 Mar 2021 10:24:04 +0000</pubDate><link>https://news.ycombinator.com/item?id=26340755</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=26340755</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=26340755</guid></item><item><title><![CDATA[New comment by xKingfisher in "Don’t blindly prefer emplace_back to push_back"]]></title><description><![CDATA[
<p>Given the whole scanf/strlen thing going on, more blog posts about API pitfalls wouldn't necessarily be a bad thing.<p>And that is basically the premise of the Abseil Tip of the week series, which are an incredible resource for learning how to work with modern C++.</p>
]]></description><pubDate>Thu, 04 Mar 2021 10:20:28 +0000</pubDate><link>https://news.ycombinator.com/item?id=26340724</link><dc:creator>xKingfisher</dc:creator><comments>https://news.ycombinator.com/item?id=26340724</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=26340724</guid></item></channel></rss>