<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: regularfry</title><link>https://news.ycombinator.com/user?id=regularfry</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Sat, 21 Nov 2026 16:08:27 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=regularfry" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by regularfry in "Formal Verification Is Not Verifiable"]]></title><description><![CDATA[
<p>Well yes, but this case is <i>specifically</i> about a time when the right thing wasn't tested for, because the team of wise, talented people attached missed it.</p>
]]></description><pubDate>Tue, 17 Nov 2026 16:31:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=50533873</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50533873</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50533873</guid></item><item><title><![CDATA[New comment by regularfry in "Formal Verification Is Not Verifiable"]]></title><description><![CDATA[
<p>In this case a reasonable conclusion would be that there are no sufficiently competent people and we should act on that basis.  Or, at least, that we have no reliable way of marrying sufficient competence with the right problems.</p>
]]></description><pubDate>Tue, 17 Nov 2026 12:16:13 +0000</pubDate><link>https://news.ycombinator.com/item?id=50530398</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50530398</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50530398</guid></item><item><title><![CDATA[New comment by regularfry in "Things That Turbo Pascal Is Smaller Than (2011)"]]></title><description><![CDATA[
<p>My first naive thought is never mind shared libraries, can touch just exec date when it has to?</p>
]]></description><pubDate>Mon, 16 Nov 2026 00:01:11 +0000</pubDate><link>https://news.ycombinator.com/item?id=50512279</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50512279</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50512279</guid></item><item><title><![CDATA[New comment by regularfry in "Things That Turbo Pascal Is Smaller Than (2011)"]]></title><description><![CDATA[
<p>Yes, but I'd say that Lazarus is the standard bearer for Pascal the way Delphi was back in the day. It would be <i>the</i> reason someone would pick up the language a lot of the time.</p>
]]></description><pubDate>Sun, 15 Nov 2026 23:59:22 +0000</pubDate><link>https://news.ycombinator.com/item?id=50512263</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50512263</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50512263</guid></item><item><title><![CDATA[New comment by regularfry in "Things That Turbo Pascal Is Smaller Than (2011)"]]></title><description><![CDATA[
<p>In a way it didn't, the Lazarus project is the current most well-known incarnation.  But that itself tells a story: it's for desktop applications, mainly.  It never really had a "here's how you do a web app" story around it.  There was a period 1998-2002 when Delphi was <i>hot</i> (and it genuinely was <i>very</i> good) but all the cool kids saw PHP3 and Flash and hopped aboard, and at that point, in retrospect, you can see the start of the decline of desktop applications.  By the time Rails landed in (I think?) 2008 "do everything on the web" was what a lot of people were already talking about, even if the truth of that didn't really land until much later.<p>Really it could have been go-lang before go-lang and I'm not sure why it wasn't other than that it wouldn't have been controllable by the folks at Google.</p>
]]></description><pubDate>Sun, 15 Nov 2026 17:54:28 +0000</pubDate><link>https://news.ycombinator.com/item?id=50509031</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50509031</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50509031</guid></item><item><title><![CDATA[New comment by regularfry in "Things That Turbo Pascal Is Smaller Than (2011)"]]></title><description><![CDATA[
<p>I was going to suggest that even that logic wouldn't account for a binary size three times the RAM of my first machine, but from the man page:<p>> DATE STRING<p>>      The --date=STRING is a mostly free format human readable date string such as "Sun, 29 Feb 2004 16:21:42 -0800" or "2004-02-29 16:21:42" or even "next Thursday".  A date string may contain items indicating calendar date, time of day, time zone, day of week, relative time, relative date, and numbers.  An empty string indicates the beginning of the day.  The date string format is more complex than is easily documented here but is fully described in the info documentation.<p>There it is.</p>
]]></description><pubDate>Sun, 15 Nov 2026 17:50:43 +0000</pubDate><link>https://news.ycombinator.com/item?id=50508994</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50508994</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50508994</guid></item><item><title><![CDATA[New comment by regularfry in "All AI-generated code should be Open Source"]]></title><description><![CDATA[
<p>There's a difference between something being released as open source and something that never gets released at all.  Open source or not only becomes relevant when making the binary available to other people.</p>
]]></description><pubDate>Sun, 15 Nov 2026 15:37:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=50507714</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50507714</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50507714</guid></item><item><title><![CDATA[New comment by regularfry in "Bad ideas that aren't bad anymore"]]></title><description><![CDATA[
<p>If your threat landscape includes "malicious insider" to any significant degree, this is not a path you can walk.</p>
]]></description><pubDate>Thu, 12 Nov 2026 09:26:55 +0000</pubDate><link>https://news.ycombinator.com/item?id=50464945</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50464945</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50464945</guid></item><item><title><![CDATA[New comment by regularfry in "Ripple-65B Preview: U.S. Open-Weight SOTA on Agentic Coding"]]></title><description><![CDATA[
<p>What'll be interesting is how it performs relative to Qwen3.8 27b when quantised to the same size. Which, of course, is a comparison they don't want to do.</p>
]]></description><pubDate>Wed, 04 Nov 2026 21:04:51 +0000</pubDate><link>https://news.ycombinator.com/item?id=50371502</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50371502</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50371502</guid></item><item><title><![CDATA[New comment by regularfry in "Stringman, an $800 open-source room scale gripper robot"]]></title><description><![CDATA[
<p>What you're missing about the attraction of humanoid robots is that you can, software permitting, drop them into an existing factory and entirely replace the human workforce without retooling. They might only be good for a single generation, but they're <i>ideal</i> for that purpose.</p>
]]></description><pubDate>Wed, 04 Nov 2026 09:13:19 +0000</pubDate><link>https://news.ycombinator.com/item?id=50363069</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50363069</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50363069</guid></item><item><title><![CDATA[New comment by regularfry in "Could a modern chess engine on a Pentium have defeated Kasparov in 1996?"]]></title><description><![CDATA[
<p>Slightly difficult to disentangle is that they might be using the verbalisation as a way to explore depth, rubber-duck style, in a way that they wouldn't over the board.</p>
]]></description><pubDate>Tue, 03 Nov 2026 13:19:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=50351629</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50351629</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50351629</guid></item><item><title><![CDATA[New comment by regularfry in "A tiny Pacific country rolls out universal basic income"]]></title><description><![CDATA[
<p>Well yes, but you're still talking ~5 years versus 40.  If the desire is for the state to step in somehow and smooth the transition (which it is, in a "UBI maybe?" discussion) then that's still providing an additional 700% training capacity.  I don't think we've got a good example from history of that working in a way that lends itself to "Keep calm. Carry on." Certainly not outside military build-ups or post-war reconstruction.</p>
]]></description><pubDate>Tue, 03 Nov 2026 12:08:28 +0000</pubDate><link>https://news.ycombinator.com/item?id=50350987</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50350987</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50350987</guid></item><item><title><![CDATA[New comment by regularfry in "Sam Altman's Case Against Your RAM Is Efficient. That's the Problem"]]></title><description><![CDATA[
<p>Efficiency and resilience are always in trade-off.  Even if there was no other argument it would be worth having local inference as a guard against systemic SPOFs in the big providers.</p>
]]></description><pubDate>Tue, 03 Nov 2026 12:00:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=50350920</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50350920</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50350920</guid></item><item><title><![CDATA[New comment by regularfry in "A tiny Pacific country rolls out universal basic income"]]></title><description><![CDATA[
<p>> I disagree. They had to attend class and pass a test to get a CDL. They were trained to drive a truck, they can be trained to perform another task. Truck drivers are not lobotomized.<p>I don't think the difficulty is with the drivers, more with the logistics of retraining them all at once. As a cohort they came up through their class and training nicely spread out over, what, a 40 year period? Making sure there are that many training spaces for other trades over a much more condensed period is non-trivial.</p>
]]></description><pubDate>Sun, 01 Nov 2026 13:12:05 +0000</pubDate><link>https://news.ycombinator.com/item?id=50325331</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50325331</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50325331</guid></item><item><title><![CDATA[New comment by regularfry in "Formlabs Studio 1 $699 Resin Printer"]]></title><description><![CDATA[
<p>"A tad messy" really undersells it. You need a solid process to manage the (sticky, carcinogenic, smelly) liquid resin both as an input and waste product, along with having to dispose of (carcinogenic, smelly, flammable) cleaning solvent.<p>The problem isn't so much the process of a successful print as how you deal with a failed one. At that point you've got a vat full of resin with some solid parts stuck to the bottom, so you need to filter and clean the damn thing which means taking it out of whatever containment you've got. And that step you <i>can't</i> use the usual solvent for because it damages the sheet. But you'll have ample time to figure all this out because you'll get loads of failed prints as you figure out how slicing and supporting for resin works: it's different to FDM, some knowledge carries over but not that much.<p>Focusing on the 10-30 minute curing process ignores the part prep step: SLA is much more reliant on supports than FDM in general, so there <i>will</i> be either rough surfaces hopefully out of sight, or (if it's not a model that orients well) scraping/sanding/filling. And you want to do support separation and get that as clean as possible before curing, while the resin is still a bit soft, but that means it's only partially cured and you need to be PPE'd up.<p>When I'm doing <i>anything</i> with my resin printer I don't even go into the room without a breathing filter, don't touch anything without gloves. It really is a very different experience to FDM.</p>
]]></description><pubDate>Thu, 29 Oct 2026 10:19:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=50286819</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50286819</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50286819</guid></item><item><title><![CDATA[New comment by regularfry in "F-Zero X on Nintendo 3DS"]]></title><description><![CDATA[
<p>Not the GP but the experience of being on the wrong end of a vociferous false positive is <i>not great</i>, and there are enough readers to make that not an unlikely outcome.</p>
]]></description><pubDate>Mon, 26 Oct 2026 10:27:51 +0000</pubDate><link>https://news.ycombinator.com/item?id=50241635</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50241635</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50241635</guid></item><item><title><![CDATA[New comment by regularfry in "LLMs are making formal software verification feasible at a much larger scale"]]></title><description><![CDATA[
<p>A trace is basically a sequence of states. If you're using TLA+ or something built on it (quint in my case) then you can get it to dump out valid sequences that you can then compare to output produced by your implementation. You have to explicitly instrument where the implementation performs the updates, so it's not free and there's potential for error there, but it clearly signals "this bit of code is responsible for that bit of the model".</p>
]]></description><pubDate>Sun, 25 Oct 2026 02:24:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=50227676</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50227676</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50227676</guid></item><item><title><![CDATA[New comment by regularfry in "Root shell on Samsung TV"]]></title><description><![CDATA[
<p>I'll probably get a kiosk panel next time.  But this time, yeah, I've just not configured the network.  All content comes from the console.  Which is a different bundle of problems and no better in a lot of ways, but it's more of a known quantity.</p>
]]></description><pubDate>Sat, 24 Oct 2026 16:23:26 +0000</pubDate><link>https://news.ycombinator.com/item?id=50222758</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50222758</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50222758</guid></item><item><title><![CDATA[New comment by regularfry in "LLMs are making formal software verification feasible at a much larger scale"]]></title><description><![CDATA[
<p>Positive note from me: one thing I've had good experience with is following the path from a project that does something similar to what I want; extract a tech-neutral prose spec from the source; convert that prose spec into a formal spec; use that spec to generate valid traces, a language-specific interface layer, and conformance checks to test whether an implementation of the interface generates good traces.  Once you've got that, implementing the system is a matter of colouring the interface in by numbers, which is really easy for a model.<p>The bonus here is that you can cut in at the prose spec stage and change it to whatever you need, as well as being able to target it at specific technical choices at the last couple of steps.<p>Where I've seen this work really well is with privilege escalation management, where you do actually want to be careful about the states the system can get into, and the APIs available for managing privileges in a system might be super specific to that system.  But I can imagine the workflow being quite a useful one any time you've got anything state-machine flavoured.</p>
]]></description><pubDate>Fri, 23 Oct 2026 12:30:53 +0000</pubDate><link>https://news.ycombinator.com/item?id=50207941</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50207941</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50207941</guid></item><item><title><![CDATA[New comment by regularfry in "The 1.5°C goal of the Paris Agreement is no longer viable"]]></title><description><![CDATA[
<p>The point is that <i>even if we did get to net zero today</i> heating would continue to rise for another 15 years. That's what the lag is. Whenever we get there, add 15 years.</p>
]]></description><pubDate>Wed, 21 Oct 2026 10:09:25 +0000</pubDate><link>https://news.ycombinator.com/item?id=50178415</link><dc:creator>regularfry</dc:creator><comments>https://news.ycombinator.com/item?id=50178415</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50178415</guid></item></channel></rss>