<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: smokel</title><link>https://news.ycombinator.com/user?id=smokel</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Sat, 10 Oct 2026 23:48:02 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=smokel" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[Mxc: Microsoft Execution Containers version 1.0.0]]></title><description><![CDATA[
<p>Article URL: <a href="https://blogs.windows.com/windowsdeveloper/2026/10/07/microsoft-execution-containers-policy-driven-containment-for-ai-agents/">https://blogs.windows.com/windowsdeveloper/2026/10/07/microsoft-execution-containers-policy-driven-containment-for-ai-agents/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=50016956">https://news.ycombinator.com/item?id=50016956</a></p>
<p>Points: 151</p>
<p># Comments: 33</p>
]]></description><pubDate>Fri, 09 Oct 2026 06:52:49 +0000</pubDate><link>https://blogs.windows.com/windowsdeveloper/2026/10/07/microsoft-execution-containers-policy-driven-containment-for-ai-agents/</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=50016956</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50016956</guid></item><item><title><![CDATA[New comment by smokel in "Classic PC demoscene productions running natively in the browser"]]></title><description><![CDATA[
<p>Second Reality ran comfortably with only 640KB RAM using the Gravis Ultrasound.<p>But it did require a 386, which typically had 2 MB or more of RAM back then.</p>
]]></description><pubDate>Thu, 08 Oct 2026 10:13:47 +0000</pubDate><link>https://news.ycombinator.com/item?id=50003986</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=50003986</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50003986</guid></item><item><title><![CDATA[New comment by smokel in "Classic PC demoscene productions running natively in the browser"]]></title><description><![CDATA[
<p>I gave up once full screen video playback was possible from CD-ROM.  After that, doing (typically worse) <i>non</i>-interactive animations in real-time, seemed a bit pointless.<p>But watching Crystal Dreams II on my phone actually made me feel young again :)</p>
]]></description><pubDate>Thu, 08 Oct 2026 10:09:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=50003945</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=50003945</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50003945</guid></item><item><title><![CDATA[New comment by smokel in "Mistral Large 4"]]></title><description><![CDATA[
<p>What was your prompt?  Most of these seem to be related to metaphors for "ideas" or thinking, or having a bright moment.<p>"Zephyr" and "breeze" might be related to forgetting everything, starting fresh.<p>So by this way of naive reverse engineering I would imagine your prompt to be "Forget everything and think about a random word".  That would prime the LLM to come up with these?</p>
]]></description><pubDate>Tue, 06 Oct 2026 18:53:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=49982495</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=49982495</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49982495</guid></item><item><title><![CDATA[New comment by smokel in "Our approach to EU text provenance rules"]]></title><description><![CDATA[
<p>Why use watermarking, and not simply add a signature?</p>
]]></description><pubDate>Mon, 05 Oct 2026 18:08:10 +0000</pubDate><link>https://news.ycombinator.com/item?id=49968284</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=49968284</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49968284</guid></item><item><title><![CDATA[New comment by smokel in "Agents don't need memory, they need documentation"]]></title><description><![CDATA[
<p>Unfortunately, that is not how many software development projects work.  A customer may request certain things, or the software might be part of a larger system.<p>Consider working on software for a coffee machine.  Why is pin 42 (GRIND) activated every now and then?  Would you really want to document that in Git commit messages?</p>
]]></description><pubDate>Mon, 05 Oct 2026 05:49:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=49961112</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=49961112</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49961112</guid></item><item><title><![CDATA[New comment by smokel in "Agents don't need memory, they need documentation"]]></title><description><![CDATA[
<p>Code typically documents the "what" and "how", not the "why".<p>Why something exists, and how it connects to the outside world may be documented in comments, but more often than not it isn't.</p>
]]></description><pubDate>Sun, 04 Oct 2026 18:08:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=49956310</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=49956310</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49956310</guid></item><item><title><![CDATA[New comment by smokel in "FTL: A new operating system for clouds"]]></title><description><![CDATA[
<p>A problem with this approach is that it would put a large burden on the application developer (or development system) to support other devices (or services) than initially planned.<p>Of course, it would be possible to add new drivers only when necessary, but that would also allow for security problems.<p>So, in theory it might work, but in practice it would require quite a bit of thought.</p>
]]></description><pubDate>Sat, 03 Oct 2026 18:18:46 +0000</pubDate><link>https://news.ycombinator.com/item?id=49946489</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=49946489</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49946489</guid></item><item><title><![CDATA[New comment by smokel in "Kolibri: A Sovereign Open-Weight Model"]]></title><description><![CDATA[
<p>Would you mind substantiating these claims a bit?</p>
]]></description><pubDate>Sat, 03 Oct 2026 17:22:24 +0000</pubDate><link>https://news.ycombinator.com/item?id=49946067</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=49946067</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49946067</guid></item><item><title><![CDATA[New comment by smokel in "With most information hidden, the game Stratego had stumped AI until now"]]></title><description><![CDATA[
<p>This approach also works for Hanabi, which is a very interesting game.  You can't see your own cards, but the other players can.  I bought the game because someone on a reinforcement learning podcast [2] mentioned it, and actually played it multiple times.<p>[1] <a href="https://en.wikipedia.org/wiki/Hanabi_(card_game)" rel="nofollow">https://en.wikipedia.org/wiki/Hanabi_(card_game)</a><p>[2] <a href="https://www.talkrl.com/episodes/jakob-foerster" rel="nofollow">https://www.talkrl.com/episodes/jakob-foerster</a></p>
]]></description><pubDate>Fri, 02 Oct 2026 19:14:03 +0000</pubDate><link>https://news.ycombinator.com/item?id=49937337</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=49937337</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49937337</guid></item><item><title><![CDATA[New comment by smokel in "With most information hidden, the game Stratego had stumped AI until now"]]></title><description><![CDATA[
<p>This puts the earlier "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", 2022 [1] in some perspective.  Apparently the "mastering" in 2022 wasn't quite there yet.  Four years later, the new approach seems to actually be better than humans.<p>[1] <a href="https://arxiv.org/abs/2206.15378" rel="nofollow">https://arxiv.org/abs/2206.15378</a></p>
]]></description><pubDate>Fri, 02 Oct 2026 19:05:55 +0000</pubDate><link>https://news.ycombinator.com/item?id=49937262</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=49937262</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49937262</guid></item><item><title><![CDATA[New comment by smokel in "Dutch computer museums (2022)"]]></title><description><![CDATA[
<p>"SAMENGEST TEKEN" is most likely an abbreviation for "SAMENGESTELD TEKEN".  I guess it's a translation of "COMPOSE KEY" and they ran out of rub-on letters somewhere.</p>
]]></description><pubDate>Fri, 02 Oct 2026 17:33:54 +0000</pubDate><link>https://news.ycombinator.com/item?id=49936218</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=49936218</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49936218</guid></item><item><title><![CDATA[New comment by smokel in "Pi 1.0"]]></title><description><![CDATA[
<p>Why do these tools have such bugs?  With AI it should be trivial to fix, no?<p>Or is it a non-trivial bug that requires a lot of refactoring, and could introduce a lot of new bugs?  That would require careful review from a human.<p>The latter may well be a reason why I don't see extreme productivity gains in larger brown-field projects.<p>(Disclaimer: I see enormous benefits in one-off greenfield projects.)</p>
]]></description><pubDate>Fri, 02 Oct 2026 09:06:22 +0000</pubDate><link>https://news.ycombinator.com/item?id=49931400</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=49931400</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49931400</guid></item><item><title><![CDATA[New comment by smokel in "Pi 1.0"]]></title><description><![CDATA[
<p>Full screen hides distractions from other applications.  Seems reasonable as something minimal.<p>Minimalism is a difficult subject.  In art, minimalists tried to strive for something that is universally minimal.  But if you look in nature for straight lines or perfect circles, you end up disappointed.  Turns out minimalism found things that were minimal with respect to how some humans think about minimalism.  For all we know, pure chaos may be more universally minimal than an empty vacuum.</p>
]]></description><pubDate>Thu, 01 Oct 2026 20:21:32 +0000</pubDate><link>https://news.ycombinator.com/item?id=49926606</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=49926606</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49926606</guid></item><item><title><![CDATA[New comment by smokel in "MicroLLM Lab – Try 7 tiny LLM's in the browser"]]></title><description><![CDATA[
<p>No, that is not what SLMs are useful for.  They have very little knowledge, and typically lack reasoning skills.<p>Useful applications include sentiment analysis, text classification, entity extraction, etc.<p>They certainly can be useful, but you shouldn't compare them with LLMs such as Opus or Fable.</p>
]]></description><pubDate>Mon, 28 Sep 2026 20:05:27 +0000</pubDate><link>https://news.ycombinator.com/item?id=49883612</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=49883612</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49883612</guid></item><item><title><![CDATA[New comment by smokel in "Show HN: A CC0 museum of retro 3D tricks you can paste into a page"]]></title><description><![CDATA[
<p>The plasma effect, for example, uses vertex shaders to calculate in the order of 8 sines <i>per pixel</i>, and then adds some squares and square roots.<p>Originally, for a plasma effect, one would calculate a sine based value only for each row and each column.  For a 320x240 pixel effect, only 320+240 "complicated" calculations were necessary per frame.  To combine these at the pixel level, simple adds and bit shifts were enough.  This was necessary, because it was entirely impossible to do so many calculations (or even memory lookups) per pixel on old hardware.<p>The actual effect is nice, but it has little to do with the old demoscene effect :)</p>
]]></description><pubDate>Sun, 27 Sep 2026 19:51:10 +0000</pubDate><link>https://news.ycombinator.com/item?id=49870190</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=49870190</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49870190</guid></item><item><title><![CDATA[New comment by smokel in "Show HN: Building a Markdown editor for Mac, iOS and web"]]></title><description><![CDATA[
<p>This question is somewhat more general than this specific application, but how can an app guarantee that files will remain local, also in future updates of the app?<p>To put it simply, an automatic update of the app might start uploading the notes to a server.  I don't think there is any security boundary in iOS, Android, or any other platform preventing this, right?<p>Other than that, congratulations on what seems to be a nice app!</p>
]]></description><pubDate>Sun, 27 Sep 2026 17:38:24 +0000</pubDate><link>https://news.ycombinator.com/item?id=49868904</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=49868904</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49868904</guid></item><item><title><![CDATA[New comment by smokel in "There are no "rogue" AI agents"]]></title><description><![CDATA[
<p>That seems like a divide and conquer attack.  The actual problem is that the electorate system leads to a two-party equilibrium.</p>
]]></description><pubDate>Sun, 27 Sep 2026 17:29:55 +0000</pubDate><link>https://news.ycombinator.com/item?id=49868807</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=49868807</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49868807</guid></item><item><title><![CDATA[New comment by smokel in "Analyzing Frontier Model Progress with My Favourite Game: Prince of Persia"]]></title><description><![CDATA[
<p>The prompts provided are atrocious.  It's amazing that the LLMs actually built something useful.<p><i>> My first prompt was simple: "in this original code there 6502 assembly code for prince of persia, use the save level files and try to do it in c# console."</i><p>Do what in the what now in the console?</p>
]]></description><pubDate>Sat, 26 Sep 2026 15:14:52 +0000</pubDate><link>https://news.ycombinator.com/item?id=49857358</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=49857358</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49857358</guid></item><item><title><![CDATA[New comment by smokel in "Frontier AI on Your Own Hardware"]]></title><description><![CDATA[
<p>The number of graduates has grown even more dramatically.  The job market not being able to keep up with that could make both your and the OP's statements correct.</p>
]]></description><pubDate>Mon, 21 Sep 2026 21:33:19 +0000</pubDate><link>https://news.ycombinator.com/item?id=49793750</link><dc:creator>smokel</dc:creator><comments>https://news.ycombinator.com/item?id=49793750</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49793750</guid></item></channel></rss>