<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: zenoprax</title><link>https://news.ycombinator.com/user?id=zenoprax</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Sun, 18 Oct 2026 03:49:33 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=zenoprax" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by zenoprax in "DeepSeek-v4.1-Flash-1017"]]></title><description><![CDATA[
<p>What is your workload?</p>
]]></description><pubDate>Sat, 17 Oct 2026 16:38:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=50132571</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=50132571</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50132571</guid></item><item><title><![CDATA[New comment by zenoprax in "DeepSeek-v4.1-Flash-1017"]]></title><description><![CDATA[
<p>I briefly tried GLM 5.3 Flash and the first thing I noticed was how slow it was followed quickly by its single-mindedness. With DS 4.1 Flash I find that it understands that not all the detail in my prompt is important. Sometimes I like to give a quick example of what I'm looking for and it can take the hint whereas GLM will seemingly reshape its worldview around this tiny detail.<p>DS just handles ambiguity and my bursts of detail with much more grace. I'm only using "max effort" and the CoT is reasonable compared to GLM.</p>
]]></description><pubDate>Sat, 17 Oct 2026 16:32:46 +0000</pubDate><link>https://news.ycombinator.com/item?id=50132519</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=50132519</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50132519</guid></item><item><title><![CDATA[New comment by zenoprax in "Someone on Torrent Sites Kept Fucking with My Sonarr"]]></title><description><![CDATA[
<p>Air-gapped Dropbox for a pre-rsync, pre-ssh world.</p>
]]></description><pubDate>Fri, 16 Oct 2026 10:33:53 +0000</pubDate><link>https://news.ycombinator.com/item?id=50116672</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=50116672</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50116672</guid></item><item><title><![CDATA[New comment by zenoprax in "Libreboot 26.10 "Ominous Olive" released with UEFI, Chromebook support"]]></title><description><![CDATA[
<p>The timeless debate of separating art from the artist.<p>In another project (literally: "Untitled") of theirs they are explicit about motivation:<p>> This is to fight against the trend of crappy github-powered websites that plague the internet.</p>
]]></description><pubDate>Thu, 15 Oct 2026 15:10:03 +0000</pubDate><link>https://news.ycombinator.com/item?id=50104838</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=50104838</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50104838</guid></item><item><title><![CDATA[New comment by zenoprax in "Incident with Git Operations, Pull Requests and Actions – Resolved"]]></title><description><![CDATA[
<p>"We don't want your passwords."<p><a href="https://tailscale.com/blog/passkeys" rel="nofollow">https://tailscale.com/blog/passkeys</a><p>Maybe running your own OpenID service would work?<p><a href="https://openid.net/developers/how-connect-works/" rel="nofollow">https://openid.net/developers/how-connect-works/</a><p>Too much work though. Easier to just switch to Nebula or Netbird and self-host the full stack.</p>
]]></description><pubDate>Wed, 07 Oct 2026 17:05:42 +0000</pubDate><link>https://news.ycombinator.com/item?id=49995640</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=49995640</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49995640</guid></item><item><title><![CDATA[New comment by zenoprax in "Show HN: Pi pod – Run your pi coding agent in sandboxes on your own server"]]></title><description><![CDATA[
<p>I looked into kata but it doesn't support Podman (my preferred baseline). I assume that touching a kernel requires some sort of rootful interface but I don't know enough about how seccomp and the other layers interact to know if it's the right balance.<p>gVisor felt a bit more "hacky" initially (at least compared to krun) but I will look into it a bit more as I saw someone else integrating it into their own setup.<p>My goal is to get 80% of the isolation for 20% inconvenience. Right now it feels like I'm getting 30-50% isolation for 10% inconvenience — I'm willing to tolerate a bit more friction if I can get a bigger jump in security.</p>
]]></description><pubDate>Sun, 04 Oct 2026 20:01:04 +0000</pubDate><link>https://news.ycombinator.com/item?id=49957280</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=49957280</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49957280</guid></item><item><title><![CDATA[New comment by zenoprax in "Show HN: Pi pod – Run your pi coding agent in sandboxes on your own server"]]></title><description><![CDATA[
<p>It depends on your threat model but I think it's a good reminder either way. I recently discovered SmolVM <<a href="https://github.com/smol-machines/smolvm" rel="nofollow">https://github.com/smol-machines/smolvm</a>> and it looks like a pretty good sweet spot for usability and isolation. There are so many sandboxing technologies to pick from and understand.<p>I've been using (rootless) Podman which gives me some basic assurances that it will stay in its designated directory and not run tools on my system directly but I have no limits on the network and with an internal UID/GID of 0:0 I have not done myself any favours. This is the same level of protection one would implement to keep a poorly written bash script from wreaking havoc and that's about it.</p>
]]></description><pubDate>Sun, 04 Oct 2026 00:03:46 +0000</pubDate><link>https://news.ycombinator.com/item?id=49949106</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=49949106</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49949106</guid></item><item><title><![CDATA[New comment by zenoprax in "CSS Bed: Classless CSS themes to use as starting points in web development"]]></title><description><![CDATA[
<p>Some really unexpected domain redirects for this one for some reason: <a href="https://www.cssbed.com/vanillacss" rel="nofollow">https://www.cssbed.com/vanillacss</a><p>All the others are as you'd expect.</p>
]]></description><pubDate>Fri, 02 Oct 2026 03:18:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=49929514</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=49929514</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49929514</guid></item><item><title><![CDATA[New comment by zenoprax in "Show HN: NSL – WSL for Linux"]]></title><description><![CDATA[
<p>I was <i>just</i> testing this and it's not clear that it works out of the box. `krun` shows a different kernel than with `crun` but it doesn't reflect the dropped capabilities in the same way. I'm probably holding it wrong but I'm not sure what to look for at the moment.</p>
]]></description><pubDate>Wed, 30 Sep 2026 02:01:47 +0000</pubDate><link>https://news.ycombinator.com/item?id=49903485</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=49903485</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49903485</guid></item><item><title><![CDATA[New comment by zenoprax in "The internet discovers TLA+. Now what?"]]></title><description><![CDATA[
<p>This makes sense to me but if I continue with the "fancy testing" analogy it seems like there is no accounting for interactions beyond the system itself: it has perfect unit and integration testing but only for a myopic amount of <i>global</i> state.<p>I am curious about the limits of what one can expect this language to solve and therefore the types of programs/domains to which it is more/less suited. Why is determining quorum the canonical use case?</p>
]]></description><pubDate>Sun, 27 Sep 2026 15:33:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=49867577</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=49867577</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49867577</guid></item><item><title><![CDATA[New comment by zenoprax in "Git-bug: Distributed, offline-first bug tracker embedded in Git"]]></title><description><![CDATA[
<p>Ah, I see what you're getting at. Reminds me of Keybase. It's mostly just a messaging app now but there was some similar efforts made 10+ years ago:<p><a href="https://keybase.io/blog/keybase-new-key-model" rel="nofollow">https://keybase.io/blog/keybase-new-key-model</a><p><a href="https://keybase.io/blog/encrypted-git-for-everyone" rel="nofollow">https://keybase.io/blog/encrypted-git-for-everyone</a></p>
]]></description><pubDate>Sat, 26 Sep 2026 02:00:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=49852450</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=49852450</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49852450</guid></item><item><title><![CDATA[New comment by zenoprax in "Git-bug: Distributed, offline-first bug tracker embedded in Git"]]></title><description><![CDATA[
<p>> In a distributed/p2p system with social interaction you need well defined identities that carries public keys. If you want to work offline, you actually need a full self-certified history of public key updates.<p>I don't understand how this is a problem. Isn't this already solved with keyservers and importing to a local keyring? My Linux distro has no problem keeping track of who is who and if they are trusted (not updating for a year or two would probably break things).<p>> so that you can backtrack to which pubkey(s) was active at the moment<p>I think GitHub solves this by simply checking at the time of the push and then never again (which is why you can change the keys without de-verifying older commits. Doesn't git's design yield blockchain-like assurance that nothing in the past has been modified?<p>---<p><a href="https://web.plc.directory/" rel="nofollow">https://web.plc.directory/</a> looks interesting, never heard of it.<p>Personally, I think signed commits should be the default. This is especially true in the age of AI where distinguishing humans from machines becomes harder every day. I would love for encrypted email/IM and sharing keys to be the norm for <i>everyone</i> but we're not there and may never be.</p>
]]></description><pubDate>Fri, 25 Sep 2026 18:51:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=49848434</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=49848434</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49848434</guid></item><item><title><![CDATA[New comment by zenoprax in "Sourcehut account takeover via build logs (XSS in ansi2html)"]]></title><description><![CDATA[
<p>Thanks, I appreciate it. Email to hn@zenoprax.com will reach me (reply will be from my real email).</p>
]]></description><pubDate>Fri, 25 Sep 2026 15:11:34 +0000</pubDate><link>https://news.ycombinator.com/item?id=49845752</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=49845752</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49845752</guid></item><item><title><![CDATA[New comment by zenoprax in "Sourcehut account takeover via build logs (XSS in ansi2html)"]]></title><description><![CDATA[
<p>Is your Forgejo+AUR workflow available? I'm looking to do the same (plus Vagrant and OCI images). Getting a headstart on hardening would be a boon.</p>
]]></description><pubDate>Fri, 25 Sep 2026 09:55:26 +0000</pubDate><link>https://news.ycombinator.com/item?id=49842308</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=49842308</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49842308</guid></item><item><title><![CDATA[New comment by zenoprax in "I don't want the details"]]></title><description><![CDATA[
<p>> "What happens when requirements change inside the launch window?"
> "The alert fired but..."<p>In the running example I think it is clear that you could stop at any point and address the problem from that point rather than continuing. Questions such as "Why do we permit last minute changes at all?" are given no space.<p>In a complex system sometimes there is no root cause (cf. aviation crash investigations and the Swiss Cheese nature of risk). By allowing one person to shoulder the entire context of the incident <i>and</i> for that person be the one to suggest lasting corrective actions it leaves little need for the VP to even be involved. How would they know if the proposed corrections are worth the time and money to implement? I don't see any leadership in the example described in the article - I just see a knee-jerk reaction rather than any kind of proactive reflection.<p>> "I thought we fixed this last time!?"<p>> "This time was different in a way that our corrective actions didn't address."</p>
]]></description><pubDate>Wed, 23 Sep 2026 20:43:31 +0000</pubDate><link>https://news.ycombinator.com/item?id=49822288</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=49822288</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49822288</guid></item><item><title><![CDATA[New comment by zenoprax in "The GitHub wiki is an anti-pattern (2022)"]]></title><description><![CDATA[
<p>I tried to make the GH Wiki work and managed to address 6 of the 7 concerns brought up in the article using a GH Action to mirror a directory of docs. It was my first GH Action and proved to be trickier than I thought. For those who just need to expose some markdown and don't want to bother with gh-pages you might find it useful:<p><a href="https://github.com/super-productivity/super-productivity/blob/master/.github/workflows/wiki-sync.yml" rel="nofollow">https://github.com/super-productivity/super-productivity/blo...</a><p>The generic-ness of the wiki wasn't an issue for me as I only intended on using it as a temporary measure to build up the bulk of the content and coordinate with the main dev on finding a balance between the flashy website with its guides and the less flashy docs. gh-pages is the obvious next step of course.</p>
]]></description><pubDate>Wed, 23 Sep 2026 16:39:07 +0000</pubDate><link>https://news.ycombinator.com/item?id=49818750</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=49818750</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49818750</guid></item><item><title><![CDATA[New comment by zenoprax in "Show HN: Drop – A rootless Linux sandbox with gVisor support"]]></title><description><![CDATA[
<p>Thanks. I think it's probably more complicated than I need at the moment but good to know it is an option down the road!</p>
]]></description><pubDate>Tue, 22 Sep 2026 22:49:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=49809379</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=49809379</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49809379</guid></item><item><title><![CDATA[New comment by zenoprax in "Show HN: Drop – A rootless Linux sandbox with gVisor support"]]></title><description><![CDATA[
<p>Can you explain a bit more about the boundary between the environment and my system? Is it just giving read-only access to /usr/lib?<p>I use an alias to launch a Podman container with opencode in $PWD. It is fully ephemeral aside from the directories I map to it (usually a couple configs). XDG_HOME is local to that working dir. My only frustration is that my image is too minimal but that can be resolved.</p>
]]></description><pubDate>Tue, 22 Sep 2026 16:32:56 +0000</pubDate><link>https://news.ycombinator.com/item?id=49803947</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=49803947</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49803947</guid></item><item><title><![CDATA[New comment by zenoprax in "Pirate Face Rescues LLM Models from Deletion"]]></title><description><![CDATA[
<p>Yeah, I just use the "netinstaller" ISOs since it's much smaller and never needs to be updated. If I had a need for air-gapped/offline installs I'd either download a larger ISO or just manually install packages from .deb as needed.</p>
]]></description><pubDate>Sun, 20 Sep 2026 20:18:19 +0000</pubDate><link>https://news.ycombinator.com/item?id=49779658</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=49779658</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49779658</guid></item><item><title><![CDATA[New comment by zenoprax in "Pirate Face Rescues LLM Models from Deletion"]]></title><description><![CDATA[
<p>Yes, models are a <i>good</i> use for P2P especially if everyone agrees to share the same torrent <i>and</i> someone (or a cohort) commit to seeding for the long haul.</p>
]]></description><pubDate>Sun, 20 Sep 2026 19:59:44 +0000</pubDate><link>https://news.ycombinator.com/item?id=49779466</link><dc:creator>zenoprax</dc:creator><comments>https://news.ycombinator.com/item?id=49779466</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49779466</guid></item></channel></rss>