<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: simon83</title><link>https://news.ycombinator.com/user?id=simon83</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 29 Oct 2026 11:10:37 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=simon83" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by simon83 in "Claude Code regularly sends each background agent’s transcript to its model"]]></title><description><![CDATA[
<p>Do you have it decompile and recompile automatically after every update? Or have you turned off auto updates?</p>
]]></description><pubDate>Thu, 29 Oct 2026 10:19:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=50286826</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=50286826</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50286826</guid></item><item><title><![CDATA[New comment by simon83 in "Claude Code regularly sends each background agent’s transcript to its model"]]></title><description><![CDATA[
<p>In my sessions it is ~16%, not using workflows due to the nature of my work (a lot of research subagents where fixed workflows don't work well)</p>
]]></description><pubDate>Thu, 29 Oct 2026 09:43:22 +0000</pubDate><link>https://news.ycombinator.com/item?id=50286551</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=50286551</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50286551</guid></item><item><title><![CDATA[New comment by simon83 in "Claude Code regularly sends each background agent’s transcript to its model"]]></title><description><![CDATA[
<p>Oh wow! I showed Opus the gist and asked it to analyze my own transcripts, and since those summary calls don't show up there, it calculated an estimate based on the 30 second timer.<p>I'm on the 20x Max plan and my quota is at 98% (resets today), so I had a quite heavy week of usage, including a bunch of orchestrated subagent runs (no workflows).<p>API equivalent spending for this week: $1,970<p>Hidden status-label calls: ~$380<p>Share of the true total: 16%<p>That's incredible...</p>
]]></description><pubDate>Thu, 29 Oct 2026 09:36:32 +0000</pubDate><link>https://news.ycombinator.com/item?id=50286504</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=50286504</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50286504</guid></item><item><title><![CDATA[New comment by simon83 in "Claude Code regularly sends each background agent’s transcript to its model"]]></title><description><![CDATA[
<p>Subagents have a default TTL of 5 minutes though, with an opt in to increase it to 1h, for a premium for writing to cache.</p>
]]></description><pubDate>Thu, 29 Oct 2026 08:29:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=50286018</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=50286018</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50286018</guid></item><item><title><![CDATA[New comment by simon83 in "Ollama now supports AMD graphics cards"]]></title><description><![CDATA[
<p>Does anyone know how the AMD consumer GPU support on Linux has been implemented? Must use something else than ROCm I assume? Because ROCm only supports the 7900 XTX on Linux[1], while on Windows[2] support is from RX 6600 and upwards.<p>[1]: <a href="https://rocblas.readthedocs.io/en/rocm-6.0.0/about/compatibility/linux-support.html#linux-supported-gpus" rel="nofollow">https://rocblas.readthedocs.io/en/rocm-6.0.0/about/compatibi...</a>
[2]: <a href="https://rocblas.readthedocs.io/en/rocm-6.0.0/about/compatibility/windows-support.html#windows-supported-gpus" rel="nofollow">https://rocblas.readthedocs.io/en/rocm-6.0.0/about/compatibi...</a></p>
]]></description><pubDate>Fri, 15 Mar 2024 18:32:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=39719127</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=39719127</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=39719127</guid></item><item><title><![CDATA[New comment by simon83 in "Marijuana addiction: those struggling often face skepticism"]]></title><description><![CDATA[
<p>> I can't reason as well or as quickly. Things go over my head. My working memory is so much smaller. I get lost in code all the time. I forget what I'm working on.<p>Is that while being high on THC, or being sober but having consumed lets say the day before?</p>
]]></description><pubDate>Mon, 31 Jul 2023 20:28:59 +0000</pubDate><link>https://news.ycombinator.com/item?id=36948322</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=36948322</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=36948322</guid></item><item><title><![CDATA[New comment by simon83 in "So you want to build your own open source chatbot"]]></title><description><![CDATA[
<p>I also don‘t think a customer facing chat bot brings much value, but an internal, employee only chat bot could be really useful, depending on the organzization of course. The company in my last position was a rather big one with an insanely huge Confluence instance. I‘ve spent (wasted) so much time searching information there. Having a chat bot, trained on all that information, would‘ve been really useful, I think.</p>
]]></description><pubDate>Sat, 29 Jul 2023 11:28:57 +0000</pubDate><link>https://news.ycombinator.com/item?id=36919156</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=36919156</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=36919156</guid></item><item><title><![CDATA[New comment by simon83 in "Valve shows a drop in Linux gamers for march, but the numbers are questionable"]]></title><description><![CDATA[
<p>As a Linux noob I decided to try Arch Linux 4 years ago because I wanted to actually learn how the various parts work that make out a whole Linux distro (things like the bootloader, systemd, various desktop environments etc). Other distros give you everything ready out of the box, which made it hard for me to actually dig deeper under the hood, because I had no clue where to start. With Arch Linux you more or less build a whole Linux system from the bottom up.<p>Following the Arch guides to get a fully working system was easy enough for a complete Linux noob like me. But you still need some basic knowledge about computing, for things like networking, file systems, drive partitioning and such. If you learned this knowledge from other operating systems then Arch Linux is fairly easy to setup by just following their guide.</p>
]]></description><pubDate>Mon, 03 Apr 2023 14:43:04 +0000</pubDate><link>https://news.ycombinator.com/item?id=35424439</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=35424439</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=35424439</guid></item><item><title><![CDATA[New comment by simon83 in "Cerebras-GPT: A Family of Open, Compute-Efficient, Large Language Models"]]></title><description><![CDATA[
<p>I understand that all of this is based on some fundamental mathematics, a couple of algorithms here, and some statistical analysis there. And I find it absolutely amazing that we can use all of that knowledge and encode it into something that resembles intelligence. This makes me think if our brains and the concept of intelligence are really as special and mysterious as we assume.</p>
]]></description><pubDate>Tue, 28 Mar 2023 20:07:55 +0000</pubDate><link>https://news.ycombinator.com/item?id=35347075</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=35347075</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=35347075</guid></item><item><title><![CDATA[New comment by simon83 in "Cerebras-GPT: A Family of Open, Compute-Efficient, Large Language Models"]]></title><description><![CDATA[
<p>Google: "confetti has left the cannon"<p>> No results found for "confetti has left the cannon".<p>I'm amazed that a "stochastic parrot" can come up with such a beautiful idiom.</p>
]]></description><pubDate>Tue, 28 Mar 2023 18:39:44 +0000</pubDate><link>https://news.ycombinator.com/item?id=35345834</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=35345834</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=35345834</guid></item><item><title><![CDATA[New comment by simon83 in "Ask HN: Can any Hetzner user please explain your workflow on Hetzner?"]]></title><description><![CDATA[
<p>> maintain security? built-in firewall and DDoS protection?<p>I have a Hetzner dedicated server (not the Cloud offering) and I setup OpnSense as an all-in-one routing and firewall solution in a separate VM. All incoming and outgoing traffic goes through this OpnSense VM, which acts as default gateway for the host system and all other VMs/Docker containers. You either need to book a 2nd public IPv4  address (or just use IPv6 for free if that is good enough for your use case, since each server comes with a IPv6 /64 subnet), or if you want to just have 1 IPv4 address you could do some Mac spoofing on the main eth interface of the host OS and give the actual Mac address and public IP to the OpnSense's WAN interface. This is necessary because Hetzner has some Mac address filtering in place, meaning only the Mac address connected to the public IP is allowed to make traffic.</p>
]]></description><pubDate>Wed, 22 Mar 2023 13:39:14 +0000</pubDate><link>https://news.ycombinator.com/item?id=35260827</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=35260827</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=35260827</guid></item><item><title><![CDATA[New comment by simon83 in "OpenAI Research Says 80% of U.S. Workers' Jobs Will Be Impacted by GPT"]]></title><description><![CDATA[
<p>You, like many others, are basing your opinion on the <i>current</i> state of LLMs/GPT. And I totally agree with you, the <i>current</i> GTP-4 version might not replace programmers. But how about future iterations? Personally I can't fathom what the next few years will bring us in this area, especially considering the jump GPT 3.5 to GPT 4 has made in such a short time frame. I'm almost convinced that it <i>will</i> make a lot of jobs obsolete in the future, including some programmer jobs, but I'm not bold enough and make a prediction when this will happen, be it in 2, 4, 10 or 20 years.</p>
]]></description><pubDate>Tue, 21 Mar 2023 14:47:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=35246970</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=35246970</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=35246970</guid></item><item><title><![CDATA[New comment by simon83 in "Ask HN: Am I getting older or did typing on the iPhone become unbearable?"]]></title><description><![CDATA[
<p>Oh wow and I thought I'm going crazy, because I also make so many typos when typing on my iPhone 11 (normal size)! On any other Android phone I don't make nearly as many typos, and it just feels so much better. I never understood why, maybe the "hitbox" on the iOS keyboard keys is just smaller? Because I could swear I'm tapping the right key but somehow it often triggers a neighboring key.<p>Also agree with the horrible cursor movement and text selection, it is wack af on iOS</p>
]]></description><pubDate>Wed, 19 Oct 2022 10:25:16 +0000</pubDate><link>https://news.ycombinator.com/item?id=33260023</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=33260023</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=33260023</guid></item><item><title><![CDATA[New comment by simon83 in "React I love you, but you're bringing me down"]]></title><description><![CDATA[
<p>I was a huge React fan myself but when I started to use Vue and its Composition API one year ago I had tears of joy in my eyes. You have a setup function that is executed exactly once. Inside this setup function you setup your life cycle callbacks, reactive state variables and so on. This is how I wish React + Hooks should have been. Nowadays I won't even accept projects that use React anymore, exactly because of all the pain points OP mentioned.<p>SolidJS is following the same principle like Vue + Composition API. I think Svelte is also in the same boat. This kind of pattern is so much easier to use and reason about. React Hooks become really painful and annoying really fast.</p>
]]></description><pubDate>Tue, 20 Sep 2022 16:45:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=32914350</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=32914350</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=32914350</guid></item><item><title><![CDATA[New comment by simon83 in "Solid.js feels like what I always wanted React to be"]]></title><description><![CDATA[
<p>I've been using React since around 2016 and when Hooks were introduced I played around with it but didn't like it at all exactly because of the reasons mentioned in this article. Sure it reduced quite a bit of boilerplate code that came with class based components and enabled better code re-usability outside of class inheritance or mixins but the disadvantages were too big for me personally to really consider using Hooks in a "greenfield" setting.<p>In my current project I'm using Vue with Composition API which gave me the exact same "aha" moment the author of this article had with SolidJS. Vue + Composition API is way more similar to SolidJS in principle than React + Hooks.</p>
]]></description><pubDate>Tue, 01 Mar 2022 10:45:31 +0000</pubDate><link>https://news.ycombinator.com/item?id=30512446</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=30512446</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=30512446</guid></item><item><title><![CDATA[New comment by simon83 in "Hetzner now provides IPv6 only dedicated servers"]]></title><description><![CDATA[
<p>For this reason I've setup a OpnSense VM on my dedicated Hetzner server where all inbound and outbound IPv4/IPv6 traffic has to go through, it acts as a gateway for the host itself and my other VMs. OpnSense itself is a pretty powerful firewall with tons of other features.<p>Of course you'll lose access to your server if the OpnSense VM breaks or doesn't boot up for whatever reasons after an update or so, but after 2 years I haven't had any problems. But in case something goes wrong Hetzner offers some nice recovery options, even if you don't have internet access to you server you can access your volumes in some kind of VM and get access to it via a VNC like interface (I had to use this feature a few times during the initial setup which consisted of a lot of trial and error I locked myself out a few times).<p>I wouldn't run this setup for anything mission critical of course, it's way too hacky and an official firewall solution would be better, but for my personal purposes as a "home lab" like setup it works perfectly fine so far.</p>
]]></description><pubDate>Tue, 07 Dec 2021 17:31:38 +0000</pubDate><link>https://news.ycombinator.com/item?id=29475509</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=29475509</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=29475509</guid></item><item><title><![CDATA[New comment by simon83 in "Microsoft Paint to be killed off after 32 years"]]></title><description><![CDATA[
<p>I wonder if you could get into legal trouble if you start to re-distribute calc.exe from older Windows version, as a free download. Probably yes, but it would look so ridiculous if MS lawyers would go after a random person doing that.</p>
]]></description><pubDate>Mon, 24 Jul 2017 18:11:04 +0000</pubDate><link>https://news.ycombinator.com/item?id=14841079</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=14841079</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=14841079</guid></item><item><title><![CDATA[New comment by simon83 in "Microsoft Paint to be killed off after 32 years"]]></title><description><![CDATA[
<p>I just wanted to add that Pinta 1.6 (which is the current stable version in the official Arch repos, for example) has awfully slow rectangle/selection tools on larger images. In my case, selecting something in a 2000x2000 image would freeze the whole program for a couple of seconds. Same with resizing or dragging a selection. It was pretty much useless.<p>But 1.7 fixes this issue. It's a development version and not considered "stable" yet (though I haven't experienced any issues so far).</p>
]]></description><pubDate>Mon, 24 Jul 2017 15:38:42 +0000</pubDate><link>https://news.ycombinator.com/item?id=14839354</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=14839354</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=14839354</guid></item><item><title><![CDATA[New comment by simon83 in "Germans in the race to bring electric cars to market"]]></title><description><![CDATA[
<p>For me a range of 100 km would be good enough. I'm just not willing to pay 20.000€+ for that. And so I'm currently looking into a Renault Twizy, which has a range of up to 100 km, for around 8700€. Only problem I have with it is that it's a one person car, and the battery rent is 50€ per month, which is more than I currently pay for gasoline.</p>
]]></description><pubDate>Mon, 24 Jul 2017 08:35:11 +0000</pubDate><link>https://news.ycombinator.com/item?id=14836779</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=14836779</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=14836779</guid></item><item><title><![CDATA[New comment by simon83 in "Germans in the race to bring electric cars to market"]]></title><description><![CDATA[
<p>> unless Tesla innovated far beyond those, it'll be a marginal car company<p>You mean just like Apple became a marginal tech company?<p>I find it fascinating how people make bold claims here in the comments, predicting the future like this...</p>
]]></description><pubDate>Mon, 24 Jul 2017 08:25:02 +0000</pubDate><link>https://news.ycombinator.com/item?id=14836741</link><dc:creator>simon83</dc:creator><comments>https://news.ycombinator.com/item?id=14836741</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=14836741</guid></item></channel></rss>