<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: radicality</title><link>https://news.ycombinator.com/user?id=radicality</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Sat, 30 May 2026 04:57:22 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=radicality" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by radicality in "245TB Micron 6600 ION Data Center SSD Now Shipping"]]></title><description><![CDATA[
<p>Yeah it sucks :( Almost exactly a year ago, I got a brand new 15.36TB Kioxia CD-6R (u.3 pcie4x4 drive) for $1450+tax from serverpartdeals.com - that same drive is now listed for ~$4600 (and it’s also out of stock there)</p>
]]></description><pubDate>Wed, 06 May 2026 16:00:21 +0000</pubDate><link>https://news.ycombinator.com/item?id=48037739</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=48037739</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48037739</guid></item><item><title><![CDATA[New comment by radicality in "How OpenAI delivers low-latency voice AI at scale"]]></title><description><![CDATA[
<p>Yeah I was quite surprised that the advanced chat gpt voice mode can’t itself go and message the frontier model underneath to retrieve data and then speak it.
I basically tried asking it for that (something like “can you go and ask gpt5.5 to research this more in depth, and while we wait, tell me about XYZ”), but apparently that’s not a thing.</p>
]]></description><pubDate>Tue, 05 May 2026 05:26:43 +0000</pubDate><link>https://news.ycombinator.com/item?id=48018406</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=48018406</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48018406</guid></item><item><title><![CDATA[New comment by radicality in "Amateur armed with ChatGPT solves an Erdős problem"]]></title><description><![CDATA[
<p>I believe so. With Pro you get “Thinking” with levels Light, Standard, Extended, and Heavy; and you also get the “Pro” model with levels “Standard” and “Extended”.<p>I don’t often go to Pro as it does take a while like you saw here, but I do often use Thinking Heavy for high quality answers.
Idk why, but i just get consistently worse results with Gemini (Gemini pro), where it’s just much lazier, eg won’t do actual searches unless explicitly told.</p>
]]></description><pubDate>Sun, 26 Apr 2026 18:02:13 +0000</pubDate><link>https://news.ycombinator.com/item?id=47912378</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47912378</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47912378</guid></item><item><title><![CDATA[New comment by radicality in "New 10 GbE USB adapters are cooler, smaller, cheaper"]]></title><description><![CDATA[
<p>I’ve been using the qnap sfp+ thunderbolt one (I think it’s a marvel/aqantia chip) for a few years now everyday with my MacBook and it’s been solid</p>
]]></description><pubDate>Sat, 25 Apr 2026 15:43:31 +0000</pubDate><link>https://news.ycombinator.com/item?id=47902282</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47902282</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47902282</guid></item><item><title><![CDATA[New comment by radicality in "Brands got worse on purpose"]]></title><description><![CDATA[
<p>Definitely feels like fraud to me too.
Kind of reminds me of Amazon listings where some seller has some decent quality product, collects a lot of high reviews, then uses the same listing and does a complete switcharoo to a much shittier more profitable product, replacing images and the description, but keeping the higher rating from earlier. Which also to me sounds like fraud.<p>At least the Chinese brands don’t try and hide it much like the companies listed here, they’ll just generate a new 5-letter new company to sell low quality crap.</p>
]]></description><pubDate>Tue, 21 Apr 2026 16:35:53 +0000</pubDate><link>https://news.ycombinator.com/item?id=47851132</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47851132</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47851132</guid></item><item><title><![CDATA[New comment by radicality in "Critical flaw in Protobuf library enables JavaScript code execution"]]></title><description><![CDATA[
<p>I think something like this: Imagine a tool for querying remote grpc endpoints - you would point at some endpoint to query it / explore it, and the tool would request the proto schema files from a potentially untrusted source - something you would not expect to be that unsafe and trigger an rce.</p>
]]></description><pubDate>Sun, 19 Apr 2026 18:16:34 +0000</pubDate><link>https://news.ycombinator.com/item?id=47826355</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47826355</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47826355</guid></item><item><title><![CDATA[New comment by radicality in "A new spam policy for “back button hijacking”"]]></title><description><![CDATA[
<p>For mobile Safari on iOS/iPad, the back button imo is just completely broken. It’s either a bug, or Apple might say I’m ‘holding it wrong’. 
One version it just stopped doing its one job correctly and it’s messing with my mental model of how I arrived at each tab. Currently:<p>Safari iOS: Be on a page, tap hold a link, click Open in new tab, go to new tab. The Back button should be grayed out and isn’t, and clicking it closes the tab. (???)<p>Chrome iOS: Be on a page, tap hold a link, click Open in new tab, go to new tab. Back button correctly grayed out as the tab has nowhere to go back to.</p>
]]></description><pubDate>Tue, 14 Apr 2026 15:41:24 +0000</pubDate><link>https://news.ycombinator.com/item?id=47767122</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47767122</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47767122</guid></item><item><title><![CDATA[New comment by radicality in "State of Homelab 2026"]]></title><description><![CDATA[
<p>Same question from me too - I do have a few services on my homelab at home - stuff like a NAS, synology surveillance, homeassistant, few lxc containers hosting random services on Proxmox - and it all works just fine for my needs with standard WireGuard vpn setup on all my devices (macbook/ipad/iphone/android). What would cloudflare tunnel get me?</p>
]]></description><pubDate>Mon, 13 Apr 2026 06:15:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=47748251</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47748251</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47748251</guid></item><item><title><![CDATA[New comment by radicality in "Project Glasswing: Securing critical software for the AI era"]]></title><description><![CDATA[
<p>Ah, thanks, makes sense, I’ll read more about this</p>
]]></description><pubDate>Wed, 08 Apr 2026 15:18:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=47691420</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47691420</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47691420</guid></item><item><title><![CDATA[New comment by radicality in "Project Glasswing: Securing critical software for the AI era"]]></title><description><![CDATA[
<p>Huh, I don’t know what “long context performance” means exactly in these tests, so completely anecdotally 
, my experience with gpt5.4 via codex cli vs Claude code opus, gpt5.4 seems to do significantly better in long contexts I think partly due to some special context compaction stored in encrypted blobs. On long conversations opus in Claude code will for me lose memory of what we were working on earlier, whereas one of my codex chats is already at >1B tokens and is still very coherent and remembers things I asked of it at the beginning of the convo.</p>
]]></description><pubDate>Wed, 08 Apr 2026 00:49:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=47683289</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47683289</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47683289</guid></item><item><title><![CDATA[New comment by radicality in "12k AI-generated blog posts added in a single commit"]]></title><description><![CDATA[
<p>About to go do that on Kagi for the linked site. Oh and also hit the “Report this site as AI generated”</p>
]]></description><pubDate>Sat, 04 Apr 2026 18:24:21 +0000</pubDate><link>https://news.ycombinator.com/item?id=47641793</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47641793</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47641793</guid></item><item><title><![CDATA[New comment by radicality in "Samsung Magician disk utility takes 18 steps and two reboots to uninstall"]]></title><description><![CDATA[
<p>Absolutely agree I hate that software. Last I remember I was trying to upgrade firmware I think of either a usbc drive, but could have been some m2 nvme drive via usb4.
Software looked so nasty that I think I managed to get it somehow working in a VM for firmware update.</p>
]]></description><pubDate>Fri, 03 Apr 2026 16:35:22 +0000</pubDate><link>https://news.ycombinator.com/item?id=47628822</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47628822</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47628822</guid></item><item><title><![CDATA[New comment by radicality in "ChatGPT won't let you type until Cloudflare reads your React state"]]></title><description><![CDATA[
<p>The way I use the products something like this. My main account on my MacBook - ChatGPT website, codex cli. Then, a Mac VM running via UTM with shared writable dir - anything more ‘shady’ in terms of permissions and for playing with new ai apps - eg ChatGPT/Codex standalone apps, Atlas, Claude desktop app etc. Seems to work decently enough.
And I do totally agree that there should be a way to opt out of all these privacy invasive measures, especially after paying $200/mo</p>
]]></description><pubDate>Mon, 30 Mar 2026 15:57:41 +0000</pubDate><link>https://news.ycombinator.com/item?id=47575929</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47575929</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47575929</guid></item><item><title><![CDATA[New comment by radicality in "Hacking old hardware by renaming to .zip [video]"]]></title><description><![CDATA[
<p>That’s just sad ugh, just the other day I was using my pre-shitty-IoT era Sous Vide machine (Anova brand, I think it might have been chefsteps recommended too, got around 2014/2015), and I was thinking how glad I am that it has zero fancy connectivity - just a wheel to set the temperature and a start/stop button and simple led display. Still works great.</p>
]]></description><pubDate>Sun, 29 Mar 2026 04:01:04 +0000</pubDate><link>https://news.ycombinator.com/item?id=47560296</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47560296</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47560296</guid></item><item><title><![CDATA[New comment by radicality in "Velxio 2.0 – Emulate Arduino, ESP32, and Raspberry Pi 3 in the Browser"]]></title><description><![CDATA[
<p>Nice. Yeah definitely do consider it, esphome is quite popular with people using HomeAssistant</p>
]]></description><pubDate>Sun, 29 Mar 2026 03:52:13 +0000</pubDate><link>https://news.ycombinator.com/item?id=47560256</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47560256</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47560256</guid></item><item><title><![CDATA[New comment by radicality in "Make macOS consistently bad unironically"]]></title><description><![CDATA[
<p>Yeah long ago when I was doing some iOS development, I can remember Apple UX responsives mantras like “don’t block the main thread”, as it’s the thing responsible for making app UIs snappy even when something is happening.<p>Nowadays seems like half of Apple’s own software blocks on their main thread, like you said things like keyboard lock up for no reason. God forbid you try to paste too much text into a Note - the paste will crawl to a halt. Or, on my M4 max MacBook, 128GB ram, 8tb ssd, Photos library all originals saved locally - I try to cmd-R to rotate an image - the rotation of a fully local image can sometimes take >10 seconds while showing a blocking UI “Rotating Image…”, it’s insane how low the bar has dropped for Apple software.</p>
]]></description><pubDate>Sat, 28 Mar 2026 17:56:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=47556856</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47556856</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47556856</guid></item><item><title><![CDATA[New comment by radicality in "Velxio 2.0 – Emulate Arduino, ESP32, and Raspberry Pi 3 in the Browser"]]></title><description><![CDATA[
<p>I often write a bunch of Esphome ‘code’ , which I then use with various esp32 based devices (mostly from M5stack) via esphome/HomeAssistant.<p>Can this project help me in any way during dev stage before uploading the code to device just to see it doesn’t work ? Eg could I use this to somehow compile&run those esphome yamls via this emulator?</p>
]]></description><pubDate>Fri, 27 Mar 2026 22:18:31 +0000</pubDate><link>https://news.ycombinator.com/item?id=47549114</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47549114</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47549114</guid></item><item><title><![CDATA[New comment by radicality in "Goodbye to Sora"]]></title><description><![CDATA[
<p>You have more info about the inflated token use? I’m using codex cli a bunch now, but the reported token usage seems like an order of magnitude higher than, say Claude code with opus.<p>Idk if it’s because I set codex to xhigh reasoning, but even then it still seems way higher than Claude. The input/output ratio feels large too, eg I have codex session which says ~500M in / ~2M out.</p>
]]></description><pubDate>Tue, 24 Mar 2026 21:17:51 +0000</pubDate><link>https://news.ycombinator.com/item?id=47509472</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47509472</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47509472</guid></item><item><title><![CDATA[New comment by radicality in "The Proliferation of DarkSword: iOS Exploit Chain Adopted by Hackers"]]></title><description><![CDATA[
<p>I think was posted here correctly: <a href="https://news.ycombinator.com/item?id=47426540">https://news.ycombinator.com/item?id=47426540</a></p>
]]></description><pubDate>Wed, 18 Mar 2026 15:42:59 +0000</pubDate><link>https://news.ycombinator.com/item?id=47427202</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47427202</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47427202</guid></item><item><title><![CDATA[New comment by radicality in "1M context is now generally available for Opus 4.6 and Sonnet 4.6"]]></title><description><![CDATA[
<p>Using Codex more for now, and there is definitely some compaction magic.
I’m keeping the same conversation going and going for days, some at almost 1B tokens (per the codex cli counters), with seemingly no coherency loss</p>
]]></description><pubDate>Sat, 14 Mar 2026 16:20:57 +0000</pubDate><link>https://news.ycombinator.com/item?id=47378214</link><dc:creator>radicality</dc:creator><comments>https://news.ycombinator.com/item?id=47378214</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47378214</guid></item></channel></rss>