<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: maille</title><link>https://news.ycombinator.com/user?id=maille</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Sat, 26 Sep 2026 14:21:33 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=maille" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[Show HN: I used an expiring Codex reset to port QtScript to Qt6]]></title><description><![CDATA[
<p>Hey HN, here is a short story that might be worth sharing.<p>I have an application that has been scriptable with QtScript for years. I can automate and extend it through js, with access to QObjects, signals, slots, properties, etc.<p>Unfortunately, QtScript disappeared with Qt6, so (after a long research) my plan was to migrate to PythonQt. That meant rewriting existing scripts, deploying a Python runtime, and praying that all our scripts would still be feasible with PythonQt.
I had no particular desire to do all that, so I had been postponing it. The thinkg about QtScript is that, contrary to Qt, online resources are very scarce, you barely find anything but old forum posts from the 2010s<p>Since I'm on the $20 Codex plan and had 2 resets that was going to expire unused. I thought I might as well spend it on something slightly unreasonable: take the Qt5.15 QtScript sources and see how far I could get on Qt6<p>-> After roughly one hour, I had a smoke test compiling, linking against Qt6 and evaluating js! (Sol/Medium did the plan and the coding)<p>Obviously that was not enough to call it a port, so the real work became validation.
I kept everything as a small patch series, reviewed the changes, rejected some questionable decisions, and treated CI as the authority, something I'm not realled used to do.
The current matrix covers Qt 6.8 LTS and Qt 6.11 on GCC and MSVC, with 29,890 ECMAScript tests and 137 V8 tests passing<p>I (we?) also ported QtScriptGenerator which was important for my use case.
While QtScript is the JavaScript engine, QtScriptGenerator generates the bindings that expose Qt classes and my application classes to JavaScript.
This means a script can take an existing button in the application, reconnect it to custom JavaScript logic, create a new QLabel from JavaScript, inject it into an existing QWidget, and send the result there. No rebuild and no new C++ plugin.
This is the part I really did not want to lose.<p>-> took the second reset an another hour to have a project that compiles and pass the smoke tests (this time I used Luna/xhigh)<p>To me, the interesting AI part was that this kind of old-code archaeology worked surprisingly well with an agent when the work was constrained and split into reviewable steps.
My workflow was: plan the change, let it implement one part, review it, test it, reject the strange ideas, continue.<p>So if someone else is still stuck on Qt 5 mainly because of QtScript or QtScriptGenerator I hope this might be useful.<p><a href="https://github.com/JulienMaille/qtscript-qt6" rel="nofollow">https://github.com/JulienMaille/qtscript-qt6</a><p><a href="https://github.com/JulienMaille/qtscriptgenerator-qt6" rel="nofollow">https://github.com/JulienMaille/qtscriptgenerator-qt6</a><p>cheers!</p>
<hr>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=49260682">https://news.ycombinator.com/item?id=49260682</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Tue, 11 Aug 2026 16:26:24 +0000</pubDate><link>https://github.com/JulienMaille/qtscript-qt6</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=49260682</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49260682</guid></item><item><title><![CDATA[New comment by maille in "[dead]"]]></title><description><![CDATA[
<p>Concrete cost data from today: rebuilding SQLite in Rust via planner/worker split cut agent spend 22x</p>
]]></description><pubDate>Mon, 27 Jul 2026 22:02:03 +0000</pubDate><link>https://news.ycombinator.com/item?id=49076044</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=49076044</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49076044</guid></item><item><title><![CDATA[Our toolchain assumes one human writer, AI agents break that illusion]]></title><description><![CDATA[
<p>Article URL: <a href="https://christophermeiklejohn.com/ai/zabriskie/agents/reliability/testing/ci/distributed/2026/07/27/one-writer.html">https://christophermeiklejohn.com/ai/zabriskie/agents/reliability/testing/ci/distributed/2026/07/27/one-writer.html</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=49075992">https://news.ycombinator.com/item?id=49075992</a></p>
<p>Points: 4</p>
<p># Comments: 0</p>
]]></description><pubDate>Mon, 27 Jul 2026 21:56:40 +0000</pubDate><link>https://christophermeiklejohn.com/ai/zabriskie/agents/reliability/testing/ci/distributed/2026/07/27/one-writer.html</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=49075992</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49075992</guid></item><item><title><![CDATA[New comment by maille in "GPT-5.5 Codex reasoning-token clustering may be leading to degraded performance"]]></title><description><![CDATA[
<p>tldr:<p>GPT-5.5 Codex model exhibits a clustering phenomenon in which <i>reasoning_output_tokens</i> cluster at fixed values spaced 518 apart.<p>These stuck responses at fixed thresholds are strongly correlated with errors in complex tasks.<p>Observed phenomenon is specific to GPT-5.5; it is much less prevalent in GPT-5.4 and almost absent in GPT-5.2 and 5.3</p>
]]></description><pubDate>Sat, 04 Jul 2026 21:51:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=48789429</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=48789429</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48789429</guid></item><item><title><![CDATA[GPT-5.5 Codex reasoning-token clustering may be leading to degraded performance]]></title><description><![CDATA[
<p>Article URL: <a href="https://github.com/openai/codex/issues/30364">https://github.com/openai/codex/issues/30364</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=48789428">https://news.ycombinator.com/item?id=48789428</a></p>
<p>Points: 372</p>
<p># Comments: 152</p>
]]></description><pubDate>Sat, 04 Jul 2026 21:51:09 +0000</pubDate><link>https://github.com/openai/codex/issues/30364</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=48789428</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48789428</guid></item><item><title><![CDATA[New comment by maille in "ChatGPT as the AOL of AI"]]></title><description><![CDATA[
<p>Intersting piece that was posted here and there in HN's comments but could deserve more visibility. Many points make sense</p>
]]></description><pubDate>Sat, 23 May 2026 08:01:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=48245749</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=48245749</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48245749</guid></item><item><title><![CDATA[ChatGPT as the AOL of AI]]></title><description><![CDATA[
<p>Article URL: <a href="https://rebecca-powell.com/posts/return-on-intelligence-02-moats/">https://rebecca-powell.com/posts/return-on-intelligence-02-moats/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=48245748">https://news.ycombinator.com/item?id=48245748</a></p>
<p>Points: 5</p>
<p># Comments: 1</p>
]]></description><pubDate>Sat, 23 May 2026 08:01:29 +0000</pubDate><link>https://rebecca-powell.com/posts/return-on-intelligence-02-moats/</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=48245748</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48245748</guid></item><item><title><![CDATA[New comment by maille in "Google Antigravity 2.0"]]></title><description><![CDATA[
<p>Looks a lot like Codex!</p>
]]></description><pubDate>Tue, 19 May 2026 19:56:07 +0000</pubDate><link>https://news.ycombinator.com/item?id=48198605</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=48198605</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48198605</guid></item><item><title><![CDATA[New comment by maille in "Qwen 3.7 Preview"]]></title><description><![CDATA[
<p>What are your use cases?</p>
]]></description><pubDate>Mon, 18 May 2026 22:44:07 +0000</pubDate><link>https://news.ycombinator.com/item?id=48186835</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=48186835</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48186835</guid></item><item><title><![CDATA[New comment by maille in "Show HN: Semble – Code search for agents that uses 98% fewer tokens than grep"]]></title><description><![CDATA[
<p>Wondering too</p>
]]></description><pubDate>Sun, 17 May 2026 23:01:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=48173915</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=48173915</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48173915</guid></item><item><title><![CDATA[New comment by maille in "Codex is now in the ChatGPT mobile app"]]></title><description><![CDATA[
<p>Does it work on windows? And how do you then remote in?</p>
]]></description><pubDate>Thu, 14 May 2026 23:28:38 +0000</pubDate><link>https://news.ycombinator.com/item?id=48142577</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=48142577</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48142577</guid></item><item><title><![CDATA[New comment by maille in "Accelerating Gemma 4: faster inference with multi-token prediction drafters"]]></title><description><![CDATA[
<p>Bad at coding, but would it be good at code review?</p>
]]></description><pubDate>Tue, 05 May 2026 22:19:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=48029500</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=48029500</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48029500</guid></item><item><title><![CDATA[New comment by maille in "We replaced RAG with a virtual filesystem for our AI documentation assistant"]]></title><description><![CDATA[
<p>If you are still ready comments, have you heard of this? <a href="https://docs.kiln.tech/docs/documents-and-search-rag" rel="nofollow">https://docs.kiln.tech/docs/documents-and-search-rag</a></p>
]]></description><pubDate>Sun, 12 Apr 2026 09:29:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=47737645</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=47737645</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47737645</guid></item><item><title><![CDATA[New comment by maille in "How many products does Microsoft have named 'Copilot'?"]]></title><description><![CDATA[
<p>How can I learn that clever prompting?</p>
]]></description><pubDate>Sat, 04 Apr 2026 22:31:32 +0000</pubDate><link>https://news.ycombinator.com/item?id=47644210</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=47644210</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47644210</guid></item><item><title><![CDATA[New comment by maille in "We replaced RAG with a virtual filesystem for our AI documentation assistant"]]></title><description><![CDATA[
<p>Are you using it? I will definitely give it a shot, any pointers to online resources will be appreciated</p>
]]></description><pubDate>Sat, 04 Apr 2026 11:28:31 +0000</pubDate><link>https://news.ycombinator.com/item?id=47638094</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=47638094</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47638094</guid></item><item><title><![CDATA[New comment by maille in "We replaced RAG with a virtual filesystem for our AI documentation assistant"]]></title><description><![CDATA[
<p>Let's say I want a free, local or free-tier-llm, simple solution to search information mostly from my emails and a little bit from text, doc and pdf files. Are there any tool I should try to have ollamma or gemini able to reply with my own knowledge base?</p>
]]></description><pubDate>Fri, 03 Apr 2026 17:57:51 +0000</pubDate><link>https://news.ycombinator.com/item?id=47629841</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=47629841</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47629841</guid></item><item><title><![CDATA[New comment by maille in "Ask HN: Read‑only LLM tool for email triage and knowledge extraction?"]]></title><description><![CDATA[
<p>So do you have a tool to suggest?</p>
]]></description><pubDate>Tue, 10 Mar 2026 23:19:03 +0000</pubDate><link>https://news.ycombinator.com/item?id=47330022</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=47330022</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47330022</guid></item><item><title><![CDATA[Ask HN: Read‑only LLM tool for email triage and knowledge extraction?]]></title><description><![CDATA[
<p>I’m looking for an LLM‑powered tool to help organize and query my email, but with a strict boundary: it must not write, send, or modify anything.<p>Standard filters miss the nuance of "middle-ground" emails, and existing AI email clients demand full inbox control or try to automate too much.<p>What I’m looking for:<p>-  <i>Strictly read‑only triage</i>: Categorizes incoming mail (e.g., “important”, “maybe useful”, “likely promo”) with zero write permissions ie. no sending, moving, deleting, or auto-archiving<p>- <i>Knowledge retrieval</i>: Builds a searchable knowledge base from scattered threads. I want to ask natural-language questions like: “What were all the decisions about Project X across the last 3 months?”<p>- <i>Follow-up & action item detection</i>: surfaces email threads that have stalled and need a nudge, while extracting explicit tasks and deadlines buried in long chains<p>- (optional) <i>Privacy-first</i>: Ideally runs locally or lets me plug in my own API keys to keep the data secure<p>Has anyone seen a tool that focuses entirely on this read-only, triage + retrieval workflow?</p>
<hr>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=47317429">https://news.ycombinator.com/item?id=47317429</a></p>
<p>Points: 2</p>
<p># Comments: 4</p>
]]></description><pubDate>Mon, 09 Mar 2026 23:53:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=47317429</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=47317429</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47317429</guid></item><item><title><![CDATA[New comment by maille in "Show HN: Beautiful interactive explainers generated with Claude Code"]]></title><description><![CDATA[
<p>What was the source of inspiration for Claude? I skimmed through the text and it does not look too bad, but devil is in the details and I need more time to go throught the fine prints. One remarks is that Young slit experiment could show what happens with a single slit vs 2</p>
]]></description><pubDate>Wed, 18 Feb 2026 19:02:24 +0000</pubDate><link>https://news.ycombinator.com/item?id=47064848</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=47064848</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47064848</guid></item><item><title><![CDATA[New comment by maille in "Show HN: Beautiful interactive explainers generated with Claude Code"]]></title><description><![CDATA[
<p>Would make one piece on optical interferometry? Inalways struggle explaining this with simple terms when asked to.</p>
]]></description><pubDate>Wed, 18 Feb 2026 07:57:10 +0000</pubDate><link>https://news.ycombinator.com/item?id=47058448</link><dc:creator>maille</dc:creator><comments>https://news.ycombinator.com/item?id=47058448</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47058448</guid></item></channel></rss>