Nvidia-smi hangs indefinitely after ~66 days
The article discusses an issue with the NVIDIA open-source GPU kernel modules, where users are reporting kernel panics and other problems when using the latest Linux kernel versions. The article provides steps for users to work around the issue and suggests that NVIDIA is working to address the problem.
Show HN: Bonsplit – Tabs and splits for native macOS apps
BonSplit is an open-source app that allows friends to easily split bills and track expenses. It simplifies the process of dividing costs and managing shared payments, making it a useful tool for groups and households.
Show HN: AutoShorts – Local, GPU-accelerated AI video pipeline for creators
This article discusses the development of a tool called AutoShorts, which automatically generates short video clips from long-form video content. The tool uses machine learning techniques to identify key moments and summarize the content, enabling efficient content repurposing and distribution.
Iran Protest Death Toll Could Top 30k, According to Local Health Officials
The article reports that senior Iranian officials have stated that more than 30,000 people have been killed in the ongoing protests in Iran, which began in September 2022 following the death of Mahsa Amini. The protests have led to a violent crackdown by the Iranian government.
The Responsibility of Intellectuals (1967)
The article discusses the role and responsibility of intellectuals in a complex, changing world. It explores the challenges intellectuals face in navigating political and ideological divides while upholding the values of free inquiry and critical thinking.
Show HN: TUI for managing XDG default applications
Author here. I made this little TUI program for managing default applications on the Linux desktop.
Maybe some of you will find it useful.
Happy to answer any questions.
Show HN: VM-curator – a TUI alternative to libvirt and virt-manager
I've long wanted to harness QEMU/KVM for my desktop virtual machines, but I'm befuddled by virt-manager's lack of support for working NVIDIA 3D acceleration, dogmatic embrace of ugly XML, and the puzzling UI decision of having to click what seems like 15 buttons to attach an ISO to a VM image. When I further learned that NVIDIA's broken 3D acceleration is the fault of libvirt as opposed to QEMU's virtio driver, I had an idea...
Behold, vm-curator! A fast and friendly VM management TUI written in Rust. You can create, configure, organize, and manage VMs directly with QEMU. No libvert. No XML. No wonky UI's. Just the right level of friendliness, customization, and speed to be really really useful.
The best part? 3D para-virtualization works with NVIDIA cards (via virtio-vga-gl!) No jumping through hoops to get GPU passthrough working!
(Disclaimer: This works great with other guest Linux VMs, but is not suitable for Windows gaming. If you want to game on Windows within a VM, passthrough is a must. vm-curator will have fast and friendly support soon.)
Looking for contributors (especially to help with the ascii art,) and donations are welcome. (Claude was a big help, but this was not a vibe-coded affair. We pair-programmed approx. 10,000 lines of code here. It was a great way to learn Rust, actually!)
Alarm overload is undermining safety at sea as crews face thousands of alerts
New research shows that crews at sea are facing an overwhelming number of daily alerts, up to tens of thousands, which is undermining their ability to respond effectively and posing a significant safety risk at sea.
TikTok is officially US-owned for American users, here's what's changing
TikTok is now officially U.S.-owned for American users, with the company's ownership shifting to Oracle and Walmart. This change aims to address national security concerns and bring the platform more in line with U.S. regulations.
What Ralph Wiggum loops are missing
The article discusses the concept of 'Ralph Wiggum loops' in programming, which are code constructs that appear to be nonsensical or inefficient but can actually serve a purpose. It explores how these loops can be used to create interesting and unexpected behaviors in software.
Hands-On with Two Apple Network Server Prototype ROMs
This article provides a hands-on review of two Apple Network Servers, exploring their features, performance, and suitability for different use cases. The author examines the servers' hardware specifications, software capabilities, and overall functionality, offering an in-depth, objective assessment.
ICE Is Not Like the Brownshirts, the Brownshirts Identified Themselves
The article draws a comparison between ICE (Immigration and Customs Enforcement) and the Brownshirts, the paramilitary group of the Nazi party, arguing that the Brownshirts were more identifiable than ICE agents who often operate in plain clothes and without clear identification.
Deaths and deportations of citizens by Trump administration
The article discusses cases of American citizens who were subjected to deaths, detentions, or deportations during the second term of the Trump administration, highlighting concerns about due process and civil liberties.
Show HN: Sightline – Shodan-style search for real-world infra using OSM Data
Hi HN,
I built *Sightline*, a Shodan-style search engine for *physical-world infrastructure*.
Shodan makes it easy to explore exposed internet services. Sightline applies the same idea to the real world, using OpenStreetMap as the data source.
You can search things like:
* “telecom towers in karnataka” * “power plants near mumbai” * “data centers in paris france”
or use structured queries:
* `type:telecom operator:airtel region:karnataka` * `type:data_center operator:google`
Sightline:
* uses Overpass API for querying OSM features * uses Nominatim for resolving countries, regions, and cities * avoids hardcoded geography * uses deterministic, rule-based parsing (no AI inference)
Repo: https://github.com/ni5arga/sightline Try it out: https://sightline-maps.vercel.app
Show HN: Notes from building a startup, shared openly
Hi HN,
I started a small newsletter to document what I’m learning while building projects — the decisions, mistakes, and experiments that usually stay behind the scenes.
It’s not a growth hack or a polished media thing. Just short write-ups about what worked, what didn’t, and how I’m thinking through problems as I go. I’ve only published a couple of posts so far, but it’s grown to 1,000+ readers, which surprised me.
Sharing here in case it’s useful to others who enjoy building in public or learning from real-world attempts rather than theory.
Happy to hear feedback or ideas on what would be most useful to write about next.
The case against ultrasonic humidifiers
This article discusses the benefits and drawbacks of using humidifiers, including their role in maintaining healthy indoor air quality, the potential risks of overuse, and guidelines for proper humidifier maintenance and use.
Bridging the Gap Between PLECS and SPICE
The article compares the capabilities of PLECS and SPICE, two popular circuit simulation tools, highlighting their strengths, weaknesses, and suitable use cases. It provides a detailed analysis of the performance, ease of use, and feature set of each tool to help readers choose the most appropriate option for their circuit design needs.
Because Coordination Is Expensive
The article discusses the costs and challenges of coordination in complex systems, highlighting that coordination is often an expensive endeavor that can limit the ability of systems to adapt and respond to changes. It emphasizes the need to consider the tradeoffs between coordination and flexibility in the design and management of complex systems.
Show HN: Lumina – Open-source observability for LLM applications
Hey HN! I built Lumina – an open-source observability platform for AI/LLM applications. Self-host it in 5 minutes with Docker Compose, all features included.
The Problem:
I've been building LLM apps for the past year, and I kept running into the same issues: - LLM responses would randomly change after prompt tweaks, breaking things - Costs would spike unexpectedly (turns out a bug was hitting GPT-4 instead of 3.5) - No easy way to compare "before vs after" when testing prompt changes - Existing tools were either too expensive or missing features in free tiers
What I Built:
Lumina is OpenTelemetry-native, meaning: - Works with your existing OTEL stack (Datadog, Grafana, etc.) - No vendor lock-in – standard trace format - Integrates in 3 lines of code
Key features: - Cost & quality monitoring – Automatic alerts when costs spike or responses degrade - Replay testing – Capture production traces, replay them after changes, see diffs - Semantic comparison – Not just string matching – uses Claude to judge if responses are "better" or "worse" - Self-hosted tier – 50k traces/day, 7-day retention, ALL features included (alerts, replay, semantic scoring)
How it works:
Start Lumina
git clone https://github.com/use-lumina/Lumina cd Lumina/infra/docker docker-compose up -d
// Add to your app (no API key needed for self-hosted!)
import { Lumina } from '@uselumina/sdk';
const lumina = new Lumina({ endpoint: 'http://localhost:8080/v1/traces', });
// Wrap your LLM call const response = await lumina.traceLLM( async () => await openai.chat.completions.create({...}), { provider: 'openai', model: 'gpt-4', prompt: '...' } );
That's it. Every LLM call is now tracked with cost, latency, tokens, and quality scores.
What makes it different:
1. Free self-hosted with limits that work – 50k traces/day and 7-day retention (resets daily at midnight UTC). All features included: alerts, replay testing, semantic scoring. Perfect for most development and small production workloads. Need more? Upgrade to managed cloud.
2. OpenTelemetry-native – Not another proprietary format. Use standard OTEL exporters, works with existing infra. Can send traces to both Lumina AND Datadog simultaneously.
3. Replay testing – The killer feature. Capture 100 production traces, change your prompt, replay them all, get a semantic diff report. Like snapshot testing for LLMs.
4. Fast – Built with Bun, Postgres, Redis, NATS. Sub-500ms from trace to alert. Handles 10k+ traces/min on a single machine.
What I'm looking for:
- Feedback on the approach (is OTEL the right foundation?) - Bug reports (tested on Mac/Linux/WSL2, but I'm sure there are issues) - Ideas for what features matter most (alerts? replay? cost tracking?) - Help with the semantic scorer (currently uses Claude, want to make it pluggable)
Why open source:
I want this to be the standard for LLM observability. That only works if it's: - Free to use and modify (Apache 2.0) - Easy to self-host (Docker Compose, no cloud dependencies) - Open to contributions (good first issues tagged)
The business model is managed hosting for teams who don't want to run infrastructure. But the core product is and always will be free.
Try it: - GitHub: https://github.com/use-lumina/Lumina - Demo video: [YouTube link] - Docs: https://docs.uselumina.io - Quick start: 5 minutes from `git clone` to dashboard
I'd love to hear what you think! Especially interested in: - What observability problems you're hitting with LLMs - Missing features that would make this useful for you - Any similar tools you're using (and what they do better)
Thanks for reading!
The Uncomfortable Truths About Immigration
The article discusses the complex and often contentious topic of immigration, exploring the nuanced economic, social, and political realities surrounding it. It presents a balanced perspective, acknowledging both the potential benefits and challenges of immigration, and encourages critical thinking on this multifaceted issue.