work lab - Moozonian Search

yaugenst.github.io• Aug 11, 2026• 1 min read

Show HN: Advect – autodiff for NumPy and Array API programsI built Advect, an automatic differentiation library for scientific Python. It provides broad NumPy API coverage, follows the Python Array API standard, and works with ordinary Python functions.In addition to dynamic differentiation, Advect can stage a function into a reusable program that can be optimized, differentiated again, serialized, and loaded later. There is also some support for SciPy, xarray, and interoperability with other autodiff frameworks.The linked playground runs entirely in the browser through Pyodide, so you can try tracing and differentiating an expression without installing anything.The first public release is available with pip install advect.Source: https://github.com/yaugenst/advect

news.ycombinator.com• Aug 11, 2026• 1 min read

Ask HN: Have you used LLMs to reinvigorate you niche programming community?I've seen a few discussions over the years about how niche languages like Lisp, Forth, Prolog, etc. would be great to use, but even for personal projects the lack of available packages makes them hard to use. However, with an LLM you can have it replicate whatever package you'd like from a more popular language. You can even port software to your favorite niche architecture. If you really want to run your homelab off a PA-RISC machine you just give an agent SSH access and tell it to make the software work.For myself, I have only done very little of this. I did add POWER vectorization to POV-Ray so that I can get a little more speed out of a machine that has more power than it needs for other things. Instead of letting it use something like Python or Go for small scripts I'll tell the LLM that the more logical choice is Prolog. But mostly I have been using mainstream languages because the code is being generated and, therefore, I'm not playing with it, so I'm interested in hearing if ot

fui-rs-demo.effindom.dev• Aug 11, 2026• 1 min read

Show HN: Replacing HTML/CSS/JS with RustI've previously launched FUI-AS (retained UI in AssemblyScript) as a browser-only canvas framework. However, the dream was always something substantially bigger. It's to have a retained mode UI framework for the web, native and iOS/Android with SDKs available in multiple languages.This is the second slice of that dream.FUI-RS (retained UI in Rust) now runs the same retained application model (via the same EffinDOM runtime as FUI-AS) in browsers (including mobile browsers, so it has extensive touch gesture support / event system) and as native desktop applications without embedding a browser, thus no Electron/WebView. Major OSes supported are Windows (D3D12), macOS (Metal) and Linux (Vulkan). Comes with packaging tools for their respective OSes too (MSIX, DMG and AppImage).In the FUI-RS demo, have a play around to see and feel what a typical web app looks like with FUI-RS. Try: - VoiceOver on macOS (yes, we have extensive semantic roles / customization support in the SDK) - bring up the

nitpicker.dev• Aug 10, 2026• 1 min read

Show HN: Nitpicler. I was quoted $1M for AI PR review – so I bulit it myselfHi HN. I wanted an AI reviewer on every PR at work. Quotes came back around $1M/year, almost entirely per-seat licensing. I built a true open source alternative. We use it internally at Postman and the cost is ~$300/m in token usage. It's open source, self hosted (also supports AWS lambda, cloudflare workers or Github actions) and can be deployed in under a minute. It is lightweight (only diff is sent to the LLM), support any LLM provider and scalable. Our busiest repo has thousands of PRs a day and we usually get PR reviews within 30 seconds.Give it a try and let me know what you think - https://nitpicker.dev

news.google.com• Aug 10, 2026• 1 min read

When will NFL Network and RedZone be available on Xfinity after end of carriage dispute? - Yahoo SportsWhen will NFL Network and RedZone be available on Xfinity after end of carriage dispute? Yahoo Sports

news.google.com• Aug 10, 2026• 1 min read

Neuro Experience Undergraduate Research Opportunity Program (NEUROP) Aims to Open Research Opportunities for Undergraduates Interested in Laboratory Work | Neurophotonics Center - Boston UniversityNeuro Experience Undergraduate Research Opportunity Program (NEUROP) Aims to Open Research Opportunities for Undergraduates Interested in Laboratory Work | Neurophotonics Center Boston University

news.google.com• Aug 10, 2026• 1 min read

Ryan Grubb: This Alabama WR is ‘one of the hardest workers on the team’ - AL.comRyan Grubb: This Alabama WR is ‘one of the hardest workers on the team’ AL.com

github.com• Aug 9, 2026• 1 min read

Show HN: Lumabri – What if LLMs worked like Napster?A while ago I started working on Colibrì to see if it was possible to run huge LLMs on a normal computer. The project grew far beyond what I expected, thanks in large part to the HackerNews community.That led me to a new question:What if we stopped thinking about one computer?This is the idea behind Lumabri.Instead of requiring a single machine to store and run an entire huge model, Lumabri treats a network of normal computers as a shared pool of resources.One machine might provide disk space, another compute, another a different part of the model. If a required block or expert isn’t available locally, the system can retrieve or execute it on a peer.This is particularly interesting for Mixture-of-Experts models. A model can have hundreds of billions of parameters, while only a fraction are activated for each token. Rather than moving huge expert weights over the network, Lumabri can send the small activation to a peer that already has the expert and let it execute it.The goal is for ma

hnhiring.azuanz.com• Aug 9, 2026• 1 min read

Show HN: HN Hiring – Search and Filter Who Is Hiringhttps://hnhiring.azuanz.comI originally built this to scratch my own itch while job hunting.The main problem for me with the monthly Who Is Hiring thread was finding relevant jobs by location. I was mainly looking around Europe, and most posts are understandably US focused.I’ve since found a new job, so I don’t really need it myself anymore, but I thought it might still be useful to others.The crawler is written in Go, but the final site is just a single HTML file plus JSON files, served through Cloudflare Pages. The crawler runs every 10 minutes on an Ubuntu VM on my existing Synology homelab.The extraction/classification part was also a bit of an experiment for me.I primarily use Sol at work, but I also had a GLM subscription that I picked up during their Christmas deal last year, so I wanted to see how well it handled both coding and this kind of structured classification task.Right now the extraction is done using sequential GLM 5.2 calls with reasoning disabled. I tried batching m

news.ycombinator.com• Aug 9, 2026• 1 min read

Show HN: Xprem: Self-hosted OTA updates and observability for Expo appsHi HN,I'm the maintainer of xprem (previously expo-open-ota).It's an open-core, self-hosted solution to manage OTA updates for Expo apps.The server implements the expo-updates protocol (https://docs.expo.dev/technical-specs/expo-updates-1/).The project started more than a year ago because I needed to ship a business app inside a highly secured network, so I had to run an "on-premise" OTA server for apps without open internet access.Since then the project has grown a lot, in particular with the addition of expo-observe support, so you can get your logs/metrics directly in your own server, stored in your own ClickHouse DB.I also spent a lot of time on performance: in a recent benchmark we simulated traffic from 1 million MAU on a single vCPU, with a p95 < 20ms at 1000 req/s (telemetry enabled).Benchmark: https://xprem.dev/benchmark (the data and the script to reproduce it are available in the repository).Website: https://xprem.devRepository: https://github.com/mercuretechnologies/xpremDo

news.ycombinator.com• Aug 9, 2026• 1 min read

The State of Google"Google Earth Pro on desktop will no longer be available for download beginning June 25, 2027."This is the message I received today while opening Google Earth. It was in the bottom left corner of the desktop application.What is happening at Google?Why is it that each of the tools I have been using for the last 20 years are being destroyed one by one? Tools that have become constant in people's lives.What are they trying to accomplish by destroying the trust gained by a generation of employees delivering the top product? Products that just worked.Did they fire all the good employees?Where are the men and women with long-term visions?What is happening???

apisubstack.com• Aug 8, 2026• 1 min read

Show HN: I updated my unofficial Substack Posting APIA couple of days ago, I released my version of an unofficial Substack API. The main difference between it and all the other solutions available on the market is that it can actually post content, create tags, add images, save drafts, and so on. In other words, it is specifically a posting API. It is not designed to fetch posts or search for content — it is designed to publish something.The API is focused on allowing users to use it for publishing on Substack. It includes a CLI and an MCP, meaning the API is fully ready to work with AI agents.However, when users started signing up, I realized that most of them were not technically advanced users. For them, it was unclear what CLI, MCP, or APIs were, what keys were, where to add them, and so on.So I decided to improve the project. I added one-click synchronization, allowing any blog from any website to be fully synchronized with Substack. All you need to do is add a link to your website, and the synchronizer will automatically find all t

martini.film• Aug 7, 2026• 1 min read

Show HN: FLUX.3 Video 1k generation test (180 mins of footage)I've always been a fan of Black Forest Labs' FLUX image models. Specifically how "un-AI" their look is. (One might even describe it the opposite of Midjourney which feels like it excels at a surreal look.)Seedance 2.5 is likely taking the spotlight today but I wanted to do a deep dive into FLUX.3 so it can get some attention.My general observation is that FLUX.3 video is the most "clay-like" video model I've played with.Specifically: - Adding and rerunning video prompts feels additive/subtractive and not completely random generations (I feel much more in control of the medium). - "Draft" and "Enhance" modes (rerunning generations with the same seed) feel like getting to work with clay before committing in a kiln. Certainly a faster and cheaper way of working with generative video.Excited to share results here.The canvas has all generations with the prompts included.Spent ≈$1,000 on this test.Disclosure: I'm the co-founder of Martini (W26), which is where I ran the test

github.com• Aug 7, 2026• 1 min read

Show HN: reverse-engineer secure-enclave fingerprint scanner to work on LinuxI own a Xiaomi Mi Pad 5 Pro 5G Android tablet, which is also pretty-much a perfect Linux tablet: has 5G support, great screen, nice folio keyboard, 8 speakers, a pen, is lightweight and relatively powerful. All of its hardware is supported at this point, although not all has been mainlined yet.It also features a Goodix SPI fingerprint scanner, except it is implemented using an app living in Qualcomm's QSEE secure enclave which handles the scanner, and which Android communicates with using a HAL and a protocol that are not documented. This 2021 article explains the problem in details: https://emainline.gitlab.io/2021/12/12/fingerprint_P1.htmlIt took me about a week of intense work of reverse-engineering it with Opus 5 and Kimi K3, and further 5 days on polishing the documentation, but I can safely say this is now done.I used Opus 5, since Fable naturally dropped out as this is security subject, and Kimi K3 for reviewing the code. I resorted to both disassembling and tracing live Android

github.com• Aug 7, 2026• 1 min read

Show HN: Certo – An open source platform to deliver Open BadgesCerto is open-source infrastructure for issuing, managing, verifying, and exchanging digital credentials.It implements Open Badges 3.0[1] and W3C Verifiable Credentials[2] which are the open standards that make credentials portable, machine-verifiable, and vendor-independent.It is designed to be:- Self-hosted: run it inside your own infrastructure, air-gapped network, or sovereign cloud - Standards-compliant: credentials it issues work with any OB3/VC-compatible system - API-first: everything the UI can do, the API can do - Exportable: your data is always yours, in open formats - Extensible: a small, stable core with a plugin interface for everything elseA demo is available at https://certo.schroedinger-hat.org/ which we're using for Schrödinger Hat[3] conferences, workshops, partner communities.I'll be around if you have any questions. Feel free to give a feedback on the project.[1] https://www.imsglobal.org/spec/ob/v3p0 [2] https://www.w3.org/TR/vc-data-model/ [3] https://schroedinge

news.google.com• Aug 7, 2026• 1 min read

‘Chaos In The Magic Kingdom:’ Disneyland ‘Threatens To Take Away’ Paid Leave From Its Costumed Characters Who Work Full-Time & Part-Time As Their Union –-Actors’ Equity - Continues To Bargain A New Contract - WNY Labor Today‘Chaos In The Magic Kingdom:’ Disneyland ‘Threatens To Take Away’ Paid Leave From Its Costumed Characters Who Work Full-Time & Part-Time As Their Union –-Actors’ Equity - Continues To Bargain A New Contract WNY Labor Today

news.google.com• Aug 7, 2026• 1 min read

Licensable picture: Aug. 7, 2026; Tuscaloosa, AL, USA; Defensive lineman Nolan Wilson works against a blocking sled during practice for the Alabama - reutersconnect.comLicensable picture: Aug. 7, 2026; Tuscaloosa, AL, USA; Defensive lineman Nolan Wilson works against a blocking sled during practice for the Alabama reutersconnect.com

bing.com• Aug 7, 2026• 1 min read

Sensitivity Labels, DLP und Insider Risk Management müssen bei Microsoft Purview technisch und organisatorisch zusammenspielen. Ein Workshop vermittelt praxisnahes Wissen dazu.Sensitivity Labels, DLP und Insider Risk Management müssen bei Microsoft Purview technisch und organisatorisch zusammenspielen. Ein Workshop vermittelt praxisnahes Wissen dazu.

news.ycombinator.com• Aug 6, 2026• 1 min read

Ask HN: Can trivial LLM calls be compiled into conventional data pipelines?We are investigating whether recurring LLM workloads can be replaced, where appropriate, by automatically constructed pipelines of regexes, deterministic parsers, traditional ML and NLP models.As an example, suppose an application repeatedly asks a frontier model to read an annual report and return all customer–supplier relationships as structured records containing a customer, supplier, and supporting evidence.A possible replacement pipeline might run named-entity recognition, entity normalization, candidate generation, entity linking, relation extraction, and schema validation. A calibrated uncertainty or out-of-distribution gate would use the pipeline for inputs inside its validated domain and escalate other cases to the original frontier model.NER → entity normalization → candidate generation → entity linking → relation extraction → schema validationOur current action space is a taxonomy of 41 atomic task types spanning classification, token and span labeling, structured extraction

bing.com• Aug 6, 2026• 1 min read

How to Evaluate an AI Video Platform: A Practical Framework for TeamsLearn how to evaluate an AI video platform with a practical framework covering features, scalability, security, collaboration, integration, output quality, and pricing to help teams choose the right ...