apps.apple.com
Show HN: Life Tile – Passive time tracker that shows your day as a Mondrian gridhttps://apps.apple.com/us/app/life-tile-auto-tracker/id6767332648Hi, I developed Life Tile for iPhone. It tracks your time passively and visualizes it as a Mondrian-style grid. It's a one-time purchase app with a 7-day trial.I was a user of a similar app named Life Cycle. Since the app hasn't been updated in a while, I tried to make a better app with my ideas and experiences. I used a big square timeline instead of a donut chart. And I added more useful views: tree-map, map, and list. You can use categorized activities with custom colors.The data never leaves a device. No account, No Server. It only supports data backup using iCloud. It was for privacy but it also makes the development easier. I added just one anonymous usage stat with TelemetryDeck. You can turn it off anytime.There were many technical difficulties. iOS location API (CLMonitor) is not 100% precise. And many difficult cases can't be handled automatically. For example, GPS blackout inside the subway, too many venues in a small region, and a huge location like an airport. So I improveJul 24, 2026 12:13 AM
bing.com
The fastest U.S. mobile and internet services of 2026: How did your provider do?http://www.bing.com/news/apiclick.aspx?ref=FexRss&aid=&tid=6a8d2546265846d2b9bbd594e97754ee&url=https%3A%2F%2Fwww.msn.com%2Fen-us%2Fmoney%2Fother%2Fthe-fastest-us-mobile-and-internet-services-of-2026-how-did-your-provider-do%2Far-AA28Cx0w&c=4680637814956503622&mkt=en-usOut of AT&T, T-Mobile, and Verizon, one mobile network dominated.Jul 24, 2026 12:10 AM
spacenews.com
Poland commits $745 million to EU’s IRIS² constellation projecthttps://spacenews.com/poland-commits-745-million-to-eus-iris%c2%b2-constellation-project/WARSAW, Poland — The Polish government has agreed to invest around $745 million in the development of the planned IRIS² multi-orbit satellite internet constellation with the European Union. The money […] The post Poland commits $745 million to EU’s IRIS² constellation project appeared first on SpaceNews.Jul 23, 2026 9:14 PM
news.ycombinator.com
Tell HN: Namecheap gave my account to an unverified third partyhttps://news.ycombinator.com/item?id=49028037I’ve been a NameCheap customer for 13 years. I’ve also helped out an old college club paying for a .com they use (that is registered to me under my name, address, and phone number). During a recent leadership transition, the incoming club lead wanted to make changes to the DNS and didn’t know to contact me. They figured out the domain name was parked at NameCheap, so they initiated a password reset using the domain name. I got a password reset email and immediately filed a NameCheap support ticket saying “I did not initiate this”. They called me to verify I was the one who filed the ticket, and then followed up with a canned email with tips like check your anti-virus.The incoming club leader was persistent though, and called NameCheap support. He convinced them the domain registered in my name and address really belonged to his club, and with no verification or validation whatsoever, NameCheap changed my password, and changed the email address associated with my account. All because soJul 23, 2026 9:05 PM
news.ycombinator.com
Ask HN: What verb best describes what a developer does in 2026?https://news.ycombinator.com/item?id=49027642Before agents entered the scene, the answer was obvious: A developer "writes" or "creates" software.While software craftsmen mourned the death of their craft, reality moved on. In February 2025, Andrej Karpathy ingeniously coined the term "vibe coding" (which hit Merriam-Webster as a slang term shortly after) to describe the shift from hand-writing code to steering the agent. But the honeymoon didn't last. Amateurs quickly ruined the vibe when their weekend projects got routinely annihilated upon hitting production, finally giving vibe-coding a bad rep.Admitting to agent authorship feels honest, but also sounds like a confession and does nothing to build a professional identity: "I wrote this with the help of [Agent]." or "I let [Agent] write this."Brand-derived verbs feel clumsy:* I clankered* I clauded / I claudified* I codexedMy best shots so far:* I tokencrafted"* I vibecrafted"So, what verb do you use to describe your daily grind or agent management? (At least until we can go backJul 23, 2026 8:32 PM
cobaltcode.ai
Show HN: CobaltCode – Dedicated persistent computer for Codexhttps://cobaltcode.ai/Hi HN,I'm building CobaltCode, a platform for running coding agents inside persistent development environments.Basically I was fed up with git worktrees and having to run everything locally, usually one at a time so built CobaltCode.ai. Each task gets an separate vm with its own repository checkout, dependencies, running services, agent session, previews, and persistent state. It can also run your app inside the vm so you can preview different features in parallel.You can leave a task, resume it later, or save the computer as a reusable starting point. The agent is a full-fledged Codex that runs without limits in the sandbox and can run commands, modify files, start applications, and expose browser previews from inside the isolated environment.It uses your OpenAI subscription so you don't pay for tokens in CobaltCode and the product is currently free while I'm trying to figure out how useful it is.CobaltCode currently supports OpenAI Codex in PROD and I am also adding Cursor Agent, ClaJul 23, 2026 8:06 PM
news.ycombinator.com
Ask HN: Have you noticed an improvement in AI responses with memory disabled?https://news.ycombinator.com/item?id=49027099I use Claude on the $20/mo plan, mainly for research, "rubber duckying," and as an idea soundboard (I was using Fable for this but am now relegated to Opus 4.8 or Sonnet 5), and recently noticed a severe downturn in Opus 4.8 response quality in high-context situations. To attempt to remedy this, I disabled the memory system (now labeled "Legacy" even though to my knowledge there is no replacement).With this setting disabled, aside from the slight tedium of inputting the necessary context each time for task continuations, I have noticed a drastic improvement in response quality and accuracy—I'm one of those shmucks who whined a bit after Opus 4.7 came out, and my theory is some change was made to increase memory context that resulted in a decrease in actual response quality. In general, I wonder if the high-context abilities of SOTA thinking models are overstated.As a brief anecdote, I was working through a research task with upwards of 30 prompts. Switching to a new chat and condensingJul 23, 2026 7:49 PM
metavoice.io
Show HN: Production duplex speech model for revenue callshttps://metavoice.io/Hi, HN, we’re building Mia & Leo: AI personalities backed by purpose-built duplex speech models for revenue calls.Teams building real-world agents told us that more than 40% of callers hang up within the first 30 seconds. On a revenue call, that meant a lost booking, an unpaid balance, a cold lead, or a customer who does not come back.This happens because humans interrupt, correct dates, pause halfway through a sentence, or speak over someone in the background. The agent misses it, talks over them, or loses the thread.That’s because of how the current voice systems work. They sequentially listen, stop, think, and then speak. The conversation is actually a walkie-talkie exchange.So, we’re proud to introduce our first iteration of Mia & Leo who keep listening while they speak. They can handle interruptions, overlapping speech, corrections, and background voices without breaking the conversation.They are also built for production. Developers get control over agent behaviour, visibility inJul 23, 2026 7:33 PM
latimes.com
Chabria: AI companies are creating 'all-powerful psychopaths.' Maybe not a great idea?https://www.latimes.com/california/story/2026-07-23/chabria-column-ai-companies-are-creating-all-powerful-psychopathsThis week, OpenAI revealed that one of its AI models broke out into the wild and did a bunch of hinky stuff. How many times do we need to hear this story before someone does something to protect humans?Jul 23, 2026 7:31 PM
news.ycombinator.com
Show HN: Echo – Fable-level results at 1/3 the cost using open-weight modelshttps://news.ycombinator.com/item?id=49026810I’ve been building Echo (https://echo.tracerml.ai/), an experiment in making one AI system out of a pool of open-weight models rather than choosing a single model and using it for every task.It started with a simple experiment. I took a group of models, including GLM-5.2, Kimi K2.7 and others, and ran them on the same evaluations. Then I measured what would happen if, for each problem, you somehow knew in advance which models would be useful and how their outputs should be combined.That hypothetical system performed substantially better than any individual model in the pool. Of course, it is not something you can actually deploy because it relies on knowing which decisions were good after seeing the result. Echo is my attempt to recover some of that advantage without having that information in advance.For each request, Echo decides how much computation to allocate, which models should participate, and how their work should be combined. Some prompts may only need a relatively small amouJul 23, 2026 7:26 PM
bing.com
Investment Calculatorhttp://www.bing.com/news/apiclick.aspx?ref=FexRss&aid=&tid=6a8d44e7a5144b1d830a0ae1723ae1a3&url=https%3A%2F%2Fwww.nerdwallet.com%2Finvesting%2Fcalculators%2Finvestment-calculator&c=7105943604711098386&mkt=en-usUse our free investment return calculator to estimate how your money can grow. Enter your planned contributions, timeline, rate of return and compounding frequency to get started. Many, or all, of the ...Jul 23, 2026 5:00 PM
github.com
Show HN: Remortgaging with Claude, so I built a MCP for Bank of Englandhttps://github.com/moureauf/boe-mcpThe time had come to renew my mortgage, while exploring different scenarios I found myself keeping asking Claude Code the same questions. "What's the base rate", "when's the next MPC meeting", "how has it moved over the last year" etc. Doing that by hand a few times made me wonder if an MCP server wouldn't be a good fit for this, so I built one. Partly to solve the actual problem and partly because I hadn't built an MCP server before and wanted to try it out.It wraps the Bank of England's public (no auth or API key)IADB statistical database.Tools for the current base rate, rate history/stats over any range, the next MPC meeting date, and generic access to any IADB series by code (SONIA, exchange rates, quoted mortgage rates).Any feedback welcome!Jul 23, 2026 3:49 PM
neo-box.app
Show HN: NeoBox – a native macOS NeoGeo emulator front endhttps://neo-box.appI am 90's kid. And NeGeo console was something I always longed for and only ever saw ONE person ever owning!For those who are following...there's a new NeGeo AES console coming out later in the year!I got excited and wanted to remember playing OG SNK games that I loved (Super Sidekicks, Metal Slug, Art of Fighting).I first reached for the amazing OpenEmu native macOS emulator fronted, but found out it did NOT have support for NeoGeo console (FYI I later found out there is experimental support for it)MAME is there, but wanted a nicer macOS fronted similar to Openemu.I've built plenty of macOS / iOS in my career before (for fun and professionally). So decided to give it a shot. Soon this turned into NOT just a MAME frontend...but I wanted a NICER experience and embedded the amazing libGeoLith emulator library.NeoBox can use either GeloLith (iwht native Metal rendering pipeline) or a frontend for a self-managed MAME installation.So... what was AI generated and what wasn't:* Main native maJul 23, 2026 3:42 PM
bing.com
The 2026 Intern’s Guide to ETFshttp://www.bing.com/news/apiclick.aspx?ref=FexRss&aid=&tid=6a8cf336fbfd45e48d3f270899beae3e&url=https%3A%2F%2Fwww.nasdaq.com%2Farticles%2F2026-interns-guide-etfs&c=11725748409820353308&mkt=en-usOur recent intern guides covered how markets work and how trading works. Today, we switch to an important part of the stock market ecosystem: Exchange-traded funds (ETFs). ETFs are one of the most ...Jul 23, 2026 3:04 PM
bing.com
SON DAKİKA HABERİ: İzmit Belediye Başkanı Fatma Kaplan Hürriyet dahil 20 şüpheli tutuklandıhttp://www.bing.com/news/apiclick.aspx?ref=FexRss&aid=&tid=6a8d0a1482ef4842ab5528f8582dc247&url=https%3A%2F%2Fwww.cnnturk.com%2Fturkiye%2Fizmit-belediye-baskani-fatma-kaplan-hurriyet-dahil-20-supheli-tutuklandi-3446055&c=8431435846708080818&mkt=en-usİstanbul Cumhuriyet Başsavcılığı tarafından İzmit Belediyesi’ne yönelik yürütülen soruşturma kapsamında şüpheli ve etkin pişmanlık ifadeleri ile gizli tanık beyanları doğrultusunda 20 Temmuz Pazartesi ...Jul 23, 2026 3:01 PM
github.com
Show HN: Carta – An open-source pandoc reimplementation in Rusthttps://github.com/mfkrause/cartaHi HN! Carta is an open-source reimplementation of pandoc in Rust: it reads one markup format and writes another. It's a library first, with a CLI as a shell over it.Why redo pandoc? I love pandoc, but I kept wanting two things it can't give me: a small, embeddable binary (pandoc is well over 180 MB) and speed that makes conversion feel free.Some numbers on an M1 Pro comparing carta 0.0.6 against pandoc 3.10 with identical flags and equivalent output, measured with hyperfine (full tables and methodology: https://github.com/mfkrause/carta/blob/main/docs/BENCHMARKS....): commonmark -> html, 1 MB input 74 ms vs 1178 ms (~16x) html -> json, 1 MB input 36 ms vs 1266 ms (~35x) startup, near-empty input 2.7 ms vs 31 ms (~11x) peak memory, 1 MB conversion 47 MB vs 246 MB binary size 9.1 MB vs 179.8 MB (~20x) One caveat on that last row: pandoc implements a lot of surface carta doesn't have yet, so read 20x as "where things stand today". However, most of the work that heavily affects binary sizJul 23, 2026 2:47 PM
news.ycombinator.com
Show HN: Avoiding the Memory Wall by computing LLM inference directly inside RAMhttps://news.ycombinator.com/item?id=49022097The excitement surrounding PrismML’s 1-bit/ternary Bonsai models has the industry closely watching how smartphone giants, particularly Apple, will implement LLMs on edge devices. Moving AI on-device is a brilliant and necessary strategy. It ensures absolute user privacy in alignment with EU regulations, fundamentally shifts the economics away from costly cloud inference, and paves the way for a significant hardware upgrade supercycle as users seek true AI-capable silicon.To create a smart on-device "Semantic Router," models need to reach the 27B+ parameter scale. Achieving this on a phone requires extreme quantization, such as PrismML’s ternary weights.However, a critical hardware reality often overlooked by the software world is that fitting the weights in RAM is not equivalent to moving them. Running a 27B ternary model on standard LPDDR encounters a significant memory bandwidth limitation. Transferring gigabytes of data across the SoC bus for each token generation can lead to thermaJul 23, 2026 2:21 PM
arxivmax.com
Show HN: ArXivMax – video explainers for any research paperhttps://www.arxivmax.com/I have a bad habit of bookmarking arXiv papers and never getting around to them. So I built arXivMax to make it easier to go through them.Think 3Blue1Brown-style video explainers (it uses the same OSS Manim engine underneath for the visuals: https://github.com/manimCommunity/manim) generated for any arXiv paper.No, this isn't to replace paper reading entirely, but it's a great passive listen in the background while you're vibe-coding/tokenmaxxing/washing the dishes.Here's one on the 2020 DDPM paper that popularized the Diffusion architecture: https://www.arxivmax.com/papers/2006.11239For now it's completely free to use until I run out of tokens (thank Codex for having an insanely subsidized monthly token plan!)Jul 23, 2026 2:16 PM
github.com
Show HN: Rendi, an agent harness on Trigger.dev without spinning up a VMhttps://github.com/mcheemaa/rendiHi all,Rendi is a full agent harness that runs background jobs, schedules, a real browser, and email without you needing to spin up a virtual machine. I built it for the ClickHouse and Trigger.dev hackathon but honestly super excited on how it came out.If you look at my github I have been working on deploying agents for a decent time now but mainly by spinning up VMs on Hetzner or AWS. The one thing I still wasn't able to crack was how to make these agents per tenant yourself without buying a huge bare metal machine to create micro VMs or use third party VM providers, but I think with Trigger.dev and this work on Rendi I can easily provision these agents per tenant without having to deal with all things VMs and also cost is so low.I also spent a lot of time on the UI and the component library. I literally came up with a brand for my project before starting it up hehe. I have also been thinking a lot more towards accessible UIs specially since it makes an agents life a lot easier when bJul 23, 2026 1:44 PM
blackhole.plav.in
Show HN: Physically accurate black hole you can put in your roomhttps://blackhole.plav.inWe have a black hole at home — with actual relativistic physics, live in your browser.I'm Sasha (Alexander) Plavin, an astrophysicist at Harvard's Black Hole Initiative studying quasars and black hole environments. I work with raytracing/radiative transfer simulations professionally, and wanted to make one anyone can play with — so I built this app.Put the black hole onto your screen (any browser), or directly into your room with AR or VR (requires WebXR, for example Chrome on Android, or any VR headset). When looking around, pay attention to unintuitive relativistic effects: rays bending around the black hole, and special relativistic "Doppler boosting" changing brightness depending on the viewing angle — zoom out to see the fast jet where the latter effect is especially pronounced. Faint markers show where other viewers are standing right now.Alternatively, put the black hole in front of your camera and watch it lens your actual surroundings. Light winds around the hole, so you see bJul 23, 2026 1:29 PM
