HN Morning Brief - March 04, 2026

Welcome to today’s Hacker News Morning Brief! Here’s a curated summary of the top 30 stories trending on HN right now.

Other

You can use newline characters in URLs

Score: 57 | Comments: 26

The WHATWG URL specification allows newline and tab characters in URLs after flagging a validation error and then removing them, so HTML with newlines in href attributes works in browsers. Other whitespace is disallowed, but data URLs go further by permitting arbitrary ASCII whitespace, enabling clean multiline SVG or base64 assets embedded directly in markup. These quirks show how parsers balance strictness with practicality, giving developers flexibility to keep code tidy.

Key Discussion Points:

  • Yeah, they might be ignored by the html parser and might “work”. Still, not a bright idea…
  • >Remove all ASCII tab or newline from input. the title is referring to inside html attributes, where they will be removed hence not affect where the link points…
  • After I read this, I started to look at the Wikipedia article on Base64 and eventually got to the article for the data URI scheme. That’s where I found a sentence that seems to a little bit at od…
  • Somewhat relatedly, GitHub Pages does support using URL-encoded newline characters %0A to reference file names with newlines,[0] but GitHub itself will omit the file from the web UI’s tree view. …

Textadept

Score: 116 | Comments: 21

Textadept is a lightweight, cross-platform text editor built around the Scintilla component, designed for speed and extensibility through Lua scripting. It offers unlimited split views and rapid performance, even in heavily monitored environments, making it ideal for quick notes and coding tasks. The editor’s simplicity and customizability attract users who want a fast, minimal environment without bloat. Its continued visibility on HN reflects appreciation for focused, efficient tools.

Key Discussion Points:

  • Glad to see this hit the front page again! I’ve been making some contributions lately and it introduced me to Lua. At its core it’s basically a very thin wrapper around the Scintilla editing component…
  • I downloaded it recently and found it to be quite useful for quick notes. And I can attest to its “fast” claim, using it on a heavily monitored corporate computer, with CrowStrike and what-n…
  • > Unlimited split views. Ok, well now I have to find out what hapoens if I get enough splits to make the width of each less than a pixel…
  • I’ve looked into TextAdept a few times. It appeals to me because it’s got a standard Qt UI, is fast and lightweight and highly customizable with Lua. But I could never commit the time to ful…
  • There is also NotepadNext which is oddly similar, sans the Lua: https://github.com/dail8859/Not…

Welcoming Elizabeth Barron as the New Executive Director of the PHP Foundation

Score: 19 | Comments: 6

The PHP Foundation has appointed Elizabeth Barron as its new Executive Director, bringing fresh leadership to organization that supports the PHP language ecosystem. With an annual budget around $500,000 USD—half allocated to developers—the Foundation plays a critical role in funding and coordinating PHP’s development. Barron’s appointment signals continued commitment to sustaining and improving PHP, a language powering a vast portion of the web. The community has welcomed move as a step toward greater inclusivity and focus.

Key Discussion Points:

  • Precise info about the PHP foundation is hard to find. They have an annual budget of ≈500k USD, half of it assigned to developers[^1]. From a 2024 presentation video, they claim that the foundation ha…
  • This is amazing news. More trans people means more inclusive programming communities…

MacBook Pro with M5 Pro and M5 Max

Score: 756 | Comments: 785

Apple introduced new MacBook Pro models featuring M5 Pro and M5 Max chips, delivering up to 4x faster LLM prompt processing compared to M4. The 14- and 16-inch laptops emphasize value for upgraders with 1TB SSD standard, while maintaining battery life and thermal efficiency. Benchmarks focus on AI workloads, leveraging a Neural Accelerator in each GPU core. Pro users get significant performance gains, especially for on-device machine learning and creative tasks.

Key Discussion Points:

  • I chased down what the “4x faster at AI tasks” was measuring: > Testing conducted by Apple in January 2026 using preproduction 13-inch and 15-inch MacBook Air systems with Apple M5, 10-co…
  • I love the following section of their copy: > Even More Value for Upgraders > The new 14- and 16-inch MacBook Pro with M5 Pro and M5 Max mark a major leap for pro users. There’s never been a bet…
  • “Scaling up performance from M5 and offering the same breakthrough GPU architecture with a Neural Accelerator in each core, M5 Pro and M5 Max deliver up to 4x faster LLM prompt processing than M4…
  • I typed “RAM” to search for it and boy they hammer home how lucky I am to be getting 1TB SSD standard, but no mention of RAM anywhere on this page. Anyway, the MacBook Pro starts with 16GB of RAM. It’…
  • I feel like Apple pulled an Instant Pot with the M1 MacBook Pro. I still haven’t had a single situation where I felt like spending more money would improve my experience. The battery is wearing o…

Intel’s make-or-break 18A process node debuts for data center with 288-core Xeon

Score: 276 | Comments: 236

Intel debuted its 18A process node with a 288-core Xeon data center CPU, using 12 chiplets on 18A stacked on Intel 3 base dies. The packaging story is notable: it’s essentially a small cluster on a single package, with 12 memory channels of DDR5-8000. Cache hierarchy groups cores into four-core blocks sharing about 4 MB of L2 per block, affecting aggregate last-level cache behavior. This core-density increase reflects Intel’s push for workloads that haven’t scaled in a year, targeting cloud and enterprise.

Key Discussion Points:

  • These sorts of core-density increases are how I win cloud debates in an org. * Identify the workloads that haven’t scaled in a year. Your ERPs, your HRIS, your dev/stage/test environmen…
  • I don’t quite follow: > From a cache hierarchy standpoint, the design groups cores into four-core blocks that share approximately 4 MB of L2 cache per block. As a result, the aggregate last-level c…
  • With packages like this (lots of cores, multi-chip packaging, lots of memory channels), the architecture is increasingly a small cluster on a package rather than a monolithic CPU. I wonder whether the…
  • I think everyone’s focusing on the core count, but the packaging story is way more interesting here. This thing is 12 separate chiplets on 18A stacked on base dies made on Intel 3, connected to I…
  • Helped a friend make a difficult career decision (cozy job vs something hard and new + moving to a new city) that ultimately ended up with him working on the project. Glad that happened. I love to see…

An Interactive Intro to CRDTs (2023)

Score: 128 | Comments: 22

Key Discussion Points:

  • I’ve spent a lot of time studying CRDTs & order theory in the last year & will publish an article too. Local-first apps are not easy to build, complex data structures (say, calendar repet…
  • Last-Write-Win CRDTs are nice, but I wish the article talked about where CRDT really shine, which is when the state truly converge in a non-destructive way, for example: 1) Counters While not really u…
  • Enjoyed the article! As someone who’s worked on bits of collaborative software (and is currently helping build one at work), I’d caution people from using CRDTs in general. A central server …
  • My absolute favorite kind of blog post and the same structure/style I use. Also a really well written piece…
  • I haven’t dug into this deeply, but to me CRDTs look like a P2P data structure abstracted to the programming language variable level. The article says they shine when you don’t want a centra…

My spicy take on vibe coding for PMs

Score: 68 | Comments: 55

Key Discussion Points:

  • Meta, and other large companies have been encouraging PMs to code, while I’ve seen many negative responses from engineers having to code review, debug, deal with production issues, etc. stemming …
  • Our job is done for. We will be shown the door, and everyone will rejoice. Everyone will live in a happy world where you’ll doddle a house and Claude will build you a next generation SaaS that ma…
  • As much as I recognize that a truly talented product manager is worth their weight in gold, I’d say the average engineer would be much more capable of learning to be an average PM than vice versa…
  • A friend at Meta — long before the age of LLMs — got paged at 3am for a site issue. When he found the PR that caused the bug, the testing section for the change simply said: YOLO! This was well into…
  • I don’t think this is a spicy take at all. A PM’s job is to prioritise, and the most important/high priority projects will naturally be handled by Engineers enabled with AI-coding workflows. The …

Motorola GrapheneOS devices will be bootloader unlockable/relockable

Score: 471 | Comments: 117

Key Discussion Points:

  • Not sure how I feel about this. Motorola seems to be the exclusive provider of encrypted cellular networks and associated devices to the Israeli military [1][2]. I’m under the impression that bas…
  • If anyone from Motorola is reading this: Please add a smaller device to your Portfolio, about max the size of a Pixel 8. I’m not hoping for an audio jack any more but at least small it could be. …
  • I’m glad to hear that. That means these devices will be a popular target, perhaps the popular target for alternative operating systems both Android-based and non-Android Linux…
  • If true. And I put a big if on that. I WILL be buying their flagship model. My go to for Graphene has been used Pixels from eBay. Because I can’t give money to Google in good conscience…
  • The biggest argument for me to buy one of these phones - when they actually arrive - next to running GrapheneOS, will be whether these phones, like all others, are way too big to use with only one han…

The largest acidic geyser has been putting on quite a show

Score: 44 | Comments: 1

Key Discussion Points:

  • Cool. Found my way to this updates page that has a summary of volcano and seismic activity from different observatories if you click the tabs: <a href=“https://www.usgs.gov/programs&#x2…

Mac external displays for designers and developers, part 2 (2022)

Score: 35 | Comments: 19

Key Discussion Points:

  • This is an older article: Published 1 April 2022. Not about the new Apple Displays…
  • I am using a Dell 4K monitor.https://www.amazon.in/dp/B0F29RSLHP?th=1 Works fairly well…
  • Which monitor is recommended for work + gaming these days? Been on the fence for a new one but couldn’t find any that is within my price range. The requirement I have is that it has to be 120Hz, …
  • The new Mac monitor supports 120hz. I have a 5k2k 44” OLED monitor from LG. Looks great on my Mac Studio…
  • LG DualUp monitor is a great monitor for developers. Vertical and wide enough to show all the toolbars and the code. Sadly they were not produced since 2024, but they are very good. BTW, if someone is…

Speculative Speculative Decoding (SSD)

Score: 37 | Comments: 5

Key Discussion Points:

  • Neat. Very similar to tree-based speculation as they point out, and they also point how to combine them. Speculative decoding: Sample a linear output (next n tokens) from draft model, submit it to a v…
  • This is interesting stuff. I wonder if these sorts of tricks are already in use at the big labs. Incidentally, I would recommend trying implementing speculative decoding yourself if you really want to…
  • > Our implementation is up to 2x faster than optimized speculative decoding baselines and up to 5x faster than autoregressive decoding with open source inference engines what about per-FLOP?…
  • Yo dawg I heard you liked speculation so we speculated your speculating…

California’s Digital Age Assurance Act, and FOSS

Score: 87 | Comments: 69

Key Discussion Points:

  • Counterpoint to peeps on this thread: * This approach is the most consistent with retaining anonymity on the internet, while actually helping parents with their issues. If any age-relevant gatekeepi…
  • What a crappy law. > Section 1798.500(e)(1) states: “Covered application store” means a publicly available internet website, software application, online service, or platform that distributes and f…
  • This is an intentionally vague law, and seems like the governor is more than happy to call for amendments: <a href=“https://www.gov.ca.gov/wp-content/uploads/2025/10&#x2F…
  • I haven’t made up my mind on whether I like this law or not, but this is a bigger condemnation of the FOSS community than anything else. This law was introduced over a year ago, it was reviewed by mul…
  • Copyright, patents, censorship, age controls etc… have never worked on kids. When it comes to technology, parents will always, always be years behind their kids. The kids will find a way to circumve…

Nobody Gets Promoted for Simplicity

Score: 173 | Comments: 67

Key Discussion Points:

  • I’ve gotten promoted for simplicity several times. “Why are we jumping through hoops to make an Xbase file on a network share be accessible to multiple users updating data at the same time? Uh, can I …
  • This hit a nerve because “simplicity” is one of those things you only notice when it’s missing, and most orgs don’t have a good way to reward “the thing that didn’t happen.” I’ve watched this exact dy…
  • I realized a long time ago that you don’t get promoted or recognized for the things you don’t do. And many times not doing something is the best option of all…
  • Early in my carrier, i saw code written by people both junior and senior. Every single time i saw a great code, it made me feel like its so simple. heck, i could have written it. It was completely opp…
  • This is just not true, people get promoted for delivering impact whether the solution is complex or simple. The best engineer I know who can work with huge complex systems in a big company usually sta…

Better JIT for Postgres

Score: 17 | Comments: 0

AI & Tech Policy

Mount Mayhem at Netflix: Scaling Containers on Modern CPUs

Score: 44 | Comments: 20

Key Discussion Points:

  • Okay, I’ll ask the dumb question: Couldn’t you also reduce the number of layers per container? Sure, if you can reuse layers you should, but unless you’ve done something very clever lik…
  • I am not familiar with the nitty gritty of container instance building process, so maybe I’m just not the intended audience, but this is particularly unclear to me: 
     > To avoid the …
  • Articles like this are pretty cool. It’s so interesting to see the behind the scenes that happens whenever we watch a Netflix movie…
  • Interesting, another case of removing HT improving performance. Reminds me of doing that on Intel CPUs of a few gens ago…
  • It took them this long to move from docker to containerd?…

Claude’s Cycles [pdf]

Score: 593 | Comments: 237

Key Discussion Points:

  • It’s fascinating to think about the space of problems which are amenable to RL scaling of these probability distributions. Before, we didn’t have a fast (we had to rely on human cognition) w…
  • I recall an earlier exchange, posted to HN, between Wolfram and Knuth on the GPT-4 model [1]. Knuth was dismissive in that exchange, concluding “I myself shall certainly continue to leave such re…
  • I didn’t expect such a misleading intro from Knuth. It reads like Claude solved Knuth’s math problem. In reality, Claude generated various example solution, and Knuth then manually generaliz…
  • > Filip also told me that he asked Claude to continue on the even case after the odd case had been resolved. “But there after a while it seemed to get stuck. In the end, it was not even able to wri…
  • I asked Claude to solve the pentominoes puzzle made famous by Arthur C. Clarke. It struggled mightily until I told it how I’d solved the problem using 64 bit unsigned integers to represent the bo…

Launch HN: Cekura (YC F24) – Testing and monitoring for voice and chat AI agents

Score: 78 | Comments: 20

Key Discussion Points:

  • The full-session evaluation framing is the right call - most teams don’t realize the failure happened in turn 2 until they’ve spent 3 hours blaming the model. One thing worth thinking about …
  • Any ideas how to solve the agent’s don’t have total common sense problem? I have found when using agents to verify agents, that the agent might observe something that a human would immediate…
  • How do you handle sessions where the correct outcome is an incomplete flow — e.g. the agent correctly refuses to move forwards because the caller failed verification, or correctly escalates to a human…
  • we treat each scenario as an explicit state machine. every conversation has checkpoints (ask for name, verify dob, gather phone) and the case only passes if each checkpoint flips true before the flow …
  • Was really fun building this - would love feedback from the HN community and get insights on your current process…

When AI writes the software, who verifies it?

Score: 208 | Comments: 207

Key Discussion Points:

  • > The Claude C Compiler illustrates the other side: it optimizes for > passing tests, not for correctness. It hard-codes values to satisfy > the test suite. It will not generalize. This is on…
  • The fundamental problem is the verification loop for the average developer is grounded not in tests, but with results. Write code, reload browser, check output. Does it work the way I want? Good. We&#…
  • I encourage everyone to RTFA and not just respond to the headline. This really is a glimpse into where the future is going. I’ve been saying “the last job to be automated will be QA” an…
  • I believe there is a Verification Complexity Barrier As you add components to a system, the time it takes to verify that the components work together increases superlinearly. At a certain point, the v…
  • Maybe I’m missing something, but isn’t this the same as writing code, but with extra steps? Currently, engineers work with loose specifications, which they translate into code. With the prop…

GPT‑5.3 Instant

Score: 337 | Comments: 264

Key Discussion Points:

  • The single biggest issue for me with ChatGPT right now is how absolutely awful it sounds in every answer. “Why it matters”, “the big picture”, “it’s not jut you”, th…
  • I’m a bit confused by this branding (never even noticed that there was a 5.2-Instant), it’s not a super fast 1000tok/s Cerebras based model which they have for codex-spark, it’s ju…
  • This malformed awful website redirects me to the French version no matter what I do. I even changed the url to en-US and it still does. Stop treating me like a child, OpenAI…
  • Since the page mentions: > Better judgment around refusals Has any AI company ever addressed any instance of a model having different rules for different population groups? I’ve seen many exam…
  • I kind of chuckled when I read the headline “GPT‑5.3 Instant: Smoother, more …” LLM companies starting to sound like cigarette advertisements…

LLMs can unmask pseudonymous users at scale with surprising accuracy

Score: 64 | Comments: 44

Key Discussion Points:

  • I thought this would be more about stylometry but it’s mostly about users literally posting the same identifiable information across multiple services, including in one example their age, dog nam…
  • There was a tool shared here that could show which accounts belong to the same person based on the writing patterns. Can’t remember the name, but it found my old accounts on HN pretty accurately…
  • Anonymous account unmasking represents a new threat to anonymity. not just this technique with llms, but the earlier text similarity one. But I think it would be generally easier to counter in the sam…
  • The internet is getting less interesting by the day…
  • As a 32 year old Ghanaian woman living in Luang Prabang and studying as an ophthalmologist, this gives me some food for thought!…

Giving LLMs a personality is just good engineering

Score: 13 | Comments: 1

Key Discussion Points:

  • ok but then why is ChatGPT’s personality so infuriating? “It’s not just X, it’s Y.” “Here it is, no extra text, no fluff.”…

Agentic Engineering Patterns

Score: 40 | Comments: 4

Key Discussion Points:

  • I’ve experimented with agentic coding/engineering a lot recently. My observation is that software that is easily tested are perfect for this sort of agentic loop. In one of my experiments I …
  • I find StrongDM’s Dark Factory principles more immediately actionable (sorry, Simon!): https://factory.str…
  • Why the Open Web Matters: A Claude Code Agent’s Case for Open Infrastructure https://news.ycombinator.com/ite…

Tech Tools & Projects

A pretty looking web for a quantum mechanics tool

Score: 10 | Comments: 1

Key Discussion Points:

  • You might try posting this under a “Show HN”, and/or reconsidering the title. This looks like an interesting and thoughtful tool with a clear purpose, and this title doesn’t convey…

On the Design of Programming Languages (1974) [pdf]

Score: 24 | Comments: 0

Show HN: Rust compiler in PHP emitting x86-64 executables

Score: 8 | Comments: 4

Key Discussion Points:

  • I wonder if the compiler runs on https://github.com/VKCOM/kphp (a PHP->C++ transpiler)…
  • Sometimes when doing offensive security work you end up in the strangest environments with limited tools, odd quirks, broken shells, and god knows what else. But you know what is almost always availab…
  • interesting proof of concept, in php, who would have thought :D…

Don’t become an engineering manager

Score: 349 | Comments: 252

Key Discussion Points:

  • I have been both an EM and a Staff software engineer at a bigtech. Both are great career choices but lately being an EM means spending over 6 hours a day in meetings and having very little agency over…
  • I cannot be alone in feeling that titles (within “tech” in particular) are almost completely arbitrary? What constitutes a “senior”, “lead”, “principal” and &qu…
  • If you are reading this and you are thinking you want to become an engineering manager, I urge you to think long term what you want that to look like. I’ve seen too often that developers who want…
  • There is a major gap in this analysis by not controlling for industry or companies. Engineering Manager is a very generic title, so this is going to get Start Ups, Big Tech, Little Tech, Enterprise, C…
  • While my friend was offered a bump with the promotion to EM, the total compensation was less than the offers he received for Senior/Staff Engineer at other startups. “At other startups”…

Voxile: A ray-traced game made in its own engine and programming language

Score: 170 | Comments: 47

Key Discussion Points:

  • The founder, Wouter, has created or helped design 10 programming languages. Voxile is built in his newest language: Lobster. Wouter has been a major contributor to WASM and LLVM while also inventing f…
  • I’m a long time Unity developer that in the past year picked up Godot. The speed at which Godot loads compared to Unity is staggering, it’s just so much faster. When I returned to Unity I ra…
  • Very cool - the post made me want to play the game, and check out lobster, but didn’t link to it - lobster is open source: <a href=“https://github.com/aardappel/lobster” rel=”…
  • Heh. I thought I remembered the name. I used to use Wouter’s E programming language on the Amiga. It was pretty good, as I recall…
  • Gorgeous. These are the graphics I wish Veloren[1] had. Maybe my machine is lacking the specs to dial up the graphics all the way… [1]: htt…

Graphics Programming Resources

Score: 71 | Comments: 7

Key Discussion Points:

  • I don’t know who posted it here. But this is not merged to the main website (it’s on the “develop” branch), and a lot of resources have not been added. I am still working on it. Cr…
  • I had hoped for some more basic stuff. I struggle for 2 months now to implement a fast line draw with width for a embeed cpu. It only has a framebuffer no gpu…

A CPU that runs entirely on GPU

Score: 48 | Comments: 14

Key Discussion Points:

  • Ya know just today I was thinking around a way to compile a neural network down to assembly. Matching and replacing neural network structures with their closest machine code equivalent. This is way co…
  • As foretold six years ago. [1] [1]: https://breandan.net/2020/06&#x…
  • This is a fun idea. What surprised me is the inversion where MUL ends up faster than ADD because the neural LUT removes sequential dependency while the adder still needs prefix stages…
  • Out of curiosity, how much slower is this than an actual CPU?…
  • Now I’ve seen it all. Time to die.. (meant humourously)…

Security & Privacy

TikTok will not introduce end-to-end encryption, saying it makes users less safe

Score: 159 | Comments: 100

Key Discussion Points:

  • Brilliant. They’re repackaging the argument governments have long made about E2EE being dangerous to children…
  • TikTok is a front for government surveillance, so it’s not really surprising that this is their position…
  • This might be off-topic but on-topic about child safety… but I’m surprised people are being myopic about age verification. Age verification should be banned, but people ignore that nowadays mos…
  • DMs are akin to private conversations in real life. Thus, every DM feature should entail E2EE. It’s ok for a platform to not feature private conversations. They should just have no DM feature at all, …
  • Why would you use TikTok for private communications anyway? It’s mostly a public short video sharing platform…

How Mossad hacked traffic cameras to track Ayatollah Khamenei

Score: 10 | Comments: 1

Key Discussion Points:

  • My link is paywalled - choose your own via google … How Mossad hacked traffic cameras to track Ayatollah Khamenei A network of operatives infiltrated the supreme leader’s inner circle. After months …

Geopolitics & War

Weave – A language aware merge algorithm based on entities

Score: 102 | Comments: 62

Key Discussion Points:

  • Some context on the validation so far: Elijah Newren, who wrote git’s merge-ort (the default merge strategy), reviewed weave and said language-aware content merging is the right approach, that he…
  • At this point, the question is: why keep files as blobs in the first place. If a revision control system stores AST trees instead, all the work is AST-level. One can run SQL-level queries then to see …
  • How does it compare to https://mergiraf.org/ ? I’ve had good experience with it so far, although I rarely even nee…
  • > This happens constantly when multiple AI agents work on the same codebase What? Is the idea of “multiple agents” of flesh and blood writing code that far fetched now?…
  • Interesting that Weave tries to solve Mergiref’s shortcomings (also Tree-sitter based): > git merges lines. mergiraf merges tree nodes. weave merges entities. [1] I’ve been using mergiraf…

History & Science

Number Research Inc

Score: 28 | Comments: 14

Key Discussion Points:

  • I’m not sure whether I’m taking too seriously something intended as a joke, but this in fact can conceivably be useful! When studying mathematical problems, sometimes you have a number that …
  • 67 has been searched 13k+ times, more than 69 and 420 combined Times are changing…
  • It seems that someone sequentially ran up to around 131k (at the moment), I can’t get any lower new number. Also please restore the input when a database error occurs…
  • The time it takes for the server to check whether the number exists is too long imo…
  • Some of the most searched numbers are surprising. Why are 8487798767697884826576, 119104105114108, or even 3551 so high up the list? See most searched here: <a href=“https://numberresearch.x…