News for hackers. Stuff that matters.

exceptione

See also: Ant: Can unprecedented revenue growth outrun irresponsibly large investment? <https://news.ycombinator.com/item?id=50326281#50335632>
Re: on Monday November 02, 2026 @10:36AM (#50338285)
As archive.is doesn't work for me, here is the full text: https://pastebin.com/KnCqQumq
Ant = Anthropic, not the Chinese Ant. Good write up. There seems to be an inconsistency in Anthropic's justification of their spend on model training. According to them, there is a feed back loop of model training to become more efficient (with statements indicating that 90% of the company's code is AI-written), while at the same time the R&D expenses (which covers the training of new models) make an enormous jump. That is a bit weird. If you expect exponential growth, the exponent matters, not the base. Maybe Anthropic thinks that gobbling up all existing hardware via e.g. leasing allows them to iterate faster, thus increasing the exponent. But to me it seems this mainly allows for increasing the base (bigger models). I would expect that improving your models means that they become smaller and more efficient, not that they grow in parameter size. So it seems like Anthropic thinks that they need to brute force intelligence by creating increasingly heavier models (which indeed might suit their business model), whereas from a customer point of view increasingly leaner models are more 'fit' in llm evolution.
Re:It wasn't AI that flipped Proton's charge on Sunday November 01, 2026 @03:34PM (#50329516)
Don't know if Proton has any shareholders, but I they do, I hope they understand the CEO's position has become untenable, no matter his other real or imaginary qualities. It is (surprisingly) not uncommon for CEO's to have no clue about the business they are leading, but most of them at a minimum understand they should not spook their own customers. If Proton wants to restore trust, they need a leader that at least does not openly trash the company's own offerings.

Aside, with age becomes a deeper understanding of how realistic the Dilbert cartoons are.

Re:Introducing Kwi, a new KDE terminal emulator on Sunday November 01, 2026 @08:48AM (#50325674)

  > "Nerd Font icon scaling"
That would be a real improvement over Konsole.

  > "Additional optimizations for running efficiently when containerized (e.g. as a Flatpak)"
I never thought about running the terminal emulator itself in a bwrap session, but I am eager to learn what those optimizations entail.
Re:OpenIDEA: The IDE you know, rebuilt in Rust on Sunday November 01, 2026 @05:17AM (#50324196)
Interesting project. It seems it uses CodeMirror [1] in the Tauri frontend (60% of total code is typescript), but uses rust in the backend (30%). Surprising to see that Tauri is apparently good enough for the three platforms it needs to support (or it uses a lot of hacks, I don't know). The advertised memory reduction is remarkable as I expected a full web view to be a bit memory heavy. Not sure how this IDE keeps up with large documents or a lot of opened views though.

___

1. The editor implementation is here: <https://github.com/l0ng-ai/openidea/blob/main/src/editor/cor...>

Congratz! And thanks for testing!
I know people hit the 10K really fast. 10K is the limit for which obsidian switches algorithms (personally, I don't understand how 10K db rows would be a problem, but ok). It is very easy to run into 100k:

- external sources like the documentation for your programming languages, frameworks etc that come as markdown

- if you store every entity as a note, like you do with obsidian bases. Think task tracking, habit trackers, book collections

- clippings, daily notes, project notes, zettels etc

So if I would design a search improvement, I would test the 100k, 200k, 500k, 1m mark with synthetic data.

So what figure exactly is tested for 10k+? Is it 12k, 100k or more like 1m? I couldn't find.

  > The malware ships preinstalled in the device firmware, and we found multiple system packages involved, depending on the device. It’s on the phone before the owner switches it on for the first time, and it can’t be uninstalled.

  > The malware runs with system-level privileges that allow it to silently install and remove apps, grant permissions, and load arbitrary code supplied remotely. This essentially means its operators could install and delete apps at will, tuning each device to their needs, including making them part of large botnets.

  > This scheme is likely designed mainly to generate revenue. The operators carry out ad and click fraud, collect device and installed-app information, and turn infected devices into residential-proxy relay nodes, making them zombies in botnets. This creates an additional revenue stream, as access to botnets for DDoS attacks is a hot commodity. The larger the botnet, the more money they can charge.
Agreed that a website should be fine, but (Dutch) banks require the app as a second factor to log into the website. Fido2 hardware tokens could replace the bullshit requirement of "choose Apple or Google to be your overlord in order to use our service".
Re:ThinkRail from JetBrains on Saturday October 24, 2026 @08:02AM (#50220471)

  > Only 5% of code will ever need to be hand typed if that.
You are talking about the 'vibing' demografic. Not all of them would use JetBrains, as not all programmers did before the advent of LLMs. But for the time they do have to be in the code, JetBrains could be their best option. They are in a unique position to cover their existing offerings in a gold finish.

  > Pretty sure following this advice would kill them really fast.
I don't think so. There is no money to be made in llm wrappers. Their market might have shrunk, sure, but on the other hand they should be able to bring down their expenses significantly while at the same time bringing their backlog to zero. If not, they have nothing to fear from AI and the hype will be over some day.
Re:ThinkRail from JetBrains on Saturday October 24, 2026 @06:43AM (#50219978)
JetBrains should focus on the artisanal engineers and the 'vibing' engineers who need to code by hand from time to time. They have a big moat already. I imagine they try to keep the user base from fleeing their platform, but I think they have already covered prompt integration. There is imho no value in yet another LLM wrapper that seems to be totally unrelated to the existing offering, at least no monetary value.

I would urge JetBrains to put all their LLM power and human brains at work to slash their IDEs backlog and deliver the GOAT of all IDEs. Building IDEs is their strength and there will always be a market for that category.

  > that it was the Russians who hacked the phone and the Dutch got a copy of the contents 
The Russian spy gave access to the GROe server, where a copy of the contents lived. Later on, in the beginning of 2018, the MIVD got direct access to the phone.
Re:What I learned from six months of being stupid on purpose on Wednesday October 21, 2026 @01:30PM (#50184606)
You are over-complicating things; you can be gay without a barbell on your chest.
Re:PHP 8.5.12, 8.4.27, 8.3.36 and 8.2.35 out on Wednesday October 21, 2026 @01:27PM (#50184556)
Those websites can easily run in a php5 container, whilst the host os is up to date.

Websites depending on an older framework like Zend Framework or libraries like Doctrine 1 do not have a clear upgrade path and are stuck at php5 or have to basically start from scratch. php5 to the latest php8 is just a jump too big.

  > The paradox of tolerance is a philosophical concept suggesting that if a society extends tolerance to those who are intolerant, it risks enabling the eventual dominance of intolerance, thereby undermining the very principle of tolerance. This paradox was articulated by philosopher Karl Popper in The Open Society and Its Enemies (1945),[1] where he argued that a truly tolerant society must retain the right to deny tolerance to those who promote intolerance. Popper posited that if intolerant ideologies are allowed unchecked expression, they could exploit open society values to erode or destroy tolerance itself through authoritarian or oppressive practices.
https://en.wikipedia.org/wiki/Paradox_of_tolerance
Re:America's $10T AI infrastructure boom is starting to fray on Sunday October 18, 2026 @01:31PM (#50143268)
Do you have an un-walled version? Content is not accessible.
Re:Dotnetup on Friday October 16, 2026 @07:55AM (#50117282)

  > "All SDKs, runtimes, and metadata live entirely within your user directory."
Make sure your user directory has enough space then. I am not sure when I need this when Arch packages work so flawlessly, but I guess that users of slacking distro's would welcome this option to install an up to date version of the .net runtime/sdk.
Let me introduce you to sports. (table tennis and curling should get a free pass though)
fyi, "Klaer Ligh­t­ende Spiegel der Ver­fkon­st" is old Dutch for "Clear Mirror of the Art of Paints", literally translated.
Wish code editors would support different fonts per semantic region. Jetbrains, if you are reading this, implement this: <https://youtrack.jetbrains.com/issue/IJPL-29552/Allow-differ...>.

The heading of Iosevka doesn't do the font right, my first thought was that this would be yet another soulless sans-serif. But it has tons of interesting variants. Hats of to Iosevka Etoile. See also <https://monaspace.githubnext.com> if you like big versatile font families.

Re:Fingerprints Are the New Cookies on Monday October 12, 2026 @06:45AM (#50054394)
The solution has to come from law. I propose we handle privacy invasion the same as rape. Body and mind aren't up for grabs. That the average internet user is unconscious isn't a valid excuse.

It would be great if a wave of lawmaking started in the USA, because most tech behemots operate from there. I would expect EU to take the lead though IF such a wave would start. The EU has still a fair share of christian/conservative MP's that keeps doing the bidding for companies by pushing chat control, but on the other hand also has some progressive and knowledgeable representatives doing great work, so there might be a glimmer of hope. We need media campaigns and public education though. Political leaders are selected for their political abilities and as such rarely have leadership qualities. If the electorate moves, the leaders will follow it. So we have to start at the bottom.

A weird thing: the exclamation mark `!` seems to be elevated from the baseline, judging from the specimen.
If the demos don't work, your browser works as it should because it is blocking third-party cookies and the site does not serve valid CORS headers for ev.shaders.com.
Re:Cram – A Compressed RAM Service on Tuesday October 06, 2026 @06:56AM (#49976716)
Link to slides at the bottom. Impressive read-speed, roughly operates at the same speed as native DRAM. Orders faster than zstd. See also:

- https://www.phoronix.com/news/Linux-CRAM-Compressed-RAM

- https://lwn.net/Articles/1059885/

Re:Effect 4.0 on Saturday October 03, 2026 @03:45AM (#49942202)
Ah thx, this is a rewrite of React's Context. `userName` would have been so much better than `value`. I had to keep the possibility in mind that `value` is a kind of implicit magic variable in atoms.

Not sure what benefit this atom stuff brings. In react I can see that `value` is an argument, which from an fp point of view is a clear pro.

Re:Effect 4.0 on Friday October 02, 2026 @03:25PM (#49937483)
Can someone explain why this is a good idea? There is a ton of magic here, since the value string is not even part of the UserName function signature. Dependency injection is not about hiding your dependencies. On a first look, this seems to be really bad, but I might be totally misunderstanding this library.

    import { make, useAtomValue } from "@effect/atom-react"
    import { Atom } from "effect/reactivity"
    import * as React from "react"
    import { renderToStaticMarkup } from "react-dom/server"

    const User = make((name: string) => Atom.make(name))

    function UserName() {
        const atom = User.use()
        const value = useAtomValue(atom)
        return React.createElement("span", null, value)
    }

    export function App() {
    return React.createElement(
        User.Provider,
        { value: "Ada" },
        React.createElement(UserName)
        )
    }

    renderToStaticMarkup(React.createElement(App)) // => "<span>Ada</span>"
(src = https://effect.website/docs/v4/api/atom-react/ScopedAtom)

Older »