jsiepkes

"I see AI and super intelligence [..]"

"super intelligence", well someone has bend a knee apparently.

Exactly. We are currently in a transition phase (of about a year iirc) where, for example, the fee will stay fixed. After that it will be a percentage of the total amount. Which is why Mollie announced[1] they would be launching a Wero alternative.

In the end this whole "buyer protection with percentage fee" is about the banks thinking: "We don't want to be a dumb digital sock for people's money. We have to provide value to them". But I want my bank to be a dumb digital sock to put my money in, nothing more. I'm really getting fat-up with all these MBA-types's thinking their company "needs to have a relation with me and add value". I mean I get emails from my utilities trying to get to know me. Seriously? I don't give a wooden nickle about my drinking water supplier. I just want to pay them and use their services and other than that I want to think about them as little as possible.

[1] https://www.accountant.nl/nieuws/2026/10/betaalbedrijf-molli...

Thats simply not true. If I want to open an account I can go to a Rabobank office and open an account. There are still plenty of Rabobank offices around. I can also access all features (even pay with Wero / iDeal) via my Rabobank scanner on my desktop via their website.

So we had something good in the Netherlands with iDeal, then they had to make it EU global and they messed it up.

For starters there is going to be percentage commission on each payment instead of just a flat fee low price like iDeal has. For some reason they felt compelled to introduce buyer protection like a credit card has, instead of just making it a digital debit payment system. And now we are also forced in to the Google ecosystem.

Great. Just great.

Why do they even feel they need the Google Play integrity API anyway? If you are really so concerned about the kind of attacks that supposedly fixes, you better ban Windows, macOS and Linux desktops and notebooks from banking online. Because their security model doesn't even come close to what Android or iOS has.

> So, nobody fuzz tested sshd before?!?!

OpenBSD probably did (which doesn't use musl), RedHat probably did (or some other distro using glibc), but they all don't use musl.

I doubt NVidia wants thriving second hand market, so I would guess grind to dust?

This about the risk of getting sued by parties with deep pockets. Antrophic is close to an IPO so that is definitely not the kind of attention they want.

> I'm not really buying that whole safe/unsafe narrative. Every language is safe if you're using it right.

Microsoft and Chromium each reported that roughly 70% of their serious security bugs were memory safety bugs. That's not a problem of beginners misusing the language. Like I said, just take a look at the Android security bulletins and count how many of these come from memory safety issues and how these numbers have improved in recent years with migrations to Rust. PHP last week? Memory safety issue.

The amount of high impact vulnerabilities coming from memory unsafe languages C and C++ proves there are too many people who "can't use it right". And these people apparently include very skilled people working at companies like Microsoft and Google on high profile projects.

We wouldn't need seatbelts if everyone would just drive safe. We wouldn't need hard hats at a construction site if everyone would just work safe.

That's kind of a weird list of languages. C and C++ are both memory unsafe languages. Every Android patch round about half of the vulnerabilities are memory safety bugs in C or C++ applications. Both Google and Microsoft independent from each other concluded that most vulnerabilities are caused by memory safety issues. Zig is also not fully memory safe and Go has a garbage collector just like Java. The Go garbage collector situation is probably worse than Java because it only has a single implementation and you have no choice. Meaning it had better fit your scenario or otherwise your screwed.

> would find any discussion based on statistically correct data of immigrants "dehumanizing"

You think the term "muselmanen" could reasonably be used in a civilized "discussion based on statistically correct data of immigrants"? You do know "muselmanen" has a very specific link to WWII concentration camps, right?

Kinda weird the say things like "the team managed to recover", have another article about the recovery and calling it "incredible" and then mention nowhere the recovery was performed bij Boskalis ( https://en.wikipedia.org/wiki/Boskalis ).

Freedom of information law's in almost all countries are about being able to access administrative documents so there can be accountability and transparency. The court ruled that a 3D scan is in this case not an administrative document.

And let's be real here; A ruling that a 3d scan of a piece of art is a "administrative document" would have been far weirder.

> How else they would force new API adoption?

And that would have been fine, if Google hadn't tried to shove the Google Play Store ecosystem down everyone's throat by throwing up barriers for side-loading apps and such.

Best selling electic truck on the market? Doesn't seem like it: https://electrek.co/2026/01/13/ford-f150-lightning-outsold-t...

And the biggest customer is SpaceX: https://www.businessinsider.com/spacex-bought-tesla-cybertru...

They didn't want their tech to be used to, I don't know, blow up a school with 120 children. But did want to allow other kinds of usage. Is that so weird?

Isn't "show stopper" a bit dramatic? What is the big problem with an SSH agent? If security is a problem you should probably store your SSH keys on a HSM like a Nitrokey or Yubikey. In which case you need an agent.

Google does this as well with Google Workspace with administrators. It's super annoying.

Not checking the "audience" of a token or misconfiguring it is pretty common. A lot of applications don't actually check it.

They backtracked on "ManBearPig" (climate change) a long time ago.

But not everything needs to be directly exposed to the internet. Framework had their data leaked because their metabase instance was hacked with a zero-day. Why was it directly exposed to the Internet? Why not require the use of a VPN like a Wireguard based solution or Nebula for these "internal" kind of apps?

It's in the article; They sold gold in NYC and with that money bought gold in London.

There are apps I cannot install via the Play Store in GrapheneOS, only via Aurora store.

> fear-mongering is not helping your cause

So we started with an OS (when Google still needed to gain market share and goodwill) which was opensource. Now Google is no longer developing Android source code in the open and decreasing the "opensourcing" to a couple of source drops each year (the interval has been decreased multiple times). Google Play was introduced (basically clamping down the app store ecosystem) and from there Google has only made it harder to get around Google Play. Introducing things like phone attestation. Supposedly a security measure, yet for some reason a Samsung phone which hasn't had an update in 4 years still passes basic level attestation. And now Google is yet again making it harder to get around the Play store.

And for some reason you seem to think it is reasonable to assume it is going to stop here?

Well, ask yourself: "Why am I on Hacker News?". You could just "as easy" use free / opensource alternatives like Lemmy, Mastodon, etc.

> and yet America voted for him a second time.

Well, a lot of them didn't vote for him. Roughly a third of eligible voters stayed home (actually the largest group) while Trump and Harris each got a bit under a third, with Trump narrowly ahead.

But yeah, I guess its still baffling about 31% voted for him.

There is literally a "Black Mirror" episode about this kind of thing.

> Everything else is essentially all the work of Google and (to some extent) Samsung.

There is plenty in Android which isn't the work of Google. For starters the Kotlin implementation and the Java implementation (OpenJDK).

I assume we are talking about DHH here? You're giving him too much credit. He is not a founder of 37 signals.

If an agent fires up NPM, takes a boatload of memory, is that memory released back to the OS after NPM shuts down in the VM?

Same here. I use this utility to make it easier on Linux to run Claude in Podman: https://github.com/mismosmi/ai-pod