Hacker News
points by Deeg9rie9usi 3 days ago | 0 comments
[–] vrganj 2 days ago
> Røsdal ignored the question about human code review and replied that he would appreciate feedback from someone else: ""Preferably from someone who actually works on real software development in 2026, someone with education in computer science, also considering that software development has changed with the introduction of AI llms"".

Yikes. What an asshole.

[–] scott01 2 days ago
The response to that was just gold. An excerpt:

> I decided to run a little experiment. So I asked a frontier model developed by my employer ...

[–] otabdeveloper4 2 days ago
Lots of "AI influencer" jackasses pretending to be software developers and/or engineers nowadays.
[–] ronsor 2 days ago
"AI llms" is an awkward phrase I've never seen before.
[–] throwaway27448 2 days ago
[–] 0rzech 2 days ago
That thread is like the "«AI-era»" version of the "MongoDB is web scale" meme video. It's absolutely worth reading all the way from the beginning to the end, especially given the plot twists. ;)
[–] wofo 2 days ago
Ugh, that's nasty. Pushing slop down the throats of Debian and Arch maintainers is _not good_.
[–] roryirvine 2 days ago
...especially when combined with an attempt at social engineering that's so clumsy that it comes close to being outright bribery!
[–] wofo 2 days ago
This quote from the mailing list is spectacular:

> If the Northstar browser is included in Debian Linux, I will do my utmost to make the browser successful and ensure it works well in Debian, and I will make sure that the Debian project and Debian contributors like you, and the four of you in particular, will be rewarded.

[–] ordu 2 days ago
It looks like the author is a socially clueless person, who was touched by an AI and realized they could write a code. No experience of shipping the code or a feel for what is acceptable behavior.

If it is so, it is a good reason to not include their browser into major distros, but in any case their advice "try to maintain it for longer and get some reputation first" make my (hypothetical) psychological portrait redundant. I mean, if I'm right, they would probably lose interest and/or completely alienate distros in a year or so.

[–] krapp 2 days ago
Unfortunately, "try to maintain it for longer and get some reputation first" is no longer feasible. Most software written now is vibe-coded by people with little or any experience writing code, much less shipping it, nor any respect for the culture they are supplanting. Reputation no longer comes from the old guard, it comes from hype entirely within the LLM community around projects that are pushed directly to prod without so much as a glance from a human being.

Even Linus is adding LLM generated code to the Linux kernel. No one cares anymore, that world is dead.

[–] 12_throw_away 2 days ago
> It looks like the author is a socially clueless person

xz-utils was 2 years ago now. It's surprising to me that people are still willing to apply any presumption of good faith and authenticity to interactions like this, especially after:

- "begging and praying" to include an unknown project in the distro

- offering what for all the world looks like an extremely clumsy bribe

- pushing a project full of contradictory and blatantly untrue claims

We don't have to - and in fact absolutely should not - evaluate projects based purely on "technical merit"! The people behind the project matter at least as much as "code quality."

And we absolutely should not be presuming good faith on the part of random people on the internet who have gotten ahold of our email addresses! Ask to see some credentials before turning your wallet over to the Wallet Inspector!

[–] ordu 2 days ago
> It's surprising to me that people are still willing to apply any presumption of good faith and authenticity to interactions like this

1. I don't think that that attempt for open bribery can mean anything except social cluelessness. Their intentions doesn't matter, because their actions as they are can achieve only a rejection of his package. We can try to find a hypothesis that would assume that the effect of their words is exactly what they wanted it to be, but I think it would be too much into the opposite territory, the territory of paranoia, "ahh all people are trying to kill me", something like that.

2. I think you are confusing a presumption of good faith with a presumption of a basic courtesy. I can be very suspicious, but (if I were a debian maintainer) I want those maillists to be functional. And it means basic courtesy and a presumption of good faith. There is no need to accuse the other side of what you suspect them to explain your suspicions to them and everyone watching.

And if my words seem to you like I have no suspicions or ignore them, well... I'll treat it as an evidence of how good I am.

If you read my message, it doesn't presume bad faith, but it doesn't presume good either. It focuses on observable behavior, proposes a hypothesis about personal traits, but it touches their motivation only in passing. "they realized they can write a code" can be seen as a description of a motivation, but for me it is not, because it doesn't explain their particular way to apply their new abilities to write code. Why browser, why to package it for debian? Why they are so insistent on it?

[–] ordu 2 days ago
edit. Ah, I forgot, there was a weak attempt at humor too. You see, the person is definitely socially clueless, but they try to social engineer their way into debian. I was laughing while reading their communications, because if they are of a bad faith, they are hopeless.
[–] c22 2 days ago
Hey! Who are you? The credentials inspector? Now hand over that wallet...
[–] qphe95 2 days ago
I personally think its pretty nasty how they treated this single developer, they could just reject him instead of writing a long hit piece on him.
[–] roryirvine 2 days ago
That would be a fair reaction if the author had stopped as soon as Debian had explained to them why their behaviour was so risky. But they didn't - they went on to do the same thing to Arch.

At that point, it became hard to write it off as mere misguided enthusiasm and LWN were being entirely responsible to write it up in the way they did.

In fact, if you search HN you can see that it's been pushed repeatedly here as well: https://news.ycombinator.com/from?site=github.com/nordstjern...

There's a pattern of behaviour on display which, though it's not entirely clear if it was intended to be malicious, does end up looking remarkably like a clumsy replay of the Jia Tan attack.

[–] augment_me 2 days ago
How do you otherwise make an example of shit you dont want in your codebase?

There is an attention cost of pushing unverified slop onto others, public shaming and bans is a valid way to prevent this behaviour.

[–] qphe95 2 days ago
Does this actually prevent this behavior? Does writing a hit piece somehow magically disable ai slop in the debian inbox or do certain contributors just enjoy going on an ego trip?
[–] nkrisc 2 days ago
In what way is it a hit piece?
[–] otabdeveloper4 2 days ago
> developer

Yes if by "developer" you mean "Claude prompt typer".

[–] totallygeeky 2 days ago
Those security holes seem damning for a project that was assuring everyone it was ready for prime time. Flatly unacceptable for something that they're actively lobbying folks to get on board with.

Maybe less time running tokens to develop the engines and more time running security passes before running to tell the world.