HackerTrans
トップ新着トレンドコメント過去質問紹介求人

i-con

no profile record

投稿

FlyTrap disables autonomous targeting drones with an umbrella

heise.de
2 ポイント·投稿者 i-con·4 か月前·2 コメント

Wiretapping scandal: Spyware manufacturer sentenced to prison in Greece

heise.de
2 ポイント·投稿者 i-con·4 か月前·0 コメント

Digital liberation: EU Parliament calls for detachment from US tech giants

heise.de
6 ポイント·投稿者 i-con·6 か月前·1 コメント

"AI eats Software": Why SaaS stocks are crashing on Wall Street

heise.de
4 ポイント·投稿者 i-con·6 か月前·1 コメント

CVE-2026-0915: GNU C Library Fixes a Security Issue Present Since 1996

phoronix.com
11 ポイント·投稿者 i-con·6 か月前·1 コメント

"Digital cash" without privacy is meaningless (2025)

medium.com
6 ポイント·投稿者 i-con·6 か月前·1 コメント

Expert: EU Commission wants an "unlimited special legal zone" for AI

heise.de
2 ポイント·投稿者 i-con·7 か月前·0 コメント

Cloud: Row in Bavaria over billion-euro Microsoft contract without tender

heise.de
9 ポイント·投稿者 i-con·7 か月前·1 コメント

Globalfoundries receives 495M euros for foundry expansion

heise.de
5 ポイント·投稿者 i-con·7 か月前·0 コメント

How a French judge was digitally cut off by the USA

heise.de
486 ポイント·投稿者 i-con·8 か月前·544 コメント

Analysis of the Digital Sovereignty Summit: Open-Source Gets Scolded

heise.de
3 ポイント·投稿者 i-con·8 か月前·0 コメント

EU Frequency Group: Upper 6 GHz Band Needed for 6G, WLAN Loses Out

heise.de
2 ポイント·投稿者 i-con·8 か月前·0 コメント

Ioema: Germany to be reconnected via subsea cable

heise.de
6 ポイント·投稿者 i-con·8 か月前·1 コメント

GEMA vs. OpenAI: Defeat for ChatGPT in Munich Court

heise.de
1 ポイント·投稿者 i-con·8 か月前·0 コメント

Apple puts App Store on the web – accidentally leaks code

heise.de
2 ポイント·投稿者 i-con·8 か月前·1 コメント

WLAN Rejection: Germany Favors Mobile for Upper 6 GHz Band

heise.de
3 ポイント·投稿者 i-con·8 か月前·1 コメント

Ask HN: Would open-source devs accept financial compensation for AI training?

2 ポイント·投稿者 i-con·8 か月前·2 コメント

コメント

i-con
·4 か月前·議論
Original publication: https://arxiv.org/abs/2509.20362
i-con
·5 か月前·議論
There's another way to explain the UB: IIRC, any value when stored to a _Bool is supposed to be converted to 0 or 1. The memset() bypasses this rule, boom.
i-con
·6 か月前·議論
Original resolution: https://www.europarl.europa.eu/doceo/document/TA-10-2026-002...
i-con
·6 か月前·議論
I'm not an expert, but roughly know the numbers. Usually with password-based key derivation, one would increase resource needs (processor time, memory demand) to counter brute forcing. Not an option for a human brain, I guess.

So the key would have to be longer. And random or a lot longer. Over 80 random bits is generally a good idea. That's roughly 24 decimal digits (random!). I guess about 16 alphanumerical characters would do to, again random. Or a very long passphrase.

So either remember long, random strings or doing a lot more math. I think it's doable but really not convenient.
i-con
·7 か月前·議論
I'm confused, is it bare metal or is it an EFI application? (bare metal used to mean that something can run without services, like those that UEFI provides)
i-con
·7 か月前·議論
Tested with WebKit and Gecko. Apparently the position gets fixed up at runtime if JavaScript is enabled. But why have dynamic elements with CSS if you need JavaScript to fix it?
i-con
·7 か月前·議論
No CSS is better than bad CSS

In my browser that "Page Contents" box is hovering above the end of the line, so I can't read the full text. Kind of ironic, that this is on w3.org
i-con
·8 か月前·議論
This, having the whole physical memory mapped all the time, reminds me of a another issue that was exploitable in KVM hypervisors [1]. I wonder what is the reason to have it all mapped? Not everybody seems to do it.

[1] https://www.vusec.net/projects/rain/
i-con
·8 か月前·議論
Hmmm, it's pretty clear but why the reminder? I read the article hoping to find some enlightenment, something that we can actually do (better) with this information in mind. But I still miss it. Does anybody know?
i-con
·8 か月前·議論
Release date 1993
i-con
·8 か月前·議論
I got to know UNIX (or rather Linux) about 12 years later. And TBH, I wasn't very impressed. I was like "oh, you have to do all that on the console". That's how green I was :D But then it caught me, and about a month later it was more "WOW, you can really do everything on the conole!"

What eventually helped me to really get into things was Linux From Scratch. If anyone wants to learn how a modern system works under the hood, and like those guys in the article know the very basic, minimal things that keep a system running, I can recommend it very much: https://linuxfromscratch.org/
i-con
·8 か月前·議論
Sorry, yes 2014. Fixed it.
i-con
·8 か月前·議論
SPARK tools are also open source. The main tool `gnatprove` is based on GCC as well. https://github.com/AdaCore/spark2014

It's not a community project, AFAICT. Few people know how to build it from source.
i-con
·8 か月前·議論
If you are looking for an open-source compiler, many distros (e.g. Archlinux, Debian and derivatives) bootstrap a full GCC (GNU compiler collection). Sometimes you have to install a particular packet, e.g. `gnat` or `gcc-ada`. There's also a language-specific packet tool `alire` that seems to aim to be somewhat like cargo. It can also install toolchains, IIRC.
i-con
·8 か月前·議論
Thanks! would be interesting to know how this is in other countries.

I don't think an implementation has to be about copyright infringement by the training itself, though. Like you say, it can also be about materials produced. And my question is really if people would like it, not how feasible it is to implement.
i-con
·9 か月前·議論
Are they right with the terminology? Do people consider mere bugs technical debt?

I haven't seen simple, visible problems being considered as technical debt so far. But rather only what the article calls architecture debt.

To me, it seems like the article indirectly proposes to use another term because the original one was used wrong too much. Do others see that, too?
i-con
·9 か月前·議論
Good point, I had no idea. I actually thought they just restrained themselves this time :D
i-con
·9 か月前·議論
Notice how they say "quantum advantage" not "supremacy" and "a (big) step toward real-world applications". So actually just another step as always. And I'm left to doubt if the classic algorithm used for comparison was properly optimised.