Hacker News: 0x_rs

New comment by 0x_rs in "JPEG XL vs. AVIF for web use-cases"

0x_rs — Sat, 17 Oct 2026 13:28:39 +0000

Browsers do not use hardware decoding with the exception of Apple for HEIC, and it's implausible any will ever do so for AVIF. Chromium tried to, and never managed for simple lossy WebPs. It'd be limited to a subset of images with a matching profile, have significant overhead for most of those "web-friendly" potato images people make pointless comparisons for, and any implementation would not come without downsides. Nobody is willingly shouldering the cost of dealing with all the broken hardware implementations out there. So it's difficult to see it as a possible advantage.

New comment by 0x_rs in "I guess I'm using WebP now?"

0x_rs — Mon, 12 Oct 2026 08:35:59 +0000

The problem really boils down to WebP having no purpose to exist for any user, therefore no reason to waste time and introduce vulnerabilities for its support, especially for feature-complete software. An image format (actually not, one of those video keyframes formats that have little to do with images, let alone photos, and in its case it's 2-in-1 too with lossless), that was already outperformed by mozjpeg on its initial release as far as human eyes are concerned, and thoroughly obsoleted by jpegli..

New comment by 0x_rs in "I could've accessed 17T Microsoft records"

0x_rs — Wed, 30 Sep 2026 15:49:56 +0000

It's not a "normal" sentence and is quite clearly produced by an LLM, it's a typical Claudeism so probably that. The entire post is also flagged by Pangram, so OP is correct.

New comment by 0x_rs in "Show HN: Real-time Solar System with 526k asteroids and all tracked satellites"

0x_rs — Wed, 30 Sep 2026 11:49:18 +0000

Middle dot (·) is an obvious telltale. Not as if it detracts from the work, but UI makes it quite clear.

New comment by 0x_rs in "KIDS Act"

0x_rs — Thu, 17 Sep 2026 20:43:23 +0000

In the page: "This can be done via the EU age verification tool or other third-party solutions that offer reliable, secure and privacy-preserving means to verify the user’s age". It's well demonstrated by now "third-party solutions" do not work, and do not have any privacy guarantees. Even banks that must abide to supposedly the most strict regulations will surrender all your personal data to anyone asking with a government email, see the latest Revolut leak. What about this "EU age verification tool?"? It's the one you may have heard about that requires you to use an Android device passing Play Integrity or iOS [0]. What would be the rationale for the EU to research and implement a truly anonymous, transparent age verification system? Even ZPKs cannot do that, and most likely won't ever. [1]

This insane proposal that has now been approved will bring an unprecedented level of mass surveillance China critics a decade ago wouldn't have even believed. Combined with other heinous ideas such as ChatControl, desperately, repeatedly pushed for earlier this year until it passed again, Europol and other surveillance lobbying groups must be salivating at the prospect of having total control over all citizen's life, regardless what the made-up excuse to implement such regulations is. "Europol Sought Unlimited Data Access in Online Child Sexual Abuse Regulation" [2]

0. https://github.com/eu-digital-identity-wallet/av-doc-technic...

1. https://balkaninsight.com/2023/09/29/europol-sought-unlimite...

2. https://www.eff.org/deeplinks/2026/08/zkps-arent-age-verific...

New comment by 0x_rs in "Europe's "Less" Is Doing More Than Anyone Gives It Credit For"

0x_rs — Sat, 12 Sep 2026 16:44:32 +0000

CTRL+F "manufacturing": 0 results for a website called "oilprice", instead it has this line: "An economy that produces more while burning less isn't stagnating…it's just getting more efficient, and there's no headline number for that, so it mostly doesn't get written about.", which is false, these are different things. EU economy shifted from manufacturing to services in the past decades and especially after the GFC (from which it never recovered in some sectors), losing millions of those jobs in the process, and an output decline in the 20% range, France leading the way with the industrialization collapse. Services are inherently less energy sensitive than manufacturing and there's no question the latter, already struggling, hit another wall in 2022 because the continent has zero energy independence. If you cannot produce your energy your manufacturing will always be at the mercy of macroeconomic events, and your consumption will always depend on another country's production. It takes 20 seconds to generate an AI article full of fabrications but orders of magnitude more to properly correct it..

New comment by 0x_rs in "How Europe is killing makers and micro-entrepreneurs"

0x_rs — Mon, 24 Aug 2026 14:18:26 +0000

>Again, EU-level politics keep assuming everyone (corporations and states) are adults. Unfortunately, everyone are just a bunch of screaming, scheming petulant children.

One in four MEPs are involved in scandals [0] ranging from harassment to organized crime and corruption. The European Parliament also remains a free out-of-jail card through abuse of parliamentary immunity, and they do abuse it [1]. EU politics is a circus.

0. https://archive.is/neTOG

1. https://www.politico.eu/article/eu-parliament-accused-overre...

New comment by 0x_rs in "How Europe is killing makers and micro-entrepreneurs"

0x_rs — Mon, 24 Aug 2026 13:53:30 +0000

It's typical eco-suicide: give eurocrats more power and fake work to do, pretend most recycling regulations aren't scams, and practically destroy trade between countries in the union because you need one representative in every single country you want to sell anything in. As if adding a 5 (soon to become 7€) fee (plus tax) for every line of item (not package) outside the union wasn't enough. It is also quite obvious both must've been heavily lobbied by major retail groups through EuroCommerce.

Feedback for this utterly deranged regulation can be seen below, over 5k entries by many businesses in EU:

https://ec.europa.eu/info/law/better-regulation/have-your-sa...

New comment by 0x_rs in "Anthropic's 'watermark' text adulteration in Claude is a perversion of writing"

0x_rs — Mon, 17 Aug 2026 00:19:49 +0000

I'd encourage reading this paper, and literature on scaling laws in autoregressive models: https://arxiv.org/abs/2303.11156

Total variation distance has been measured to decrease as you scale a model, and that is the primary mechanism "watermarking" as discussed in the Anthropic announcement relies on. It becomes more difficult to reliably detect text as a fixed sample count without tweaking the distribution further. Either way, it's a minor problem that will be addressed over time, compared to the issue of who can detect this without guessing or developing their own sets: providers not releasing a way to detect any such watermarks without going through them makes this entire approach hostile to the public. The EU regulation on this subject is interesting, although again most certainly not the primary driver for these practices:

"1.1.2: Signatories will ensure that AI-generated or manipulated content is marked with an imperceptible watermark, with the exception of very short text. For free-form text longer than 200 tokens, watermarking still needs to be applied, even though it may have lower reliability compared to that of watermarking very long text"

A proper, effective and useful law would have required providers to regularly release datasets to run your own verification on any text released within a fixed interval of time, presumably once out of rotation. Instead, it only talks about exposing an user interface going through their own services:

"Signatories will ensure access to their detection solution through a user interface appropriate for the audience of end-users that may eventually be exposed to the content generated or manipulated by their AI system. [...] Any restriction to the access will be limited in time until more reliable and robust detection mechanisms have emerged and have been adopted as the state of the art for detection mechanisms for the watermarking of free-form text evolves."

Most interestingly, in line with the EU's mass-surveillance program, an alternative solution to watermarking where it may not be sufficient is also suggested, although only optional for now:

"Where appropriate and taking into account potential trade-offs related to privacy and security, as well as scalability challenges and costs, Signatories may implement as an optional supplementary measure fingerprinting or logging solutions for AI-generated or manipulated content which allow for checking whether content has been generated or manipulated by their AI system. For example, direct logging may be appropriate for text content, whereas fingerprinting approaches may be preferable for audio and visual content."

Adversarial Paraphrasing: Attack for Humanizing AI-Generated Text (2025)

0x_rs — Fri, 14 Aug 2026 14:52:50 +0000

Article URL: https://arxiv.org/abs/2506.07001

Comments URL: https://news.ycombinator.com/item?id=49299505

Points: 2

# Comments: 0

New comment by 0x_rs in "Single log line is 49KB+ (ext4) / 110KB+ (btrfs) of systemd-journald disk writes"

0x_rs — Thu, 13 Aug 2026 23:38:55 +0000

journald is awful for many reasons, but what makes it worse is that everything running on your machine thinks it has any rights to dump all the logs it wants unprompted. Open a file picker and kio will decide it's a good idea to spam tens or hundreds of thousands of entries into it a day, listing every single file you have in a directory with some log such as "No node found for item that was just removed" and that has zero impact to the user whatsoever. You almost need to keep a script tracking all the journal floods for every new service to make sure it's not treating your system log as its dumping ground. To be fair, the kernel and usb peripherals can also have a bad day and spam 3 million lines an hour into it, think input irq status -75.

It's too much of a chore to keep up with all the program-level configs (if they have them) and service files, but LogFilterPatterns in systemd can help in an unintended way: you can make one log blacklist with a .conf file in /etc/systemd/system/service.d/, and put in there all the patterns that spam your journal one by one, don't even have to chase misattributed loglevels. It just looks something like:

[Service]

LogFilterPatterns=~I am a completely useless log entry

LogFilterPatterns=~I am another useless log entry

But it doesn't pick up on identifiers and doesn't do anything for kernel spam. It's only great to make some messages shut up. Also, I'd consider any btrfs install that does not have nocow on cache, journal etc. to be defective.

New comment by 0x_rs in "How Claude marks AI-generated content"

0x_rs — Tue, 11 Aug 2026 19:04:06 +0000

If we go by "fool's errand" as "needless or profitless endeavor", https://arxiv.org/abs/2303.11156 may already be a good enough answer to the paper you cited, so their work is already laid out for them. The green token idea is thoroughly attacked with much more effective techniques than those in the original paper through "recursive paraphrasing". Among some hypotheses in the paper, one is particularly interesting:

>These experiments provide empirical evidence that more advanced LLMs can lead to smaller TV distances. Thus, based on Theorem 1, reliable AI text detection would become increasingly difficult

New comment by 0x_rs in "How Claude marks AI-generated content"

0x_rs — Tue, 11 Aug 2026 18:24:02 +0000

It's safe to assume so. It says "the output can carry a Claude mark even if the underlying ideas, text, or data originated from another source" specifically about translations among other tasks. It's not strictly about generation but processing, and what level of processing is involved is entirely arbitrary, that is to say: you must feed your text into proprietary black box machines to figure it out, because it's not something you should be supposed to tell otherwise.

New comment by 0x_rs in "How Claude marks AI-generated content"

0x_rs — Tue, 11 Aug 2026 14:00:55 +0000

Is the detection mechanism going to be open, free, and possible to run locally without prostrating to an opaque third-party company that will do whatever they want with the text content provided (including using it for training), and take no responsibility in case of false-positives for which there can exist no proof or evidence against by the victim? This is another useless, if not actively harmful, performative EU regulation, for which they ought to take the full blame despite the fact "AI" companies have been researching and working on watermarks, including in text, for a while now. Just copy-and-paste everything you see and let a machine decide for you if what you're reading is slop or not. Real propaganda machine doesn't care about inane rules and won't waste time with gimped mainstream models either.

>Claude may not be the original author. People often use Claude to proofread, translate, summarize, or convert files. The output can carry a Claude mark even if the underlying ideas, text, or data originated from another source;

Such models already struggle not making any unnecessary or unwanted changes to a corpus, this makes them unable to by design.

New comment by 0x_rs in "Gentoo bugzilla closed due AI bot scraper overload"

0x_rs — Sat, 08 Aug 2026 18:53:15 +0000

Cloudflare is working exactly on enabling that, like it or not.

https://blog.cloudflare.com/monetization-gateway/

New comment by 0x_rs in "Europe's fires are just the start"

0x_rs — Thu, 30 Jul 2026 22:11:29 +0000

Neither of those two have anything to do with the suggested technique(s). Humans are already engaging in climate engineering through existing at any meaningful rhythm necessary to sustain 8 billion people on the planet, it's merely by side-effect, aimless, and lacking any coordination and agency. You are not getting billions of people to "consume" and "produce" less. What you're engaging in right now is no different from what oil companies did decades ago in a coordinated effort, shifting blame onto the consumers, and ultimately the swindling of the century when it comes to environmentalism, see [0] for a quick example. You're free to succumb through inaction, the rest of us would rather not, and never depend on failing, moving emission targets dependent on the whims of politicians and the masses electing them, proven to be incapable of entertaining any abstract long-term planning, let alone sticking to it. It's unfathomable some would prefer entertaining the idea of humanity fundamentally altering its nature rather than actively making up for the effects they are influencing while addressing them, through technology available today, but you must've already given up to believe this is "the new normal" like you said.

0. https://www.npr.org/2020/09/11/897692090/how-big-oil-misled-...

New comment by 0x_rs in "Europe's fires are just the start"

0x_rs — Thu, 30 Jul 2026 16:03:40 +0000

It's been reduced before through cracking down on pollution. The results are directly suspected to be responsible for an increase in warming rate, both locally and worldwide. The only way in which it would be necessary to keep releasing it, and in increasingly higher quantities, is only if no other measure or action is taken, which is obviously not the intended goal: its purpose is to make up for the damage already done that will linger for longer than anyone currently alive.

The emission reductions in our simulations correspond closely to the emissions reductions realized in East Asia over the period 2010–2023, in magnitude and geographical location. This allows us to put our results in the context of the recent uptick in the observed rate of global mean surface warming. Here, we find that emissions reductions in East Asia have contributed up to 0.05 °C/decade since 2010, explaining a large fraction of the observed increase of 0.06 °C/decade over the same period, after filtering out the effects of interannual variability. (from https://www.nature.com/articles/s43247-025-02527-3)

New comment by 0x_rs in "Europe's fires are just the start"

0x_rs — Thu, 30 Jul 2026 15:01:41 +0000

Europe (and the world at large, really) can either wait for its demise, through increasingly oppressive climate hostile to human occupation and de-industrialization while the rest of the world continues on the current trajectory, or start getting its shit together and immediately commence climate engineering on a continental scale. Start with stratospheric aerosol injection, and mandate it not unlike "sustainable" aviation fuels on aircraft flying at the minimum altitude required to have any effect, as it should be possible [0], expedite research into other climate engineering efforts including space-based ones, promptly address the complete and utter reliance on fossil fuels and those profiting from it.. if climate change isn't something worth approaching as a war would be, then there's no point discussing it at all. Researchers and journals have been warning of crossing the point of no return and temperature targets for years now, you either submit to it and keep chasing emission reduction targets, or take action. The idea of simply "adapting to it getting worse" is suicidal, hard to sell to the public and depends on bureaucrats not getting kicked out.

0. https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/202...

New comment by 0x_rs in "Android May Soon Restrict On-Device ADB"

0x_rs — Sat, 25 Jul 2026 14:29:03 +0000

I don't think the bandcamp analogy applies. I may purchase a device with the intent of eventually unlocking the bootloader, but not doing so immediately for whatever reason, such as it still being supported by official updates, or the nth feature not being yet removed. Then the possibility of the option being taken away before you've exercised it remains.

New comment by 0x_rs in "Android may soon restrict on-device ADB"

0x_rs — Sat, 25 Jul 2026 13:59:50 +0000

>What's the problem here? Are you roaming all the time and that imposes a unreasonable cost on you? You want to stay totally off the grid and using VPN/tor isn't enough?

It depends on a remote endpoint that can go down at any point, and depends on the company's policies present and future. It is functionally asking for permission to another party to unlock it, and that would not fall under the umbrella of ownership through being able to install "my own OS" on it, at least for me. And a simple OTA update is all it takes to add multiple verification steps to it, such as providing your ID/using your verified Google account, or just take it out.