oktoberpaard

Do you use E2B or E4B?

I’d say it’s very easy to hurt someone with pixels on the screen by spreading these generated images of actual people online.

Local software could be stealing plaintext secrets from your encrypted disk. Physical access is not the only attack vector.

> In general, of 3rd party blockers, uBlock Origin isn't even the best, AdGuard is.

Why? I thought uBlock Origin on Firefox was the most effective combination available (assuming that you use the same filter lists).

Have you actually seen the contents of the comment? Do you actually have an example of anyone that went to jail for calling someone fat? Also: https://www.theguardian.com/us-news/2025/dec/18/tennessee-ch....

It runs on another device and connects to the YouTube app(s) as a client.

HDR photos taken on iOS or Android devices are displayed as SDR images when opened on Windows. The gain map that they contain (see ISO 21496-1) is ignored. Before the ISO standard it didn’t even work between iOS and Android. This is what OP’s frustration is about.

What you are taking about is also called HDR, but has nothing to do with what the other person is talking about. The other person is talking about the still image equivalent of HDR video formats. When displayed on an HDR capable monitor, it will map the brightest parts of the image to the extended headroom of the monitor instead of tone mapping it to be displayed on a standard SDR monitor. So to be even more clear: it defines brightness levels beyond what is normally 100%.

Apparently I was wrong about that part. Only the part about cents still being legal tender was correct. So you can pay the exact amount, but not demand the exact change.

Thanks for the correction. So only the part about legal tender was correct, which is probably what I was confused with. The relevant part:

> The one and two cent coins will remain legal tender, and retailers can choose whether or not to participate. The Netherlands cannot declare the smallest coins worthless on its own in Europe; this must be arranged in Brussels.

The government page about this: https://www.rijksoverheid.nl/onderwerpen/betalingsverkeer/vr...

> Of course, budget stores like Lidl and Aldi still use them but any other corporation is not going to care.

What do you mean with this? Surely the Lidl and Aldi don’t give exact change by default in the Netherlands in 2025?

In the Netherlands cash payments get rounded to the nearest 5 cents, in both directions. Card payments are not rounded. If I’m not mistaken, you can still demand exact change according to the law and you’re allowed to pay the exact amount (cents are still legal tender). Most merchants wouldn’t be able to give you exact change, so it depends on the situation what would happen. I’ve never heard of such a situation happening, though.

> You could also have a Safari content blocker with an optional WebExtension for additional functionality with no usage of DeclarativeNetRequest.

That’s exactly what AdGuard and some other content blockers do. The result is that content blocking works everywhere, but it’s most effective in Safari. As a user, I prefer that over the approach of uBlock Origin Lite, which is a pure WebExtension and doesn’t do anything outside of Safari. Too bad, because I prefer using uBlock Origin on other platforms.

Without having any experience with the APIs to back up my claim, I believe that the WebExtensions API is more powerful in the sense that it allows more complex blocking rules. AdGuard seems to include both options simultaneously, where you have "advanced protection" (WebExtensions API) that only works in Safari and separate blocking lists (old API) that work in both Safari and WebView. This is precisely what keeps me from using uBlock Origin Lite.

I got RSI using a touchpad, which then transferred to using a normal mouse. I couldn’t do anything with a mouse for longer than a few minutes. I completely fixed it with a DXT Mouse 2 (and later 3).

My normal (mechanical) keyboard doesn’t give me any issues, as long as it’s narrow enough (75% layout or less) that I can keep my mouse close enough to the center of my body that I don’t have to rotate my shoulder outwards too much.

My natural resting position on my keyboard is with both arms coming in diagonally without bending at the wrists. I do a lot of work on the terminal, including programming, so I’m using it quite intensively and I never feel any strain.

I tried to get used to a split orthogonal keyboard, but I couldn’t use it for extended periods without getting RSI like symptoms. Would I have persisted, I might have gotten used to it. Or maybe it just wasn’t the right one for me.

Anyway, this is not to counter any of the things you said, because I basically followed the same path with my mouse, but it shows how different postures and usage patterns can lead to different outcomes. I’m still interested in ergonomic keyboards, though, and I might try one again in the future.

What are you talking about? We’re below average already[1].

[1] https://ec.europa.eu/eurostat/statistics-explained/index.php...

Zooming in and out worked on my iOS device to trigger it to load properly.

For me it’s not so much about the readability in sunlight, but about being able to glance at your watch without moving your wrist and about the watch not emitting light in dark environments. I find that distracting and I like the stealthiness of MIPS. That being said, if the minimum brightness is low enough and battery life with always-on high enough I think I could live with it, but with wrist gestures completely disabled except during activity.

I’m running Ollama on 2 eGPUs over Thunderbolt. Works well for me. You’re still dealing with an NVDIA device, of course. The connection type is not going to change that hassle.

It gives weird results for me. I’m using Qwen3-32B with 32K context length at Q4_K_M, with 8 bit KV cache fully offloaded to 24GB VRAM. According to this calculator this should be impossible by a large margin, yet it’s working for me.

Edit: this might be because I’ve got flash attention enabled in Ollama.

Or it could have another reason: https://www.quantamagazine.org/why-are-plants-green-to-reduc....

You might as well use uBlock Origin Lite. The point is that all of these options are less powerful because of the limitations of manifest v3. Instead of downgrading the effectiveness, they’ve opted to release a separate less powerful option so that it’s clear to the end user that it’s less effective than what was available with manifest v2.

FWIW: you can disable DoH without uninstalling the profile. General -> VPN & device management -> DNS -> automatic.

Safari on iOS does support adblockers, in case you didn’t know. And I’m not taking about DNS level blocking, I really mean browser extensions. I’m using AdGuard, but there are other popular options too.

With a 128K context length and 8 bit KV cache, the 27b model occupies 22 GiB on my system. With a smaller context length you should be able to fit it on a 16 GiB GPU.

Yet it can’t block anything from a hostname that also serves regular content (that is, without also blocking the regular content). A good example is YouTube ads. I’m also a happy NextDNS user, but I don’t browse the internet without an adblocker extension.

This is the first response that I got from Claude with an exact copy and paste of the question:

Let me think about this step by step:

1. We're comparing two masses: * Steel ingots: 9.99 pounds * Cotton: 10.01 pounds

2. The cotton bag has a mass of 10.01 pounds, while the steel bag has a mass of 9.99 pounds

3. While steel is much denser than cotton (meaning it takes up less space per pound), this doesn't affect the weight - a pound is a pound regardless of material

4. 10.01 is greater than 9.99 by 0.02 pounds

Therefore, the 10.01-pound bag of fluffy cotton is heavier than the 9.99-pound bag of steel ingots. The cotton may take up much more space, but when we're just comparing weights, density doesn't matter - only the mass does.

FYI: the third url links to an AI generated image.

Firefox + uBlock origin on the desktop and Safari + AdGuard (browser extension, not DNS) on iOS.

I also use the OISD (https://oisd.nl) blocklist for DNS level blocking with NextDNS. OISD prioritizes functionality over blocking, which is exactly the way I like it for DNS level blocking. Never had to manually whitelist anything.

I saw an estimation somewhere that Lily Allen’s label gets a daily payout of $4000, which is probably much more than what she makes from 1000 OnlyFans subscribers. The issue is indeed that the amount that she gets is probably much lower than that. That begs the question: is it Spotify that is the problem, or the record industry.

> That's the point is that it's up to the shell to make sure that you don't use your work if you do the wrong thing.

And how do you envision this would work inside a non-interactive Bash script? A shell option or environment variable that removes the confirmation step? With alternative syntax this would indeed work, because then it’s explicit.

> this one command says "I'm going to make your pipeline terminate, this step will be un-pipeable."

And your method would work like tee and not only write to the file, but also pass it on to the next step in the pipe? If not, the pipe still terminates at that point.