Well, to reference Julia Evans another time:

head and HEAD are specifically the third meaning of ‘branch’, i.e. the newest commit on a branch, but can also refer to a commit not on a branch, when in that detached head state.

And while I’m not enamored with these names either, I can’t think of a word that I like better for this meaning.

Recently, I saw a toddler ask her mom why a snowman decoration was wearing a scarf. And the mom decided to go the playful route and said that the snowman is probably cold. So, the toddler followed up asking why the snowman is out here and not inside the house, where it’s warm? To which the mom replied that it would melt, if it was inside…well, uh, not this one, as it’s a decoration made out of wood.

And like, yeah, these are some tough questions.
Do you just explain to a toddler that humans put up decorations, which don’t always make infinite sense?

You say that like anyone knows how Fahrenheit even works.

It took maybe 10 minutes or so for a 256 GB hard drive for me, if I remember correctly.

That was an SSD, though, so yeah, mileage would definitely vary on an HDD.

Hmm, what does that full format do? Write zeros over everything?

Personally, I would run shred on the root filesystem. It’s a tool specifically intended for properly deleting data (overwrites it with random data multiple times).

I came into this comment section wanting to make the same argument, but I guess, you could also be carrying around a USB-C-to-audio-jack adapter in addition to your wired headphones…

They do have a history of such things happening, yes, which is why my comment exists in the first place. Normally, I would assume this to just be the result of regular shitty management practices paired with regular shitty profit motives.

The history makes it look like they might genuinely have a higher motive here, and I’m saying I still don’t think so, because it would be far too petty and I don’t see them benefitting that much from it.

Most of it?

Google is blocking popular instances these days, so yeah, you basically need to find an unpopular instance, which usually means it’s new and may not live for long, or it will quickly become popular, because it works, which will cause Google to block it.

Yeah, Google started blocking popular instances of Invidious and Piped in May this year: https://github.com/iv-org/invidious/issues/3822

Every so often, it may start working again when those instances get a different IP address, but it usually doesn’t last more than a few days…

The thing is, I really don’t think, Google would care about Firefox. Firefox is sitting at negligible percentages of usage share. The only real competitor to Chrome is Safari and that’s because of iOS.
I guess, they might impact Safari on macOS with this, but someone would have to try this out to actually see, and ultimately, this could still just be a dumb mistake.

Having said that, Google holds a near-monopoly in both video content and web browsers. They have a special duty to not disadvantage competitors and even if this was an honest mistake, I do think, it deserves a slap on the wrist.

I don’t think, there’s currently any plans to introduce a non-JS API for accessing the DOM. It would just take an insane amount of implementation work + documentation.

But frameworks can generate access code for you, so you don’t actually need to write any JS yourself. Rust is quite far ahead in this regard, thanks to the wasm-bindgen library.

I mean, so far, all of them require tons of humanly produced data.

Discriminative AI (deep learning et al) requires humans to label data for hours on end, per use-case.
And generative AI (LLMs et al) require just insane amounts of human works to copy from, albeit not necessarily limited to individual use-cases.

I guess, what I’m saying is that the ratio of how much labor humans (involuntarily) invested into AIs, compared to the labor these AIs actually perform, is likely a lot higher than 70%.

It’s a thing here in Europe. I’m guessing, because our walls are generally concrete, we usually either throw on decorative plaster or a wallpaper, to make it feel a bit warmer and have a uniform surface which accepts paint more readily.

It’s even quite common that if you rent an appartment, that the walls have wallpaper on them, which is painted with a fresh coat of white paint every time someone moves out and the next folks move in.
And then some people, after they move in, will just paint (some of) the walls in a different color, if they feel like not living in pure white…

I mean, at this rate, I’m imagining Microsoft will have hollowed out OpenAI in a few years, but I could see them buying Boston Dynamics, too, yes

If we’re talking passwords, that’s a no. If we’re talking enough personal data that you could use it for spear phishing, identity theft or targetted malvertising, that’s a no.

Honestly, no matter how innocous the information you want is, I would be extremely suspicious why you’d want it. And I’m certainly not turning off my ad blocker either.

I guess, the real question is: Could we be using (simplistic) LLMs on a phone for predictive text?

There’s some LLMs that can be run offline and which maybe wouldn’t use enormous amounts of battery. But I don’t know how good the quality of those is…

You’re in the No Stupid Questions community. Think about rule 7 in particular.

🦀

Ah, true. Thanks.

Theoretically, it was supposed to be pseudo-code, secretly inspired by Rust, but I did get that one mixed up.

And I am actually even a fan of the word unwrap there, because it follows a schema and you can have your IDE auto-complete all the things you can do with an Option.
In many of these other languages, you just get weird collections of symbols which you basically have to memorize and which only cover rather specific uses.