Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
rafram
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
17 ms
·
421.
▲
by
rafram
9mo ago
Two numbers from the USDA: - Canned, drained and rinsed: 7g protein / 100g [1] - Boiled: 9g protein / 100g [2] Not sure what explains the discrepancy (though the second number is much older), but both are considerably higher than
422.
▲
by
rafram
9mo ago
An adult who weighs 75 kg, so is targeting about 75 grams of protein intake per day, would only need to eat 833 grams of cooked chickpeas (which are 9% protein by weight) to get there. That is indeed a lot of chickpeas! But a lot less than
423.
▲
by
rafram
9mo ago
> Meat is actually super easy for humans to digest and it has no downsides to it. In moderate amounts, sure. But frequently eating red meat (more than two or three servings a week) is terrible for you. There's "a clear link bet
424.
▲
by
rafram
9mo ago
> Build scripts communicate with Cargo by printing to stdout. Oh lord.
425.
▲
by
rafram
9mo ago
Run `npm install react` and see how many packages it says it added. (One.)
426.
▲
by
rafram
10mo ago
An “aural” deaf school? This seems like a fairly harmful approach. I know that approaches to deaf education are quite fraught, but pushing students to communicate orally and not allowing sign language in the classroom seems like it’ll set a
427.
▲
by
rafram
10mo ago
https://github.com/WebKit/WebKit/wiki/Safer-CPP-Guidelines
428.
▲
by
rafram
10mo ago
Based on the prose style, I'm assuming you copy-pasted a ChatGPT "deep research" answer?
429.
▲
by
rafram
10mo ago
We don’t need a “deterrent” against things being redacted in publicly released documents. We can have transparency without the whole world finding out the names of victims and witnesses, people’s phone numbers and SSNs, etc., every time a d
430.
▲
by
rafram
10mo ago
This is wrong - Apple did implement multipoint pairing. I’ve had Bose headphones and cheapo Amazon headphones and both switched perfectly between various Apple devices.
431.
▲
by
rafram
10mo ago
It doesn’t need “the cloud” (switching works offline) but it does need to verify that the device it’s switching to belongs to you, which it does using a keypair associated with your account.
432.
▲
by
rafram
10mo ago
https://amtraker.com/map
433.
▲
by
rafram
10mo ago
Among others: https://www.pbs.org/newshour/amp/politics/how-a-fox-news-rep... https://www.bbc.com/news/world-us-canada-68353305.amp
434.
▲
by
rafram
10mo ago
To be clear, there’s no actual evidence that he’s being evicted. Talking about “his eviction” is pretty premature. It also seems like he will receive the reward.
435.
▲
by
rafram
10mo ago
Their handwriting benchmark is not useful. The test cases aren’t even handwritten! https://www.codesota.com/ocr/best-for-handwriting
436.
▲
by
rafram
10mo ago
I think it’s very fair. Anubis generated a lot of buzz in tech communities like this one, and developers pushed it to production without taking a serious look at what it’s doing on their server. It’s a very flawed piece of software that doe
437.
▲
by
rafram
10mo ago
This is a terrible idea.
438.
▲
by
rafram
10mo ago
Thanks, this is good to hear!
439.
▲
by
rafram
10mo ago
I’m seeing a lot of AI tells in the cover art (like for War of the Worlds) and book descriptions. Didn’t see any discussion of where the narration comes from on the site. Are all your audiobooks restorations of classic readings? Do you hire
440.
▲
by
rafram
10mo ago
That seems unlikely to me. One side is made up of lots and lots of entrenched interests with sympathetic figures like authors and artists on their side, and the other is “big tech,” dominated by the rather unsympathetic OpenAI and Google.
441.
▲
by
rafram
10mo ago
> It looks for GLOBAL or STACK_GLOBAL instructions referencing dangerous modules like os.system, subprocess, or socket. This seems like a doomed approach. You can’t make a list of every “dangerous” function in every library.
442.
▲
by
rafram
10mo ago
They hadn't positively identified them, but they knew exactly how old they were? It seems much more likely that they had identified them, but they hadn't gone through the full set of procedures (notifying family members, etc.) tha
443.
▲
by
rafram
10mo ago
Gemini now uses SynthID to detect AI-generated content on request, and people don't know that it has a special tool that other chatbots don't, so now people just think chatbots can tell whether something is AI-generated.
444.
▲
by
rafram
10mo ago
It doesn't seem like Nvidia even has any 3nm GPUs on the market. But sure. When you control for power efficiency, it turns out there's no difference at all!
445.
▲
by
rafram
10mo ago
That chart shows that M4 achieves 25% of the Geekbench scores of GPUs that pull >10x more power. That's definitely efficient.
446.
▲
by
rafram
10mo ago
31% on the economy specifically. Unbelievably (to me), a full 41% of the country still believes he’s doing a good job in general.
447.
▲
by
rafram
10mo ago
I personally think Goyal should focus on making Calibre’s UI less unpleasant before adding yet another bolted-on AI chat textbox, but this doesn’t feel very productive or useful.
448.
▲
by
rafram
10mo ago
Most people aren't willing to sacrifice half their screen real estate 100% of the time, or deal with a significantly thicker phone, just to get a physical keyboard. The market for that is very small.
449.
▲
by
rafram
11mo ago
https://en.wikipedia.org/wiki/Lee_Jun-seok
450.
▲
by
rafram
11mo ago
You write: > I want to see what it would look like if someone like this had more power and responsibility. You were attracted to his campaign because of his policy positions, which most people see as dangerous and offensive.
More ›