Goes-19 weather satellite enters Safe Hold mode 6 days ago
HN user
lolc
4,967 karma
Posts1
Comments2,179
Alman: German, Simplified 7 days ago
To native me, Alman sounds like an US American talking German. Unsurprising, if we look at the originator of the proposal.
I've been repeating this to anybody unwilling to listen: What appears like needless complexity in a language may improve the latency and stability in understanding a message. Redundancies are a feature of language, not just baggage.
Certainly, whether the German declination of articles and pronouns is worth it is very much up for debate. But looking at this example, Alman drops a lot of flexibility in how sentences can be arranged, and where nouns may be omitted. Imagine losing relative pronouns in English. That would slow down native speakers but make learning English easier. Good tradeoff? Of course not!
Codex starts encrypting sub-agent prompts 9 days ago
They must think they have some secret sauce they don't want others to learn. How to optimally instruct sub agents for example. If they hide the sub agent prompts, other models cannot be trained to emulate.
Oh and you can't even use local models or other providers for the sub-agents. You're locked-in.
Road to Elm 1.0 15 days ago
Great hack, thanks! And yea I was missing some native string api when I abandoned Elm at 0.19.
Road to Elm 1.0 16 days ago
Haha, yeah I didn't realize there would be a monkey patch route to FFI in Elm. Can't control the runtime. Is there a good read up on this technique?
The whole point of HTML's design is that it's incredibly easy for a human to write correctly.
A lot of people (me included) used text editors to write HTML. The process was not easy, and the results mostly not correct.
HTML at the time was intended as an application of SGML. This is the first example of HTML from RFC 1866 that laid out HTML 2.0 in 1995:
<!DOCTYPE html PUBLIC "-//IETF//DTD HTML 2.0//EN">
<title>Parsing Example</title>
<p>Some text. <em>*wow*</em></p>
Yes that's where the word comes from. I was describing how the word was used here. You quibble with established usage. https://www.merriam-webster.com/dictionary/analogue#h1
Analogue means having the same function. It does not say which came first.
I was lucky my pharmacist didn't listen to my doctor when I showed up with a prescription for a misdiagnosed disease. Would have delayed getting effective treatment for a life threatening disease. If I'd read forums I could've spotted it too. Don't just listen to your doctor.
... which has adversely affected public health or safety.
Why would they tack that on at the end of a very long sentence? Because they don't want to talk about the loss of USS Scorpion. They mention the sub once on the whole page and even misspell it as "Scorpian". Would not trust them as a source.
I'm under the assumption that Starlink will never provide more than one percent of the mobile bandwidth in a developed country. Because they can't match dense antenna webs in cities. Am I wrong?
From my reading the German court's ruling is that Google are responsible both for claims they made up and claims they rephrase from somewhere else. Because the claims appear as a statement of fact made by Google. If they showed it as "User X on site Y claims Z" they would not be held responsible. Because that's how search engines are understood to work. If they make a wrong quote, they would be responsible for that though.
So my understanding is their unreliable AI summaries are a legal liability for Google in Germany and people can request corrections through the courts.
Wait do I have to choose what search provider I use for every search? They don't get merged by Uruky?
Agents are fixing bugs so quickly and at a scale humans can't do already.
The metric is how many defects are introduced per defect fixed. Being fast is bad if this ratio is above one.
Start each meeting with a prayer to thank the transcription gremlins for their accurate work.
Notice: I love the new AI accurate transcription feature in this meeting!
Eight vaccines linked to a lower risk of dementia 3 months ago
Weird article because it offhandedly mentions the term "healthy vaccinee effect[0]" but does not explain it. So while I was reading and wondering whether all these studies they mention were observational, they drop the term somewhere in the middle and then just continue their breathless listing of vaccines' inverse correlation to dementia.
The word interventional is not found anywhere in the article and it is hard for me to believe there are no interventional studies in this field.
Multiple large observational studies have found that routine adult vaccines are associated with a reduced risk of dementia, with some showing risk reductions of 25% to 40%.
The observed difference is most likely caused by people with declining health being less likely to vaccinate. My assumption is that the article was compiled to show a conclusion that is not supported if one looks only a little bit further.
[0] https://www.science.org/content/blog-post/healthy-vaccinee-e...
As a diabetic I have done this exact exercise: Look at photos and guess carbs. Two slices of bread is easy mode.
Why assume trick ingredients?
At long last, InfoWars is ours 3 months ago
The victims went after a slanderer who systematically profited from his lies. Don't see why we should compare him to randos on Twitter.
"Great idea wise customer, I will certainly mock one out just for you!"
Yea sorry I didn't follow the thread properly.
IPv6 is the only way forward 3 months ago
You haven't answered my question. When do I switch addresses in DNS?
Not clear what your point is. The Signal server wakes up the app via an empty message. At most the info this conveys is that a Signal app got a message to pull.
IPv6 is the only way forward 3 months ago
At what point could I confidently publish the addresses 1:2:3:4:0:0:0:1 and 1:2:3:4:0:0:0:2 in DNS records for people to reach those two servers? After my ISP has switched, or after everybody's ISP has switched?
The idea that any ISP would do a Dagen H is very alien to how an ISP thinks. https://en.wikipedia.org/wiki/Dagen_H
I read through some of the discussion on Wikipedia. The operator of the bot comes across as agreeable and arrogant at the same time.
Questioned about it, he's asking his rig why it did something and quotes verbatim from the generated text. Then when a Wikipedian asks how the bot logged in, berates them how it's all ephemeral code and he could only guess.
If you want a glimpse into the mindset, read this interview: https://www.niemanlab.org/2026/03/i-was-surprised-how-upset-...
The overall attitude is that this was going to happen anyway and we should feel lucky he's so helpful. I rather agree with another commenter here that this was "pissing in the fountain". Whatever pure motivations there may have been, cleanup was left to others.
Why would it need to? Users police themselves :-)
What I want to see is a dating service where I can pledge some money to some charity. Everybody on the site can check what I've pledged, and I can release it anytime. The service can take their cut, fine.
Now when I meet somebody through the service, and we think it's serious, we can release that money. And we can check whether the other did too!
Sure there will still be profiles with people that don't pledge, because they're just testing the waters, or poor, or scammers. Whatever. Point is I can send a signal that at some point I want to be done with the service, and then pay them for that.
How the AI Bubble Bursts 4 months ago
I don't think it's as easy as looking at open weight API prices. We don't know whether the operators are making a profit on all the hardware they bought. Maybe the prices we pay just cover electricity. And it's not even certain that running costs are covered by API prices: The operators may be siphoning content and subsidize from selling that.
In the current volatile environment, the API prices are more of a baseline where we can assume it can't be much cheaper to operate these models.
This, to me, is the strongest argument to offer these slop generators. It provides an incentive to follow the robots.txt.