frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Software factories and the agentic moment

https://factory.strongdm.ai/
39•mellosouls•3h ago•32 comments

Al Lowe on model trains, funny deaths and working with Disney

https://spillhistorie.no/2026/02/06/interview-with-sierra-veteran-al-lowe/
36•thelok•2h ago•3 comments

Hoot: Scheme on WebAssembly

https://www.spritely.institute/hoot/
95•AlexeyBrin•5h ago•17 comments

First Proof

https://arxiv.org/abs/2602.05192
46•samasblack•2h ago•34 comments

OpenCiv3: Open-source, cross-platform reimagining of Civilization III

https://openciv3.org/
787•klaussilveira•20h ago•241 comments

StrongDM's AI team build serious software without even looking at the code

https://simonwillison.net/2026/Feb/7/software-factory/
29•simonw•2h ago•35 comments

Stories from 25 Years of Software Development

https://susam.net/twenty-five-years-of-computing.html
37•vinhnx•3h ago•4 comments

Reinforcement Learning from Human Feedback

https://arxiv.org/abs/2504.12501
59•onurkanbkrc•5h ago•3 comments

Start all of your commands with a comma (2009)

https://rhodesmill.org/brandon/2009/commands-with-comma/
456•theblazehen•2d ago•163 comments

The Waymo World Model

https://waymo.com/blog/2026/02/the-waymo-world-model-a-new-frontier-for-autonomous-driving-simula...
1037•xnx•1d ago•587 comments

France's homegrown open source online office suite

https://github.com/suitenumerique
496•nar001•4h ago•231 comments

Vinklu Turns Forgotten Plot in Bucharest into Tiny Coffee Shop

https://design-milk.com/vinklu-turns-forgotten-plot-in-bucharest-into-tiny-coffee-shop/
12•surprisetalk•5d ago•0 comments

Vocal Guide – belt sing without killing yourself

https://jesperordrup.github.io/vocal-guide/
174•jesperordrup•10h ago•65 comments

Coding agents have replaced every framework I used

https://blog.alaindichiappari.dev/p/software-engineering-is-back
182•alainrk•5h ago•269 comments

A Fresh Look at IBM 3270 Information Display System

https://www.rs-online.com/designspark/a-fresh-look-at-ibm-3270-information-display-system
27•rbanffy•4d ago•5 comments

The AI boom is causing shortages everywhere else

https://www.washingtonpost.com/technology/2026/02/07/ai-spending-economy-shortages/
59•1vuio0pswjnm7•6h ago•56 comments

72M Points of Interest

https://tech.marksblogg.com/overture-places-pois.html
17•marklit•5d ago•0 comments

Unseen Footage of Atari Battlezone Arcade Cabinet Production

https://arcadeblogger.com/2026/02/02/unseen-footage-of-atari-battlezone-cabinet-production/
107•videotopia•4d ago•27 comments

Where did all the starships go?

https://www.datawrapper.de/blog/science-fiction-decline
56•speckx•4d ago•62 comments

Show HN: Look Ma, No Linux: Shell, App Installer, Vi, Cc on ESP32-S3 / BreezyBox

https://github.com/valdanylchuk/breezydemo
267•isitcontent•20h ago•33 comments

Monty: A minimal, secure Python interpreter written in Rust for use by AI

https://github.com/pydantic/monty
280•dmpetrov•21h ago•148 comments

Learning from context is harder than we thought

https://hy.tencent.com/research/100025?langVersion=en
196•limoce•4d ago•105 comments

Making geo joins faster with H3 indexes

https://floedb.ai/blog/how-we-made-geo-joins-400-faster-with-h3-indexes
152•matheusalmeida•2d ago•46 comments

British drivers over 70 to face eye tests every three years

https://www.bbc.com/news/articles/c205nxy0p31o
165•bookofjoe•2h ago•150 comments

What Is Stoicism?

https://stoacentral.com/guides/what-is-stoicism
9•0xmattf•2h ago•4 comments

Ga68, a GNU Algol 68 Compiler

https://fosdem.org/2026/schedule/event/PEXRTN-ga68-intro/
37•matt_d•4d ago•12 comments

Hackers (1995) Animated Experience

https://hackers-1995.vercel.app/
547•todsacerdoti•1d ago•266 comments

Sheldon Brown's Bicycle Technical Info

https://www.sheldonbrown.com/
422•ostacke•1d ago•110 comments

Show HN: I spent 4 years building a UI design tool with only the features I use

https://vecti.com
365•vecti•22h ago•167 comments

Show HN: If you lose your memory, how to regain access to your computer?

https://eljojo.github.io/rememory/
339•eljojo•23h ago•209 comments
Open in hackernews

Greeting Vocalizations in Domestic Cats Are More Frequent with Male Caregivers

https://onlinelibrary.wiley.com/doi/10.1111/eth.70033
92•JumpCrisscross•2mo ago

Comments

lawlessone•2mo ago
>it is therefore possible that male caregivers require more explicit vocalizations to notice and respond to the needs of their cats, which in turn reinforces cats' tendency to use more directed and frequent vocal behavior to attract their attention.

This is very funny. Ig Noble prize stuff. https://en.wikipedia.org/wiki/Ig_Nobel_Prize

x3n0ph3n3•2mo ago
Does the study mention anything about how much male caregivers meow back?
lawlessone•2mo ago
It says females are more verbally interactive, possibly for cultural reasons , study was done in Turkey.

Anecdotally i can see that being similar in europe though

markhahn•2mo ago
cats being cats, perhaps a little verbal reticence is what it takes to inspire them ;)
abetusk•2mo ago
N=31

> We acknowledge that our sample size limits the generalizability of our findings on cat greeting behaviors.

psunavy03•2mo ago
That smacking sound you heard was stats professors facepalming across the world.
moozilla•2mo ago
I also found this bit funny:

> We also suspect that the geographical and cultural factors may have influenced interaction patterns, given that all our participants were residing in Türkiye.

ddoolin•2mo ago
Anecdotally, my cats meow at me a lot. But they're my cats (others live in the house and help care for them). I also meow back more than anyone else. In fact, I might be the only one to do so. :)
alexjplant•2mo ago
I used to meow back to my cat when he was younger for vocal modulation and pitch practice (lots of minor seconds and perfect fourths). This might have been a mistake as he's now very talkative, particularly when I'm on the phone.
leoc•2mo ago
Already, a solid 2026 Ig Nobel contender.
gradus_ad•2mo ago
Or maybe they're more excited to see the male caregivers? Or maybe the male caregivers are louder themselves so they copy them? Or maybe...
ASalazarMX•2mo ago
Maybe

"Greeting Vocalizations in [These 31] Domestic Cats Are More Frequent with Male Caregivers"

casey2•2mo ago
Maybe

""

You couldn't write a book containing the context needed to qualify a factual statement about any of these cats. It seems even the article authors couldn't be bothered, writing only 5 pages after failing meet their nonsensical objective.

siliconpotato•2mo ago
If you read the paper, it suggests Turkish female caregivers are more vocal with their cats and understand their vocal cues more intuitively/don't need telling twice.
SecondHandTofu•2mo ago
22 behavioral measures looking for one that is <0.05?

Unless they pre-registered that prediction, isn't this just the Texas Sharpshooter fallacy?

psunavy03•2mo ago
p-hacking is a thing . . .
EnPissant•2mo ago
Yep! It's green jelly beans cause acne almost exactly.
byronic•2mo ago
we asked seventy-four cats and you won't BELIEVE number eight
canjobear•2mo ago
They did a Bonferroni correction to the p values, which is supposed to compensate for this kind of thing.
observationist•2mo ago
I can't wait for all the research papers from the "They Can Talk" talking buttons studies. Absolutely fascinating stuff coming out - pets apparently learn to understand not only simple words, but complex constructions and grammar, when sufficiently modeled by their people. They even chain together words in meaningful ways to augment their vocabulary and communicate complex ideas.

Simple vocalizations are cool, but that's pretty limited communication, and if they're capable of so much more, it's pretty likely that they're using individual meows in much the same way a human would under the same constraints. Imagine only being able to say "HEY!" and having to use context and body language in every situation.

TheyCanTalk have got n>10k animals so far, mostly cats and dogs, with a handful of pigs, rabbits, goats, and others. I've seen a few horses, cows, and others on social media, too, and they've got a lot more customers than study participants.

OneDeuxTriSeiGo•2mo ago
Yeah it varies from individual to individual but the running theory I've seen is that cats can understand and interact with language at the level of a 3-4yo.
phainopepla2•2mo ago
This is an absolutely wild theory. 3-4 year-olds can speak in full sentences, and understand quite complicated language
magixx•2mo ago
I would love to believe this but given my experience with my cat even single words are tough for them. For example my cat knows how to fetch her favorite toy but doesn't respond to the word "fetch" at all even with positive reinforcement (treats). The word "fetch" just doesn't mean anything for her even though I've paired it with the action of her fetch many times we play.
bckr•2mo ago
I’m afraid you’re conflating understanding with obedience.
toast0•2mo ago
The thing with cats is they can do a lot of things, but that doesn't mean they will.

My cats will regularly come when called, if they feel like it.

But... what's she going to do if you say fetch? If she wants to fetch, she'll bring you the toy and if she doesn't want to fetch, you telling her that you would like to isn't going to change her mind.

rendall•2mo ago
After many years of living with cats, I have come to believe a cat actively refuses a command or request in a way that looks to humans like the cat does not hear or understand the request.
veeti•2mo ago
Just like me...
jasonjmcghee•2mo ago
Anecdotally, the pitch/tone/voice you say something seems to matter more than the words themselves, to cats.
stavros•2mo ago
I don't know about cats (I haven't tried training) but my dog definitely knew a few nouns and verbs. She understood "food", "water", "walk", "bone", "ball", "bear" (her toys), and could distinguish between "point", "fetch", and "drop". With "fetch ball" she would go get the ball, whereas with "point food" she would point (paw) at the food, and with arbitrary combinations of these verbs and nouns.

It's astonishing, I didn't think they could do that, but apparently they can.

tigeba•2mo ago
I have two cats that I have clicker trained. They can: shake, high-five, fist bump, stand up, and will "go here" and come and touch your finger. They won't do any of the tricks without treats :)

One of the cats had seemed interested in fetch for several years but it never quite worked. I tried a bunch of different toys and finally found a plastic spring that she will fetch very well.

magixx•2mo ago
I think we have the same cat. I also did the clicker training for paw/fist tricks and sitting pretty quick. I took a pause with the stick pointing because she traumatized herself with a stick toy and is quite avoidant to sticks. Sprint toys are her favorite and she'll fetch those or bring them to me to throw when she wants treats.
stavros•2mo ago
Really? That's interesting, I didn't think cats could be trained so well. Maybe I'll try mine!
jonners00•2mo ago
In our case:

over there; bed; basket; lead; walk; wait at the door; go to [room]; go to [family member]; stay; stay on the path; paw; paws up; sit; lie down; heel; don't pull on the lead; no more poo poo; no more wee wee; this is the commute; up; where's your piggy?; treat?; buscuit?: up for a cuddle?; want to come up?; permission; okay; no; stop; this way; stay close; hungry for some food?; water?; down; don't lick; who did this?; come here; shall we go the pub; let's go;

and probably 5 or 6 more words/phrases. I don't include words like catch/fetch where accompanying action may trigger the response.

It's definitely not a three year old's grasp of language, but it's probably a point that a baby crosses somewhere between 9 and 15 months old.

05•2mo ago
Treats are good and all but you will probably get better results with clicker training and splitting the job into simple parts. Fetch is pretty complicated and even dogs often have issues with it - our dog did just fine on the 'chase the toy' part - mostly due to strong prey drive - but then just ran away with it :). Took months to get it working.. Also cats usually lose interest quickly, so short sessions repeated more often will give better results.
magixx•2mo ago
Yeah I did clicker training for her to learn fetch and she know how to. She just seems to have zero association between the "fetch" word and action.
red-iron-pine•2mo ago
"no give, only throw"
jandrewrogers•2mo ago
I think you might be conflating cats understanding and cats responding. When cats choose to engage they can be quite intelligent and take direction, I’ve seen it many times. But most of the time they DGAF and will ignore you.

Unlike dogs, cats aren’t seeking your approval. It is a different kind of relationship.

magixx•2mo ago
I'm impressed by the degree of selective catS possess. They can responds to like 5 different names if in the mood while other times won't event flinch even after repeating 20 times.

While my cat doesn't seek my approval, she is definitely highly food motivated to the point where she was faking wanting to go outside and thus put a harness on (which she does not enjoy) just for the treats she would get.

topspin•2mo ago
Merely an anecdote: I had one female house cat that clearly understood a number of words. She could easily and consistently pick out "catnip" in a sentence. "Cow", "get up", "tuna" and several other words and phrases were all understood.

This is unique in my personal experience. I've haven't seen this in other cats.

nkrisc•2mo ago
Having a 3-year old and two cats (and many more previously in my life), that sounds absurd.

I can give my 3-year old (38 months) multi-part instructions and he can even ask clarifying questions back when my instructions are ambiguous or he doesn’t understand them. He’s also being to ask insightful questions as he observes the world around him.

To suggest cats are capable of anything remotely close to that (with humans) requires extraordinary evidence.

bombcar•2mo ago
My two year old can completely understand a sentence and then say, “No!” or ignore me.

I’ve had cats that could do the same.

lcnPylGDnU4H9OF•2mo ago
> he can even ask clarifying questions back when my instructions are ambiguous or he doesn’t understand them

This is kinda moot in this discussion given cats have neither the facial structure nor lung capacity for this. Others are essentially saying that a cat would (have the mental capacity to) do these things if they were not so physically constrained.

nkrisc•2mo ago
Again, extraordinary evidence required.

If it’s true, someone ought to be able to demonstrate it.

lcnPylGDnU4H9OF•2mo ago
> If it’s true, someone ought to be able to demonstrate it.

This is not a truism.

> Again, extraordinary evidence required.

Yeah, whatever, dude. I'm not trying to convince you, just pointing out the irrelevance of the point that humans are physically capable of speaking like humans. That's moot to the claim that had been made in the parent comment.

I could go on about other points you made. For example, you seem to be conflating working memory ("multi-part instructions") with understanding of language, given that was what the parent commenter had claimed.

> He’s also being to ask insightful questions as he observes the world around him.

Cats don't have the same mental capabilities as a 3-4 year-old but that is not under contention. That doesn't strictly mean they don't have a similar capacity for language as a 3-4 year-old. Put another way, you were not talking about your son's capacity for understanding language but rather his general mental capacity, which is not an apt refutation of what was claimed.

This isn't to convince you of some belief (indeed, you're entitled to your opinion, friend) but to point out the illogic in your argument.

Antibabelic•2mo ago
I'm not sure I understand your distinction between understanding of language and general mental capacity. As I (and the other person responding, I believe) understand, the two are inseparably connected in humans.
nkrisc•2mo ago
Ok, then what does “ a similar capacity for language as a 3-4 year-old” mean if it’s not everything else you say it isn’t.

So they can’t speak nor understand us, nor remember several instructions, nor formulate questions, so what then do they have that puts them on par with a 3-4 year old?

tkfoss•2mo ago
I haven't seen a single convincing example yet, do you mind sharing some?
nullstyle•2mo ago
I’ve seen some BilliSpeaks videos that absolutely convince me cats have more language capabilities than previously thought.

When you say “convincing” what are you looking for? Do you think these youtubers are just using editing tricks and traps to convince a gullible internet? Genuinely curious.

guizadillas•2mo ago
I assume they need convincing in the way it proves pets can communicate complex language or it is just the owner interpreting whatever they want
willis936•2mo ago
Cats don't talk in complex ways to owners. Listen to what a male cat vocalizing while trying to woo a female cat. They have something to say and a means to say it. They don't need that kind of communication for reminding the food dispenser what time it is.
guizadillas•2mo ago
That's exactly my thought, it is no different than teaching my dog a trick for treats
MangoToupe•2mo ago
> pets can communicate complex language

Who was even trying to push this narrative in the first place?

guizadillas•2mo ago
the first comment: "complex constructions and grammar"
MangoToupe•2mo ago
> a single convincing example

On which particular thesis?

randycupertino•2mo ago
Not necessarily "convincing" but the "ma-AH" aka "come here/how to speak cat" video is very cute! https://old.reddit.com/r/fixedbytheduet/comments/1os4uid/hav...

I tried it on my cat and it worked although she may have just come over to investigate why I was making weird noises at her!

The rest of the original guy's "how to speak cat" series is also fun.

https://www.youtube.com/watch?v=Qyc7Krs0Fi8

observationist•2mo ago
https://cclab.ucsd.edu/pet-cognition-communication/

Check out their research. It's very solid - there are all sorts of conditioned response tricks to see on social media, but the UCSD research has recorded examples of multiple pets using the same untrained behavior where they string together 2 or more words to indicate a new idea. Not only are the pets talking, but there are emergent and convergent abilities being documented that demonstrates a common mode of cognition that arises from the underlying cognitive capacity of these animals, across species.

There's a great example of a dog on youtube where it used "squeaky car" to point out a firetruck with the sirens going. There are examples of dogs "laughing" at farts, calling out smells, telling owners about injuries or burrs stuck in paws, discomfort or pain in a leg or stomach or ear, and so forth. They indicate affection, anger, impatience, happiness, fear, confusion. They identify group membership - stranger, family, neighbor, friend, human, animal, dog, etc.

Video examples on YouTube are ubiquitous at this point - take a look at the training methods, the way vocabulary is modeled and taught, and then go in with a skeptical perspective. For those owners that follow the teaching methodology validated by the UCSD research, they're providing their animals with a legitimate skillset that empowers the pet with language understanding, the ability to communicate and creatively produce novel and complex ideas.

This shouldn't be too surprising - based on neuroscience research, it looks like the engine of human cognition is the mammalian neocortex. Lions, tigers, and bears (and dogs, cats, whales, primates, and nearly all mammals) are likely able to use language, given the apparatus and training to do so. Humans have a unique capacity for complex vocalization and extremely sophisticated culture, allowing rapid uptake of language skills. Given the opportunity, it looks likely that nearly all mammals will have some facility with language, simply because the neocortex function is so powerful and plastic.

NetMageSCW•2mo ago
I am Groot.
canjobear•2mo ago
Do you have examples for complex grammar?
observationist•2mo ago
https://cclab.ucsd.edu/in-the-news/

There are a lot of examples, but also I am using "complex" generously - "we go friend walk later" and things like that, the limitations seem to be 3-4 direct ideas chained together, with person+activity+time being about as complicated as things get. It's probably possible with the right curriculum to expand on that significantly, but it's really hard to get sufficient repetition on abstract things, and dog attention is highly selective and limited. If they get overexcited, they're not in word-learning mode anymore, they're just fixated on the ball, or the pigeon, or the weird smell that just happened.

Once BCI for dogs is a thing, it wouldn't shock me to have a dog eventually write simple sentences or even engage in some sort of ideogram based text communications, but you're never getting a canine Shakespeare. You could definitely get narco k9s that deliberately say things like "I smell cocaine on this guy's shoe", or "I don't smell anything bad". With that and bodycam, you'd have a record and a way of auditing k9s.

One of the potential spin-offs of this canine cognition research is the potential for a structured standard vocabulary for working dogs in various professions, with obvious benefits in police accountability, disability accommodation and assistance dogs, and so on. If all dog trainers worked from a common curriculum, then it could be standard that dogs are taught a consistent 30-40 baseline words, so if a dog escapes, it can interact with animal control officers or the public in useful ways.

The biggest thing is that they can use language at all, I think. It's wonderful, and I hope it ends up being a big boost for animal welfare in general.

blindriver•2mo ago
"n = 31" tells you how valuable this research paper is.
jacquesm•2mo ago
What the cats really are saying is 'unshaved human alert, unshaved human alert'.
amanaplanacanal•2mo ago
In my house the occupants are divided into fur-faces, and bare-faces.
jacquesm•2mo ago
That could be the answer right there: kinship.
fat_cantor•2mo ago
A lazy error [0] in the abstract suggests that the paper may have been written by a domestic cat (Felis catus). Given that male experimenters tend to stress out mice [1], it is plausible that the opposite effect occurs in cats. Mostly, it would be fun to know whether cat greetings reflect different motivational or emotional states, since reasonable people might believe that they don't.

[0] "We also tested whether demographic factors such as the influenced the amount of greeting behavior expressed by household cats."

[1] https://www.nature.com/articles/nature.2014.15106

ChrisArchitect•2mo ago
Previously: https://news.ycombinator.com/item?id=46132439

https://news.ycombinator.com/item?id=46077728

casey2•2mo ago
They are not.
tim333•2mo ago
That was probably true with our cat. It was very 'talkative' when I came home. It also definitely distinguished between men and women, prefering to sit on the lap of women, preferably old ones.
foxyv•2mo ago
This makes me wonder if we could invent a constructed language that can be used by both cats and people. If people learned how to speak such a language, could they teach some of the basic vocabulary to cats too? I imagine it would be a pretty weird phonetic scheme. Cats have a ton of cool sounds they can make because they use them to lure in prey.
telesilla•2mo ago
Because our cat knows who the big softie is when asking for snacks (not me the woman of the house, who is more concerned that kittie will get too fat to fit through her cat door).
stevenalowe•2mo ago
Is it because men don’t listen?