News for hackers. Stuff that matters.

MattCruikshank

There's a free trial.
I'd be curious how it compares to BeyondCompare by Turtle Software. I've always been really impressed with their diffs and 3-way merges.
Re:Google Search is now less than half of Alphabet’s revenue on Thursday October 29, 2026 @02:20PM (#50293607)
Do they control the Trademark "Google"? Do they control the Domain Name "Google.com"?

Do they control the default Search engine for Android? For Chrome?

Did they pay to be the default Search engine on iPhone? On Firefox?

Are those rare resources? I'd say yes.

And you're conflating "illegally-maintained Monopoly" with "Monopoly." Whether they got there through using illegal tactics or not is a separate question from whether they are a Monopoly now. 90% - is that a high enough market share to qualify as a monopoly, when their annual revenue is like $224 billion? What percentage would be enough to convince you? Does it need to be 100%?

Re:Google Search is now less than half of Alphabet’s revenue on Thursday October 29, 2026 @12:50PM (#50292098)
Great, now they won't mind so much when we break Search out of Alphabet, because it's a monopoly... :D
Re:Alphabet Announces Third Quarter 2026 Results [pdf] on Wednesday October 28, 2026 @04:33PM (#50280500)
...and the stock falls 8% in after-hours trading.
I'm not looking for a $3000 camera that I can install Roblox on.

I'm looking for a $300-$500 camera that has a better-than-cellphone sensors, and better-than-cellphone optics, and as an actual Android device can trivially talk to the apps I share photos and videos on.

I also want to upload to Google Photos automatically. And have the ability to share to Facebook Messenger, post to Facebook Wall, live stream to Instagram or Youtube or whatever... It adds up, pretty quickly.
I'm surprised there aren't Android Cameras? At a reasonable-ish price?

Like a Pixel 6a is $249 new right now.

So, where's that same caliber of hardware, but as a camera-first?

Re:The Kevin Bacon on Saturday October 24, 2026 @04:02PM (#50224596)
I mean... He's been directed by absolute A-list directors, opposite absolute-A++-list actors... Yeah, I'd definitely call him A-list.

* Footloose - How many movie-featured songs get played at MORE weddings?

* Apollo 13 - directed by Ron Howard

* Mystic River - directed by Clint Eastwood

* Sleepers

* X-Men: First Class

* A Few Good Men - directed by Rob Reiner

* JFK - Oliver Stone

* Murder in the First

* She's Having A Baby - directed by John Hughes

* Planes, Trains and Automobiles - directed by John Hughes

* Animal House

* Balto

Re:The Kevin Bacon on Friday October 23, 2026 @11:28PM (#50218124)
If you count the random cut-ins, maybe.

I feel like A Few Good Men would do really well.

Re:The Kevin Bacon on Friday October 23, 2026 @02:41PM (#50213558)
It was probably Amitabh Bachchan at one point, right?
Re:The Kevin Bacon on Friday October 23, 2026 @01:35PM (#50212592)
Sure, but what's the Basis Set movie? If you only had one actor in a movie, and you wanted to have the most connectedness in it, you'd choose Sam Jackson right now. (The one from Pulp Fiction, not the one from Ted.)

But what if you could have TWO actors in a movie? Would Sam Jackson (again, the one from Jurassic Park, not the lawyer) still be one of them? Or is it possible that in aggregate, there are two actors who would achieve a better connectedness than Sam Jackson (the one from The Long Kiss Goodnight, not the one who was in Veronica Mars) plus one of those actors?

What if you could have THREE actors in a movie?

Or what about, who is the perfect sidekick for Sam Jackson (stop asking, yes, of course I mean the one with the purple lightsaber, not the one from In Time)? Who, if you add them to a Sam Jackson (the director of S.H.I.E.L.D., not the Mean Girl) movie, gives them the most aggregate connectedness?

Enquiring minds want to know!

Also, who's the Kevin Bacon of Bollywood?

Re:Google cleared 300 hectares of Finnish forest on Thursday October 22, 2026 @01:54PM (#50198847)
Mandatory Scientology Joke:

That's impressive - I didn't even know you could hook an e-meter up to a tree!

Re:Yet another way to invoke functions without parentheses on Friday October 16, 2026 @12:47PM (#50120845)
I honestly thought of typing that one out here, lol.

You have the camera... I have the monitor... How about I flash QR codes? Oh, the camera is on top of the monitor, facing the room? Well, at night, I could FLASH the monitor full white in binary, and you can see the light on the chair and wall, right?

I'll leave the disc head on the sector number. You read the disc head position.

I'll exit my function on a time modulo a number to encode the remainder. You see what time you were called, and can decode the remainder.

I'll warm the CPU to a temperature... you read the temperature...

The list goes on and on...

"The Exfiltration Manual"...?

Re:Yet another way to invoke functions without parentheses on Friday October 16, 2026 @10:08AM (#50118596)
I want to write a super pedantic book showing every way I can think of to move data from here, to there. In the same program, or between programs.

Global variable.

Parameter to a function.

Pass a functor/lambda, get the value.

Environment variable.

Registry.

File.

Store it in a ping response.

VRAM texture that you write and then read.

Database.

REST server.

I'm pretty sure I could come up with a thousand ways.

But... why? Why do I want to write this book? Why has it been on my mind?

Re:Wayland desktop for Linux with skeuomorphic design on Thursday October 15, 2026 @03:17PM (#50108811)
Looks like an 80s video game.
So, this was a horrible idea I had back in 2012...

"Save A Kid"

It costs $100 to buy the game, and it's $15 per month to play.

When you pick a character, you get to pick a Country of origin, and then an age and a sex. Then you get to start customizing, by looking for a character that has a given race or religion, if you care about that. For newbies, I'd definitely recommend picking a character that speaks a language you can speak fluently. And if you're an Expert, you can play a character that has different Disadvantages. It doesn't give you any bonus attributes, but you get bragging rights for playing one of them well. There's like, Orphan and Kid-With-AIDS, and then there are harder ones like "Scarred by War" where the kids have missing limbs, etc. Anyway, newbies should probably avoid those at first.

So, anyway, it's kind of like The Sims, or a Tamagotchi, or Farm Town, or something.

You log on every day, and depending on the character you picked, you either get an email, or maybe a voice mail or a video mail. The character tells you about how much it appreciates the new computer and the food. The first couple levels, it's all about gaining Health Points. You can totally nerf your character by spending some of your monthly allotment of extra gold points on buffing the kid's parents, if he has them. The parents start generating more gold points on their own, which makes it a lot easier to play. Some people even buy buffs online, which I think is cheating, but for a newbie, it's an easy way to keep your kid alive.

Anyway, you and your friends can make a clan, and all play characters in the same village. I've even heard that some really crazy rich players can buy buffs for the whole village, like a potable water and electricity system.

There are Events in the game, where like the country the kid is in goes into War. That sucks, because sometimes you lose your character, and the game is perma-death, which is lame. They do a really good job making the game realistic, because sometimes you even see in the news that the real country is in war. Most of the time, they just make stuff up though, because most of the wars never show up on TV, but your character won't shut up about them.

Anyway, it's a sandbox game, which some people don't like. But the AI is incredible. Your character will like ask you how they can keep their food from spoiling, and if you look it up online, and explain to them how to do it, they can save some of their food. Or like, how they can build a windmill from scrap metal.

Anyway, you play the same character as long as you want to, but most people get bored after the kid has a job and their own family, and they start over at level one. Restarting costs another $100 though.

The game developers say if too many people play the game, they might even run out of characters to play. I don't buy it. They'll always make up some new War, or some world-shattering Disaster like a typhoon or something, and then a bunch of new characters will open up. Expansion packs, and all that.

A lot of people play this game like 40 hours a week, or something ridiculous. Some people say that playing this game has changed their lives or whatever, and they give up their jobs and play it full time.

Anyway, I guess you can already play most of those mini-games in other forms, but until they package it up in a nice box that you can buy at Best Buy, and they set up some dedicated servers, and set up a scoring ladder (good grades in school, etc.) and maybe some achievements ("high school diploma", "got a job," etc.), oh and a nice technical support line to answer questions and give advice, I don't think the game will really take off.

Is there a sci-fi novel yet, where everyone teaches their own AI how to correctly predict how they answer moral questions, and how they prioritize their principles? And then their AI actually represents them in Congress? Representative AI Democracy?

(Yes, this is a potential Torment Nexus.)

Re:Gitframes on Monday October 05, 2026 @10:13AM (#49965165)
Why does it keep saying "in the browser," but then there's no way to see these running in the browser?

Wouldn't that be trivially easy to do?

Nicholas Polson is an anagram for Nonsocial Slop.
Re:Pirating the Pirates on Monday September 28, 2026 @05:49PM (#49884904)
Dear copyright holders,

You deserve to get compensated. You should get compensated.

But stop acting against the will of the consumers. And stop fighting the technology.

Remember when you thought VHS and DVD would kill the movie industry? You were wrong then, and you're wrong again, now.

Make a very simple website where people can search for actors, search for directors, search keywords, search awards...

And then show us torrent links for all of the content, all of the versions, all of the audio formats, all of the video resolutions, all of the languages, the director's cut, with or without subtitles baked in, theatrical release, re-release, fan edits, all of it. Make it easy for someone else to document a new version, and share the torrent. Set up Popcorn Time or whatever, if you're smart.

I download the content on my own dime. I'll even seed it for a bit, because I like it, and why not?

And then, let me buy a decryption key. Make the price reasonable. Make it Pay What You Want, if you're really smart. When you're setting that price, think real hard about the fact that the illegal option is literally one click away for anyone technically savvy. That decryption key should be shared for all of the different versions available to me. I pay once, I get all the formats, now, and forever. Let me save the key to my account, if you want to set up accounts and recommendations, and etc. Let me save the key in my Plex or Jellyfin server. Or as file metadata, or in a json in a zip file at the end of the file...

Say it's legal for me to pass that license on to one of my kids, when I die.

If you're real smart, grant me the right to legally loan that key to a friend, so they can watch the movie, and then return the key (or I get to auto-revoke it after a number of days).

Hell, tell me a price to have a private (or public?) screening of the movie at a theater that rents out screens.

Cocoon is not available on streaming. That's so dumb. It will enter the public domain in 2081. Until then, earn some reasonable money, by making a legal way for people to do what they're already doing. As frictionless as possible.

And if you're a genius, come up with a way for someone to show you their DVD, Blu-Ray, VHS copy of a movie, and give them that decryption key, for free. They already bought your movie once. Don't make them regret it. Reward them for it. Send it out to anyone who "bought" the movie on AppleTV, Google Play Movies, YouTube, Amazon, etc. But I doubt you're that smart.

Re:Cloudflare Quick Tunnels on Friday September 18, 2026 @03:06PM (#49758781)
That's why I keep using tsnet directly, where you can have as many virtual machines, doing sharing or funneling, as you want to.

The CLI doesn't let you do as much as the go library does.

Re:Why I'm still bearish on LLMs after navier-stokes on Thursday September 17, 2026 @10:12AM (#49741058)
If an LLM writes AlphaZero, and it competes with itself, and is dominant (and beats stockfish!!!), with no book positions cribbed from its learning...

The LLM has a process to beat chess.

Just like, if it doesn't inherently know how to multiply 13 * 17 without using Python to do it... I don't really care.

Maybe you do care. Maybe you want an LLM to be able to do work, only in its head.

But I kind of can't understand the desire for that limitation...

I mean, I do. But it seems ridiculously arbitrary. Like driving a car in 2nd gear and complaining that it gets terrible mileage and can't go fast enough. The Drive gear is literally right there.

Re:Why I'm still bearish on LLMs after navier-stokes on Wednesday September 16, 2026 @10:38PM (#49735789)
So if you need a scratchpad to solve a problem, does that mean you cannot solve that problem?
Re:Why I'm still bearish on LLMs after navier-stokes on Wednesday September 16, 2026 @02:18PM (#49730881)
Do I care if Richard Feynman was only able to do nuclear physics with the help of an abacus?

Sure, a Spelling Bee is a fun thing to have. Little kids work so hard. They practice for hours. There's joy and heartbreak. Prized, sometimes. Notoriety. But in the real world, computer-assisted spelling is by far the norm.

Sometimes you care about the Bee, sometimes you care about the results.

Re:Why I'm still bearish on LLMs after navier-stokes on Wednesday September 16, 2026 @12:19PM (#49729319)
I feel like you're saying something as odd as "Transistors still aren't good at playing chess."

I'm pretty sure Fable could write AlphaZero, which has no lineage in common with stockfish.

Re:Why I'm still bearish on LLMs after navier-stokes on Wednesday September 16, 2026 @12:15PM (#49729256)
I agree that it's a fascinating to crawl inside an LLM, and also to crawl inside of a human, and try to understand the processes and limitations. Like, Phineas Gage is one of the most remarkable learning opportunities we ever had.

That said, it's really weird to me when people use (and judge) LLMs one way... and won't try using them another way.

Like, to judge their utility, I think we should be open to letting them write code, and use the code they produce.

Otherwise, it's like judging a Chromebook without an internet connection. Like, this was one of the most dishonest ads I've ever seen: https://www.youtube.com/watch?v=gDy9AUQJ3Fg

This lamp, without a working power outlet? It really doesn't do anything...

I'm kind of exhausted...

I want to sick Claude on porting tsnet to a C/C++ library that can run in a fat binary Actually Portable Executable.

And then I want to use tailscale share / tailscale funnel APEs for... everything...

Like... everything. So many things.

Re:Why I'm still bearish on LLMs after navier-stokes on Wednesday September 16, 2026 @10:59AM (#49728076)
Do you want to measure the ability of the box, or measure the ability of the box with one hand tied behind its back?

More to my point, I think it's stupid to have LLMs do work that should be done by programs... programs potentially written by LLMs.

I'm advising people that they should think about this distinction, themselves, when they have data and want answers.

Re:Why I'm still bearish on LLMs after navier-stokes on Wednesday September 16, 2026 @10:12AM (#49727344)
> The fact that the human would have a much harder time writing a useful program is irrelevant.

Why?

There's a box.

You give it a problem, and it comes up with a solution.

Why does it matter to you if the box is strictly a LLM, or if the LLM can write code that it executes?

Even neater if the box is self-contained with a local model. You provide electricity, and it comes up with solutions. Why does it matter if it can do chess "in its head", or if it has to use scratch paper?

Older »