HNHacker News
TopNewBestAskShowJobs

ndriscoll

5,424 karma · joined March 8, 2022

submissionscomments
ndriscoll··on Getting debug access on a Brother printer over Telnet
> Disables tracking dots

> Only person to do this, creating a unique signature

ndriscoll··on If IPv4 were land
What problem does 128 bits cause? Just that you don't like :: for a run of 0s? Do you need to care about ipv6 addresses in your work? Because I can tell you that when I'm trying to design simple, efficient routing and firewall rules (with e.g. 100s of thousands of endpoints across different datacenters on different continents), it's really nice to be able to subnet things the way I'd like instead of the way I must to squeeze things into the space I have, and not have to worry about backing myself into a corner later because I gave the wrong number of bits to various dimensions. And worst case, if you somehow got it wrong, just give yourself an entire /64 or more to try again, with it being compatible with what you did (no need to do a phased migration that requires reclamation as you go along).

Like I only tangentially interact with this stuff, but I can't imagine how a network operator wouldn't prefer we live in an all ipv6 world.

ndriscoll··on Start Doing Graph Theory
You can't just make up an identity, claim you have a PhD from a well-known university, and appear on someone else's (e.g. the university's) website. I asked ChatGPT:

> The author of https://vertex.degree/ claims to have a phd and taught courses. Can you find evidence from the named institutions?

And it gave me links to .edu sites corroborating the claims (e.g. he's on a registry of PhDs[0]). I asked if the institutions link to this site, and it said his personal site links to this site[1], and that CMU links to his personal site[2]. If anything it's far easier to check this stuff now. Took me 30 seconds.

[0] https://www.cmu.edu/math/grad/phd/past-recipients.html

[1] https://misha.fish/

[2] https://www.math.cmu.edu/users/math/mlavrov/

ndriscoll··on If IPv4 were land
How does 192.168.0.1 talk to 192.0.0.0.0.168.0.1? Like, literally, how does it make a packet with that destination address?

The entire ipv4 space is already mapped to reserved subsets of ipv6, by the way. e.g. ::ffff:0:0/96 (i.e. 0.0.0.0.0.0.0.0.0.0.255.255.w.x.y.z) or 64:ff9b::/96. v6 devices can already "talk to" v4. The problem is the ipv6 space is obviously not a subset of ipv4, so v4 doesn't have a clean way to talk to or even about v6.

ndriscoll··on Start Doing Graph Theory
Well, this person claims on their website to have a PhD in math from CMU and to have taught the topic at multiple state universities. You could dig some into whether that's true, and then rely on the reputation of the named institutions if you don't know anything about them.

Just like reputation already worked. People have associations with each other. People endorse each other. Endorsements carry some transitive weight (I suppose you could use graph theory to model this). This has nothing to do with AI. Yes, Amazon is full of constantly changing "brands" with no reputation. That is different from a person or company that does have a consistent reputation that you can look up, even if you hadn't heard of them until today. Even Amazon also has e.g. Samsung or Asus if you don't want to buy AHIKJASFDIUHER.

ndriscoll··on Start Doing Graph Theory
That doesn't really answer anything though. There are tons of human cranks out there too who are perfectly willing to spend tons of time on their "theories." Humans make errors too. If you don't have a way to evaluate the source material or some endorsement from someone you have a reason to trust, then AI doesn't change that. If you do have an endorsement (e.g. Terence Tao says some AI book is extremely well done, or puts out his own book that he openly used AI to write), then why would you discard it because AI?
ndriscoll··on Start Doing Graph Theory
What do you mean "the" or "our" current system? That's entirely your interpretation, and we've always had a sliding scale of "non-authorship" from editors to ghostwriters. AI models are the same; you can use them in a heavily directed way (basically, as an editor, or in an outsourcing capacity for e.g. illustrations), or in a hands off way that you just endorse.
ndriscoll··on Start Doing Graph Theory
Well, if we're going to have the conversation: so what if it were AI? Shouldn't the question be around correctness and good pedagogy? Shouldn't the question always be about vouching for those things? If it's actually done well (e.g. they were able to generate better illustrations than they otherwise could've), why exactly should someone think twice?

The author put his name and reputation to it. If it's excellent work, it's excellent work. If it's slop, it's slop.

ndriscoll··on R/opensource bans AI-made and AI-focused projects and non-English speakers
Assuming they allow it, post your thoughts in the language you form them in. Browsers have local translation built in now. The Babel Fish goes in your ear, not your mouth.
ndriscoll··on Ask HN: How to deal with shipping AI-written code you don't understand?
Sour grapes is when you can't have something, so you say it must be bad as a cope. I am successful in tech. It just took until my mid 30s to be a millionaire working on infrastructure instead of my late 20s working on surveillance, but I am perfectly happy with that. I have everything I could want, and didn't need to make those compromises. I say the thing is bad because it's bad.

I might be lucky (in a sense I am, but in another sense "all" I've done is work boring normal jobs and have low expenses), so maybe my path is not perfectly reproducible, but that's kind of the point of temptation: it has to be tempting. And quick, easy, sure money is the bait. You'll never even be held accountable. The only cost is looking around at the bad things in the world and knowing it wouldn't have been there without you.

Personally, the only reason I have to continue working these days is to give my kids a better future, so creating a hellscape for them to inherit would defeat the purpose.

ndriscoll··on Anthropic S-1
If you had been paying any attention at all over the last year, there were 32GB AMD cards at like $1200-1400 and Intel at a little over $1000. I actually hadn't yet noticed they've gone up in price and had last checked only a couple months ago. But even at like $2000, that's still comparable to the nominal price of a mid 90s consumer PC for actual magic.
ndriscoll··on Root shell on Samsung TV
In my experience, those "alignment meetings" are to convey what the outcome is supposed to be, and e.g. how important it is as a revenue source, and then people work to figure out the how and when vs competing priorities.

You don't have to hide what's happening from people that work in ads. There are tons of people who willingly work at e.g. Facebook or Google knowing every single thing they will do is about ads. They might not use the word "spying," but only because they delude themselves into thinking people want the spying and propaganda (e.g. you don't have to look much to find people on this forum asking how else people would know what to buy. Nevermind that a significant portion of the population takes active steps to block them). Like they completely unironically think "you know, I am like a Big Brother. Why would anyone not want a Big Brother? "

ndriscoll··on Ask HN: How to deal with shipping AI-written code you don't understand?
I'm actually doing quite well economically and am generally quite happy with my career!
ndriscoll··on Root shell on Samsung TV
My idea, as I've repeated multiple times, is to blame everyone, and yes, I think they should all be criminally liable. You're the one who seems to think only business people can do wrong, and are perfectly willing to go to bat for the people who actually design and build these systems.

We're not limited to a bounded amount of blame to divvy out. We can blame the mob boss and the hitmen.

ndriscoll··on Root shell on Samsung TV
I think you'll find that I'm the only one in this thread that's not excusing anyone. There is no "instead." They all deserve 50 million years of prison. They all know what they're doing.

Maybe not the engineer who designed a microphone, but certainly the ones who integrated it into the spying platform. The TV doesn't accidentally send that data back. The data doesn't accidentally store itself, and it doesn't accidentally send itself to partners or feed into an ad model. The CEO is also probably not personally working the sales funnel. There's an entire line of people who all knew exactly what was happening.

ndriscoll··on Root shell on Samsung TV
Okay, but if you have a 120Hz panel, as you said, you can just hold your 24Hz signal for 5 frames. And a minipc and TV can happily negotiate a 24Hz HDMI signal. Flicker fusion also has nothing to do with this unless you have like BFI or something.

If your source is 24Hz, it will never be 120Hz end-to-end. It's already 24Hz. You can just send that to the display. If your panel can't cleanly display that, it's not going to be able to with a built in app either. The app is also getting a 24Hz source video.

ndriscoll··on Root shell on Samsung TV
Whether you were well compensated for your wrongs doesn't affect whether you did them. They are all responsible.
ndriscoll··on Root shell on Samsung TV
It makes plenty of sense: they made the devices spy. They collected and stored the data. They built everything. It's like saying it doesn't make sense to go after the gangster shaking down store owners. Yes it does. You go after the boss for ordering it and the person who actually did the bad thing for doing it.
ndriscoll··on Root shell on Samsung TV
Don't excuse your peers' actions. Engineers are smart people. They know what they're doing. And C levels aren't putting anything together as a product. They are giving direction and paying others (engineers) to do that.

Everyone involved is to blame (including management), but it is an engineer somewhere that actually makes these things. Devices don't magically start spying because a business person makes a deal. The data doesn't magically end up in the hands of their 500 partners. There are large datastores to scale and integrations that require work to build.

And even if they're never held accountable (I expect they won't be), you should at least be acknowledging that they have no ethics and shunning them, not defending them for actually being the ones that built surveillance hellworld.

ndriscoll··on Root shell on Samsung TV
My minipc has no issue outputting at 24 or 23.976 or 30 or 29.97 or 59.94 Hz either. And again that big brother series on the smart TV app is going to be garbage quality anyway. It's like setting up a 4k240 OLED and 5090 to run an N64 emulator to play games with a 20 FPS limit and prerendered backgrounds with no texture replacers or anything.

And for "other uses like gaming" you are using an external computer over HDMI, not smart TV apps.

ndriscoll··on Root shell on Samsung TV
Executive class? I expect there was at least one engineer involved that should be facing millions of counts of wiretapping, so a few million years in prison.
ndriscoll··on Root shell on Samsung TV
No streaming app is giving you a video feed with 48Gbps of data (more like 5-15 Mbps), and no one produces 120 or 144Hz video (most movies/TV are still 24Hz aren't they?), so there's no point in sending that kind of signal to a TV (you'll just have a crisp, high refresh signal of the low refresh blurry compression artifacts), but even ignoring that, cheap n100 minipcs can do 4k60 or 1k240 just fine (I use a minipc as my desktop with a 4k240 OLED).
ndriscoll··on Samsung TVs look for microphones in your house via Bluetooth and then spy events
Definitely not with a TV remote vs a keyboard/mouse.
ndriscoll··on UUIDv7: Timestamps aren’t time-ordered and other oddities
That doesn't do anything to help with the issue being discussed, which is index fragmentation and random IO affecting performance. It just means now you have two indexes to maintain.
ndriscoll··on UUIDv7: Timestamps aren’t time-ordered and other oddities
I suppose that's fair, though in my experience nobody seems to think it's crazy if the computer takes over 1s to respond to a human, so 200 ms is kind of small potatoes, and for most cases you can have a local read replica, so only writes need be slow.
ndriscoll··on Why is AI generated language so hard to understand?
Yeah, I've never used Claude, and generally use the codex CLI with whatever base prompt that implies, and I've found that it's always been very clear and concise without goofy "-isms." It just comes across as an intelligent colleague. It even makes effective use of bold and italics for emphasis. I'm a little mystified reading all of these comments.
ndriscoll··on UUIDv7: Timestamps aren’t time-ordered and other oddities
On the other hand, one modern server is 10,000x faster than a 1990s server with 10,000x the capacity, and modern RDBMSs are better at executing queries and give you more options for index types, so the 1990s single-server approach goes very, very far.

I suppose that over time we'll also see more data sovereignty laws, which gives a pretty natural partition key (i.e. for most data you needn't have a global serial because putting them in one database where you'd be worried about conflicts would be illegal anyway, and if you do have some place where you can do those joins, you can adjoin a partition id).

ndriscoll··on UUIDv7: Timestamps aren’t time-ordered and other oddities
Yes, you cannot be in one order and also a different order at the same time. You can make another index, but that's something else entirely.
ndriscoll··on UUIDv7: Timestamps aren’t time-ordered and other oddities
How would it internally sort your index by (timestamp, ID) without you later having to provide the timestamp to locate the ID?
ndriscoll··on UUIDv7: Timestamps aren’t time-ordered and other oddities
It's been a while since I've touched db stuff, but pg still has to write the indexes before commit completes, no? So it's still having to go load and write a gazillion pages. And then unless your filesystem has atomic page writes like ZFS, it also writes the whole page to WAL to ensure that it doesn't do torn writes. Writes to the same page within a checkpoint don't need full page write, so benefit.
← PreviousPage 2 of 34Next →