HNHacker News
TopNewBestAskShowJobs

dgroshev

1,778 karma · joined September 30, 2014

submissionscomments
dgroshev··on Stuffed-Na(a)N: stuff your NaNs
I love it, how did I not think about this!
dgroshev··on Stuffed-Na(a)N: stuff your NaNs
Yes! Since any function returning a float might need to return NaN, NaNs are represented as a special kind of float, with most bits unused.
dgroshev··on A decision to eject from a failing F-35B fighter and the betrayal in its wake
I read it and I know about that case. However, as I said I'm struggling to see your point.

I guessed you meant that that one case proves something about air superiority or Iran having an advantage over USAF, so I responded with a historical parallel.

dgroshev··on A decision to eject from a failing F-35B fighter and the betrayal in its wake
I don't understand what that link is supposed to prove. A single F-117A getting shot down in Yugoslavia due to complacency and chance doesn't negate Nighthawks bombing SAM defended areas with impunity. Iraq had a strong and integrated AD network for the time.

Air superiority alone doesn't, but it's a massive force multiplier.

dgroshev··on A decision to eject from a failing F-35B fighter and the betrayal in its wake
$2t is the price tag of the entire program [1], from its conception to the jets retiring decades from now (note that programs tend to get extended, so the retirement year is probably too pessimistic). For comparison, the US is still producing F-16s, with the first F-16 produced in 1976.

F-35s has a much lower crash rate than F-16s during their first 20 years in service [2] and just recently passed 1 million flight hours [3]. The program has its problems, but it resulted in an incredibly capable fighter plane. Practically every US ally that has access to the F-35 run their evaluations and concluded that the F-35 is the best option (eg [4], quote: "F-35A offers highest overall benefit at lowest cost by far").

[1]: https://www.gao.gov/blog/f-35-will-now-exceed-2-trillion-mil...

[2]: https://www.aerotime.aero/articles/are-new-fighter-jets-more...

[3]: https://theaviationist.com/2025/03/04/f-35-one-million-fligh...

[4]: https://www.admin.ch/gov/en/start/documentation/media-releas...

dgroshev··on A decision to eject from a failing F-35B fighter and the betrayal in its wake
> As Ukraine shows, peer conflicts are won by other means, not stealth air superiority

I don't think you can conclude that when neither of the belligerents has the capability. As Gulf War shows, training and capabilities (including stealth) do enable SEAD/DEAD to an extent that unlocks air superiority.

dgroshev··on Building a BitTorrent client from the ground up in Go (2020)
Yeah. I was working on [0] for a while (I might get back to it when I have more time) and implemented a chunk of the protocol with an eye for a more scalable libtorrent replacement, and it's not entirely trivial. Even things like the piece picker get trickier at scale.

[0]: https://dgroshev.com/blog/torrentdyne-encryption/

dgroshev··on Building a BitTorrent client from the ground up in Go (2020)
It's not impossible, it's just hard to be better than libtorrent, otherwise a new client only makes sense as an exercise.

Arvid Norberg is hardly a "beginner developer", yet look how hard it is to make a performant implementation of uTP: https://github.com/arvidn/libtorrent/issues/3542

dgroshev··on Building a BitTorrent client from the ground up in Go (2020)
Few more challenges:

- seeding a million active torrents (this is entirely possible and needed but drives a lot of architecture)

- a UI for a million active torrents

- be nice to trackers (keep-alive, batched reports; again, architecture gets tricky, iirc libtorrent can't do that still)

- correct pipelining

- congestion control on uTP

- I/O scheduling/batching

- testing all of the above

It is quite a lot.

dgroshev··on Open Euro LLM: Open LLMs for Transparent AI in Europe
From experience, regulation as an explanation for EU startup competitiveness is overused so much it's almost meaningless. Can you point out specific laws that you consider existential for startups?

What I find matters way, way more is two factors:

- Concentration of capital. The US has an ecosystem of wealthy people that want to put their money somewhere. This is good for startups, but can also backfire as we can see in the news.

- Unified market. EU is not a single market, it's several dozen markets with different regulations, different languages, and different cultures. You can't sell the same B2C product with the same marketing in Germany, Spain, and Sweden as easily as you can in California, Ohio, and Texas.

dgroshev··on Open-R1: an open reproduction of DeepSeek-R1
A judgement is by necessity subjective. Have a look at some sentencing remarks, for example this, point 19 forward: https://www.judiciary.uk/wp-content/uploads/2021/09/Wayne-Co...

Someone has to make a call. The weight of the call rests on the person's life experience, their understanding of the context and the cost to the society, their empathy to both the defendant and the accused, and their conscience. Treating it as a black box exercise misses the point completely.

dgroshev··on Open-R1: an open reproduction of DeepSeek-R1
This is nonsense though. What does "better" mean in this case? A judge is not a black box with an input (the case) and an output (the verdict), the entire point of having a judge is to have empathy, conscience, and personal responsibility built into the system.

It's a blind spot that too many people have because we take those qualities for granted. LLMs unbundle them, so we need to start recognising the inherent value of humans, fast. I wrote a few words about it here: https://dgroshev.com/blog/feel-bad/

dgroshev··on It's Time to Switch to OpenTofu
That's not about nationality though. That PR is about (re)enabling OpenTofu to work more smoothly with Russian SaaSes, which are either already sanctioned or are likely to be sanctioned.

Everything is political, being "apolitical" is a political choice. You can't escape politics.

dgroshev··on Britain doesn't need to walk a US or EU path on AI
https://archive.ph/MpOOZ
dgroshev··on Supernovae evidence for foundational change to cosmological models
Previously (includes informed critique of the paper): https://news.ycombinator.com/item?id=42495703
dgroshev··on Apple Photos phones home on iOS 18 and macOS 15
That's precisely what https://en.wikipedia.org/wiki/EPrivacy_Regulation was supposed to be! As you can imagine, there are strong incentives to lobby against it, so it's almost a decade late already.

Whoever came up with an idea to attach CSAM scanning provision to it is an evil genius, what an incredible way to make sure it's not going to pass any time soon.

dgroshev··on Supernovae evidence for foundational change to cosmological models
Previously: https://news.ycombinator.com/item?id=42495703
dgroshev··on A 10-Year Battery for AirTag
I'm almost sure they are 3D printed on a nylon SLS machine, what a lovely example of high quality printing enabling relatively small scale, high number of SKU production!
dgroshev··on Three Mistakes from Dart/Flutter's Weak PRNG
There is nothing wrong with PRNGs, they are perfectly cryptographically secure when used correctly. In fact, AES-CTR is just XORing the plaintext with a pseudo-random generator (AES'd counter + initialiser). The problem is bad and misused PRNGs.
dgroshev··on What we know about CEO shooting suspect
Was it a governmental agency or a private entity that denied coverage in their case?
dgroshev··on A transformer supply crisis bottlenecks energy projects
It's unfortunate that the article doesn't point at the most obvious cause of the shortage: Russia deliberately and systematically destroying Ukrainian power grid for three years, and Western unwillingness to put a stop to that.

The scale of necessary repairs is mind boggling. Here's what just Europe alone shipped to Ukraine to repair the damage [1]:

- 8 827 power generators… with an estimated 1 258 590 kVA in power generators

- 3 597 transformers + 5 autotransformers

- 3 364 generators and 14 transformers from the rescEU stockpile

Then there is US support:

- 18 autotransformers "with more on the way" [2]

- 59 transformers just for Kharkiv [3]

…and that is just scratching the surface and is barely enough to keep the lights on, with Ukrainians installing private batteries and generation up and down the country.

Just last night Russia launched another massive attack at the energy infrastructure [4]. More expensive, rare, long lead time transformers getting blown up, and the raid doesn't even get any attention in the West.

[1]: https://erccportal.jrc.ec.europa.eu/ECHO-Products/Maps#/maps...

[2]: https://x.com/USAmbKyiv/status/1819094722754797600

[3]: https://x.com/USAmbKyiv/status/1792791129513218539

[4]: https://www.reuters.com/world/europe/russia-launches-large-s...

dgroshev··on A transformer supply crisis bottlenecks energy projects
I genuinely don't understand what compelled you to make this comment. One is a tiny enclave with just a handful of transformers and a rudimentary power grid. Another is the biggest country in Europe, home to about 40 million people, with a vast energy grid systematically targeted for three years, and harsh winters. It's +19 in Gaza and -4 in Kyiv right now, without electricity Ukrainians will very literally freeze to death in a very short time.

This is such a glaring disparity that the only explanation I can see is some kind of moral point, since clearly one is much more relevant to the global shortage of transformers than the other. What is the moral point you're making? What are you actually trying to say?

dgroshev··on What we know about CEO shooting suspect
Any medical system inevitably has limits of what they can spend per patient. Do you prefer the limit to be set and enforced by the government that is amenable to political process, or anonymous profit-seeking insurance company board members, like in the sibling comment case https://news.ycombinator.com/item?id=42375998 ?
dgroshev··on Ukraine's three nuclear power plants have restored electricity production
Extraordinary claims require extraordinary evidence, and hearsay and unnamed "statistical estimates" are hardly extraordinary.

The claim is definitely extraordinary since both wars are extensively documented, and we just know (like, we have verifiable videos) that Russia has much, much less regard for civilian casualties, employs much less precise fires, and Ukraine is more populous and dense than Iraq.

Besides, the numbers I could find are the other way around. This study [1] estimates 3,750 civilian deaths in Iraq, and this UN report [2] says 10,582 recorded civilian deaths (that is, it's definitely an undercount, unlike the estimation in Iraq).

[1]: https://www.comw.org/pda/0310rm8.html#3.3.%20Total%20Iraqi%2...

[2]: https://www.ohchr.org/sites/default/files/2024-02/two-year-u...

dgroshev··on Ukraine's three nuclear power plants have restored electricity production
That's not how escalation and escalation management works, and "Mutually Assured Destruction" is mostly a lie.

Putin is a paranoid coward and fears for his life like few people you know. No he's not going to "destroy the world", it would endanger his own life and the palaces he built for himself.

No it's not binary, there's a vast, vast gulf between limited nuclear use and a full-on attack on CONUS. In fact, any nuclear use by Russia would likely bring its rapid defeat, since their main allies (India and China) are nuclear armed countries with a border dispute and are very interested in the nuclear taboo staying intact.

No it's not "Mutually Assured Destruction", the mainstream of nuclear targeting is counter-force disarming strikes, and the US is perfectly capable of winning a nuclear war. "Winning" here means ending up in a situation where cities are mostly intact on both sides, but only one side is capable of threatening them. Moreover, being credibly capable of winning a nuclear exchange was (during the Cold War) and is necessary to keep European and Asian allies safe. No, militaries are not deliberately glassing cities if push comes to shove, it doesn't make any military or political sense. You don't kill a hostage to win a negotiation.

No, a nuclear war is not a "suicide". It's very bad news for everyone (although chances are we will see battlefield nuclear use in our lifetime), but there's an incredible amount of nuance and complexity. "Putin presses the red button and everyone's dead" is a take that is not even wrong.

dgroshev··on Ukraine's three nuclear power plants have restored electricity production
Can you please provide a source for the 10s of thousands claim? I find it extremely hard to believe, considering the way the war was persecuted. The US did not indiscriminately shell cities with cluster munitions [1], and Iraqi cities are less dense and smaller than commie block filled Ukrainian ones.

[1]: Russia routinely fires cluster munitions into densely populated cities, eg https://www.bbc.com/ukrainian/features-60941459

dgroshev··on Ukraine's three nuclear power plants have restored electricity production
What do you gain from hunting civilians? [1] [2]

[1]: https://www.ft.com/content/31b630b3-2639-456c-ba50-3caea7a9b...

[2]: https://archive.ph/v5qeA

dgroshev··on Meta Fined $843M by EU over Marketplace Ads
> why do you think Facebook won't simply sell the ad space to the highest bidder, whether it is AcmeShop or a Marketplace seller?

Depends on which placement you mean.

"Boosted listings" in Facebook Marketplace? AcmeShop can't bid on that.

In-Marketplace ads like [0]? They are similarly in-Marketplace, I don't think it's technically possible to bid on them.

In-feed Marketplace ads like [1]? They are specifically Marketplace, I have no evidence that they enter the same bidding pool, and it seems improbable.

In-feed generic auto-placed ads like [2]? Maybe, maybe not.

> Remember, Facebook doesn't even make money if you sell something locally and pay in cash.

Facebook doesn't just make money from boosted listings, they also benefit from people just using Facebook. Even if Facebook didn't make any money directly from Marketplace, it'd still be worth it for them and it'd still be anti-competitive. If they can make themselves the main marketplace, it creates a large captive audience scrolling through their ads.

> But I think it has to be shown that Facebook is actually engaging in that kind of anti-competitive behavior. We can't just assume it.

Aren't we commenting under an article about a fine for anti-competitive behaviour?

[0]: https://storage.googleapis.com/website-production/uploads/20...

[1]: https://newsfeed.org/wp-content/uploads/Marketplace_boosted_...

[2]: https://storage.googleapis.com/website-production/uploads/20...

dgroshev··on Meta Fined $843M by EU over Marketplace Ads
Alright, let's reframe this:

- AcmeShop is paying Facebook money to run a targeted marketing campaign on Facebook

- to do this, in practice AcmeShop is forced to share visit data with Facebook (this bit could be done differently, but for some reason that's how Facebook does it)

- a customer goes to AcmeShop, doesn't make an immediate purchase, leaves

- AcmeShop is paying Facebook to run a remarketing campaign for that customer, showing them the same product leading to AcmeShop

- at the same time, Facebook is using information that they obtained in the process of providing paid services to AcmeShop to show goods on their marketplace platform to the customer. For Facebook, it's not "remarketing", it's just "marketing" at this point. Not only that, but they control which ads they are showing, and naturally with Marketplace-only placements Facebook has an advantage over AcmeShop (again, using information that they effectively forced AcmeShop to provide)

Does this reframing make the problem of Facebook being both an ad seller, an ad manager, and offering the same service as their customers any clearer?

dgroshev··on Meta Fined $843M by EU over Marketplace Ads
When Facebook is showing ads for items on Facebook Marketplace based on a customer visiting AcmeShop, for Facebook it's not "remarketing". It's only "remarketing" when AcmeShop does it. For Facebook, it's just targeting based on data they obtained during providing a paid service to AcmeShop. The customer didn't express interest but then abandon their purchase from Facebook yet.

Google doing the same thing is hardly a good excuse.

In your analogy, you don't just "see". You're employed by a nearby stall because you're one of the few plumbers in the town, so you're in that shop a lot, but you're also making photos of their customers and then use those photos to find them on the street and advertise your frobulators. You really don't see anything weird about this scenario?

Besides, if a shop has only one chance ever, what are remarketing ads (marketed by Facebook to businesses) for?

← PreviousPage 8 of 15Next →