HNHacker News
TopNewBestAskShowJobs

axegon_

3,418 karma · joined November 9, 2018

email: ping@axegon.com
submissionscomments
axegon_··on I would enjoy an HN chat. Is there one?
This. I've had my email for years. Admittedly my engagement on HN is chaotic and somewhat rare but in all those years, I think I've had half a dozen one-off emails.
axegon_··on Ask HN: Data engineers, What suck when working on exploratory data-related task?
Oh don't even get me started on cities, states, towns and addresses :D https://www.google.com/maps/place/95450+Us,+France/@49.10531...
axegon_··on Ask HN: Data engineers, What suck when working on exploratory data-related task?
This isn't wildly different from what I've done but it's the sheer volume of crap that's scattered through all the different fields. The countries are the least of my problems. There are others where I'm faced with tens of millions of different combinations. The countries are a relatively trivial problem in comparison.
axegon_··on Ask HN: Data engineers, What suck when working on exploratory data-related task?
Not a data engineer but my work revolves around processing a ton of data(let's call it partial data engineering). Much of the data I get is inputted by humans from different sources and platforms and countries. My biggest pain in a nutshell - the human factor. Believe it or not, people have managed to misspell "Austria" over 11,000 times(accents, spaces, different encodings, alphabets, languages, null characters and so on. Multiply that by 250-something countries and multiply that by around 90-100 other fields which suffer from similar issues and multiply that by 2.something billion rows and you get the picture.
axegon_··on What's your "coworker didn't lock screen" procedure?
None but the best story I have is from around a decade ago - a guy did not lock his screen and had his facebook opened. His manager changed his birthday to the current day. The best part is when his aunt wished him a happy birthday.
axegon_··on Every wondered how Facebook spoofs Gmail message list snippet text?
Disclosure: I do not have a facebook account and I am not a lawyer. But... I am pretty sure that's in breach of several EU regulations if this is happening to EU citizens and EU residents: DSA, GDPR, ePrivacy directive, consumer protection and possibly misleading advertising and unfair commercial practices.
axegon_··on Ask HN: What are you working on? (May 2025)
Same thing as comparing a phone camera to a professional camera really. For the average person, nothing really changes beyond a steep learning curve. But better quality video, faster(not fpv fast of course), can carry a heavier payload, a lot more options when making videos/capturing photos(which will likely be my next biggest challenge - geting the wiring from the drone to the gimbal and camera. But one bridge at a time - the gimbal is proving to be insanely challenging as it is so far.
axegon_··on Ask HN: What are you working on? (May 2025)
Fully open source cinematography drone. Spoilers: I only started a few weeks ago and I've got a long way to go still. Currently prototyping the gimbal for more context and wasting a ton of PLA in the process.
axegon_··on Using unwrap() in Rust is Okay (2022)
I saw that article in 2023 and I have mixed feelings. In general there are cases where you want a program to crash and it's a preferable option as opposed to a never-ending chain of handling. For that purpose, unwrap is actually perfect. It is also great for examples, documentation, so that people can see what to expect and how a certain block should behave.

But there's the other case(which I've also suffered from). It is tempting to do the fast thing when you are on a tight schedule. One such instance was when we had to deliver something which was arguably undeliverable within the deadline we had. There were two services which were communicating over gRPC, and to make matters worse, not tonic but some custom implementation. But those worked surprisingly well against all odds. But there was a third service where we had to use a regular, plain ol' boring http requests. "I'm returning you the ID as a string inside the response, just cast it to a 32 bit integer and you're good to go" they said. "Fair", I thought: response_id::<i32>().unwrap(), compile and deploy. Needless to say, the ID was not a 32 bit integer. Point is, if you control the entire ecosystem you are working on, unwrap, although undesirable in many cases, is fine. Anything else - handle the errors.

axegon_··on Roto: A Compiled Scripting Language for Rust
> who has pondered their ideal scripting language

Et tu, brute? :D

axegon_··on Everyone Is Cheating Their Way Through College
Idk, maybe not cheating but a very unfair advantage - everything I ever did took me a fraction of the time compared to everyone else - while everyone was going nuts last minute, I had long forgotten about anything I had to do and was spending days or even weeks out playing pool in the local pub.
axegon_··on Everyone Is Cheating Their Way Through College
For a number of reasons I got my university degree in Spanish(and Spanish speaking country), where very few people spoke English - I was one of two in total as a matter of fact. I always felt like I was cheating my way through cause I had access to infinitely more and better quality resources and all I had to do was translate them into Spanish and I was instantly 10 steps ahead of everyone else. With all the LLM bollocks these days, I'm suddenly feeling a lot less guilty.
axegon_··on Elon Musk is wrong about GDP
> "elon is wrong about {X}"

People are (very) slowly waking up. The number of X-es is large to say the least.

axegon_··on Try Switching to Kagi
No, thanks, I'll stick to qwant: https://old.reddit.com/r/ukraine/comments/1gvcqua/psa_the_ka...
axegon_··on Privacy folks – what's your take on using LLMs at work?
Honestly? I'm somewhat OK with Mistral, in that it's a French company, therefore they need to comply (on paper anyways) with GDPR. And if they do not, there are serious consequences. But overall - locally. If it is not running on your hardware and it requires an internet connection to work, then the data is not yours. For the time being, building an AI rig at home is relatively cheap on ebay: if you play your cards well, it is doable for 1000 bucks or less.
axegon_··on What If We Could Rebuild Kafka from Scratch?
Having used both Kafka and Redpanda on several occasions, I'd pick Redpanda any day of the week without a second thought. Easier to setup, easier to maintain, a lot less finicky ans uses a fraction of the resources.
axegon_··on Atuin Desktop: Runbooks That Run
Oh, that's really neat! Thanks for sharing!
axegon_··on America Underestimates the Difficulty of Bringing Manufacturing Back
The true and sad truth is that manufacturing can be moved anywhere but the people that keep parroting about it's importance are in complete denial about the primary issue: costs. These costs have very little to do with infrastructure and building factories or logistics. Those are a contributor, sure, but that only scratches the surface. While China has seen insane growth in the last 20 years, that growth is at the expense of workers. No doubt they have a lot of value in terms of skills(which take a long time to acquire) but you also need to remember that there is a difference between the significance of working in Asia and Europe/north America. To us Europeans (and North Americans) work brings stability and security. In Asia, work is the difference between life and death, regardless of how skilled you are-you are legally expandable. Does anyone seriously believe that iPhones will be made in the US? The basic salary at foxconn is just under 320$/month or $1.81/hour. That is around 10x less than the US. This is ignoring the atrocious working conditions and far above the 40 hour work week. If we do factor in that as well, the difference is likely in the 25-30x range. I come from a country with a minimum hourly wage of around $3.6, let me tell you, as soon as the clock hits 18:00, people will drop everything where they stand and go home. The only way to compete with China is to automate everything and let machines do all the work, which is not a terrible idea but also nearly impossible to achieve. And even if you spend two decades doing all that, there are costs to all the R&D to get there. No one is going to buy a $15k iPhone, nor will they buy a $20K laptop. A logo that says "Made in X" won't justify the price. This comes from someone that uses a dual-xeon workstation as a personal computer.

Here's another example: a market that has been completely dominated by China: consumer drones. Believe me when I say this, I hate DJI and while I have one, I refuse to use it because of all the security implications. How many European and US companies are competing with them? Quite a few actually but the big names off the top of my head are Parrot and Skydio. I own both a Parrot and a Skydio and the quality of both is amazing. Yet they are still barely keeping up with DJI and at 5x the cost despite the demand - DJI still holds 90% of the market share. I can justify the price because of my privacy concerns but that's 1/1000 people. For most people it's always going to be a trade-off between price and quality+privacy.

If you want to enforce all that through tariffs, just put 5000% tariffs so that the local manufacturing cost will be the same as the cheap import and you solved the problem. How many people will be willing to spend 100 bucks for a pair of socks? That's a different story. The soviet union attempted something similar for several decades while trying to copy western technology. Anyone that knows a bit of history can tell you how that ended. Spoilers: not a success story.

axegon_··on The thing about Europe: it's the actual land of the free now
True but the mass adoption and flooding social media was far more gradual compared to the US still - all the big social media came out of the US really, so did all the massively centralized services. Computers in general have a steep learning curve so they are at large useless for remote rural areas in the US(in Europe you are never too far from a large city with universities and all the modern day creature comforts). Smartphones on the other hand are basic commodity with no entry barrier. 200 bucks and you have access to infinite and unfiltered information, which is what ultimately seems to have caused the issue: Across Europe(like mentioned before) we were all(and sadly still are) collectively playing catch up. And again - Nordics are all too familiar with how russia operates, whereas the US is still far from figuring it out. Hence the reason Europe is far more immune to it, although many are still falling victim. Europe also knows how to protest, which is still at large something the US has no clue about. Imagine if the leader of any large European country shamelessly pulls a pump and dump the way Trump did (twice as a matter of fact) this week. People in Europe pull out their pitchforks for infinitely less.
axegon_··on The thing about Europe: it's the actual land of the free now
The sad thing is I've always seen (and still do see) the US as a friend, despite the fact that the current government is trying really hard to turn the US into a dictatorship(shutting down medias that criticize them, cozying up with literal criminals, murderers and worse and so on and calling it "freedom" - the exact same approach lenin had). I know things will turn around eventually since this is not the first such instance in the history of the US. Coming from eastern Europe, the US was the place to be when I was growing up and this carried on into the first half of my 20's. I immigrated into western Europe in my teens and continued my education there but as soon as my country entered the EU, somehow the appeal of going to the US slowly started to vanish. The idea of dealing with immigration offices and endless bureaucracy was really not pleasant. And surprisingly after I completed my studies, I moved back home and all things considered, this was probably the best decision I've ever made.

Now let's set something straight: Europe is bureaucratic hellhole, which is the reason why entrepreneurship is fairly uncommon. Here we are taught to go through the procedures, get whatever licenses and paperwork is required, make sure we follow all steps and cover everything up and then and only then start building a product. The US philosophy is the complete opposite approach: start building and figure it out along the way, which is the reason why Europe is always behind the curve.

This however ended up being both a curse and blessing in disguise for us: the digitalization of everyday life took longer, whereas everyone and their dog in the US had a smartphone and social networks from day 0. In addition, here in Europe we've been exposed to bad actors such as russia for centuries and many of us can navigate through their tactics, whereas the US instantly swallowed everything that was thrown at them with the oldest trick in their book: "media is lying to you, see this picture/video only here". A decade and a half of actively trying to discredit establishments and it ended where it's at. And as much as GDPR is a pain in the ass, companies here are very well aware how badly things can go for them if things are not kept in check. Which, as a citizen, is a great thing.

The other problem with the quick rise of digitization in the US in conjunction with practically non-existent regulations around privacy are grifters, which truly swarmed the internet. Business strategy: make a dumb video "owning someone", share it around so people learn who you are, get them roiled up against one another, having half of them become your worshipers, release a book, merchandise or courses and you are set for life. From practical nobodies all the way up to presidents - it's a guaranteed success. But you have to have 0 moral values to do that and at some point you will need people around you. Thing is that it's only a question of time before everyone around you starts realizing that no one has moral values and start screwing each other up. The US is currently in that stage.

All things considered, the US has fallen really far behind in terms of freedom in single digit months.

Edit: Some time ago my mum was watching some interview with an analyst who really summed it up: "I used to believe the US was about 100 years ahead of China in terms of innovation. I was wrong, it's probably closer to 50. And if you ask me where we are in Europe? Preoccupied with coming up with more inconvenient bottle caps to solve a problem we don't have".

axegon_··on Owning my own data, part 1: Integrating a self-hosted calendar solution
It is next in the pipeline but I need to clean up my home lab first cause it's become a bit of a mess.
axegon_··on Owning my own data, part 1: Integrating a self-hosted calendar solution
I too have been on a very aggressive mission to decloudify myself lately(as in, since comrade musk and co took over everything and I have 0 reason to trust him or anyone around him). And to be fair, for all the hundreds of gigabytes of data I have, the migration took no longer than two weeks, on and off, and it works infinitely better than I was hoping. The only service I've outsourced is email since I hate dealing with MX records, IMAP/SMTP. For everything else, the one thing that did wonders was an Asus PN40 with a dual core celeron CPU. It's biggest strength is that it's both accessible and has a slot for an M.2 NVME and a SATA drive so you can have a safe backup in one spot. I too share the opinion that CalDAV is kind of sketchy. That said, Nextcloud covered most of my needs(although it can be slow for some things, even with redis backing it up). Calendars, dropbox/google drive, docs, photos, backups and automatic syncing - it covers everything really well. That, combined with several other open source solutions truly cover all my needs(music and video streaming, monitoring, alerting, vpn, readers, management, and various other utilities). And as under-powered this mini PC is, it handles everything insanely well with lots of room to spare, given that my ISP is a solid 10/10. In fact it would be perfectly capable of handling another 5-6 people like me with ease.
axegon_··on How much do you think it costs to make a pair of Nike shoes in Asia?
There was an interview with the local marketing managers for Coca Cola and Pepsi when I was a kid. I vaguely remember it (and also the idea was somewhat foreign to me since the world had just started opening for eastern Europe and it was the first time many of us heard the names such as Coca Cola, McDonald's, etc). The one thing that stood out was the Pepsi manager raging about campaigns, failed products and whatnot. The Coca Cola manager calmy looked at him and said: "Sir... Look at our product... Who in their right mind would pay for a rusty-colored liquid as a beverage? Absolutely no one. It's all about the advertisement".

Same story here - would you buy a pair of shoes from anyone, seeing the horrible conditions they work in, sleep deprivation and poverty if you knew they'd still get close to nothing? Probably not. You see a fancy store (or website these days), you see a product you like and you're invited to open up your wallet for a fancy pair of shoes. Deep down you still know they came from the people working in horrible conditions but, you know, out of sight, out of mind. And thus, you are willing to pay the extra price.

axegon_··on Ask HN: Do you still use search engines?
Qwant has been reasonably good with programming for me though or maybe I'm more commonly going straight to the documentation or source code, idk. My only complaint is that wikipedia tends to be pretty far down in the results for non-programming questions. LLMs(all of them) have been terrible with programming questions too. Imaginary APIs, leaks, code that outright breaks your data, you name it. Oh and don't get me started on optimization and performance(which is ultimately one of my main concerns on day to day basis). They are absolutely horrible.
axegon_··on Ask HN: Do you still use search engines?
Yes, I do. I'd never use an LLM for any meaningful or important information because by design, they will simply shove the most likely next token and you get a ton of responses which are pure nonsense if you start digging into it. Mind you, I've been noticing that Google has been terrible for a long while now. Kagi seemed alright at first but it also gave a lot of nonsense. The final straw for Kagi was the fact that they are backed by yandex and by extension fund the russian terrorist state. Lately I've switched to Qwant and so far it covers almost all of my needs.
axegon_··on Less Htmx Is More
> interns handle that

Reality has entered the chat.

axegon_··on Less Htmx Is More
Oh yeah a 15 second spinning circle and 250mb memory is sooooo much better than a 500ms refresh </sarcasm>
axegon_··on Less Htmx Is More
> Updates that users would not expect to see on a refresh (or a new page load)

I always hated this idea. As a user, a refresh indicates that something is happening and it's abundantly clear when something is wrong. People don't always handle errors and in all fairness they shouldn't - a developer has no way of knowing what custom stuff I have on my browser, whether I'm using any blockers or pi-holes or whatever and they should not know. Simple navigation, refreshes and server side rendering is something which worked great, the web was fast and could run on anything with a graphical output. These days a single page eats up 150+ mb while it loads. All that so the page doesn't "refresh".

axegon_··on The slow collapse of critical thinking in OSINT due to AI
OSINT is a symptom of it. When GPT-2 came along, I was worried that at some point the internet will get spammed with AI-crap. Boy, was I naive... I see this incredibly frequently and I get a ton of hate for saying this (including here on HN): LLMs and AI in general is a perfect demonstration of a shiny-new-toy. What people fail to acknowledge is that the so called "reasoning" is nothing more then predicting the most likely next token, which works reasonably well for basic one-off tasks. And I have used LLMs in that way - "give me the ISO 3166-1 of the following 20 countries:". That works. But as soon as you throw something more complex and start analyzing the results(which look reasonable at first glance), the picture becomes very different. "Oh just use RAGs, are you dumb?", I hear you say. Yeah?

class ParsedAddress(BaseModel):

    street: str | None

    postcode: str | None

    city: str | None

    province: str | None

    country_iso2: str | None

Response:

{

  "street": "Boulevard",

  "postcode": 12345,

  "city": "Cannot be accurately determined from the input",

  "province": "MY and NY are both possible in the provided address",

  "country_iso2": "US"
}

Sure, I can spend 2 days trying out different models and tweaking the prompts and see which one gets it, but I have 33 billion other addresses and a finite amount of time.

The issue occurs in OSINT as well: A well structured answer lures people into a mental trap. Anthropomorphism is something humans have fallen for since the dawn of mankind and is doing so yet again with AI. The thought that you have someone intelligent nearby with god-like abilities can be comforting but... Um... LLMs don't work like that.

axegon_··on Doge staffer's YouTube nickname accidentally revealed his teen hacking activity
You basically just described rebellious youth, which we have all been at some point in our lives. And this statement is complete and utter bs. The problem is that comrade muskov is not a technical person, he is just a "marketing strategist" (quotation marks since I don't want to offend people who actually know what they are doing). Here's the painful truth: most governments have appalling security practices and it's a miracle that the world still exists. Finding vulnerabilities or leaks is not a hard task and it only requires patience - there are dozens of such examples in my own country alone and the only reason no one has sounded the alarm is because two things are going to happen: media scandal for 2 days, then everyone will forget about it and the second thing that will happen is that whoever rings the alarm will be dragged in by security agencies for years to get questioned about what they were looking for and they will not accept "I know you are morons and I don't trust you with handling my data in a service that is completely open to the general public". "Run-ins with the law" has nothing to do with talent and if anything, it's the absence of a talent if that resulted in "run-ins with the law". That is a clear indication that the people involved have no idea what they are doing and just managed to connect two simple dots after a "complete 2 hour hacking crash course - FREE, pls subscribe for more". The simple fact is that musk has no goddamn clue about what he is doing or talking about: "the government does not run relational databases". Right comrade, cause precisely the usa spending service which you are referring to is not a django app using psql: https://github.com/fedspendingtransparency/usaspending-api/b...
← PreviousPage 7 of 34Next →