varun_ch

I’m ambivalent towards AI generated poster or other business design. It might leave a bad taste in my mouth but it’s whatever. While I’m against it, I can understand why for many businesses it’s an easy and convenient choice.

What makes me feel almost physical disgust is the food in AI generated posters or menus. There’s something deeply uncannily unappetizing about the food. And once I can identify it as AI generated, it’s even worse… there’s something almost creepy about knowing the image is a computers best approximation of something humans want to ingest.

I cannot understand how a food establishment that takes themselves seriously would want to use generative assets like this… can’t they see how unappetizing their own food looks??

This is pretty cool. iOS could do with more 'do one thing and do it well' apps without paywalls or subscriptions...

Does HN still use simple files for storing its data?

I think that really depends on how you use the cards. I know medicine students who clearly are passionate about their subject, but just need to get the reminders on individual terms because there’s such a wide range of topics they need to be prepared for, and lots of very similar words.

I also know that in math courses, there’s certain formulas with similar names I simply need to remember because deriving them on a test would be too time consuming. Or I could imagine in chemistry there’s various special cases that you’d want to be able to recite on a test.

I’ve never tried it for language learning but people talk highly of it, so there’s definitely some truth in their utility there.

I think the trick is flash cards are fantastic supplementally with other studying strategies.

  function getPercentileText(survived) {
      if (survived === 18) return 'Top 1% of players today ';
      if (survived === 17) return 'Top 2% of players today ';
      if (survived === 16) return 'Top 3% of players today ';
      if (survived === 15) return 'Top 5% of players today ';
      if (survived >= 12 && survived <= 14) return 'Top 10% of players today ';
      if (survived >= 10 && survived <= 11) return 'Top 20% of players today ';
      if (survived >= 8 && survived <= 9) return 'Top 50% of players today ';
      return 'No trophy earned today ';
  }

This is really cool. I think a lot of people will initially dismiss your demo as ‘slop’ unless the landing page drops some of the Claude-isms (text and design) because it’s clear this is a project with passion in it, but the demo page and landing page both look like something Claude might’ve made with a single prompt.

we’re on hacker news! this is the right place for the adventurous people to ‘try out all these new things’ so that the new stuff can emerge victorious :P

genuine question: is latency that big of a deal wrt tools for agents? I see this a lot with these AI-native tooling replacement projects, but surely the inference will be orders of magnitudes slower?

on the topic of Swiss Internet: everyone I know in Zurich’s home internet gets a mostly-static IPv4. It’s not a promised feature or anything, but my IP didn’t rotate for years. This is super handy for self hosting.

to be fair, if you went to a movie store, bought a movie (and unbeknownst to you there was some note on the counter that said the store reserved the right to take your movie from you), when the stores employees break into your house one day and forcefully take the DVDs from you, you might use a similar word…

that’s fair! It definitely looks good and modern!! I just wonder if it compromises the initial impressions of the project in some way.

maybe slightly unrelated but the new htmx homepage (https://four.htmx.org/) feels a little ironic, seemingly written with tailwindcss and a full JS ecosystem Astro build system. It also has the ‘vibey’ ‘hypey’ landing page design that’s hard to describe but you’ll find on any web framework, rather than dropping you to docs like the old site.

Compared to the original simple HTML site it’s really surprising to see from the grugbrain.dev author!

surely the weights for the model & the equipment to run them make it logistically challenging enough to deter that… also I’m sure models have leaked in their APIs before but those would be pretty easy and quick to catch/fix.

It seems like being a security researcher in 2024 vs 2026 is a wildly different experience across the board (presumably not just with J&J).

I wonder what changed…

When I took the digital SAT a couple years ago, we had access to the Desmos Graphing Calculator during the whole math section.

The entire point of the exam was to test whether you can read a math question, input it into the calculator and select the option that matches the result within 60 seconds. If you get a couple questions wrong, you drop hundreds of points. I don’t think it was a valuable test whatsoever (and of course, it biases to students who can afford time/money for thousands of practice questions to improve this “skill” through repetition)

The English reading/writing section was much more interesting, but again, the time limitations make it a skimming test more than anything else.

Many universities allow you to ‘superscore’ multiple attempts, to combine a math and RW score from different SATs. So again, scores bias towards students who can afford to take one test dedicated to math, and another dedicated to English.

this might be the only place where a redirection to bing provides a much better UX than what was there previously!

I think most people don’t need to worry about their host abusing its power to impersonate them, but the cool thing is, the people who do need to/want to worry (journalists, politicians, celebrities, activists, open source maintainers, etc etc etc) can self host a PDS and be a lot safer, and still interact with everyone else.

the person who made this prototyped an e-ink framework 13 and it looks awesome... https://x.com/zephray_wenting/status/1535041457035280392

tangled.org!!!!

I think GitHub and GitLab have special agreements

Doing this for Neovim at work and it’s so much fun. I justify the slopped config to myself because I want to get work done rather than learn the config of some random Neovim package and how it interacts with the rest of the system.

I can just ask Claude, “make leader / toggle a terminal at the bottom of the screen , and leader t swaps it to the right side” and it’ll just do that. I liked a theme but I wanted it to also change the status line of the focused window, and it just did that for me. I had an issue where the neo tree would freeze for a few seconds, and Claude figured out it was a bad interaction with the git in our toolchain, etc etc.

I have my very own text editor that I customized in plain English! I could’ve learnt spent hours learning Neovim’s style of Lua, researching packages, debugging, etc, but this gets the same stuff done way faster and lets me get to work. This was the biggest thing that kept me going back to either a preconfigured Vim setup like LazyVim or vscode. Definitely recommend.

There is an awesome YouTube video about this from the person who made it: https://m.youtube.com/watch?v=nHbA2-_qzH4

I think this device isn’t so much about a pen. It seems like it could be a really nice typing or coding or reading display. Maybe a future model could improve on the pen

If I trusted my friend to hold onto my childhood photo collection as a favor, and then came back to find him asking for $5 to return it, and I reluctantly agree only to find him empty handed, I wouldn't consider that providing value.

The world would be a better place if you made money by providing value to people. Instead of extorting them.

It’s also relevant that LLMs have so much JavaScript training data that I don’t see a world where we’re not still using JavaScript.

The flight simulator was always my favorite part of the desktop app, Google Earth Pro(?). (I guess apart from it having a really old version of chromium embedded)

I found out recently that you can just buy books. There's businesses who sell books. Not any specific book, but just books to fill shelves to decorate rooms. You can even buy colour coordinated books.

https://booksbythefoot.com/

You’re right that lots of Windows apps were designed with Keyboard only workflows in mind. It’s a shame that MacOS has so many points where if you don’t have a mouse you’re out of luck.

There is one major improvement you can do on Mac, at least for menus:

https://varun.ch/posts/macos-keyboard/

Hi! I was a high school student in your exact position a year ago. I just finished my first year in university (Computer Science). I don't think anyone here can give you a definitive answer, but for me personally I think I'm still making the right choice: (this is what I tell myself)

1. learning stuff is fun. even if AI drastically changes what it means to be a software engineer, as it stands, we still need software engineers. You can go to university, learn CS/Coding by hand, and on the side keep up with what it means to be productive with AI tools. That way, you're still employable and you get the gift of getting to learn stuff (perhaps at great cost, but maybe you live somewhere where the education is affordable)

2. the underlying principles aren't changing. Computer Science is still Computer Science. computers still have memory and the basic data structures are always going to be what they are. I think it's important to know what we're building on top of. (ie. a React developer should probably understand the DOM. A C developer should understand what their code is compiling to). I don't think it's any different with using AI to write code. Learning programming/computer science will still be important even with AI because it's important to have people who understand the full stack that we build technology on top of.

3. you could work on the AI. People still need to understand the math that builds AI. You could be one of those people.

4. AI is great at making things that already exist. but we will still need to make _new_ things. Humans do that.

my main thing is, if I wasn't in school to learn Computer Science, what would I be doing instead? I certainly don't want to be a someone who's job is genuinely replaceable by an agent. I don't think all programmers will be like that.