kaesve

I was also curious, looks like open cascade, and a pretty good range of operations supported (see https://github.com/Fluid-CAD/FluidCAD/tree/main/lib/oc). Super cool!

My wife's Etsy shop (https://www.etsy.com/shop/LittleLanternShop) is starting to pick up. She is leaning into creating digital sewing patterns for decorative felt crafts. We have had Etsy success before with 3d printed products, but managing printers and fulfilling orders can be stressful and time consuming, and she was hoping to build up a more passive income stream. She made over $1000 in the past month, which beat both our expectations

(We'll get back into 3d printing once life slows down a _little_ bit again)

I like https://thebookofshaders.com/ . It’s unfinished and I don’t think it’s been updated in years, but what’s there is pretty good

I'd use this on linux as soon as it's available

https://github.com/jonas/tig is one of the first things I install on a new dev machine. It's a really nice UI for staging files or hunks. Since it's just a companion to the git CLI, it feels much more focused than full-blown git GUIs, and doesn't do anything magical.

Don't forget Knuth wrote the paper on complexity of song lyrics https://dl.acm.org/doi/pdf/10.1145/1008354.1008355

There's https://arxiv.org/abs/2310.00166, which uses an LLM for intrinsic rewards for a RL agent. They use it on nethack. It was discussed on the TalkRL podcast: https://www.talkrl.com/episodes/pierluca-doro-and-martin-kli...

neat! I've wanted to add irregular grids to my infinite minesweeper. Playing your version, it definitely adds something. How do you generate your grid? Is it voronoi cells on top of blue noise?

  Location: Philadelphia/New York
  Remote: Possible
  Willing to relocate: No
  Technologies: Python, TensorFlow, PyTorch,  JS/TS, NodeJS, Angular/React/Vue
  Résumé/CV: https://hireken.at/hacker-news/
  Email: mail@kaesve.nl

Proving Ground: The Untold Story of the Six Women Who Programmed the World’s First Modern Computer (by Kathy Kleiman)

I enjoy reading up on computer history, and this is a pretty good retelling of an important story. However, the parts that really struck me are the personal accounts of the author in trying to recover this history. The ENIAC 6 played an important role in the history of the first computer and of programming as a vocation, but their story was almost forgotten. To the point that these women were not even invited to celebrate the 50th anniversary of the ENIAC. That's shameful, and I'm happy that they finally have started to get the recognition they deserve.

If you're interested in this history, Jean Jennings Bartik also wrote an autobiography, that tells the story from her perspective.

  Location: New York/Philadelphia
  Remote: Yes
  Willing to relocate: No
  Technologies: Python, Tensorflow/Torch, Javascript/Typescript, Node, Angular/React/Vue, Docker, SQL
  Résumé/CV: (working on it)
  Email: mail@kaesve.nl

I'm from the Netherlands, but moving to the US (Princeton) soon, and should have my green card by the end of the year, so I'm looking for something new.

I'd love to find a new challenge in the ML field, but I'm open to other opportunities.

There is also "The Algorithmic Beauty of Seaweeds, Sponges and Corals" (https://link.springer.com/book/10.1007/978-3-662-04339-4), and, my personal favorite in the series, "The Algorithmic Beauty of Sea shells" (https://link.springer.com/book/10.1007/978-3-540-92142-4).

I really enjoyed reading the latter last year, because it does a great job slowly building up. By the end, I felt like I didn't just understand the models, but also how the author found them. I can recognize these patterns in nature, and figure out how they were generated.

I spent some time last year implementing all the models described in the book in JS: https://kaesve.nl/projects/shells .

Not the OP, but I still start most of my projects without TypeScript, especially if I'm working on it by myself. I do appreciate TypeScript, and I will happily use it in any bigger project or when I have collaborators. But there's just something that I really enjoy about being able to start a project without any build steps or installing stuff. I love to be able to just jam out some code and open a file in a browser to run it, or rsync it to my server and share it with my friends.

There's a ton of value in building stuff without tools. It's somewhat similar to learning two frameworks; seeing React's philosophy makes me a better Vue programmer. In a similar vain, building stuff without tools teaches me a lot about what those tools actually offer me. What stuff is really hard? What stuff is really easy to abstract out? How do Vue/React/Whatever do these things?

It also teaches you more about your platform. In the end, all these web tools and frameworks are running in that same browser. I find especially with newer web developers, it's hard to solve something in React/Vue/.. because they don't know how to break up the problem. If you'd roughly know how to solve the problem without those frameworks, it becomes much easier to figure out how to solve it with them too.

Finally, I also just get toolchain fatigue. If I want to build a simple landing page, I don't want to learn new stuff, I don't want to read what changed in the latest versions, what the correct way of the day is to set up a project. Honestly, I don't even want to have a build pipeline and module swapping dev server, that inevitably needs to be configured. Just some static html/css/js is often good enough. And as I said, you also learn which problems frameworks solve well, so you'll know when some static files isn't the right choice.

I'm like this. I have started literal hundreds of projects, most of them game ideas. Projects are never _really_ finished, but through the years I have become better at leaving them at a satisfying point, even making some of my projects public. For me, this was a journey. Here's a few lessons I learned for me:

- Figure out why you are doing this in the first place. Usually, my core desire behind a project idea is not at all "to make money" or "help people", but something more simple. It's to figure out how to do something, or to quench my own curiosity. - I realized I have a near-unending pool of ideas, but I can just do one at a time. Which one to pick? For a while, I chose them based on project size; how easily could I gratify that core desire? How quickly could I be happy with this project? Almost any project that I'd think would take more than a weekend would be sent back to the pile. Sometimes it took a while to distill the core desire behind a project idea, to really see why it was interesting to me. - Doing this also taught me a lot about how to make projects smaller. This was extremely freeing. I don't have to worry about losing work -- I'll lose maybe up to a day of typing, and the real value was working on the project anyway, not necessarily the produced artifact. I also realized I can do many projects with nothing but a browser and a text editor, and the more I used this small tool set, the deeper my understanding got and the more I could do in a weekend project.

Over time, I've 'finished' many of these projects, and I've also started to learn that finishing really is a skill in itself -- something that takes practice. More of my projects nowadays are at least somewhat ready for public consumption, and I'm still getting better at this. I still start projects that I don't finish. If an idea is cool enough, I'll still give it a try. Some project ideas were so cool/fun to work on that they would just be able to capture my attention for the weeks or sometimes even months they asked for to get them somewhere (most recently, that's https://kaesve.nl/projects/shells/). Some projects can't capture all that attention at once, but are still fun enough to revisit. They take years, but slowly move forward. And if I abandon a project, that's okay too; it usually means I found a new, shorter-term project, or that I realize that that core desire behind that project idea wasn't so satisfying after all.

How is relatedness measured? Using some embedding space? I often disagree with the measurements, the worst one being "punch" and "bowl" only relating 12%.

The concept is very fun though. I might try to make my own version, as it also seems like a fun side project and a way to explore different word embedding spaces. Could be fun to maybe also have a visualization of the embedding space.

https://neonarcade.games

Just silly games I make in my free time. Ever wanted to play wordle and snake at the same time? I got you. That connect-the-pipe game, but also tetris? Sure!

They are mostly novelty games, with a focus on fun-to-make instead of fun-to-play.

I make small web games, usually by twisting up one or two existing games. Like wordle, but also snake? Why not play both at the same time! Remember that pipe game? What if it was also tetris?

A lot of these projects make for terrible games, but really fun side projects. I've been putting up the more playable ones on https://neonarcade.games

A 'trick' that I've grown fond of is for iterating over a circular list where the current and previous items form an edge, with a special case for the first edge, that is composed of the first and last items of the list (making it circular).

The 'normal' way to write a loop over this is something like:

for (let i = 0; i < ls.length; i++) {

    const j = i > 0 ? i - 1 : ls.length - 1;

    ...

A trick I got from sean barretts website (https://nothings.org/), is to use the for-loop initialization clause for the special case:

for (let i = 0, j = ls.length - 1; i < ls.length; j = i++) {

  ...

The work I do usually does not care about the performance improvement if there is any, but it feels good to avoid that per-iteration check.

I'm working through "The algorithmic beauty of seashells" by Hans Meinhardt and it's fantastic. It's been on my list for quite a while, but I only got around to it about 2 months ago. I'm through the first 90%, which means I got through the seashell pattern modeling. I'm reading the last part more slowly, both because it's not building up the model and because I'm building an interactive implementation of the models.

I'm also reading Seveneves by Neal Stephenson. Started less than a month ago, am at p256/867. It's not my 'priority book', but I've been enjoying it when my brain needed a break from the seashells.

I recently put up https://neonarcade.games where I put my own hobby project games; an astroid-like game, infinite hexagonal minesweeper, and a game where you have to play two arcade classics at the same time. There are a few more games in the pipeline that should be added soon.

They are not of the highest quality, but I continue to work on them. They are all built from scratch with only a (my own) math library that I reuse. If something's broken and you bug me about it, I should fix it soon. Same story for adding mobile support where it's not there yet.

I decided I am not going to do ads or tracking. I have fun making these, so I just hope someone else can have fun playing. There might still be some google font references, but I'm planning to take those out too. The same should go for https://masterpieces.app, where you can do jigsaws with the Rijksmuseum collection.

is the cellular automata tiled? Did you implement it yourself, or did you grab an existing one? Did you have to do any work to get it to run so smooth? It's very cool.

And GPT-J -- are you running it yourself? Or are you using a hosted API somewhere?

What's wrong with the current theme? I quite like the simplicity. It's easy to navigate and read and it doesn't feel bloated like a lot of websites do.

The theme does what it needs to, so I'd only change it if you want to make it more fun for _you_, or make it feel more personal to you -- and in that case it's 100% personal what you go for.

My current project is https://neonarcade.games, a collection of silly casual games. 90% of the motivation to make these games is to have fun in the process, so the focus is on funny ideas and DX over UX and good game design. The canvas api is excellent for these kinds of projects, and you really don't need much more. Most games I make start as a single HTML file and a math.js, which I regularly start from scratch, and I often use emoji as a proxy for real sprites.

Though the focus is on making, I have been playing my own minesweeper clone almost every day. In some sense, that makes it my most successful side project to date.

I've actually been building a [jigsaw puzzle app](https://kaesve.nl/projects/masterpieces) because wanted to do something with their online collection. They not only put paintings and other artworks online for free, but also their meta data, through an api or bulk download (see https://data.rijksmuseum.nl/).

I love my HHKB to death. I can also really recommend getting a reprogramable controller off of hasu from geekhack (https://geekhack.org/index.php?PHPSESSID=58s2o9bvije7q36shi8...). It's really easy to install and customize the firmware, and it adds a whole new dimension to how much I enjoy my keyboard.

I wrote an implementation using WebGL (https://kaesve.nl/projects/mold), two weekends ago. The original author has been posting a lot of variations on his twitter, including a stunning multiscale implementation: https://twitter.com/mxsage

I personally really like to just use plain javascript and the canvas API. I found the tutorial on mdn very useful (https://developer.mozilla.org/en-US/docs/Web/API/Canvas_API/...), and I enjoy the simplicity of not using libraries and tools. I've also used EaselJS, which is also pretty good (but doesn't offer me anything I really need), and PixiJS, which is supposed to be very fast, by using WebGL where possible.

Are you talking about "The Geometry of Innocent Flesh on the Bone: Return-into-libc without Function Calls (on the x86)"(https://cseweb.ucsd.edu/~hovav/dist/geometry.pdf)? Great paper with a great title.

This is giving me a strong Renowned Explorers vibe :p Want to play the new content when I get home.